Title: Brewing science and practice

Brewing 

Science and practice 

Dennis E. Briggs, Chris A. Boulton, Peter A. Brookes and 

Roger Stevens 

Copyright © 2004 Woodhead Publishing Limited and CRC Press, LLC

Published by Woodhead Publishing Limited, Abington Hall, Abington 

Cambridge CB1 6AH, England 

www.woodhead-publishing.com 

Published in North America by CRC Press LLC, 2000 Corporate Blvd, NW 

Boca Raton FL 33431, USA 

First published 2004, Woodhead Publishing Limited and CRC Press LLC 

ß2004, Dennis E. Briggs, Chris A. Boulton, Peter A. Brookes and Roger Stevens 

The authors have asserted their moral rights. 

This book contains information obtained from authentic and highly regarded sources. Reprinted 

material is quoted with permission, and sources are indicated. Reasonable efforts have been 

made to publish reliable data and information, but the authors and the publishers cannot assume 

responsibility for the validity of all materials. Neither the authors nor the publishers, nor anyone 

else associated with this publication, shall be liable for any loss, damage or liability directly or 

indirectly caused or alleged to be caused by this book. 

Neither this book nor any part may be reproduced or transmitted in any form or by any 

means, electronic or mechanical, including photocopying, microfilming, and recording, or by 

any information storage or retrieval system, without prior permission in writing from the 

publishers. 

The consent of Woodhead Publishing Limited and CRC Press LLC does not extend to 

copying for general distribution, for promotion, for creating new works, or for resale. Specific 

permission must be obtained in writing from Woodhead Publishing Limited or CRC Press LLC 

for such copying. 

Trademark notice: Product or corporate names may be trademarks or registered trademarks, and 

are used only for identification and explanation, without intent to infringe. 

British Library Cataloguing in Publication Data 

Acatalogue record for this book is available from the British Library. 

Library of Congress Cataloging-in-Publication Data 

Acatalog record for this book is available from the Library of Congress. 

Woodhead Publishing Limited ISBN 185573 490 7(book) 185573 906 2(e-book) 

CRC Press ISBN 0-8493-2547-1 

CRC Press order number: WP2547 

The publisher's policy is to use permanent paper from mills that operate asustainable forestry 

policy, and which have been manufactured from pulp which is processed using acid-free and 

elementary chlorine-free practices. Furthermore, the publisher ensures that the text paper and 

cover board used have met acceptable environmental accreditation standards. 

Project managed by Macfarlane Production Services, Markyate, St Albans, Hertfordshire 

(e-mail: macfarl@aol.com) 

Typeset by MHL Typesetting Ltd, Coventry, Warwickshire 

Printed by TJ International, Cornwall, England 


Contents 

Preface 

1 An outline of brewing 

1.1 Introduction 

1.2 Malts 

1.3 Mash tun adjuncts 

1.4 Brewing liquor 

1.5 Milling and mashing in 

1.6 Mashing and wort separation systems 

1.7 The hop-boil and copper adjuncts 

1.8 Wort clarification, cooling and aeration 

1.9 Fermentation 

1.10 The processing of beer 

1.11 Types of beer 

1.12 Analytical systems 

1.13 The economics of brewing 

1.14 Excise 

1.15 References and further reading 

1.15.1 The systems of malting and brewing analysis 

1.15.2 General references 

2 Malts, adjuncts and supplementary enzymes 

2.1 Grists and other sources of extract 

2.2 Malting 

2.2.1 Malting in outline 

2.2.2 Changes occurring in malting grain 

2.2.3 Malting technology 

2.2.4 Malt analyses 

2.2.5 Types of kilned malt 

2.2.6 Special malts 


vi Contents 

2.2.7 Malt specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 

2.3 Adjuncts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 

2.3.1 Mash tun adjuncts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 

2.3.2 Copper adjuncts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 

2.4 Priming sugars, caramels, malt colourants and Farbebier . . . . . . . . . . 45 

2.5 Supplementary enzymes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 

2.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 

3 Water, effluents and wastes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 

3.2 Sources of water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 

3.3 Preliminary water treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 

3.4 Secondary water treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 

3.5 Grades of water used in breweries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 

3.6 The effects of ions on the brewing process . . . . . . . . . . . . . . . . . . . . . . . . . 65 

3.7 Brewery effluents, wastes and by-products . . . . . . . . . . . . . . . . . . . . . . . . . 68 

3.7.1 The characterization of waste water . . . . . . . . . . . . . . . . . . . . . . . . 69 

3.7.2 The characteristics of some brewery wastes and 

by-products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 

3.8 The disposal of brewery effluents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 

3.8.1 Preliminary treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 

3.8.2 Aerobic treatments of brewery effluents . . . . . . . . . . . . . . . . . . . 75 

3.8.3 Sludge treatments and disposal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 

3.8.4 Anaerobic and mixed treatments of brewery effluents . . . . . 79 

3.9 Other water treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 

3.10 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 

4 The science of mashing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 

4.2 Mashing schedules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 

4.3 Altering mashing conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 

4.3.1 The grist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 

4.3.2 Malts in mashing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 

4.3.3 Mashing with adjuncts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 

4.3.4 The influence of mashing temperatures and times on 

wort quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 

4.3.5 Non-malt enzymes in mashing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 

4.3.6 Mashing liquor and mash pH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 

4.3.7 Mash thickness, extract yield and wort quality . . . . . . . . . . . . . 116 

4.3.8 Wort separation and sparging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 

4.4 Mashing biochemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 

4.4.1 Wort carbohydrates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 

4.4.2 Starch degradation in mashing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 

4.4.3 Non-starch polysaccharides in mashing . . . . . . . . . . . . . . . . . . . . 136 

4.4.4 Proteins, peptides and amino acids . . . . . . . . . . . . . . . . . . . . . . . . . 142 

4.4.5 Nucleic acids and related substances . . . . . . . . . . . . . . . . . . . . . . . 146 

4.4.6 Miscellaneous substances containing nitrogen . . . . . . . . . . . . . . 146 

4.4.7 Vitamins and yeast growth factors . . . . . . . . . . . . . . . . . . . . . . . . . 149 

4.4.8 Lipids in mashing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 


4.4.9 Phenols 

4.4.10 Miscellaneous acids 

4.4.11 Inorganic ions in sweet wort 

4.5 Mashing and beer flavour 

4.6 Spent grains 

4.7 References 

5 The preparation of grists 

5.1 Intake, handling and storage of raw materials 

5.2 The principles of milling 

5.3 Laboratory mills 

5.4 Dry roller milling 

5.5 Impact mills 

5.6 Conditioned dry milling 

5.7 Spray steep roller milling 

5.8 Steep conditioning 

5.9 Milling under water 

5.10 Grist cases 


6 Mashing technology 


6.2 Mashing in 

6.3 The mash tun 

6.3.1 Construction 

6.3.2 Mash tun operations 

6.4 Mashing vessels for decoction, double mashing and temperatureprogrammed 

infusion mashing systems 

6.4.1 Decoction and double mashing 

6.4.2 Temperature-programmed infusion mashing 

6.5 Lauter tuns 

6.6 The Strainmaster 

6.7 Mash filters 

6.8 The choice of mashing and wort separation systems 

6.9 Other methods of wort separation and mashing 


6.11 Theory of wort separation 


7 Hops 


7.2 Botany 

7.3 Cultivation 

7.4 Drying 

7.5 Hop products 

7.5.1 Hop pellets 

7.5.2 Hop extracts 

7.5.3 Hop oils 

7.6 Pests and diseases 


7.6.1 Damson-hop aphid 

7.6.2 (Red) Spider Mite 

7.6.3 Other pests 

7.6.4 Downy mildew 

7.6.5 Powdery mildew 

7.6.6 Verticillium wilt 

7.6.7 Virus diseases 

7.7 Hop varieties 


8 The chemistry of hop constituents 


8.2 Hop resins 

8.2.1 Introduction 

8.2.2 Biosynthesis of the hop resins 

8.2.3 Analysis of the hop resins 

8.2.4 Isomerization of the -acids 

8.2.5 Hard resins and prenylflavonoids 

8.2.6 Oxidation of the hop resins 

8.3 Hop oil 


8.3.2 Hydrocarbons 

8.3.3 Oxygen-containing components 

8.3.4 Sulphur-containing compounds 

8.3.5 Most potent odorants in hop oil 

8.3.6 Hop oil constituents in beer 

8.3.7 Post fermentation aroma products 

8.4 Hop polyphenols (tannins) 

8.5 Chemical identification of hop cultivars 


9 Chemistry of wort boiling 


9.2 Carbohydrates 

9.3 Nitrogenous constituents 


9.3.2 Proteins 

9.4 Carbohydrate-nitrogenous constituent interactions 

9.4.1 Melanoidins 

9.4.2 Caramel 

9.5 Protein-polyphenol (tannin) interactions 

9.6 Copper finings and trub formation 


10 Wort boiling, clarification, cooling and aeration 


10.2 The principles of heating wort 

10.3 Types of coppers 

10.4 The addition of hops 


10.5 Pressurized hop-boiling systems 

10.5.1 Low-pressure boiling 

10.5.2 Dynamic low-pressure boiling 

10.5.3 Continuous high-pressure boiling 

10.6 The control of volatile substances in wort 

10.7 Energy conservation and the hop-boil 

10.8 Hot wort clarification 

10.9 Wort cooling 

10.10 The cold break 

10.11 Wort aeration/oxygenation 


11 Yeast biology 

11.1 Historical note 

11.2 Taxonomy 

11.3 Yeast ecology 

11.4 Cellular composition 

11.5 Yeast morphology 

11.6 Yeast cytology 

11.6.1 Cell wall 

11.6.2 The periplasm 

11.6.3 The plasma membrane 

11.6.4 The cytoplasm 

11.6.5 Vacuoles and intracellular membrane systems 

11.6.6 Mitochondria 

11.6.7 The nucleus 

11.7 Yeast cell cycle 

11.7.1 Yeast sexual cycle 

11.8 Yeast genetics 

11.8.1 Methods of genetic analysis 

11.8.2 The yeast genome 

11.9 Strain improvement 


12 Metabolism of wort by yeast 


12.2 Yeast metabolism ±an overview 

12.3 Yeast nutrition 

12.3.1 Water relations 

12.3.2 Sources of carbon 

12.3.3 Sources of nitrogen 

12.3.4 Sources of minerals 

12.3.5 Growth factors 

12.4 Nutrient uptake 

12.4.1 Sugar uptake 

12.4.2 Uptake of nitrogenous nutrients 

12.4.3 Lipid uptake 

12.4.4 Ion uptake 

12.4.5 Transport of the products of fermentation 


12.5 Sugar metabolism 

12.5.1 Glycolysis 

12.5.2 Hexose monophosphate (pentose phosphate) pathway 

12.5.3 Tricarboxylic acid cycle 

12.5.4 Electron transport and oxidative phosphorylation 

12.5.5 Fermentative sugar catabolism 

12.5.6 Gluconeogenesis and the glyoxylate cycle 

12.5.7 Storage carbohydrates 

12.5.8 Regulation of sugar metabolism 

12.5.9 Ethanol toxicity and tolerance 

12.6 The role of oxygen 

12.7 Lipid metabolism 

12.7.1 Fatty acid metabolism 

12.7.2 Phospholipids 

12.7.3 Sterols 

12.8 Nitrogen metabolism 

12.9 Yeast stress responses 

12.10 Minor products of metabolism contributing to beer flavour 

12.10.1 Organic and fatty acids 

12.10.2 Carbonyl compounds 

12.10.3 Higher alcohols 

12.10.4 Esters 



13 Yeast growth 


13.2 Measurement of yeast biomass 

13.3 Batch culture 

13.3.1 Brewery batch fermentations 

13.3.2 Effects of process variables on fermentation performance 

13.4 Yeast ageing 

13.5 Yeast propagation 

13.5.1 Maintenance and supply of yeast cultures 

13.5.2 Laboratory yeast propagation 

13.5.3 Brewery propagation 

13.6 Fed-batch cultures 

13.7 Continuous culture 

13.8 Immobilized yeast reactors 

13.9 Growth on solid media 

13.10 Yeast identification 

13.10.1 Microbiological tests 

13.10.2 Biochemical tests 

13.10.3 Tests based on cell surface properties 

13.10.4 Non-traditional methods 

13.11 Measurement of viability 

13.12 Assessment of yeast physiological state 



14 Fermentation technologies 


14.2 Basic principles of fermentation technology 

14.2.1 Fermentability of wort 

14.2.2 Time course of fermentation 

14.2.3 Heat output in fermentation 

14.3 Bottom fermentation systems 

14.3.1 Choice, size and shape of vessels 

14.3.2 Construction of cylindroconical vessels 

14.3.3 Operation of cylindroconical vessels 

14.4 Top fermentation systems 

14.4.1 Traditional top fermentation 

14.4.2 Yorkshire square fermentation 

14.4.3 Burton Union fermentation 

14.5 Continuous fermentation 

14.5.1 Early systems of continuous fermentation 

14.5.2 The New Zealand system 

14.5.3 Continuous primary fermentation with immobilized yeast 

14.6 Fermentation control systems 

14.6.1 Specific gravity changes 

14.6.2 Other methods 

14.7 Summary 


15 Beer maturation and treatments 


15.2 Maturation: flavour and aroma changes 

15.2.1 Principles of secondary fermentation 

15.2.2 Important flavour changes 5 

15.2.3 Techniques of maturation 

15.2.4 Flavour, aroma and colour adjustments by addition 

15.2.5 Maturation vessels 

15.3 Stabilization against non-biological haze 

15.3.1 Mechanisms for haze formation 

15.3.2 Removal of protein 

15.3.3 Removal of polyphenols 

15.3.4 Combined treatments to remove protein and polyphenols 

15.3.5 Hazes from other than protein or polyphenols 

15.4 Carbonation 

15.4.1 Carbon dioxide saturation 

15.4.2 Carbon dioxide addition 

15.4.3 Carbon dioxide recovery 

15.5 Clarification and filtration 

15.5.1 Removal of yeast and beer recovery 

15.5.2 Beer filtration 

15.6 Special beer treatments 

15.6.1 Low-alcohol and alcohol-free beers 

15.6.2 Ice beers 

15.6.3 Diet beers 


15.7 Summary 


16 Native African beers 


16.1.1 An outline of the stages of production 

16.1.2 Bouza 

16.1.3 Merissa 

16.1.4 Busaa and some other beers 

16.1.5 Southern African beers 

16.2 Malting sorghum and millets 

16.3 Brewing African beers on an industrial scale 

16.4 Attempts to obtain stable African beers 

16.5 Beer composition and its nutritional value 


17 Microbiology 


17.2 The microbiological threat to the brewing process 

17.3 Beer spoilage micro-organisms 

17.3.1 Detection of brewery microbial contaminants 

17.3.2 Identification of brewery bacteria 

17.3.3 Gram negative beer spoiling bacteria 

17.3.4 Gram positive beer spoiling bacteria 

17.3.5 Beer spoilage yeasts 

17.3.6 Microbiological media and the cultivation of 

micro-organisms 

17.4 Microbiological quality assurance 

17.5 Sampling 

17.5.1 Sampling devices 

17.6 Disinfection of pitching yeast 

17.7 Cleaning in the brewery 

17.7.1 Range of cleaning operations 

17.7.2 CIP systems 

17.7.3 Cleaning agents 

17.7.4 Cleaning beer dispense lines 

17.7.5 Validation of CIP 


18 Brewhouses: types, control and economy 


18.2 History of brewhouse development 

18.2.1 The tower brewery lay-out 

18.2.2 The horizontal brewery lay-out 

18.3 Types of modern brewhouses 

18.3.1 Experimental brewhouses 

18.3.2 Micro- and pub breweries 

18.4 Control of brewhouse operations 

18.4.1 Automation in the brewhouse 


18.4.2 Scheduling of brewhouse operations 

18.5 Economic aspects of brewhouses 

18.6 Summary 


19 Chemical and physical properties of beer 

19.1 Chemical composition of beer 

19.1.1 Inorganic constituents 

19.1.2 Alcohol and original extract 

19.1.3 Carbohydrates 

19.1.4 Other constituents containing carbon, hydrogen and 

oxygen 

19.1.5 Nitrogenous constituents 

19.1.6 Sulphur-containing constituents 

19.2 Nutritive value of beer 

19.3 Colour of beer 

19.4 Haze 

19.4.1 Measurement of haze 

19.4.2 Composition and formation of haze 

19.4.3 Prediction of haze and beer stability 

19.4.4 Practical methods for improving beer stability 

19.5 Viscosity 

19.6 Foam characteristics and head retention 

19.6.1 Methods of assessing foam characteristics 

19.6.2 Beer components influencing head retention 

19.6.3 Head retention and the brewing process 

19.7 Gushing 


20 Beer flavour and sensory assessment 


20.2 Flavour ±taste and odour 

20.3 Flavour stability 

20.4 Sensory analysis 


21 Packaging 


21.2 General overview of packaging operations 

21.3 Bottling 

21.3.1 Managing the bottle flow 

21.3.2 Managing the beer flow 

21.3.3 Managing plant cleaning 

21.3.4 Materials for making bottles 

21.4 Canning 

21.4.1 The beer can 

21.4.2 Preparing cans at the brewery for filling 

21.4.3 Can filling 

21.4.4 Can closing (seaming) 


21.4.5 Widgets in cans 

21.5 Kegging 

21.5.1 The keg 

21.5.2 Treatment of beer for kegging 

21.5.3 Handling of kegs 

21.5.4 Keg internal cleaning and filling 

21.5.5 Keg capping and labelling 

21.5.6 Smooth flow ale in kegs 

21.6 Cask beer 

21.6.1 The cask 

21.6.2 Handling casks 

21.6.3 Preparing beer for cask filling 

21.6.4 Cask filling 

21.7 Summary 


22 Storage and distribution 


22.2 Warehousing 

22.2.1 Principles of warehouse operation 

22.2.2 Safety in the warehouse 

22.3 Distribution 

22.3.1 Logistics 

22.3.2 Quality assurance 

22.4 Summary 


23 Beer in the trade 


23.2 History 

23.3 Beer cellars 

23.3.1 Hygiene 

23.3.2 Temperature 

23.3.3 Lighting 

23.4 Beer dispense 

23.4.1 Keg beer 

23.4.2 Cask beer 

23.4.3 Bottled and canned beer 

23.5 Quality control 

23.6 New developments in trade quality 

23.7 Summary 


Appendix: units and some data of use in brewing 

Table A1 SI derived units 

Table A2 Prefixes for SI units 

Table A3 Comparison of thermometer scales 

Table A4 Interconversion factors for units of measurement 

Table A5 Specific gravity and extract table 


Table A6 Equivalence between Institute of Brewing units of hot 

water extract 

Table A7 Solution divisors of some sugars 

Table A8 Some properties of water at various temperatures 

Table A9 The density and viscosity of water at various temperatures 

Table A10 Some more properties of water 

Table A11 The relationship between the absolute pressure and the 

temperature of water-saturated steam 

Table A12 The solubility of pure gases in water at different 

temperatures 

Table A13 Salts in brewing liquors 

Table A14 Units of degrees of water hardness 

Table A15 Characteristics of some brewing materials 

Table A16 Pasteurization units 

Fig. A1 The relationships between ethanol/water mixtures and the 

densities of the solutions. 

References 


Preface 

ThetwovolumesofthesecondeditionofMaltingandBrewingScienceI,MaltandSweet 

Wort and II, Hopped Wort and Beer, by James S. Hough, Dennis E. Briggs, Roger 

Stevens and Tom W. Young, appeared in 1981 and 1982. This book provided the 

framework for the M.Sc. in Malting and Brewing Science, the course that was offered by 

the British School of Malting and Brewing in the University of Birmingham (UK). It also 

provided the backbone of many other courses. After more than 20 years the demand for 

these volumes has continued, although they are increasingly out of date. Malts and 

Malting, by Dennis E. Briggs, appeared in 1998, and Brewing Yeast and Fermentation, 

by Chris Boulton and David Quain, became available in 2001. These books cover their 

named topics in depth. However, the need for an up-to-date, integrated textbook on 

brewing, comparable in scope and depth of coverage to Malting and Brewing Science, 

remained. 

Brewing:Scienceandpracticeisintendedtomeetthisneed.Decidingonthedetailsof 

the coverage has given rise to some anxious discussions. Practically it is impossible to 

describe all aspects of all the varieties of brewing processes in depth, in one moderately 

sized volume. Inevitably it has been necessary to assume some background knowledge of 

physics, chemistry, biology, and engineering. However, the book is understandable to 

people without detailed knowledge in these areas. The references at the end of each 

chapter provide guidance for further reading. Since the wide range of kinds of brewing 

operations, from simple, low-volume, single-line breweries to extremely large, highly 

complex, multiple-line installations, does not allow a single description of brewing 

activities, the book concentrates on the principles of the various brewing processes. 

Brewing is carried out all over the world and, unsurprisingly, different terminologies 

and methods of measurement and analysis are used. The different systems of units and 

analyses are explained in the text and conversion factors (where valid) and some other 

useful data are given in the Appendix. Alist of abbreviations is included in the index for 

reference. The index also includes a list of formulae 

First of all the authors warmly thank our wives, Rosemary, Wendy, Stella and Betty, 

for their unfailing patience and good-humoured support. We have also been given a great 

deal of help from our colleagues and friends. We are grateful to Mrs Doreen Hough for 


permission to use some of the late Professor Jim Hough's diagrams. Permission to use 

other diagrams is acknowledged in the text. We would like to thank: Mrs Marjorie 

Anderson, Dr John M. H. Andrews, Mrs Marjorie Anderson, Dr Raymond G. Anderson, 

Mr David J. Banfield, Mr Zane C. Barnes, Herr Volker Borngraber, Mr Andy Carter, Dr 

Peter Darby, Mr J. Brian Eaton, Dr David L. Griggs, Dr Paul K. Hegarty, Mrs Sue M. 

Henderson, Mr James Johnstone, Mr Roy F. Lindsay, Dr G. C. Linsley-Noakes, Dr David 

E. Long, Mr John MacDonald, Dr Ray Marriott, Mr P. A. (Tom) Martin, Dr A. Peter 

Maule, Ms Elaine McCrimmon, Dr Philip Morrall, Dr Ray Neve, Dr George Philliskirk, 

Dr David E. Quain, Mr Trevor R. Roberts, Mr Derek Wareham and Dr Richard D. J. 

Webster. We also wish to thank Coors Brewers for the use of the Technical Centre, 

Burton-on-Trent. 

We apologise if any acknowledgements have been omitted. 


1 

An outline of brewing 


Beersandbeer-likebeveragesmaybepreparedfromrawcerealgrains,maltedcerealgrains 

and (historically) bread. This book is primarily concerned with beers of the types that 

originated in Europe, but which are now produced world-wide. However, an account is 

given of Àfrican-style' beers (Chapter 16). The most simple preparation of European-style 

beers involves (a) incubating and extracting malted, ground up cereal grains (usually 

barley) with warm water. Sometimes the ground malt is mixed with other starchy materials 

and/or enzymes. (b) The solution obtained is boiled with hops or hop preparations. (c) The 

boiled solution is clarified and cooled. (d) The cooled liquid is fermented by added yeast. 

Usually the beer is clarified, packaged and served while effervescent with escaping carbon 

dioxide. In this chapter the preparation of beers is outlined and the brewers' vocabulary is 

introduced. Beers are made in amounts ranging from afew hectolitres (hl) aweek to 

thousands of hl. They are made using various different systems of brewing. 

1.2 Malts 

Malts are made from selected cereal grain, usually barley, (but sometimes wheat, rye, 

oats, sorghum or millet), that has been cleaned and stored until dormancy has declined 

and it is needed. It is then germinated under controlled conditions. Their preparation is 

outlined in Chapter 2, and is described in detail in Briggs (1998). The grain is hydrated, 

or `steeped', by immersion in water. During steeping the water will be changed at least 

once, air may be sucked through the grain during `dry' periods between immersions, and 

may be blown into the grain while immersed. After steeping the grain is drained and is 

germinated to a limited extent in a cool, moist atmosphere with occasional turning and 

mixing to prevent the rootlets matting together. During germination the acrospire 

(coleoptile) grows beneath the husk and rootlets grow from the end of the grain, enzymes 

accumulate and so do sugars and other soluble materials. The dead storage tissue of the 

grain, the starchy endosperm, is partly degraded, or `modified', and its physical strength 


is reduced. When germination and `modification' are sufficiently advanced they are 

stopped by kilning. The `green malt' (green in the sense of immature, it is not green in 

colour) is kilned, that is, it is dried and lightly cooked, or cured, in acurrent of warm to 

hot air. Pale, `white' malts are kilned using low temperatures and in these enzyme 

survival is considerable. In darker, coloured malts, kilned using higher temperatures, 

enzyme survival is less. In extreme cases the darkest, special malts are heated in a 

roasting drum and contain no active enzymes. After kilning the malt is cooled and 

`dressed', that is, the brittle rootlets (`culms', sprouts) are broken off and they and dust 

are removed. The culms are usually used for cattle food. Pale malts are usually stored for 

some weeks before use. In contrast to the tough, ungerminated barley grain malt is 

`friable', that is, it is easily crushed. 

1.3 Mash tun adjuncts 

Mash tun adjuncts are preparations of cereals (e.g., flaked maize or rice flakes, wheat 

flour, micronized wheat grains, or rice or maize grits which have to be cooked separately 

inthebrewery)which maybemixed withground maltinthemashingprocess. Theuseof 

an adjunct alters the character of the beer produced. An adjunct's starch is hydrolysed 

during mashing by enzymes from the malt, so providing a(sometimes) less expensive 

source of sugars as well as changing the character of the wort. Sometimes microbial 

enzymes are added to the mash. In afew countries the use of adjuncts is forbidden. In 

GermanytheReinheitsgebotstipulatesthatbeermaybemadeonlywithwater,malt,hops 

and yeast. 

1.4 Brewing liquor 

In brewing, water is commonly known as liquor. It is used for many purposes besides 

mashing, including beer dilution at the end of high-gravity brewing, cleaning and in 

raising steam. Water for each purpose must meet different quality criteria (Chapter 3). 

The brewing liquor used in mashing must be essentially `pure', but it must contain 

dissolved salts appropriate for the beer being made. The quality of the liquor influences 

the character of the beer made from it. Famous brewing locations gained their 

reputations, at least in part, from the qualities of the liquors available to them. Thus 

Burton-on-Trent is famous for its pale ales, Dublin for its stouts and Pilsen for its fine, 

pale lagers. It is now usual, at least in larger breweries, to adjust the composition of the 

brewing liquor (Chapter 3). 

1.5 Milling and mashing in 

The malt, sometimes premixed with particular adjuncts, is broken up to acontrolled 

extent by milling to create the `grist'. The type of mill used and the extent to which the 

malt (and adjunct) is broken down is chosen to suit the types of mashing and wortseparation 

systems being used (Chapter5).If drymillingisused the grist, possibly mixed 

with adjuncts, is collected in a container, the grist case. 

At mashing-in (doughing-in) the grist is intimately mixed with brewing liquor, both 

flowing at controlled rates, into a mashing vessel at an exactly controlled temperature. 


The resulting `mash', with the consistency of athin slurry, is held for aperiod of 

`conversion'. The objective is to obtain amash that will yield asuitable `sweet wort', a 

liquidrichinmaterialsdissolvedfromthemaltandanyadjunctsthathave beenused.The 

dissolved material, the èxtract', contains soluble substances that were preformed in the 

grist and other substances (especially carbohydrates derived from starch), that are formed 

from previously insoluble materials by enzyme-catalysed hydrolytic breakdown during 

mashing. 

1.6 Mashing and wort separation systems 

The major mashing systems are, broadly, (a) the simplest, nearly isothermal, infusion 

mashing system, (traditional for British ale brewers); (b) the decoction system, 

(traditional for mainland European lager brewers); (c) the double mash system, (common 

in North American practice); (d) the temperature-programmed infusion mashing system 

that is being widely adopted in the UK and mainland Europe (Chapters 4and 5). Amash 

should be held at a chosen temperature (or at successive different temperatures), for predetermined 

times, to allow enzymes to `convert' (degrade) the starch and dextrins to 

soluble sugars, to cause the partial breakdown of proteins, to degrade nucleic acids and 

other substances. At the end of mashing the sweet, or unhopped wort (the solution of 

extractives, mainly carbohydrates; the èxtract') is separated from the undissolved solids, 

the spent grains or draff. 

Infusion mashing is carried out in mash tuns. Mash conversion and the separation of 

the sweet wort from the spent grains take place in this vessel. The coarsely ground grist, 

made with a high proportion of well-modified malt, is mashed in to give a relatively 

thick, porridge-like mash at 63 67 ëC (145.4 152.6 ëF). After a stand of between 30 

minutes and two and a half hours the wort (liquid) is withdrawn from the mash. The first 

worts are cloudy and are re-circulated, but as the run off is continued the wort becomes 

`bright' (clear), because it is filtered through the bed of grist particles. When bright the 

wort is either collected in a holding vessel (an underback) or it is moved directly to a 

copper to be boiled with hops. Most of the residual extract, initially entrained in the wet 

grains, is washed out by sparging (spraying) hot liquor, at 75 80 ëC (167 176 ëF ) over 

the goods. 

Decoction mashing is carried out with more finely ground grists, originally made with 

malts that were undermodified. These mashes are relatively `thin', so they may be moved 

by pumping and can be stirred. Decoction mashing uses three vessels, a stirred mash 

mixing vessel, a stirred decoction vessel or mash cooker and a wort separation device, 

either a lauter tun or a mash filter. In one traditional mashing programme the grist is 

mashed in to give an initial temperature of around 35 ëC (96 ëF). After a stand a decoction 

is carried out, that is, a proportion of the mash, e.g., a third, is pumped to the mash 

cooker, where it is heated to boiling. The boiling mash is pumped back to the mash 

mixing vessel and is mixed with the vessels contents, raising the temperature to, e.g., 

50 ëC (122 ëF). After another stand a second decoction is carried out, increasing the 

temperature of the mixed mash to about 65 ëC (149 ëF). A final decoction increases the 

mash temperature to about 76 ëC (167 ëF). The mash is then transferred to a lauter tun or a 

mash filter. The sweet wort and spargings are collected, ready to be boiled with hops. 

Typically, double-mashing uses nitrogen- (`protein-') and enzyme-rich malts and 

substantial quantities of maize or rice grits. It also involves the use of three vessels. Most 

of the malt grist is mashed into a mash-mixing vessel to give a mash at around 38 ëC 


(100.4ëF). The grits, mixed with asmall proportion of ground malt and/or apreparation 

of microbial enzymes, are mashed in aseparate vessel called acereal cooker. The 

contents are carefully heated with mixing, and arest at about 70ëC (158ëF), to 100ëC 

(212ëF) to disperse the starch and partly liquefy it. The adjunct mash is pumped from the 

cereal cooker into the malt mash, with continuous mixing, to give afinal temperature of 

about 70ëC (158ëF). After astand the mash is heated to about 73ëC (163.4ëF), then it is 

usually transferred to alauter tun for wort collection. 

Temperature-programmed infusion mashing is increasingly displacing older mashing 

systems. The grist is finely ground and the mash is made `thin' to allow it to be stirred. 

The grist is mashed into astirred and externally heated mash-mixing vessel to give an 

initial temperature of 35ëC (95ëF) for apoorly modified malt or 50ëC (122ëF), or more, 

for abetter modified malt. The mash is heated, with `stands' typically at 50ëC (122ëF), 

65ëC (149ëF) and 75ëC (167ëC). Then the sweet wort is collected using alauter tun or a 

mash filter. 

1.7 The hop-boil and copper adjuncts 

The sweet wort is transferred to avessel, acopper or kettle, in which it is boiled with hops 

or hop preparations, usually for 1±2 hours. Hops are the female cones of hop plants. They 

may be used whole, or ground up, or as pellets or as extracts. The choice dictates the type 

of equipment used inthe next stage of brewing. Pelleted powders are often preferred. Hops 

contribute various groups of substances to the wort. During boiling anumber of changes 

occurinthewortofwhichthemoreobviousarethecoagulationofproteinas`hotbreak'or 

`trub', the gaining of bitterness and hop aroma and the destruction of micro-organisms 

(Chapters9and10).Evaporation ofthe wort,reduces the volumeby,say,7±10%, andsoit 

is concentrated. Unwanted flavour-rich and aromatic volatile substances are removed. 

When used, sugars, syrups and even malt extracts (copper adjuncts) are dispersed and 

dissolve in the wort during the copper boil. During the boil flavour changes and a 

darkening of the colour occurs. Caramels may be added at this stage to adjust the colour. 

The hop-boil consumes about half of the energy use in brewing. 

1.8 Wort clarification, cooling and aeration 

At the end of the boil the transparent, or `bright' wort contains flocs of trub (the hot 

break) and suspended fragments of hops. If whole hops were used then residual solids are 

strained off in a hop back or other filtration device and the bed of hop cones filters off the 

trub, giving a clear, hopped wort. However, if powders, hop pellets, (which break up into 

small particles), or extracts were used then hop fragments (if present) and the trub are 

usually separated in a `whirlpool tank'. The clear `hopped wort' is cooled to check 

continuing darkening and flavour changes and so it can be inoculated (`pitched') with 

yeast, and can be aerated or oxygenated without a risk of oxidative deterioration. The 

heated cooling water is used for various purposes around the brewery. During cooling a 

second separation of solids occurs in the wort. This `cold break' is composed mostly of 

proteins and polyphenols and some associated lipids. It is often, but not always, 

considered desirable to remove this material to give a `bright', completely clear wort. The 

wort is aerated or even oxygenated, to provide oxygen for the yeast in the initial stages of 

fermentation. 


1.9 Fermentation 

Fermentation may be carried out in many different types of vessel (Chapter 14; Boulton 

and Quain, 2001). Fermenters may be open or completely closed or they may allow part 

of the yeast to be exposed to the air for part of the fermentation period. The variety of 

fermenters remains because yeasts working in different vessels produce beers with 

different flavours. Wort fermentation is initiated by pitching (inoculating) the cooled, 

hopped wort with aselected yeast. In afew cases mixtures of yeasts are used. Brewery 

yeast is amass of tiny, single, ovoid cells (Saccharomyces cerevisiae, the `sugar fungus 

of beer'). Yeast strains vary in their properties and the flavours they impart.In avery few 

cases, as with Belgian Gueuze and Lambic beers, (or some African beers; Chapter 16), 

fermentation occurs `spontaneously' and a complex mixture of microbes is involved. The 

yeast metabolizes extract substances dissolved in the wort. More yeast cells and `minor' 

amounts of many substances are produced, some of which add to the beer's character. 

The major products of carbohydrate metabolism are ethyl alcohol (ethanol), carbon 

dioxide and heat. The yeast multiplies around 3±5 times. Some is retained for use in 

subsequent fermentations, while the surplus is disposed of to distillers or the makers of 

yeast extracts. 

Traditionally, ales are fermented with `top yeasts' which rise to the top of the beer in 

the head of foam. These are pitched at about 16 ëC (61 ëF) and fermentation is carried out 

at 15 20 ëC (59 68 ëF) for 2±3 days. Traditional lagers are fermented with `bottom 

yeasts', which settle to the base of the fermenter. These are pitched at lower temperatures 

(e.g., 7 10 ëC; 44.6 50 ëF) and fermentations are also carried out at lower temperatures 

(e.g., 10 15 ëC; 50 59 ëF), consequently they take longer than ale fermentations. As 

wort is converted into beer the removal of materials (especially sugars) from solution and 

the appearance of ethanol both contribute to the decline in specific gravity. The initial or 

original gravity, OG, the final or present gravity at the end of the fermentation, FG or PG, 

and the final alcohol content, are important characteristics of beers. 

Yeasts are selected with reference to: 

1. their rate and extent of growth 

2. the rate and extent of fermentation 

3. the flavour and aroma of the beer produced 

4. in older fermentation systems it is imperative that top yeasts rise into a good head of 

foam and bottom yeasts sediment cleanly. 

Substances (finings) may be added to promote yeast separation at the end of 

fermentation. However, in some modern systems `powdery' yeasts are employed that 

stay in suspension until the beer is chilled or until collected by centrifugation. 

1.10 The processing of beer 

When the main, or `primary' fermentation is nearly complete the yeast density is reduced 

to a pre-determined value. The `green' or immature beer (it is not green in colour, but has 

an unacceptable, ìmmature' flavour) is held for a period of maturation or secondary 

fermentation. During this process the flavour of the mature beer is refined. Sometimes 

`priming' sugar or a small amount of wort is added to boost yeast metabolism and the 

`maturation', `conditioning' or `lagering' process. (Lagern is German and means stored 

or deposited). In traditional lager brewing the immature beer was stored cold, e.g., at 


2 ëC (28.4 ëF), for extended periods, sometimes months, when a very slow secondary 

fermentation occurred and yeast and cold trub settled to the base of the storage vessel. 

Conditioning is carried out in various ways. The primary and secondary fermentations 

were carried out in separate, special vessels but increasingly single vessels are used. 

Traditionally, ales are run from fermenters into casks or bottles with a little sugar, finings 

and a regulated amount of yeast. The secondary fermentation `conditions' the beer in the 

container, charging it with carbon dioxide. The ale is dispensed from above a layer of 

settled yeast. Such naturally conditioned beers are now made in only small amounts. 

These beers are not stable for extended periods and they require careful, intelligent 

handling. 

Now, after conditioning in bulk, most beers are chilled and filtered or centrifuged to 

remove residual yeast. These completely bright beers are then carbonated, that is, their 

carbon dioxide content is adjusted, they are transferred into bottles, cans, kegs, or bulk 

tanks. Nitrogen gas is sometimes added to the package, so the beer contains both this and 

carbon dioxide, but as far as possible air is excluded. Before packaging the beer may be 

sterile filtered, a process that avoids flavour damage but it follows that all subsequent 

beer movements must be made under rigidly aseptic conditions. More often the beer is 

pasteurized, that is, it is subjected to a carefully regulated heat treatment. This may be 

applied to the filled bottles or cans or to the flowing beer as it moves to fill a sterile 

container. With the notable exceptions of some dark stouts and wheat beers, such beers 

should (a) be brilliantly clear, (b) develop a stable white foam, or head, when poured into 

a clean glass, and (c) their flavours and gas-contents should remain steady. 

The careful selection of raw materials and processing conditions help brewers to 

approach these objectives. However, it may be necessary to employ other techniques. For 

example, the plant proteolytic enzyme papain may be added to beer, or the beers may be 

treated with insoluble adsorbents to remove haze precursors. In addition substances may 

be added to reduce the dissolved oxygen content of the beer, to maximize its haze and 

flavour stability. Other substances may be added to stabilize beer foam. 

1.11 Types of beer 

There is no truly satisfactory classification of beers. `Clear, European-style beers' may be 

distinguished by the raw materials used in their preparation, the ways in which the 

brewing operations are carried out, whether top, bottom or `bulk' fermentation is used, 

how the product is conditioned, whether it is chilled and filtered and carbonated or is 

conditioned in bottle or cask and how it is packaged. Stouts, porters and wheat beers, 

which are produced in conventional ways, are often not transparent. A beer may also be 

distinguished by its OG and degree of attenuation or alcohol content, colour, acidity, 

flavour and aroma, by its `body' or `mouth feel', by its head (foam) characteristics and by 

its physiological effects. How a drinker perceives a beer is influenced by many factors, 

including the manner in which it is served, its temperature, clarity and colour, flavour, 

aroma and `character', the ambience, and whether or not it is being taken with food and 

what has been consumed before. 

Within each grouping, `class' or `style', individual beers may be quite distinct and 

brewers aim to produce distinctive products. In North America most beer is pale, lightly 

hopped and served very cold (often at about 0 ëC; 32 ëF). Many new, small breweries have 

been set up and these make a wide variety of beers based on styles from around the world. 

In Europe, for about a century, British brewing practices diverged from those of mainland 


producers, but in recent years convergence has started. For example, in Germany most 

beers(notall)weremadeusingdecoctionmashing,bottomfermentationandlongperiods 

of cold storage (lagering). Increasingly, temperature programming, infusion mashing, 

bulk fermentation and shorter periods of lagering are being used. Under many 

circumstances the use of adjuncts is still not used. Although under EEC legislation the 

use of adjuncts is allowed, most German brewers still abide by the Reinheitsgebot for 

domestic beers. 

ThemaingroupsofbeersaretheverypalePilsentypes,palegolden-brownViennatypes, 

and the darker, rich Munich types. Other beers include MaÈrzen, Oktoberfest, wheat beers, 

ryebeersandsmokedbeers.IntheUKlagerworts areproduced inmanyways,buttheyare 

bottomfermented.TheBritish`lagers'areallpalebeers.Alesaretraditionallymadewithan 

infusion mashing system. They are moderately strongly hopped and atop fermentation 

systemisused.Traditionalgroupsarethe(progressivelydarker)paleales,mildales(usually 

darker,sweeterandlessstronglyhopped),brownales(darkerformsof`mild'),andstoutsor 

`porters'. The distinctionsbetween ale andlager breweries are increasingly blurred assome 

brewers adopt similar wort production and fermentation systems. 

Less common products include wheat beers, low-alcohol and alcohol-free beers 

(which may be carbonated worts, underfermented beers or beers from which the alcohol 

hasbeenremoved),andbeerswithexceptionallyhighalcoholcontents(e.g.,barleywines 

and Trappist beers, with 9% ABV, or more). In low carbohydrate (lite, light or dietetic) 

beers, prepared by using special mashing conditions and added starch-degrading 

enzymes, essentially all the starch-derived dextrins are degraded to fermentable sugars 

and are utilized by the yeast. African opaque beers (Chapter 16) and kvass (Russian) are 

distinct products. Some unusual beers, made in Belgium, include Lambic, Gueuze and 

fruit-flavoured beers (kriek, flavoured with cherries; framboise, flavoured with 

raspberries). These are all made using spontaneous fermentations which involve mixtures 

of organisms. 

Beer strength may be defined in several ways; by the specific gravity of the wort before 

fermentation (the OG), by the alcohol content of the final beer (% alcohol by volume or 

ABV) or even by the content of hop bitter substances. The fermentability of extract 

depends on many factors. There is no fixed relationship between the OG and the alcohol 

content of abeer. In Britain the specific gravity of awort or beer is usually quoted times 

1,000 so, for example, water has aSG of 1000.00 and wort with aspecific gravity (s.g.) of 

1.040 at 20ëC (68ëF) has aSG of 1040.00. In the past, extract was calculated as brewer's 

pounds per barrel, and the excess weight (in lb.) over water was referred to as brewer's 

poundsgravity.Thusabarrelofwater(36imp.gallons,UK)weighs360pounds(lb.),buta 

barrel of wort at SG 1040 weighed 374.2lb. So this wort had agravity of 14.2lb. Outside 

the UK concentrations are often expressed in terms of concentrations of sucrose solutions 

of the same gravity (see appendix). Thus, in von Balling's tables of 1843 wort of aspecific 

gravity of 1040 is equivalent to a sucrose solution of 9.95% (w/w). Von Balling's tables 

were revised by Plato in 1918 and gravity is often expressed as degrees Plato. Increasingly 

beer strengths are being given as the concentration of alcohol % ABV that they contain. 

1.12 Analytical systems 

For both trading and quality-control purposes all the materials used in making beers, the 

liquor, the sweet and hopped worts and the beers themselves are analysed. Not all the 

methods used are standardized and, regrettably, there are at least four àgreed', but 


discordant, sets of methods in use. The methods used in the different sets differ 

significantly and give different results. In many instances there are no valid or reliable 

conversion factors to interconvert analytical results. The most commonly used methods 

are those of the Institute of Brewing (IoB; now the Institute and Guild of Brewing, IGB), 

the European Brewery Convention (EBC), the American Society of Brewing Chemists 

(ASBC) and the methods of the MitteleuropaÈischen Analysen Kommission (MEBAK). 

The methods are frequently revised, successive versions being distinguished by their 

dates.Inthisbookthemostrecent unitsareusedwhereverpossible.By2005themethods 

of the IGB and the EBC should have been merged. The number of units of measurement 

in use is large. Here metric units have been used where possible, with British (UK) 

equivalents so, for example, hectolitres and imperial gallons. It should be noted that the 

American gallon (US) has only about 0.8 of the volume of an imperial gallon. Systems of 

units and conversion factors are given in the Appendix. 

1.13 The economics of brewing 

The economics of brewing are influenced by many factors, including the manning levels 

required, the local costs of labour, raw materials, how brewing practices are influenced 

by governmental regulations and how the products are taxed. The scales of brewery 

operations vary widely, from units that produce < 10 barrels (imp. brl, approx.16.4 

hectolitres, hl) per week to > 30,000 imp. brl (49,092 hl) per week. Thus savings per imp. 

brl that are trivial to the small-scale brewer are worthwhile to a larger operator. Breweries 

that operate continuously, for 24 hours a day, use their capital investment in plant to the 

best effect and they can also make other savings, for example, by using heat-recovery 

systems that are not suitable for breweries that operate intermittently. There are strong 

and increasing pressures to minimize water use, to minimize the production of wastes and 

effluents and the release of heat and odorous gases (such as vapours from hop-boiling), 

and `greenhouse gases' such as carbon dioxide and refrigerants, to utilize raw materials as 

efficiently as possible, and to utilize fuels and power efficiently. 

In order to use plant at peak efficiency it is necessary to have it well engineered, 

instrumented, automated and maintained so that it can operate nearly continuously. To 

make such investment worthwhile the capacity of the plant must be large and, in 

consequence, the manpower needed to produce a given volume of beer is lower than is 

needed with less sophisticated plant. The personnel needed to operate modern plant 

successfully must be highly trained. Such plant is most efficient at making large volumes 

of relatively few beers. Smaller, more labour-intensive plants are often better suited for 

making a wide variety of beers in smaller amounts. Large brewing companies tend to 

produce fewer beers in larger and larger plants. The problems of `product matching', of 

trying to make large volumes of one beer in different breweries, are notorious. Smaller 

breweries, making smaller volumes, often of more `specialist', and even èccentric' 

beers, are appearing all the time. Smaller breweries usually deliver beer over a small area, 

and so have lower transportation costs relative to larger breweries, which must deliver to 

larger areas to market the larger amounts of beer that they produce. 

Energy and water requirements per unit volume of beer produced vary widely. In part 

this is due to differences in the efficiencies of production plants, but it also depends on 

the production processes used and on how the beer is packaged. Thus decoction mashing 

uses more energy than temperature-programmed infusion mashing. Not all breweries 

recover heat from the vapours in their mash-cooker or copper-stacks, and the efficiency 


of heat recovery varies with the sophistication of the equipment used. The heat, power 

and water usage in bottling and canning halls (which not all breweries have) is high, 

because of the amount of washing carried out, the conveying and the heat used by the 

pasteurizers. 

A widely adopted technique for improving the economics of a brewery is `high gravity' 

(HG) brewing, in which concentrated worts are produced and processed. The concentrated 

beers produced are diluted for sale. Thus a larger volume of beer is produced, per brew, 

than would have been the case had the plant been operated in the conventional way. HG 

brewing is a technically sophisticated process. There are difficulties with preparing 

concentrated worts unless the addition of sugars or syrups to the copper is allowed. Almost 

all the stages of the brewing process have to be adjusted, and the water used to dilute the 

HG beer must be very carefully sterilized, deoxygenated and carbonated. 

1.14 Excise 

Beer is usually taxed. In Britain malt was taxed and the regulations imposed, to maximize 

the tax receipts, fossilized the malting and brewing processes. The malt tax was 

withdrawn in 1880, but the styles of beers that had been produced using well modified ale 

malts were established as `traditional' and continued in use. Only recently have newer 

methods of brewing been widely adopted. Next, tax was levied on the gravity and volume 

of the brewer's wort, after boiling and cooling. The consequent economic need to convert 

as high a proportion of this wort as possible into saleable beer influenced the designs of 

fermentation vessels and yeast propagators, the recovery of beer from harvested yeast, 

from filters, and so on. At present in the UK, and many other countries, excise is levied 

on the volume and alcohol content (ABV) of the beer leaving the brewery. Sometimes 

beers are classified according to the alcohol band (range of strengths) in which it falls. 

Each band is taxed at a different rate and the tax increases with the alcohol content. There 

are countries where the beer is taxed by volume only. 

1.15 References and further reading 

1.15.1 The systems of malting and brewing analysis 

ASBC (1992) The American Society of Brewing Chemists. Methods of analysis (8th edn, revised), ASBC, 

St. Paul, Minn. 

EBC (1997; 1998) European Brewery Convention. Analytica-EBC (5th edn, with revisions), Fachverlag 

Hans Carl, NuÈrnberg. 

IoB (1997) The Institute of Brewing, Recommended Methods of Analysis (2 volumes, and revisions), The 

Institute and Guild of Brewing, London. 

MEBAK (1993) Brautechnische Analysenmethoden: Methodensammlung der MitteleuropaÈischen 

Brautechnischer Analysenkommission (5 volumes). 

1.15.2 General 

BAMFORTH, C. W. (1998) Beer; tap into the art and science of brewing, London, Insight Press, 245 pp. 

BAMFORTH, C. W. (2002) Standards of Brewing: a practical approach to consistency and excellence, 

Boulder, Colorado, Brewers' Publications, 209 pp. 

BOULTON, C. and QUAIN, D. (2001) Brewing Yeast and Fermentation, London, Blackwell Science, 644 pp. 

BRIGGS, D. E. (1998) Malts and Malting, London, Blackie Academic and Professional/Gaithersburg, 

Aspen Publishing, 796 pp. 

COULTATE, T. P. (2002) Food, the chemistry of its components, (4th edn), Cambridge, The Royal Society 

of Chemistry, 432 pp. 


HLATKY, C. and HLATKY, M. (1997) Bierbrauen zu Hause, Graz, Leopold Stocker Verlag, 177 pp. 

HORNSEY, I. S. (1999) Brewing, Cambridge, The Royal Society of Chemistry, 231 pp. 

HORNSEY, I. S. (2003) A History of Beer and Brewing, Cambridge, The Royal Society of Chemistry, 742 

pp. 

KUNZE, W. (1996) Technology, Brewing and Malting (International edn, translated Wainwright, T.), 

Berlin, VLB, 726 pp. 

LEWIS, M. L. and YOUNG, T. W. (2003) Brewing (2nd edn), New York, Kluwer Academic, 398 pp. 

MEISEL, D. (1997) A Practical Guide to Good Lager Brewing Practice, Hout Bay, South Africa, The 

Institute of Brewing, Central and Southern African Section. 

MOLL, M. (1994) Beers and Coolers (English edn, translated Wainwright, T.), Andover, Intercept, 495 pp. 

SYSILAÈ , I. (1997) Small-Scale Brewing. Brew your own beer, Helsinki, Limes, 278 pp. 

WAINWRIGHT, T. (1998) Basic Brewing Science, Reigate, Wainwright, 317 pp + appendices. 


2 

Malts, adjuncts and supplementary enzymes 

2.1 Grists and other sources of extract 

The sources of extract used in brewing are materials used in the mash and materials 

dissolved during the hop-boil (Chapter 1). In addition, small amounts of sugars may be 

added to beers as primings or for sweetening. Caramels, coloured malt extracts and 

Farbebier may also be added to adjust colours. Supplementary enzymes, derived from 

non-malt sources, may be added to the mash or at later stages of beer production. Malt is 

thetraditional source ofenzymesandtheextractproducedinmashing(Chapters1and4). 

The contents of this chapter are discussed in more detail elsewhere (Briggs, 1998; 

Brissart et al., 2000). 

2.2 Malting 

2.2.1 Malting in outline 

Barley (Hordeum vulgare) is the cereal grain most often malted. Wheat (Triticum 

aestivum)andsorghum(Sorghumvulgare)arealsomaltedinnotablequantities (thelatter 

in Africa), but small amounts of rye (Secale cereale), oats (Avena sativum) and millets 

(various spp.) are also used. The barley grain or corn has acomplex structure (Briggs, 

1978, 1998, Figs 2.1 and 2.2), and is asingle-seeded fruit (a caryopsis). Barley varieties 

differ in their suitabilities for malting. Barley plants are annual grasses. Some are planted 

in the autumn (winter barleys) while others are planted in the spring (spring barleys). 

Grains are arranged in rows, borne on the head, or ear. The number of rows varies, being 

two in two-rowed varieties and six in six-rowed forms. In mainland Europe winter 

barleys are usually of poor malting quality, but some of the two-rowed winter varieties 

grown in the UK (such as Maris Otter, Halcyon and Pearl) are of outstandingly good 

quality. Good spring malting barleys include Alexis, Chariot, Optic and Prisma. Grains 

vary in size, shape and chemical composition. 

It is important to understand that malts consist of mixtures of grains with differing 

properties. This heterogeneity, which is reflected in the malt, can give rise to problems in 


Awn 

Starchy endosperm 

Distal 

end 

Dorsal side 

Aleurone 

layer Lemma Pericarp Testa 

Sub-aleurone 

region Palea 

Crushed 

cell layer 

Ventral furrow side 

Acrospire 

Embryo 

Scutellum 

Rootlets 

Coleorhiza 

Rachilla 

Micropylar region 

Broken pedicel 

Scutellar epithelium 

Proximal 

end 

Fig. 2.1 A schematic longitudinal section of a barley grain, to one side of the ventral furrow and 

the sheaf cells (after Briggs et al., 1981). 

Sheaf cells 

Pigment 

strand 

Dorsal side 

Rachilla Ventral 

furrow 

Ventral side 

Lemma 

Vascular 

bundles 

Pericarp and testa 

Palea 

Aleurone layer 

Sub-aleurone 

layer 

Central region 

Vascular bundle 

Starchy 

endosperm 

Fig. 2.2 A diagram of a transverse section of a plump barley grain, taken at the widest part (after 

Briggs et al., 1981). 

brewing. Barley dimensions vary, usually in the ranges: lengths, 6 12 mm, 0.24 

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. 

Two-rowed malting barley grains may have one thousand corn dry weights (TCW) in the 

range 32 44 g, and some six-rowed barleys have values of about 30 g. Differences 

between grain sizes must be allowed for when setting brewer's mills. The barley corn is 

elongated and tapers at the ends (Figs 2.1, 2.2). The dorsal, or rounded side is covered by 

the lemma, while the ventral, grooved or furrow side is covered by the palea. Together 

these units constitute the husk. The lemma has five longitudinal ridges, or `veins' running 

along it while the palea has two. In threshed grain the apical tip of the lemma is crudely 

broken off. In the unthreshed grain this is where the extended awn is attached. At the base 

of the grain, where it was attached to the plant, the rachilla, or basal bristle, lies in the 


ventral furrow. Rachillae vary greatly in their shapes and sizes, and are of use in helping 

to identify grain variety. The husk protects the grain from physical damage. In wheat, rye, 

sorghum and millets (and in some few `naked' barleys, which are not malted) husks are 

absent in threshed grain, so the corns are easily damaged. 

Within the husk the multi-layered pericarp also has a protective function. Finally, the 

testa is the layer that `seals' the interior of the grain from the exterior and limits the 

inward and outward movements of dissolved substances, such as sugars, amino acids, 

salts and proteins. This layer invests the entire interior of the grain except at the embryo, 

where its structure is modified in the micropylar region, and in the furrow, where the two 

edges are sealed together by the pigment strand. The testa consists of two cuticularized 

layers between which polyphenolic proanthocyanidins usually occur. At the base of the 

grain, over the embryo and between the pericarp and the husk, there are two small, hairy 

structures, the lodicules. During steeping these may distribute water over the embryo, by 

capillarity. Their varied forms make them valuable aids in identifying a grain's variety. 

Within the testa, at the base of the grain, is the small embryo. This is situated towards 

the dorsal side of the grain. The embryonic axis consists of the coleoptile (the maltster's 

àcrospire') pointing towards the apex of the grain and the root sheath (coleorhiza) which 

surrounds several (typically five) embryonic roots. This appears at the end of the grain, at 

the onset of germination, as the `chit'. The axis is the part of the embryo that can grow 

into a small plant. It is recessed into an expanded part of the embryo called the scutellum 

(Latin, `little shield'). Unlike the scutellum in oats, in barley this organ does not grow. Its 

inner surface, which is faced with a specialized epithelial layer, is pressed against the 

largest tissue of the grain, the starchy endosperm. With the exception of the embryo all 

the tissues mentioned so far are dead. All the surface structures, outside the testa, are 

infested with mixed populations of micro-organisms. 

The starchy endosperm is a dead tissue of thin-walled cells packed with starch 

granules embedded in a protein matrix. The granules occur in two size ranges (usually 

with diameters 1.7 2.5 m and 22.5 47.5 m), which behave differently during malting 

and brewing. The cell walls are mainly -glucans, with some pentosans and a little 

holocellulose. This tissue contains most of the grain's reserves, although others are 

present in the embryo and in the aleurone layer. In transverse section the cell walls radiate 

outwards from a `crest' of sheaf cells that run along the grain, above the pigment strand. 

These sheaf cells are devoid of contents and consist of cell walls pressed together, at least 

in the dry grain. They are not part of the endosperm tissue, the cell walls of which are 

more readily degraded by enzymes (Briggs, 2002). The outer region of the starchy 

endosperm, the sub-aleurone layer, is relatively richer in protein (including -amylase) 

and small starch granules but poor in large starch granules. Where the starchy endosperm 

fits against the scutellum the cells are devoid of contents and the cell walls are pressed 

together, comprising the crushed-cell or depleted layer. The starchy endosperm, away 

from the sheaf cells, is surrounded by the aleurone layer (which botanically is also 

endosperm tissue). On average it is about three cells thick. The cells are alive but do not 

multiply or grow during germination, have thick cell walls and contain reserves of lipids 

(fat) and protein, sucrose and possibly fructosans, as well as a full range of functional 

organelles. They do not contain any starch. A reduced layer of aleurone tissue, a single 

layer of flattened cells, extends partly over the surface of the embryo. The estimates are 

approximate, but on a dry weight basis (d.b.) a two-rowed barley corn may consist of 

husk + pericarp + lodicules, 9 14%; testa, 1 3%; embryo, 2 3.5%; aleurone layer, 

about 5%; starchy endosperm + sheaf cells, 76 82%. Malting can be understood only by 

reference to the grain structure and the interactions which occur between the tissues. 


Barley is purchased in large amounts. The grain delivered must be of the correct 

quality, i.e., it must match or exceed in quality asample seen in advance or an agreed 

specification. The evaluation of the grain involves both visual and laboratory 

assessments. Each delivery should be checked before it is unloaded. Delivery may be 

by railway, barge or (most usually in the UK) by lorry. The grain will be uncovered and 

inspected for infesting insects, local wetting, admixture of varieties, the presence of ergot 

sclerotia (poisonous, grain-sized structures produced by the fungus Claviceps purpurea), 

or any sign of heavy fungal attack. If any of these faults is noted the load is likely to be 

rejected and, if insects are present, the load will be ordered off the premises. With the 

exceptionofvarietieswithblue-pigmentedaleuronelayers,(whichappear greenishasthe 

blue is viewed through the yellow husk), grain should appear `bright', with aclean strawyellow 

colour. Discoloration is caused by heavy microbial contamination. 

Samples of the grain bulk are drawn and sent to the laboratory. The moisture content 

willbedetermined.IntheUKthegrainwillbeinspectedtocheckthatitispredominantly 

(e.g., >97%) of one specified variety, that its viability or germinative capacity (GC; 

checked by tetrazolium staining) is equal to or exceeds the specified limit (at least 98%) 

and that the total nitrogen content (TN) or crude protein content (6.25 TN) is within 

specified limits. Grain moisture and nitrogen contents are usually checked using nearinfra-red 

spectroscopy (NIR), but slower methods may be used. The grain will also be 

checked for `pre-germination', since grain that has already started to germinate will not 

keep or malt well. Asample will be graded (screened) by shaking on aset of slotted 

sieves, usually with slot widths of 2.2 or 2.25, 2.5, and 2.8mm. In North America the slot 

sizes are 7/64 in., 6/64 in. and 5/64 in. (about 2.78, 2.38 and 1.98 mm, respectively). The 

sample must have an acceptable size distribution. Grain, dust and rubbish passing the 

2.2mm or other agreed screen is regarded as `screenings', or thin corns. It will not be 

malted and so will have to be removed, collected and sold as animal feed. If screenings 

exceed aspecified weight percentage the load may be rejected or purchased at areduced 

price. 

Each lorry-load of grain (typically 20 25t) will be evaluated on afew hundred grams 

of grain. For the results to have any statistical validity, because of the inherent 

inhomogeneity of grain, the samples must be drawn, mixed and sub-divided strictly in 

accordancewiththerulessetoutinthesetsofanalyticalmethods(Section1.15.1,p.9).If 

the load is acceptable it will be unloaded and transferred to a `green grain' store. The 

grain is not green in appearance but at this stage it has not been pre-cleaned, dried, 

screened or further graded. Grain is best handled and stored in batches, separated by 

variety, TN, and grade. After thorough cleaning, drying and perhaps more extended 

storage the grain will receive a more thorough laboratory evaluation. These check 

procedures take days, compared for the checks carried out at grain intake, for which only 

a few minutes are available. Efforts are made to ensure that the grain does not carry 

unacceptable levels of residues of insecticides, fungicides, plant growth regulators, or 

herbicides by checking the grain's history with suppliers. Some grain samples will be sent 

to specialized laboratories to check residue levels. 

2.2.2 Changes occurring in malting grain 

Before malting, grain is screened and aspirated to remove large and small impurities and 

`thin' corns. To initiate malting it is hydrated. This is achieved by `steeping', immersing 

the grain in water or `steep liquor'. Later, the moisture content may be increased by 

spraying or `sprinkling' the grain. The steep-water temperature should be controlled. At 


elevated temperatures water uptake is faster but microbial growth is accelerated and the 

grain may be damaged or killed. The best temperature for steeping immature (partly 

dormant) grain is low (about 12 ëC, 53.6 ëF). For less dormant grain a value of 16 18 ëC 

(60.8 64.4 ëF) is often used. As the grain hydrates it swells to 1.3 1.4 times its original 

volume. To prevent it packing tightly and wedging in the steep it may be loosened and 

mixed by blowing air into the base of the steeping vessel. This also adds oxygen to the 

steep liquor. The oxygen is rapidly taken up, both by the grain and by the microbes that 

multiply on the grain and in the liquor. Material is leached from the grain and enzymes 

from the microbes start to degrade the materials in the grain surface layers. Thus the 

liquor contains an increasing number of microbes, microbial metabolites and dissolved 

substances, it becomes yellow, gains a characteristic smell and may froth. 

Some of the substances and the microbes in the liquor check grain germination. 

Infestations of microbes are undesirable. They compete with the grain for oxygen and 

reduce the percentage germination and germination vigour. Some produce plant growth 

regulators (including gibberellins) which stimulate or inhibit malting, others may produce 

mycotoxins which damage yeasts and/or are toxic to human beings. Some produce agents 

which cause beer to gush (over-foam), they produce some hydrolytic enzymes which may 

improve malt performance in the mash tun. Contamination with bacteria may give rise to 

worts and beers which are hazy with suspended, dead microbes (Schwarz et al., 2002; 

Walker et al., 1997). 

Steep water, which checks grain germination and growth if re-used, is periodically 

drained from the grain and replaced with fresh. The minimum acceptable number of 

water changes are used since both the supply of fresh water and the disposal of steep 

effluent are costly. Sterilants are not routinely used in steeps, but many substances, 

including mineral acids, potassium and sodium hydroxides, potassium permanganate, 

sodium metabisulphite, slaked lime water and slurried calcium carbonate and 

formaldehyde, have been used, as has hydrogen peroxide. `Plug rinsing' grain in the 

steep by washing downwards with a layer of fresh water, (with or without hydrogen 

peroxide or other substances), as the steep is drained is an economical possibility for 

removing suspended microbes, their nutrients and other substances (Briggs, 2002). 

To control the microbes which produce mycotoxins and gushing-promoting agents it 

has been proposed that they should be swamped with `harmless' microbes which will 

outgrow the problem-causing species. Species investigated include lactobacilli and 

strains of Geotrichum yeast (Boivin and Malanda,1998; Haikara et al., 1993; Laitila et 

al., 1999). The results appear promising, but these microbes will also compete with the 

grain for oxygen. Their use might be combined with a washing procedure (Briggs, 2002). 

Air rests are used between steeps. After a steep has been drained air, which should be 

humid and at the correct temperature, is sucked down through the grain. Such downward 

ventilation, or `CO2 extraction', assists drainage, provides the grain with oxygen, 

removes the growth-inhibiting carbon dioxide and removes some of the heat generated by 

the metabolizing grain. In consequence, and in contrast to traditional practice, barley 

leaving the steep has usually started to germinate. When the grain is immersed it is partly 

anaerobic, and it ferments, forming carbon dioxide and alcohol (ethanol), a proportion of 

which enters the steep liquor. Under such conditions the grain will not germinate. Under 

aerobic conditions fermentation is repressed and germination can occur. During 

immersions air may be blown into the base of a steep, providing some oxygen and 

lifting and mixing the grain. 

The onset of germination is indicated by the appearance of the small, white `chit', the 

root sheath (coleorhiza) that protrudes from the base of each germinated grain. At this 


stage the grain is transferred to a germination vessel (or floor in older maltings) or, if it is 

in a steeping/germination vessel, the equipment will be set into the germination mode. 

The grain grows, producing a tuft of rootlets (culms) at the base of the grain and, less 

obviously, the coleoptile or àcrospire' grows along the dorsal side of the grain, beneath 

the husk. The extent of acrospire growth, expressed as a proportion of the length of the 

grain, is used as an approximate guide to the advance of the malting process. Variations 

in acrospire lengths indicate heterogeneity in growth. The living tissues respire and 

carbon dioxide and water are generated resulting in a loss of dry matter. The energy 

liberated supports growth and is liberated as heat. 

Many hydrolytic enzymes, which are needed when malt is mashed, appear or increase in 

amount. Some of these catalyse the physical modification of the starchy endosperm. In the 

initial stages of germination these hydrolases are released from the scutellum. However, 

after a short lag the embryo releases gibberellin hormones (GA 1 and GA 3, gibberellic acid). 

These diffuse along the grain triggering the formation of some enzymes in the aleurone layer 

and the release of these and other enzymes into the starchy endosperm. Here they join the 

enzymes from the embryo in catalysing modification. As germination progresses the starchy 

endosperm softens and becomes more easily `rubbed out' between finger and thumb. When 

the malt has been dried the modified material is easily crushed and `friable', and is easily 

roller-milled, in contrast to the tough barley. The stages of physical modification are the 

progressive degradation of the cell walls of the starchy endosperm, which involves the 

breakdown of the troublesome -glucans and pentosans, followed by the partial degradation 

of the protein within the cells and the partial or locally complete breakdown of some of the 

starch granules, the small granules being attacked preferentially. The extent of breakdown is 

limited by the availability of water. 

Modification begins beneath the entire `face' of the scutellum. In a proportion of 

grains it advances more rapidly on the ventral side of the endosperm, adjacent to the sheaf 

cells, while in others it advances roughly parallel to the face of the scutellum (Briggs, 

1998). When enzymes from the aleurone layer have been produced modification 

progresses more rapidly, particularly adjacent to the aleurone layer. Thus modification 

begins adjacent to the embryo and advances towards the apex as germination proceeds. In 

well-made malt only a small proportion of grains are undermodified, and contain large 

amounts of undegraded (unmodified) starchy endosperm tissue. The products of 

endosperm breakdown, sugars, amino acids, etc., together with materials from the 

aleurone layer (phosphate, metal ions, etc.), diffuse through the endosperm and a 

proportion support the metabolism of the living tissues, while the remainder accumulates. 

The growth of the embryo is at first supported by its own reserve substances and later 

by soluble materials from the modifying starchy endosperm, so there is a net migration of 

materials into the embryo. The levels of soluble materials that accumulate are regulated 

by the balance between their rates of formation in the endosperm and their rates of 

utilization by the embryo. In the finished malt these materials may be estimated as the 

cold water extract, CWE, or the `pre-formed solubles'. The accumulation, with time, of 

enzymes and the physical modification of the grain, permit the increasingly greater 

recovery of hot water extract up to a maximum value. When the acrospires have grown to 

about 3/4 to 7/8 the length of the grain the hot water extract, the cold water extract and 

the level of soluble nitrogenous substances cease to increase with increasing germination 

time, and the fine-coarse extract difference has almost stopped decreasing although 

friability is still increasing and the viscosity of grain extracts may still be declining. 

Enzyme levels may or may not be increasing, depending on the malting conditions. 

Usually germination is terminated at this stage by kilning. Longer germination periods 


waste malthouse capacity and result in extra malting losses. A correct dose of gibberellic 

acid, GA3, which, when used, is usually applied at the end of steeping, accelerates the 

growth of the embryo but stimulates most of the processes of modification relatively 

more so that malt can be prepared more rapidly and in better yield. Where permitted the 

excessive accumulation of soluble nitrogenous substances that can occur in GA 3-treated 

grain may be limited by the application of sodium or potassium bromate, which inhibits 

the activity of some proteolytic enzymes. By checking the growth of rootlets this agent 

also increases malt yields. 

The processes that occur in germination are regulated by controlling the moisture 

content of the grain, the quality of the grain and the temperature programme of the grain 

during steeping and the germination period. Nearly all malt is made using `pneumatic' 

malting plant in which the grain is ventilated with a stream of humidified and 

temperature-adjusted air to remove excess heat and carbon dioxide and to supply oxygen. 

In the UK, limited amounts of high-quality malts are still made by traditional floor 

malting. From time to time the piece (batch) will be turned or stirred to separate the 

grains by untangling the rootlets and allowing the easier passage of the conditioning 

airflow. 

Malting losses can be defined in several ways. If they are defined in terms of the losses 

in dry weight, which occur when cleaned barley entering the steep is recovered as kilned 

malt and has been de-culmed (dressed), then the losses sustained in making conventional 

malts are usually in the ranges: steeping losses, 0.5 1.5%; germination losses, 

3.5 7.5%; rootlets, 2.5 5.0%. These divisions are artificial, since some respiration 

and growth occur in the steeping phase and in the initial stages of kilning. Rootlets are 

sold, usually for use in animal feeds, but the cash value is less than that of an equal 

weight of malt. Malting losses are larger when coloured malts are being produced. 

Modification of the green malt is likely to be more extensive if it is to be used in making 

darker malts. It will probably have been germinated with a relatively high moisture 

content and will be rich in soluble sugars and soluble nitrogenous compounds that will 

react during kilning to generate melanoidins, and so generates colours and characteristic 

flavours and aromas. 

Most modern kilns hold a bed of grain about one metre deep, (which is not turned 

during kilning), through which a current of air is fan-driven from below. This air is heated 

either directly, using low-NOx burners fuelled with oil or gas (which generate little or no 

oxides of nitrogen), or indirectly by heat exchangers. Oxides of nitrogen are avoided to 

prevent the formation of potentially harmful nitrosamines. When making pale malts the 

airflow is rapid and the àir-on' temperature is low during the initial, drying phase. As the 

air rises through the bed of malt it becomes saturated with moisture and it is cooled by the 

need to provide energy to evaporate the water. So above the drying zone the air is 

saturated with moisture and the grain can continue to grow, generate enzymes and modify 

while being warm. In the initial stages the air-on temperature may be about 50 ëC 

(122 ëF), and the air-off temperature is around 25 ëC (77 ëF), which will be the 

temperature of all the green malt above the top of the drying zone. To economize with 

fuel this air should be passed through a heat exchanger to pre-heat incoming air. As the 

outgoing air is cooled moisture condenses on the tubes of the heat exchanger, liberating 

its heat of condensation, which is passed to the incoming air. 

With the passage of time the drying zone extends upwards through the bed of malt 

until it reaches the surface of the grain bed. At this time, at the `break point', the relative 

humidity of the àir-off' falls and the temperature rises. When this occurs the airflow is 

reduced and the air-on temperature is increased to begin the curing (cooking) stage. As 


the malt dries the temperature is progressively increased to the maximum, `curing' 

temperature. As the fuel used in kilning is costly, the procedure is adjusted to save heat. 

During curing a progressively higher proportion of the air may be re-circulated. 

Alternatively the hot air may be diverted to asecond, `linked' kiln in which the malt is in 

the drying stage. Here it is mixed with more heated air to provide the large volume of air 

required during drying. Many pale malts are cured at about 80ëC (176ëF), but some will 

be `finished' at higher temperatures, up to 105ëC (221ëF). 

While enzyme destruction occurs at these elevated temperatures some enzymes 

survive provided that the malt has first been dried at low temperatures to alow moisture 

content. Under these conditions colour formation is minimized. In the manufacture of 

some coloured malts the temperature is increased while the grain is still comparatively 

wet to promote the formation of free sugars and amino acids and the interaction of these 

and other substances form the coloured melanoidins and flavoursome and aromatic 

substances. In these malts enzyme levels are comparatively low and, in extreme cases, 

enzyme destruction is complete. 

Most special malts are now `finished' in roasting drums. These metal cylinders may 

have capacities ranging from 0.5tto 10t. As they turn the contents are mixed by internal 

vanes and may be heated indirectly, by heating the outside of the drum, or directly, when 

hot air is passed through the interior of the cylinder, drying the contents. Depending on 

the malt being made adrum is loaded either with pale, kilned malt or green malt, and the 

processing is varied. The cylinder is heated while rotating and the contents are subjected 

to acarefully chosen temperature regime. The colour and physical state of the grain is 

frequently checked on samples. At exactly the correct time heating is stopped, in some 

instances water is sprayed into the cylinder, and the malt is withdrawn and cooled. Green 

malt is usually used in making crystal or caramel malts (which are not all dark in colour) 

while pale, kilned malts are used in making other types which vary in colour from amber 

to black (Briggs, 1998 and Sections 2.2.5, 2.2.6). 

After kilning malts are dressed (de-culmed or de-rooted and cleaned). The cooled malt 

is agitated to break up the brittle rootlets and these, and dust, are separated by sieving and 

aspiration with air currents. Pale malts are usually stored for 4 6 weeks before use when, 

for unknown reasons, the brewing values often improve. Coloured and special malts 

should be brewed with as soon as possible, because during storage their special aromas 

(and perhaps flavours) decline. Malts are stored in ways intended to minimize the pickup 

of moisture, and to exclude birds, rats, mice and insects. It is important to prevent malts 

being mixed or being contaminated with un-malted barley during handling or storage. It 

is impossible to make successive batches of malt that have precisely the same analysis. 

Each batch should be stored separately and different batches should be blended so that the 

mixture meets the brewer's requirements. Different batches of malt made from one 

variety of barley, in the same way and intended to meet the same specification can be 

safely blended. Brewers take different views regarding what other blending is 

permissible. Specifications may stipulate that no other blending should occur, while 

others will accept blends of any varieties (even involving malts from two- and six-rowed 

barleys) provided that the analyses of the mixture are as stipulated. 

Before dispatch the malt will be cleaned by screening, aspiration, passage over 

magnetic separators to remove fragments of iron, and (often) through a gravity separator/ 

de-stoner. Usually malt is delivered in bulk, (often 25 t batches), but for export some 

batches will be packed in very large sacks (1 t capacity) which will be transported in 

containers. Some maltings will provide smaller breweries with malt in smaller sacks that 

are made with several layers of different materials and are strong and waterproof. 


Rootlets are usually sold for cattle feed. However, they have been used to provide 

nutrients for microbial cultures and in making composts for growing mushrooms. Other 

markets are being sought. Because of their low bulk density, inconvenience in handling 

and a strong tendency to pick up moisture, rootlets are now usually pelletized, together 

with malt and grain dust and sometimes with thin grains. Rootlets (culms, coombes, 

cummins, malt sprouts) vary in their nature depending on what malt they came from and 

in particular how strongly they were kilned (Briggs, 1978,). Commonly analyses are in 

the ranges: non-protein extract, 35 50%; crude protein, 20 35%; ash, 6 8% and fibre, 

9 15%. They are rich in low molecular weight nitrogenous substances and B vitamins. 

2.2.3 Malting technology 

Many types of malting plant are in use, but only the most common types will be 

described. Malting is comparatively safe, provided that certain precautions are observed. 

Some, sometimes unfamiliar, risks are due to carbon dioxide and to grain and malt dust. 

As grain is steeped and germinated it liberates carbon dioxide. This heavy gas can `pool', 

so it is essential to check that vessels and confined spaces are ventilated before they are 

entered. Dust must be confined and cleaned away not only because it becomes damp and 

a breeding ground for insects and microbes, but also because when it is breathed it can 

cause allergies and fungal lung infections and it can form explosive mixtures when mixed 

with air. All handling equipment must be earthed (grounded) to prevent sparks, which 

might trigger an explosion, and all conveyors, ducts, etc., should have explosion vents. 

Modern malt factories process large batches of grain, (often 200 300 t batches), and 

the process stages are highly automated so that processing conditions are reproducible 

and the manpower needed to produce each tonne of malt is minimized. In large maltings 

grain is delivered in bulk, by ship or barge or train or lorry. Before unloading begins the 

bulk should be inspected and sampled for analysis. When the quality has been agreed 

unloading begins. Grain is usually sucked from the holds of vessels, and this pneumatic 

system may be used to empty rail wagons or lorries, or these may be emptied under 

gravity. Lorries usually unload into an intake pit by tipping or from a hopper. The grain 

runs into the pit, which is ventilated to remove dust, and is equipped with a coarse 

moving screen (sieve) to catch and remove coarse impurities such as straw and large 

stones. Each lorry is weighed on to the site and off after unloading. The difference in 

weights gives the amount of grain unloaded. 

During malting grain will be moved several times. The equipment used varies, but will 

usually include bucket elevators, helical screw conveyors (worms), belt conveyors and 

chain and flight conveyors. Less usually the grain will be moved using pneumatic 

conveyors. It is highly desirable that, to avoid cross-contamination, the equipment used to 

move grain is entirely separate from that used to move malt. The freshly delivered barley 

is conveyed to a `green grain' bin for temporary storage. Here it will remain, usually 

being ventilated with fresh air, until it can be precleaned and, if necessary, dried. 

Precleaning involves rapid screening to remove gross impurities, such as sand, straw, 

stones and string, which are either appreciably larger or smaller than the grains, and 

aspiration with air to remove dust. The dust from this and other locations is trapped in 

cyclones and textile-sleeve filters. The grain also passes over magnetic separators, which 

retain iron and steel impurities. 

In Northern Europe the grain usually needs to be dried (to 12% moisture, or less) 

before it can be safely stored. Drying and pre-cleaning may be carried out before the 

grain is delivered but, because of the risk of heat damage caused by inexpert drying, some 


maltsters do not allow this. The drying temperatures used are lower for more moist grain, 

because wetter grain is more easily damaged by heat. Batch drying can be carried out in 

malt kilns or in steeping, germination and kilning units (vessels; SGKVs) or in dedicated 

batch driers. In these the grain rests on a perforated floor or deck and warm air is passed 

through it, e.g., for eight hours, until the grain has been dried sufficiently. The grain then 

may or may not be cooled, depending whether it is to be committed to long-term storage 

or it is to be stored warm for a short period to overcome dormancy (i.e. to hasten postharvest 

maturation). In flow-through dryers the grain passes downwards under gravity in 

a stream that is regulated by valves. The grain passes through a series of zones in which it 

meets air at different temperatures and is successively warmed, dried and cooled. If there 

is to be a period of warm storage the cooling may be limited or omitted, so that the grain 

reaching store is at 30 40 ëC (86 104 ëF), rather than 15 ëC (59 ëF) or less, which is 

desirable for long-term storage. 

The dried grain may now be thoroughly cleaned either immediately or after warm 

storage. This process is less rushed than pre-cleaning and so is more thorough. The grain 

is screened to remove thin corns and sometimes it is graded into size classes (e.g., above 

and below 2.5 mm width), which are malted separately. The screens used may be flat and 

oscillate horizontally or they may be rotating cylinders. At present the quality of the grain 

on delivery in the UK is so good that apart from aspiration, screening and passage over 

magnetic separators, this is all the cleaning required. However, with less clean samples it 

may be necessary to remove light impurities with air classification and foreign seeds and 

broken grains with Trieur cylinders or Carter-Simon disc separators (Briggs, 1998). The 

clean barley may be stored in flat-bed stores, bins or silos. If storage is to be for an 

extended period then the grain can be treated with an approved insecticide. If the grain is 

held relatively moist (> 12%) it will have to be ventilated. At a 12% moisture content 

grain can be stored for some months at or below 15 ëC, but for periods over about six 

months a moisture content of 10% is safer. Stores must be regularly inspected for signs of 

insect infestation and fungal attack and depredations by birds or rodents. The temperature 

of the grain, determined by probes positioned at various sites and depths, should be 

recorded weekly and any undue increase acted on as a sign of deterioration. 

Grain is weighed on its way to the steep(s). If abrasion (limited physical battering or 

rubbing the grains together) is to be employed this is carried out in advance of steeping as 

grain can be treated at rates of only 10 12 t/h and malting batch sizes are often as high as 

300 t, and so this amount of treated grain must be accumulated before steeping can begin. 

Historically, steeps were barrels or shallow troughs in which grain rested, under water, at 

depths of 1 2 ft. (0.31 0.62 m). Numerous patterns of steeping vessels have been used. 

Those preferred now are either flat-bed or conical-bottomed steeps. Flat-bed steeps are 

circular in plan view, and the grain is supported on a perforated deck above the true base, so 

there is a plenum beneath the deck. For a 200 t batch size the steep might have a diameter of 

15 m (49.2 ft.; Gibson, 1989). Initial depths (before the grain swells) may be 1.5 1.8 m 

(approx. 4.9 5.9 ft.). Grain is loaded in from above, dropping through sprays of water that 

quench the dust, falling into water. The bed is levelled with a rotating spreader, called a 

giracleur. Air may be blown in beneath the deck while the grain is immersed, and in dry 

periods air may be sucked down through the grain. The steeped grain is discharged through 

ports impelled by the giracleur. Such steeps allow relatively even grain treatment, since the 

bed depth is comparatively shallow, but the water used to fill the plenum is `waste' and so 

effluent volumes are large. In addition it is difficult to keep the plenum chamber clean. 

Newer maltings usually employ various types of conical-bottomed steeps. As each 

steep should contain less than 50 t, to avoid deep cones with excessively high pressures 


on the grain in the cone base, aset of steeps is required for each batch of grain. Typically 

each steep consists of avertical cylinder, closed below with acone. Grain is loaded into 

each steep from above. The base of the cone contains avalve that retains the grain and 

water, aperforated zone through which the water can be drained while the grain is 

retained, an outlet through which air can be drawn during CO 2extraction during air rests 

and inlets through which compressed air can be supplied when the grain is being aerated 

whenunderwater.Suchsteepsare`self-emptying'.Theconeangleissufficientlyacuteto 

ensure that when the valve is opened the grain falls out in to aconveyor. This is one way 

of `dry-casting' the grain. An alternative method is to `wet-cast' the grain, pumping it to 

the germination compartment slurried in water. In steep-germination and steepgermination-kilning 

vessels (SGVs and SGKVs) no transfer is required. If additives, 

such as gibberellic acid and/or sodium bromate are to be added it is convenient and 

economic to spray on solutions as the grain is conveyed from the steep. 

In floor malting the steeped grain is spread on afloor in aroom having acool, humid 

atmosphere. Germination is controlled by turning the `piece' (batch) and thickening or 

thinning the layer of grain to allow temperature rises or falls as needed. Fine malts can be 

made in this way, but only in small quantities (ca. 10t/batch) and with substantial 

manpower. Modern maltings are of the pneumatic type, in which the grain is turned 

mechanically and the grain temperature is controlled by forcing astream of attemperated 

and water-saturated air through abed of grain. Newer germination vessels are usually 

rectangular `Saladin boxes' or circular compartments. In these vessels steeped grain is 

formed into abed, usually 0.6 1.0m(approx. 2.0 3.3ft.) deep. The grain rests on a 

perforated deck, through which the conditioning airflow is driven. Some of the air is 

recirculatedandmixedwithfreshair.Theairisdrivenbyafanandisusuallyhumidifiedby 

passagethroughspraysofwater.Airtemperaturemaybecontrolled,byregulatingthewater 

temperature, sometimes augmented with heating or cooling by heat exchangers. The grain 

lifted and partly mixed, and the rootlets are separated by passing arow of vertical, contrarotating 

helical screws through the bed. The bed is `lightened' and the resistance to the 

airflow is reduced. Bed temperatures of 15 19ëC (59 66.2ëF) are common, with 

temperaturedifferentialsbetweenthetopandbottomofthebedof2 3ëC(3.6 5.4ëF).The 

turner arrays are usually fitted with sprays to allow the grain to be moistened. 

In some plants the grain is first germinated in acircular, stainless-steel lined vessel, 

then it is transferred to a germination and kilning unit (or vessel; GKV). When 

germination is sufficiently advanced the cool airflow, which may or may not be 

humidified, is discontinued and hot air is supplied from afurnace or heat exchanger. In 

old kilns the malt was dried in thin layers and with periodic turning. In modern kilns the 

grain beds are relatively deep and are not turned. The kilns may be directly or indirectly 

heated. They are instrumented so that correct temperature differentials are maintained 

between the air-on and the air-off and that at the break point the airflow is reduced and 

subsequent air re-circulation, temperatures and flow rates are correct. As noted before, 

heat should be conserved by `linking' kilns and/or using heat exchangers. Further heat 

can be recovered from the outgoing air by using aheat pump, but at present this is not 

economic because the capital and maintenance costs are high. 

2.2.4 Malt analyses 

Malt analyses are carried out according to one of the several sets of agreed methods 

(Section 1.15.1, p. 9). As with barley, analysis of amalt lot is useless unless the samples 

used are properly drawn and handled. Because of differences between the methods, both 


in the conditions and calculations used, the results obtained often differ, both in the 

values obtained and the ways in which the results are expressed. Some of the most 

important methods will be discussed. Others are considered elsewhere (Briggs, 1987; 

1998). The moisture contents of malts are usually in the range 1.5 6%, expressed on a 

fresh weight (fr. wt.; as is) basis. Primary analyses are by oven-drying methods, but NIR 

(near infra-red) analysis is the usual, more rapid, secondary method. Brewers normally 

specify an upper moisture limit. Because malt is hygroscopic it will normally have a 

lower value when dispatched, to allow for moisture uptake while in transit. Brewers use 

malt às is', and so they pay attention to the extract of the undried malt. However, for 

comparative purposes extracts are mostly given on a dry weight basis (on dry). A malt 

sample must not contain more than a certain percentage of thin corns, because thin corns 

are not broken up in mills with rollers set relatively far apart to achieve a coarse grist. 

When malt is hammer milled this consideration does not apply. Cold water extracts 

(CWE) are used by some ale brewers. The value of this determination is disputed. The 

grain is ground and extracted at 20 ëC (68 ëF) with water made alkaline with ammonia (to 

inactivate enzymes). The specific gravity of the extract is a measure of the `preformed 

soluble substances' present in the malt. Departure from customary values warns that the 

malt lot is different from its predecessors. 

The hot water extract or extract (HWE or E) value is the single most important 

measurement in judging malt quality. The HWE method used by the IoB was designed 

for use by traditional ale brewers, and involves an isothermal laboratory mash (65 ëC; 

149 ëF) made with distilled water and a comparatively coarsely ground grist. After one 

hour the mash is cooled and adjusted to either a volume of 515 ml or to a weight of 450 g 

and the specific gravity of the liquid, obtained by filtration, is measured at 20 ëC (68 ëF). 

Using the appropriate formula the extract is calculated from the excess specific gravity 

(i.e., the gravity 1000 above water, taken as 1000.00 at 20 ëC (68 ëF)) as litre-degrees/ 

kg malt (l ë/kg). Sometimes the HWE of a finely ground grist is also determined. Extract 

is obtained in greater yields from finely ground malt and the smaller the fine-coarse (f.-c.) 

extract difference the better the malt is modified. Because this value is the difference 

between two large numbers and is small relative to the errors involved in measuring the 

extracts, the determination must be replicated to obtain a reliable value, which is 

laborious. Another `non-standard' proposal is to use the difference in extract yield from a 

fine grind mash and a concentrated, very coarse grind mash, This `f.-c. conc.-extract 

difference' method gives larger differences than the f.-c. grind method and appears to be 

an improvement on it, the values obtained being inversely related to extract recoveries in 

a brewery (Bourne and Wheeler, 1982; Briggs, 1998). 

The determination of extract, E, by the EBC and the very similar ASBC methods 

differs considerably from the IoB method. They were developed for traditional lager 

brewers but the temperature programme used does not resemble that of most old lager 

breweries or that of breweries which employ temperature programmed mashing. In the 

EBC method finely ground malt is mashed in at a low temperature (45 ëC; 113 ëF), with 

continuous stirring. The temperature is then increased, at 1 ëC (1.8 ëF)/min., until it 

reaches 70 ëC (158 ëF). This temperature is now maintained and more water, also at 70 ëC, 

is added. After one hour, during which the `saccharification time' is determined (see page 

24), the mash is cooled and adjusted to 450 g. The specific gravity of the wort is 

determined. Using tables that relate the strengths of sucrose solutions with their specific 

gravities, the weight of extract in the laboratory wort is calculated, assuming that the 

dissolved extract solids change the specific gravity to the same extent as sucrose. The 

EBC method uses Plato's tables while the ASBC method uses Balling's tables 


(Appendix). The extract of the malt is expressed as apercentage (%). This calculation 

makes some unwarranted assumptions, so the values are unreliable in absolute terms, but 

it gives useful relative values. There are no conversion factors that allow the calculation 

of extracts determined by one method to be accurately expressed as the extracts 

determinedbyanothermethod,eventhoughtheresultsareroughlycorrelated. Extracts of 

pale malts determined by the EBC method are usually in the range 77 83%, (on dry), 

whilevaluesfortheIoBmethodareusually300 310lë/kg(ondry).Similarlythetypical 

ranges for darker, malts are 75 78% and 255 285 lë/kg respectively. Sorghum malts, 

which are used to make pale lager-style beers in tropical Africa as well as opaque 

African-style beers, are not reliably analysed by the extract determination methods 

developed for barley malts, because the gelatinization temperature of sorghum starch is 

higher than that of barley (see Table 2.3 on page 38). 

Various special, but apparently unstandardized, analytical mashing programmes are in 

use (Briggs, 1998). The laboratory mashes differ from brewery mashes in anumber of 

important ways. Unlike brewery mashing liquor the water used is distilled and contains 

no salts, nor is the mash pH adjusted. Also the grist is prepared by using mills that work 

differently from brewery mills. The laboratory mashes are dilute compared to brewery 

mashes andattheendofmashingthegristisnot spargedwithhotwater.Severalattempts 

have been made to devise more `brewery-like' laboratory mashes, but they have not been 

accepted. Each brewer discovers the relationship between amalt's `lab. extract' and the 

extract recovered from this malt in the brewery. The extract determinations described 

apply to pale malts. Different methods are necessary for special, highly coloured malts 

that lack enzymes. For example a50:50 mix of acoloured malt with an enzyme-rich pale 

malt may be mashed and the extract of the coloured malt is calculated, making the 

assumption that the pale malt gives half of the extract it yields when mashed alone. 

More information is gained from analysing the mash and laboratory worts. The rate of 

wort filtration from alaboratory mash does not give agood indication of the brewery 

wort run off. `Mashing columns' are needed to achieve this, and these devices are not 

suitable for routine analyses (e.g., Webster, 1981). Wort colours are determined in 

different ways, either visually with colour comparators, or at asingle wavelength, or at 

three different wavelengths (the tri-stimulus method, Chapter 19) using a spectrophotometer. 

All these approaches have limitations since extracts from different malts not 

only differ in colour intensity but also in their spectral characteristics. Because worts 

darken to various extents during the hop-boil it is sometimes desirable to measure the 

colours of boiled laboratory worts. One problem with the IoB methods is that the wort 

from the 450 g mash is more concentrated, and therefore more deeply coloured, than that 

from the 515 ml mash, and so the mashing method employed must be stated when results 

are given. 

The pH of wort is often routinely recorded. The values obtained vary with the type of 

malt. Unusually acid wort (low pH) can be caused by a heavy infestation of microbes on 

the malt (Stars et al., 1993). 

Malts are analysed for their nitrogen (`protein') contents and the laboratory worts are 

also analysed for dissolved nitrogenous materials. The values are expressed as nitrogen, 

N, in the IoB methods and as `protein' (crude protein ˆ 6.25 N) in the EBC and ASBC 

methods. Because of the differences in the mashing conditions the last two mashes 

generally give more soluble nitrogen in the worts than the IoB method. Brewers are 

concerned that a malt has a total nitrogen (TN; `protein') content in the specified range 

since, outside this range, the wort obtained may differ significantly in quality and there 

may be problems with extract recovery. 


Total soluble nitrogen (TSN) and free amino nitrogen (FAN) values are determined. 

The TSN needs to be sufficiently high so that the `body' and mouth-feel of the beer is 

adequate, and the beer foam (or `head') will be stable. The soluble nitrogen ratio (SNR; 

TSN/TN) of the malt (or the soluble protein ratio or Kolbach Index of the ASBC and 

EBC methods (in each case soluble protein/total protein) serve as measures of 

modification and avalue is often included in malt specifications. FAN values (chiefly 

amino acids and small peptides) must be sufficiently high to ensure that lack of 

nitrogenous yeast nutrients does not limit fermentation. FAN has been determined by 

different methods, which gave different results, so it is essential that the method used is 

specified. 

Previously,nitrogenousyeastnutrientswereassayedas`formol-nitrogen'.Overtheyears 

thepreferrednitrogencontentsforBritishpalealemaltshaverisenfromaround1.3 1.45% 

toaround1.65%.Perhapsthisisdueinparttonewervarietiesofbarleyandchangedbrewing 

practices. For enzyme-rich malts, needed with high-adjunct brews as in some North 

American breweries, TN values of 2.2% (13.8% protein) or more may be preferred. In the 

past,othermeasuresweremadesuchaspermanentlysolublenitrogen(PSN)andcoagulable 

nitrogen, respectively the nitrogen in the substances remaining in the wort and those 

precipitated when the wort was boiled. This method has fallen out of use. 

In EBC analysis the time in minutes taken after the mash has reached 70ëC (158ëF) 

for samples to stop giving a positive iodine test for starch is recorded as the 

`saccharification time'. This is really arough measure of the time taken for the starch to 

be dextrinized, and is largely dependent on the -amylase content of the malt. The odour 

of the mash is noted as, less usually, is the flavour. Both should be normal for the type of 

malt being analysed. The appearance of the wort is noted, whether it is clear, opalescent 

orturbid.Theactivityofthemixtureofthestarch-degradingenzymesinmaltisestimated 

as the `diastatic power', or DP. The enzymes are collectively referred to as diastase. In 

principle, soluble starch is incubated with a malt extract and the degree of starch 

breakdown is estimated after aperiod of incubation at acontrolled temperature. The 

results are not highly reproducible and represent the joint activities of several enzymes 

that are present in different proportions in different malts. Results are expressed in 

different units including ëL (degrees Lintner) and ëW-K (Windisch-Kolbach units). The 

values indicate to brewers if the enzyme content of amalt is adequate. 

The level of -amylase in malt extracts is determined by one of several methods. The 

level of activity of this enzyme must be adequate if the starch from adjuncts is to be 

liquefied in a mash. The ratio of fermentable to non-fermentable sugars is largely 

regulated by the activities of the diastatic enzymes during mashing. The fermentabilities 

of worts should be constant when brewing a particular beer. Analytically the 

fermentabilities of the HWE or Eworts may be determined. However, the fermentability 

of these worts increases with storage time as malt enzymes that have survived the 

mashing process continue to break down dextrins to simpler, fermentable sugars. Thus 

laboratory wort should be boiled to inactivate the enzymes (as occurs in the hop-boil). 

Thenitisinoculatedwithapureyeastanditisincubatedunderanaerobicconditionsuntil 

fermentation is complete. The fall in the specific gravity of the wort allows the 

calculation of the attenuation limit of the wort and its fermentability (Chapter 4). 

When malts contain substantial levels of -glucans modification is incomplete and 

the polysaccharide itself may cause problems in the brewhouse. The -glucans may be 

assayed using methods based on enzymes degrading them to glucose, which is measured, 

or by the fluorescence of the complex between the polysacchride and the reagent 

Calcofluor. The activity of the enzymes in malt that degrade -glucans, the - 


glucanases, are measured either by following the decline in viscosity of a solution of - 

glucan incubated with an extract of malt containing the enzyme, or by following the 

breakdown of an artificial substrate, a colour-labelled -glucan. Under some 

circumstances it is desirable that malts should contain adequate levels of this enzyme, 

which is readily inactivated when green malt is kilned using any except the lowest 

temperatures. Other substances that may be estimated include the dimethyl sulphide 

precursor (DMS-P), which is S-methyl methionine (SMM), and N-nitrosodimethylamine 

(NDMA). Different styles of beer require different amounts of dimethyl sulphide in the 

final product. As the precursor can be destroyed during kilning it is important that its 

levels are regulated. NDMA, and other less volatile nitrosamines, are suspected of being 

carcinogens. When discovered, high levels of NDMA were found in malts, particularly 

those from directly fired kilns. However the precautions now taken ensure that the 

amounts present are usually below the levels of detection. The levels of NDMA are still 

monitored. 

The growth of acrospires roughly parallels the advance of modification in malting 

grains. The evaluation of acrospire growth in grains in a sample of malt can indicate that 

a malt has been made with irregularly germinated material or that good and less good 

malts have been mixed. In North American practice the acrospire lengths of corns in a 

sample of malt are classified, by length relative to corn length, as 0 0.25, 0.25 0.50, 

0.50 0.75, 0.75 1.0 and over 1. Grains in which the acrospire has grown out from 

beneath the husk, i.e., over 1.0 in length (so-called overgrown corns, huzzars, cockspurs 

or bolters) are undesirable in most kinds of malt as they are deficient in extract. 

However, overgrown malts may be relatively rich in enzymes. Specifications may call 

for 86 95% of acrospires to fall in the 0.75 1.0 category and less than 5% to be 

overgrown (1+). 

The physical modification of malt grains is traditionally assessed by crushing a series 

of corns between finger and thumb. Well-modified grains crush to a powder while the 

presence of `hard ends' indicates that the apices are not modified and completely hard 

grains are unmodified. A number of other methods have been used. In the traditional 

`sinker test' a handful of malt corns is thrown into water. Barley corns sink, fully 

modified malt corns float horizontally and partly modified corns float with the apical 

ends downwards. This test is unreliable. The resistance of malt corns to grinding has been 

measured, as has the resistance of corns to cutting or penetration by blunt needles. 

A device that has achieved wide acceptance is the Friabilimeter. In this a 50 g sample 

of malt is broken up between a rotating wire sieve and a spring-loaded roller. The friable 

material and the husk fragments escape through the sieve, and the material remaining 

after a set period is weighed. The friability is the percentage (by weight) of material that 

passes through the sieve. Investigation of material remaining on the sieve can be 

informative and can indicate if the malt corns generally contain unmodified material or if 

a substantial proportion of wholly unmodified grains is present. From this an estimate of 

the homogeneity of the malt can be made. Other approaches give indications of the 

patterns of modification that have occurred. Samples of malt are stuck to a flat support 

and a proportion of the grains is ground away with a mechanical sander. In one method 

the exposed grain interiors are treated to suppress autofluorescence and then they are 

treated with Calcofluor. Under UV light this fluoresces when associated with the - 

glucans in the unmodified regions of the endosperm cell walls but there is no 

fluorescence in the modified regions. Thus the percentage area modified in each exposed 

area of endosperm can be assessed, preferably with a specialized automatic scanner. In 

the other method about 0.25% of the grains is sanded away and then they are exposed to 


an alcoholic solution of the dye methylene blue. This penetrates only into the modified 

regions of the grain. The stained grains are dried and are sanded further. The blue regions 

of the endosperm are modified and the white regions are not. Again, the relative areas 

modified can be estimated. This data allows the degree of modification and its 

heterogeneity to be estimated. 

The values for `nitrogen modification' (SNR; Kolbach Index) do not always parallel 

the estimates of physical modification, and indeed the relationships between the two can 

differ to an important extent when different varieties of barley are malted in parallel. 

Brewers do not want undermodified or overmodified malt. With undermodified malts 

extract recoveries in the brewery are unduly low, wort separation can be slow, the worts 

obtained may be cloudy, the hot break may form poorly, the wort may have a low 

fermentability and ferment slowly, the beer may be difficult to filter and the filters may 

become blocked quickly causing high pressures to build up and giving short filtration 

runs. Finally the beers may quickly become hazy. In extreme cases -glucan gels may 

form and deposit. On the other hand overmodified malts have their disadvantages. Malt 

breakage and losses (as dust) are high and wort separation may be slowed by the large 

proportion of fine particles in the grist. Head retention may be poor, and yeast growth can 

be wastefully excessive. The hot and cold breaks may be heavy, the wort may contain 

finely divided material that is hard to remove by filtration and, because of the excessive 

levels of reducing sugars and amino acids present, the wort may darken too much on 

boiling, due to the formation of melanoidins. 

From time to time other analyses may be performed. Thus levels of arsenic, lead, 

cadmium and iron may be determined to check for the absence of contamination. Zinc 

can be measured but, as there is no clear relationship between this and the amount of zinc 

available to the yeast in the wort, this is rarely done. Levels of microbes, especially 

Fusaria, may be determined and several tests for agents causing gushing have been 

devised (Donhauser et al., 1991; Vaag et al., 1993). Sometimes samples may be analysed 

for halogenated contaminants (such as chlorinated substances), or residues of insecticides 

or agricultural chemicals. Where the use of added gibberellic acid is forbidden residues of 

this substance may be sought on the malt's surface. As this substance occurs naturally 

within the grains the results of such tests must be suspect. 

2.2.5 Types of kilned malt 

Malt types are not clearly distinct. The descriptions given here are representative. 

Different breweries specify distinctly different malts giving them the same title (`pale 

lager', mild, ale, etc.). The question of what constitutes a sensible malt specification is 

discussed later. Extensive sets of malt analyses are available (Briggs, 1998; Narziss, 

1976; 1991). In this chapter malts are described with emphasis on the aspects most of 

interest to brewers. Where the use of adjuncts is forbidden, as by the German 

Reinheitsgebot, chit malts and short grown malts may be used. These are less expensive 

to produce than `normal' malts. They retain some raw grain characteristics and have 

some of the advantages that are gained from using unmalted grains as adjuncts. These 

malts are made by steeping barley to a low moisture content and then, either as soon as 

the grain has chitted or after a short period of germination, the `green malt' is dried at a 

low temperature. The malting losses occurring in making these materials are small and, 

because of their low moisture contents, they are comparatively inexpensive to kiln. The 

products have moisture contents of 2 5% and contain some hydrolytic enzymes but the 

endosperms are incompletely modified. Their Kolbach indices are low and extracts may 


e 77 80% (on dry, EBC) with fine-coarse extract differences of 6 12%. These malts 

provide less expensive extract and better beer foam stability than conventional malts, but 

they enhance wort and beer viscosity, slow wort separation and reduce the rate of beer 

filtration and the length of filter runs. Short grown green malts have been flaked before 

use, without being kilned, which facilitates extract recovery but destroys the enzymes 

originally present. 

Kilning is expensive, so attempts have been made to brew with green, unkilned malt. 

This material is unstable and must be used as soon as it is ready. It is exceptionally rich in 

enzymes and yields highly fermentable, proanthocyanidin-poor wort with a good extract. 

Its high content of hydrolytic enzymes allow it to convert a high proportion of adjuncts in 

mashes (Briggs et al., 1981). This material has rarely been used both because of its 

instability and because it imparts unpleasant flavours in the finished beers. Roots remain 

attached to green malts. A compromise would be to have malts dried at a low temperature 

to 7 8% moisture. Roots can be removed from such material, which can be stored for 

some weeks. It contains high levels of hydrolytic enzymes and, because less intensive 

drying is needed on the kiln, is less expensive to produce than normally kilned malt. Such 

material seems not to give unwanted flavours to beers. Despite these advantages such 

malt is apparently not in use. 

Lager beers are widely produced, and all malts making these beers are, by definition, 

lager malts. In Germany, where most beers are lagers, the palest to the darkest malts are 

lager malts. In the UK lager malts and lager beers are all pale. In North America the 

lager-style beers are usually brewed using high levels of unmalted adjuncts in the mash 

and so the malts used differ significantly from European lager malts. The characteristics 

of European lager malts have changed during the last century, and the differences 

between British pale ale malts and lager malts have become indistinct. North American 

malts may be made from two- or six-row barleys or mixtures of both. In general their 

nitrogen contents are high (TN 1.7 2.3%; 10.6 14.4% protein), with soluble to total 

protein ratios of 43 48% and high levels of FAN, and hydrolytic enzymes (DU values of 

30 45 or even 50 for -amylase), all characteristics needed with adjunct-rich mashes. 

The high nitrogen contents are associated with lower extracts (77 81% on dry) but this is 

of less significance when so much of the extract is derived from adjuncts. Such malts are 

pale (1.4 1.9 ëLovibond) and have moisture contents of 3.7 4.3%. 

The palest of the European products are Pilsen malts (Pilsener Malz). In the past these 

were undermodified but now they are fully modified and are prepared from barleys 

having moderate nitrogen contents. They are kilned at low temperatures to minimize 

colour formation. Typical analyses are E, at least 81% (EBC, on dry), fine-coarse extract 

difference 1 2%; TN, 1.68 (10.5% protein); Kolbach index 38 42%; moisture less than 

4.5%; -amylase 40 DU; DP 240 300 ëW.-K.; saccharification time 10 15 min.; colour, 

2.5 3.4 ëEBC; boiled wort colour, 4.2 6.2 ëEBC; wort pH, 5.9 6.0. Helles (pale; light) 

malts are rather similar, but are made from barleys richer in nitrogen. British lager malts 

are all pale and well modified. Analyses are usually in the ranges: HWE 300 310 lë/kg 

(on dry), TN, 1.55 1.75%; TSN, 0.5 0.7%; SNR, 31 41%; DP, not more than 70 ëIoB; 

moisture less than 4.5%; saccharification time less that 15 minutes. Colour may be 

3.0 ëEBC. Because of the low temperatures used in kilning lager malts (finishing curing at 

e.g., 70 ëC; 158 ëF) are rich in enzymes and so sometimes give slightly higher extracts 

than pale ale malts, which are cured at higher temperatures (finishing at 95 105 ëC; 

203 221 ëF), and have more characteristic flavours but lower enzyme activities. 

In the last 50 60 years the moisture contents of the best pale ale malts have been 

allowed to rise from 1.5% to a maximum of 3% and preferred TN values have increased 


from as little as 1.35% to around 1.65%. Other preferred current analyses are: HWE, 

306 310 lë/kg (on dry); TN from less than 1.55% to less that 1.70%; TSN, 0.5 0.7%; 

SNR, 31 42%; FAN, 0.1 0.12%; CWE, 18 22%, colour 4 6 ëEBC; and DP, 

44 65 ëIoB. The malt will have a high friability. Many ale and lager malts are made 

to meet customers' particular specifications. However, it is more economical for 

maltsters to make larger volumes of `standard' malt having specifications close to those 

of many of their customers and offer these at lower prices. The `standard malts' from 

different suppliers have different specifications and these are likely to change when 

barley quality changes. Mild ale malts are generally made from lower-quality barleys and 

will be slightly less well modified and will be more strongly coloured than pale ale malts. 

A representative mild malt might have a HWE of 305 lë/kg (on dry), and a fine-coarse 

extract difference of about 5 lë/kg; a moisture content of 3.5%; CWE, 18 21%; TN, 

1.55 1.75%; TSN, 0.6 0.7%; SNR, 36 40%. DP, 40 60; colour, 6 9 ëEBC. Thus in 

UK malts the colours increase as one progresses from lager to pale ale to mild. 

In German practice the next type is Viennese malt (Wiener Malz), which is used for 

making `golden' lagers. This is made from normally modified green malt kilned to a final 

temperature of about 90 ëC (194 ëF), giving a colour of 5.5 6.0 ëEBC. Munich malt 

(MuÈnchener Malz) is relatively dark, very well modified and aromatic and is made by 

germinating nitrogen-rich barley, steeped to a high moisture content, so that it is well 

grown (all acrospires at least three-quarters grown) and finishing germination warm, at 

25 ëC (77 ëF). Kilning involves some stewing and curing is finished at 100 105 ëC 

(212 221 ëF), conditions causing appreciable enzyme destruction. This malt has a colour 

of 15 25 ëEBC. The wort is rich in melanoidin precursors and darkens on boiling, e.g., 

from 15 to 25 ëEBC. Other typical analyses are: E, 80%, (on dry); fine-coarse extract 

difference 2 3%; total protein 11.5% (TN, 1.84%); Kolbach index, 38 40%; 

saccharification time, 20 30 min.; wort fermentability, about 75% (compared to wort 

from Pilsen malt of about 81%). -Amylase and DP values are low, at 30 DU and 

140 ëW.-K. respectively. Analyses of a British made, Munich-style malt are: HWE, 300 

lë/kg, (on dry); moisture 4.5%; TN, less than 1.65%, TSN less than 0.65%, colour about 

15 ëEBC and DP at least 30 ëIoB. 

Brumalt (BruÈhmalz) is an even darker German malt, which is made by steeping a 

nitrogen-rich barley (e.g., TN 1.84%; protein 11.5%) and germinating it at an 

exceptionally high moisture content, about 48%. When the grain is well grown it is 

held in a closed container for, say, 36 hours so that the oxygen is used up and carbon 

dioxide accumulates. The temperature rises to 40 50 ëC (104 122 ëF) and the grain 

contents soften and become pulpy and rich in reducing sugars and amino acids. These 

melanoidin precursors interact when the green malt is kilned, at 80 90 ëC (176 194 ëF), 

to give a highly aromatic and melanoidin-rich material with a high colour, usually of 

30 40 ëEBC. Such malt gives characteristic rich flavours to beers and these are said to be 

stabilized by the reductones from the malt. Sometimes this general kind of material is 

called rH malt or melanoidin malt. 

Some malts are made from barleys lacking proanthocyanidins (anthocyanogens; 

Briggs, 1998; Sole, 2000). The absence of the polyphenolic haze precursors means that 

beers made from these malts and using polyphenol-free hop extracts are unlikely to 

become hazy. Other `conventional' malts are in use. Some special beers are brewed using 

a proportion of smoked malt (Rauch Malz) to gain a `smoky' flavour, but these contain 

elevated levels of NDMA. Such flavours were once common, when malts were all kilned 

using direct-fired, wood-burning kilns. Now the smoke from a wood-burning furnace is 

led into the hot stream of air entering the bed of green malt on the kiln. 


Acid- or lactic-malts have been used sporadically in the UK and more frequently in 

Germany to adjust the mash pH. Originally they were used to offset the effects of 

bicarbonate-rich mashing liquor. These malts carry 2 4% of non-volatile lactic acid. 

They are made in several different ways, for example by steeping or spraying green malt 

with solutions of biologically prepared lactic acid during germination and before kilning. 

Application of lactic acid to germinating green malt can check rootlet growth and the rise 

in malting losses and favour the accumulation of soluble nitrogenous substances. In 

another system pale malt is steeped in water at 45 47 ëC (113 116.6 ëF) for an extended 

period. Sugars are leached into solution and thermophilic lactic acid bacteria convert 

much of them to lactic acid. The grain is re-dried, and the acidic steeping water is reused. 

The use of such malts (as 5 10% of the grist) lowers mash pH and, at least with 

some of these malts, the CWE is increased as is the HWE, the TSN and the FAN. A 

typical analysis (IoB) is HWE, over 297 lë/kg (on dry); moisture, 5 6%; TSN, 

0.8 1.2%; colour, 10 25 ëEBC; lactic acid, 2.1 2.5%. Such malts are often used where 

the addition of chemically prepared acids for pH adjustment is forbidden. 

Some brewers prefer to add lactic acid, prepared biologically from wort (using 

Lactobacillus delbruÈckii), for mash pH adjustment. Various other materials have been 

added to malts. For example, added formaldehyde was not readily detectable, but beers 

made from such malt were low in proanthocyanidins and unlikely to form non-biological 

haze. In the Belmalt process green malt was sprayed with a solution of glucose syrup 

(3.5 kg/100 kg original barley) some hours before kilning. The malts were more acid and 

gave higher extracts and levels of soluble nitrogen than controls and the worts had higher 

attenuation limits. Residual glucose would account for the higher extract and 

fermentability and the conversion of some of the glucose to lactic acid by bacteria on 

the grain would cause mash acidification and so increase the TSN. Gum arabic has been 

sprayed onto malts to increase the head retention of beers made from them. Solutions of 

thermostable enzymes, probably including amylase and -glucanase, have been applied 

to malts, presumably to enhance their apparent quality. These enzymes will not penetrate 

into the interiors of grains and so must exert their effects when carried forward into the 

mash. Generally it is better for brewers to keep these kinds of additions under their own 

control and make them at mashing or later in brewing. 

Malts are made from cereals other than barley (Briggs, 1998; Byrne et al., 1993; 

Narziss, 1976; Taylor and Boxall, 1999). Wheat malts are generally pale, although dark 

wheat malts are `made to order'. In mainland Europe wheat malts make up the major 

parts of the grists (up to 80%) of special beers, including the German top-fermented 

Weissbier (white beer) and Weizenbier (wheat beer). In the UK small amounts of wheat 

malt (3 10%) may be included in grists to improve the foam formation and head 

retention of the beers. Other benefits claimed are improved beer clarity and palatefullness. 

The flavour of wheat extract is relatively `neutral'. Wheat malts tend to be 

undermodified and their inclusion in the mash can lead to slower wort run off and 

sometimes fining problems in beers. Wheat has a naked grain, so it is easily damaged 

during handling and the acrospire (coleoptile) is not protected by a husk during 

germination. Water uptake is rapid during steeping. Usually soft wheats (TN less than 

1.9%) are malted and, like barley, they will respond to applications of gibberellic acid. 

Because of the absence of husk, which yields no extract, extracts of wheat malts can be 

relatively high, e.g., 328 lë/kg (on dry), 86%. Wheat malts may have moisture contents of 

5%, colours of 2 6 ëEBC; a TN of 1.87%, an SNR of 38 40%, a Kolbach index of up to 

or over 50%, and high levels of diastatic enzymes. Rye varieties have thin, naked grains 

and the malts made from them can confer unusual flavours (toffee, caramel) to beers, 


with changes to the mouth-feel (palate; smoother; more mellow) and a slight 

improvement in head retention. Rye malt may give ared tinge to beer. Rye varieties 

differ greatly in their suitability for making malts. Compared to barley the thin, naked 

grain takes up water quickly. Pale rye malts (2 8ëEBC) have exceptionally high but 

variable extracts (about 315 lë/kg, on dry, even 85 90% EBC) and very varied nitrogen 

contents (1.2 2.3%) and varied levels of diastatic enzymes. The malts are only 

occasionally used in special beers. 

Malts have been made from the `synthetic' wheat-rye hybrid triticale (Triticosecale) 

which, like its parents, has naked grains and takes up water rapidly during steeping. 

There are wide varietal differences in the malting quality of triticale cultivars. Although 

extracts can be very high (values of 335 lë/kg, (on dry) and 88 90% EBC have been 

reported) yet triticale malts are not used, possibly because the grain has atendency to 

have high TN values and the malts yield hazy worts rich in soluble nitrogen, including 

suspended and finely divided protein largely derived from prolamines. The worts are 

very viscous because of dissolved pentosans (Blanchflower and Briggs, 1991; Byrne et 

al., 1993). When milled, triticale malts give rise to afine flour which impedes wort 

separation. In the past, stouts were made with high proportions of oat malts in the grists. 

The reasons for using oat malts are not clear and fermentation problems (foaming 

fermentations and cloudy worts) were often encountered. However, the high husk 

content of oats meant that the `husky' grist favoured wort run off at the end of mashing. 

Small amounts of oat malts are now used to give character to special beers (toasted, 

biscuit-like aroma and an intense mouth-feel). The malts have some unusual 

characteristics. Extracts are low (e.g., 230 234 lë/kg on dry), the lipid content of the 

malts is high with the danger of the material becoming rancid and the beer having poor 

head retention and flavour instability. The TN value may be moderate (e.g., 1.6%), and 

the SNR is low (e.g., 18%). Diastatic power may be about the same as that of abarley 

malt having asimilar nitrogen content. 

Although malts are made from several tropical cereals only those made from sorghums 

have attracted much attention. In places in Africa millets have been malted mixed with 

sorghum grain, probably for the extra starch degrading enzymes provided by the malted 

millet (Chapter 16). Sorghum grain is steeped and watered during germination. Often the 

grain is treated with substances such as formaldehyde or sodium hydroxide in attempts to 

control the surface microbes. Relative to barley, sorghum is malted warm (e.g., 25 ëC; 77 ëF). 

The rootlets and the shoots are removed from the dried malt when lager-style beer is being 

made but not when the malt is for making opaque beer. The seedling tissues are rich in 

hydrogen cyanide. The information available on malted sorghum is inconsistent. This is 

probably because malted sorghum is used in making opaque beers (Chapter 16) as well as 

lagers and the requirements for these processes are different because of the large differences 

between different varieties of sorghum and because the methods of analysis and mashing 

that are applicable to barley malts are not all suitable for sorghum malts. This is primarily 

because the gelatinization temperatures of sorghum starches are higher than the starches of 

barley or wheat malts. Thus the extract of a sorghum malt determined using an inappropriate 

method (mashing at 65 ëC; 149 ëF) was found to be 112 lë/kg while mashing in an appropriate 

way gave an extract of 268 lë/kg. Elsewhere extracts from sorghum malts of between 65 and 

85% were found, with fine-coarse extract differences in the range 0.5 18.2%. Estimates of 

other analyses also varied widely. It has been generally accepted that sorghum malts are 

deficient in enzymes (this is not a problem when making traditional, opaque beers), but this 

may not be not true for malts made from some varieties. 


2.2.6 Special malts 

The malts already described are all `finished' on kilns. However, there is a group of malts 

which are finished in roasting drums (Bemment, 1985; Briggs, 1998; Gretenhart, 1997; 

Maule, 1998; Narziss, 1976). All these special malts are used as small proportions of 

grists to give particular colours, flavours and aromas (i.e., to impart characters) to beers. 

They can be considered in two groups; those that are prepared by a simple heating 

process, such as amber, diamber, brown, chocolate and black malts (and, by tradition in 

the UK, roasted barley), and crystal and caramel malts in which the wet malts are 

`stewed' so that the endosperm contents are liquefied before they are dried and cooked. In 

each group a wide range of colours occurs. As the colour ranges are continuous and as the 

qualities of the starting materials can be varied as, to some extent, can the roasting 

regimes, it follows that the number of malts that might be made is unlimited. The more 

usual types and divisions are described here, but intermediate types can be made and 

sometimes are. Because these malts are required primarily for the characters and colours 

that they provide, extract and colour are the analyses which, together with moisture 

content, are usually specified. Sometimes coloured malts are made from wheat or other 

cereals but only the barley malts are in common use. Unlike `white' malts, coloured malts 

should be used as fresh as possible, storage time being minimized, to retain their aromas. 

During their preparation the heating is so intense that no enzymes survive. As the colour 

increases in a series of malts so the malt extracts decline slightly as the extra colour is 

generated by more extreme or more prolonged heating. For example, in a series of 

German caramel malts the extracts and colours were: Carapils, E 78%, colour, 

2 5 ëEBC; Carahell, E 77%, colour 20 25ë EBC; CaramuÈnch, E 76%, colour 

50 300 ëEBC. As the colour increases so the wort pH values tend to decrease and the 

Kolbach indices decline. 

Amber malts are prepared by roasting pale ale or mild malts or, after drying, well 

modified green malts. Heating programmes begin at about 48 ëC (118.4 ëF) and rise to 

about 170 ëC (338 ëF). The `normal' colour range varies, but is usually 40 85 ëEBC. 

Moisture contents are 3.5% or less. Extracts vary between 270 and 285 lë/kg. These malts 

are valued for giving characteristic dry palates and baked or biscuit-like flavours to 

golden-coloured ales. Diamber malts are probably not made now, but modern brown 

malts are similar to amber malts prepared at higher temperatures. Such malts may have 

extracts of 260 280 lë/kg and colours of 90 150 ëEBC. Chocolate and black malts and 

roasted barley are also prepared in roasting cylinders but, relative to amber and brown 

malts, the heating is much more severe and there is a risk that the grain may catch fire. 

The process must be regulated so that no charring occurs and that when cut the grains are 

evenly coloured and have a floury texture, with no glassiness, and have the correct colour 

throughout. Well-modified green malt (TN 1.5 1.7%) is carefully dried and dressed. The 

material is loaded into a roasting cylinder and the temperature is increased from about 

75 ëC (167 ëF) to 175 ëC (347 ëF) and then more slowly to 215 ëC (419 ëF) for chocolate 

malts and to 225 ëC (437 ëF) for black malts. During roasting, unpleasant fumes are 

released and these must be eliminated by scrubbers or after-burners. Roasted barley is 

finished at a higher temperature, 230 ëC (446 ëF). Towards the end of roasting, when the 

heaters are switched off, the temperature of the load continues to rise as heat is generated 

in the grain. At this stage the operator checks colour every 2 3 min. and at the correct 

moment quenches the load with a spray of water. 

Roasted barleys usually have colours in the range 1200 1500 ëEBC. These are used in 

making some stouts and impart `sharp', `dry', àcidic' or `burnt' notes to the product. In 

contrast to roasted malts roasted barley gives no hint of sweetness. The roasted grains 


should be reddish-black, shiny and swollen and a proportion will be split. Extracts are 

HWE, 260 275 lë/kg, and moisture contents will be less than 2%. Pale chocolate malts 

will have colours of 500 600 ëEBC, and the more usual chocolate malts 

900 1100 ëEBC. Black malts have colours of 1200 1400 ëEBC. All have moisture 

contents of 2% or less and extracts of 255 275 lë/kg. Flavour descriptions of these 

materials are not satisfactory, but they include `dry', `burnt', àcid' and àstringent' but 

when chewed they have a residual sweetness which is distinct from the flavour of roast 

barley. The hot water extracts of chocolate and roasted malts and roasted barley are 

determined on finely ground samples mashed with boiling water at 100 ëC (212 ëF) in the 

IoB method, so enzymolysis is not involved. 

Crystal and caramel malts are unique in that during their preparation the endosperm 

contents are deliberately mashed, stewed and liquefied and, when cut, the finished malts 

should be hard and all the grains should be glassy or `crystalline' in appearance. These 

malts are prepared in a wide range of colours, some of which are named. They impart rich 

and delicious and other characteristic flavours and they give body to beers and are 

believed to improve beer stability. Sound barley, sometimes with a high nitrogen content 

of 1.7 2.0%, is malted. When it is well modified either the green malt is taken to a 

roasting drum directly or, less economically, it is lightly kiln dried. The green malt, or the 

re-wetted, kilned malt is warmed and held moist at a temperature of 60 75 ëC 

(140 167 ëF) until the contents of the grains are liquefied and the liquid contents can be 

squeezed out. The temperature is then increased and the grain is ventilated with hot air so 

that both cooking and drying occur. The liquefaction step ensures that starch, and 

possibly the endosperm cell-walls, are degraded and sugars and other soluble materials 

accumulate. Thus on heating and drying and depending on the exact conditions a 

concentrated sugar solution is produced together with various amounts of melanoidins 

and flavour and aroma substances. 

The finished product is rapidly cooled, and the contents solidify to a sugary, solid 

mass. Moisture contents are 3 7.5% and extracts are 260 285 lë/kg; 76 80%. Preferred 

colour ranges are about 20, 120 140 and 300 500 ëEBC. While the palest crystal malts 

are sweet, the darker malts have more complex flavours with caramel-, toffee-, malty-, 

aromatic-, honey-like and luscious characters becoming more apparent until in the 

darkest products harsher, burnt flavours appear. These products are variously called 

caramel or crystal malts. It has been shown that the flavour spectra can usefully be varied 

(Chandra et al., 1999). 

2.2.7 Malt specifications 

When brewers purchase malt they require it to be excellent in quality and moderate in 

price. They expect the extract yield and quality will be good and that beer production will 

run smoothly and yield a good product. Malts have different properties and are used to 

produce different types of beer. Brewers need to decide what analyses define the best 

malt with which to make a particular beer, and to agree with maltsters that this is what 

can and will be supplied. The analyses available do not reliably predict a malt's 

brewhouse performance and brewers have yet to agree on what set of analyses should be 

used to specifically define a malt. Cheap, poorly made malts are often undermodified 

and/or inhomogeneous and brewing with them can give rise to costs resembling those 

arising from mashing with excessive levels of particular adjuncts. For example, failure to 

recover the expected extract in the brewhouse or the need for a lengthened mashing 

programme, slow wort separation prolonging the lautering stage and so disrupting the 


production programme, excessive break-formation in the hop-boil, short filter runs and 

slow beer-filtration rates so the production cycle is further disrupted and extra filter aids, 

e.g., kieselguhr (diatomaceous earth) may be needed. Furthermore, there may be a need 

to use extra beer stabilization treatments and/or add extra enzymes to the mash. 

Inadequate yields of small nitrogenous molecules (FAN), that then limit yeast growth and 

fermentation, may also occur as may low carbohydrate fermentability (too few 

fermentable sugars) that ensures that alcohol yield is depressed. All these problems 

cause disruption in the production schedule and increase costs. 

In practice, brewers use different analyses in attempts to ensure that malts meet their 

requirements and the situation is complicated by the ongoing search for and introduction 

of improved methods (Aastrup et al., 1991; Briggs, 1998; Buckee, 1997; Copestake, 

1998; Gromus, 1988; Hyde and Brookes, 1978; Seward, 1992). The brewer may specify 

the variety(ies) of barley from which the malt may be made, and the harvest year, 

whether or not abrasion and/or additives may be used, details of the kilning cycle, and a 

minimum (or maximum for coloured malts) period between manufacture and delivery. A 

specification will contain an upper limit to screenings (thin corns) and dust, a maximum 

moisture content, a measure of the laboratory extract coupled to a lower limit, sometimes 

a preferred range for the fine-coarse extract difference, a total nitrogen (protein) limit or 

range, a range or limit for the total soluble nitrogen (protein) value and for the SNR or 

Kolbach Index, and often a lower limit for the free amino nitrogen. Values (maximum, 

minimum or ranges) may be specified for DP, -amylase and saccharification time, and 

limits may be set on the concentration of the DMS precursor. 

In addition, an upper limit or a range will be set for the colour of the laboratory wort, 

often before and/or after boiling. To these may be added specified limits for the other 

characteristics of the laboratory wort, including smell, clarity, pH, viscosity and -glucan 

content, and estimates of malt -glucanase, friability, homogeneity, and any others, 

including wort fermentability. As many of these values can be determined in more that 

one way and the results of analyses may be expressed in non-standard ways, even 

including non-standard units, a maltster in international trade may need to recognize 

nearly 300 analytical values, a situation so bizarre as to be ridiculous. In addition a 

guarantee may be needed to indicate that the malt is not contaminated with lead, arsenic 

or nitrosamines, mycotoxins or unapproved agricultural chemicals, insecticides or 

fumigants. 

Two other kinds of problem arise. The first relates to the brewer specifying 

combinations of malt characteristics that cannot be combined in one product. For 

example, it is not possible to produce a pale malt with a very rich flavour, or an enzyme 

rich malt that has a high colour. Malts with low SNRs cannot be made highly friable. 

Barleys with low nitrogen (protein) contents cannot be malted to give products 

exceptionally rich in enzymes, high nitrogen contents cannot be combined with high 

carbohydrate extracts. Malts with poor physical modification cannot have low -glucan 

contents, and so on. These facts are the inevitable consequences of the composition of 

barley and the integrated way in which changes in the grain occur during malting. The 

second kind of difficulty arises from drawing up specifications that are too inflexible, or 

`tight', so that they cannot be met routinely. For example, it is ridiculous to specify a 

particular analytical value without taking account of analytical variations and the 

variations that occur in barley. It is meaningless to specify that a malt's nitrogen content, 

TN, must be 1.65% exactly, when the repeatability and reproducibility values for the 

analysis are 0.049% and 0.085% respectively according to the Recommended Methods of 

the Institute of Brewing. Realistic specifications must be agreed between maltsters and 


ewers, probably annually, taking into account the changing varieties of barley being 

grown and the quality of the barleys available from the current harvest. 

2.3 Adjuncts 

Adjuncts are materials, other than malt, that are sources of extract (Briggs, 1998; Byrne 

and Letters, 1992; Letters, 1990; Lloyd, 1986, 1988a, b; Martin, 1978; Stowell, 1985). 

They are used because they yield less expensive extract than malt and/or they impart 

desirablecharacteristicstotheproduct.Forexample,theymaydilutethelevelsofsoluble 

nitrogen and polyphenolic tannins in the wort, allowing the use of high-nitrogen (protein 

rich) malts and the production of beer less prone to form haze. Some adjuncts enhance 

head formation and retention. The higher proportion of adjuncts used in amash the more 

difficult it is to achieve good extract recoveries and also wort viscosity is often increased, 

run off is slowed and fermentability is reduced. The addition of soluble sugars or syrups 

to the wort effectively increases the capacity of the brewhouse and provides asimple 

method for generating high-gravity worts and adjusting wort fermentability. Solid, `mash 

tun' adjuncts may be added to the grist and the starch they contain will be hydrolysed by 

enzymes from the malt or from other sources (Section 2.5). Other soluble preparations, 

sugars and syrups, otherwise `copper'- or `kettle'-adjuncts, are dissolved in the wort 

during the hop-boil. In addition to these abrewer may add other sugars to the beer as 

`primings', and caramels or other materials may be added to adjust beer colour. 

Adjuncts are analysed according to the official sets of methods (Section 1.15.1, p. 9). 

Since, like special malts, mash tun adjuncts are largely or wholly lacking in hydrolytic 

enzymes, analytical mashes are made with adjunct, which may be pre-cooked and mixed 

50:50 with an enzyme-rich malt. The extract yield of the adjunct is calculated, assuming that 

the extract recovered from the malt is half that which is obtained from an all-malt mash. The 

copper adjuncts are dissolved and the characteristics of the solutions are determined. The 

analytical values of interest include the specific gravity and hence the extract, the colour and 

clarity of the wort (often the colour is very low), the yield of soluble nitrogen, the oil content, 

the sulphur dioxide content, the pH, the ash content, levels of heavy metals (such as iron, 

copper, lead and arsenic), the level of microbes, the spectrum of sugars present and the 

fermentability of the mixture, the flavour, aroma and purity of the material and the absence 

of any deterioration. The amounts of adjuncts used vary widely. In some places their use is 

forbidden. In North America 60% of the extract in a brew may be derived from adjuncts, 

while elsewhere 10 20% is more usual. It is feasible to make beers with up to 95% of the 

grist being raw barley (Briggs et al., 1981; Wieg, 1973, 1987). 

The choice of adjunct(s) requires care. The material chosen must be regularly 

available in adequate amounts and be of good quality. The use of this material must 

enhance, or at least not reduce, the quality of the beer being made. It is difficult to switch 

between different kinds of adjunct. Apart from the risk of altering the nature of the beer, 

changing adjuncts may require alterations in the brewery equipment. For example, the 

handling plant needed for syrups is completely different from that needed for any mash 

tun adjunct and the equipments needed to handle flours, flakes and grits are all different. 

2.3.1 Mash tun adjuncts 

Mash tun adjuncts fall into three classes, those that can be mixed into the grist without 

pre-cooking, such as wheat flours, those that are pre-cooked before mashing begins (e.g. 


flaked maize, torrefied wheat) and those that are cooked in the brewery as part of the 

mashing programme, such as maize-, rice- and sorghum-grits (Tables 2.1, 2.2). The type 

of adjunct that astarch-rich material produces is largely determined by the gelatinization 

temperature of its starch (Table 2.3; Chapter 4). If the starch granules swell and lose their 

structure and become susceptible to rapid enzyme attack (i.e. gelatinize) at temperatures 

low enough for the malt enzymes to remain active, then that material (e.g. wheat flour) 

need not be pre-cooked. However, if the starch has ahigh gelatinization temperature (e.g. 

maize) the material must be cooked at ahigh temperature to gelatinize the starch (either 

by flaking or in acooker at the brewery site) before it is mixed with the main malt mash 

at atemperature at which the malt enzymes can act. 

Raw barley grain has been used as an adjunct after hammer-milling or other kinds of 

dry-milling or wet-milling. It is an advantage to wash the grain before use (Briggs et al., 

1981;Wieg,1973,1987).Theviability ofthisgrainisirrelevant.Itcontains -amylase (a 

proportion of which is insoluble) and some other hydrolases, as well as proteins 

inhibitory to some -amylases, proteases and limit dextrinase. Mashes containing much 

raw barley often need to be supplemented with enzyme mixtures from microbes 

containing -amylase, protease and -glucanase to convert the starch, to provide 

sufficient amounts of FAN and to degrade the relatively large amounts of -glucans that 

are present. Coarsely ground grain gives poor extract recoveries, but finely ground grain, 

while giving higher yields of extract, causes problems, in particular even greater 

quantities of -glucans are extracted. Because of practical difficulties, including the need 

for prolonged temperature programmed mashes and problems with the lack of desired 

character and raw-grain flavours in the products, interest in `barley brewing' has 

declined. 

Wheat has been processed in various ways, but most wheat is used as flour. Wheat 

flour milling is aspecialized process that, by aseries of roller milling and sieving steps, 

can produce material that is nearly pure endosperm tissue. By removing the germs and 

bran the starch percentage in the product is increased while the protein, ash and oil 

contents are reduced. Brewing flours are prepared from soft wheats, and often the 

nitrogen contents of the flours are high. By using air classification fractions can be 

obtained that are depleted in protein and enriched in starch and so yield higher extracts. 

For example, air classification of aflour containing 9.5% protein gave afraction with 

only 7% protein. The nitrogen-reduced material is used in brewing while the nitrogenenriched 

material is used in making biscuits. Handling flour is not simple. Special 

hoppers, usually equipped with vibrating feeds, are needed to ensure that the flour flows 

and specialized conveying equipment (vibrating or pneumatic) is needed. Flour dust 

mixed with air can form explosive mixtures and so all the usual precautions must be 

taken. 

To minimize handling problems, flours with `clumps' of starch granules, cell walls 

and protein have been prepared, with particle diameters of about 100 m(rather than the 

more usual 17 35 m) but perhaps the most convenient preparations are those in which 

the flour particlesareàgglomerated', thatis,boundtogetherwith asoluble bindersothat 

the material produces little dust and is handled in agranular form. In the mash the 

granules disintegrate releasing the flour. Wheat flour has ahigh extract content (340 lë/ 

kg, on dry, Table 2.1), and its use favours haze stability and especially head formation 

and retention. Raw or pre-gelatinized rye or millets, used in relatively small amounts, 

support head retention better than wheat (Stowell, 1985). However, wheat flour retards 

wort separation both because pentosans increase the viscosity of the wort and because 

they and proteins form fine particles that block the mash bed. Lipid micelles may also 


Table 2.1 Typical analyses of some starch-rich adjuncts 

Adjunct Moisture 

Hot water extract 

TN TSN Bulk density 

% (1ëkg as is) (1ëkg d.b.) (% d.b.) (% d.b.) (% d.b.) (kg/l) 

Maize grits 12 301 342 90 1.5 ± 0.76 

Maize flakes 9 313 344±355 ± 1.5 0.04 0.46±0.66 a 

Starches, refined 10 347 380±390 102±105 < 0.1 Neglible 0.60±0.70 

Rice grits 11 316 355 93 1.0 ± 0.85 

Rice flakes 9 325 357 ± 0.85 ± 0.30 

Wheat flour 11 304 342 ± 1.5 0.36 0.51 

Wheat, torrefied or micronized 4±9 291 310±315 ± 1.6±2.0 0.15 0.55 

Wheat flakes 5±8 287 302 ± 1.8 0.12 0.37±0.40 

Wheat, raw 12 260 295 ± 1.6 ± 0.77 

Barley, torrefied 5±6 254 267 72 1.8±2.2 ± 0.37±0.40 

Barley, flaked 

Barley, raw 

9 

12 

253 

250 

278 ± 1.8 0.12 0.25±0.26 

b 

284 b 

± 1.8 Variable b 

0.65±0.66 

Analyses IoB (1993) except HWE (%), determined by the ASBC method (ASBC, 1992). 

a Depends on the degree to which they are crushed. 

b Depends on added enzymes. 

From Lloyd (1986, 1988a); Brookes and Philliskirk (1987); Briggs (1998). 


Table 2.2 Analysis (ASBC) of various adjuncts 

Adjunct Moisture 

Extract 

Protein Fat/oil Fibre Ash pH 

Gelatinization 

temperature range 

(% as is) (% as is) (% on dry) (% as is) (% as is) (% as is) (% as is) (ëC) (ëF) 

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 

Corn (maize) flakes 

Refined maize grits 

4.7±11.3 82.1±88.2 91.0±93.4 ± 0.31±0.54 ± ± ± ± ± 

(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 

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 

142±172 a 

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 

Wheat flour 11.5 80.1 90.7 11.4 0.7 ± 0.8 ± ± ± 

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 

Torrefied wheat 4.9 74.4 78.2 12.2 1.0 ± ± 6.2 ± ± 

Torrefied barley 6.0 67.9 72.2 13.5 1.5 ± ± 5.9 ± ± 

a Variance in US rice types: 65±68 ëC (149±154.4 ëF); 71±74 ëC (c. 160±165 ëF) (long grain rice). 

Data of Canares and Sierra (1976); Coors (1976); Bradee (1977); Canales (1979); through Briggs (1998). 


Table 2.3 Some reported gelatinization temperature ranges of starches (Briggs, 1998; Reichelt, 

1983; Various). Reported values often disagree, probably because different methods have been used 

to determine them (Bentley and Williams, 1996). The values reported in ëF are only approximate 

equivalents of the temperature in ëC 

Gelatinization temperature range 

Starch ëC ëF 

Maize (Corn)* 62±77 143±171 

Waxy Maize* 62±80 143±176 

Sorghum* 69±75 156±167 

Millets* 54±80 129±176 

Barley 60±62 140±144 

Barley, small granules 51±92 124±198 

large granules 60±65 140±149 

Barley Malt 64±67 147±153 

Wheat 52±66 126±151 

Rye 49±61 120±142 

Oats 52±64 126±147 

Rice* 61±82 142±180 

Rice, short grain* 65±68 149±155 

Rice, long grain* 71±74 160±165 

Potato 56±71 133-160 

Tapioca 63±80 145±176 

Arrow root (Maranta) 67±85 152±185 

* Starches or adjuncts made from these materials must always be cooked before mashing. The other materials 

may be converted better if first cooked. 

contribute to these filtration problems, as they do with the filtration problems 

encountered with wheat starch hydrolysates (Matser and Steeneken, 1998). Sometimes 

these problems can be reduced by adding microbial pentosan-degrading enzymes to the 

mash. 

Wheat flour is now used to about 5 10% malt replacement in some British breweries 

although, in the past, much higher replacement rates, of 25% or even 36%, were used 

(Briggs et al., 1981). Wheat flour is used directly in infusion mashes, but higher extracts 

may be obtained if the flour is pre-soaked or is pre-cooked. From time to time purified 

starches from wheat, potatoes, manioc and other sources are used in mashing, depending 

on local economics. It might be expected that wheat starch would be fully converted in 

the mash, but it has been reported that better extract recovery occurs if the material is 

cooked at 96ëC (c. 205ëF), possibly because the small starch granules are gelatinized 

only at the high temperature. The material is not boiled to avoid frothing. 

Flours are produced, as by-products, during the manufacture of maize, rice and 

sorghum grits. Like the grits these flours must be cooked before being mixed in with the 

malt mash. The extract yields of refined starches are high since, due to the uptake of the 

water of hydrolysis during conversion, 100 units (on dry) of starch give rise to 

103 105% units of dry sugars and dextrins, so extracts of 380 390 lë/kg (on dry), or 

102 105% (on dry) are obtained. Generally, purified maize starch is not used directly in 

breweries although when it is it is cooked. Most brewing sugars are prepared from maize 

starch. 

Pre-cooked adjunctsused inmashinginclude micronizedandtorrefied whole grains or 

flaked wheat or barley or flaked maize grits or flaked rice grits or flaked pearl barley 

(Tables2.1, 2.2). These materials are easily handledand,becausethey have been cooked, 

they yield better extracts than the raw materials because their starches are gelatinized and 


the hemicelluloses are partly degraded. They are easily broken up in malt mills. Whole 

grains of wheat or barley, graded to remove thin grains and with adjusted moisture 

contents, are cooked in hot air at 220 260ëC (428 500ëF). During cooling the softened 

material becomes firm and has amoisture content of about 4%. Torrefied barley may 

have an extract (on dry) of 267 lë/kg or 72%, while for torrefied wheat the values may be 

310 315 lë/kg or 78 80%. It is advantageous to use grains with low nitrogen contents, 

since a1% increase in nitrogen content (6.25% crude protein) will reduce the extract 

yield by 5%. Micronized cereal grains have similar properties. These are prepared by 

cooking in athin layer, moving beneath gas flame heated ceramic tiles which give out 

radiant heat (infra-red radiation). The grain, which should be carefully conditioned and 

not heated for too long aperiod to obtain the best quality product, reaches atemperature 

of about 140ëC (284ëF) (Brookes and Philliskirk, 1987; South, 1992). Micronized grains 

may be cooled and mixed with the malt before it is milled. While it is still hot, 

micronized grain may be rolled to form flakes, which do not need to be dried. 

The older process, for flaking whole grains, or pearl barley or maize- or rice-grits, 

began by adjusting the moisture content of the material then, after a period of 

conditioning, cooking at 90 100ëC (194 212ëF), flaking it by passing it between hot 

rollers. Theflakes were dried inastream ofhot air,before cooling. At present (2004) it is 

not economic to use flaked rice but it, like flaked maize, is awell-liked adjunct that gives 

up its extract easily, in good yield, even in simple infusion mashes (Table 2.1). Flaked 

barley and, to an even greater extent, flaked pearl barley (grains from which the husk and 

surface layers have been removed by abrasive milling; Briggs, 1978) give problems in 

brewing largely because they contain comparatively large amounts of -glucan. Flaked 

barley has been prepared sprayed with asolution of bacterial enzymes containing - 

amylase, -glucanase and probably protease. The product had an appreciable cold water 

extract and did not give rise to highly viscous worts or any of the other problems 

associated with -glucans. In the past flaked oats were used in making some stouts. They 

were described as being greyish, with low extracts of 252 282 lë/kg, and were rich in 

husk, protein and in oil that could readily become rancid. Experimentally it has been 

shown that milled, cooked and extruded cereals are convenient adjuncts (Briggs et al., 

1986; Dale et al., 1989; Laws et al., 1986). These preparations seem to be used only in 

the preparation of some African beers. 

Grits are preparations of nearly pure starchy endosperm in which the starch granules 

are invested with protein and are enclosed with cell walls (Johnson, 1991). For brewing 

purposes these are prepared from maize (`corn'), rice or sorghum(Tables 2.1, 2.2). These 

gritsmustbecookedbeforebeingmixedwiththemainmaltmash.Thehightemperatures 

used (up to boiling or, when processed under pressure, even over 100ëC; 212ëF) disrupt 

the cellular structure of the grits and gelatinize the starch. The -amylase included in the 

mixture liquefies some of the starch, reducing the viscosity of the mixture and preventing 

retrogradation. The -amylase may be bacterial in origin or it may be from asmall 

amount of highly enzymic, ground malt. Brewing rice is usually aby-product of grain 

beingprepared forhumanconsumption.Thismaterialhasbecome tooexpensivetousein 

manyareas,butitisstillusedinAsia.Preferredricegritsarelessthan2mm(0.079in.)in 

diameter, have moisture contents of about 13%, and extracts of 88 90 or even 95% (on 

dry). Typically they contain 5 8% protein and 0.2 0.4% oil and about 0.9% ash. The 

flavour imparted to beers by rice are described as neutral, `dry', `light' and `clean'. 

Different grades of rice behave very differently in mashing, so that wort separation 

timesmayvarybyfactorsof2or3andthegelatinizationtemperaturesofthestarchesvary 

widely (Table 2.3). When rice grits are slurried in water and are progressively heated it is 


found that the gel point of each sample (the temperature at which the viscosity suddenly 

increases) is related to the difficulties of using the material in the brewhouse (Teng et al., 

1983). It is advantageous to use athermostable bacterial -amylase when cooking rice. It 

seems that all rice grits should be heated to boiling. Rice grits, like maize grits, may be 

cooked and flaked. Flaked preparations are used in brewing without the need for acereal 

cooker.Typicalanalysesofmaizegrits,whicharepreparedfromyellowdentcornbyadry 

milling process, are: moisture 13%; protein 7 9%; fat, 0.7 1%; ash, 0.5 0.7%; extract 

about 90% (on dry). Usually particles have diameters of 0.3 1.5mm, 0.012 0.059 in. 

(Tables 2.1, 2.2; Johnson, 1991). These grits impart afuller flavour to beer, compared to 

rice grits. In some areas, notably Africa and Mexico, sorghum grits are used. In quality 

theycloselyresemblemaizegrits,givingextractsof91%(ondry),withmoisturecontents 

of 11 12%. When first used sorghum grits gave unpleasant flavours to beers, but 

improved milling techniques, processing large yellow or white, low tannin grains, now 

produce fully acceptable materials. Pearl barley is analogous to grits, being almost pure 

endosperm tissue. It does not need to be cooked but it is now little used in brewing. Grits 

can be prepared from several millets, but this is probably not done commercially. 

2.3.2 Copper adjuncts 

Copper adjuncts come in two categories (Table 2.4). First, wort extenders, which add 

essentially only carbohydrates (such as sucrose, invert sugar and hydrolyzed starch 

syrups) and wort replacements such as malt extracts and syrups made from hydrolyzed 

cereals. These materials add carbohydrates and acomplex mixture of other substances to 

the process stream. The formulae of sugars are given in Chapter 4. 

Sucrose (`sugar'), derived from either sugar cane or sugar beet, is awell-liked copper 

adjunct, used either as asolid or in solution and either as the disaccharide sucrose ( -Dglucopyranosyl-(1,2)- 

-D-fructofuranose) or as the hydrolysis product, the equimolecular 

mixture of glucose and fructose, ìnvert sugar', so called because as the sucrose is 

hydrolysed the optical rotation of the solution decreases and becomes negative and 

ìnverted'. At present, with the exception of Australia, sucrose-based materials are little 

used because they are costly. Beet sugar must be used pure, because the impurities have 

unpleasant flavours. While pure cane sugar is perfectly acceptable, partially purified 

preparations have been preferred because of their luscious flavours. These preparations 

may contain small quantities (perhaps 5%) of unfermentable di- and tri-saccharides 

(Table 2.4). Sucrose is extremely soluble (see Appendix), solutions containing over 63% 

solids being attainable. However, concentrated solutions of pure sugars are liable to 

crystallize. Solutions of invert sugar containing 83% solids can be prepared. Some 

brewer's preparations contain both sucrose and invert sugar. Yeasts ferment these sugars 

easily so, as the sugars dissolve completely in the wort, extract recovery is 100% and, 

with the pure preparations, the added sugars are 100% fermentable. The sugars may be 

provided in solution or as solids. A sugar syrup may give an extract of 258 lë/kg (fresh 

wt.), have a specific gravity of about 1.33 and a colour of 3 12 ëEBC. Nitrogen contents 

(e.g. 0.01%) are negligible. An invert sugar preparation may have an extract of 318 lë/kg 

(fresh wt.; Table 2.4). To prevent crystallization and to reduce the viscosity, so improving 

handling characteristics, these sucrose or invert sugar syrups are handled and stored 

warm, at 40 50 ëC (104 122 ëF). They cannot be stored for long periods, and so must be 

delivered shortly before use. 

Sugar adjuncts used in small amounts include lactose (from whey; a sweet, nonfermentable 

sugar), honey and maple syrup (Wainwright, 2003). Many sugar preparations 


Table 2.4 Typical analyses of sugar-rich brewing adjuncts, priming sugars and caramels 

Preparation Hot water Total nitrogen Colour Fermenability Specific gravity 

extract (f.wt.) (%) 10% w/v solution solids (20 ëC) 

(f.wt.) (EBC units) (%) 

(1ë/kg) a 

Sugar preparations 

Raw cane sugar syrup 258 0.01 3 95+ 1.33 

Invert sugar liquid 318 0.01 3±12 95+ 1.43 

Sugar primings 310 0.01 3±12 95+ 1.42 

Refined maize-starch hydrolysates 

Brewing syrup 310 0.02 Colourless or 77±78 1.42 

adjusted b 

Confectioners' glucose 318 < 0.01 Colourless 30±50 1.43 

Solid brewing sugar 310 0.02 Colourless or 86±87 ± 

adjusted b 

Glucose chips 318 0.01 20±50 82 ± 

Other materials 

Grain-based syrup 302 0.4±0.8 4 65±70 1.40 

Malt extract 302 0.65±1.3 4 70 1.40 

Caramel, 46,000 liquid 242 ± 4600 ± 1.29 

Caramel, 32,000 liquid 284 ± 3200 ± 1.36 

a 

Dry, solid sugar preparations, e.g., sucrose, have values of 382±386 1ë/kg. 

b 

FAN values of 0.01±1.15%. The colour may be adjusted to specification by the addition of other sugar-based products. 

Analyses IoB. 

After Lloyd (1986, 1988a); through Briggs (1998). 


are made from `refined grits', refined maize starch (corn flour). This is prepared by a 

continuous wet milling process. Maize grain is soaked in a solution of sulphur dioxide 

and is then broken up. The oil-rich germs and hulls are separated and the remaining 

endosperm tissue is milled and the gluten and starch granules are separated. The starch is 

recovered as a 35 40% suspension in water. This material may be dried. The powdery 

product is dusty and must be handled with all the precautions used with flours. 

Sometimes starch is added to the cereal cooker with grits, but it is usually converted into 

solutions of hydrolysis products by specialist manufacturers. A sample of maize starch 

had an amylose:amylopectin ratio of 28:72, a moisture content of 11 12%, a crude 

protein content of 0.35% and a fat content of 0.04 to 0.5%. The dry component of this 

material was mainly polysaccharide. Most maize starch is used in brewing, after 

hydrolysis, as syrupy copper (kettle) adjuncts. The starch is treated in two stages. In the 

first stage it is cooked to disrupt the granules and the material is treated with a mineral 

acid or a thermostable bacterial -amylase (stabilized by additions of calcium salts) to 

`liquefy' the polysaccharide, degrading it to a mixture of dextrins, oligosaccharides and 

sugars. In part the high cooking temperature is needed to disrupt amylose-lipid 

complexes, making the polysaccharide more easily degraded. The liquefied mixture 

flows and has a comparatively low viscosity, in contrast to cooked, but not liquefied, 

starch which is very viscous and sets to a gel on cooling. The liquefied material is partly 

purified by treatment with active charcoal and/or ion exchange resins to remove lipids 

and ionic substances. If mineral acid was used then this must be neutralized if the next 

process is to be enzyme-catalysed. In the second stage the liquefied material is 

saccharified to produce the mixture of carbohydrates finally required. Saccharification 

may be carried out with mineral acid or, after adjustment of the pH, with one or more 

enzymes. 

A very wide range of products, varying in salt content, sugar spectra and 

fermentability, are available. The materials may be classed as acid/acid, acid/enzyme 

or enzyme/enzyme products. Those prepared using acid hydrolysis may have high salt 

contents and, because of side reactions occurring during hydrolysis, may contain 

oligosaccharides containing unusual inter-sugar linkages, and may be coloured and have 

characteristic flavours. Thus acid/acid hydrolysis can yield confectioner's `chip sugar', 

which is rich in glucose and with colour in the range 200 500 ëEBC. Acid/enzyme and 

enzyme/enzyme products may be produced with little colour and with closely controlled 

compositions. They may be dried and delivered as solids or in solution as liquid syrups. 

Generally, like the sucrose-based syrups, these syrups are kept warm (at 50 ëC; 122 ëF) or 

above) to prevent crystallization and the separation of solids from the mix and to reduce 

the viscosity. They may be delivered and stored at 60 70 ëC (140 158 ëF). The surfaces 

of the stored materials are often ventilated with sterile air to remove water vapour which 

otherwise might condense and drip back onto the surface of the syrup, so locally diluting 

it and allowing the proliferation of microbes, notably osmophilic yeasts. The headspace 

may be filled with nitrogen or be illuminated with sterilizing, ultraviolet light. 

Often syrups contain sulphur dioxide as a preservative (2 40 mg/l), and brewers 

specify an upper concentration. Syrups are described as having reducing dextrose 

equivalent (DE) values. However, as different mixtures of sugars and dextrins can have 

the same DE values, these are of limited use to brewers. Preparations can be obtained 

with fermentabilities ranging from 30 95%, but usually values are about 75 85%. 

These syrups can be used to adjust the final fermentability of wort. However, the 

fermentability of a syrup is not a sufficient characterization, the spectrum of sugars 

present is also significant. Thus the fermentable material may be rich in glucose, or be 


nearlyentirelyglucose.Thismaybeundesirablesinceyeastsinwortsrichinglucosemay 

not be able to adapt to metabolize maltose and maltotriose, leading to slow or `hanging' 

fermentations. Glucose-rich syrups are usually made with the enzyme amyloglucosidase, 

sometimes mixed with adebranching enzyme to accelerate the hydrolysis of the starch 

and dextrin -1,6-linkages. This problem does not arise if most of the fermentable 

carbohydrate is maltose. Maltose-rich syrups are made by incubating liquefied starch 

with a -amylase (a plant enzyme or the enzyme derived from Bacillus polymixa) and a 

debranching enzyme such as pullulanase. 

Starch-derived syrups and malt extracts and syrups prepared from cereal grains are 

introduced into the wort during the hop-boil, as are solid sugars (Chapter 10). All these 

materialsmustbedissolvedandfullydispersed.Ifthisisnotachievedandmaterialsettles 

in the copper, the sugars can burn on to the heating surfaces with the creation of heating 

and cleaning problems, aloss of extract and perhaps the generation of unwanted flavours 

and colours. As prepared these syrups are very pale but, if required, makers may add 

caramel to give aspecified colour. 

Other copper adjuncts are malt extracts or syrups obtained by hydrolysing cereal 

grains (Briggs, 1978, 1998; Tables 2.4 and 2.5). These materials contain both 

carbohydrates and a complex mixture of substances including nitrogenous materials, 

minerals and yeast growth factors. Additions of these materials to the wort are equivalent 

to adding concentrated wort to the beer production stream. Malt extracts are made by 

grinding the malt, mashing it, with or without mash tun adjuncts and supplementary 

enzymes, and separating the wort, then concentrating it using triple effect vacuum 

evaporators. Many types of material can be produced depending on the grist, the mashing 

programme employed and the evaporation conditions used. By mashing enzyme-rich 

malts at low temperatures and concentrating the worts at low temperatures, enzyme-rich 

malt extracts may be obtained. At the other extreme, by heating the wort strongly, 

sometimes at a reduced pH, before concentration a product lacking enzymes can be 

prepared. Extracts can contain 75 82% solids (SG values 1400 1450), the more 

concentrated materials being used in the tropics. To keep the preparations liquid they 

need to be kept warm (e.g. 50 ëC, 122 ëF). At this temperature the material will slowly 

continue to darken and its other characteristics will change, so it should be used promptly. 

A representative extract is 302 lë/kg (fresh wt.). Colours may range from 3 520 ëEBC, 

have varied enzyme contents (DP 0 400 ëL), and fermentabilities in the range 56 93%. 

Some of the mash grists contain large proportions of raw cereal or cereal adjuncts, and 

to obtain adequate extracts the mashes may be supplemented with microbial enzymes and 

long, rising temperature programmes may be used. These products are best termed cereal 

syrups. A distinct product was `liquid malt'. This was made by mashing green barley 

malt, so eliminating the cost of kilning. The wort was concentrated in the usual way and 

the unwanted flavour components were evaporated during the concentration stage. 

According to German law this material is not an adjunct and so, like conventional malt 

extracts, its use is permitted. Potentially such syrups are highly fermentable, can be 

enzyme rich and low in proanthocyanidins (anthocyanogens), and so their use favours 

haze stability in beers. Malt extracts and cereal syrups are used less by large-scale 

breweries than was once the case. In contrast syrups made by hydrolysing starches are 

widely used. While malt extracts were once added to supplement the enzyme contents of 

mashes this highly uneconomic practice has long been discontinued, at least in largescale 

brewing. However, 3.7 volumes of malt extract give about the same amount of 

extract as 10 volumes of malt, making it a very compact source of extract, and it has been 

the practice to send malt extract (pre-hopped or not) to be fermented to make beer at 


Table 2.5 The carbohydrate compositions (%) of two worts and several syrups prepared from starches (after Wainwright, 2003) 

Infusion Decoction Low fermentable 63 DE High maltose Very high High dextrose 

mash wort mash wort syrup* syrup* syrup* maltose syrup* (glucose) syrup* 

Glucose + Fructose 10 11 4 38 2 3 94 

Sucrose 5 2 0 0 0 0 0 

Maltose 45 52 10 33 55 68 3.5 

Maltotriose 15 12 12 6 16 18.5 0 

Dextrins y 25 23 74 23 27 10.5 2.5 

Fermentability (%) z 72 74.8 23.6 75.6 69.8 85.8 97.5 

* Starch hydrolysates do not contain fructose or sucrose. 

y Dextrins are not fermentable. 

z Calculated fermentability. 


emote locations or on ships. In addition, hopped or unhopped malt extracts are used by 

many home brewersand small-scale or`micro' brewers toavoid the inconvenience of the 

mashing and wort separation operations. Copper adjuncts effectively increase the 

production capacity of abrewery. They are convenient for preparing high-gravity worts 

and for adjusting wort fermentability. Most add insignificant amounts of nitrogenous 

substances or polyphenols, or flavours or colours to wort. The approved sets of analytical 

methods specify ways of evaluating copper adjuncts. In contrast to malts and mash tun 

adjuncts all the potential extract of acopper extract is recovered in the wort provided that 

it is completely dissolved. 

2.4 Priming sugars, caramels, malt colourants and Farbebier 

The materials described in this section are not regarded as adjuncts. However, they all 

add extract to the wort or the product and so they are considered here. Priming sugars are 

addedtobeersthataretobecask-orbottle-conditioned.Theobjectistoprovidetheyeast 

with asupply of easily fermented sugars that can indirectly supply the carbon dioxide 

needed to carbonate the beer, and `bring the beer into condition'. Since the sugars are 

mostly fermented their nature is not important; sucrose, invert sugar and glucose- or 

maltose-rich syrups will serve. However, if the preparation contains aproportion of 

unfermentable material this will remain in the beer and may alter its character. To utilize 

some of the residual dextrins present in beer, enzymes have been added to catalyse the 

hydrolysis of aproportion into fermentable sugars, aprocedure which removes the need 

for priming sugars. Various enzymes have been used for this purpose. Amyloglucosidase 

was unsatisfactory since it is too stable and so is not reliably destroyed by the 

temperatures reached during pasteurization. Consequently the enzyme continues to act 

and sweeten the beer when its activity is no longer required. Less stable enzymes such as 

fungal -amylase, or pullulanase with -amylase have been more successful, but the 

problem of deciding when the correct degree of dextrin degradation has occurred, and so 

when the enzymes must be inactivated, still remains. Sugars may be added to some 

filtered and sterile beers to sweeten them. If this is done then sucrose or high-fructose 

preparations are probably to be preferred. 

Caramels are used to adjust colour by adding them to the wort or beer (Chapter 9). 

Caramels are made in different ways and not all types are suitable for brewing purposes 

(Comline, 1999). The class III, electropositive-ammonia caramels, the caramels used in 

beers, are made by heating sugars (usually high glucose syrups) with ammonia. Complex 

reactions occur and the product is a mixture of high molecular weight coloured 

substances and lower molecular weight substances which impart flavour and aroma. The 

preparations may have colours up to 35,000 ëEBC, contain 65 75% solids, 2.5 5% 

nitrogen and have pI values of 6.0 6.5. (A pI value of a substance is the pH at which it is 

50% ionized). By using ultrafiltration the coloured and flavoured components can be 

separated, permitting beer colour and flavour to be adjusted separately (Walker and 

Westwood, 1991). The specifications of brewing caramels usually include values for 

colour, pH, extract content, and stability when dissolved in worts and beers. 

Sometimes the use of caramels is forbidden but it is permissible to use extracts from 

coloured malts. Crystal, chocolate or black malts (or roasted barley, where allowed) are 

extracted with hot water and the extracts are concentrated. Colours (of 10% solutions) of 

850 1,700 ëEBC may be obtained. It is not clear how widely these malt colourants are 

used. In Germany beer colour may be adjusted using Farbebier. This `colour beer' is 


produced by specialist manufacturers (Narziss, 1992; Kunze, 1996; Riese, 1997). A 

mixture of, say, 60% pale malt and 40% dark malt is mashed to give awort with avery 

high density (e.g. 18 20ëPlato, approx. SG 1074 1084). The extract is boiled and 

fermentedinaspecialwaytogiveaproductwithacolourofabout8,000ëEBC.Itmaybe 

concentrated under vacuum. This material is undrinkable, but it is added to wort or beer 

to adjust the colour. Sometimes the material is treated with active charcoal to reduce 

bitter flavour. 

2.5 Supplementary enzymes 

Enzymes derived from sources other than malt may be used at various stages during 

brewing, provided that thisis allowed by local regulations (Bamforth,1986; Briggs et al., 

1981; Byrne, 1991; Godfrey and Reichelt, 1983). Enzymes are also used in the 

production of some adjuncts (Section 2.3). These enzymes are mostly prepared using 

liquid suspension cultures of various microbes (bacteria and fungi), but a few are 

obtained from plants and at least one was obtained from animal sources. The 

preparations, which may be dry powders or solutions, must be approved for use in 

foodstuffs. They are not `pure', and will usually contain residual materials from the 

nutrient medium in which the microbes were cultured, other enzymes besides the one(s) 

specified, diluents, extenders or carriers, and preservatives. They should not contain 

viable microbes. The preparations available have a wide range of characteristics. 

Different suppliers describe their preparations in different ways so that it is difficult to 

make comparisons between them. The lack of standard analyses is a source of difficulties. 

The temperature and pH optima of enzymes are so influenced by incubation conditions, 

and the conditions used in different breweries and at different stages of brewing are so 

varied, that it is not possible to give useful values. Consequently the effectiveness of the 

addition of an enzyme preparation must be determined by brewers under their particular 

processing conditions. 

The activities of `named' enzymes in preparations are standardized by suppliers. 

However, this is not true of other enzymes that may be present. The presence of these 

additional enzymes may be advantageous or harmful. For example, the presence of - 

glucanase in preparations of bacterial -amylase may be beneficial when added to a 

mash, particularly if undermodified malt or barley or oats adjuncts are used in the grist. 

On the other hand, while the presence of protease activity may be an advantage if more 

FAN is needed, it is most undesirable if it elevates the levels of soluble nitrogen too far 

and/or if the degradation of protein leads to a reduction in foam formation or stability. 

The presence of some àdditional' enzymes can easily be detected (Albini et al., 1987). 

Enzyme preparations are not stable, so they should be stored cool and used fresh. 

Different enzymes in a mixture will usually have different half-lives, so the ratios of 

enzyme activities in a preparation will alter with storage time. This may generate 

problems. Many of the enzymes used in the manufacture of starch- and cereal-derived 

syrups may also be used in breweries. Usually enzymes, where used, may be added to the 

mash or the cooker, or they may be added to the wort or beer. Used intelligently they can 

improve extract recovery, wort collection rate, the rate of beer filtration and the length of 

filtration runs, wort fermentability, and the resistance of the beer to haze formation. 

Added enzymes can minimize the presence of residual starch or gums in the wort. Other 

uses are indicated later. The enzymes of most interest in brewing are those which catalyse 

the hydrolysis of starch and dextrins, those which attack hemicelluloses and gums (both 


-glucans and pentosans), and those which degrade proteins. However, other enzymes 

may be of interest. While some brewers may use added enzymes on a routine basis others 

use them only to combat unforeseen production problems. 

-Amylases used in brewing are from different sources and, because of their different 

properties, they are suited to different purposes. They are all stabilized by elevated levels 

of calcium ions and by their substrates, starch and dextrins. They are all endo-acting 

enzymes, that is they catalyse the hydrolysis of the -(1, 4)-links within the dextrin, 

amylose and amylopectin chains. However, the range of hydrolysis products differs 

significantly, and the enzymes differ in thermal stabilities to a remarkable extent. Fungal 

enzymes (usually from Aspergillus spp.) have pH optima in the range 5.0 6.5, and 

temperature optima of around 60 65 ëC (140 149 ëF), or 55 ëC (131 ëF). Despite these 

low values these preparations have been added to mashes where the complex mixture of 

èxtra' enzymes (which commonly include hemicellulases and proteases) may be of 

value. This type of enzyme, which is inactivated by pasteurization, has been added to 

beers to hydrolyse dextrins and so obviate the need for priming sugars. It produces 

appreciable amounts of maltose among its products. 

Several types of bacterial -amylase are in use. The enzyme from Bacillus subtilis has 

a pH optimum between 6.0 and 7.5, but it is usefully active at mash pH values, 5 6. The 

temperature optimum is around 65 70 ëC (149 158 ëF), but is strongly dependent on the 

presence of starch, which stabilizes it. The enzyme may act briefly at temperatures up to 

80 ëC (176 ëF). Usually preparations of this enzyme, like those other bacterial enzymes, 

contain protease and -glucanase activities. While the alkaline protease may have little 

action under mashing conditions the neutral protease does. The enzyme from another 

bacterium, Bacillus subtilis, var. amyloliquefaciens is appreciably more heat stable. This 

-amylase has a reported pH optimum at 5.7 5.9 (at 40 ëC; 104 ëF). Although its 

temperature optimum is about 70 ëC (158 ëF) this enzyme is able to liquefy a 35 40% 

starch slurry at 85 90 ëC (185 194 ëF), and so it is useful for liquefying the starch when 

adjuncts are cooked, since it is so much more stable than the malt enzyme. In contrast the 

-amylase from Bacillus licheniformis is too heat stable for some brewing purposes. This 

enzyme, which has a wide pH optimum around 6, has a temperature optimum at 90 ëC 

(184 ëF) at high calcium ion concentrations. It can act briefly at 115 ëC (239 ëF), and it is 

not reliably destroyed by boiling unless the solution is slightly acid and the calcium and 

starch concentrations are low. These conditions can be met when the enzyme is used to 

liquefy starch during the manufacture of sugars and syrups, but cannot be reliably 

achieved in brewing. 

Debranching enzymes are used in the manufacture of copper adjuncts, and they have 

been investigated for use in the brewhouse. Two types of enzyme have been investigated. 

Isoamylase is able to hydrolyse the -(1,6)-links in amylopectin but not in dextrins. This 

enzyme seems not to be of value in brewing. However, pullulanase, an enzyme produced 

by the bacterium Klebsiella pneumoniae (Aerobacter aerogenes), hydrolyses -(1,6)links 

in both amylopectin and in dextrins, including limit dextrins. The enzyme is 

thermolabile, and is used at 45 55 ëC (113 131 ëF), when saccharifying dextrins with 

amyloglucosidase or -amylase in making glucose- or maltose-rich syrups respectively. 

The enzyme has been added to cooled mashes in experimental brewing, and it has been 

used, together with -amylase, to replace priming sugars in beer. As it is readily 

inactivated by heat this process can be stopped by pasteurizing the beer. The pH optimum 

has been given as 5.5 6.0, but the enzyme has been used at values as low as 4. 

-Amylases may be obtained from plants or particular bacteria. Enzymes from cereals 

(including flours), soya beans and sweet potatoes have been used to saccharify dextrins, 


with or without the addition of other hydrolases. The pH optima are about 5.3, but the 

useful pH range is about 4.5 7.0. These enzymes attack the penultimate -(1,4)-links in 

starch chains, releasing the disaccharide maltose. They are readily denatured by heat, and 

have temperature optima around 55ëC (131ëF). These enzymes have been added to 

mashes to increase the wort fermentability, and they have been added to wort for the 

same purpose and to beers to replace priming sugars. These enzyme preparations often 

contain -glucosidase (which is generally ignored) and may contain the unwanted 

enzyme lipoxygenase (LOX) as well as other enzymes. The bacteria Bacillus polymixa 

and Bacillus cereus, var. mycoides, produce both pullulanase and -amylase which, 

acting together, have been used when making high maltose syrups. 

Amyloglucosidase (syn. glucoamylase; AG; AMG) is prepared from several different 

fungi (e.g. Aspergillus spp., Rhizopus spp.). Some preparations also contain -amylase 

and/or transglucosidase. The latter is undesirable as it catalyses the formation, by 

transglucosylation, of unwanted and unfermentable oligosaccharides such as isomaltose 

and panose. Amyloglucosidase attacks the non-reducing ends of starch chains and 

dextrins releasing glucose. Its attack on -(1,4)-links is comparatively rapid relative to 

the attack on -(1,6)-links, so the conversion of starch into glucose by this enzyme is 

accelerated by the addition of pullulanase. The optimal pH range is 4.0 5.5, and the 

enzyme will act for extended periods at 60 65ëC (140 149ëF). It has been added to 

mashes (particularly mashes containing large proportions of adjuncts) to increase the 

fermentability of the wort. It is regularly used in the production of glucose and has been 

added to beer to replace priming sugars. It is no longer used for this, being replaced by 

more thermolabile enzymes. 

There is aproposal to add aglycosyl transferase to mashes to increase the levels of 

unfermentable isomaltooligosaccharides in the wort to produce abeer with areduced 

alcohol content but with a full body. In contrast, the same enzyme added to cool, 

fermenting wort increases the fermentability and hence the final alcohol content 

(Robinson et al., 2001). 

Whenundermodified orinhomogeneousbarleymaltsareusedorwhenbarley(oroats) 

mashtunadjunctsareemployed,problemscan ariseinthebreweryandtheseareoften,at 

least partly, due to residual, high molecular weight -glucans. Similarly, when problems 

arise from the use of wheat, rye or triticale adjuncts or wheat malt the problems are often 

attributed to pentosans. The problems include slow wort separation, slow beer filtration 

and short filter runs and sometimes the separation of hazes and gelatinous precipitates in 

the beer. The enzymes used to degrade -glucans may be divided into -glucanases and 

cellulases. Sometimes these preparations contain complex mixtures of enzymes. Because 

the structures of pentosans are complex (Chapter 4) mixtures of enzymes may be needed 

to obtain substantial degradation of these materials. 

The -glucanase of Bacillus subtilis is a well characterized enzyme, with an optimal 

pH range of 6.0 7.5 and temperature range of 50 60 ëC (122 140 ëF). In temperatureprogrammed 

mashes it acts best at about 50 ëC (122 ëF). However, the enzyme will act in 

brewery mashes, at least briefly, at about pH 5.3, at temperatures up to 75 ëC (167 ëF). 

This enzyme is specific in that it attacks only mixed-linked -(1,3;1,4)-glucans. It has 

been used in mashes made with barley adjuncts, and it is usually accompanied by - 

amylase and two proteases. Fungal -glucanase preparations (e.g. from Aspergillus spp.) 

have varied properties, but usually have inconveniently low temperature optima 

(45 60 ëC; 113 140 ëF) for mashing but have convenient pH optima in the range 

3.5 6.0. They probably contain a complex mixture of hydrolases, and are not clearly 

distinguished from the cellulases. 


Apreparation from Humicola insolens is active at degrading -glucans at up to 75ëC 

(167ëF). Cellulases used in brewing include those from Trichoderma spp. (T. reesei; T. 

viride), with temperature optima of 50 55ëC (122 131ëF) and pH optima in the range 

3.5 5.5. Such preparations are useful in temperature programmed brewery mashes. They 

contain mixtures of enzymes, including amylases and pentosanases. Cellulase preparationsfromPenicilliumfuniculosumhaveactivityinthepHrange4.3 

5.0,andfunctionat 

temperatures of 65ëC (149ëF). Preparations from P. emersonii are more heat stable, with 

an optimal temperature of 80ëC (176ëF) and auseful optimum pH range of 3.7 5.0. The 

enzyme mixture attacks not only mixed link barley -glucans but also holocellulosic 

material in barley, starch and pentosans. It is well suited for addition to mashes. 

Pentosanases need to be complex mixtures of enzymes and contain acetyl esterase, 

feruloylesterase, -L-arabinofuranosidase, exo-(xylobiase) andendo-xylanase activities. 

Preparations usually contain starch-, cellulose- and -glucan-degrading activities. 

Preparations have been made from Disporotrichum, Trichoderma and Aspergillus spp. 

Usually these are used in temperature-programmed mashes, being active at about 50ëC 

(122ëF). They are particularly useful when wheat, rye and triticale adjuncts are used. 

Non-maltproteolytic enzymes areusedfor twopurposes inbrewing.First, byaddinga 

protease to the mash, the amount of nitrogenous yeast nutrients (FAN; formol-N) in the 

wortisincreasedand,secondly,byaddingaproteasetobeer,polypeptidehazeprecursors 

are degraded. Pepsin, an animal protease with an acidic pH optimum, was added to beer 

as astabilizing agent, but this function is now carried out by thiol-dependent plant 

proteases, in particular papain, from the latex of the pawpaw (Carica papaya), bromelin 

from the pineapple (Ananas spp.) and ficin from figs (Ficus spp.). These enzymes differ 

alittle in their properties. All are destroyed by pasteurization and, while they degrade 

hazeprecursors,theyapparentlydonotdegradethedesirablefoam-formingpolypeptides. 

In contrast the bacterial proteases do destroy foam precursors. Protease activities are 

often present in fungal enzyme preparations, with pH optima in the range 3 6, and 

temperature optima around 50ëC (122ëF). Probably the bacterial proteases are of most 

interest, since these have been added to mashes to increase the FAN levels. Bacillus 

subtilis, like some other Bacilli, produces proteases having neutral and alkaline pH 

optima. The mixture works at mash pH values, and has an optimal temperature of 

45 50ëC (113 140ëF). When mashes are made that are rich in raw barley and are 

supplemented with bacterial enzymes an extended stand is needed at 50ëC (122ëF) to 

obtain an adequate level of soluble nitrogen in the wort. Raw barley contains an inhibitor 

of bacterial neutral protease. 

Lipases, nucleases, phosphatases (including phytase), oxidases, transglycosylases, - 

and -glucosidases are enzymes of potential interest. It is generally considered that the 

presence of lipases (fat hydrolysing enzymes) and lipoxygenase is undesirable. It has 

been proposed that the addition of `tanninases' to wort or beer should hazeproof the beer. 

How, or if, such enzymes might act on barley and hop proanthocyanidins is not clear, 

since these are not hydrolysable tannins. The enzyme -acetolactate decarboxylase may 

be added to beer to break down its substrate to carbon dioxide and acetoin. Thus by 

destroying aprecursor of diacetyl the flavour stability of the beer is improved (Section 

12.10.2). Glucose oxidase has been added to beer to `scavenge' oxygen, which is utilized 

to convert glucose to gluconolactone and hydrogen peroxide. This latter compound is 

itself an oxidizing agent and its presence is probably undesirable. It is degraded by the 

enzyme catalase to oxygen and water. The oxygen (half the amount initially used) is 

again used by the glucose oxidase, and so the amount present is progressively reduced. 



AASTRUP, S., BRANDT, J. and RIJS, P. (1991) Louvain Brewing Letters, 4 (1), 16. 

ALBINI, P. A., BRIGGS, D. E. and WADESON, A. (1987) J. Inst. Brewing, 93, 97. 

BAMFORTH, C. W. (1986) European Brewery Convention Monograph ± XI. E. B. C. ± Symposium on Wort 

Production, Maffliers. p. 149. 

BEMMENT, D. W. (1985) The Brewer, Dec., p. 457. 

BENTLEY, I. S. and WILLIAMS, E. C. (1996) in Industrial Enzymology (2nd edn, Godfrey, T. and West, S., 

eds), Macmillan Press, London. pp. 339±57. 

BLANCHFLOWER, A. J. and BRIGGS, D. E. (1991) J. Sci. Food Agric. 56, 103, 117, 129. 

BOIVIN, P. and MALANDA, M. (1998) Proc. 25th Conv. Inst. Brewing (Asia Pacific Sect.), Perth. p. 30. 

BOURNE, D. T. and WHEELER, R. E. (1982) J. Inst. Brewing, 88, 324. 

BRIGGS, D. E. (1978) Barley. Chapman & Hall, London. 612 pp. 

BRIGGS, D. E. (1987) in Cereals in a European Context (Morton, I. D. ed.). Ellis Horwood, Chichester. p. 

119. 

BRIGGS, D. E. (1998) Malts & Malting. Blackie Academic & Professional, London. 796 pp. 

BRIGGS, D. E. (2002) J. Inst. Brewing, 108, (4), 395. 

BRIGGS, D. E., HOUGH, J. S., STEVENS, R. and YOUNG, T. W. (1981) Malting & Brewing Science, Vol I, Malt 

& Sweet Wort (2nd edn). Chapman & Hall, London. 387 pp. 

BRIGGS, D. E., WADESON, A., STATHAM, R. and TAYLOR, J. F. (1986) J. Inst. Brewing, 92, 468. 

BRISSART, R., BRAÈ UNINGER, U., HAYDON, S., MORAND, R., PALMER, G., SAUVAGE, R. and SEWARD, B. (2000) 

European Brewery Convention Manual of Good Practice. Malting Technology. Fachverlag Hans 

Carl. NuÈrnberg. 224 pp. 

BROOKES, P. A. and PHILLISKIRK, G. (1987) Proc. 21st Congr. Eur. Brew. Conv., Madrid. p. 337. 

BUCKEE, G. K. (1997) J. Inst. Brewing, 103, 115. 

BYRNE, H. (1991) Brew. Distill. Internat., 22 (11), 24. 

BYRNE, H. and LETTERS, R. (1992) Proc. 22nd Conv. Inst. Brewing (Australia & New Zealand Sect.), 

Melbourne, p. 125. 

BYRNE, H., DONNELLY, M. F. and CARROLL, M. B. (1993) Proc. 4th Sci. Tech. Conf. Inst. Brewing (Central 

& Southern Africa Sect.), p. 13. 

CHANDRA, S., BOOER, C., PROUDLOVE, M. and JUPP, D. (1999) Proc. 27th Congr. Eur. Brew. Conv., 

Cannes. p. 501. 

COMLINE, P. (1999) Brew. Distill. Internat. Sept., p. 14. 

COPESTAKE, C. (1998) The Brewer, Feb., p. 79. 

DALE, C. J., YOUNG, T. W. and MAKINDE, A. (1989) J. Inst. Brewing, 95, 157. 

DONHAUSER, S., WEIDENEDER, A. and GEIGER, E. (1991) Brauwelt Internat., 141, (4), 294. 

GIBSON, G. (1989) in Cereal Science & Technology (Palmer, G. H. ed.). p. 279. The University Press, 

Aberdeen. 

GODFREY, T. and REICHELT, J. (eds) (1983) Industrial Enzymology. The application of enzymes in 

industry. Macmillan, Basingstoke. 582pp. 

GRETENHART, K. E. (1997) MBAA Tech. Quart., 34 (2), 102. 

GROMUS, J. (1988) Brauwelt Internat., (II), 150. 

HAIKARA, A., ULJAS, H. and SUURNAÈ KKI, A. (1993) Proc. 24th Congr. Eur. Brew. Conv., Oslo. p. 163. 

HYDE, W. R. and BROOKES, P. A. (1978) J. Inst. Brewing, 84, 167. 

JOHNSON, L. A. (1991) in Handbook of Cereal Science & Technology (Lorenz, K. J. and Kulp, K. eds). 

Marcel Dekker, New York. p. 55. 

KUNZE, W. (1996) Technology Malting and Brewing. (Internat. ed., Wainwright, T. transl.). VLB, Berlin. 

726 pp. 

LAITILA, A., SCHMEDDING, D., VAN GESTEL, M., VLEGELS, P. and HAIKARA, A. (1999) Proc. 27th Congr. 

Eur. Brew. Conv., Cannes, p. 559. 

LAWS, D. R. J., BAXTER, E. D. and CRESCENZI, A. M. (1986) European Brewery Convention Monograph- XI. 

E. B. C.-Symposium on Wort Production, Maffliers, p. 14. 

LETTERS, R. (1990) Louvain Brewing Letters, 4 (3/4) 12. 

LLOYD, W. J. W. (1986) J. Inst. Brewing, 92, 336. 

LLOYD, W. J. W. (1988a) Brewer, 74 (882), 147. 

LLOYD, W. J. W. (1988b) Brewers' Guard., 47, (5), 23. 

MARTIN, P. A. (1978) Brewers' Guard., Aug., p. 29. 

MATSER, A. M. and STEENEKEN, P. A. (1998) Cereal Chem., 75 (3), 189. 

MAULE, A. P. (1998) Ferment, 11 (1), 23. 

NARZISS, L. (1976) Die Bierbrauerei. Vol. I. Die Technologie der Malzbereitung. (6th edn). Ferdinand 

Enke Verlag, Stuttgart. 382 pp. 

NARZISS, L. (1991) Brauwelt Internat., (4), 284. 

NARZISS, L. (1992) Die Bierbrauerei, Vol. II. Die Technologie der WuÈrzebereitung. (7th edn). Ferdinand 

Enke Verlag, Stuttgart. 402 pp. 

REICHELT, J. R. (1983) in Industrial Enzymology. The application of enzymes in industry, (Godfrey, T. and 


Reichelt, J. R. eds), Macmillan, Basingstoke. 

RIESE, J. C. (1997) MBAA Tech Quart., 34 (2), 91. 

ROBINSON, N., AMANO, H. and MIZUNO, A. (2001) Eur. Brew. Conv. Monograph no. 31. Symposium on 

flavour and flavour stability. 

SCHWARZ, P. B., JONES, B. L. and STEFFENSON, B. J. (2002) J. Amer. Soc. Brew. Chem., 60 (3), 130. 

SEWARD, B. J. (1992) Ferment, 5 (4), 275. 

SOLE, S. M. (2000) Ferment, 13 (4), 25. 

SOUTH, J. B. (1992) MBAA Tech. Quart., 29, 20. 

STARS, A. C., SOUTH, J. B. and SMITH, N. A. (1993) Proc. 24th Congr. Eur. Brew. Conv., Oslo, p. 103. 

STOWELL, K. C. (1985) Proc. 20th Congr. Eur. Brew. Conv., Helsinki, p. 507. 

TAYLOR, D. G. and BOXALL, J. (1999) Ferment, 12 (6), 18. 

TENG, J., STUBITS, M. and LIN, E. (1983) Proc. 19th Congr. Eur. Brew. Conv., London, p. 47. 

VAAG, P., RIIS, P., KNUDSEN, A.-D., PEDERSEN, S. and MEILING, E. (1993) Proc. 24th Congr. Eur. Brew. 

Conv., Oslo, p. 155. 

WAINWRIGHT, T. (2003) Brewers' Guard., 132 (2), 20. 

WALKER, M. D., BOURNE, D. T. and WENN, R. V. (1997) Proc. 26th Congr. Eur. Brew. Conv., Maastricht, 

p. 191. 

WALKER, M. D. and WESTWOOD, K. T. (1991) J. Amer. Soc. Brew. Chem., 50, 4. 

WEBSTER, R. D. J. (1981) J. Inst. Brewing, 87, 52. 

WIEG, A. J. (1973) MBAA Tech. Quart., 10 (2), 7. 

WIEG, A. J. (1987) in Brewing Science, Vol. 3. (Pollock, J. R. A. ed.). p. 533. Academic Press, London. 


3 

Water, effluents and wastes 


Breweries use large amounts of water, (`liquor' in the UK). The actual amounts of water 

used ranging from three to (exceptionally) 30 times the volumes of beer produced. As 

beers usually have water contents of 91 98% (or even 89% in the cases of barley wines), 

and the amounts lost by evaporation and with by-products are relatively small it follows 

that large volumes of waste water are produced. Sometimes large volumes are produced 

because of operational inefficiencies but breweries operating in efficient but different 

ways, and with different product ranges, have substantially different water requirements. 

Apart from brewing, sparging and dilution liquors, water is used for a range of other 

purposes. These include cleaning the plant using manual or cleaning-in-place (CIP) 

systems, cooling, heating (either as hot water or after conversion into steam in a boiler), 

water to occupy the lines before and after running beer through them, for loading filter 

aids such as kieselguhr, for washing yeast and for slurrying and conveying away wastes 

as well as for washing beer containers such as tankers, kegs, casks and returnable bottles. 

The acquisition and treatment of liquor and the disposal of the brewery effluents are 

expensive processes and have long been studied. 

While water is the major component of beer the brewery takes in many other materials 

such as bottles and other packaging materials, malts, adjuncts and hops, and during the 

brewing and packaging processes `pollutants' and `wastes' are generated. These include 

broken glass, damaged cans, packaging materials such as cardboard and plastic, spent 

grains, spent hops, trub, tank bottoms, carbon dioxide, spilled or spoilt beer, wort, noise, 

odours, domestic wastes and heat. All these must be dealt with and, where possible, 

disposed of at a profit. This chapter is primarily concerned with the acquisition and 

preparation of water of the grades needed in the brewery and the disposal of the dirty 

water, or effluents. However, the treatments or actions needed to deal with some other 

wastes or by-products are discussed (Anon., 1988; Armitt, 1981; Bak et al., 2001; Benson 

et al., 1997; Comrie, 1967; Crispin, 1996; Eden, 1987; Eumann, 1999; Grant, 1995; 

Hackstaff, 1978; Harrison et al., 1963; Hartemann, 1988; Heron, 1989; Mailer et al., 

1989; Moll, 1979, 1995; Taylor, 1989; Theaker, 1988). 


3.2 Sources of water 

Thesourcesofavailablewatercanbeunderstoodwithreferencetothewatercycle.Water 

evaporates from the land, plants, fresh water and the sea. In time this forms clouds and 

precipitates as rain, snow or hail, falling back onto the land or into the sea. Of that falling 

ontothelandaproportionevaporates,somerunsoffassurfacewaterandsomepenetrates 

intothesoil.Thesurfacewatermaybecollectedinlakes,riversorbehinddamsandsobe 

available for use. Water from these sources is variously contaminated. Even rain-water is 

not pure, as it contains oxides of nitrogen and sulphur, dust, soot, pollen, microbes and 

industrial wastes. Collected on the ground it may be further contaminated with industrial 

and domestic effluents, spillages, drainings from dumps, rotting plant materials, farm 

animal wastes, leached agricultural materials (fertilizers, pesticides, and herbicides) and 

so on. The water which penetrates the soil is progressively filtered as it sinks downwards 

and so contains less of some surface-derived contaminants and micro-organisms. On the 

other hand salts may be dissolved from the pervious strata through which it passes. Thus 

surface waters will be comparatively `soft', i.e., will contain little in the way of dissolved 

salts, in contrast to waters recovered from underground, which may be either `soft' or 

`hard'. Water which passes through chalk or limestone becomes enriched with calcium 

bicarbonate, while in other areas it may contain calcium sulphate or salt. When the water 

meets an imperviouslayer the pervious layers above become saturated with water and are 

called aquifers. Water can be drawn from some of these. Near the sea the soil may be 

saturated with brine, with less dense fresh water layered above. In these areas fresh water 

must be withdrawn only slowly or the saline water table may be drawn up and brine will 

enter the well. Waters with very different characterstics may be available within one area 

(Rudin, 1976). 

Historically, different regions became famous for particular types of beer and in part 

these beer types were defined by the waters available for brewing (Table 3.1). Thus 

Pilsen, famous for very pale and delicate lagers has, like Melbourne, very soft water. 

Burton-on-Trent, with its extremely hard water, rich in calcium sulphate, is famous for its 

pale ales while Munich is well-known for its dark lagers, and Dublin (which has similar 

soft water) for its stouts. Breweries may receive water from different sources, which may 

be changed without warning. Water supplies may vary in their salt contents between day 

and night, from year to year and between seasons (Rudin, 1976; Byrne, 1990). It is now 

usual for breweries to adjust the composition of the water they use. In some few regions 

of the world saline water must be used, even sea water. In principle, several desalination 

methods might be used, but in practice it seems that purified water is obtained from sea 

water either by a highly thermally efficient distillation (Briggs et al., 1981), which is very 

costly, or by reverse osmosis (see below). Usually breweries obtain their water either 

from their own wells, springs or boreholes (surface waters are avoided where possible) or 

they may obtain them from water companies. 

Boreholes may extend downwards for 200 m (approx. 656 ft.), or more, and be fitted 

with an immersed pump to drive the water to the surface (Bak et al., 2001; Kunze, 1996). 

Water is drawn from an aquifer via a filter. A bore is sealed to prevent surface water or 

water from upper soil levels rapidly leaking down to the aquifer being used. While water 

from water companies is typically of a high standard of purity and is `potable', that is, it 

is fit for domestic use and is safe to drink, it is costly and is not necessarily fit to use in 

brewing (Baxter and Hughes, 2001). In addition, its composition and temperature are 

likely to vary and limits may be set on its use. Brewers' own water supplies will be more 

uniform, and will be substantially cheaper. However, there are likely to be charges for the 


ight to abstract the water and the volumes and rates of abstraction will probably be 

limited to avoid exhausting the available ground water or seriously disturbing the water 

table. 

Most regions have strict regulations, which must be met before water is classified as 

being potable, and these provide the minimum standards for brewing waters (Armitt, 

1981; Bak et al., 2001; Baxter and Hughes, 2001; Moll, 1979, 1995). These regulations 

are often reviewed, the upper permitted limits for specified substances are frequently 

reduced and the numbers of substances mentioned are increased. Table 3.2 indicates how 

complex these `minimum standards' can be. The requirements may be grouped as 

àesthetic' (colour, turbidity, odour and taste), microbiological standards (particularly the 

absence of pathogens), the levels of organic and inorganic materials that are in solution 

and the presence of radioactive materials. Some of these standards require comment. 

Drinking water must be safe, and so it must contain no pathogenic bacteria, protozoa, or 

viruses. However, the water is not necessarily sterile and so free of any organisms that 

can infect wort or beer, which must be the case for brewing water. The limits set for 

dissolved salts may be exceeded in some brewing waters. For example in Burton-on- 

Trent well waters the levels of calcium and sulphate ions may be very high (Table 3.1). 

Limitations on ammonia/ammonium and nitrogen levels are set since these are often 

indicators of contamination with decomposing organic matter. Nitrate levels, which vary 

widely, are a cause of concern as water sources are increasingly contaminated by nitrate 

from leached agricultural fertilizers. The fear is that during the preparation of the beer or 

in the consumer the nitrate may be reduced to nitrite (also limited, Table 3.2) and this, in 

turn, may give rise to carcinogens. The need to limit amounts of toxic ions is obvious 

although yeast needs trace amounts of many of them including copper, zinc, manganese 

and iron. These trace elements can be obtained from the brewers' grist. The minimum 

levels for total hardness and alkalinity are set to limit corrosion in pipework. Fluoride is 

often added to drinking water, but at the levels used it is harmless and without influence 

on fermentation. 

The organic contaminants mentioned (Table 3.2) deserve comment. Acrylamide, vinyl 

chloride and epichlorohydrin are toxic substances used in the manufacture of organic 

polymers and their presence indicates that unsafe disposal practices have been used. 

Aldrin, dieldrin, heptachlor and heptachlor epoxide are insecticides or their metabolites. 

In other countries limits on other substances, including selective herbicides such as 2, 4-D 

(2, 4-dichloro phenoxyacetic acid) and diquat may be specified. Some of the polycyclic 

aromatic hydrocarbons are carcinogenic and the trihalomethanes confer unwanted 

flavours and may be toxic. The trihalomethanes, THMs, are unfortunately named since 

the organic substances in this group are not all based on the methane carbon skeleton and 

not all are tri-substituted with halogens. Chlorine and bromine are the usual halogen 

substitutes (Cowan and Westhuysen, 1999; Grant, 1995; McGarrity, 1990; Taylor, 1989). 

THMs may be industrial solvent residues or they can arise from organic materials in the 

water when this is sterilized by chlorination. Thus they can be formed during water 

treatment in the brewery. 

Organic materials are particularly likely to be present in surface waters and may be 

dissolved or present as colloidal or suspended materials. Humic and fulvic acids are crude 

mixtures of organic materials with molecular weight ranges of 500 2,000,000 and 

200 1,000 respectively. These are particularly likely to give rise to THMs during 

chlorination. The bromine substituents can be added when the chlorinated water contains 

bromide ions. The composition of the THM group varies. It includes chloroform, 

bromomethane, carbon tetrachloride, 1, 1-dichloroethane, 1, 1, 2-trichloroethane and 


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 

(Moll, 1995; Mailer et al., 1989) 

Parameter Pilsen Burton-on-Trent MuÈnchen (Munich) Dortmund London Wien (Vienna) Melbourne 

Total dry solids 51 ± 1226 536 273 984 320 984 25 

Calcium (Ca 2+ ) 7.1 352 268 109 80 237 90 163 1.3 

Magnesium (Mg 2+ ) 3.4 24 62 21 19 26 4 68 0.8 

Bicarbonate (HCO3 ) 14 320 ± 171 ± 174 ± 243 ± 

Carbonate (CO3 2 ) ± ± 141 ± 164 ± 123 ± 3.6 

Sulphate (SO 4 2 ) 4.8 820 638 7.9 5 318 58 216 0.9 

Nitrate (NO3 ) tr. 18 31 53 3 46 3 tr. 0.2 

Chloride (Cl ) 5.0 16 36 36 1 53 18 39 6.5 

Sodium (Na + ) ± ± 30 ± 1 ± 24 ± 4.5 

tr. ˆ Traces. 

± ˆ Not given. 


Table 3.2 A list of the maximum (minimum) concentrations of substances that may not be 

exceeded in drinking water in the UK in 2001 (courtesy of J. MacDonald). Compare Bak et al., 

(2001); Baxter and Hughes (2001) 

Parameter Units Concentration or value 

Colour mg/l (Pt/Co scale) 20 

Turbidity Formazin units 1 

Odour Dilution number 3 at 25 ëC 

Taste Dilution number 3 at 25 ëC 

Temperature ëC 25 

pH (limits) pH units 6.5±10.0 

Conductivity S/cm at 20 ëC 2500 

Permanganate value O2, mg/l 5 

Total organic carbon, TDC C, mg/l no significant increase 

Total coliform bacteria number/100 ml 0 

Faecal coliform bacteria number/100 ml 0 

Faecal Streptococci, Enterococci number/100 ml 0 

Clostridium perfringens number/100 ml 0 

Sulphate reducing Clostridia number/20 ml 1 

Colony counts number/ml at 25 or 37 ëC no significant increase 

(In some regions tests are also 

carried out for Protozoa, such as 

Cryptosporidium and Guiardia) 

Radioactivity (total indicative dose) MSv/year 0.1 

Tritium Bq/l 100 

Boron B mg/l 1 

Chloride Cl, mg/l 250 

Calcium Ca, mg/l 250 

Total hardness Ca, mg/l 60 (minimum) 

Alkalinity HCO 3, mg/l 30 (minimum) 

Sulphate SO4, mg/l 250 

Magnesium Mg, mg/l 50 

Sodium Na, mg/l 200 

Potassium K, mg/l 12 

Dry residues (after 180 ëC) mg/l 1500 

Nitrate NO3, mg/l 50 

Nitrite NO2, mg/l 0.5 

Ammonia, ammonium ions NH 4, mg/l 0.5 

Kjeldahl nitrogen N, mg/l 1.0 

Dissolved or emulsified hydrocarbons 

Mineral oils g/l 10 

Benzene g/l 1 

Phenols C6H5OH, g/l 0.5 

Surfactants (detergents) as lauryl sulphate, g/l 200 

Aluminium Al, g/l 200 

Iron Fe, g/l 200 

Manganese Mn, g/l 50 

Copper Cu, mg/l 2 

Zinc Zn, mg/l 5 

Phosphate P, mg/l 2.2 

Fluoride F, mg/l 1.5 

Silver Ag, g/l 10 

Arsenic As, g/l 10 

Bromate BrO3, g/l 10 

Cadmium Cd, g/l 5 

Cyanide CN, g/l 50 

Chromium Cr, g/l 50 


Table 3.2 Continued 

Parameter Units Concentration or value. 

Mercury Hg, g/l 1 

Nickel Ni, g/l 20 

Lead Pb, g/l (will be reduced 

in 2013) 25 

Antimony Sb, g/l 5 

(Elsewhere limits are set on other substances, such as thallium, beryllium, uranium and asbestos) 

Acrylamide g/l 0.1 

Vinyl chloride g/l 0.5 

Epichlorohydrin g/l 0.1 

Aldrin g/l 0.03 

Dieldrin g/l 0.03 

Heptachlor g/l 0.03 

Heptochlorepoxide g/l 0.03 

Other pesticides g/l 0.1 

Pesticides, total g/l 0.5 

Polycyclic aromatic hydrocarbons* g/l 0.1 

Benzo(a)-3,4-pyrene ng/l 10 

1,2-Dichloroethane g/l 3 

Tetrachloromethane g/l 3 

Trichloroethane g/l 10 

Tetrachloroethane & trichloroethene g/l 10 

Trihalomethanes, total y g/l 100 

Substances extractable in chloroform mg/1, dry residue 1 

* Sum of individual concentrations of members of a list of substances benzo[b]fluoranthene, benzo[k]fluoranthene, 

benzo-11,12-fluoranthene, benzo[ghi]perylene and indeno-[1,2,3-cd]pyrene. 

y Sum of chloroform, bromoform, dibromochloromethane and dibromodichloromethane. 

tetrachloroethane together with arange of other substances. Some THMs are also VOCs, 

(volatile organic compounds). Their volatility is the basis of their partial or total removal 

during gas-stripping processes (as when removing carbon dioxide after dealkylation, or 

oxygen removal) or during mashing and in the copper boil. THMs are also removed by 

active carbon filtration and partly removed during reverse osmosis. Chlorination of 

aromatic, organic materials can give rise to other undesirable materials, including 

medicinally flavoured chlorophenols. 

3.3 Preliminary water treatments 

Most brewers find it necessary to treat the water coming into the brewery. The variety of 

substances that may occur in water is large, and different treatments are needed to deal 

with them (Fig. 3.1). Different waters require different treatments and brewers require 

grades of water treated in different ways depending on the uses to which it will be put. In 

some instances it may only be necessary to pre-treat the liquor, while in other cases 

extensive further treatment will be needed. Preliminary treatments may involve aeration, 

sedimentation (with or without the prior addition of coagulants and flocculating agents, 

which initially require vigorous stirring, followed by more gentle stirring to encourage 

the build up of flocs), flotation, filtration and sterilization. Some of these treatments may 

be used more than once (e.g. sterilization) during the preparation of liquors. Aeration 

with compressed air (with or without ozone) is used to oxidize ferrous ions to ferric 


Diameter (μm) 

Types of substance 

Water treatments 

10 –4 

10 –3 

10 –2 

10 –1 

1 10 10 2 

10 3 

Ions and molecules 

in true solution 

Macromolecules Microparticles Fine particles 

Salts 

Sugars 

Colloids 

Suspended solids 

Viruses Bacteria 

Pollen 

Sand 

Yeasts 

Biological uptake 

Chemical precipitation 

Particle filtration Screening 

Ion exchange 

Microfiltration 

Sedimentation 

Ultrafiltration 

Sand filtration and 

Nanofiltration 

microscreening 

Reverse osmosis 

Adsorption 

Coagulation and filtration 

Fig. 3.1 The sizes of dissolved, colloidal and suspended materials that must be considered in 

water purification and the methods used in removing them (Bak et al., 2001; Briggs et al., 1981). 

oxide/hydroxide (which separates from solution), to remove volatile organic substances, 

hydrogen sulphide, and carbon dioxide from water. Water is sprayed onto the top of a 

column filled with plastic packing, and flows downwards against a counter-current flow 

of air, which carries away the unwanted volatile substances. 

Measured amounts of ferric chloride or aluminium sulphate, with or without some 

organic polyelectrolyte, may be added to water to act as coagulants. The salts hydrolyse, 

giving rise to voluminous precipitates of hydrated ferric hydroxide or aluminium 

hydroxide. After thorough mixing the precipitates are allowed to settle, carrying down 

inorganic and some organic suspended matter. The comparatively clear supernatant is 

removed, leaving the sludge to be collected and dumped. Untreated water may also have 

a sedimentation treatment to allow the denser suspended materials to settle, or it may be 

filtered. If the water is rich in dissolved iron or manganese these should be removed. Iron 

in particular can deposit oxides as slime which blocks pipes and can clog filters. In the 

brewing process iron ions give colours with polyphenols and, probably acting as 

oxidation catalysts, promote flavour and haze instability. Aeration or treatment with 

oxidizing agents, such as chlorine, converts ferrous ions to ferric ions which then separate 

as ferric hydroxide. Oxidizing agents are also needed to convert manganese to a 

precipitable form. Sedimentation or flotation are generally used before filtration. To 

achieve sedimentation the water is passed into a large tank in which it moves quietly and 

slowly to allow the solids to precipitate or the water may pass through lamellar 

separators, or centrifuges or hydrocyclones (Bak et al., 2001). Alternatively, flocculated 

material may be removed by flotation in which finely divided bubbles of air rise from the 

base of a vessel and carry the flocs to the surface, where they accumulate and are 

removed by skimming. 

Often suspended materials are removed from water by coarse filtration. It may be 

passed through a bed of sharp, calcined sand that may be 2 3 m (6.6 10 ft.) thick or it 

may pass through successive layers of granular plastic (3 5 mm), anthracite (2 3 mm) 

and sand (0.5 1.5 mm). When the filter becomes blocked, as signalled by a rising 

resistance to the water flow, it is back-flushed with a reverse stream of water to carry 

away the blocking particles. In special BIRM filters the sand is mixed with manganese 

dioxide, which catalyses the oxidation, and so the precipitation, of ferrous ions as ferric 

hydroxide. A newer device is the fibrous depth filter. Fibres are firmly twisted together 

around a support to form a tube, creating an efficient filter, which is able to exclude more 


than 98% of particles over 2 m in diameter. To clean the filter the tension is reduced and 

the enlarged spaces between the fibres are cleaned with a back-wash. Cartridge filters, 

which exclude particles over 5 m, may also be used. 

Water in breweries may be sterilized more than once at different stages. Chemical 

sterilants are chlorine, hypochlorites, chlorine dioxide, ozone and, less often, silver. 

Physical sterilants used are exposure to ultraviolet light, sterilizing filtration and, rarely 

(except during the hop-boil), heat. Chlorine, used as the green-yellow gas or as sodium, 

potassium, or calcium hypochlorites, is a commonly used sterilant. One recommendation 

is that the level of available chlorine should initially be 5 mg/l that the water should be 

held at least for 30 min., to allow the sterilant to act and at this time the level of free 

chlorine should not have fallen below 1 mg/l. This recommendation emphasizes that 

chlorine is a highly reactive compound and a strong oxidizing agent that is used up during 

water sterilization but that it remains in solution long enough to have a useful `residual' 

sterilizing effect. Disadvantages of using chlorine include the formation of chlorophenols 

and THMs from organic substances in the water. Unwanted residual chlorine may be 

removed by aeration, evaporation, by filtration through active carbon, or by adding 

bisulphite or sulphite to the liquor, when the chlorine is reduced to chloride ions while the 

sulphite is oxidized to sulphate. If the water contains ferrous ions chlorine will oxidize 

them to ferric ions, which will then form flocculent ferric hydroxide. 

Chlorine dioxide, ClO2, an unstable, yellow, explosive gas that is generated on site 

immediately before use, from hydrochloric acid and sodium chlorite: 

5NaClO4 ‡ 4HCl ! 4ClO2 ‡ 5NaCl ‡ 2H2O 

It is a strong oxidizing agent. Unlike chlorine, it does not chlorinate organic substances 

and so does not give rise to THMs or unwanted flavour compounds. Indeed it destroys the 

off-flavours given by some chlorophenols. Its `residuals' last for a shorter period than 

those of chlorine and so this agent is less effective at preventing re-infection. A contact 

time of 15 min. is desirable. Ozone, O 3, is formed on site by passing dry air or oxygen 

through an electrical generator. This gas is a strong oxidizing agent, but its lifetime is 

short and so it gives almost no residual protection against re-infection. This agent is said 

to be more effective against Giardia, cysts of other protozoa and some viruses and 

bacteria than chlorine or chlorine dioxide. It is also effective at destroying some taints 

and odours. Treated water should initially contain 1 3 g ozone/m 3 . Treatment should be 

extended from 3 to 15 min. and the higher doses should be used if iron and/or manganese 

ions are to be oxidized. Ozone is toxic and should be degraded before waste gases are 

vented to the atmosphere. Silver ions, generated by the electrolytic `katadyn process' are 

also effective sterilants under some conditions but their use is not permitted everywhere 

and impurities in the water can reduce their effectiveness. All the sterilants mentioned 

must be handled with care, as they can be dangerous. 

Sterilization with ultraviolet light, UV, relies on the fact that emissions at wavelengths 

around 260 nm are absorbed by nucleic acids, which are then disrupted. Thus UV light from 

low-pressure mercury lamps is able to kill microbes, including viruses, but of course the 

treatment leaves no sterilizing residue. The long tubular lamps are housed in quartz tubes 

and the water flows past in a tubular metal housing which limits the length of the UV light 

path. The water must be clear and colourless and not give deposits to avoid blocking the 

UV radiation. The dwell time in the radiation chamber must be sufficient for sterilization to 

be complete. UV treatment of water containing dissolved ozone is more effective than 

either agent alone, and chlorinated hydrocarbons are fully oxidized, to carbon dioxide and 


hydrochloric acid, and the COD of the water is reduced. The lamp tubes must be checked 

regularly and must be replaced as they approach the end of their working lives. Personnel 

must not be exposed to this radiation. Bacteria and fungi (but not viruses) can be removed 

by sterile filtration through special membranes (wound membranes or hollow fibres) 

having, for example, notional pore sizes of 0.2 or 0.45 m. Such membranes can easily be 

blocked and so the water must be free of components that can deposit sludge or scale or 

contain fine suspended matter therefore the water to be sterilized must be pre-treated and 

carefully filtered. Pasteurization is rarely applied to water, but is used on some beers 

(Chapter 21). Other treatments, such as flocculation or reverse osmosis, deplete or remove 

microbes, but these processes are primarily used for other purposes. 

3.4 Secondary water treatments 

Waterusedinbreweriesisusuallytreatedtoadjustitscomposition(Baketal.,2001;Benson 

etal.,1997;Blackmann,1998;Comrie,1967;Maileretal.,1989;Moll,1995;Taylor,1989). 

Treatments may reduce levels of organic compounds in solution or adjust the ionic 

compositionoftheliquor.Inthepastthissubjectwasconfusedbywidelydifferingmethods 

of expressing salt concentrations (Moll, 1979,1995; Appendix). Here units of mg ion/l will 

be used. Ions in beer can influence its flavour (see below) and calcium ions in particular 

influence the mashing process (Chapter 4). Discussions of water composition often involve 

the term `hardness'. `Soft' water contains low concentrations of dissolved salts, particularly 

salts of calcium and (with less emphasis) salts of magnesium. `Hard' water contains high 

concentrations of salts, usually mainly calcium bicarbonate or calcium sulphate. `Temporary 

hardness' is caused chiefly by calcium bicarbonate and is so-called because if the water is 

boiled the bicarbonate is converted to the carbonate, which precipitates leaving the clarified 

water `softened'. In contrast `permanent hardness' is mainly caused by calcium sulphate, 

and this remains in solution when the water is boiled. The distinction is important if the 

liquor is to be used for mashing or, even more, for sparging. 

While temporary hardness can be removed by boiling water, this process is costly and 

is usually avoided although it may be beneficial in other ways, such as sterilizing the 

water, driving out the dissolved oxygen and evaporating volatile contaminants such as 

THMs. The decomposition of the bicarbonate occurs as: 

Ca(HCO 3† 2 ! CaCO3 # ‡ H2O+CO2 " 

Magnesium bicarbonate is also decomposed by boiling, but magnesium carbonate is 

appreciably soluble, and so its removal is incomplete. The calcium carbonate precipitates 

and the carbon dioxide is driven off. Boiling also accelerates the oxidation of ferrous ions 

to ferric ions, which precipitate as the hydroxide. Treatments with lime water may be 

used to remove temporary hardness from water. A calculated amount of lime-water, or a 

slurry of lime in water, is mixed with the water. Calcium carbonate is precipitated: 

Ca…HCO3† 2 ‡ Ca…OH† 2 ! 2CaCO3 # ‡ 2H2O 

CO2 ‡ Ca…OH† 2 ! CaCO3 # ‡ H2O 

In older plant the precipitate of calcium carbonate settles slowly, but in a more modern 

and fully automated plant the calcium carbonate is deposited on crystalline granules of 

the same material 0.1 2.5 mm in diameter. In either case residual suspended calcium 

carbonate is removed, for example by sand filtration. The calcium carbonate is used in 


agriculture, spread on fields to reduce soil acidity. After the lime treatment the water is 

alkaline and, for brewing purposes, must be adjusted to about pH 7. In Germany, and 

elsewhere where the use of mineral acids is forbidden, this is achieved by adding carbon 

dioxide. However, where it is permitted, the alkalinity is reduced by additions of food 

grade acids, commonly mineral acids but sometimes lactic acid. When the water contains 

an appreciable amount of magnesium bicarbonate it may receive a `split treatment'. A 

portion of the water is dosed with a high level of lime. Calcium carbonate precipitates and 

magnesium precipitates as the hydroxide: 

Mg…HCO3† 2 ‡ 2Ca…OH† 2 ! Mg…OH† 2 # ‡2CaCO3 # ‡H2O 

This partly treated water is strongly alkaline, with a pH of about 12. It is mixed with the 

remainder of the water (about two-thirds of the amount being treated), precipitating the 

calcium bicarbonate as the carbonate. After clarification the pH of the water is adjusted. 

Thus calcium bicarbonate and some of the magnesium salts are removed. Both the lime 

water treatments also precipitate iron and manganese ions, as hydroxides, and the 

precipitates entrain and remove some organic contaminants. Another way of removing 

bicarbonate ions from solution is to acidify the water and then remove the carbon dioxide 

formed with aeration. Thus: 

Ca…HCO3† 2 ‡ H2SO4 ! CaSO4 ‡ 2 H2CO3 

H2CO3 ! H2O ‡ CO2 " 

The food-grade acid used depends on flavour, safety and operational considerations. 

After acidification the water is passed down a packed tower against an upward stream of 

air that carries away the carbon dioxide. Incidentally, it also removes some volatile 

organic compounds and chlorine, if these are present. 

Several types of ion exchange treatments may be applied to brewing waters. Modern 

ion exchange resins are now used rather than the old mineral ion exchangers, such as 

zeolites. The resins are beads of varying porosities, often of cross-linked polystyrene, 

which carry acidic or basic groups. Ion exchange treatments may be fully automated. 

Resins must be free of flavoured, low molecular weight organic materials, and they must 

not be exposed to chlorine, which will attack them. Iron and manganese must have been 

removed from the feed water and this is carefully filtered to prevent the resin beds 

becoming blocked. The costs of ion exchange treatments include the costs of regenerating 

the resins and of disposing of the regeneration liquid chemical wastes. The treatments 

may be divided into dealkalization, softening and demineralization. In dealkalization, 

which removes temporary hardness, the water is passed through a packed column of a 

weakly acidic cation exchange resin, which carries carboxylic acid groups. This resin 

exchanges hydrogen ions for calcium and magnesium ions in the water. The hydrogen 

ions combine with bicarbonate ions in the water forming carbonic acid and this then 

dissociates reversibly to carbon dioxide and water: 

Res 2 2H 2 ‡ Ca 2‡ ! Res 2 Ca 2‡ ‡ 2H ‡ 

H ‡ ‡ HCO3 $ H2CO3 $ CO2 " ‡ H2O 

The water is then passed down an aeration tower where the carbon dioxide is removed 

together with some volatile organic compounds, VOCs, and chlorine. 

After ion exchange treatment water is often passed through active carbon filters as a 

precaution to remove any unwanted off-flavoured compounds that may be released from 


the resins. In time the exchange capacity of the resin is exhausted and it has to be 

regenerated. Both the resin and the acid used to regenerate it are comparatively 

inexpensive and indeed this treatment may be used before demineralization to reduce the 

cost of this latter process. The waste regeneration liquid is acidic. Water softening can be 

carried out by adding sodium carbonate to water containing, for example, calcium 

sulphate. Calcium carbonate precipitates leaving the more soluble sodium sulphate in 

solution. More often softening is carried out by ion exchange. A strongly acidic ion 

exchange resin, carrying sulphonic acid groups, is loaded with sodium ions. When the 

water passes through the resin this exchanges sodium ions for the more strongly bound 

divalent metal ions, like those of calcium and magnesium. The softened water is used 

where the use of hard water might give rise to scales and deposits of sludge, for example 

in cooling water, in boilers and in rinsing water. 

Demineralization involves treating the water with strongly acidic and strongly basic 

resins loaded with hydrogen and hydroxyl ions respectively. The water may go through 

two resin beds working in series or mixed bed resins may be used. Aeration may be used, 

either after passage through the strongly acidic resin or after the entire treatment, to 

remove liberated carbon dioxide. Mixed bed resins are returned to the makers for 

regeneration. Using this system all the positively charged ions in the water are exchanged 

for hydrogen ions and all the negatively charged ions are exchanged for hydroxyl ions. 

For example: 

Res 2‡ 2OH ‡ SO4 2 ! Res 2‡ SO4 2 ‡ 2OH 

Res 2 2H ‡ ‡ Ca 2‡ ! Res 2 Ca 2‡ ‡ 2H ‡ 

H ‡ ‡ OH $ H2O 

The hydrogen and hydroxyl ions combine to give water. Demineralization can give very 

pure water. Very strong basic resins can even remove silicate ions and some organic acids 

and the resins can at least partly remove some herbicides and their breakdown products. 

Temporary and permanent hardnesses are removed and so are all ions, including nitrate. 

It is now commonplace for brewing water to be demineralized and then for the 

compositions of the process streams to be adjusted to meet the different process 

requirements. This is convenient, since fluctuations in the composition of the incoming 

water become irrelevant and different brewing liquors can be prepared as needed for 

different beers. The processes of demineralization and reverse osmosis are in direct 

competition. If the levels of the total dissolved salts are comparatively low (TDS < 1,000 

mg/l) then demineralization is likely to be chosen, despite the cost of regenerating the 

resins. Typically some 10 15% of the incoming water is used in the regeneration 

processes and goes to waste. 

Reverse osmosis (RO) is attractive if the water to be purified is high in total dissolved 

solids (> 1,000 mg/l; Benson et al., 1997; Berkmortel, 1988a. 1988b; McGarrity, 1990; 

Thompson, 1995). There have been great advances in the technology of making 

membrane units either in the form of hollow fibres or as spirally wound sheets. The semipermeable 

membranes used in reverse osmosis are permeable to water but they are 

impermeable to microbes, ions and organic substances of molecular weight > 200. If pure 

water is separated from a salt solution by a semi-permeable membrane (i.e. permeable to 

water but not to solutes) there is a net migration of water through the membrane into the 

salt solution. If the pressure on the salt solution is increased then at a particular value it 

can balance the osmotic pressure, the tendency of the water to migrate into the salt 

solution, and no net movement of water will occur. If the pressure on the salt solution is 


increased still more water will be driven through the membrane from the salt solution, 

which is concentrated, and the permeate will be substantially pure water. High-pressure 

pumpsareneededtodrivethisprocess.Inanextremecaseapressureof25bar(367.5psi) 

is needed to desalinate sea water with aTDS of 35,000mg/l. This extreme degree of 

desalination may be carried out in two steps. The water under pressure flows across the 

membranes and about 75% is recovered as purified permeate and 25% as concentrated 

`saline'. In many cases this `saline water' is still of use for hosing down, etc. 

InalessextremecaseafeedwaterwithaTDSof1200mg/ltreatedbyreverseosmosis 

(RO) gave risetoapermeate with 1 8mg/l TDS.RO plant hasno regenerationcostsand 

can be cleaned automatically (CIP). On the other hand many of the membranes must be 

protected from chlorine and, to prevent membrane blockages, the feed water must be free 

of suspended solids, manganese or iron salts or materials that can form scales or sludges. 

Therefore the water may need pre-treatment and it must be filtered, probably through a 

sand filter and then through afine filter removing particles >10 mdiameter (Braun and 

Niefind, 1988). Filtration also protects the high-pressure pumps from damage byabrasive 

particles. As the salt concentration of the water increases so does the cost of treatment, 

but not to aproportional extent. Sets of membranes are expected to last about five years 

therefore atreatment could involve filtration through aBIRM filter, acidification, the 

removal of carbon dioxide in an aeration tower, very fine filtration and then RO. Reverse 

osmosis is now in widespread use. In contrast, electrodialysis, acompeting technology, 

seems not to have found favour. 

Brewing water is often passed through layers of active carbon. This `carbon filtration' 

is used to remove residual chlorine, humic and fulvic acids, many aromatic organic 

substances, some pesticides, some THMs and phenolic substances, and unwanted 

coloured, flavoured and odorous materials. Charcoals from different sources differ in 

their adsorbtive capacities and the types of substances that they remove best (Gough, 

1995). Bituminous coal, anthracite or coconut shells, as examples, are pyrolysed, giving 

products that are predominantly microcrystalline graphite. These are then àctivated' by 

one of several methods. The material chosen for use must have the correct particle sizes, 

be strong enough to resist some wear, be of a `food grade', and have the correct 

adsorbtive characteristics. Acharge for afilter should last for five to seven years. The 

liquorreachingthefiltershouldbesterileandwellfilteredandhavebeentreatedsothatit 

does not give deposits. With the passage of time the filter will tend to become blocked 

and asource of microbiological infection. In addition, its adsorbtive capacity will tend to 

become saturated and so the charcoal must be cleaned, sterilized and regenerated. As a 

routine, carbon filters are backwashed with chlorinated water and then drained and 

steamed. Sterilization of the liquor coming from acarbon filter must be by atechnique 

that leaves no residues. Consequently UV radiation is often used or, less often, ozone. 

Brewersareincreasinglyconcernedtoexcludeair,orrathertheoxygenintheair,from 

their beers and from the production stream. To help to achieve this several methods for 

deoxygenating water are in use (Andersson and Norman, 1997; Benson et al., 1997; 

Cleather, 1992; Kunze, 1996). In some instances the carbon dioxide and/or nitrogen 

levels of the liquor are adjusted as the oxygen is removed. As the temperature of water 

rises so the amount of oxygen that it will hold in solution declines (Table A12 on page 

844) therefore water can be at least partly deaerated by boiling or stripping with (clean) 

steam. This approach is costly. Another method is to heat the water to 85 ëC (185 ëF) or 

more and then spray it into a vacuum chamber to remove the dissolved gas. More than 

one treatment may be needed. Another approach is to saturate, under pressure, the water 

to be deoxygenated with carbon dioxide or nitrogen. The water is then transferred into a 


low-pressure chamber when the gas bursts out of solution carrying the dissolved oxygen 

with it. The chamber may be ventilated with the carrier gas. Merely bubbling carbon 

dioxide or nitrogen through the water is not sufficient. 

Several methods seem to be preferable to those described; water can be sprayed into 

the top of atower packed with aplastic filler. As the water trickles down as athin film it 

meets an upflow of the stripping gas, carbon dioxide or nitrogen. The stripping gas may 

be sterile filtered. The liquor loses its oxygen to the stripping gas, which removes it 

efficiently as the system works on the counter-current principle. An alternative approach 

is to pass the water across ahydrophobic but gas-permeable membrane (hollow fibre or 

spirally wound format. Brown et al., 1999). The non-water side of the membrane may be 

avacuum or aflowing stripping gas. The water loses its oxygen through the membrane. 

Several of the methods mentioned ensure that the water is saturated with the chosen 

carrier gas. Another method for reducing dissolved oxygen, DO, levels is its catalytic 

reduction with hydrogen. The DO level of astream of flowing water is sensed and the 

appropriate amount of hydrogen is dosed into the water stream and is dissolved under 

pressure. After thorough mixing the water passes over apalladium catalyst (supported on 

beads of an ion exchange resin) and the oxygen is reduced to water. Using this technique 

the oxygen content of the water can be reduced to as little as 0.002mgO2/l. In some 

instances, e.g., in some boiler waters, sulphite salts may be added to react with, and so 

`scavenge', oxygen. The sulphite is oxidized by the oxygen, becoming sulphate. 

3.5 Grades of water used in breweries 

The mixture of dissolved substances judged suitable in liquor used for mashing may 

differ from those present in sparge liquor or dilution liquor and will certainly differ from 

those preferred in cleaning or boiler waters. Mashing liquor may not be completely 

sterile, but its microbial count must be low. Increasingly, brewers employing newer types 

of plant will mash with oxygen-reduced or oxygen-free water and under conditions such 

that oxygen pick-up is minimal. The mixture of salts present in the liquor may have been 

supplemented or adjusted. The addition of calcium sulphate and/or chloride, `Burtonization', 

is common and when demineralized or reverse osmosis processed water is used as 

the base allthe saltspresent willhave been added. Thesaltsused andtheir concentrations 

are decided with reference to their functions in mashing (Ch. 4; cf. Table 3.1) and their 

flavours. To reduce the pH of a malt mash by 0.1 unit requires the addition of 300 g 

calcium sulphate or 250 g calcium chloride/100 kg malt, the salts being added in the 

brewing and sparging liquor (Comrie, 1967). 

The flavours of chloride and sulphate ions are different. It is recommended that brewing 

liquors for Burton style pale ales should have a sulphate to chloride ratio of 2:1 to 3:1. For 

mild ales the concentration of calcium should be less and the ratio should be about 2:3, 

while liquor for stouts should contain little or no sulphate. Sparge liquors may resemble 

mashing liquors, but it is desirable that the bicarbonate levels are very low, otherwise there 

is an undesirable rise in the pH of the last runnings as the buffering substances are leached 

from the mash. Deaerated and sterile water is required for pre-run and chase water 

preceding and following beer through pipework and for carrying slurried kieselguhr when 

forming the pre-coat on filters. Water that is sterile, deoxygenated, correctly carbonated and 

has the correct ionic composition and pH is used to dilute `high gravity' beers to their final 

strengths. Sterile water is also used to slurry and wash yeast. When water is to be heated 

during use, as in cooling water or in the pasteurizer, it needs to have been softened, 


demineralized or otherwise treated to prevent the deposition of sludges and scale, which 

can cover surfaces and interfere with heat exchange and may even block pipework. In 

addition, carbon dioxide and oxygen should be removed to minimize corrosion and 

antimicrobial agents may be added. In some instances the pH of heating water is adjusted 

with phosphate salts and scale-softening agents, such as tannins, may be added. 

While some high-pressure boilers require asupply of fully de-ionized, oxygen-free 

water, low-pressure boilers may operate with softened water sometimes dosed with 

chelating agents, such as EDTA or polyphosphates, to prevent the deposition of calcium 

salts on the heat-exchange surfaces (Ibbotson, 1986). Sludges are removed by `blowingdown', 

that is, ejecting them from the boiler to waste. Water being cooled in cooling 

towers should be treated with biocides to check the build up of populations of 

undesirable organisms, including beer-spoilage organisms and Legionella. For cleaning 

in vessels, pipework, bottles, kegs, etc., the water used should be sterile and it may 

contain traces of sterilant (e.g. ClO2). It must not leave deposits after draining. Water 

used for general cleaning, but that does not come into contact with the microbe-free 

surfaces that will contact wort or beer, can be of alower quality and need not be sterile. 

Clearly, supplying liquors of the correct grades for different uses around abrewery can 

be acomparatively complex process. The objective must be to obtain supplies of the 

various grades as simply and inexpensively as possible. The ways in which this is 

achieved are very varied. 

3.6 The effects of ions on the brewing process 

Ions present in brewing water have arange of effects on the production process and the 

quality of the product (Bak et al., 2001; Comrie, 1967; Moll, 1995; Taylor, 1981,1989). 

Inthis section the roles of major ions willbe consideredinturn. It willbe understoodthat 

other ions are added to the process stream from the grist and from the hops. In addition 

solid salts may be added directly to the mash or to the wort. Calcium ions (Ca 2+ ,at. wt. 

40.08) serve several important functions in brewing. They stabilize the enzyme - 

amylase during mashing and, by interacting with phosphate, phytate, peptides and 

proteins in the mash and during the copper boil, the pH values of the mash and the wort 

are usefully reduced. For example: 

or 

3Ca 2‡ ‡2HPO4 2 ‡2OH !Ca3…PO4† 2 #‡2H2O 

3Ca 2‡ ‡2HPO4 2 !Ca3…PO4† 2 #‡2H ‡ 

If bicarbonate ions are also present (the water has temporary hardness) these can more 

than offset the effect of calcium and cause a rise in pH (Chapter 4). Perhaps the 

concentration of calcium ions should not greatly exceed 100 mg/l in the mashing liquor as 

no great advantage is gained from higher doses and there is the risk that too much 

phosphate may be removed from the wort, and the yeast may then have an inadequate 

supply. Another recommendation is that calcium should be in the range 20 150 mg/l, 

depending on the beer being made. 

Calcium oxalate, Ca(COO)2, is deposited as beer stone during fermentations, and an 

adequate level of calcium ions ensures that the deposition is nearly complete. Crystals of 

calcium oxalate formed later in packaged beer provide nuclei for the breakout of carbon 


dioxide and so can cause gushing and haze. In mashing the fall in pH caused by calcium 

ions favours proteolysis and so an increase in FAN, and faster saccharification. The more 

acid conditions also reduce wort colour, hop utilization and favour a reduction in 

astringent flavours. Calcium ions favour the formation of agood, flocculent hot break 

(trub) and yeast flocculation, but they seem to have little effect on flavour. 

Magnesium ions (Mg 2+ ,at. wt. 24.32) are needed by many yeast enzymes, such as 

pyruvate decarboxylase. In some respects the effects of this ion resemble those of the 

calciumion, but the effects onpH from interactions with phosphates are less pronounced, 

being about half, because the salts are more soluble. While high concentrations of 

magnesium ions are unusual, they can impart asour or bitter flavour to beer. High, 

laxative concentrations are not reached. An upper limit of 30mg magnesium ions/litre 

has been proposed. 

Sodium ions (Na + ,at. wt. 23.0) occur in some waters and sodium chloride is the main 

solute in saline waters. Sodium ions can impart sour/salty flavours at high concentrations 

(over about 150 mg/litre, which is also aproposed maximum concentration) and sodium 

chloride may be added tobrewing liquors (75 150mg/l) to enhance `palate-fullness' and 

a certain sweetness. Sometimes potassium chloride is added instead, at low 

concentrations, to achieve a less sour flavour. Excess potassium ions ((K + , at. wt. 

39.1) >10mg/l) can have laxative effects and impart asalty taste. 

Hydrogen ions (H +, at. wt. 1.01) and hydroxyl ions (OH ,at. wt. 17.01) are always 

present in water, which is neutral when these ions are present in equimolecular amounts, 

[H + ] = [OH ]. The negative log10 of the hydrogen ion concentration, expressed in 

molarity, is the pH. As the temperature rises the dissociation of the water increases, the 

hydrogenionconcentrationincreases,andsothepHofwateratneutrality declines (Table 

A8 on page 842). 

Iron ions (Fe 2+ , ferrous and Fe 3+ , ferric; at. wt. 55.9) can occur in solution, for 

example, as ferrous bicarbonate or complexed with organic materials. Ferrous water is 

undesirable for brewing purposes, since it can deposit slimes (probably after oxidation, as 

red-brown hydrated ferric hydroxide), which can block pipes, filters, ion exchange 

columns, reverse osmosis equipment, etc. In addition, iron ions can confer dark colours to 

worts and beers by interacting with phenolic substances from the malt and hops and can 

convey metallic, astringent tastes to beers, give hazy worts and inhibit yeasts. The ions, 

possibly because of their ability to act as oxidation/reduction catalysts, favour haze 

formation and flavour instability. At concentrations of > 1 mg/l iron ions are harmful to 

yeasts. Perhaps concentrations should be reduced to less than 0.1 mg Fe/l. For all these 

reasons, and because of the difficulties that they can cause in some water treatments, it is 

usual to reduce the levels of dissolved iron early in a water treatment process. 

Copper (Cu 2+ , at. wt. 63.5) presented problems in brewing when vessels and pipework 

were made of copper but since these have come to be made of stainless steel there have 

been fewer problems with dissolved copper in breweries. Copper ions are toxic and 

mutagenic to yeasts, which accumulate them and develop `yeast weakness'. Another 

source of copper ions was the older, copper-based fungicides applied to hops. Copper 

ions are oxidation/reduction catalysts and their presence favours flavour instability and 

haze formation in beer. Brewing liquor should contain < 0.1 mg copper/litre. 

Manganese (Mn 2+ , manganous; Mn 4+ , manganic; at. wt. 54.9) levels in brewing water 

should be low, (< 0.2 mg/litre or even < 0.05 mg/litre) but trace amounts of this element 

(and copper and iron) are needed by yeast. Like copper and iron these ions are oxidation/ 

reduction catalysts and have adverse effects on flavour and beer colloidal stability. 

Manganese is less easily removed from water than iron. 


Ammonia(NH3,m.wt.17.03)andammoniumions(NH4 + ,m.wt.18.04.pKaˆ9.25at 

25ëC) occurs almost entirely as ammonium ions under brewing conditions. Ammonia in 

water indicates that the water may be contaminated with rotting organic matter and a 

proposed maximum concentration is 0.5mg/litre. However, ammonia can escape from 

refrigeration equipment and this, very water-soluble, gas is toxic. 

Zinc (Zn 2+ ,at. wt. 65.4), if present in appreciable amounts in brewing water, usually 

inicates thatthision hasbeen pickedupduringtransferorstorage.Highconcentrations in 

ground watersare unusual.Athigh levels this substancecan betoxic, the upper permitted 

concentration in potable water is 5mg/l (Table 3.2). High concentrations are damaging to 

yeasts but small amounts are essential. Not infrequently the levels of zinc in worts are 

insufficient to maintain good fermentations and in these cases the worts may be 

supplemented with additions of zinc chloride (0.15 0.2 mg/l). The recommended range 

in brewing liquor is 0.15 0.5 mg/litre. 

Bicarbonate (HCO3 , m. wt. 61.02) and carbonate ions (CO3 2 , m. wt. 60.01). The 

stages of the ionization of carbonic acid, formed by the hydration of carbon dioxide, are: 

CO2 ‡ H2O $ H2CO3 $ H ‡ ‡ HCO3 $ 2H ‡ ‡ CO3 2 

The pKa values of the first and second dissociations, at 25 ëC, are 6.4 and 10.3 

respectively. Thus at brewing pH values carbon dioxide is present as the gas, as carbonic 

acid and as the bicarbonate ion. High levels of bicarbonate ions in brewing water are 

undesirable since they cause unwanted increases in pH during mashing and sparging and 

in the hop-boil. Probably the concentration of bicarbonate ions in brewing liquor should 

never exceed 50 mg/l. 

Sulphate ions (SO4 2 , m. wt. 96.07); sulphate is the major counter ion to calcium and 

magnesium ions in permanently hard water. The ion contributes a drier, more bitter 

flavour to beers that should be balanced by appropriate amounts of chloride ions. Yeasts 

metabolize sulphate producing, inter alia, small amounts of hydrogen sulphide, (H 2S), 

sulphur dioxide, (SO 2), and other substances that contribute to the aromas of beers 

brewed with sulphate-rich water. The classic example is the `Burton nose' of the ales 

brewed at Burton-upon-Trent. Acceptable sulphate concentrations are in the range 

10 250 mg/litre. 

Chloride ions (Cl , at. wt. 35.5) occur at high levels in saline waters. High levels are 

reported to limit yeast flocculation but to improve beer clarification and colloidal 

stability. Chloride ions contribute to the mellow, palate-full character of beer. The ratio of 

chloride to sulphate helps to regulate the saline/bitter character of beer. Ratios and 

concentrations for different types of beers have been proposed (see above; Comrie, 1967). 

A reasonable maximum concentration is 150 mg/litre. 

Nitrate (NO3 , m. wt. 62.01) and nitrite ions (NO2 , m. wt. 46.01); there is concern 

about the rising levels of nitrate ions being found in ground waters. These ions are 

derived from agricultural fertilizers being leached from the topsoil and filtering down to 

the aquifers that supply water. Other sources are sewage and rotting organic matter. Even 

if the nitrogen is initially present as ammonium ions these are quickly oxidized to nitrate 

in the soil. Potable water usually has an upper limit of 50 mg nitrate/l and a limit of 0.1 or 

0.5 mg nitrite/l, and these limits must not be exceeded in beers. However, brewers require 

lower levels in their brewing water since nitrate will be added to wort from the hops. The 

concern is that bacterial contaminants may reduce nitrate to nitrite. Nitrite ions can be 

toxic, they can give colours with tannins and, of most concern, can give rise to potentially 

carcinogenic nitrosamines. 


Phosphate ions (e.g. HPO3 2 , m. wt. 80.01); the stages of ionization of phosphoric 

acid are: 

H3PO4 $ H ‡ ‡ H2PO4 $ 2H ‡ ‡ HPO4 2 $ 3H ‡ ‡ PO4 3 

The pK a values for the successive dissociations, at 25 ëC, are 2.0 2.2, 6.7 6.8 and 12.4. 

Thus at mashing, wort and beer pH values, around pH5, most phosphate is present as the 

H2PO4 ion. There are regulatory limits on the concentration of phosphate phosphorus 

that may be present in potable water, and a suggested maximum in brewing liquor is 1 

mg/litre. Most of the phosphate in beer is derived from malt, although phosphoric acid or 

acid phosphate salts may be used to adjust the pH or to release carbon dioxide from 

bicarbonate-rich waters. Worts can also contain the organic phosphate phytate (salts of 

phytic acid), derived from malt. Phosphates are important pH buffers in brewing and 

interactions between calcium ions and phosphates, and other substances, usefully reduce 

the pH in mashing and during the hop-boil. Phosphoric acid is also used for acid-washing 

yeasts. 

Silicate ions (e.g., SiO3 2 , , m. wt. 76.09); silica can dissolve to form a range of ions, 

with various silica to oxygen ratios. Reportedly high concentrations of silicates can 

damage yeasts and give rise to hazes in beers but if so, these events must be infrequent. 

The most significant effect of silicates is the deposition of scales, formed between silicate 

and calcium and magnesium ions, when the water is heated. An upper acceptable 

concentration of silicate may be 40 mg/litre. 

Fluoride ions (F , at. wt. 19.00) occur in some ground waters. They rarely reach toxic 

levels. Potable waters have upper concentration limits of about 1.5 mg/litre. Small 

amounts of fluorides may be added to water for domestic use to reduce the incidence of 

dental caries. Even at substantially higher levels (10 mg F/litre) the ions are without 

perceptible effects on brewing. 

3.7 Brewery effluents, wastes and by-products 

Breweries generate wastes, by-products, pollutants and effluents (Armitt, 1981; Brooks et 

al., 1972; Huige, 1994; Isaac, 1976; Isaac and Anderson, 1973; Klijnhout and Van Eerde, 

1986; Meyer, 1973; Rostron, 1996). These must be dealt with in the least costly way or, 

in one or two instances, profitably. Into these categories come noise, heat, odours, dusts 

(from malt and adjuncts), cullet (broken glass), waste aluminium cans, plastic waste, 

domestic and laboratory wastes, carbon dioxide, trub, spent grains, spent grain drainings 

and pressings, surplus yeast, used kieselguhr from filters, waste beer, wort, waste water, 

boiler blow-down sludge, and acids, alkalis and detergents (from CIP and other cleaning 

systems), labels, lubricants, copper condensate, PVPP and ion exchange regeneration 

reagents, and so on. Dealing with these is expensive. Breweries generate large volumes of 

waste water, and most of the remainder of this chapter is concerned with its disposal. 

However, all wastes and by-products must be disposed of quickly in the interest of saving 

space and minimizing the risks of microbial contamination. Broadly, all wastes are either 

dumped or disposed of for recycling or sold or discharged into a sewer or a waterway or 

the sea. Once it was normal to wash as much waste as possible down a sewer. In many 

countries this is now too costly and so attempts are made to recycle materials, to 

minimize waste production, to àdd value' to by-products or at least to sell them and to 

partly or completely treat liquid wastes. 


3.7.1 The characterization of waste water 

Brewery waste water is chiefly contaminated with putrescible organic matter and so it is 

comparatively easily purified by biological treatments. However, the flow of waste water 

from a brewery varies greatly with the time of day, often with the day of the week and 

with the time of year. Worse, the water will vary widely in its temperature, pH, load of 

suspended solids and the amounts of organic and inorganic materials in solution. If the 

waste is being discharged to a public sewer the operating authority will usually set limits 

on the composition, volume, rate of flow, temperature and pH of the effluent, with 

swingeing penalties if the limits are exceeded. Evidently the detailed composition of the 

water is variable and very complex. However, for treatment purposes water analyses are 

simplified, and these analyses are adequate for charging for treatments and for deciding 

what the correct treatments should be (Armitt,1981; Benson et al., 1997; Briggs et al., 

1981). 

Suspended, or settleable, solids, SS, are usually reported as mg dry matter/litre. 

Sometimes SS are defined as particles retained by a 1 m filter. Their presence results in 

a reduced water clarity and they may deposit and create blockages or be abrasive and 

cause wear to pumps. Some may be partly destroyed or removed during biological 

treatments. Suspended solids which are not biodegradable usually finish as sludge in 

treatment plants. Total dissolved solids, TDS mg/litre, dissolved organic carbon, DOC 

mg/litre, and total organic carbon, TOC mg/litre are measures sometimes used but much 

more emphasis is placed on the biological oxygen demand, BOD, and the chemical 

oxygen demand, COD. 

Because most older waste-water treatments were based on microbiological aerobic 

oxidations of dissolved organic matter, the biological, or biochemical, oxygen demand of 

the waste was of prime importance. The BOD5 20 is the weight of oxygen (mg) taken up as 

the organic substances in the water (1 litre) are oxidized by a mixture of micro-organisms 

in five days at 20 ëC, in the dark. The test must be in darkness to prevent the growth of 

algae, which generate oxygen. This important test is slow and not very precise. In 

consequence it is increasingly being replaced with determinations of the chemical oxygen 

demand, COD. This is normally calculated from the amount of dichromate used up when 

the water is boiled with the acid reagent for two hours in the presence of a silver sulphate 

catalyst. The results are expressed as mg oxygen/litre water. However, as the dichromate 

oxidizes more substances than the microbes the COD values are greater than the BOD 

values. The ratios between these values vary widely, but for most mixed brewery wastes 

the COD/BOD ratio is 1.6 1.8. For domestic sewage the value is about 2.5. The 

permanganate value, PV, an alternative determination of oxidizable organic matter, is 

now used less. 

In the UK the costs of having effluent treated in a municipal sewage works is usually 

calculated with reference to the `Mogden Formula', (named after a London sewage 

works) or a modification of it. 

C ˆ R ‡ V ‡ …OT=OSxB† ‡ …ST=SSxS† 

Where C ˆ the total charge/m 3 (unit volume) for the trade effluent discharge. R ˆ the 

cost of conveying and receiving the effluent + overhead costs. V ˆ unit cost of 

volumetric and primary treatments (screening and settlement). O T ˆ COD of the trade 

effluent after settlement at pH7. OS ˆ COD of the average settled sewage and trade 

effluent (i.e. a reference strength). B ˆ the unit cost of the biological treatment of the 

mixed and settled trade effluent and sewage. ST ˆ total suspended solids (SS) of the 


mixed trade effluent. SSˆThe SS of the mixture of the sewage and trade effluent, 

(standard strength). Sˆthe unit cost of the treatment and disposal, of the sludge. Thus 

increases of COD, SS or volume lead to greater charges. 

Extracharges willbe madefor effluents discharged outside stipulated`consent limits', 

such as peak volume flows, pH or temperature values, COD, SS, soluble nitrogen, and so 

on. The actual charges made vary from place to place and are continuing to rise, 

pressurizing brewers to find ways of reducing these costs. Using particular examples, 

with 1987 figures, Table 3.3 illustrates how increasing water usage, and the inevitable 

extra effluent production, can be very expensive. Examples of reported ranges of BOD 

(mg/litre) are beers, 60,000 120,000; spent grain press liquor, 60,000; waste yeast with 

entrained beer, 200,000; trub, 70,000; spent kieselguhr, 80,000; fermenter washings, 

20,000 30,000. Suspended solids, SS, represent 12 23% of brewery effluent BOD 

(Armitt, 1981; Benson et al., 1997). Examples of the average characteristics of brewery 

waste water are: BOD5(mg/litre), 400 1750 (900 2000); mean COD(mg/litre), 

1200 3,000; BOD5(kg/hl product), 0.45 0.95; SS(mg/litre), 93 772; SS (kg/hl 

product), 0.17 0.40; pH4.1 11.5; temperature 13 49ëC (55.4 120.2ëF), ratio of 

effluent volume produced/volume of beer produced, 4 33. Soluble nitrogen (mg/litre), 

30 80; phosphorus (mg/litre), 10 30. However, shock discharges, in particular effluent 

flows may be pH 2 3to 10 13 and BOD 170,000 500,000. These should be prevented 

from reaching the sewer (see below). Abrewery making 10,000 30,000 barrels of beer/ 

week may release in the effluent 2 5.6tCOD/day and 0.46 0.87tSS/day. 

Sometimes brewery effluent loads are expressed as population equivalents. Different 

equivalentsmaybeused(Armitt,1981).However,abrewerywithanoutputof10 6 hlwill 

produce about as much effluent as 50,000 people. The quality of treated water that may 

be discharged depends on whether this is to the sea, to an estuary, or to ariver, and what 

the minimum dilution rate will be. Some suggested limits are COD, 127mg/litre; BOD, 

25mg/litre; SS, 35mg/litre; nitrogen, 10mg/litre; phosphorus, 1mg/litre. Exceeding 

these, or other appropriate limits, will risk causing algal blooms, or the overgrowth of 

Table 3.3 Examples of calculated water and effluent costs, using mean UK Water Authority 

charges for 1987 (data of Askew, 1987). A brewery has an annual production of 1 10 6 barrels 

(36 10 6 imp. gal.; 1.637 10 6 hl). With a water : product ratio of 8 : 1, an effluent product ratio of 

5.5 : 1, a mean effluent BOD of 1000 mg/l, a mean COD of 1800 mg/l, and an effluent mean SS of 

500 mg/l then the water demand p.a. is 288 10 6 imp. gal., the effluent volume is 198 10 6 imp. 

gal, the BOD load p.a. is 900 tonnes (t), the COD load p.a. is 1620 t/p.a. and the SS load is 450 t/p.a. 

Assuming different efficiencies of water usage the three examples show the variations in water and 

effluent costs 

Measure Case 1 Case 2 Case 3 

Water: product ratio 8 : 1 9.23 : 1 12.9 : 1 

Annual water consumption (m 3 ) 1.309 10 6 

1.51 10 6 

2.11 10 6 

Cost (p/m 3 ) 27.8 27.8 27.8 

Cost (£/year) 363,902 419,780 586,580 

Effluent: product ratio 5.5 6.43 9.7 

Effluent volume (m 3 ) 900,000 1,052,180 1,587,270 

Effluent COD (mg/l) 1,800 3,580 4,500 

Effluent SS (mg/l) 500 291 400 

Effluent cost (p/m 3 ) 43.52 62.85 77.95 

Effluent cost (£/year) 391,680 661,295 1,237,277 

Total water and effluent costs (£/year) 755,582 1,081,075 1,823,857 

Since 1987 costs have risen to a considerable extent. At present (August, 2004) £1 ˆ 1.48 Euros ˆ $ (USA) 1.80 


microbescausing the receiving water tobecome anaerobicsokillingallhigherlifeforms. 

Clearly brewery wastes must be treated before discharge to waterways. The treatment 

may be carried out, in whole or in part, at the brewery or at asewage works. 

3.7.2 The characteristics of some brewery wastes and by-products 

The immense range of ratios of water taken in to beer produced is chiefly due to different 

efficiencies of water use, although some breweries, e.g., those that bottle a high 

proportion of their beer in returnable bottles, (which must be cleaned), are at a 

disadvantage.Tocontrolwasteitisnecessarytometerthevolumeandcompositionofthe 

effluent from every department and the brewery as a whole, to detect and prevent 

wasteful practices. To be of use effluent streams must be sampled in statistically valid 

ways. Time-proportional or volume-proportional sampling may be used. Apparently 

minor leaks, or leaving hoses running, or having CIP programmes operating with 

excessive water rinsing or without the re-use of final rinse liquor for the first rinse can 

cause substantial losses (Horrigan et al., 1989). The pH of effluent may become extreme, 

for example, when alkaline cleaning liquids or PVPP or ion exchange regeneration 

liquors are released. Precautions should be taken to trap these liquors and release them 

slowly at ametered rate with the general effluent to dilute them to an acceptable level, or 

their pH values may have to be adjusted before release. It has been proposed that alkaline 

liquors should be neutralized with carbon dioxide from furnace gases or from the 

fermenters rather than with mineral acids. 

Much effort has been spent in finding better ways of dealing with brewery by-products 

and wastes (Brooks et al., 1972; Horrigan et al., 1989; Huige, 1994; Penrose, 1985; Reed 

and Henderson,1999/2000; Vriens et al., 1986, 1990). Many breweries collect some of the 

carbon dioxide from the fermenters and use it to carbonate beer (Chapter 15). Others allow 

much of it to escape and yet others may use it to neutralize alkaline waste liquors. Furnace 

gases, which may also be used for neutralizing alkaline effluents, contain carbon dioxide 

and acid oxides of sulphur. Efforts are now made to save heat in breweries (re-use of 

cooling water, warming water by condensing vapours from the hop-boil, etc.), and so less 

heat escapes in effluents. In principle, heat could be withdrawn from effluents using heat 

pumps, but probably this is not economic. It has been suggested that warm, clean water 

from a brewery might be used in a fish farm. Trub (hot break) and spent hops usually 

contain some wort. If washed into the sewer they add substantially to the BOD and SS. 

Normally the entrained wort is recovered, for example, by adding the trub and spent hops to 

the lauter tun. Alternatively, the spent hops and trub may be added directly to the spent 

grains. Some 30% of the trub solids are digestible proteins that add to the feed value of the 

spent grains. Spent hops have been used as a mulch or as low-grade fertilizer. 

Sometimes trub is mixed with surplus yeast intended for animal feed. Surplus yeast is 

collected from fermenters, and is also present in tank bottoms. It can add substantially to 

the BOD and SS of effluents and, with a crude protein content of about 47% d.m., a 

carbohydrate content of 43% d.m., and a mixture of vitamins surplus yeast is potentially a 

valuable by-product. Often it is sold to companies that debitter it and turn it into food 

supplements (Putman, 2001). Surplus brewing yeast is used by distillers. On a 

comparatively small scale the yeast may be used as a source of biochemicals, particular 

proteins, enzymes and glutathione, and yeast extracts are used in culture media for 

microbes. Yeast tablets are used as vitamin supplements. Sometimes the yeast cake is 

washed and the yeast is returned to the production stream by adding it into the mash. 

Alternatively, it may be autolysed and then added to the spent grains to enhance their 


valueasanimalfeed.Maltdusthasbeenaddedtomashesorhasbeenmixedinwithspent 

grains. 

Used kieselguhr, from beer filters, used to be flushed into the sewers, but its large 

contribution to SS and BODs makes this undesirable and probably most is dumped, with 

or without first mixing it with quick-lime, at landfill sites. Small amounts have been used 

as asoil improver. Sometimes the used filter-aid has been mixed with the spent grains, 

but this is not always acceptable. Dumping is becoming costly and other disposal 

methods are being tested. Filter aid has been regenerated by chemical cleaning and by 

calcining to burn away organic residues. The regenerated material has been used again as 

afilter aid, as afertilizer carrier, and as afiller in paints and varnishes. 

`Waste' beer or wort may include last runnings from the mash, press-liquor from the 

spent grains, rinsingsfrom vessels orpipework,returnedbeer, spillages from bottlingand 

canning plants and rinsings from returned containers. All these residues may 

(expensively) be consigned to the sewer. Some brewers believe that returning last 

runnings and liquor from the spent grains to the mash constitutes arisk to the quality of 

the beer. Others have used these materials after clarification by centrifugation, with or 

without treatment with active charcoal, and keeping them for not more than two to three 

hours at 80ëC before adding them to the mashing liquor or to the copper (Coors and 

Jangaard, 1975). In these instances no deterioration of beer quality was detected. 

Returned beer may be blended or discarded. Returned beer and weak worts have been 

used to make alcohol or vinegar, to grow other microbes, such as fodder yeasts (Torula 

spp., Candida spp., Aspergillus spp.), for use in foodstuffs. Yet others have dried the 

liquids. The product, dried brewers' solubles, has been added to foodstuffs. Probably 

most of these processes are not usually economically viable. 

Spent grains are produced at the end of every mash. They are of value as afoodstuff, 

particularly for ruminants, but they are bulky, and they soon begin to decompose, so they 

mustberemovedfromthebrewerypromptly.Thehandlingequipmentmustbekeptclean 

to prevent the growth of spoilage organisms. Depending on the grists their composition is 

variable. In one example the composition reported, on adry weight basis, was crude 

protein, 27%; fat, 6 7%; ash, 4 5%; crude fibre, 15%; N-free extract, 46%. The 

moisture content of spent grains varies widely depending on the wort separation system 

used (Chapter 6). Thus grains from aStrainmaster may contain 87 90% moisture, and 

they are sloppy and are easy to pump, but they are so wet that liquid drains from them and 

they must be de-watered before removal. Water drains from grains with moisture contents 

above 80%. The drainings are an excellent medium for unwanted microbes. Grains from 

lauter tuns can contain 75 85% moisture and those from pressure filters contain as little 

as 50 55%. Exceptionally, these grains may be dried further in a current of hot air, to c. 

8% moisture, when the dried material is stable. Grains may be sold for animal feed either 

wet or after de-watering. Using continuous screw or roller presses the moisture contents 

can be reduced to 63 72%, producing a more easily handled solid material and the 

squeeze liquor. The liquor may contain up to 3.5% dissolved solids and 5% SS as it has a 

high BOD. As it contains valuable extract it should be returned to the process stream, 

where possible. In one case this saved about 1% of brewer's extract and reduced water 

use by 5% (Coors and Jangaard, 1975). 

The grains used for animal feed must be stored by farmers and the drier they are the 

easier they are to handle. Sometimes they are ensiled and additions of propionic acid or 

sorbic acid with phosphoric acid have been used to act as preservatives, and seeding with 

lactic acid bacteria has been proposed. As noted, autolysed yeast, trub, spent hops and 

even used kieselguhr may be added to spent grains. A range of other uses for spent grains 


has been proposed, as asource of biogas and soil conditioner produced by anaerobic 

digestion, disposal by burning (giving heat), as asource of `secondary worts' generated 

by acid orenzymic hydrolysis, as asourceof protein, as asourceof food-gradefibre,asa 

basis for mushroom compost, as asoil conditioner and organic fertilizer, as amedium for 

growing earthworms to use in poultry food, and in fish food. None of these alternative 

uses seems to be widely employed. 

3.8 The disposal of brewery effluents 

Aconsideration of the sources of brewhouse effluents shows that large volumes, BOD 

and SS are generated at different stages of the brewing process (Table 3.4; Askew, 1975, 

1987; Askew and Rogers, 1997; Huige, 1994; Love, 1987; Robertson et al., 1979). In 

some breweries all the `waste' water is collected into a common sewer. However, it is 

senseless for surface run-off (storm water) to be directed to a treatment plant and clean, 

but warm water that has been used for cooling should find a use in the brewhouse for 

mashing or cleaning, both to minimize water and effluent charges and to conserve heat 

and so reduce heating costs. If the effluents are treated, in whole or in part, at the brewery 

then it is usually advisable to separate `weak' and `strong' effluents and to treat them 

separately. The costs of treating effluents in municipal sewage treatment works are high. 

Brewers try to minimize the volumes and strengths (BOD, COD, SS) of the effluents. 

Sometimes it is economical to partly, or even extensively, treat brewery effluents on site. 

Secondary treatment may allow the water to be discharged into a watercourse. If costs 

continue to rise it may become worthwhile to purify some effluents further, using tertiary 

treatments, so that the water is pure enough to be used again for some purposes in the 

brewhouse and so, by recycling, avoid acquisition and disposal costs. Usually effluent 

purification is undertaken with reluctance. Treatment plant takes up space, it is costly, 

and it requires well-trained staff to operate it successfully. 

Whether effluents are treated at a brewery site or elsewhere the objectives are the 

same, to reduce the temperature to a moderate level (often under 40 ëC, 104 ëF), to restrict 

the pH to a specified range (e.g. 6 10) and to reduce the BOD, COD and SS levels to 

below specified levels (e.g. 25, 125 and 35 mg/l, respectively) so that the water can be 

released into a stream, river or estuary. Some breweries do not carry out any treatment on 

site, and there is no uniform system of treatment among the others. Historically, after 

preliminary screening, water was purified, using oxidative, aerobic biological systems. In 

recent years partial treatments using some anaerobic processes are being used. It is 

convenient to divide treatments into preliminary treatments, `primary' treatments, 

`secondary' treatments (aerobic, anaerobic or a combination of the two), and `tertiary' or 

`polishing treatments'. 

3.8.1 Preliminary treatments 

Many preliminary treatments are in use (Armitt, 1981; Benson et al., 1997; Huige, 1994; 

Klijnhout and Van Eerde, 1986; Vereijken et al., 1999; Vriens et al., 1986, 1990; Walker, 

1994). First the effluent should be screened to remove labels, bottle caps, floating plastic 

items and spent grains. These screens may be of many types, for instance, hyperbolic bar 

screens or screens of woven stainless steel mesh. It is desirable that the water also flows 

through a settling tank in which the reduction of the flow-rate permits cullet, grit, sand 

and some SS to settle. The deposit formed is removed at intervals by scrapers, to be 


Table 3.4 Estimated biological oxygen demand and suspended solids loads in a brewery employing yeast recovery (Robertson et al., 1979) 

Process Flow BOD SS 

(hl/day) (gal/day) (kg/day) (lb/day) (kg/day) (lb/day) 

Mash mixer 69.4 1526 3.6 8 7.3 16 

Lauter tun* 68.0 1496 189.6 418 122.5 270 

Brew kettle (copper) 97.2 2137 27.7 61 3.6 8 

Hot wort tank (whirlpool)* 26.6 584 284.0 626 98.0 216 

Wort cooler 24.1 531 0.5 1 0 0 

Fermenters* 290.0 6386 240.4 530 155.1 342 

Ageing tank (maturation) 407.7 8962 93.0 205 137.9 304 

Primary filter* 81.9 1801 46.3 102 231.3 510 

Secondary storage 385.1 8471 40.4 89 59.0 130 

Secondary filter* 92.6 2037 8.2 18 42.2 93 

Bottling tank 39.9 877 0.5 1 0 0 

Filler 922.4 20291 23.1 51 9.5 27 

Pasteurizer 3582.6 78808 34.5 76 3.2 7 

Bottle washer* 2933.5 64529 68.0 150 29.0 64 

Cooling water 4332.2 95298 0 0 0 0 

Miscellaneous flows 92.2 2028 0.5 1 0.5 1 

*Major sources of BOD and/or SS. Gal ˆ imperial gallons. (1 imp. gal ˆ 1.201 US gal ˆ 4.546 litres). 


dumped. Suspended kieselguhr and some yeast may be removed at this stage. The 

effluent may also pass through oil traps. The compositions, characteristics and flows of 

brewery effluents are highly variable and need to be èvened out' if treatments are to be 

successful. Strongly acid or alkaline cleaning or regeneration solutions are sometimes 

released immediately into the effluent stream. It is better practice to hold these solutions 

in a special receiver to mix acidic and alkaline solutions and, perhaps after adjusting the 

pH, release them into the main flow of effluent over a period of time to dilute them to 

such an extent that they are harmless. Sometimes such material may be removed by a 

specialist contractor. An additional `calamity tank' may be provided to hold sudden 

unexpected flows of liquid that can be released later, over a period, and so even out 

variations in the composition and flow rate of the effluent. All breweries, whether or not 

they treat their own effluents, should have a balancing or conditioning tank. This should 

be stirred and possibly be aerated to prevent the formation of odours. The size should be 

decided with reference to the brewery operations and should have a residence time (12 

hours and 36 hours have been suggested, as well as much longer periods) selected to truly 

àverage out' effluent flow rate and composition, so that it will not be harmful to the 

microbes that carry out the next stage of treatment. The pH of the effluent may be 

adjusted automatically as it leaves this tank and if it is going to a biological treatment 

plant on site some microbial nutrients may be added at this stage. 

Sometimes a different type of preliminary treatment may be carried out. For example, 

the effluent may be treated with a slurry of lime or another inorganic coagulant, and 

perhaps a synthetic polyelectrolyte. Then the mixture is given a flotation treatment. Air is 

dissolved in the effluent under pressure and this is directed into the base of the flotation 

vessel. The air separates as a cloud of fine bubbles and carries the coagulated materials up 

to the top of the vessel from which they are skimmed. In one case the removal of the SS 

was 96% and of the COD, 6% while in another instance the values were 60% and 45% 

respectively (Hughes, 1987; Lunney, 1981). 

3.8.2 Aerobic treatments of brewery effluents 

Many aerobic systems have been tried for treating brewery effluents and these are well 

understood (Armitt, 1981; Benson et al., 1997; Klijnhout and Van Eerde, 1986; 

KuÈhtreiber and Laa-Thaya, 1995; Reed and Henderson, 1999/2000). The effluent, 

preferably of uniform composition and having the correct pH and levels of supplementary 

nutrients (nitrogen and phosphate), is aerated in the presence of `sludge', a mixed 

population of micro-organisms. These multiply and grow. About 30% of the BODgenerating 

substances is oxidized to carbon dioxide and water, while the remainder is 

assimilated into microbial mass, the sludge. Surplus sludge is collected and must be 

disposed of. Sludge treatment and disposal is often the most difficult part of aerobic 

treatments. Broadly aerobic systems are operated in two different ways, although 

intermediate methods may be used. In the `low load' methods, effluent `lightly' loaded 

with BOD is supplied to a well-aerated mass of microbes and the contact time is 

extended. Under these conditions a type of sludge develops that is easy to handle and 

settles readily. BOD removal may exceed 98%. In `high load' systems the ratio of BOD 

supplied/unit of biomass is high. BOD removal is less, for example 80%. A large mass of 

`putrescible sludge' is formed. With effluents having high BOD values `sludge bulking' 

occurs, involving the excessive growth of filamentous microbes that do not readily settle. 

This sludge is difficult to handle and de-water. This system is less resistant to `shock 

loads' than the low load system. 


From these considerations it can be concluded that for optimal treatments brewery 

effluents should pass through a buffering tank (or two tanks if the high BOD and low 

BOD effluents are to be treated separately), and effluents with high BODs should receive 

a preliminary treatment (aerobic or anaerobic; see below) before they are treated further 

in a `low-load' system. Aerobic treatment systems are considered in two groups, the 

activated sludge systems in which the biomass is in suspension and systems in which at 

least most of the biomass is attached to a solid support. 

Perhaps the `single tank systems' are the most simple of those using activated sludge. 

In these effluent is progressively loaded into an aerated tank which already contains 

biomass which is mixed by the aeration process, usually achieved by blowing compressed 

air into the base of the tank through dispersers which release it as fine bubbles. When the 

tank is full the effluent is diverted to the next tank in line. Aeration is continued in the 

first tank until the BOD has been sufficiently reduced. Then aeration ceases and the 

active sludge is allowed to settle and the clear, treated effluent is drawn off from above 

the sludge. Any surplus sludge is removed, usually to a settling tank, and aeration is 

resumed and the tank is ready to begin receiving the next charge of effluent. 

Where large areas of ground are available effluents may be treated in large lagoons. 

Usually two lagoons operate in series and, because they are large, they constitute a lowload 

system. The lagoons are aerated, preferably with bubbles rising from the bottom, 

(surface aerators are inefficient and can create microbe-laden aerosols) and each may be 

followed by a sludge-settling tank. Effluent then flows into a lagoon, is diluted and aerated. 

After an average holding period, it flows into a settling tank. The clarified effluent flows to 

the next lagoon or out of the system while a proportion of the sludge is transferred back to 

the inlet at the entrance of the lagoon and is mixed with the incoming effluent, creating an 

initial high sludge concentration. With this and other aerobic systems, provided that 

aeration is adequate, the higher the sludge concentration the faster the effluent can be 

treated. In one case it was reported that BOD removal was about 95% in the first lagoon 

and increased to 99% in the second. The comparable COD values were 94 and 98%. 

Numerous more compact activated sludge systems, in which effluent flows through a 

series of aerated tanks and settling tanks, have been described. Several designs have been 

used in breweries. The Pasveer ditch consists of a shallow continuous ditch, roughly 

elliptical in plan, and with a trapezoidal cross-section, into which effluent flows and from 

which it flows to a settling tank. Some of the settled sludge is returned to the effluent 

inlet. The liquid is kept aerated and flowing around the ditch by brush beaters. The 

average retention time of effluent is often 2 3 days. When overloaded with effluent there 

are problems with sludge bulking, the plants occupy a large area and are in the open. 

Often breweries must use the least ground possible and the plant must not be obtrusive, 

particularly if situated in a town. Tall, thin, fully enclosed aerobic reactors may be 

chosen. An example is the deep-shaft reactor. These are typically 0.5 2.0 m 

(1.64 6.56 ft.) in diameter and extend 100 200 m (329 658 ft.) into the ground. In 

the centre of the shaft there is a downflow pipe and the annular space between the pipes 

acts as the riser. Initially the internal circulation is started by filling the shaft with effluent 

primed with activated sludge and then injecting air into the outer shaft. When the flow is 

established air is injected into the inner, down-shaft and as the bubbles are carried 

downwards it is efficiently dissolved as the pressure rises. As the pressure declines during 

the rise in the outer, annular shaft air and carbon dioxide break out of solution and the 

bubbles act as a gas-lift and maintain the circulation. 

High removals of BOD from brewery effluents have been obtained (Lom and 

Fedderson, 1981; Vriens et al., 1990). A comparable, above-ground tower system is 


described later (Fig. 3.2). One type of atwo-stage aerobic system is the Artois Unitank 

(Eyben et al., 1985; Vriens et al., 1986, 1990). In this system the effluent flows through a 

rectangular equalization tank, then through two rectangular aerobic/sludge settlement 

tanks working in series and so comprising a high-loaded tank and a low-loaded tank. 

Each aerobic/settlement tank is incompletely partitioned into three sections, so that liquid 

can move from one section into the next, but free mixing between the sections is 

prevented. The effluent flows from the equalization tank into the first compartment of the 

high-loaded treatment tank. It moves from compartment to compartment, meeting 

aeration and activated sludge in the first two. In the third compartment there is no 

aeration, the sludge settles and the treated effluent leaves, over a sludge-retaining weir, 

and flows to the low-loaded tank. At intervals of about two hours the flow in a tank is 

reversed and aeration now occurs in what was previously the settlement compartment 

(which is rich in settled sludge) and in the central compartment, but what was previously 

the inlet compartment is no longer aerated and becomes the settlement compartment. The 

third, low-loaded tank is operated in a similar way. BOD removal in the first tank is 

reported to be 80 88%, and to exceed 98% after the second tank treatment. At intervals 

surplus sludge is removed to a thickening tank. Promising trials with a novel, pilot-scale 

membrane reactor have been reported (Ward, 2000). The effluent from the enclosed, 

aerated and stirred chamber passed out by way of a membrane filter that retained all the 

suspended solids and microbes. Ferric sulphate, which acted as a coagulant, and nutrients 

were added to the aerated chamber and the pH was adjusted. Because the concentration of 

suspended microbes was exceptionally high and because no washout of activated sludge 

could occur, BOD removal exceeded 99%. 

In some other aerobic treatment plants the biomass is attached to a supporting solid, 

and is in contact with the aerated effluent. The oldest of this type of system is the 

trickling bed filter, in which effluent is sprayed over the surface of a bed of rough solids 

(such as gravel, broken rocks, or coke) and trickles downward over the solid's surfaces, 

meeting an up-flow of air. The beds may be circular or rectangular in plan and 2 3 m 

(6.56 9.84 ft.) deep. The bed packing becomes coated with a very complex mixture of 

microbes and other organisms, which oxidize dissolved organic materials and reduce the 

BOD. As one pass is insufficient the liquid is recirculated. At intervals surplus sludge 

sloughs away and is collected in a settling tank. To cope with BOD-rich effluents several 

filters may operate in sequence. Such filters are expensive to build, they occupy a large 

area and they have a limited capacity, being liable to `ponding' if overloaded. Sometimes 

problems arise from flies which multiply on the biomass. These filters have increasingly 

been replaced by high-rate biofiltration towers. These towers, which may be 4.3 6.1 m 

(14 20ft.) high, are packed with plastic units with shapes designed to support the 

biomass and to have a large surface to volume ratio but to be resistant to blockage. 

Effluent is sprayed into the top of the tower and as it trickles downward it meets an upflow 

of air and is oxidized by the film of microbes on the plastic. Towers have capacities 

about ten times those of trickling filters covering the same areas. If desired the liquid can 

be re-circulated and a BOD removal of 60 65% is obtained, so they achieve a partial 

treatment. Towers are followed by settling tanks, which retain sludge. If followed by 

further treatment in a low-loaded activated sludge plant 97% BOD removal and 94% SS 

removal has been achieved. If towers are overloaded they, like trickling filters, can give 

rise to unpleasant odours. 

Rotating disc contactors are another type of plant in which most of the biomass is 

supported on sets of lightweight plastic discs mounted side by side along a rotating axle. 

The surfaces of the slowly rotating discs are alternately immersed in a trough of effluent 


and then, together with the film of liquid coating the surface, are exposed to air. Strips 

projecting from the faces of the discs form chambers, which cause air to be carried down 

below the surface of the liquid (KuÈhtreiber and Laa-Thaya, 1995). Discs are about 45% 

submerged. Much of the biomass is supported on the discs but some may float free in the 

liquid in the trough. Other forms of rotating contactors have also been used. The energy 

requirements of these devices are much lower than those using compressed air for 

aeration. BOD removal of 80 90% is possible. 

Other types of aerobic treatment plants have been tried. For example, fluidised beds 

have been tested and some are in use (Section 3.2.4, Fig. 3.1). In these the biomass is 

supported in agranular, porous and relatively dense material. The mass, lifted into the 

flowing effluent by the aeration bubbles, achieves ahigh biomass density because the 

material supporting the biomass is too dense to be swept out of the aeration chamber, 

removingthe needforasettlementtank.Inanotherapproach,inapilotplant,the biomass 

was supported in small pieces of plastic sponge (Leeder, 1986). When these were 

overloaded with biomass the surplus was removed by collecting the pieces of sponge and 

squeezing them between rollers. 

3.8.3 Sludge treatments and disposal 

Aerobic effluent treatments inevitably produce considerable amounts of surplus biomass, 

`sludge', which retains about 70% of the mass of the substances that contribute to the 

BOD. The treatment and disposal of this sludge is inconvenient and contributes 

substantially, about 50%, to the cost of the treatment (Armitt, 1981; Huige, 1994; Vriens 

et al., 1986, 1990). Sludges differ in their characteristics, such as the ease with which 

they will settle and how compact they are. Normally sludges are collected by 

sedimentation when, with the addition of coagulants, they may have asolids content 

of 1 2% dry matter. After afurther period of settling the solids content will increase to, 

say, 2 4%. After each concentration treatment the liquid that has separated is returned to 

the effluent treatment plant. The sludge may be consolidated further by centrifugation, 

vacuum filtration or in a pressure filter. Each consolidation reduces the volume of sludge 

to be handled and so reduces the transport costs as the material is carted away for disposal 

(Table 3.5). An alternative is to treat the sludge with coagulants followed by flotation and 

collection by skimming. In some circumstances the sludge may be `stabilized', for 

example by aeration or by anaerobic digestion. Both of these treatments reduce the bulk. 

Anaerobic digestion is accompanied by the generation of methane-rich biogas, which can 

be used as a fuel. It is doubtful if this technique is used by brewers. It is used at some 

large sewage works. In some places dried sludge is incinerated, but again this does not 

seem to be suitable for brewers unless they are in a hot and dry region where the material 

can be dried spread on earth beds to dry in the open. Brewers usually have the sludge 

Table 3.5 The influence of dewatering on the volume of sludge (Lloyd, 1981) 

Stage of dewatering Moisture content (%) Equivalent volume 

From settling tanks 98.5 100 

After further settling 97.0 50 

After decantation 95.5 33 

After centrifugation 80.0 7 

From vacuum filter 70.0 5 

From filter press 65.0 4 


emoved by contractors who will bury it in landfill sites or may arrange to have it spread 

on farmland. The latter method of disposal is only available at certain times of the year. 

Because of its composition the sludge is potentially valuable. Markets for it have been 

sought, and it has been used (with or without admixture with lime) as a soil conditioner 

and has been used as a supplement in animal feeds. 

3.8.4 Anaerobic and mixed treatments of brewery effluents 

It was once considered impracticable to treat brewery effluents by anaerobic digestion, 

but methods for doing this have been developed (Anderson and Saw, 1986; Benson et al., 

1997; Driessen et al., 1997; Eder, 1982; Fuchs, 1995; Gerards and Vriens, 1996; 

Hellriegel, 1996; Huige, 1994; Klijnhout and Van Eerde, 1986; Langereis and Smith, 

1998; Love, 1987; Martin and Sanchez, 1987; Mayer and Eeckhaut, 1997; Pipyn et al., 

1983; Schumann, 1999; Schur et al., 1995; Swinkels et al., 1985; Vriens et al.,1986, 

1990). The effluent supplied to an anaerobic digestion plant must be carefully regulated 

in terms of its pH, flow, temperature, and BOD. These plants operate best on a steady 

flow of BOD-rich effluents and they are easily put out of commission by `shocks' or 

traces of toxic substances, so preliminary mixing and buffering treatments must have 

been applied to the effluent input. Start-up times are slow (4 10 weeks) because the 

anaerobic micro-organisms multiply slowly. Despite these stringent requirements 

anaerobic plants are being used both because they produce very little sludge, and 

because they are comparatively compact (about 70% less space than aerobic plant) and 

cheap to construct and run (no aeration plant is needed) and because biogas is generated 

(0.3 0.5 m 3 /kg COD removed) and this may be used as fuel in the boiler house. 

Anaerobic systems never completely remove BOD (50 95%, usually 70 85%), and 

COD removal is generally 60 75%. Essentially no nitrogen or phosphate is removed by 

anaerobic treatments. They are best regarded as preliminary treatments for strong 

effluents that must be followed by an aerobic treatment for weak effluents, separated 

from the strong effluents at the brewery, and the effluent from the anaerobic treatment. 

This may occur either at the brewery or at the sewage works. 

Anaerobic treatments occur in three stages, but often the first two take place in a 

single vessel. In the first stage, which may be aerobic, microbes generate and release 

hydrolytic enzymes that degrade the complex molecules present in the effluent to 

smaller molecules, e.g., polysaccharides to simple sugars, proteins to amino acids and 

free fatty acids are liberated from lipids. These simple molecules are easily assimilated 

by microbes. In the second, or àcidification' stage, which is strictly anaerobic, many of 

these molecules are converted to organic acids. At this stage some biogas, containing 

hydrogen and carbon dioxide, is produced. The products of the acidification stage are 

good substrates for the microbes needed in the next stage, but the best cultural conditions 

are different. After the acidification process the pH (6.6 7.6), temperature, and nitrogen 

and phosphate levels are adjusted as the liquid flows to the `methanization' vessel. 

Methanization is a strictly anaerobic process. It may be carried out at ambient 

temperatures (18 20 ëC; 64.4 68 ëF) or with thermophilic organisms at about 50 ëC 

(122 ëF) but in practice mesophilic conditions are usually chosen (about 35 ëC, 95 ëF). 

The effluent to be treated may be warmed by steam, partly produced by burning the 

biogas, and partly by heat recovered from the brewery. The biogas produced contains 

55 75% methane, 1 5% hydrogen, 25 40% carbon dioxide and 1 7% nitrogen, as 

well as traces of ammonia and hydrogen sulphide. This gas needs to be scrubbed with 

sodium hydroxide to remove the carbon dioxide and this may need to be supplemented 


with hydrogen peroxide to oxidize the hydrogen sulphide. Alternatively, the sulphide 

may be removed as iron sulphide or be adsorbed by silica gel. Escaping gases may be 

deodorized by passage through abiofilter. 

Variouskindsofequipmentareinusefortheanaerobic digestionofbrewery effluents. 

Stirred vessels have been used, with or without the microbes being supported on solid 

carrier materials. The preferred systems now use the UASB or upward-flow anaerobic 

sludge blanket system in which acidified effluent iscontinuously mixed into the base of a 

reaction vessel. As the liquid rises it passes through alayer of agglomerated microbes 

which have formed spheres 2 5mm in diameter. The process automatically selects for 

microbeswhichwillclumptogether.Theorganicmaterialsinsolutionareattackedbythe 

microbes, which release biogas. Very little extra sludge is formed. The biogas rises, 

carrying some of the granular microbial blanket. Towards the top of the vessel the rising 

mixture meets athree-phase separator. The gas is collected and taken from the vessel and 

the granular microbial material, which had been carried up by the gas bubbles, settles 

back into the body of the reactor. The liquid flows up and over one or more weirs and 

leaves the vessel. After passing through ade-foamer and being scrubbed the biogas, now 

chiefly methane, may be stored in agas-holder, be flared off or be directed to aboiler. In 

one case biogas provided 8% of the boiler's fuel requirements. The reactor requires a 

constant flow of liquor so if there is acheck in the inflow of effluent the liquor is recirculated. 

It is advantageous to have two reactors working in series. 

Anexampleofananaerobicandaerobicplantforpartialeffluenttreatmentisshownin 

Fig. 3.2. This plant, which was designed to occupy asmall ground area, has anumber of 

novel features (Driessen et al., 1997; Meijer, 1998). The IC Õ reactor is atwo-stage 

methanization tower, which can be regarded as having two UASB reactors mounted one 

above the other. In the aerobic, Circox Õ tower the microbes are supported on grains of 

Fig.3.2(opposite) Diagramofananaerobic/aerobiceffluenttreatmentplant(afterDriessenetal., 

1997). The fully enclosed plant is situated on a confined site (200 m 2 ) in a built-up area. Screened 

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, 

in the case of alkaline or other èxtreme' wastes, to a calamity tank (150 m 3 ), from where it may be 

metered into the buffer tank or to the sewer. Mixed effluent is transferred to the pre-acidification 

(PA) tank (500 m 3 ; 25 m high) where acidification and the hydrolysis of polymeric materials occur 

and where the pH and nutrient levels are adjusted. Phosphoric acid and urea may be added if 

required. The acidified effluent goes to the base of the IC Õ reactor tower (390 m 3 ; 20 m high). This 

is, in effect, two strictly anaerobic UASB reactors mounted one above the other. The influent enters 

the base (6) of the high-loaded reactor (1) and is mixed in with the surrounding fluid and granular 

microbial sludge. It rises through the sludge blanket in the reactor and meets the first three-phase 

separator (3). The gas rises to a de-foaming unit generating a gas-lift, and the liquid and entrained 

sludge return to (6) via a `downer' pipe, (5). The sludge, from the first three-phase separator, settles 

back into the body of the reactor while the liquid rises into the second, low-loaded reactor (2), 

moves through the second sludge blanket and meets the second three-phase separator (4). The 

biogas is scrubbed, collected and utilized, the sludge is retained in the reactor and the effluent 

moves on to the Circox Õ reactor (230 m 3 ; 19 m high). This second reactor is an aerobic tower, with 

an internal circulation through the inner, `riser' pipe (7) and the outer, annular `downer' channel. 

The circulation is maintained by the air-lift generated by the aeration air injected into the base. The 

ventilation air is drawn from all parts of the plant and sulphides in the air are oxidized to sulphate. 

The biomass is supported on granular basalt, so it readily sediments when it reaches the settling 

space (10). The gases rise through a pipe (9), and are vented. The effluent rises over weirs, which 

retain the biomass, and is directed by way of a cyclone tank that retains some suspended solids, but 

not fine suspended matter, like kieselguhr, which passes with the liquid into the sewer. The system 

is compact and has a rapid throughput with a low production of surplus sludge. The average 

reductions of the total COD and the soluble COD are 80% and 94% respectively. 


Waste water 

Screen 

0.5 mm 

Ventilation air 

Solid 

waste 

NaOH 

HCl 

500 m 3 

Buffer 

tank 

Copyright © 2004 Woodhead Publishing Limited and CRC Press, LLC 

Calamity 

tank 

To 

sewer 

500 m 3 

Preacidification 

tank 

Biogas 

4 

2 

3 

1 

Defoam 

tank 

IC reactor 

390 m 3 

Air 

5 

Gas/liquid 

separator 

Effluent 

6 

Gas 

scrubber 

10 

7 

Gasholder 

10 m 3 

Vented air 

9 

10 

8 

Aerobic 

Circox 

230 m 3 

Flare 

320 m 3 

Cyclone 

tank 

Effluent 

to sewer 

Biofilter 

14 m 2 

Effluent 

to sewer

asalt, which readily settle and so preventing the biomass being washed out of the 

column. The plant removes 80% of the total COD and 94% of the soluble COD. In this 

particular plant the kieselguhr is not removed and is discharged in the final effluent, 

which is delivered to the sewer. Other combined anaerobic and aerobic plants have been 

described (e.g. Eyben et al., 1985, 1995; Gerards and Vriens, 1996; KuÈhbeck, 1995; 

Mayer and Eeckhaut, 1997; Vriens et al., 1990). Under aerobic conditions microbes take 

up phosphate, and many store it as polyphosphate, and so this is removed from the 

effluent. The situation with nitrogen removal is more complex (Vriens et al., 1990). 

Aerobic conditions are needed to oxidize ammonium ions to nitrate, then anaerobic 

conditions are required for the nitrate to be reduced to nitrogen gas which is removed 

from the system (Eyben et al., 1995). 

3.9 Other water treatments 

Reedbeds have been used to treat some crude or partially purified malting effluents or 

even some sludge (Maule et al., 1996; Walton, 1995). They are essentially shallow tanks 

filled with a permeable support such as limestone chips on which reeds, e.g., Phragmites 

spp. and perhaps yellow flag iris are planted. The effluent percolates through the beds, 

among the roots of the plants. Suspended solids have been reduced from 80 to less than 

20 mg/l, and COD from 60 to less than 10 mg/l. The removal of nitrogen and phosphate 

appears to be satisfactory. A problem with reed beds is that they occupy large areas of 

ground. 

Treated effluents have been polished in various ways, including passage through reed 

beds, passage through large, shallow lagoons (with or without forced aeration), and 

recycling through percolating filters. As the cost of acquiring water continues to rise and 

the pressure to treat effluents to higher standards continues to increase it is inevitable that 

brewers working in large breweries will frequently consider purifying their effluents 

sufficiently to allow the water to be recovered and used, at least for some purposes, so 

minimizing water acquisition and effluent disposal charges. The re-use of effluent is 

likely to involve a two-stage aerobic or an anaerobic/aerobic treatment, followed by a 

polishing treatment, sterilization, sand filtration possibly followed by finer filtration, 

carbon filtration and perhaps ultrafiltration or demineralization by ion exchange. 


ANDERSON, G. K. and SAW, C. B. (1986) Pauls Brewing Room Book, 1986±1988 (8th edn), Pauls' Malt, 

Ipswich, p. 37. 

ANDERSSON, L. E. and NORMAN, H. (1997). Brew. Distill. Internat., 28 (7), 18. 

ANON. (1988) European Brewery Convention Monograph XIV. EBC Symposium. Water in the Brewing 

Industry, Zoeterwoude. 

ARMITT, J. D. G. (1981) in Brewing Science 2 (Pollock, J. R. A., ed.) Academic Press, London, p. 551. 

ASKEW, M. (1975) Process Biochem., 10 (1), 5. 

ASKEW, M. (1987), Brewer, 73, Nov., p. 500. 

ASKEW, M. and ROGERS, S. (1997), Brewers' Guard., 126 (5), 24. 

BAK, S. N., EKENGREN, OÈ ., EKSTAM, K., HAÈ RNULV, G., PAJUNEN, E., PRUCHA, P. and RASI, J. (2001) European 

Brewery Convention Manual of Good Practice. Water in Brewing. Hans Carl, NuÈrnberg, 128 pp. 

BAXTER, E. D. and HUGHES, P.S. (2001) Beer: quality, safety and nutritional aspects. Cambridge. The 

Royal Society of Chemistry, 138 pp. 

BENSON, J. T., COLEMAN, A. R., DUE, J. E. B., HENHAM, A. W., TWAALFHOVEN, J. G. P. and VINCKX, W. (1997) 

European Brewery Convention Manual of Good Practice: Brewery Utilities. Hans Carl, NuÈrnberg, 

p. 137. 


BERKMORTEL, H. A. VAN DEN. (1988a) European Brewery Convention Monograph XIV. EBC Symposium. 

Water in the Brewing Industry, Zoeterwoude, p. 49. 

BERKMORTEL, H. A. VAN DEN. (1988b) MBAA Tech Quart., 25 (3), 85. 

BLACKMANN, B. (1998) Brewers' Guard. 127, Apr., p. 30. 

BRAUN, G. and NIEFIND, H.-J. (1988) European Brewery Convention Monograph XIV. EBC Symposium. 

Water in the Brewing Industry, Zoeterwoude, p. 68. 

BRIGGS, D. E., HOUGH, J. S., STEVENS, R. and YOUNG, T. W. (1981) Malting and Brewing Science, Vol I, 

Malt and Sweet Wort (2nd edn). Chapman and Hall, London. p. 194. 

BROOKS, R. B., HALFORD, M. H. and SKINNER, R. N. (1972) Proc. 12th Conv., Inst. of Brewing (Australia 

and New Zealand Sect.), Perth, p. 73. 

BROWN, J. W., BOTT, N.J., BOWERMAN, R. and SMITH, D. (1999) Proc. 27th Congr. Eur. Brew. Conv., 

Cannes, p. 225. 

BYRNE, H. (1990) Ferment, 3 (2), 90. 

CLEATHER, T. G. (1992) Proc. 5th Internat. Brew. Tech. Conf., Harrogate, p. 287. 

COMRIE, A. A. D. (1967) J. Inst. Brewing, 73, 335. 

COORS, J. H. and JANGAARD, N. O. (1975) Proc. 16th Congr. Eur. Brew. Conv., Nice, p. 311. 

COWAN, J. A. and WESTHUYSEN, J. VAN DER (1999) Proc. 7th Sci. Tech. Conv. Inst. Brewing, Africa Sect., 

Nairobi, p. 193. 

CRISPIN, P. (1996) Brewers' Guard., 125 (9), 34. 

DRIESSEN, W., HABETS, L. and VEREIJKEN, T. (1997) Ferment, 10 (4), 243. 

EDEN, G. E. (1987) Brewer, 73, Nov., 504. 

EDER, L. J. (1982) MBAA Tech Quart. 19 (3), 111, 138. 

EUMANN, M. (1999) Proc. 7th Sci. Tech. Conv. Inst. Brewing, Africa Sect., Nairobi, p. 169. 

EYBEN, D., GERARDS, R. and VRIENS, L. (1995) MBAA Tech. Quart., 32 (3), 142. 

EYBEN, D., VRIENS, L., FRANCO, P. and VERACHTERT, H. (1985) Proc. 20th Congr. Eur. Brew. Conv., 

Helsinki, p. 555. 

FUCHS, C. B. (1995) MBAA Tech. Quart., 32, 85. 

GERARDS, R. and VRIENS, L. (1996) Proc. 24th Conv. Inst. Brewing, Asia-Pacific Section, Singapore, 

p. 192. 

GOUGH, A. J. E. (1995) MBAA Tech. Quart., 32, 195. 

GRANT, A. P. (1995) Ferment, 8 (4), 252. 

HACKSTAFF, B. W. (1978) MBAA Tech. Quart., 15 (1), 1. 

HARRISON, J. G., LAUFER, S., STEWART, E. D., SIEBENBERG, J. and BRENNAN, M. W. (1963) J. Inst. Brewing, 

69 (4), 323. 

HARTEMANN, P. (1988) European Brewery Monograph XIV. EBC Symposium, Water in the Brewing 

Industry, Zoetewoude. p. 37. 

HELLRIEGEL, K. (1996) Brauwelt Internat., 14 (5), 422. 

HERON, J. R. (1989) Ferment, 2 (2), 118. 

HORRIGAN, R., LLOYD, W. J. W. and YOUNG, I. M. (1989) Project No. 59. Report of the Joint Maker/User 

Committee, Inst. Brewing and ABTA. Conservation of water and reduction of effluent. 

HUGHES, D. A. (1987) Brewer, 73 (872), 266. 

HUIGE, N. J. (1994) in Handbook of Brewing (Hardwick, W. A., ed.), Marcel Dekker, New York. p. 501. 

IBBOTSON, G. E. (1986) Brewer, 72 (859), 169. 

ISAAC, P. G. (1976) Process Biochem. 11 (2), 17. 

ISAAC, P. C. G. and ANDERSON, G. K. (1973), J. Inst. Brewing, 79, 154. 

KLIJNHOUT, A. F. and VAN EERDE, P. (1986) J. Inst. Brewing, 92, 426. 

KUÈ HBECK, G. (1995) Proc. 25th Congr. Eur. Brew. Conv., Brussels, p. 751. 

KUÈ HTREIBER, F. and LAA-THAYA, A. (1995) Brauwelt Internat. (1), 45. 

KUNZE, W. (1996) Technology Brewing and Malting, VLB, Berlin, p. 60. 

LANGEREIS, W. H. and SMITH, C. G. (1998) Proc. 25th Conv. Inst Brewing, Asia-Pacific Sect., Perth, 

p. 187. 

LEEDER, G. I. (1986) Brewer, 72, p. 213. 

LLOYD, W. J. W. (1981) Brewer, 67, 396. 

LOM, T. and FEDDERSON, C. C. (1981) Proc. 18th Congr. Eur. Brew. Conv., Copenhagen, p. 121. 

LOVE, L. S. (1987) MBAA Tech. Quart., 24 (2), 51. 

LUNNEY, M. J. (1981) Brewers' Guard, 110 (1), 13. 

MAILER, A., PEEL, R. G., THEAKER, P. D. and RAVINDRAN, R. (1989) MBAA Tech. Quart., 26, 35. 

MARTIN, S. and SANCHEZ, R. (1987) Proc. 21st Congr. Eur. Brew. Conv., Madrid, p. 655. 

MAULE, A. P., TRUMPESS, C. R. and THURGOOD, S. D. (1996) Proc. 6th Internat. Brew. Tech. Conf., 

Harrogate. Brewing Technology, the Market and the Environment. p. 408. 

MAYER, W. and EECKHAUT, M. (1997) Brauwelt Internat., 15 (5), 414. 

MCGARRITY, M. J. (1990) Louvain Brewing Lett., 4 (3/4), 3. 

MEIJER, D. (1998) Brauwelt Internat., 16, 455. 

MEYER, H. (1973) Proc. 14th Congr. Eur. Brew. Conv., Salzburg, p. 429. 

MOLL, M. (1979) in Brewing Science, 1 ( Pollock, J. R. A., ed.) London, Academic Press. p. 539. 


MOLL, M. (1995) in A Handbook of Brewing (Hardwick, W. A. ed.). New York, Marcel Dekker, p. 133. 

PENROSE, J. D. F. (1985) Brewers' Guard., 114, 25. 

PIPYN, P., OMBREGT, J. P. and TOYE, J. (1983) Proc. 19th Congr. Eur. Brew. Conv., London, p. 587. 

PUTMAN, R. (2001) Brewer Internat., 2 (4), 27. 

REED, R. and HENDERSON, G. (1999/2000) Ferment 12 (6), 13. 

ROBERTSON, J. L., BROWN, L. C. and MURPHY, K. L. (1979) MBAA Tech. Quart. 16 (1), 33. 

ROSTRON, J. (1996) Brewers' Guard., 125, (5), 20. 

RUDIN, A. D. (1976) Brewers' Guard., 105 (12), 30. 

SCHUMANN, G. (1999) Brauwelt Internat., 17 (5), 374. 

SCHUR, F., BHEND, D., BUCHER, A. J. and WETZEL, E. (1995) Proc. 25th Congr. Eur. Brew. Conv., Brussels, 

p. 741. 

SWINKELS, K. T. M., VEREIJKEN, T. L. F. and HACK, P. J. F. (1985) Proc. 20th Congr. Eur. Brew. Conv., 

Helsinki, p. 563. 

TAYLOR, D. (1981) Brew. Distill. Internat., 11, 35, 42. 

TAYLOR, D. G. (1989) Ferment, 2 (1), 76. 

THEAKER, P. D. (1988) Brewers' Guard., 117 (3), 22. 

THOMPSON, J. (1995) Ferment, 8 (3), 177. 

VEREIJKEN, T. L. F., DRIESSEN, W. J. B. and YSPEART, Y. (1999) Proc. 7th Sci. Tech. Conf. Inst. Brewing, 

Africa Sect., Nairobi, p. 174. 

VRIENS, L., EYBEN, D. and VERACHTERT, H. (1986) Proc. Symp. J. De Clerck Chair II; Microbiology and 

the Brewing Industry from Barley to Beer. Leuven/Louvain, p. 128. 

VRIENS, L., VAN SOEST, H. and VERACHTERT, H. (1990) CRC Crit. Revs. Biotechnol., 10 (1), 1. 

WALKER, M. J. (1994) Proc. 4th Aviemore Conf. On Malting, Brewing and Distilling. London, Inst of 

Brewing, p. 345. 

WALTON, C. (1995) Brew. Distill. Internat., 26 (7), 29. 

WARD, J. A. (2000) Proc. 26th Conv. Inst. Brewing, Asia-Pacific Sect., Singapore, p. 122. 


4 

The science of mashing 


Mashing consists of mixing ground malt (usually amixture of malts) and other prepared 

grist materials (appropriate adjuncts and sometimes salts and sometimes supplementary 

enzymes,Chapter2)withacarefullycontrolledamountofliquoratachosentemperature. 

In some few instances the mash may be made with mainly, or entirely, unmalted 

preparations of cereals mixed with industrial enzymes. The liquor is nearly always prepurified 

and contains achosen mixture of dissolved salts (Chapter 3). In the simplest 

systems, after a `stand' or stands of various durations, at one or more selected 

temperatures, the liquid, or`sweet wort' isseparatedfrom the residual solids of the mash, 

the spent grains or draff. In other cases portions of the mash are withdrawn and heated 

before being added back to the `main mash' while in other cases cereal preparations, 

cooked in a separate vessel, are transferred and mixed into the main malt mash. 

Representative mashing schedules are considered below. The sweet wort goes forward to 

the kettle or copper, to be boiled with hops, while the spent grains are disposed of 

(Chapter 3). The, apparently simple, processes of mashing conceal a very complex 

mixture of physical, chemical and biochemical changes. An understanding of these 

changes has been essential to permit the logical development of mashing conditions for 

the preparation of desirable and uniform worts in rapid and reproducible ways. Thus the 

purpose of mashing is economically to prepare wort of the correct composition, flavour 

and colour in the highest practical yield and in the shortest time. 

Thewortis partly characterized by its`strength', the amount ofsolids, orèxtract'that 

are in solution and the volume of liquid in which the solids are dispersed. Unfortunately, 

the concentration of wort is expressed in avariety of different units (Appendix). In the 

simplest case the specific gravity is used as a measure. The higher the specific gravity the 

more concentrated the solution of wort solids. For example, a wort might have a specific 

gravity (s.g.) of 1.040 at 20 ëC (68 ëF), relative to pure water at 1.000, or the same wort 

has an SG of 1040 relative to water as 1000.0. In other systems the amounts of solids in 

solution are estimated from the s.g. and reference to tables relating to the s.g. values of 

sucrose solutions, either from the table of Balling (ASBC) or the table of Plato (EBC; see 


Appendix). So awort with an SG of 1040 has aconcentration of solids of about 9.99% 

(w/w) assuming that the wort solids influence the SG in an identical way to sucrose. In 

the old British system (pre-1977) the extract in a barrel of wort at 15.5 ëC (60 ëF) was the 

excess weight of a barrel of wort in pounds (lb.) over the weight of a barrel of water, 

360 lb. So the SG ˆ (excess Brewer's lb. + 360)/360. So in the case given, SG ˆ 1040 (at 

15.5 ëC), the excess SG ˆ 40 so the extract, in Brewer's lb. ˆ 40 0.36 ˆ 14.4. The 

efficiency of mashing is often estimated by comparing the extract recovered in the 

brewery with that obtained in laboratory mashes when the hot water extract (HWE) of the 

grist is determined. Reliable estimates of extract recoveries are not freely available, but 

while in old mash tuns the value might be about 85 95%, in newer mash filters the value 

may equal or just exceed 100%, a commercially valuable advantage provided that wort 

quality is maintained. 

The importance of enzymes in mashing is illustrated by the fact that a cold water 

extract of a pale malt (CWE; preformed soluble materials), prepared in a cool, dilute 

solution of ammonia to stop enzyme activity, will be in the range 15 22%, dry basis, 

while the hot water extract (HWE) will be 75 83%, dry basis. Thus, by holding a smallscale 

mash for between one and two hours under conditions such that enzyme activity is 

favoured, some 53 68% more of the malt solids are brought into solution as the result of 

enzyme-catalysed reactions. While about half of the CWE solids are fermentable by 

yeasts typically 75% and up to about 85% of the HWE is fermentable. The enzyme 

catalysed changes that occur during mashing are more complex than those normally 

investigated by biochemists, who usually study each enzyme acting in isolation, with a 

homogeneous substrate, at one temperature, with an unchanging pH and with a large 

excess of substrate to maintain enzyme activity. 

In contrast, in mashing, a very large number of enzymes act simultaneously on the 

components of the grist (malt and mash tun adjuncts) under conditions that are far from 

optimal for many of them in terms of substrate concentration and accessibility, pH and 

enzyme stability. Enzymes are inactivated at different rates depending on the 

temperature, the pH, the presence of substrate and other substances (such as tannins 

and cofactors such as calcium ions) in solution. Starch, proteins, nucleic acids, lipids and 

other substances are attacked, usually by hydrolytic reactions, but other reactions, such as 

oxidations, also occur. Not only are enzymes progressively inactivated but substrate 

concentrations alter and, in the case of starch for instance, are nearly totally degraded. 

Solid starch granules are not readily degraded until gelatinization temperatures are 

approached and starch grains are disrupted. The polysaccharide cell walls and the starch 

granules are coated with proteins that seem to impede their enzymic degradation. Where 

grist particles are relatively large and the cell walls are intact, as in unmodified fragments 

of malt and some adjuncts, the cell walls prevent enzymes reaching and degrading the 

cell's contents. In many instances the products of hydrolysis competitively inhibit 

enzyme activity and in a number of instances (e.g. -amylase, limit dextrinase and some 

proteolytic enzymes) proteins occur which partially or largely inhibit enzyme activities. 

In addition some enzymes occur bound to insoluble materials in the mash, preventing 

them diffusing and so limiting their ability to reach their substrates. This is true of - 

amylase, proteases and -glucosidase in malts. The significance of insoluble enzymes 

and the presence of endogenous enzyme inhibitors in mashes has been widely ignored 

A knowledge of the properties of enzymes is essential for an understanding of mashing 

regimes, yet the traditional methods for studying their properties are of limited use. 

Brewers make use of the concept of òptima' in considering enzyme activities in mashes. 

This is helpful, but it must be realized that a temperature or a pH optimum is not a true 


constant. Under any particular set of conditions an enzyme may be stable and so, when it 

is incubated with asubstrate, product(s) will be formed at aconstant rate until substrate 

concentration is reduced to asignificant extent. If the temperature is increased then the 

reaction catalysed will proceed faster. However, if at this higher temperature the enzyme 

is progressively inactivated then the rate of formation of the product(s) will decline and 

will ultimately stop even if some substrate remains. At progressively higher temperatures 

the reaction rates catalysed by the `native', undenatured enzyme will increase, but the 

rate of enzyme inactivation also increases and so the production of product will cease 

sooner. Thus the amount of product formed, at aparticular temperature in agiven time, 

will depend on the rates of catalysis and on the rate of enzyme inactivation. The longer 

the reaction period the lower the optimum temperature, that is, the temperature at which 

the most product will have been produced (Fig. 4.1). In mashing, the rate of appearance 

of asubstance depends on the mixture in the grist, the mash thickness, the fineness of the 

grind and so the particle size distribution in the grist. Similarly the optimum pH of a 

reaction varies with the test conditions. The situation with pH is complicated by the 

frequentfailure totakeaccountofthepHchangesthatoccurasthetemperatureisaltered, 

and the difficulty of measuring pH at the elevated temperatures used in mashing (see 

below). 

Sweet wort is viscous, sweet, dense, sticky and more or less coloured. Its composition 

is highly complex (probably thousands of components are present). No wort has ever 

been completely analysed. Substances present include simple sugars, dextrins, -glucans, 

pentosans, phosphates, dissolved inorganic ions, proteins, peptides and amino acids, 

nucleic acid breakdown products, lipids, yeast growth factors (vitamins), organic acids, 

bases and phenolic substances.Sometimes it is desirable toanalyseachemical fraction in 

detail, but this is not always the case. A `typical' sweet wort may contain solids 

consisting of about 90 92% carbohydrates, 4 5% nitrogen-containing substances and 

Product 

t1 t2 t3 

Incubation time 

Fig. 4.1 Graphs illustrating the appearance of products in idealized enzyme-catalysed reactions 

carried out with the initial conditions the same but at different incubation temperatures (ëC) (after 

Dixon and Webb, 1958). As the temperature is increased so the initial reaction rate, at time zero, 

increases. In the sample, at 40 ëC the enzyme is stable and so the reaction carries on at a steady rate 

(substrate is present in excess), and product is formed linearly with time. However, at higher 

temperatures the enzyme is less stable and so, although the initial reaction rates are more rapid, 

enzyme is progressively inactivated and the rates of product formation decline. If the maximum 

amounts of product formed at different times are noted it can be seen that at the shortest time, t1, the 

òptimum' is at 70 ëC, at t2 it is at 60 ëC while at the longest time, t3, it is at 50 ëC. 


50° 

60° 

70° 

40° 

80°

1.5 2% ash (MacWilliam, 1968). As yeast ferments wort the simpler sugars are partly 

converted into ethyl alcohol and carbon dioxide, and the specific gravity of the mixture 

progressively declines until fermentation is complete. This final value is the attenuation 

limit of the wort and mainly depends on the carbohydrate spectrum of the wort (Chapters 

12 and 14). During the hop-boil the fermentability of the wort and its strength may be 

adjusted by the addition of carbohydrate adjuncts (Chapter 2). Enzymes remain in sweet 

wort, slowly increasing its fermentability, which is not fixed until the wort is boiled. 

Fermentability values are calculated from the changes in specific gravity. 

Wort colour mustbe within specification, and so mustthe nitrogen (crude protein ˆN 

6.25) fractions of the wort. Total soluble nitrogen (TSN) is self-explanatory. Free 

amino nitrogen (FAN) is ameasure of the low molecular weight substances, mainly 

amino acids, which are needed to support yeast growth and metabolism. Older measures 

included coagulable-N (nitrogen containing material that was precipitated when the wort 

was boiled) and permanently soluble nitrogen, (PSN) which remained in solution. 

Formol-N was an older method of estimating the amino acid and peptide fraction. 

4.2 Mashing schedules 

Mashingschedulesvarywidely.Oneischosenwithreferencetothebeertypetobemade, 

thewayithasbeenmadeinthepast,theplantandrawmaterialsthatareavailableandthe 

energy consumption and speed of the process. There is atrend towards temperatureprogrammed 

infusion mashing but some brewers have not been able to `match' their 

traditional products when changing from older types of mashing programmes, and so 

thesehavebeenretained.Atpresentarangeofbrewingproceduresareinuseandtheyare 

carried out in many different types of equipment (Chapters 5and 6). In many small 

breweries it is normal to mash not more than once a day, but in some large production 

units the target is to mash 12 14 times every 24 hours. This has necessitated many 

changes in equipment and mashing practices. It is convenient to distinguish between 

traditional infusion mashing, decoction mashing, double mashing, temperatureprogrammed 

infusion mashing and àll-' or `mainly-adjunct' mashing, although the 

distinctions between these classes are not absolute. Some `mixed' mashing systems are 

used in Belgium (Briggs, et al., 1981; De Clerck, 1957; Kunze, 1996; Narziss, 1992a; 

Hind, 1950; Wright, 1892). The mechanisms of grist preparation and mashing are 

discussed in Chapters 5 and 6. 

Before 1945 the infusion mashing carried out in the UK typically involved making a 

thick mash with well modified and comparatively coarsely ground malt or malts, mixed 

with 5 15% of flaked maize or flaked rice adjunct. The grist was mixed with hot liquor 

(water) at a temperature (`liquor heat' or `striking heat') chosen to give a particular 

ìnitial heat' or mash temperature. After a stand of about 30 minutes, when the mash gave 

a negative iodine test for starch, an underlet (hot water added into the bottom of the mash) 

might be given to raise the temperature then, after a total period of 2 3 h, wort collection, 

recirculation and sparging would begin. Typically on mashing in the liquor/grist ratio 

would be 2.15 2.42 hl/100kg grist (2 2.25 imp. brl/Qr.) and the temperature would be 

63.4 67.2 ëC(146 153 ëF). After underletting, with additions of hot water of 0 1.34 hl/ 

100kg (0 1.25 imp. brl/Qr.) the temperature of the mash would be 66.6 68.8 ëC 

(152 156 ëF). Finally, the wort would be collected and the goods (residual solids) were 

sparged with 3.76 4.30 hl/100 kg (3.5 4 imp. brl/Qr.) of liquor at 75 77 ëC 

(167 170.7 ëF). Thus the whole process from mashing in to finishing collecting the 


wort would take at least six hours. In one brewery the whole process took 18 hours. This 

process cannot be greatly accelerated. 

The choice of good quality malt minimizes the chance of a set mash and allows the 

stand to be shortened to 1 1.5 h, and a shortening of the wort separation time and the 

addition of some hydrolytic enzymes can accelerate wort separation. Mashing directly in 

a lauter tun, rather than a mash tun, allows the use of a more finely ground grist and faster 

wort separation. By shortening the stand period and by accelerating sparging, time can be 

saved but at the risk of reducing extract recovery and altering the quality of the wort. The 

total volumes of liquor used were 6.98 7.52 hl/100 kg grist (6.5 7.0 imp. brl/Qr.). 

Modern infusion mashes are made with 1.6 3.2 hl liquor/100 kg grist (1.5 3.0 imp. brl/ 

Qr.). Torrefied cereals or wheat flour are commonly used adjuncts. The initial 

temperature is usually in the range 63 67 ëC (145.5 152.7 ëF) and is best held for 

1 3 h. The temperature of the mash rises during sparging. This type of mashing does not 

allow air to be excluded from a mash, and indeed the entrained air bubbles cause much of 

the mash to float. This is no disadvantage, and may even be desirable, for making 

traditional, cask-conditioned British beers. However, with other beers, intended to have 

very long shelf-lives, efforts are increasingly being made to exclude air from the mash 

and the hot wort. To achieve this equipment other than a mash tun must be used. In 

special cases, when alcohol production is to be minimized, the mashing-in temperature is 

increased to, e.g., 75 ëC (167 ëF) to allow -amylase to liquefy and dextrinize the starch 

while minimizing saccharification by -amylase and so producing a less fermentable 

mixture of carbohydrates. 

In traditional continental European decoction mashing a thin mash (3.5 5 hl liquor/ 

100 kg. grist; 3.26 4.66 imp. brl/Qr.) is made from undermodified malt that is 

comparatively finely ground. The thin mash is necessary to permit it to be stirred and 

pumped between mashing vessels. In this, and the other mashing systems to be 

considered, the mash conversion processes are carried out in vessels that are separate 

from the devices (lauter tuns or mash filters) in which the wort is separated from the 

residual spent grains. Because the mash is stirred and portions of it are pumped between 

vessels air is not entrained and the solids do not float. When portions of the mash are 

boiled the starch is gelatinized and becomes susceptible to enzymic attack, residual 

cellular structures are disrupted, proteins are denatured and precipitated, enzymes are 

inactivated, chemical processes are accelerated, flavour substances (not necessarily 

desirable) appear in the wort and the wort darkens. Unwanted substances such as 

pentosans and -glucans are extracted. Boiling portions of the mash is expensive because 

it involves the consumption of energy. The successive temperatures, which occur in the 

`main, mixed mash', allow key enzymes to act at or near their optimal temperatures. In 

decoction mashing the grist is mashed into the mash-mixing vessel, which has a stirrer 

and may have heat-exchanging surfaces to allow the temperature of the contents to be 

increased. At intervals aliquots of the mash are withdrawn to the decoction vessel where 

they are heated, rapidly or slowly as the programme requires, with or without `rests' at 

particular temperatures, to boiling. After a period of boiling the hot material is pumped 

back into, and is mixed with, the main mash raising its temperature at a predetermined 

rate to a pre-chosen value. Before a decoction is carried out the stirrer in the mash-mixing 

vessel may be turned off and the mash allowed to settle. Then part of the settled `thick 

mash' is pumped to the decoction vessel. 

If adjuncts are permitted they may be cooked, with some of the malt mash and 

possibly added microbial enzymes, in the decoction vessel. The mash is allowed to stand 

until the next temperature rise, created either by another decoction, or by direct heating or 


y sparging. At the end of the mash conversion period the mash is transferred either to a 

lauter tun or to a mash filter for wort separation. With undermodified malts double 

decoction mashing is said to recover 2% more extract than an infusion mash and a single 

decoction process recovers 1.5% more extract. These gains are made at the expense of 

higher energy costs, as boiling part of a mash requires heat. With well-made malts the 

advantages, if present, are very small if temperature-programmed infusion mashing is 

employed. The brewing problems created by using poorly modified malts are such that 

their use is now avoided where possible and so the need for decoction mashing is going. 

On the other hand some mainland European beers have their full and desirable range of 

characteristics only if their worts are prepared by decoction mashing. 

In British infusion mashing, carried out with well modified malts, extract yield is 

likely to be limited by the extract recovery from the mash, rather than the extent of the 

mash conversion. In other words extract will remain in the spent grains. Decoction 

mashing schedules are very flexible and are easily adjusted (Kunze, 1996; Narziss, 

1992a, b). In the classical three-decoction process (Fig. 4.2) light beers are made with a 

liquor/grist ratio of 4.8 5.4 hl/100 kg grist (4.5 5 imp. brl/Qr.) while dark beers are 

made with thicker mashes, 3 4 hl/100 kg (2.75 3.75 imp. brl/Qr.). In the decoctions 

used when making light beers, the boiling periods are shorter than when dark beers are 

being made. The grist may be mashed in with cold water and the temperature is raised to 

35 40 ëC (95 104 ëF) either by adding hot water, or by direct heating, while the mash is 

stirred. The main mash may be allowed to remain at this temperature for about two hours. 

During this stand heat-labile enzymes, such as -glucanase, maltase, proteases and 

phytase, have a chance to act. The pH of the mash may fall, partly due to the activities of 

lactic acid bacteria. After about one hour into this period a third of the mash (stirred `thin' 

or settled and `thick') is transferred to the decoction vessel and is heated to boiling, often 

Temperature (°C) 

100 

75 

50 

25 

1 2 3 

Proteolysis active in mash-mixing vessel 

Saccharification active 

0 32 

0 1 2 3 4 5 6 

Time (h) 

212 

167 

Enzyme 

inactivation 

Fig. 4.2 The temperature changes occurring in a typical, traditional triple-decoction mashing 

programme (after Hind, 1950). ÐÐÐ , temperatures in the main mash vessel. temperatures 

in the mash copper during the first, second and third decoctions (1, 2 and 3). About one-third of the 

mash is used in each decoction. A `thick mash' or a mixed mash may be used. The `proteolysis' and 

other periods are the oversimplified, traditional names for the divisions of the process. Process 

duration, about six hours. 


122 

77 

Temperature (°F)

with arest at 65 70ëC (149 158ëF) to allow -amylase to liquefy the starch. After a 

period at 100ëC (212ëF), say 15min. for pale beers and 45min. for dark beers, the hot 

mixture is added back to the stirred main mash, increasing its temperature to 50 53ëC 

(122 127ëF). During the next rest the surviving enzymes begin to attack the gelatinized 

and liquefied starch and proteolysis continues relatively quickly. Asecond decoction, 

also with about athird of the mash, which may or may not have a`rest' during heating at 

about 65ëC (149ëF) to liquefy starch, increases the temperature of the main mash to 

about 65 70ëC (149 158ëF).Afinal decoctionincreasesthe temperature toabout 76ëC 

(169ëF) then, after ashort rest, the mash is transferred to the lauter tun or filter and wort 

collection begins. Athree-decoction mash may last six hours. As with other mashes the 

exact temperatures chosen, the duration of the rests and boils and the rates of mash 

heating and mixing can be varied. However, with well-modified malts such aprocess is 

unnecessary. It is too long, too complex and the three boils are too expensive. 

Many faster and more economical double- and single-decoction procedures are used. 

For example, malt may be mashed in at 35ëC (95ëF) and, after a short rest, the 

temperature of the mash is raised to around 52ëC (126ëF; Fig. 4.3). Two successive 

decoctions, each with aquarter of the mash, increase the main mash temperature to 

65 70ëC (149 158ëF) and then 76ëC (169ëF). The whole process takes about 4.5h. 

Many morerapid double-decoction processeshave been describedandineach case better 

modified malts are needed and the processes more nearly approach the conditions used in 

infusion mashing. For example, amash is prepared at 63ëC (145ëF) and after ashort 

stand about aquarter of the mash, possibly a`thick mash', is withdrawn and boiled (Fig. 

4.4). When it is mixed into the main mash the temperature is increased to 70ëC (158ëF). 

After arest of 45 60min. asecond decoction, also with about aquarter of the mash, is 

used to increase the main mash temperature to about 77ëC (171ëF). After about 30min. 

wort collection can begin. The whole process takes 2 3h. 

Single decoction mashes are even simpler. For the preparation of dark beers the mash 

may be given a preliminary long stand at a low temperature. This can allow the 


100 

75 

50 

25 

Proteolysis active 

1 2 


0 0 1 2 3 4 

Time (h) 

Enzyme 

inactivation 

Fig. 4.3 Atemperature scheme for atypical two-decoction mashing programme (after Hind, 

1950). The key is in Fig. 4.2. About aquarter of the mash is used in each decoction. Process 

duration, four and a half hours. 


212 

167 

122 

77 

32 

Temperature (°F)


100 

75 

50 

25 

Proteolysis 

1 2 


0 32 

0 1 2 3 

Time (h) 

Enzyme 

inactivation 

Fig. 4.4 Ashortened double-decoction process, lasting about two and ahalf hours (after Hind, 

1950). For the key, see Fig. 4.2. About aquarter of the mash is used in each decoction. 

multiplication of unwanted microbes and the development of unwanted flavours. Often 

the single decoction is coupled with atemperature-programmed period. For example, a 

mash is made at 35ëC (95ëF), then the temperature is successively raised to 50ëC 

(122ëF)andthen65 67ëC(149 152.7ëF)withrestsatthesetemperatures.Thenabouta 

third of the mash (a thick mash) is taken and heated, with arest at 70ëC (158ëF), to 

boiling. This is added back to the main mash, increasing the temperature to 75ëC 

(167ëF). An interesting variant is where amash is made at 35 or 50ëC (95 or 122ëF) and 

after astand the mash is allowed to settle and the relatively clear liquid, which contains 

enzymes,isheldwhilethethickmashisstirredanddirectlyheatedwithrestsat63 65ëC 

(145 149ëF) and 70 75ëC (158 167ëF), then the thick mash is heated to boiling. This 

boil will disrupt any solids and gelatinize any remaining starch granules but will 

inactivate enzymes. Then the thick mash is cooled to 65ëC (149ëF) and is recombined 

with the thin mash, which provides enzymes to attack the disrupted materials. The 

recombined mash is initially at about 67ëC (152.7ëF), and is successively warmed, by 

direct heating, to 70 and then 75ëC (158 and 167ëF). However, the cooling process 

wastes heat. 

Various special mashing programmes are used in Germany (Kunze, 1996; Narziss, 

1992a,b). In the jump-mash system (Springmaischverfahren), which is used to produce 

wort with alow fermentability, athick mash is prepared at 35 40ëC (95 104ëF). Then 

boiling water is stirred in over a15min. period to give atemperature of 72ëC (161.6ëF). 

By this means the grist is hydrated and some of the thermolabile enzymes have achance 

to act before the temperature is increased to permit starch liquefaction and dextrinization 

while minimizing saccharification. The mash temperature is increased to about 78ëC 

(172.4ëF) before wort collection. The wort has an attenuation limit of only about 40%. In 

the Kubessa process the grist is divided into flour, grits and husk fractions. The husk 

fraction is mashed separately at 50ëC (122ëF) and is held at this temperature while the 

flour and grits are mashed using a rising temperature programme, with rests at 

appropriate temperatures, until the mix is boiled. Then the two mashes are combined to 

giveamixed mashatabout70ëC(158ëF).Afterastandthetemperatureisraised to78ëC 

(172.4ëF) and the wort is collected. This process, which is little used, avoids boiling the 

huskmaterial and gives beer with abetter flavour. Inthe preparation oflow-carbohydrate 


212 

167 

122 

77 

Temperature (°F)

eers it is necessary to ferment as much of the wort carbohydrate as possible. Where the 

use of microbial enzymes is permitted this is achieved by adding fungal -amylase, with 

or without pullulanase, to the fermenting beer. In Germany highly fermentable wort may 

be made by using an exceptionally ìntensive' temperature-programmed mash, with rests 

at 50 ëC (122 ëF)/30 min.; 62 ëC (143.6 ëF)/45 min.; 65 ëC (149 ëF)/45 min.; 68 ëC 

(154.4 ëF)/30 min.; 70 ëC (158 ëF)/30 min.; 72 ëC(161.6 ëF)/15 min. and then mashing off 

at 73 74 ëC (163.4 165.2 ëF). This process takes 3.5 4 h. Even so the wort is not fully 

fermentable and it is necessary to add powdered highly diastatic malt or malt extract to 

the fermenting beer. These additions risk contaminating the beer with spoilage 

organisms. 

Decoction mashing is convenient when small amounts of adjuncts need to be cooked, 

since cooking can be carried out in the decoction vessel. The use of large quantities of 

adjuncts, such as rice, maize and sorghum grits, that require thorough cooking, combined 

with the availability of high-nitrogen, enzyme-rich malts gave rise, initially in North 

America, to the double-mash system. With this two mashes are prepared and then they 

are combined, often in a third vessel (Fig. 4.5). The adjuncts, which may comprise 

25 60% of the grist, are mashed in, in a cereal cooker, with a proportion (5 10%) of a 

highly enzymic malt (80 200 ëL) or a heat-stable bacterial -amylase, at about 35 ëC 

(95 ëF). The temperature is raised to about 70 ëC (158 ëF) and the malt starch is liquefied 

and liquefaction of the adjunct starch begins. The temperature is traditionally increased to 

boiling and is held at this temperature for about 45 min. However, depending on the grist 

and the enzymes used, it may be preferable to hold the temperature at 85 ëC (185 ëF), and 

so save the cost of the fuel needed for boiling. While the adjunct mash is in progress the 

malt mash is prepared by mashing-in (doughing-in) at about 35 ëC (95 ëF). After a rest of 

about one hour the two mashes are combined and mixed achieving a temperature of about 

68 ëC (154.4 ëF). After a stand of 15 30 min., when all the starch is saccharified, the 

mash is heated to around 73 ëC (163 ëF) by adding hot liquor or by steam injection, then it 

is lautered. 


100 

75 

50 

25 

Saccharification 

in first mash 


in second mash 

0 32 

0 1 2 3 

Time (h) 

Enzyme 

inactivation 

Fig. 4.5 A scheme of the temperatures found in a double-mash programme using a grits cooker 

(after Hind, 1950). the temperature of the adjuncts/grits mash in the cereal cooker. ÐÐÐ the 

temperature in the malt mash and in the combined mash during and after mixing. 


212 

167 

122 

77 

Temperature (°F)

In parts of Africa no barley malt is available and the sorghum malt is not suitable for 

making lager-types of beers with mashing schedules designed for barley malts (Chapter 

16). Under these circumstances mashes have been made with maize or sorghum grits 

converted with microbial enzymes. Sometimes 10 20% of sorghum malt has been 

included in these grists to provide soluble nitrogenous materials, but they are virtually 

àll-adjunct' mashes. 

When making some traditional Belgian top-fermented beers the temperature of the 

infusion mash may be increased by steam injection (De Clerck, 1957). Mashing-in may 

be at 45 50 ëC (113 122 ëF) then, after a 30 45 min. stand the temperature is increased 

to 62 63 ëC (143.6 145.4 ëF). After another rest of 30 45 min. the temperature is raised 

to 70 ëC (158 ëF) and then, after 30 45 min., to 75 ëC (167 ëF). After a pause wort 

collection is started. Temperature-programmed infusion mashing is being more widely 

used both in ale and lager breweries. The rising temperature programmes may be adjusted 

in many ways, allowing rests at any desired temperature. The stirred mash is heated in 

one vessel, sometimes with precautions to exclude air, and so the costs of a decoction 

vessel and heating parts of the mash to boiling are avoided, although an adjunct cooker 

may be needed. Boiling thick mashes is not practical and so, if undermodified malts or 

mashes with particular adjuncts are used, the mashing programmes must be extended and 

it is sometimes necessary to add microbial enzymes. For a mash made with an 

undermodified malt the temperature/time sequence might be 35 ëC (95 ëF)/30 min.; 50 ëC 

(122 ëF)/30 min.; 65 ëC (149 ëF)/30 min.; 70 ëC (158 ëF)/30 min., 75 ëC (167 ëF)/15 min. 

then mashing off, with the temperature rising between the rests at the rate of 1 ëC (1.8 ëF)/ 

min. For mashes being made with better modified malts the programme might start at 

48 50 ëC (118.4 122 ëF) and the durations of the different rests may be varied to 

achieve the desired quality of wort. Typically these mashes last 2 3 h. Often the mashing 

programme is chosen to produce a wort that is closely similar to one made by a decoction 

mashing programme (Hug and Pfenninger, 1979; Fig. 4.6). Temperature-programmed 

Temperature 

134.6°F 

125.6°F 

57°C 

52°C 

52°C 

52°C 

57°C 

72°C 

72°C 

62°C 

143.6°F 

62°C 

62°C 

168.8°F 

161.6°F 

76°C 

72°C 

100°C 

72°C 76°C 

100°C 100°C 

72°C 

76°C 

0 20 40 60 80 100 120 140 160 180 

Time (min.) 

Fig. 4.6 The temperature programmes of three mashes that yield very similar worts (after Hug and 

Pfenninger, 1979). The uppermost scheme is for a temperature-programmed infusion mash while 

the central scheme is for a single-decoction mash and the lowest is for a double-decoction mash. 


mashing is easily automated and is said to use 30 50% less energy than asimilar 

decoction mashing programme. However, while wort produced by different programmes 

may often be matched, this is not always the case. Decoction mashes tend to give darker 

worts with lower TSN levels and higher viscosities due to the non-starch polysaccharides 

dissolved during boiling. The husks are not boiled in infusion mashing. This is said to 

improve the flavour of the resultant beer. 

Novel `mixed' mashing systems are used in Belgium for preparing some traditional 

beers (De Clerck, 1957). These involve the use of large amounts of unmalted wheat. In 

one system the wheat is boiled in acopper, and then cooled and adiastatic extract is 

added to liquefy the starch. Separately the malt is mashed in amash tun and the wort is 

drawn off from the base. The wheat mash is transferred onto the top of the malt mash and 

the liquid is collected after it has filtered through the layer of malt. Other mashing 

systems involve producing `turbid worts' by mashing malt and wheat, mixed together, in 

a mash tun at 50ëC (122ëF), standing and then collecting the wort and the turbid 

supernatant in acopper. The re-mashing is repeated two or three more times, using 

progressively hotter liquor. Each time the wort is added to the copper. The copper 

contents are heated and held at 70ëC (158ëF) to saccharify suspended starch. This 

process, which produces turbid final worts, is needed to give Lambic beer its correct 

character. 

4.3 Altering mashing conditions 

4.3.1 The grist 

Maltandsomeadjuncts,suchastorrifiedwheat,mustbebrokenupbeforemashing.Until 

recently this was nearly always achieved by roller milling, but now hammer-milling is 

sometimes used (Chapter 5). The objective of milling is to break up the grist to give an 

acceptable range of particle sizes. The acceptable range is determined by the wort 

separation system being used. Often milling is carried out in such away as to minimize 

the break-up of husk material, as husk fragments help to give the mash an open structure 

andaidwortseparation.Indeed,manyyearsagooat huskswere addedtomashestoòpen 

them up'. Mash tuns require the coarsest grists, followed by various types of lauter tun, 

and then mash filters. While older types of mash filters required afine, roller-milled grist 

the newest designs use hammer-milled grists that are very fine indeed. There are reasons 

for using the most finely ground grist that can be processed with the equipment available. 

The finer the particles the faster they hydrate on mashing, the faster the pre-formed 

soluble substances dissolve and the faster the extract leaches from the particles during 

sparging. Furthermore, the enzymes have more ready access to their substrates in 

thoroughly disrupted grists. The surface/volume ratio of amaterial is larger the smaller 

and more numerous the particles into which it is divided and so afinely divided grist 

provides alarger surface area on which enzymes can act and across which substances can 

diffuse. To varying extents the particles will be pervious, permitting enzymes, substrates 

and the products of hydrolysis to enter and leave. In practice, finer grinding gives grists 

that, up to agiven `degree of fineness', yield higher extracts (Tables 4.1, 4.2). Finer 

grinding is less advantageous with better modified malts (this is the basis of the analytical 

fine-coarse extract difference determination), but it is beneficial with many (perhaps all) 

mash tun adjuncts. 

For many years attempts have been made to find ways of producing worts from finely 

ground, or `pulverized' grists. Decanter centrifuges, belt filters and rotary vacuum filters 


Table 4.1 Average laboratory extract values, obtained from 12 malts mashed isothermally at 

65 ëC (159 ëF) for one hour, using three settings of the BuÈhler-Miag disc mill. The smaller the gap 

between the grinding surfaces the finer the grind (Martin, 1979) 

Gaps between milling surfaces (mm) 0.2 0.5 0.7 

Hot water extract (l ë/kg) 296.2 293.5 291.4 

Confidence limits (95%) 2.0 1.9 1.9 

(In a previous trial the HWE values were 305.4, 302.3 and 300.0 respectively.) 

Table 4.2 Some analyses of two short-grown, experimental malts and a commercial malt ground 

with different degrees of fineness (Wackerbauer et al., 1993) 

Malts (days germination) 3 5 7 (Commercial) 

Friability (%) 62 83 91 

Whole corns (friabilimeter, %) 3 2 1 

Extract (coarse grind EBC,%) 73.9 78.6 79.7 

Extract (fine grind EBC, %) 80.2 80.8 80.9 

Extract (hammer milled, %) [78.6]* 81.4 81.7 

* The reason for this atypical low value is not clear. Possibly the mixing in of the very fine grist was uneven. 

(used with kieselguhr filter aids) have been tried for separating the wort from the spent 

grains.Theadventofthenewesttypesofmashfilters(Chapter6)haspermittedtheuseof 

very fine grists. The introduction of very fine grinds necessitates the alteration of the 

mashing schedule if an established product is to be `matched'. More finely ground grists 

are `converted' more quickly, saccharify faster, give higher extracts and sometimes the 

worts obtained are more fermentable and less turbid. Narziss (1992b) reported that malts 

ground coarsely, finely and powdered gave samples with extracts (% dry wt.) of 78.9, 

80.7 and 82.4, the worts having real attenuation limits (%) of 66.1, 65.1 and 65.3 

respectively. By using very fine grists the whole mashing process can be carried out more 

quickly and with better extract yields. The levels of TSN and FAN increase and, at least 

with some grists, the levels of soluble -glucans increase and wort viscosities increase 

(Pollock and Pool, 1968; Narziss, 1992a, b; Kunze, 1996). Sometimes the flavour of the 

beer produced is improved, perhaps because the shortened mashing times allow less 

poorly flavoured material to be extracted. It may be possible to use a higher proportion of 

adjuncts in a finely divided grist, and the fine grind is helpful when processing undermodified 

malts. With modern mash filters the extract recoveries from finely ground grists 

can equal, or exceed, laboratory extracts. 

Malt grists can be fractionated by sieving (screening) and/or by air classification. The 

fractions have different compositions and yield different worts when mashed separately. 

So, for example, fine flour derived mainly from the inner starchy endosperm saccharifies 

well when mashed alone, and yields an exceptionally high extract (e.g. 96%), and gives 

pale beers with very fine, pure and fresh flavours but lacking in body, low in phenolic 

tannins and resistant to the formation of chill haze. In contrast the fraction enriched with 

the outer parts of the starchy endosperm yields more soluble nitrogen, has an extract of 

about 80%, an intermediate colour and gives a beer that is full bodied and with a fresh 

flavour but having a harsh, clinging, astringent or bitter after-taste (Kieninger, 1969, 

1972). The husk fraction was obtained in smaller amounts. By combining fractions in 

different proportions different types of beer could be made. Also, by removing part or all 

of the husk material, brewing with the remainder of the grist would be more rapid, the 

beers produced would be paler and have a higher haze stability and a `finer' flavour 

(Vose, 1979). These fractionation processes take time and involves extra costs and 


produce ahusk-enriched fraction that needs to be used. Its use in cattle food would seem 

to be uneconomic and to use it, added to the grist, in the production of `normal lagers', 

would compromise their quality. Probably grist fractionation processes are not in 

commercial use. 

By using ahigh-impact mill, running at areduced speed, and then fractionating the 

gristovera1.60mmscreenKrottenthaleretal.,(1999)separated thegrist,whichinitially 

had an unusually wide range of particle sizes, into 79% finely ground and 21% more 

coarsely ground. The husk contents of the fractions seemed to be the same, and the 

extracts were almost identical. However, on analysis the coarse fraction took 55 60min. 

tosaccharify,comparedtothe10min.bythefinefraction,andthecoarsefractionyielded 

less FAN and total soluble nitrogen but substantially more -glucan and amore viscous 

wort. Thus the coarse material was derived from the under-modified portions of the malt. 

By mashing in the coarse material at 35ëC (95ëF) then temperature programming to 

65ëC (149ëF) then, after arest during which starch conversion should have occurred, 

combining the `coarse mash' with the mash of the fine fraction, that had been made at 

25ëC (77ëF) in asecond vessel, acombined mash temperature of 45ëC (113ëF) was 

obtained. After arest, during which the temperature-sensitive -glucanase surviving in 

the fine fraction mash should have operated, the temperature of the combined mash was 

increasedto65ëC(149ëF).Then,afterafurtherrestduringwhichthestarchfromthefine 

fraction should have been converted, the temperature was increased to 72ëC (161.6ëF), 

which was held for afurther period, allowing some final -amylolysis, before lautering. 

Comparedtothestandardmashingprogramme,slightlymoreextractwasrecoveredusing 

the fractionated grist mash (83.0 compared to 83.5%) the viscosity of the wort was 

usefully reduced (1.61 compared to 1.52 mPa.s for an 8.6% wort) and the -glucan 

content wasroughlyhalved.Suchatechniquemay beattractivewheretheReinheitsgebot 

or similar restrictions are in force. Asimilar result could be obtained more simply by 

adding afungal -glucanase/cellulase preparation to the unfractionated mash. 

4.3.2 Malts in mashing 

The choice of malts is dictated by the type of beer to be made and by quality 

considerations. Some qualities of different types of malt are indicated in Chapter 2. The 

brewer is faced with the problem that malts with the same traditional analyses may be 

different, and the differences can give rise to major problems in the brewery and in beer 

quality. Coloured and special malts' flavours change and decline with age and so these 

materials should be used fresh and their lab worts should be tasted and smelled to see that 

they are `normal'. Although chemical `marker' substances, such heterocyclic, nitrogencontaining 

Maillard products, have been sought, to allow flavour to be quantified 

indirectly by chemical analyses, this approach has had little success. Most attention has 

been paid to pale malts, since these make up the greater part of malt requirements, 

whether or not adjuncts are also used. In the first place each batch of malt should be as 

nearly identical as possible to earlier batches of the same type used successfully to make 

a particular product. This is not necessarily easily achieved. The old belief that some pale 

malts behave better when mashed after several weeks storage has been confirmed 

(Rennie and Ball, 1979). The ease of wort separation improves over a period of about 

three weeks and the clarity of the worts improves, but the reason(s) are unknown. 

As the available varieties of barley change the problem arises that they produce 

different malts when malted in one way. Some varieties give malts that give higher, or 

lower levels of hydrolytic enzymes, TSN or FAN relative to the yield of extract, or that 


give worts differing in fermentability or flavour and so on. Even comparatively small 

differences between samples of one variety of barley can cause differences during 

malting and in the quality of the malt produced. Irregularities in germination can lead to 

inhomogeneity, which is not always easy to detect, and may not be suspected until it has 

caused problems in the brewery. 

Brewers require amalt that mills easily to give the correct rangesof particle sizes, that 

converts in the `standard' time to consistently produce their standard wort, and allows 

easy, rapid and repeatable wort separation from the mash. The recovery of the extract 

should be as high as the equipment in use allows. These points are particularly important 

in automated plantand/orwhereahigh, fixed number of brews should becompleted each 

day and there is no spare time. The wort should have the correct characteristics for the 

beer being made and the beer made from it should be easy to filter and require minimum 

`stabilization treatments' (e.g. with silica hydrogel, PVPP adsorbents or additions of 

enzymes) to minimize haze formation or flavour deterioration and so have amaximum 

shelf-life.Thewortshouldhaveallthecomponents theyeastneedstoachieve arapidand 

complete fermentation. 

Brewers require analyses of each batch of malt. Regrettably different brewers have 

different requirements (Chapter 2). Normally analyses will include moisture content, 

colour, laboratory extract (HWE or E), total nitrogen (TN; protein), soluble nitrogen (or 

protein) and free amino nitrogen (FAN, determined by astated method). Sometimes, and 

often if adjuncts are being used, they will also ask for estimates of the diastatic power 

(DP) and -amylase activity. -Glucanase estimations may also be required. These are of 

minimum value if isothermal infusion mashing is being used unless barley adjuncts are 

included in the grist. However, they can be of use when deciding on the decoction or 

temperature-programmed mashing sequence to be used with under-modified malts or 

grists containing barley adjuncts. 

The most important analysis is the laboratory extract as, in general, the higher this is 

the better the quality of the malt. Allowance is made for the nitrogen content, which is 

inversely related to the extract yield, and for the variety of barley from which the malt is 

made. High total nitrogen contents are related to better foam characteristics in the beer 

andtohardermalts.AnadequateFANlevelisneededinaworttoensurethatyeastgrows 

well and that fermentation proceeds rapidly and completely. Too high values are not 

desired as this can lead to excessive and wasteful yeast growth. Higher values are 

required for malts that are going to be used together with adjuncts that act as nitrogen 

diluents as they contribute relatively little or no soluble nitrogen to the wort. Colour, or 

boiled wort colour, gives agood estimate of the colour of the beer and indicates what 

colour adjustment may be needed. 

Other tests are now often used in attempts to overcome the deficiencies of the 

traditional analyses (Briggs, 1998). Heterogeneity may be determined by scoring stained 

grain sections or by the determination of partially unmodified grains using the 

friabilimeter. The friability of the grain indicates the type of result that will be obtained 

when the malt is milled. Wort -glucan may be determined as may the residual -glucan 

in the malt and the viscosity of the wort. High values for fine grind-coarse grind 

laboratory extract differences indicate that malts are under-modified and that they may 

give rise to brewing problems (Table 4.2). At least one brewery has found that the fine 

grind, coarse grind and concentrated mash extract difference and the total malt -glucan 

content are inversely related to extract recovery in the brewhouse and that the viscosity of 

wort from a 70 ëC laboratory mash is correlated with the viscosity of strong brewery 

worts and so high values give warning of possible beer filtration problems and the 


occurrence of -glucan hazes and gels (Bourne and Wheeler, 1982, 1984; Bourne et al., 

1982). In another brewery, using traditional ale isothermal infusion mashing, the extract 

recoveries in the brewery were correlated with the hot water extracts and fine-grindcoarse-grind 

extract differences. So, from the laboratory measurements it was possible, 

by entering the values in the appropriate equation, to predict the brewhouse yield of 

extract (Maule and Crabb, 1980). Many correlations between laboratory analyses and 

brewery performances have been reported, but the correlation coefficients seem to vary 

significantly or tofail in different years and/ornot to be applicable todifferent breweries. 

Thus the performance of each brewing line needs to be evaluated and ways to predict its 

performance need to be assessed individually. 

`Problem malts', malts which give rise to brewing difficulties, are usually 

characterized by their `wrong' degrees of modification, either in all the grains or in a 

proportion of the grains, when the malt is inhomogeneous. As noted, the `correct' malt 

characteristics vary with the way it is to be used. In addition to inadequate enzyme levels 

under-modified malts are characterized by the inadequate breakdown of the endosperm 

cell walls. These unmodified regions resemble raw barley, and the problems associated 

with their presence resemble the difficulties encountered when raw barley is used as an 

adjunct. They are tough and, when the malt is milled, they give rise to coarse grits. The 

intact cell walls contain -glucan and pentosans and, as they `box in' the starch and 

protein of the cells, these are not degraded because enzymes cannot pass through the cell 

walls except where these are disrupted by milling or heating, as in decoctions. So 

undermodified malts give low extract recoveries, and the worts are often poorly 

fermentable. The levels of soluble nitrogen are low, the worts are viscous and rich in - 

glucans and wort run-off is slow. The -glucans may, or may not, deposit as gels or give 

rise to hazes, but they always seem to give difficulties with beer filtration. These 

problems can be minimized by adding microbial enzymes to the mash. In addition the 

beers may show protein-polyphenol haze and flavour instabilities. Nor are over-modified 

malts desirable. Besides the high malting losses accumulated during their production 

these give rise to too powdery grists when the malts are milled (impeding wort run-off) 

and although the yield of extract is good the quality can be poor. In particular the beer 

made from this malt is likely to lack body, the flavour may be poor, and the foaming 

characteristics will be bad. 

Different types of malt have different characteristics. For example, the more highly 

cured amalt is the lower its enzyme content (Table 4.3). The extract is slightly reduced 

by more curing, and the levels of soluble nitrogen are reduced, as shown by the decline 

in the nitrogen index of modification. S-Methyl methionine, SMM, the precursor of 

dimethyl sulphide, DMS, is destroyed and so kilning may be used to regulate the levels 

of this compound. The fermentability of the wort is reduced, giving beer with a lower 

alcohol content and more residual carbohydrate. The colour of the worts from more 

highly kilned malts are darker. Crystal malts and black malts are enzyme free and their 

inclusion in a mash reduces the fermentability of the wort. In making low-alcohol beers 

it is usual to mash well-cured malts with caramel malts at high temperatures to minimize 

saccharification, and so reduce the production of fermentable sugars. In addition 

experiments have been made in steaming green malts to cause enzyme destruction 

before kilning (Briggs, 1998). Curiously, the use of malts dried at low temperatures 

(40 ëC, 104 ëF) to a moisture content of 7 8%, which have high enzyme contents, seems 

not to occur although the reduction in kilning costs should make them less expensive. 

The use of undried, `green' malts is impractical in production brewing. Malts made with 

barleys containing a mutation that prevents the formation of anthocyanogen 


Table 4.3 Some effects of malt kilning on wort and beer analyses. The green malt was freeze 

dried before analysis. The kiln-dried samples were removed at successive stages of kilning (data of 

MacWilliam, 1972) 

Malt properties 

Freeze-dried Lightly Kilned Strongly 

kilned kilned 

Colour ± 3 6 13 

Moisture (%) 41.3* 3.7 2.9 2.3 

Hot water extract (lb/Qr) 104.0 102.9 102.2 101.9 

(l ë/kg) y 308 305 303 302 

Cold water extract (%) 19.4 19.5 19.1 17.1 

Diastatic power ( ëL) 131 98 68 48 

Total nitrogen content (%) 1.59 1.55 1.56 1.47 

Index of nitrogen modification (%) z 43.7 43.9 41.6 38.7 

[* Moisture content before freeze drying. y Approximate equivalents to the values in the older units. z PSN/TN]. 

Beer properties 

Total carbohydrate (g/l) 13.3 16.7 21.6 27.4 

Residual fermentable sugars (g/l) 1.1 2.8 1.3 4.8 

Non-fermentable carbohydrate (g/l) 12.2 13.9 20.3 22.6 

Total soluble nitrogen (mg/l) 526 645 593 580 

Bitterness (EBC units) 22.3 25.8 22.6 23.6 

Head retention (half-life, seconds) 91 106 93 101 

Colour (EBC units) 12 15 17 27 

Flavour (bottled) `Green malt' `Lager-like' `Pale ale' `Mild ale' 

polyphenols, which contribute to the formation of protein-tannin hazes, give beers that 

are extremely resistant to haze formation, are in limited use. However, proposals to 

make malts from low -glucan barley mutants or barleys that have been genetically 

modified to contain more heat-stable -amylase or -glucanase have not been carried 

out. In part this may be due to the sentiment opposing the use of genetically modified 

materials in brewing. 

Several brewing problems are associated with microbial infections of malts. Offflavours 

may occur and there is always the concern that mycotoxins may be present on 

poor malts. Particular attention has been paid to the possible presence of aflatoxins, 

ochratoxin, zearalenone, deoxynivalenol, fuminosins and citrinin, which can be produced 

by a range of fungi infecting barley (Scott, 1996). Some, such as citrinin, do not survive 

the brewing process, but others, such as deoxynivalenol, can survive into beer. Fungi also 

produce factors that cause gushing (over-foaming) in beers. The solution seems to be to 

avoid making malts from cereals that are heavily infected with fungi. Fungal infections 

are a considerable problem in tropical areas. High levels of bacteria on malt can also give 

problems. Bacteria multiply very greatly during malting, especially on the substances 

leached from split grains. Malts made with heavily infested barleys have, on mashing, 

given rise to very slow wort filtrations, possibly due to microbe-produced polysaccharides 

clogging the grain bed. Other malts have given worts having persistent hazes due to 

suspended dead bacteria, about 0.6 m in diameter (Walker et al., 1997). Another problem 

caused by microbial infestations of malts are the wild, and unpredictable fluctuations in 

the pH of worts (e.g. pH 5.45±6.06; Stars et al., 1993). Multiplication of lactic acid 

bacteria on the growing malt and particularly in the initial stages of kilning high-moisture 


green malts, was largely to blame. The malting process had to be modified to minimize 

this problem. 

The most uniform homogeneous malts are made by malting grain of one variety and 

onegrade.However,successivebatchesofmalt,madetomeetonespecificationinevitably 

differ slightly and so they may be mixed, or `blended', to meet abrewer's specification. 

Many European brewers regard blending malts of one grade but made from different 

barleys as unacceptable, even though in some other areas mixtures of barleys are malted. 

Brewers commonly mix different malts (pale, caramel, brown, etc.) to obtain the mix 

appropriate for aparticular beer. Narziss (1991, 1992a) gives examples of malt mixtures 

used to make many European types of beers. At the start of anew season the old season's 

malt, of any type, will increasingly be diluted with the new season's malt of the same type 

so that any consequent small differences in beer quality will not be noticed. Some types of 

blending are never acceptable. For example, suppose abeer is made with acoloured malt 

to give acolour 10. If the usual coloured malt is not available it is not acceptable to blend 

50:50 two malts of colours 5and 15. The colours may match in intensity (but probably 

not quality) but the flavour of the product will not match the original, since the àverage' 

mix of flavour substances will be different from that in the original malt. When two 

similarmaltsareblendeditisnecessarytobeabletopredictthewortqualitythattheblend 

will give (Moll, et al., 1982; Yamada and Yoshida, 1976). In general the extract and FAN 

values of mixtures vary linearly between the values of the individual malts according to 

their proportions in the mixures. But the fermentabilities of the worts will be better than 

predicted from simple proportions (synergism is shown) as enzymes from the more 

enzyme-rich malt partly compensate for the inadequate levels in the other. Since in one 

case the increase above the fermentability was due to increased levels of maltose and 

maltotriose this could have been due to the activity of limit dextrinase. 

4.3.3 Mashing with adjuncts 

The characteristics of commonly used adjuncts are summarized in Chapter 2and Briggs 

(1998). Like special, highly coloured barley malts mash tun adjuncts are deficient or 

totally lacking in the enzymes needed to convert the starch or degrade nitrogenous 

substances during mashing. Like the unmodified regions of pale malts, and especially chit 

malts, many adjuncts retain their cellular structure which must be disrupted by cooking or 

milling to allow enzymes access to the starch and protein during mashing. Malt extracts, 

sugars and syrups may be used as copper adjuncts to provide more extract, to create highgravity 

worts, to `dilute' soluble nitrogen and to adjust wort fermentabilities. In general, 

adjuncts are used to provide relatively inexpensive extract, to modify beer character 

(often resulting in less body, a more `delicate' and bland flavour or a less strong 

character), and to dilute the levels of lipids, polyphenols and soluble proteins from the 

malt giving a more haze-resistant, and sometimes more flavour-stable beer and to 

improve the head of the beer (Briggs, 1998; Martin, 1978). 

The proportions of adjuncts used in mashes vary widely, even within one country. In 

Bavaria no adjuncts may be used, in the UK 0 25% of wort extract may be derived from 

adjuncts (including copper adjuncts), while in the USA and Australia levels of 40 50% 

are common and sometimes may be higher. It is not possible to switch between adjuncts 

at will because the handling equipment and processing conditions that each require are 

different. Mash tun adjuncts can cause some brewing problems and extra costs, including 

the need for extended mashing schedules and/or the provision of a cereal cooker, the need 

for additions of microbial enzymes, slow wort separations and difficulties with beer 


filtration. As the proportion of adjunct in the mash increases so, at some stage, the 

quantities of enzymes available from the malt become inadequate for achieving a good 

starch conversion, and the recovery of extract declines, the fermentability of the wort 

falls, the levels of soluble nitrogen and free amino nitrogen, and even some inorganic 

substances may fall to such a level that yeast growth and fermentation may be impaired. 

In North America, using the double-mashing system (Section 4.2), up to 60% of maize or 

sorghum grits may be used in conjunction with highly enzymic, nitrogen-rich malt. Even 

with this system it may be advantageous to use microbial enzymes, such as heat-stable - 

amylase, to liquefy the grits in the cooker or a fungal saccharogenic amylase, pullulanase 

or amyloglucosidase to adjust wort fermentability. 

Green malt, used experimentally in simple infusion mashes, can convert large 

proportions of adjuncts exceptionally well giving highly fermentable worts with low 

levels of proanthocyanidins that give highly haze-resistant beers (MacWilliam et al., 

1963; Briggs et al., 1981). However, the difficulties of handling green malt, coupled with 

`raw-grain' flavour, have prevented this being used. In practice, highly enzymatic, highnitrogen 

pale malts made from selected varieties of barley, are best for converting 

adjuncts (Allen, 1987; Halford and Blake, 1972). 

The adjuncts used in simple infusion mashes either contain starches with low 

gelatinization temperatures or have been pre-cooked to pre-gelatinize the starch. These 

include raw barley, (sometimes regarded as the `natural adjunct'), wheat flour, torrified 

barley or wheat and flaked maize or rice grits. In addition, flaked barley or wheat may be 

used. Flaked maize and flaked rice were very popular with British brewers, but they are 

little used now, because of their costs. Some pre-cooked cereals give lower extracts than 

expected because during cooking some of the amylose is induced to crystallize and so 

become enzyme-resistant (Home et al., 1994). Provided that adjuncts are not used in 

excessive amounts, they are well mixed in with the malt at mashing in and the malt 

contains an adequate level of enzymes, they are relatively easy to use. The torrified grains 

can be milled with the malt. However, raw barley is tough and may need to be wet-milled 

or ground separately in a specially adjusted mill. `Barley brewing' is considered later. 

Wheat flour, like raw barley, can slow wort separation to a serious extent and although it 

has been used at malt replacement levels of up to 36% (Briggs et al., 1981; Geiger, 1972), 

in conventional infusion mashing levels of 5 10% are more usual. The inconvenience of 

the slow wort separation is partly offset by enhanced head formation and beer stability 

Wort filtration problems are reduced by grinding the wheat more coarsely, (e.g by 

hammer-milling it with a 2 3 mm screen) and by adding a mixture of pentosanase and 

cellulase enzymes to the mash (Forrest et al., 1985). 

Highly purified wheat starch is not troublesome but the technical grade usually used has 

associated protein and other materials which slow run-off. Higher extracts are recovered if 

wheat starches or flours are pre-soaked or pre-cooked at about 85 ëC (185 ëF) to allow 

gelatinization and liquefaction. The material is not boiled to avoid frothing. The flours are 

mashed-in at cool temperatures to avoid clumping. Raw barley, flaked barley and, to a 

lesser extent, torrified barley, like under-modified barley malts, release -glucan into the 

wort. This is associated with (but is not necessarily the major cause of) slow wort 

separation, but it causes worts to be viscous, it slows beer filtration, and sometimes causes 

hazes and the separation of polysaccharide gels from strong beers. To minimize these 

problems it is necessary to use malt containing adequate levels of -glucanase, or to add 

microbial -glucanase or cellulase to the mash (Crabb and Bathgate, 1973). In temperatureprogrammed 

mashing an appropriate low-temperature rest is used to allow the malt enzyme 

to act, or a heat-stable microbial -glucanase or cellulase can be added to the mash. 


In double-mashing the adjuncts used are usually maize-, or rice-grits, although sorghum 

grits, cleaned sorghum grains and other cereal preparations can be used. Rice grits tend to 

give a`drier' character to abeer, while maize grits confer a`rounder' and more `full' 

impression. Cooking is essential when the adjuncts have starches with high gelatinization 

temperatures(Table2.3).Ricegritsandfloursareveryvariableintheirqualitiesandcooked 

viscositiesmayvary100-foldandrun-offtimesthree-fold(Tengetal.,1983).Ricegritsare, 

perhaps, the most difficult adjuncts to use. On cooking, arice-mash may set to agel and 

becomeunmanageable.Sometimesitmaybepossibletobuyriceofanamedvariety,having 

favourablebrewingcharacteristics,butoftenthisisnotpossible,asonlymixturesofricesof 

unnamed varieties are available. Extract recovery is inversely proportional to the gel point, 

so ameasurement of the latter can be used to give an indication of the brewing value of a 

batch and how it should be mashed (Teng et al., 1983). If barley malt is the source of the 

liquefying -amylase then at temperatures over 78ëC (172.4ëF) the enzyme is quickly 

inactivated while some of the starch will not have been gelatinized. It may be necessary to 

employcomplexmashingscheduleswithheating,coolinganddecoctionsandmorethanone 

addition of malt (Kunze, 1996; Narziss, 1992a). Where the use of microbial enzymes is 

permittedtherice-mashmaybesupplementedwithacalciumsalt(e.g.100mgCa/litre)and 

the pH adjusted to 6.0. Then aheat-stable bacterial -amylase is added and the mixture is 

heated, with arest at the best temperature for the enzyme (e.g. 85ëC; 185ëF) to boiling. It 

may be preferable, and more economic, to use two enzyme preparations, aless expensive 

preparation of -amylase with acomparatively low temperature optimum (75ëC; 167ëF) 

and athermostable enzyme with an optimum temperature of 90ëC (194ëF), and then carry 

out the heating with rests at the two temperature optima. Cooking may be carried out under 

pressure, but at the elevated temperatures attained unwanted flavours, involving the 

formation of sulphur compounds, may be formed. At the end of the cooking period the pH 

should be adjusted to about 5.5, before mixing with the malt mash. 

Sorghum and maize grits are used as adjuncts with barley malt mashes. In some areas 

of Africa there is pressure to use indigenous cereals to make beers and the use of 

imported barley malt has been restricted or prevented. Some clear, lager-style beers have 

been made using malted sorghum. The poor and irregular quality of this material and the 

high gelatinization temperature of its starch combined with adeficiency of desirable 

enzymes means that microbial enzymes are routinely used in making most or all of the 

commercially prepared beers (Demuyakar et al., 1994; Hallgren, 1995; Lisbjerg and 

Nielsen, 1991; Little, 1994; MacFadden, 1989; Muts, 1991; Chapter 16). Since added 

enzymes are needed it is economic to use little or no sorghum malt and to mash 

exclusively with whole sorghum grains or sorghum grits and/or maize grits. With these 

mashes it is necessary to use mash filters for wort separation. The composition of 

sorghum grains is very variable, but may yield extracts of about 70%. Grits can yield an 

extract of over 90% and, at 1%, their lipid content is less than that of the grains at about 

3.5%. It is best to use pale grains. If the tannin-rich, coloured, `birdproof' types must be 

used then they should be given an alkaline wash with calcium hydroxide, sodium 

hydroxide or ammonia solutions to extract as much tannin as possible or a formaldehyde 

wash may be used to bind the tannins, although this is now disliked. Failure to control the 

tannins results in the inhibition of enzymes during mashing and possibly flavour 

problems. 

Sometimes a proportion of sorghum malt is included in the mashes to enhance the 

levels of assimilable nitrogen in the wort (Bajomo and Young, 1993). When making an 

all-grits mash the problem is to obtain the maximum extract, having the correct degree of 

fermentabiliy, with an adequate level of FAN to support yeast growth and fermentation. 


This is achieved by a careful choice of enzymes and temperature programmes. For 

example, the grits suspension is adjusted to pH6, and a calcium salt is added at 50 ëC 

(122 ëF; Lisbjerg and Nielsen, 1991), followed by an addition of a thermostable bacterial 

-amylase. The mash is heated to 76 ëC (168.8 ëF) then, after a 15 min. rest, is heated to 

boiling and is boiled for 30 min. It is then cooled to 52 ëC (125.6 ëF) by adding cold water. 

Then preparations of a protease and a saccharogenic fungal amylase are added. Following 

a 60 min. rest the pH is adjusted to 5.5, then the temperature is raised to 60 ëC (140 ëF) 

and this is held for 60 min. Finally, after increasing the temperature to 78 ëC (172.4 ëF) 

and a 10 min. hold, the mash is filtered. Such worts, if no sorghum malt has been used, 

are deficient in assimilable nitrogen and minerals and so need to be supplemented with 

`yeast foods' or autolysed yeast and perhaps mineral salts, including zinc, to achieve 

good fermentations. 

In the 1970s efforts were made to use increasing amounts of raw barley as an adjunct. 

Moderate amounts of various barley adjuncts are still used, but the use of 70%, or more, 

raw grain mashes (with attempts to use 100%), has been discontinued. The problems 

encountered in developing `barley brewing' can be summarized as difficulties in milling, 

inadequate extract recoveries, poorly fermentable worts, insufficient levels of assimilable 

nitrogen, poor foaming characteristics of the final beers, problems due to high levels of - 

glucans (slow wort separation, viscous worts, slow beer filtration, hazes and gel 

formation), higher wort pH values, lack of the correct beer characters and colours and the 

need for microbial enzymes and excessively long mashing times. On the other hand hop 

utilization increased when the worts were boiled and the beers were more resistant to 

haze formation. The barley had to be carefully screened and cleaned (and sometimes 

washed), and preferably had a low nitrogen content. Some varieties were preferable to 

others but inexpensive feed-grade grain might be used. In general the use of up to 50% 

barley in grists was regarded as successful, but the use of larger proportions seemed to 

increase the problems unduly. Wet milling was widely employed but dry milling of 

moisture conditioned grain and other methods were used (Wieg, 1987; Allen, 1987; 

Button and Palmer, 1974; Brenner, 1972). 

Various enzyme mixtures were employed. The -amylase, -glucanase and protease 

mixture from Bacillus subtilis was widely used. Wort fermentabilities were adjusted with 

additions of fungal saccharogenic amylase, pullulanase or amyloglucosidase or the use of 

higher proportions of pale, enzymic malts. Since that time other useful enzymes have 

become available. A successful mashing schedule, using 70% raw barley and additions of 

bacterial enzymes, involved wet milling and adding enzymes at mashing-in, which was at 

room temperature. Then the temperature was increased to 50 ëC (122 ëF). This was held 

for 60 min. Further increases were followed by rests at 63 ëC (145.4 ëF)/75 min., 65 ëC 

(149 ëF)/40 min., 68 ëC (154.4 ëF)/20 min. and then, after heating to 75 ëC (167 ëF) 

transferring to the lauter tun (Button and Palmer, 1974). The duration of this process was 

about five hours, which is too long to be economic for many breweries. 

4.3.4 The influences of mashing temperatures and times on wort quality 

A knowledge of the influences of mashing temperatures and times is essential to allow 

the logical choice of mashing conditions. The results of studies, made over many years, 

do not agree exactly. This is to be expected since different malts, grinds and thicknesses 

of the mashes were used. This section concentrates on all-malt mashes. Relative to 

laboratory mashes, brewery mashes are more concentrated (have a lower liquid/grist 

ratio), they are made with different degrees of milling, with salts in the brewing liquor 


and sometimes with pH adjustments. Under brewhouse conditions many key enzymes 

remain active for much longer than expected from laboratory studies. Some enzyme 

inhibition can occur through product inhibition and alimiting supply of `free' water (see 

next section). The simplest mashes are isothermal, and these are considered first. In 

production mashing, temperatures are often adjusted to allow temperature-sensitive 

enzymes to act before the temperature is increased to destructive levels. Temperature 

adjustments may be achieved by direct steam injection, underletting, decoctions, direct 

heating through the walls of the vessel, or the addition of hot water. At the end of the 

process, during wort recovery, the temperature is increased by the hot sparge liquor. 

Increasing the mash temperature increases the rate of chemical and enzyme catalysed 

reactions, acceleratestheratesofdenaturationand precipitationofproteins(including the 

inactivation of enzymes), accelerates dissolution and diffusion processes, accelerates 

mixing and, at least above acertain temperature, causes the gelatinization of starches and 

(at least during decoctions and adjunct boiling) disrupts the cellular structure of 

unmodified cereal endosperm tissues. Mixtures of enzymes are active in mashing and 

thesehavearangeofwidelydifferenttemperaturesensitivities.Enzymeactivitiesdecline 

as mashing proceeds, and so the temperature òptima' occur at lower temperatures as 

mashing proceeds (Fig. 4.1). Figures 4.7 and 4.8 illustrate that the optimum temperature 

for the production of permanently soluble nitrogen, which is dependent on amixture of 

enzymes, alters with the pH. Some of the substances extracted during mashing are 

preformedinthemalt.Thesemayormaynotbealteredasmashingproceedsandtheyare 

joinedbymaterialssolubilizedbyenzyme-catalysedhydrolyticreactions.Someestimates 

of temperature optima are shown in Table 4.4. 

Permanently soluble nitrogen (mg/100 g malt) 

300 

250 

200 

150 

100 

50 

3 h 

2 h 

1 h 

0.5 h 

0.25 h 

Mashing temperature (°F) 

86 104 122 140 158 176 

0 

0 10 20 30 40 50 60 70 80 

Mashing temperature (°C) 

Fig. 4.7 The influence of mashing time on the temperature optimum of the production of 

permanently soluble nitrogen (data of Windisch and Kolbach; after Moll et al., 1974). With 

increasing incubation time the temperature òptimum' declines (compare Fig. 4.1). 


Mashing time

Fig. 4.8 The influence of mashing temperature on the pH optimum of permanently soluble 

nitrogen formation (data of Windisch and Kolbach; after Moll et al., 1974). 

Table 4.4 Temperature optima for some mash processes, carried out for 2±3 h. Data from various 

sources (Briggs et al., 1981; Hind, 1950; Hopkins and Krause, 1947). The values can be 

substantially different under different mashing conditions 

Process ëC ëF 

Highest extract (mainly starch conversion) 65±68 149±154.4 

Fastest saccharification (dextrinization) 70 158 

Highest yield of reducing sugars 60±63 140±145.4 

Highest yield of fermentable extract 65 149 

Highest percentage fermentability (%) 63 145.4 

Highest yield of permanently soluble nitrogen 

(Mash times 1±3 h. Higher temp. optima for 

more concentrated mashes) 50±55 (60) 122±131 (140) 

Highest yield of formol-nitrogen 50±55 122±131 

Highest PSN minus formol-N 55±60 131±140 

Highest yield of àcid buffers' 45±55 113±131 

Maximum activity of -amylase 70 158 

Maximum activity of -amylase 60 140 

Maximum activity of -glucanase 40 104 

Maximum activity of phytase 50±60 122±140 

As noted previously, these optima are not true `constants'. At mashing temperatures 

the survival and activity of the proteolytic system of enzymes is extremely dependent on 

mash thickness. -Amylase is less stable and both -amylase and limit dextrinase are 

more stable than might be predicted from their behaviours in pure solutions. Most of the 

extract formed during mashing comes from the conversion of starch to a mixture of 


Extract (%) 

70 

50 

30 

10 

95 80 70 60 50 


soluble sugars, oligosaccharides and dextrins. It is not true that starch can be converted 

only at or above its gelatinization temperature, although this occurs most rapidly after the 

granules have been disrupted. It has been known for many years that in mashes made at 

55ëC (131ëF) over 90% of the potential extract can be recovered in two hours, although 

this temperature is well below the gelatinization temperature of barley starch (Table 2.3). 

Even at lower temperatures some starch conversion occurs, so malt enzymes can slowly 

degrade malt-starch granules (Fig. 4.9). When the time-courses of extract and 

permanently soluble nitrogen formation are followed it is seen that in amash made at 

65.5ëC (150ëF) the extract rises rapidly at first and most of the extract has been 

recovered in one hour but, in this instance, the maximum is not recovered until after 

1.5 2h. (Table 4.5). This limit is set by the nearly complete solubilization of the starch. 

In contrast the extract of the 49ëC (120.2ëF) mash is still rising after three hours, and 

extract recovery is not nearly complete. The amount of nitrogen solubilized during 

mashing never approaches all the nitrogenous substances in the malt, so the halt in the 

riseinPSNinthe65.5ëCmashafterabouttwohoursisduetoenzymeinactivation.Inthe 

49ëC mash the PSN is still increasing after three hours, and the value exceeds that 

achieved at 65.5ëC, provingthatadepletionofinitially insolublenitrogenous materials is 

not limiting. Even in three hours at 49ëC less that 40% of the total nitrogen has been 

40 

80 

Time (min.) 

Fig. 4.9 The interrelationships between the yields of extract, the mashing duration and the 

mashing temperature (after Schur et al., 1975). 

Table 4.5 Changes in yields with time of extract and permanently soluble nitrogen (PSN) in 

mashes made at two different temperatures (data of H. T. Brown, via Hind, 1950) 

Time (min.) 15 30 60 90 120 180 

Mash at 49 ëC (120 ëF) 

Extract (%) 20.3 24.8 28.2 30.3 34.0 37.1 

PSN (% TN) 24.6 28.0 32.2 34.6 36.5 39.0 

Mash at 65.5 ëC (150 ëF) 

Extract (%) 63.4 67.1 69.4 69.7 69.7 69.0 

PSN (% TN) 27.7 30.7 32.9 33.7 34.6 34.6 


160

Table 4.6 Extracts and fermentable extracts obtained in isothermal mashes during different 

incubation periods. (Data of Windisch, Kolbach und Schild, via Hopkins and Krause, 1947) 

Mashing period (min.) 15 30 60 120 180 

60 ëC (140 ëF) 

Extract (%) 50.2 53.4 57.2 60.7 62.2 

Fermentable extract (%) 36.0 39.0 43.1 47.9 50.2 

65 ëC (149 ëF) 

Extract (%) 60.6 62.2 62.8 63.6 63.6 


70 ëC (158 ëF) 

Extract (%) 61.2 62.5 62.9 63.4 63.6 


brought into solution. However, extract recoveries can be nearly complete at 

temperatures as high as 80ëC (176ëF), but fermentability may be as low as 30%, 

compared to 70%, or more, in worts from mashes made at 65ëC (149ëF) (Fig. 4.13 on 

page 117; Muller, 1991). 

At 80ëC (176ëF) sufficient -amylase remains for starch liquefaction and partial 

dextrinization to occur while -amylase and other heat labile enzymes are so rapidly 

destroyed that little saccharification can take place. In Table 4.6 the maximum extract 

recovery was nearly achieved in three hours, mashing at 60ëC (140ëF) and was achieved 

in about two and ahalf hours at 65 and 70ëC (149 and 158ëF). On the other hand the 

fermentable extract was still rising after three hours in the lowest temperature mash, was 

rising more slowly in the 65ëC (149ëF) after three hours, but had stopped increasing 

between one and two hours in the mash made at the highest temperature. Thus mash 

temperatures must be very carefully controlled if worts of one quality are to be produced. 

When the extract yields of isothermal mashes, made at temperatures between 0ëC (32ëF) 

and 80ëC (176ëF), are compared it is seen that the extracts increase relatively little with 

increasing temperature until 45 50ëC (113 122ëF) is reached. Then extracts rise 

sharply with increasing temperature up to about 55 60ëC (131 140ëF), then the rate of 

increase falls sharply and the maximum is achieved at 62 66ëC (143.6 150.8ëF). At 

higher temperatures there is aslow decline in extract recovery, at least up to 80ëC 

(176ëF;Windischetal.,1932). Therelationshipbetweenextract yieldandamorenarrow 

temperature range for different times is shown in Fig. 4.9. By using different series of 

increasing temperatures, in temperature-programmed mashes, wort composition can be 

adjusted to auseful extent, for example, to produce beers with different alcohol contents. 

Mashing for extended periods at low temperatures favours the formation of soluble 

nitrogen, the optimum temperature depending on the time (Fig. 4.7). Carbohydrates are 

the main contributors to extract. The formation of the major groups of carbohydrates 

during atemperature-programmed mash are illustrated in Fig. 4.10 (Enevoldsen, 1974). 

Notice that starch hydrolysis must have begun before the temperature reached 63 ëC 

(149 ëF). The `rest' temperatures used in mashing are chosen with reference to the 

temperature optima of key groups of enzyme-catalysed reactions. The low-temperature 

rests at about 45 50 ëC (113 122 ëF) are needed with undermodified malts when the 

breakdown of proteins and -glucans is to be encouraged. The rests at about 65 ëC 

(149 ëF) are to maximize starch conversion and production of fermentable sugars. The 

fermentable group of sugars includes glucose (4.1), fructose (4.2), sucrose (4.3) and 

maltotriose (4.5), but the major component is maltose (4.4). The non-fermentable 


Extract (g/100 ml) 

15 

10 

5 

52°C 

(125.6°F) 

63°C 

(145.4°F) 

0 

0 30 50 110 160 

Fig. 4.10 

Mashing time (min.) 

Changes in the yield of total extract (TE), the yield of total carbohydrate (TC), the 

fermentable sugars (FS) and the unfermentable dextrins (D) during atemperature-programmed 

mash (after Enevoldsen, 1974). 

carbohydrate fraction is chiefly a complex mixture of dextrins. Mannose (4.8) and 

galactose (4.9) are not released into solution. The extract increases at 52ëC (125.6ëF) but 

increases much more rapidly as the temperature increases to 63ëC (145.4ëF) and is held 

at this temperature. It continues to increase slightly up to the final temperature of 78ëC 

(172.4ëF) (Fig. 4.10). 

The changes that occur in yields of soluble nitrogen during temperature-programmed 

mashing are illustrated in Fig. 4.11. In the mash made at 35ëC (95ëF) the highest level of 

TSN was achieved after about 100min., then as the temperature continued to increase the 

level fell, due to the denaturation and precipitation of solubilized proteins. Beginning the 

programme at ahigher temperature, 50ëC (122ëF) yielded the same final amount of TSN 

at the end of the mash, but mashing in at 65ëC (149ëF) reduced the final amount of TSN 

significantly. The amounts of amino nitrogen formed fell as the mashing in temperature 

increased, so the mixture of nitrogenous substances in solution was altered by changes in 

the mashing regime. This result is due to the different temperature sensitivities of the 

enzymes involved. Prolonged mashing regimes are usually needed when under-modified 

malts are used, and the advantage in terms of yield of TSN is also shown in Fig. 4.11. 

The production of amino acids (measured by two methods) during a different 

temperature-programmed mash is shown in Fig. 4.12. It was long believed that 

proteolysis in mashing ceased at temperatures above about 60ëC (140ëF). This is not so. 

Proteolysis is most rapid at lower, `conventional' protein rest temperatures, but it does 

not cease immediately the temperature rises. At these lower temperatures phytase and - 

glucanase also continue to act. By mashing amalt in different ways worts with different 

qualities are produced (Table 4.7). By encouraging proteolysis the colours of worts are 

increased, probably because the elevated levels of nitrogen-containing substances favour 

melanoidin formation during the copper boil. More complete proteolysis reduces the 

amount of break formed and this enhances the hop utilization, since less of the bitter 

substances are deposited with the coagulated protein (trub). Less trub (sludge) is formed 


T.E. 

T.C. 

D 

F.S. 

78°C 

(172.4°F)

TSN (mg/100 g of dry malt) 

α-Amino N (mg/100 g of dry malt) 

850 

800 

750 

700 

650 

600 

550 

500 

200 

190 

180 

170 

160 

150 

(a) 

(b) 

35°C 

95°F 

35°C Undermodified malt 

0 60 

Time (min.) 

120 180 

50°C 

122°F 

65°C 

149°F 

158°F 

75°C 

167°F 

0 60 

Time (min.) 

120 180 

Fig. 4.11 Increases in the levels of the (a) total soluble nitrogen and (b) the -amino nitrogen 

during temperature-programmed mashing. The different mashes were started at the temperatures 

shown and then the programme was completed (after Narziss, 1977). The discontinuous line in (a) 

shows the results obtained with a different, undermodified malt. 

in worts from decoction mashes, probably because some of the protein has been 

denatured and precipitated in the boiled parts of the goods. 

4.3.5 Non-malt enzymes in mashing 

The mainly microbial, non-malt enzymes used in brewing are indicated in Chapter 2. In 

this section only the addition of enzymes to mashes is considered. Because of the absence 

of agreed methods of assay and the presence of èxtra' unspecified and/or unquantified 


70°C 

35°C 

50°C 

65°C 

35°C 

50°C 

65°C

Amino acids (μmoles/g dry wt) 

80 

70 

60 

50 

45°C 

(113°F) 

55°C 

(131°F) 

TCA-sol. α-NH 2 

0 0 30 60 90 120 150 


70°C 

F.A.A. 

(158°F) 

Fig. 4.12 The formation of amino acids during a temperature-programmed mash (after Enari, 

1974). The àmino acids' were estimated in two different ways, which gave discordant results. 

Although amino acid liberation proceeded fastest at low temperatures it was still going on at the end 

of the mash at the highest temperature. (FAA, free amino acids; TCA sol- -NH 2, amino groups 

soluble in trichloroacetic acid.) 

enzymes in enzyme preparations it is essential that brewers test each preparation on the 

laboratory and pilot-plant scales before introducing it into the brewhouse. The dose-rate 

is critical for achieving a particular result, and this must be financially worthwhile. 

Difficulties arise when a preparation that has been correctly standardized on one enzyme 

activity is used but the activity of an important `secondary' enzyme differs between 

batches or declines at a different rate during storage. It is possible to store particular 

preparations at `room temperature' for some months but in principle, all enzyme 

preparations should be stored cold to minimize rates of deterioration. The use of added 

enzymes when mashing adjuncts has been considered and in the cases of barley- and 

grits-mashing there is a need for enzymes to cause starch liquefaction and 

saccharification, -glucan degradation and protein hydrolysis. The raw barley contains 

bound and soluble -amylase, the bound form of which may, at least experimentally, be 

partly activated and solubilized by additions of the amino acid cysteine hydrochloride, or 

other thiol-containing substances or sulphite, making this enzyme more useful in 

mashing. Raw barley contains proteins which can inhibit some microbial proteases. The 

difficulty in obtaining sufficient assimilable nitrogen from barley- and grits-mashes 

underlines the absence of inexpensive peptidase preparations suitable for use in mashing. 

Enzymes may be added to mashes to speed starch conversion, to degrade -glucans, or 

to accelerate wort run-off. Various enzymes are adequately active at mash pHs but these 

often do not coincide with the pH optima of the enzymes. Usually the mash pH is adjusted 

only once at the onset of mashing. The exception is when grits are being cooked the pH is 

adjusted to 6 to allow a heat-stable bacterial -amylase to act. Afterwards the pH must be 


Table 4.7 A comparison of worts made in a Bavarian brewery, with an undermodified, 1931 malt (u) and a well-modified, 1932 malt (m) using four different 

mashing programmes (data of LuÈers et al., 1934). Pi, inorganic phosphate as % of the total in the malt. TSN, total soluble nitrogen, and amino-nitrogen as % of dry 

extract. In all heating periods, including decoctions, there were rests at 50 ëC (122 ëF), and 65±68ëC (149±154.4 ëF). Worts were collected at 76 ëC (168.8 ëF) 

Wort Wort 

Extract attenuation viscosity TSN Amino-N Pi Sludge 

(% dm) (% dm) (%) (relative) (% dry extr.) (% dry extr.) (%, total) (kg dry) 

Malt modification u m u m u m u m u m m u m 

Two-mash process. Mash in at 50 ëC 

(122 ëF). Decoctions to 67 ëC (152.5 ëF) 

and 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 

Three-mash process. Mash in at 35.5 ëC 

(96 ëF). Decoctions to 50 ëC (122 ëF), 

65 ëC (149 ëF) and 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 

`High-quick mash process'. Mash in 

at 68 ëC (154.5 ëF), then raise to 76 ëC 

(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 

Temperature-programmed, infusion-mash. 

Mashed in at 35 ëC (95 ëF). Heat with 

steam-coils, with rests at 50 ëC (122 ëF), 

67 ëC (152.5 ëF) and 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 


eadjusted downwards to about 5.5 before adding to the malt mash or saccharifying with a 

fungal saccharogenic amylase or -amylase, with or without pullulanase. The low 

temperature stabilities of these enzymes is inconvenient and may necessitate cooling the 

mash to about 50 55ëC (122 131ëF) to allow them to act to an adequate extent. The 

choice of which bacterial -amylase to use is largely dictated by their stabilities. The 

Bacillussubtilisenzymeworkswellat70ëC(158ëF)andthemorestableenzymefromthe 

B. subtilis var. amyloliquefaciens, is useful in cookers. It is inactivated on boiling, and this 

is an advantage. If amore stable enzyme, from B. licheniformis, is used it must be under 

conditionsthatensureitsinactivationduringthecopperboilorthecompositionofthewort 

andbeerwilldriftwithtimeasdegradationoftheresidualdextrinscontinues.Theaddition 

of uneconomically large amounts of amixture of bacterial -amylase and -glucanase 

(specific for mixed-link -glucans) can increase the laboratory extract of asound malt 

obtained with an extended mashing schedule, e.g., from 308 to about 313 lë/kg (Albini et 

al., 1987). The major use of these preparations is to offset the deficiencies of undermodified 

malts and the presence of barley adjuncts. Most preparations also contain 

proteases which, while they can usefully increase the level of soluble nitrogen in the wort, 

can reduce the head formation and stability in the final beer. 

In modern breweries with very short `turn-round times' it is essential that wort 

separation occurs rapidly. Slow wort separation may be associated with using poor or 

inhomogeneous malt. Of the various enzymes tested the preparations of the relatively 

heat-stable cellulases, e.g., from Trichoderma viride (50 55ëC; 122 131ëF), Penicilliumfuniculosum(65ëC;149ëF)andP.emersonii(80ëC;176ëF)seemtobesuccessfulin 

accelerating wort separation. These preparations contain complex mixtures of enzymes 

that attack mixed-link glucans and holocellulose (polysaccharides insoluble in hot 

sodium hydroxide solution) and may also have pentosanase activity. These preparations 

sometimes improve extract recovery. In mashes containing awheat adjunct it may be 

necessary to supplement the mash with a mixture of cellulase and pentosanase 

preparations to achieve asufficiently fast wort separation. 

4.3.6 Mashing liquor and mash pH 

Mashing liquor must be free of taints, and must be potable. In addition it must be free of 

many substances and organisms which might reduce beer quality. The quality of the 

water received must be checked regularly, whatever the source and, if necessary, it must 

be treated to convert it to the proper quality for the beers to be brewed. For mashing and 

beer dilution liquors oxygen may have been removed and dilution liquors may have been 

charged with carbon dioxide, with or without nitrogen (Chapter 3). The ratio of 

temporary to permanent hardness, the total amount of hardness and the amounts of ions 

that may influence flavour must be regulated (Narziss, 1992a; MacWilliam, 1975; Taylor, 

1981). Interactions between calcium (and to a lesser extent magnesium) ions and wort 

components have an important effect on mash and wort pH values. Thus bicarbonate ions 

effectively remove hydrogen ions and so, indirectly, raise the pH: 

HCO3 ‡ H ‡ $ H2CO3 ! CO2 " ‡ H2O 

Calcium ions (and magnesium ions to a lesser extent) interact with mash components 

such as inorganic phosphate, phytic acid and less phosphorylated inositol phosphates, 

peptides, proteins and probably with other substances displacing hydrogen ions into the 

mash and reducing the pH. For example, 


3Ca 2‡ ‡2HPO4 2 !2H ‡ ‡Ca3…PO4† 2 # 

The calcium phosphate tends to precipitate and precipitates more rapidly from wort at 

higher temperatures. Thus the pH of mashes decline faster during decoction mashing and 

later the pH of the wort declinesfurther during the hop-boil.Of the calcium ions added to 

amash 40 60% are retained in the spent grains. 

Amajor difficulty follows from the habit of measuring the pH of worts or mashes at 

roomtemperatureandassumingthatthesevaluesapplyathighertemperatures,whenthey 

do not (Hopkins and Krause, 1947). Weak acids, like water (see Appendix), dissociate 

more as the temperature rises and so the pH values of their solutions fall, like the pH 

values of mashes (Table 4.8). Thus at 65 ëC (149 ëF) the pH of a wort is likely to be about 

0.35 pH unit lower than at room temperature and 0.45 lower at 78 ëC (172.4 ëF). As the 

temperature of a mash changes (decoctions, temperature programming, sparging) so will 

the pH. These differences are significant, yet in many reports it is unclear if pH values 

have been determined at wort- or mash-temperatures or on cooled samples. Probably the 

latter is most usual. The pH optimum of -amylase, determined at room temperature, is 

about pH 5.3, but its optimum estimated from mashing experiments is often reported to be 

about 5.7. This error is due to the pH having been determined on the mash after it was 

cooled, when the pH had risen. Because of this difficulty the pH optima of changes 

occurring in mashes are a little uncertain (Table 4.9). 

Mashing pale malt in distilled water usually gives a wort with a pH of about 5.8 6.0, 

this value being maintained by the buffer substances (including phosphates and proteins) 

Table 4.8 Changes in the pH values of two mashes, made with distilled water and moderately 

carbonated water, measured at the temperatures shown (after Hopkins and Krause, 1947) 

Temperature of measurements pH values of the mashes 

( ëC) ( ëF) (Distilled water) (Carbonated water) 

18 64.4 5.84 6.03 

35 95 5.70 5.90 

52 125.6 5.65 5.80 

65 149 5.50 5.70 

78 172.4 5.40 5.55 

Table 4.9 Òptimal' pH values for `normal' isothermal infusion mashes made with pale malts 

lasting 1±2 h. at 65.5 ëC (150 ëF). Data from various sources. As far as possible the temperatures 

(mash temperature, m , and cooled wort, w), at which the pH values were determined are indicated 

(see text for a warning) 

Characteristic Òptimal' pH 

Shortest saccharification (dextrinization) time 5.3 m±5.7 w 

Greatest extract obtained 5.2±5.4 m? 

Greatest extract from a decoction mash 5.3 m±5.6 m? 

The most fermentable wort 5.1±5.3 m?; 5.4±5.6 w? 

Mash impossible to filter < 4.7 

-Amylase most active (+ Ca 2+ ) 5.3 m±5.7 w 

-Amylase most active 5.1±5.3 (4.7?) 

Maximum yield of PSN 4.4±4.6 m; 4.9±5.1 w 

Maximum yield of formol-N 4.4±4.6 m, 4.9±5.2 w 

Maximum protease activity (depends on substrate) 4.3 m; 4.6±5.0 m? 

Maximum phytase activity about 5.2 m 

Carboxypeptidase activity maximal 4.8±5.7 


from the grist. Infusion mashes are best carried out at pH 5.2 5.4 (mash temperature), 

and so will give cooled worts with pH values of about 5.5 5.8. It has been recommended 

that decoction mashes should not give worts with pH values less than 5.5. Lowering the 

pH too much results in increases in the soluble nitrogenous materials but lengthens the 

saccharification time and lowers the yield of extract. Lowering the pH to the correct 

extent, by additions of calcium salts or other means, speeds the rate of starch degradation, 

enhances the activities of other carbohydrases and the proteolytic mixture of enzymes so 

that TSN and FAN values are increased and wort colour is reduced. The solubility 

characteristics of some proteins are altered, the buffering power of the wort is increased 

and, at later stages of brewing, hop utilization is decreased. Conversely the increase in pH 

caused by mashing with waters rich in bicarbonate ions is generally undesirable (Table 

4.10). The pH values of mashes are conveniently lowered by mashing with `permanently 

hard' water, either natural or that has been `Burtonized' by additions of calcium sulphate 

and/or calcium chloride (Tables 4.11, 4.12). Often about 100 mg of calcium is added to 

each litre of liquor. However, other means may be adopted such as the direct addition of 

sulphuric, phosphoric or lactic acids (where this is permitted) or by the use of lactic acid 

malts or acidified worts. 

Table 4.10 The effects of water hardness on the pH values of the cold water extracts and cooled 

worts prepared by decoction mashing (after Hopkins and Krause, 1947) 

Nature of water used pH of CWE or wort 

Distilled water; cold water extract (CWE) 6.2±6.3 

Wort, water with temporary hardness (about 15 grains CaCO 3/gal., 

214 mg/litre) 5.89 

Wort, distilled water 5.76 

Wort, water with permanent hardness (about 4 grains CaSO 4/gal., 

57 mg/litre) 5.65 

Table 4.11 The amounts of salts (shown in the anhydrous forms) added to some British brewing 

liquors (Comrie, 1967). Larger amounts of calcium salts may be added to offset the presence of 

bicarbonate ions. The amounts are varied to allow for alterations in mash thickness and to achieve 

the desired final flavours. (1 grain/imperial gallon (UK) ˆ 14.21 mg/litre) 

Pale ales Mild ales Stouts 

Salts grains/gal mg/l grains/gal mg/l grains/gal mg/l 

CaSO4 15.7±31.5 223±448 5±10 71±142 none none 

CaCl 2 6.9±13.7 98±195 7±14 99±199 5.5±11 78±156 

MgSO4 2.5 36 2.5 36 2.5 36 

NaCl 2.5±5.0 36±71 5±10 71±142 7±12 99±171 

Table 4.12 The effects of calcium ions, added as calcium chloride, on the pH, extract and soluble 

nitrogen fractions given by mashes made with one malt at 65 ëC (149 ëF; Taylor, 1981) 

pH Extract TSN FAN 

(mg/litre) (litreë/kg) (ppm) (ppm) 

AddedCa 2+ 

0 5.74 287 904 188 

100 5.48 291 973 195 

200 5.39 292 983 207 

300 5.28 292 1062 220 

The pH values were measured at room temperature. TSN and FAN were adjusted to a wort concentration of SG 

1040. 


Table4.13 Someoftheinfluencesofaddinggypsum(calciumsulphate,CaSO4.2H2O. This contains 

23.28% Ca, by weight) to the liquor when mashing malt (data of Hind, via Briggs et al., 1981) 

Gypsum added Extract (l ë/kg) Unboiled wort Boiled wort 

(mg/litre liquor) 

(Apparent) (Corrected Ash Phosphates Phosphate 

for ash (mg/100 ml) as P2O5 as P2O5 

content) (mg/100 ml) (mg/100 ml) 

0 296.7 286.4 138 70 68 

380 300.5 289.3 148 63 59 

760 302.9 290.4 167 56 54 

1140 305.7 290.6 197 54 50 

Additions of calcium ions to the mash reduce the quantities of phosphates in solution 

but apparently not to undesirable extents (Table 4.13). In addition to the advantages 

achieved by favourable pH adjustments the calcium ions stabilize -amylase during 

mashing, accelerate wort separation and run-off from the mash, assist in break formation 

in the hop-boil and the beer clarifies better, yeast flocculation is favoured and calcium 

oxalate crystals (which can be deposited on the walls of fermenters as `beer stone') are 

precipitated and so the potentially toxic oxalic acid (4.151) does not go forward. In the 

beer the calcium oxalate may give rise to haze or initiate gushing. Where the 

Reinheitsgebot and similar laws operate, the addition of `chemicals' is not permitted and 

with some beers (e.g. Pilsen-style lagers) the brewing liquor must be soft. In these cases 

the adjustment of mash pH values is achieved by the use of biologically prepared lactic 

acid introduced into the mash either as acid malt or as acidified wort. Acid malts are 

prepared in various ways (Briggs, 1998), and carry 1 5% (typically 2%) lactic acid and, 

on being mashed alone, give awort with apH in the range 3.8 4.4. The pKa of lactic 

acid is 3.86 at 25ëC. Usual additions are about 5% of the grist but larger quantities may 

be used. Some beers derive part of their character from the lactic acid they contain. An 

alternative is to acidify unhopped first wort by incubating it with thermophilic lactic acid 

bacteria (Lactobacillus delbruÈckii, L. amylolyticus) at 45 47ëC (113 116.6ëF) for 

8 71h(Oliver-Daumen, 1988). Batch, semi-continuous and continuous acidification 

plants are available, the batch type being the most common. It is possible to calculate the 

amount of acid needing to be added to achieve adesired reduction in pH. 

During lautering (wort collection) buffers are washed out of the mash and there is a 

tendency for the pH to rise, particularly if bicarbonate is present in the hot (e.g. 

75 80ëC; 167 176ëF) sparge liquor. This is highly undesirable as at the higher pH 

unwanted polyphenols and flavour substances are leached from the goods. This may 

make it necessary to treat the weaker last runnings with active charcoal (to remove 

unwantedsubstances), beforethey areadded backtoasubsequentmashoraretransferred 

tothecoppertobeboiled.ItispreferabletomakesurethatanyriseinpHduringsparging 

is minimal by excluding the use of water containing bicarbonates and by ensuring that 

adequate levels of calcium ions are present. 

4.3.7 Mash thickness, extract yield and wort quality 

Changes in mash thickness (liquor/grist ratio) have significant effects on mash 

performance (Hind, 1950; Hopkins and Krause, 1947; Harris and MacWilliam, 1961; 

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), 

with a duration of 180 min. (Data of Windisch, Kolbach and Schild, via Hopkins and Krause, 1947) 

Concentration of mash (water:malt) 2:1 2.7:1 4.0:1 5.3:1 

Extract (% dry malt) 71.7 77.0 80.0 79.9 

Fermentable extract (% dry malt) 52.3 56.3 58.5 57.8 

Fermentable extract (% total extract) 72.9 73.1 73.1 72.3 

Permanently soluble nitrogen (% dry malt) 0.57 0.56 0.54 0.53 

Formol nitrogen (% dry malt) 0.22 0.21 0.20 0.19 

down, worts are more concentrated and viscous, TSN and FAN are increased and more 

high molecular weight nitrogenous substances remain in solution, but a lower proportion 

of hydrophobic peptides (relative to the amount of extract) are present, causing `high 

gravity' beers to have poor head retentions (Bryce et al., 1997). In the concentrated 

mashes both the enzymes and their substrates are more concentrated. Some enzymes 

(proteolytic enzymes, disaccharidases) are more stable in concentrated mashes producing 

higher proportions of TSN and hexose sugars respectively. At high mashing temperatures 

thicker mashes give worts with higher fermentabilities (Muller, 1991; Fig. 4.13). On the 

Starch (%) Fermentability (%) Carbohydrate (%) 

12 

10 

8 

6 

80 

60 

40 

20 

0 

1.6 

1.2 

70°, 75°, 

80°, 85 °C 

70°C 

75°C 

80°C 

85°C 

85°C 

0.8 

80°C 

0.4 

75°C 

70°C 

0 

50 100 150 200 250 300 350 400 

Water (ml)/malt (50 g) 

Fig. 4.13 The influences of mash thickness and mash temperature, during 1 h. isothermal 

mashing, on (upper) the yield of carbohydrate in the wort, (middle) the fermentability of the wort 

and (lower) the starch present in the wort (after Muller, 1991). 


Extract (L°/kg) 

300 

250 

200 

150 

130 

0 1 2 3 4 5 6 7 

Water (ml)/Grist (g) ratio 

Potato starch 

Wheat starch 

Controls (all malt) 

Barley 

Maize starch 

Fig. 4.14 The effects of mash thickness on the extract recoveries from mashes made with all-malt 

or 50:50 mixtures of malt and the adjuncts indicated (data of Muller, 1991). 

other hand, at `normal' mashing temperatures weaker mashes give more fermentable 

worts. The high concentrations of sugars and dextrins present in thick mashes can inhibit 

the amylases. Enzyme inhibition is due to the reduced availability of free water as well as 

to the sugars acting as competitive inhibitors. Brewery worts contain 0 40% more 

soluble nitrogen than laboratory analytical worts. It was reported that mashes made with 

39% solids give worts with maximum extract yields while worts with the highest 

fermentabilities are given by mashes made with 16 32% solids. The effects of mash 

concentration on extract yield are also present when adjuncts are included in the mash 

(Harris and MacWilliam, 1961; Muller, 1991; Fig. 4.14). 

As the grist hydrates water is bound, and there is a rise in temperature caused by the 

release of heat (the `heat of hydration'). As the mash proceeds water is utilized in 

hydrolyses, a water molecule being consumed when any bond is split. Some water is 

more or less firmly bound (by hydrogen bonding) to starch, to sugars in solution, to - 

glucans, to pentosans and to other substances reducing the concentration of `free' water. 

In all-malt mashes and mashes made with 50 : 50 malt and barley or wheat starch the 

extract recovered falls very sharply as the liquor/grist ratio is reduced below about 2.5 

(Fig. 4.14). Generally, altering the liquor/grist ratio at values over 3 has comparatively 

minor effects, but these are not necessarily negligible. In a particular case mashing with a 

liquor/grist ratio of 2.5 : 1 gave an extract of 291 l ë/kg, while at a ratio of 7 : 1 the extract 

was 311 l ë/kg. The extent of water binding becomes progressively greater as mashes 

become more concentrated and there is insufficient free water to permit the gelatinization 

of much of the starch. The addition of more enzymes to a very thick mash does not 

quickly convert the ungelatinized starch and so does not enhance the extract obtained. 

The situation with the maize starch (Fig. 4.14) is complicated because its gelatinization 

temperature (70 75 ëC; 158 167 ëF) is above that of the mashing temperature (65 ëC; 

149 ëF) and so the conversion of the starch into extract is relatively slow. The potato 

starch had a wide gelatinization temperature range (56 69 ëC; 132.8 156.2 ëF), which 

spanned the temperature of the mash, and the pattern of extract recovery was different 

again (Fig. 4.14). 


4.3.8 Wort separation and sparging 

At the end of the mash the wort is separated from the residual solids. This may be arapid 

process, asinmashfilters,oritmaytake1.5 2.5hinsomelautertunsor4 18hinmash 

tuns. An extended run-off period allows residual enzymes to continue acting for at least 

part of the time. When amash tun or lauter tun is used, the first wort to emerge is diluted 

with the water that was originally under the plates. The first runnings are generally 

returned to the top of the mash and the wort is recycled until it is completely clear and 

`runs bright'. Then wort collection begins, the strong wort emerges and gradually the 

mash settles onto the plates. Sparging is started and the liquor, sprayed onto the surface 

fromrotatingsparge-arms,permeatesdownthroughthegoods,progressivelyleachingout 

and carrying away the remaining extract. In mash tuns this raises the temperature, so the 

temperature of the final wort is about 74ëC (165ëF). In contrast, in two- and three-vessel 

mashing systems the temperature of the whole mash is usually raised to the sparging 

temperature and after wort recirculation (if this isused), the first wort is collected and the 

sparge liquor is applied at the same temperature (e.g. 75 78ëC; 167 172.4ëF). 

Although sparging temperatures of up to 80ëC (176ëF) may be used, and the use of 

even higher temperatures has been proposed, these are usually avoided because 

undesirable flavours and unwanted substances, such as undegraded starch and 

hemicelluloses, may be eluted from the goods. This is particularly likely if undermodified 

malt or raw cereal adjuncts have been used. At these elevated temperatures 

enzyme destruction is rapid, the rates of diffusion of extract materials from the grist 

particles is rapid, the rate of wort separation (`filtration') occurs faster, more protein 

aggregation occurs and wort viscosity is reduced. 

As run off progresses the quality and concentration of the wort declines. The last 

runnings contain extract that has acomparatively poor quality (Hind, 1950; Figs 4.15, 

4.16).Relativetotheextractmorehigh-andlow-molecularweightnitrogenousmaterials, 

ash (including phosphates), silicates (mostly from the silica in the malt husk), 

polyphenols and astringent substances are dissolved, all these being favoured by the 

increasing pH. The specific gravity of the wort rises then declines as the sparge liquor 

emerges. As the wort is diluted the fermentability initially increases and finally falls 

sharply. Often the pH rises, (e.g. by 0.2 0.7), as the buffering substances are eluted from 

the goods. The rise is particularly marked if abicarbonate sparge liquor is used. This rise 

is undesirable and should be checked and the calcium ion concentration of the liquor 

should be maintained (Laing and Taylor, 1984). Experimental thick mashes (liquor/grist 

2.5/1, i.e. 28.6%) would not run off unless ahigh concentration of calcium ions (200mg/l) 

was used. Thus the last worts are weak, and are relatively rich in poorly flavoured 

extractives and potential haze-forming substances. These last runnings, like the press 

liquor from the spent grains (Chapter 3), may be stored hot for ashort period (to prevent 

spoilage by micro-organisms) and then be added to a subsequent mash to recover the 

extract. However, to maintain the quality of the beer the weak wort may need to be 

clarified by centrifugation to remove suspended solids (particularly lipids) and/or may be 

treated with active charcoal (doses of 10 50 g/hl have been suggested) to reduce the 

levels of tannins, nitrogenous substances, colour and harsh flavours before it is added to a 

later mash (Morraye, 1938; Prechtl, 1967). 

The faster the wort is run off the higher its content of suspended solids, lipids (which 

favour flavour instability, i.e. beer staling), and -and -glucans which may give 

problems when the beer is filtered (Muts and Pesman, 1986; Whitear et al., 1983). The 

lipid contents of strong worts, separated in various devices, were in increasing order mash 

tun < lauter tun < Strainmaster < older pattern mash-filter, and were given as 10 < 50 


Contents (mg/100 g of extract) 

1400 

1000 

500 

200 

a 

b 

c 

d 

e 

Fermentation 

limit 

Maltose 

19 15 10 5 4 3 2 1 0 

Extract in wort (% of dry matter) 

impossible without extra manipulations, such as under-letting) in any equipment except 

mash filters. 

Rates of wort separation are faster with more coarsely ground grists but with these 

extract recoveries are less good than from finely ground grists (Section 6.9). The more a 

mash is stirred the more fines are produced, the more oxidations are likely to occur 

(including the cross-linking of gel-proteins), but if stirring is inadequate temperature 

gradients may occur and mash may settle and burn onto the containing vessel's heating 

surfaces, so there is a critical, `compromise' stirring speed (Laing and Taylor, 1984). The 

deeper the mash-bed the slower wort or sparge liquor will flow through it. Progressively 

shallower beds and more finely ground grists are used in mash tuns, in lauter tuns and in 

mash filters. Run-off is impeded by fine particles in the grist (from the malt, cereal flours 

or formed in the mash) and it is favoured by keeping the malt husk as intact as possible to 

give the mash bed a more òpen' structure. Mashing under nitrogen gas, experimentally 

adding bisulphite (which is a reducing agent), adding heat-stable cellulase, maintaining 

adequate levels of calcium ions (particularly in thick mashes), using well-modified malt, 

experimentally adding cationic poly-electrolyte flocculants (such as boiled or unboiled 

lysozymze or partly de-acetylated chitin) and collecting wort at elevated temperatures all 

favour rapid wort run-off. Malt may contain endogenous flocculants and others may be 

present from the fungi present on the surface of the grains (Anderson, 1993). In contrast, 

mashing under air or oxygen gas, experimentally adding the oxidizing agent potassium 

bromate, omitting calcium ions and using poorly modified malt, all favour slow wort runoff 

(Anderson, 1993; Barrett et al., 1973, 1975; Crabb and Bathgate, 1973; Laing and 

Taylor, 1984; Muller, 1995; Muts et al., 1984). 

Poorly modified malts are rich in non-starch polysaccharides (NSPs; pentosans and - 

glucans) and undegraded proteins which are rich in thiol groups. The oxidation of 

cysteine side chains produces disulphide links between protein chains that can produce an 

insoluble, jelly-like mass of `gel-protein'. Reduction of this material should disperse the 

protein gel. An inverse correlation has been established between the gel-protein content 

of malt and wort separation rate. The hemicellulosic polysaccharides may also form a gel, 

which can be attacked by -glucanases, and so improve wort separation (Crabb and 

Bathgate, 1973). Fine aggregates of protein, small starch granules, cell-wall non-starchy 

polysaccharides (NSPs) and lipids can form `high flow-resistant' layers in a mash and 

particularly in the poorly permeable gel-like layer (the Oberteig) which forms on the 

surface of mashes in lauter tuns. Removal of this layer reduces the pressure differential 

across the bed, increases the flow-rate of the wort but reduces extract recovery (Muts and 

Pesman, 1986). The composition of this layer is variable; two examples contained, 

respectively, 18 and 20% protein and 65 and 79% polysaccharides. The composition of 

the small aggregates which form in mashing is also variable; for example, small starch 

granules, 4 21%, -glucan, 3 19%, pentosan, 5 31% and protein 26 42%. In the case 

of particles from an all-malt mash, which contained 29% starch, the free lipid content was 

5% and the bound lipid was 17% (Barrett et al., 1975). The bound lipid may have been 

associated with the starch. 

These particles probably contribute to, or constitute, the Oberteig. The formation of 

this material is favoured by oxidizing conditions. Small starch granules are often firmly 

invested with protein which, when oxidized, presumably firmly binds them into the 

particles. During mashing the greater part of the malt is dissolved, and some proteins are 

dissolved and, particularly as the temperature rises in temperature-programmed or 

decoction mashes, a proportion of the protein is denatured, aggregates and precipitates 

(Lewis and Oh, 1985; BuÈhler et al.,1996). The finer particles (< 1 150 m) tend to block 


the pores of the mash and impede run-off, but as the particles aggregate and enlarge so 

theirobstructiveeffectbecomesless.Largerparticlesarefavouredbycationicflocculants 

and (apparently) adequate concentrations of calcium ions. Aggregation is better at higher 

temperatures, and so in three mashes that had been not been heated above 65ëC (149ëF) 

offered 3, 3or 3.7 times the specific resistance to the flow of the wort offered by mashes 

that had been heated to 80ëC (176ëF; BuÈhler et al., 1996). 

`Models' of liquid flow through mashes, for instance based on Poiseuille's equation, 

which relates to flow through parallel capillary tubes, or the Carman-Kozeny equation or 

Darcy's law, that relate to beds of spherical particles, all emphasize that the rate of flow 

through abed of particles is proportional to aconstant, the pressure difference across the 

bed, the channel radius to the power 4, to the diameter of the particles squared, and 

inversely proportionaltothedepthofthebedandtothe viscosity oftheliquid(Anderson, 

1993; Bathgate,1974; Huite and Westermann,1974; Laing and Taylor, 1984; Meddings 

and Potter, 1971; Webster, 1978). 

As the temperature is increased so wort viscosity falls to comparatively low levels and 

the small particles aggregate and increase in size. Both changes favour faster wort 

separation. While the viscosity of wort (caused mainly by dissolved sugars and, to 

varying extents, by polysaccharides and perhaps other materials) is not unimportant the 

major limitation in wort separation for abed of agiven depth is the àverage' particle 

diameter, d. Because the flow rate is proportional to d 2 ,as dbecomes smaller so the flow 

rate rapidly declines (dˆ1, flowˆ1; dˆ0.5, flowˆ0.25; dˆ0.1, flowˆ0.01, etc.). 

Determining an àverage' diameter, dis impracticable for the particles in amash, and in 

any case it is not the depth of the entire mash but the characteristics and depth of the 

surface Oberteig layer that are often limiting. By using aderived formula that relates to 

compressible beds it is possible to find bed permeabilities and so test the factors that may 

influence them (Laing and Taylor, 1984). 

4.4 Mashing biochemistry 

4.4.1 Wort carbohydrates 

The complex mixture of carbohydrates in wort makes up about 92% of the solids in 

solution. The most important sugars and dextrins in wort are made of glucose, which also 

occurs free (4.1). Thus maltose (4.4), maltotriose (4.5) maltotetraose (4.6) and 

maltopentaose (4.7) are made of D-glucopyranose units joined by -(1,4) links (Table 

4.15). In cellobiose (4.18) and laminaribiose (4.19) the glucose residues are linked by - 

(1,4) and -(1,3) bonds, respectively. (An introduction to carbohydrate chemistry is 

given by Coultate, 2002). The precise composition of the mixture will depend on the 

make-up of the grist and the mashing conditions. In some `conventional' sweet worts the 

carbohydrate spectra are surprisingly similar, whether or not mash tun adjuncts are used 

(MacWilliam, 1968). The exceptions are when mashing is carried out to produce `lowalcohol' 

beers or `low-carbohydrate' beers when the conditions are chosen to obtain 

poorly fermentable and maximally fermentable worts, respectively (Chapter 15). Wort 

fermentability may be increased by adding amyloglucosidase or, preferably, other 

microbial enzymes, such as pullulanase and -amylase or fungal saccharogenic amylase 

to the mash and/or to the fermenter. 

The conversion of barley into malt involves aconsiderable loss of potential extract 

(Briggs, 1998; Table 4.16). Sugars fermentable by most yeasts are the monosaccharides 

glucose (4.1) and fructose (4.2), the disaccharides sucrose (4.3) and maltose (4.4) and the 


Table 4.15 The major wort carbohydrate fractions compared with the `potentially extractable' 

carbohydrates of malt (data of Hall et al., 1956). The brewery extract of the malt was 102.2 lb/Qr 

(about 77.5%; 303.5 lë/kg). The laboratory extract was 104.6 lb/Qr (about 78.9%; 309.1 lë/kg). The 

values are given as hexose equivalents, that is, as if each fraction had been fully hydrolysed to yield 

its component hexoses, and so the reported weights are greater than the weights of the unhydrolysed 

materials. The carbohydrates made up 91.8% of the wort solids 

Malt carbohydrates (hexose equivalents as Wort carbohydrates (hexose equivalents as 

% total wort solids) % wort solids) 

Starch 85.8 Dextrins, glucans and pentosans 22.2 

Glucans and pentosans* 2.5 ? 

Fructans y 1.4 ? 

± ± Maltotetraose 6.1 

Maltotriose 0.6 Maltotriose 14.0 

Maltose 1.0 Maltose 41.1 

Sucrose 5.1 Sucrose 5.5 

Glucose 1.7 Glucose + Fructose 8.9 

Fructose 0.7 

Total 98.8 97.8 

* These `gums' were soluble in water at 40 ëC (104 ëF). The maltose fraction in the wort contained a trace of 

unfermentable isomaltose. y Fructans in the wort were included in the other fractions but were certainly present. 

Maltotetraose was essentially absent from the malt. 

trisaccharide maltotriose (4.5). Typically maltose is the most abundant sugar in wort. 

Sugars in a 12% wort, in g/100 ml, were glucose + fructose, 0.9 1.2; sucrose, 0.4 0.5; 

maltose, 5.6 5.9; and maltotriose, 1.4 1.7; total 8.3 9.3 (Evans et al., 2002). Some 

yeasts only attack maltotriose (4.5) to a limited extent, while other `super-attenuating' 

strains may also utilize maltotetraose (4.6) and dextrins. The monosaccharides are 

Table 4.16 The carbohydrate composition (% dry basis) of Carlsberg barley (TN 1.43%) and a 

floor-malt made from it (recalculated from Hall et al., 1956). The supposed `structural 

carbohydrates' that are not involved with extract formation have been ignored 

Barley Malt 

Glucose 0.04 1.31 

Fructose 0.07 0.55 

Sucrose 0.77 3.73 

Maltose 0 0.73 

Maltotriose 0 0.42 

`Glucodifructose' 0.08 0 

Raffinose 0.15 0 

Fructans 0.58 1.00 

Glucans and pentosans* 2.10 2.45 

Starch 65.86 58.90 

Total 69.65 69.09 

* Non-starch polysaccharides soluble in warm water at 40 ëC (104 ëF). The raffinose went during 

germination, but the `glucodifructose' could not be determined in the malt, and will have been 

included in another fraction. 

The thousand corn dry weights of the barley and the malt were 39.1 g and 35.3 g respectively, so the malt yield 

was 35.3 100/39.1 ˆ 90.3%. 

100 g barley contained 65.86 g starch hexose, while 90.3 g malt (from 100 g barley) contained 58.90 0.903 g 

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 

the starch going during germination was 19.2%. 


HO·CH 2 

H 

H 

O 

OH 

CH2OH H O OH 

4 

HO 

OH H 

Fructose; β-Dfructofuranose 

(4.2) 

6 

5 

H 

OH 

H 

H OH 

3 

H 

2 

* 

1 

CHO 

* 1 

H C OH 

HO C H 

H C OH 

H C OH 

CH 2OH 

β-D-Glucopyranose D-Glucose, 

aldose form 

OH 

CH 2OH 

H 

CH2OH O H HO·CH2 O 

HO 

H 

OH 

H 

H OH 

O 

2 

3 

4 

5 

6 

(4.1) 

Sucrose; α-D-glucopyranosyl-(1→2)β-D-fructofuranoside 

CH2OH H O H 

HO 

H 

OH 

H 

H OH 

(4.3) 

O 

H 

CH2OH H O H 

HO 

OH 

H 

OH 

H OH 

OH 

H 

H OH 

H 

CH 2OH 

H 

CH2OH O 

* 

H, OH 

Maltose; α-D-glucopyranosyl-(1→4)- 

D-glucopyranose 

(4.4) 


H 

OH 

H 

H 

O 

CH2OH H O H 

H 

OH 

H 

H OH 

CH2OH H O H 

4 

HO 

6 

H 

5 

OH 

H OH 

3 

H 

OH 

α-D-Glucopyranose 

Maltotriose (n = 1); maltotetraose (n = 2); maltopentaose (n = 3) 

n 

O 

2 

* 

1 

CH2OH H O 

H 

OH 

H 

H OH 

(4.5) (4.6) (4.7) 

CH2OH H O H 

HO 

H 

OH OH 

H H 

HO 

CH2OH O H 

H 

OH H * 

H OH 

H OH 

(4.9) 

H, OH * 

Mannose; α-Dmanmopyramose 

(4.8) 

* 

OH 

Galactose; α-Dgalactopyranose

H 

5 

H O OH 

H 

6 OH 

HO 

H * 

H 

H OH 

3 

2 

Xylose; β-Dxylopyranose 

(4.10) 

1 

HOCH2 5 

H O OH 

4 

OH H * 

H 

H OH 

3 

Arabinose; α-Larabinofuranose 

(4.11) 

fermented the most rapidly, while maltotriose is fermented slowly and sometimes 

incompletely so traces may remain in beer. Dextrins, derived from the partial 

degradation of starch, are not fermentable and neither are pentosans nor are -glucans. 

Sometimes the `fermentable sugars' and `dextrins' groups are determined and the results 

are used to calculate the carbohydrate fermentability of the wort as the fermentable 

carbohydrates as apercentage of the total carbohydrates. Values in the range 64 77% 

are common. The fermentable carbohydrates are the major energy source of the yeast 

and alcohol and carbon dioxide are the major metabolic products. The major source of 

extract in an all-malt wort is starch, but preformed sugars are also important andof these 

sucrose(4.3)isthemostabundant(Tables4.15,4.16).Mannose(4.8)andgalactose(4.9) 

occur combined in malt but neither is released during mashing. In worts maltose (4.4), 

along with many other substances, is produced during mashing by the partial hydrolysis 

of starch. 

Table 4.15, which quantifies the major groups of carbohydrates, indicates the major 

sources of the carbohydrates in the extract. However, depending on the grist, isothermal 

mashes may yield about 2% of extract from non-starch polysaccharides while this value 

may reach 6% in decoction mashes. The value is likely to be higher when the mashes are 

supplemented with microbial enzymes, which attack non-starch polysaccharides. The 

levels of the monosaccharides in wort should not be abnormally high (as can be the case 

when amyloglucosidase is added to the mash) because this can interfere with the uptake 

and metabolism of maltose (4.4) and maltotriose (4.5) by yeast, causing a `sticking 

fermentation' (the premature cessation of the fermentation). While the major simple 

sugars are as indicated other sugars occur in minor amounts. Pentoses are present (e.g. 

xylose (4.10) at 1.5 mg/100 ml, arabinose (4.11) at 1.4 mg/100 ml and ribose (4.12) at 

0.2 mg/100 ml) compared to maltose (4.4) at 4000 to 6000 mg/100 ml. 

The pentoses are more abundant in decoction and (probably) temperature-programmed 

worts than in simple infusion worts, since the conditions in infusion mashes do not favour 

the enzyme-catalysed breakdown of pentosans. Other sugars detected in tiny amounts are 

isomaltose (4.13), panose (4.14), isopanose (4.15), nigerose(4.16) and maltulose(4.17). 

Apparently cellobiose (4.18) and laminaribiose (4.19), expected breakdown products of 

-glucans, have not been detected. -Glucans and pentosans are always present, but in 

varying amounts. Carbohydrates also occur in glycolipids, in nucleic acids and 

nucleotides as well as glycoproteins. These are important in various ways, but are 

insignificant in terms of extract yield. The calorific value of beer is due to the 

unfermentable carbohydrates and the ethanol (ethyl alcohol). Small amounts of fructans 

occur in malt. These can be regarded as sucrose molecules (4.3) to which one or more 

fructose residues have been attached. Simple examples are kestose (4.20), isokestose 

(4.21) and bifurcose (4.22). The fate of fructans in mashing is not known. 


2 

1 

HO·CH 2 

O OH 

H H * 

H H 

OH 

Ribose; β-Dribofuranose 

(4.12) 

OH 

1


CH2OH H OH 

H 

H 

OH 

O 

H 

O 

OH 

H 

H 

H, OH * 

HO H H O 

H OH 

CH 2OH 

Cellobiose; β-D-glucopyranosyl-(1→4)- 


(4.18) 

H 

CH2OH O H 

H 

HO·CH2 O H 

OH 

HO 

H 

O H HO 

CH2· O 

H OH 

OH H 

CH2OH H CH2OH H 

H 

OH 

O 

H 

O 

OH 

H 

H O 

HO 

H 

H 

OH 

H 

H, OH 

OH * 

Laminaribiose; β-D-glucopyranosyl-(1→3)- 


(4.19) 

Kestose (6-kestose); α-D-glucopyranosyl-(1→2)-β-Dfructofuranosyl-(6→2)-β-D-fructofuranoside 

(4.20) 

CH2OH H O H 

H 

OH H 

HO 

HO·CH 2 

H 

H 

HO·CH 2 

H OH 

HO H 

O 

H 

H 

HO·CH 2 

OH H 

HO 

CH2OH 4.4.2 Starch degradation in mashing 

Starch makes up the greatest proportion of malt, often about 58% (dry basis), and is 

present in greater proportions in some mash tun adjuncts. The breakdown products of 

starch make up most of the extract in worts. In malt the starch is practically confined to 

the starchy endosperm where, in the undermodified regions, it is enclosed in the cell 

walls. It is often invested with protein, which may impede its breakdown. Starch is 

H 

O H 

iso-Kestose (1-kestose); α-D-glucopyranosyl-(1→2)-β-Dfructofuranosyl-(1→2)-β-D-fructofuranoside 

O 

O 

(4.21) 

O 

OH 

CH2 HO 

HO H 


CH2·OH H O H 

H 

OH H 

HO 

H OH 

O 

HO·CH 2 

Bifurcose, the basic unit of the two series of fructosans. 

Extended in direction A, with β (2→1)-linked fructofuranose 

units; kritesin (inulin-type) fructosans. Extended in direction 

B with β (2→6)-linked fructofuranose units: hordeacin (phlein type) 

H 2·C·O·CO·R 1 

HO·C·H O 

H 

O 

OH 

CH2OH OH H 

CH 2 

H 

+ 

H2C·O·P·O·CH2·CH2·N(CH3) 3 

O – 

O H 

O 

OH H 

HO 

CH2·O (4.22) 

Lysophosphatidylcholine 

R 1·CO- is a fatty acid residue 

(4.23) 

deposited in organelles called amyloplasts, and presumably residues of these also 

surround it. It occurs as granules. In barley the granules occur in two populations, the 

larger Agranules that may have diameters of 22 48 m, and Bgranules with diameters 

of 1.7 2.5 m. The large granules make up 10 20% of the granules by number but 

85 90% by weight. The large granules have alower gelatinization temperature range 

than the small granules. 

As the sugar concentration of the surrounding liquid increases, so the gelatinization 

temperatures of starches increase (Bathgate and Palmer, 1972; Briggs, 1978, 1992, 1998; 

Eliasson and Tatham, 2001; Letters, 1995a, b; Stone, 1996; Tester, 1997). Starches from 

other sources may differ significantly, both in size, shape and physical properties 

(Chapter 2). The major components of the granules are the polysaccharides amylose and 

amylopectin, together called `starch'. However, the granules are not pure starch, but also 

contain some protein, ash and lipids. Typically amylose makes up 22 26% of the 

polysaccharide, the balance being amylopectin. Amylose is amixture of predominantly 

linear -(1,4)-linked chains of D-glucopyranose, about 1600 1900 residues long (Fig. 

4.17). The presence of occasional branch-points, formed by -(1,6)-links, is indicated by 

the incomplete hydrolysis of amylose by -amylase. In solution amylose can retrograde, 

that is crystallize and separate from solution. This retrograded material is comparatively 

resistant to enzymic attack, and so will not readily be converted into extract if formed 

during mashing. Amylose adopts a helical shape (six glucose residues/turn) and inclusion 

compounds can be formed with polar lipids or iodine being contained in the helix. The 

complex with iodine has a characteristic blue-black colour. The lipid inclusion 

H 

B 

HO·CH 2 


H 

O H 

OH H 

A 

HO 

CH2OH

complexes, which occur in barley starch, involve mainly lysophosphatidyl choline, LPC 

(4.23), and are not readily degraded by enzymes. The polar choline moiety projects from 

the end of the helix. The presence of lipids slows, or prevents, retrogradation. Other 

cerealstarchescontaininclusioncompoundswithfreefattyacids.Thelipidcomplexesdo 

not give colours with iodineand so incompletely degraded starch that is complexedis not 

detected by the iodine test unless the lipid is first removed, for example, with butanol. 

Thus the iodine test, as usually applied to samples from the mash or to spent grains, is 

unreliable and does not detect all undegraded starch. Each amylose molecule (molecular 

weight 26 31 10 4 )will have anon-reducing chain end, in which the terminal glucose 

residue is unsubstituted on position C-4, (4.1) and areducing chain end where the 

terminal glucose has afree C-1 position. 

Amylopectin is amixture of highly branched molecules, the -(1,4)-linked chains, 

around 26 glucose units long on average, are joined through -(1,6) branch-points, 

which may number c. 6% of the bonds in the molecules (Fig. 4.17). Molecular weights 

may be very high, e.g., 2 10 6 4 10 8 .Amylopectin is less soluble in water than 

amylose. The chains may also adopt helical configurations which, with iodine, give a 

red-violet colour. Each amylopectin molecule has only one reducing chain end but 

numerous non-reducing chain ends (Fig. 4.17). The polysaccharide molecules are 

ordered in the starch granules, which have apartly crystalline structure, as shown by Xray 

diffraction. In the granules the crystalline regions alternate with the amorphous 

regions, which are more easily attacked by enzymes. The amylose molecules are 

supposed to be mixed in among the amylopectin, but the arrangements proposed are 

tentative (Fig. 4.18; Imberty et al., 1991). The ordered molecular structure of the 

granules is also demonstrated by their birefringence. In polarized light the granules 

appear to have adark `maltese cross' on alight background. As the starch is heated in 

water and gelatinization begins so the swelling granules begin to lose their birefringence 

and the crosses disappear, a fact that allows an estimation of the gelatinization 

temperature range of the starch. Gelatinized starch is readily attacked by enzymes, but 

this should take place before retrogradation of any of the amylose occurs. The amyloselipid 

complexes of some starches are not disrupted in cooking until temperatures of 

90 120ëC (194 248ëF) are reached, which may explain the need to cook some 

adjuncts at temperatures above the gelatinization temperatures of their starches. On 

cooling inclusion complexes can slowly reform, and so the enzymic degradation of the 

`liberated' amylose should not be delayed. 

`Diastase', the mixture of malt enzymes that catalyses the hydrolytic breakdown of 

starch, has been studied for many years, but even now there are uncertainties about the 

roles of some of the component enzymes (Fig. 4.19; Briggs, 1992, 1998). Of these 

enzymes only the activity of -amylase correlates well with the determination of diastatic 

power, DP, as it is usually determined. Older studies attributed the conversion of starch 

during mashing to the activities of the - and -amylases. While these are the enzymes 

chiefly involved it seems that, at least in temperature-programmed mashes, other 

enzymes play significant roles. 

Malt -amylase is a mixture of different molecules (isoenzymes), having slightly 

differing properties and these are formed during malting; they are essentially absent from 

sound, ungerminated barley. The three `classes' observed each contain multiple forms. - 

Amylase-I occurs in comparatively small amounts in malt. It is relatively resistant to acid 

conditions and chelating agents because it binds calcium ions very strongly. It is inhibited 

by small amounts of heavy metal ions, such as copper. -Amylase-II is the `classical' 

malt enzyme. It is comparatively resistant to heat, particularly in the presence of excess 


Fig. 4.17 Idealized diagrams of (a) amylose and (b) amylopectin. The chains of D-glucopyranose 

residues (hexagons) are joined by -(1, 4)-links in the amylose and the short chains of the 

amylopectin which are joined together through -(1, 6)-links, which create branch points, in the 

amylopectin. The reducing chain ends are marked with an asterisk, while the non-reducing chain 

ends are indicated with solid hexagons. While the straight chained amylose molecule has only one 

reducing group and one non-reducing chain end, the highly branched amylopectin has one reducing 

group but numerous non-reducing chain ends per molecule. The dashed line around the amylopectin 

indicates the approximate limit of the -limit dextrin remaining after the molecule has been 

attacked by pure -amylase. The shortened branches, of two or three glucose residues, are readily 

hydrolysed by limit dextrinase or pullulanase to release maltose or maltotriose. The debranched 

dextrin can be degraded further by -amylase. 

calcium ions, and to heavy metal ions, but it is inhibited by calcium-binding `chelating 

agents', such as phytic acid. It isnot completely stable in mashes. It has apH optimum of 

about5.3,anditisunstableatvaluesbelow4.9.` -Amylase-III'isacomplexbetween - 

amylase-II and another small protein, BASI, (barley amylase/subtilisin inhibitor), which 

limits the activity ofthe enzyme.Thiscomplexis probably disrupted at starch conversion 

temperatures. The -amylase mixture from malt attacks the -(1,4)-links within the 

starch chains, producing arange of products. Attack is slower at the chain-ends and 

ceases near the -(1,6) branch points. The products of extensive -amylolysis include 

glucose (4.1), maltose (4.4), and a complex mixture of branched and unbranched 

oligosaccharides and dextrins (Fig. 4.19). Because this kind of attack reduces the starchiodine 

colour rapidly, but increases the reducing power of the digest comparatively 

slowly this enzyme is often referred to as a`dextrinogenic' amylase. It is also able to 

liquefy starch gels. This enzyme is capable of degrading intact starch granules as is - 

glucosidase. In mashing, -amylase liberates the dextrins that are the substrate for the 

`saccharogenic' -amylase. 


Amorphous 

Crystalline 

Amorphous 

Crystalline 

Amorphous 

Fig. 4.18 Adiagram of the ways in which amylopectin molecules may be packed together in a 

starch granule to create amorphous and crystalline regions, which differ in their susceptibilities to 

enzymolysis (after Imberty et al., 1991). Many of the side chains are wound together in double 

helices. The single reducing chain end is at the top of the diagram. Linear amylose molecules, some 

of which are in the form of helices and inclusion compounds with polar lipids are, in some way, 

interspersed with the amylopectin in the granules, possibly in the amorphous regions. 

-Amylase occurs in barley in insoluble and soluble forms, and is of value when raw 

barley is used as amash tun adjunct. During malting the proportion of the `free', soluble 

enzyme increases as does the ease of extraction. Little or no more enzyme is formed 

during germination. The soluble enzyme contains multiple forms having various 

molecular weights, including dimers in which enzyme is linked to the, enzymically 

inactive, protein Zby disulphide bonds and in which high proportions of the monomers 

seem to be partially proteolytically degraded forms of at least two genetically distinct 

isozymes. The insoluble enzyme can be partly released by proteases, such as papain, by 

agents which reduce, and so break, the disulphide bonds between the enzyme and other 

insoluble proteins, and by àmphipathic' detergents that disrupt hydrophobic bonds 

(ButtimerandBriggs,2000).Comparedto -amylase, -amylaseisrelativelysensitiveto 

heat and heavy metal ions and is resistant to mild acidity and chelating agents. 

-Amylase from different barleys differs in its temperature sensitivity. Barleys with 

the more stable enzyme give malts which yield the most fermentable worts (Evans et al., 

2002). This enzyme is readily inactivated by chemical agents that react with thiol groups. 

It has abroad pH optimum around 5.0 5.3, (but which alters with the buffer used in the 

activity measurements), catalyses the hydrolysis of the penultimate -(1,4)-link of the 

non-reducing chain ends of amylose and amylopectin, with the release of the reducing 

disaccharide maltose (4.4), the most abundant sugar in wort (Fig. 4.19). However, the 

enzyme willnothydrolyse -(1,4)bondsnear to -(1,6)branch-pointsinamylopectinor 


Phosphorylase 

G–G–G–G–G–… + Pi G–1–P + G–G–G–G–… 

Non-reducing 

chain end 

β-Amylase 

Non-reducing 

chain end 

α-Glucosidase 

α-Amylase 

Inorganic 

phosphate 

Glucose-1phosphate 

Debranching enzyme (limit dextrinase; R-enzyme) 

Shortened 

chain 

G–G–G–G–(G) N–G… + H 2O G–G* + G–G–(G) N–G… 

Maltose Shortened 

chain end 

G–G* or G–G–G* + H2O G* + G* or G* + GG* 

Maltose Maltotriose Glucose Glucose Maltose 

G* or G–G* 

G 

G 

Isomaltose 

Dextrins, etc. 

+ H 2O 

G–G–G–G–(G) N –G–G–G–G–(G) M –G* + H 2 O 

G–G–G–G–G–G 

G* + G* or G* + G–G* 

Glucose Glucose Maltose 

Branched and unbranched chains Mixture of sugars, 

oligosaccharides and dextrins 

Transglucosylase reactions 

(These may be catalysed by various enzymes, as ‘side reactions’) 

G*, G–G*, G–G–G* 

G–G–G–G–G–G–G–G* 

G–G–G–G 

G–G–G–G–G–G–G–G–G–G* + H 2O G–G–G–G–G–G–G–G–G–G* 

G–G–G–G 

Amylopectin; branched dextrins 

G–G* 

Maltose 

Glucose + various products 

G–G–G–G–G* 

Amylose; products without 

α(1→6)branch-points 

Isomaltose 

G–G–G–G–G–G–G–G–(G) N –G–G–G* G–G–G–G–G–G–G–G–G–G* 

G–G–G–G–G–G–G 

Amylose 

Amylopectin 

Fig. 4.19 A summary of how the enzymes, that together make up diastase, act on the 

polysaccharide components of starch or their breakdown products (after Briggs, 1998). G, 

glucopyranose residues ± , (1, 4)-link |; (1, 6)-link; *, reducing group. 

-D- 


G* 

G

dextrins. Thus if dispersed starch is attacked by -amylase acting on its own amylose 

molecules are broken down, with the liberation of maltose, until one of the occasional 

branch-points is encountered, while amylopectin is degraded to maltose and a -limit 

dextrin in which all the non-reducing chain ends are within two or three glucose residues 

ofbranchpoints(Fig.4.17).Inmashingafewofthe -(1,6)-linksmaybebrokenbylimit 

dextrinase, with the release of maltose (4.4) or maltotriose (4.5), or -glucosidase, with 

the release of glucose (4.1), and links within the chains can be broken by -amylase. In 

each case the effect is to expose anon-reducing chain end that the -amylase can attack. 

The extensive degradation of starch that occurs during mashing depends on the concerted 

action of the mixture of enzymes present and is often limited by enzyme destruction 

rather than the absence of substrates for possible enzyme attack. 

Phosphorylase, the enzyme that catalyses the cleavage of the terminal -(1,4) links in 

non-reducing chain ends with inorganic phosphate to release glucose-1-phosphate, is 

present in barley and green malt (Fig. 4.19). Apparently its possible role in mashing has 

never been investigated. Like -amylase this enzyme can degrade chains until it comes to 

a branch point. As wort contains inorganic phosphate this enzyme may be active and as 

phosphatases are also present the glucose-1-phosphate generated would be hydrolysed to 

glucose and phosphate, the overall effect being the same as degradation by -glucosidase. 

-Glucosidase is present in barley and increases in amount during malting (Briggs, 

1998). The enzyme seems to have several forms differing in their substrate specificities and 

a proportion is insoluble but still able to catalyse the hydrolysis of small molecules such as 

maltose (4.4) (Fig. 4.19). The enzyme has a pH optimum of about 4.6 and a temperature 

optimum of 40 45 ëC (104 113 ëF). This enzyme is probably active in the early stages of 

temperature-programmed mashing. Like -amylase this enzyme attacks starch granules 

(Sun and Henson, 1990), and acts synergistically with -amylase in this respect. This 

enzyme attacks maltose (4.4), isomaltose (4.13), oligosaccharides, dextrins and starch at 

the ends of the non-reducing chains, hydrolysing -(1, 4) links preferentially and -(1, 6) 

links more slowly (Fig. 4.19). Like some other carbohydrases this enzyme can catalyse 

transglucosylation in strong solutions of sugars, generating small amounts of different 

materials. The data on the role of this enzyme in mashing is inadequate, but indirect 

evidence suggests that its action can be significant in temperature-programmed mashes. 

Debranching enzyme catalyses the hydrolysis of -(1, 6) links in amylopectin and 

dextrins. Previously it was thought that two enzymes were involved and these were 

referred to as limit dextrinase and R-enzyme. When the enzyme acts on -limit dextrins 

maltose (4.4) and maltotriose (4.5), but not glucose (4.1), are released. While the role of 

the malt enzyme in brewer's mashes is uncertain the value of using the similar bacterial 

enzyme pullulanase in making highly fermentable worts is clear (Enevoldsen, 1975). The 

survival of this enzyme in malt is strongly dependent on the kilning conditions. The 

enzyme occurs in insoluble and soluble forms and much of the soluble enzyme is 

inhibited by one or more associated proteins. It was thought that as most of the -(1, 6) 

links initially present in the mash survived into the wort the activity of debranching 

enzyme must have been insignificant (Enevoldsen, 1975). However, more recent work 

indicates that in spite of earlier reports the enzyme is at least as stable as -amylase in 

mashes and its activity may be significant in some circumstances (Bryce et al., 1995; 

Sissons, 1996; SjoÈholm et al.,1995; Stenholm and Home, 1999; Stenholm et al., 1996). 

Clearly, brewers or distillers who require highly fermentable worts, desire high levels of 

active debranching enzyme and low levels of the inhibitory proteins in their mashes. 

Experiments with soluble starch and enzyme preparations in buffered solutions are 

unrealistic `models' for mashes in that the conditions are significantly different and this 


alters the stabilities of the enzymes and indeed takes no account of the insoluble enzymes 

present in malt. -Amylase acting alone on starch for an extended period, at various 

temperatures around 65ëC (149ëF), produces poorly fermentable worts (about 20%) and 

this result is comparatively temperature insensitive. If the -amylase is supplemented 

with increasing amounts of -amylase (probably contaminated with other enzymes) then 

wort fermentability increases. However, the higher the temperature the lower the 

fermentability. Thus, as expected, wort fermentability is dependent on both the mashing 

temperature and the amounts of enzymes present in the mash. At higher temperatures the 

heat-labile -amylase is destroyed faster and so the `dextrins' produced by the -amylase 

areless-well`saccharified'bytheother,moreheatlabileenzymes.Thisisconsistentwith 

experience with the results of experimental isothermal mashes (Table 4.17; Fig. 4.20). 

Alterations in isothermal mashing temperatures and durations and malt quality influence 

extract yield and its quality (Fig. 4.9; Tables 4.6; 4.17). The changes occurring in 

temperature-programmedmashingare more complex (Stenholmet al.,1996; Gjertsenand 

Hartlev, 1980; Schur et al., 1973; Table 4.18). Thus, during amash programmed with 

restsat48ëC,63ëC,72ëCand80ëC(118.4,145.4,161.6and176ëF)extractrose,butata 

decreasing rate, until the temperature rise from 63 to 72ëC when it became constant (Fig. 

4.21). Fermentability stopped rising during the 63ëC rest, when about 0.7 of the - 

amylase initially present had been destroyed. Destruction was completed as the 

temperature rose to 72ëC. -Amylase destruction was not complete until 80ëC had been 

reached. 

Whilethepatternsofsugarformationduringmashingreflectenzymeactivitiestheydo 

not unambiguously demonstrate which enzymes are active. Thus fructose (4.2) may 

originate from the hydrolysis of sucrose (4.3) or higher oligosaccharide fructans, or from 

free sugar initially present in the grist. Glucose (4.1) occurs in malt and during mashing 

may be formed by the activity of -amylase, -glucosidase or -glucosidase. Maltose is 

undoubtedly mainly formed by -amylase but there may be contributions from 

debranching enzyme (acting as a limit dextrinase) and -amylase. In temperatureprogrammed 

mashes the rise in glucose at low temperatures suggests that -glucosidase 

isactive(Schur et al.,1973). Thepatternsofsugarspresent inisothermalmashes madeat 

different temperatures indicate that glucose production is maximal at about 57ëC 

(approx. 135ëF), which is consistent with -glucosidase activity (Taylor, 1974; Fig. 

4.20). 

When starch-containing adjuncts are added in increasing amounts to mashes, wort 

fermentability usually declines before extract recovery indicating that saccharogenic 

activity becomes limiting before dextrinogenic activity, essentially -amylase. However, 

increasing the amount of -amylase in these mashes can increase wort fermentability. In 

Table 4.17 The influence of isothermally mashing two malts, at three different temperatures, on 

wort fermentability and content of nitrogenous substances (Hudson, 1975) 

Malt 1 (Diastatic power, Malt 2 (Diastatic power, 

Mashing temperature 33 ëL; TN, 1.3%) 90 ëL; TN, 1.8%) 

ëC ëF Ferm. (%) TSN (mg/100 ml) Ferm. (%) TSN (mg/100 ml) 

68 155 72 73 77 96 

65.5 150 76 78 86 106 

63 145 79 84 88 113 

Abbreviations; Ferm., fermentability. TSN, total soluble nitrogen. TN, total nitrogen content of the malt (% on 

dry). 


Sugar (% w/v) 

10 

9 

8 

7 

6 

5 

4 

3 

2 

1 

TFS 

G2 

G 3 

SG 

1050 

1046 

1042 

1038 

1034 

0 

120 125 

S 

130 135 

G4 140 145 150 155 

1030 

160 

Mashing temperature (°F) 

50 55 60 65 70 

G 

Mashing temperature (°C) 

Fig. 4.20 The specific gravity of worts (l ÐÐ l SG) and the levels of total fermentable sugars 

(ut ± ± ± ut TFS) and the sugars present in the worts prepared by isothermal 2h. mashes at the 

temperatures shown (after Taylor, 1974). Key to the sugars: G, glucose ± ± ± ; G2, maltose 

4 ± ± ± 4 ; G 3, maltotriose s ± ± ± s ; S, sucrose r ± ± ± r ; G 4, maltotetraose n ± ± ± n. 

Only traces of fructose were detected. 

Table 4.18 The influence of three mashing conditions on the -glucan contents of worts prepared 

from three malts (Narziss, 1978). In the extended programme the malt grist was mashed in at 35 ëC 

(95 ëF). After a 30 min. rest the temperature was raised (during 15 min.) to 50 ëC (122 ëF) and this 

was held for 30 min. Temperature rests were subsequently at 65 ëC (149 ëF)/30 min., 70 ëC (158 ëF)/ 

30 min. and 75 ëC (167 ëF)/5 min. In between the rests the temperature was increased at 1 ëC 

(1.8 ëF)/min. Total time, 180 min. The shortened mashes began at 50 ëC (122 ëF) or 65 ëC (149 ëF), 

and in each case the rest of the temperature programme was as before, so mashing times were 135 

and 90 min. respectively 

Malt modification Excellent Normal Poor 

Fine/coarse extract difference (% EBC) 1.1 2.0 3.8 

Endo- -glucanase activity (mPa/s) 0.343 0.315 0.096 

Mashing in temperature Duration -Glucan yield (mg/100 g dry wt.) 

ëC ëF (min.) 

35 95 180 17 58 390 

50 122 135 33 82 595 

65 149 90 86 231 645 


Specific gravity of wort (SG)

Extract or fermentable sugars (% w/v) 

20 

15 

10 

5 

48°C 

118.4°F 

63°C 

145.4°F 

β-Amylase 

72°C 

161.6°F 

0 0 40 80 120 


part this may be due to the fermentable sugars produced by this enzyme, but the faster 

liquefaction of the starch and faster production of dextrins produces more accessible 

substrate that -amylase and other thermolabile enzymes can attack before they are heat 

inactivated. The problem with starches with high gelatinization temperatures is that by 

the time they have been liquefied at the necessarily high temperatures used, all the 

saccharogenic enzymes have been destroyed. Hence the need to cook and liquefy such 

starches, and adjuncts containing them, then to cool the mixture and mix it with malt at 

temperatures at which saccharification is still possible. The gelatinization temperature of 

malt starch can vary by as much as 6 ëC (10.8 ëF; Bourne, 1998). Worts prepared from 

these differing malts had varying fermentabilities. 

4.4.3 Non-starch polysaccharides in mashing 

The non-starch polysaccharides in the grist (NSP) are the fructans, the hemicelluloses, the 

gums and the holocellulose. Pectins seem to constitute negligible proportions of grist 

materials, although small amounts of combined uronic acids are present. Apart from 

sucrose (4.3), which seems to undergo little hydrolysis during mashing, the fate of the 

fructans is unknown. However, as the levels of fructose do not increase appreciably 

during mashing, it is likely that the fructans, which are very soluble, are not hydrolysed 

and so they remain with the unfermentable carbohydrates. In many plants fructans are 

metabolized by transglycosylation reactions. Holocellulose is the polysaccharide material 

which remains undissolved after extracting grist with hot water and solutions of caustic 

alkalis. This fraction comes mainly from the husk in malt, where it is associated with 

lignin and it makes up about 5% of barley. However, small amounts are found in all parts 

of grains. This material is thought not to undergo any alterations in mashing. There is no 

80°C 

176°F 

Extract 

Fermentable 

sugars 

α-Amylase 

Fig. 4.21 The changing yields of extract and fermentable sugars (glucose, maltose and 

maltotriose) and the declining levels of -amylase and -amylase during temperature-programmed 

mashing (data of Stenholm et al., 1996). 


400 

200 

0 

α-Amylase (U/g) 

1000 

500 

0 

β-Amylase (U/g)

evidence that pure cellulose (poly- -(1,4)D-glucan) is present in malt, although some 

could be present in the holocellulose fraction, which contains combined glucose (4.1), 

mannose (4.8) and lesser amounts of galactose (4.9). 

The remaining fractions, usually grouped as the gums and hemicelluloses, have been 

extensively studied (Briggs, 1998; MacGregor, 1990; Fincher, 1992; Letters, 1995a,b; 

HanandSchwarz,1996).Togethertheymakeupabout10% ofbarley,butduringmalting 

the -glucancomponentissubstantiallydegradedwhilethepentosansincrease.Gums are 

soluble in water, while the residual hemicelluloses are soluble in hot solutions of caustic 

alkali. If the extraction of gums is carried out with water of increasing temperatures the 

quantity of gum recovered increases at the expense of the residual hemicellulose. Thus 

these fractions are aseries of materials with arange of solubilities. Chemically there are 

two major groups of substances in these fractions, the -glucans, that have been 

exhaustively studied, and the less-studied pentosans. Minor amounts of other 

polysaccharides are probably present. These substances occur in the cell walls of barley 

and malt. The major, but not the only source of these substances, is the cell walls of the 

starchy endosperm in malt and barley and wheat adjuncts. In barley the carbohydrates of 

the endosperm cell walls contain about 70 75% -glucan, 20 25% pentosan and 2 4% 

holocellulose. 

During malting the -glucan in the grain is preferentially degraded. In under-modified 

malts, chit malts, inhomogeneous malts and barley adjuncts the undegraded gums (and 

possibly hemicelluoses) present give rise to production problems if they are not 

adequately broken down during mashing. Problems with mashes made with wheat (or 

rye,ortriticale)adjunctsareoftencausedbypentosans.Wortseparationmaybeslow,the 

wort will be too viscous, extract recovery is likely to be low, beer filtration will be slow 

and will use large amounts of filter-aid material, the beers may become hazy and even 

deposit gels. These materials are also deposited if beer is frozen. However, other 

polysaccharides, including -glucan dextrins (derived from starch), yeast glycogen and 

cell-wall glucomannans may also be involved (Forage and Letters, 1986; Letters, 

1995a,b). Many of the problems attributed to -glucans are probably partly due to 

pentosan materials (Han and Schwarz, 1996). These polysaccharides may also have 

beneficial effects on beer qualities, adding to palate-fullness and foam stability. Elevated 

levels of -glucans indicate that malt is under-modified and so, in addition to other 

consequences, may lack adequate `nitrogen (protein)-modification' and levels of the 

enzymes needed in mashing. 

-Glucans are families of molecules consisting of linear chains of -D-glucopyranose 

units, of various lengths (molecular weights) linked in various ways. These chains are 

unsubstituted and, despite speculations to the contrary, there is no evidence for crosslinking 

via peptides or other materials. The major class in barley contains amixture of 

1,3- and 1,4-bonds in which about 90% consists of cellotriosyl and cellotetraosyl units 

(in which the respectively 3and 4glucose units are linked by -(1,4)-bonds) are joined 

by single -(1,3)-links (Fig. 4.22). In general the frequency of the (1,3)- to (1,4)-links is 

about 3 to 7, but this varies. This material resembles the seaweed polysaccharide lichenin. 

Some longer runs of -(1, 3)-links do occur and sequences of up to 14 -(1, 4)-linkages 

have been reported. In addition small amounts of an exclusively -(1, 3)-linked glucan, 

which resembles laminarin, are also present in barley. Apparently, its presence in malt 

has not been investigated. The chain lengths of the -glucan molecules vary greatly, the 

longer chain materials giving rise to more brewing problems and very viscous solutions. 

The enzymes involved in the hydrolytic breakdown of -glucans during malting are 

indicated in Fig. 4.22. In mashing the amount of -glucan that dissolves is increased by 


H 2O 

β-Glucosidases 

H2O β-glucan 

solubilase 

activities 

mashing 

G–G–(G)···G–G···G···(G)–G–G–G–G···(G)–G–G– 

Limited attack 

laminarinase 

(β(1→3) 

glucanase) 

Insoluble, mixed link β-glucan 

H 2O 

Limited attack 

microbial 

cellulase 

(β(1→4) 

glucanase) 

Investing 

protein; crosslinking 

peptides? 

H 2O 

G–G–(G) a···G–G···G···(G) m–G–G–G–G–G···(G) a···G–G···G 

Soluble, mixed-link β-glucan 

H 2O 

Endo-β-glucanase(s) 

G–G···(G) p···G–G–G···G 

G···G···(G) q···G–G–G 

H 

G–G–(G) 

2O 

r···G 

Exo-β-glucanase(s) β-Glucan oligosaccharides 

(?) 

H2O Endo-β-glucanase(s) 

G–G or G···G 

Cellobiose Laminaribiose 

H 2O 

β-Glucosidases 

G Glucose 

Laminarinase 

H2O H2O β-Glucosidases 

Acidic 

carboxypeptidase 

Degradation 

products 

G···G···G···(G) n···G···G···G···G···G···G 

Laminarin-like β-glucan 

H2O Laminarinase 

(β(1→3)glucanase) 

G···G···(G) n···G···G 

‘Laminarin’ oligosaccharides 

Fig. 4.22 A summary of the activities of the enzymes believed to be involved in the hydrolytic breakdown of 

-glucans in germinating grain (Briggs, 1998). G, -D-glucopyranose residue; ± -(1, 4)-link; , -(1, 3)-link. 


the activity of aheat-stable enzyme (or mixture of enzymes) now termed -glucan 

solubilase (Luchsinger et al., 1958; Scott, 1972; MacGregor and Yin, 1990; Bamforth et 

al., 1997). The nature of this enzyme has not been established, and the enzyme activity 

has been attributed to avariety of enzymes acting separately or in concert such as - 

(1,3)-glucanase, cellulase ( -(1,4)-glucanase),andpeptidases.Theseactivitiesmayexert 

their effects by partial degradation of the glucan chains to shorter, more soluble 

fragmentsortotheremovalofìnvesting'materialsassociated withthecellwallstructure 

(such as proteins and/or pentosans). The most significant enzymes are the endo- - 

glucanases, which degrade the chains by random attack on the susceptible bonds, 

reducing the viscosity of their solutions. The malt endo- -(1,3)-glucanase, sometimes 

called laminarinase, is relatively heat stable but will only attack polysaccharides within 

sequences of consecutive (1,3)- -links. Two isozymes have been noted, with the same 

pH optimum ofabout 5.6. Their significance inmashing isuncertain. Malts contain small 

and variable amounts of cellulase(s), enzymes that can only attack -glucan with 

consecutive runs of (1,4)-bonds. It is probable that much of this enzymic activity 

originates in microbes on the malt. 

The more important enzyme, with two isozymes, is properly called endo-(1,3; 1,4)- 

-glucan 4-glucanohydrolase, but is usually referred to as (malt) -glucanase. These 

enzymes have pH optimaof 4.7andhydrolyse the -(1,4)-bond adjacent toasubstituting 

-(1,3)-link, giving rise to oligosaccharides which, since these are not known to 

accumulate during mashing, are presumably rapidly degraded further by -glucosidases. 

Only isozyme EII survives kilning and remains in malt (Fincher, 1992). This enzyme is 

heat labile and amounts remaining in pale malts are very variable. Low levels of malt - 

glucanase can give rise to aneed to supplement mashes with preparations of bacterial - 

glucanases or fungal glucanases. The -glucosidases, which can hydrolyse laminaribiose 

(4.19) and cellobiose (4.18) as well as other -glucosides and which will attack the nonreducing 

chain ends of the -glucan chains, seem to occur as isoezymes having 

significantlydifferentspecificities.Theirroleinmashingisuninvestigated.Theexistence 

of exo- -glucanases, which were postulated to attack the non-reducing chain ends and 

give rise to cellobiose (4.18) and laminaribiose (4.19), has not been confirmed. To be 

effective in mashing the malt used must have been carefully kilned to allow the survival 

of active -glucanase. In many well-modified, strongly cured ale malts little or none of 

this enzyme remains. In all-malt mashes, it is not needed. The enzyme is needed when 

inhomogeneous or under-modified malt (including chit malt) is used or -glucan-rich 

adjuncts are included in the grist. Then, because the enzyme is heat-labile, the mashing 

temperature programme must beadjustedtogive theenzymetimetoact.From aseries of 

isothermal mashes, made at different temperatures, it is seen that at temperatures above 

45ëC (113ëF) increasing amounts of -glucan are extracted into the wort as higher 

temperaturesareused(Fig.4.23).Thiseffectisduetothegreatersolubilityoftheglucans 

at higher temperatures combined with the earlier destruction of the -glucanase at higher 

temperatures. 

In commercial, isothermal mashes it is found that even at temperatures of about 65 ëC 

(149 ëF), it is beneficial to have appreciable -glucanase activity in the malt if flake barley 

is being used as an adjunct even if, as has been estimated, the enzyme activity survives for 

only 2 5 min. Steamed flakes are probably the barley adjunct that most readily releases - 

glucan during mashing while micronized barley releases least, perhaps because the 

polysaccharide is partly degraded by heat during the preparation of the adjunct. Elevated 

wort viscosity, which is an indication of faulty wort production, is often used as a warning 

that too much -glucan is in the wort, and has even been used to `measure' the 


β-Glucan (mg/l) 

600 

400 

200 

55°C 

40°C 

45°C 

50°C 

52°C 

70°C 

(158°F) 

50°C 

(122°F) 

0 40 80 

(104, 113°F) 

40, 45°C 

120 


polysaccharide present. This simplistic view is mistaken (Pierce, 1980; Bathgate and 

Dalgliesh, 1975). Many wort components contribute to its viscosity, including dextrins, 

pentosans, and sugars. The increase of viscosity with increasing -glucan content is not 

linear but is more nearly alogarithmic relationship and the viscosity contributions of the 

wort components are not simply additive. Temperature-programmed mashing is attractive 

inthat,byallowinga`rest'atornearthetemperatureoptimumof -glucanaseitsactivityis 

favoured(Table4.18).Howeverextendedlow-temperaturerestswillalsoallowotherheatlabile 

enzymes, such as phosphatases, glycosidases and proteases, to continue acting, 

possibly with undesirable consequences, such as unduly elevated levels of TSN. 

The fate of the pentosan polysaccharides in mashing has not been adequately 

investigated. Pentosans vary in their sizes and detailed structures (Briggs, 1998; 

Fincher, 1992). However, they are all based on chains of xylose ( -D-xylopyranose 

(4.10)) molecules joined through (1, 4)-links. The xylose chains are variously 

substituted with arabinose ( -L-arabinofuranose (4.11)) units. A xylose residue may 

be unsubstituted, or be substituted on C-2, C-3 or in both positions. The substitutions 

occur irregularly along each chain. Arabinoxylans from malt tissues other than the 

starchy endosperm may also be substituted with D-glucuronic acid and perhaps 

galacturonic acid residues. In addition some of the arabinose substituents may have a 

single xylose residue attached. Many studies have treated the pentosans as though they 

were pure polysaccharides. However, a proportion of the arabinose units are substituted 

with phenolic acids, overwhelmingly ferulic acid (4.131), and the molecules are 

acetylated. It is assumed that the acetyl residues are attached to the xylose units. The 

80°C 

(176°F) 

60°C 

(140°F) 

(131°F) 

55°C 

52°C 

(126.6°F) 

Fig. 4.23 The extraction, in time, of -glucan, ÐÐÐ , and the activity of -glucanase, ..........., in 

isothermal mashes made at various temperatures (data of Home et al., 1993). 


120 

100 

80 

60 

40 

20 

0 

β-Glucanase activity (U/kg)

H 2O 

β-Xylopyranosidase 

H 2O 

X 

X 

X 

X 

Exoxylanase 

(?) 

F 

A 

A A 

X–X–X–X–X–X–X–X–X–X–X–X–X–X 

Ac A Ac Ac Ac A A 

Acetic H2O F 

acid Ferulic 

acid 

H2O Deacetylase Deferuloylase 

A A 

X–X–X–X–X–X–X–X–X–X–X–X–X–X–X 

A 

A A 

X–X–X–X–X–X–X–X 

Xylan 

H 2O 

H 2O 

Arabinosidase 

A Arabinose 

Endoxylanase 

consequences of these substitutions on the polysaccharides probably include increased 

hydrophobicity, with consequent hydrophobic binding between molecules, including 

some proteins, and decreased solubility. 

Esterase enzymes must be present to remove the acylating (acetic acid and ferulic 

acid) substituents and so expose the polysaccharide to attack by carbohydrases. Mixtures 

of microbial esterases and carbohydrases act synergistically to break down pentosans. 

Pentosans bind large amounts of water, and it is this characteristic of the hemicellulose, 

present in the fine particles, that is believed to contribute to their slowing wort separation 

from mashes. The enzymes believed to be involved in the degradation of pentosans 

during malting are indicated in Fig. 4.24. The esterases are unstable in buffer solutions 

H 2O 

F 

Arabinoxylan 

Endoxylanase 

A 

H 2O 

X–X–X–X–X–X–X 

A A 

Arabinoxylan 

oligosaccharides 

H 2O 

Arabinosidase 

A Arabinose 


Shorter chains, Xylan oligosaccharides 

X–X 

Xylobiose 

X 

Xylose 

H 2O 

β-Xylopyranosidase 

H 2O 

X–X–X 

Xylotriose 

Endoxylanase 

H 2O 

A 

Deferuloylase 

Acetic 

acid 

X–X–X–X–X 

A 

F 

A 

Fate ? 

A 

F 

F 

A 


A 

F 

A 

H 2O 

A 

F 

Arabinosidase 

A Arabinose 

H 2O 

Endoxylanase 

F 

A 

X–X–X–X–X + X–X 

A 

F 

A 

A 

F 

Acetylated and 

feruloylated 

arabinoxylan 

oligomers 

Fig. 4.24 A scheme of the activities of the enzymes believed to be involved in the hydrolytic 

breakdown of the grain pentosans during malting (Briggs, 1998). X, -D-xylpyranose residues 

(4.10) linked (1, 4) in chains. A, -L-arabinofuranose residues (4.11) Ac and F, acetyl and feruloyl 

substituents, acetic acid and ferulic acid (4.131) attached to the polysaccharide through ester links. 


over 30ëC (86ëF) (Humberstone and Briggs, 1998). However, feruloyl esterase is active 

in mashing with atemperature optimum of around 45ëC (113ëF) (McMurrough et al., 

1984, 1996; Narziss et al., 1990). 

The ferulic acid (4.131) liberated is apotential anti-oxidant and if decarboxylated 

during boiling or by bacteria or pof + yeast strains, gives rise to 4-vinyl guaiacol (4.134, 

Fig. 4.33 on page 158), astrongly flavoured substance that is undesirable in most beers. 

Which other enzymes are involved in pentosan degradation during mashing is not clear, 

but the endo-xylanases, which are relatively heat stable, are probably involved and the 

greater amounts of arabinose (4.11) and xylose (4.10) found in temperature-programmed 

mashes, relative to isothermal infusion mashes, indicates that heat-labile glycosidases are 

active, at least at lower temperatures. Malt contains an inhibitor of xylanase (Debyser et 

al., 1997b). Some 70 90% of malt gums are pentosans, the remainder being chiefly - 

glucans. Quoted values are not consistent, but malt may contain 6.4 6.9% pentosan, of 

which 0.49 0.69% is water-soluble (Debyser et al., 1997a). Some beers contain 

0.3 0.5% (w/v) non-starch polysaccharides, (which may comprise 10% of the beer 

carbohydrate), of which c. 70% is pentosan, 23% is -glucan and 7% is other materials 

(Han and Schwarz, 1996). In 15 other beers the arabinoxylans were in the range 

514 4211mg/l, while -glucans were in the range 0.3 248mg/l (Schwarz and Han, 

1995). 

4.4.4 Proteins, peptides and amino acids 

In some analytical systems the nitrogen content of amaterial 6.25, (or some other 

factor), is reported as the protein content. This is incorrect and misleading, since many 

substances besides proteins contain nitrogen. Proteins consist primarily of chains of 

amino acids joined by peptide links (Figs 4.25 and 4.26). Aprotein may consist of one or 

more long polypeptide chains which may or may not be covalently cross-linked through 

disulphide bonds between two cysteine (4.31) residues (cystine; 4.32). The chain(s) are 

folded together in particular ways and the biological activities of proteins, such as 

enzymes or lectins, depend on the folding being correct, that is the protein is in its 

`native' state. If the folding is disrupted, for example by heat, then the biological activity 

is lost, the protein is `denatured' and, if it was in solution, it may precipitate. Protein 

solution followed by thermal denaturation and precipitation occurs in temperatureprogrammed 

mashing and more aggregation and precipitation occurs during the hop-boil, 

giving the trub. Native proteins may be soluble or insoluble. Proteins may be substituted 

with various molecules, such as sugars in the case of glycoproteins or haem or other 

prosthetic groups in the cases of some enzymes, such as peroxidase. 

The proteins of cereals are often considered in groups defined by their solubilities 

(Briggs, 1998). Globulins are soluble in pure water, while both albumins and globulins 

are soluble in salt solutions. Many enzymes occur in these soluble fractions but insoluble 

enzymes also occur in malt. The hordeins are soluble in hot solutions of aqueous 

alcohols, and their solubility is enhanced if reducing agents are also present (e.g. 70% 

ethanol containing 2-mercaptoethanol, or 60% n-propanol containing sodium borohydride). 

The glutelins are soluble only in solutions of strong alkalis. The last two fractions 

are largely either reserve materials or, in the case of the glutelins, they may have 

structural functions. Some cereal proteins have unexpected properties, for example, 

solubility in light petroleum (thionins). When degraded by hydrolysis proteins give rise to 

peptides, shorter chains of amino acids (Fig. 4.25), and eventually free amino acids (Fig. 

4.26 (4.24±4.49)). 


R 6 

R 1 –CH–COOH 

NH 2 

Generalized amino acid 

R 1 –CH–CO–NH–CH–COOH 

NH 2 

Generalized dipeptide 

Generalized tripeptide 

Generalized protein 

The amino acids differ in their properties and, when joined in peptide chains, their 

side-chains largely define the properties of the peptide or protein of which they form a 

part. The amino acid cysteine (4.31) is of interest since the thiol (-SH) on the side chain 

can be oxidized to give the `di-amino acid' cystine, containing adisulphide link (4.32). 

Thus peptide chains can be cross-linked by covalent disulphide bonds formed by 

oxidizing cysteine residues in the chains, a fact that is probably important in 

polymerizing gel-proteins during mashing (Van den Berg and Van Eerde, 1982; Muller, 

1995). Conversely thesebondsmay be split by reducing the disulphide bridges topairs of 

thiols. The reduction may be brought about with sodium borohydride, or by thioldisulphide 

exchange or by cleavage with bisulphite ions. Presumably, the enhanced rate 

of wort run-off caused by the experimental addition of bisulphites to mashes is caused by 

the altered structure of the proteins in the fine particles or Oberteig following the 

breaking of disulphide cross-links. 

Thousands of proteins have been detected and from the brewing point of view it is 

usually most convenient to consider them in groups defined by particular properties. 

However, it is sometimes necessary to consider individual proteins, for example, because 

of their enzymic capabilities, or because they bind to particular sugars (i.e. they are 

lectins) or they bind to lipids or because they are involved in foam formation and 

stability, or because they are involved in binding to polyphenols giving adducts which 

can form hazes in beers. These `haze-forming proteins' can be selectively removed from 

beer by adsorbtion onto silica hydrogels, or can be selectively degraded by proteolytic 

enzymes such aspapain.These proteins and polypeptidesappeartobedistinct from those 

which add to the `body' of the beer and those which help form and stabilize foam. 

Surprisingly some malt proteins or modified products (protein Z, 40kDa, and LTP 1, a 

lipid transfer protein, 10kDa, are examples) partly survive mashing and boiling and 

appear, sometimes partly modified, in beer. 

R 2 

R 3 –CH–CO–NH–CH–CO·NH·CH–COOH 

NH 2 

R 4 

R various ( ) n 

NH 2–CH–CO– NH–CH–CO –NH–CH–CO–NH–CH–COOH 

Fig. 4.25 Generalized formulae of an -amino acid, adipeptide, atripeptide, and asection of a 

polypeptide chain, as found in proteins and their degradation products. The various side-chains, R, 

(Fig. 4.26) differ in their reactivities and in some cases may be substituted, for example with 

carbohydrates. 


R 7 

R 5 

R 8

(4.24) 

(4.25) 

(4.26) 

(4.27) 

(4.28) 

(4.29) 

(4.30) 

(4.31) 

(4.32) 

(4.33) 

(4.34) 

(4.35) 

(4.36) 

(4.37) 

CH 3·CH(NH 2)·COOH 

α-Alanine* 

NH 2·CH 2·CH 2·COOH 

β-Alanine 

HOOC·(CH 2) 2·CH(NH 2)·COOH 

α-Aminoadipic acid 

NH 2·(CH 2) 3·COOH 

γ-Aminobutyric acid 

NH 2 

HN 

Arginine* 

HOOC·CH 2·CH(NH 2)·COOH 

Aspartic acid* 

NH 2·CO·CH 2·CH(NH 2)·COOH 

Asparagine* 

HS·CH 2·CH(NH 2)·COOH 

Cysteine* 

Cystine* 

HOOC·(CH 2) 2·CH(NH 2)·COOH 

Glutamic acid* 

NH 2·CO(CH 2) 2·CH(NH 2)·COOH 

Glutamine* 

NH 2·CH 2·COOH 

Glycine* 

C·NH·(CH 2) 2·CH(NH 2)·COOH 

S·CH 2·CH(NH 2)·COOH 

S·CH 2·CH(NH 2)·COOH 

N 

HO 


NH 

Histidine* 

COOH 

N 

H 

Hydroxyproline* 

* Occurs in proteins. 

(4.38) 

(4.39) 

(4.40) 

(4.41) 

(4.42) 

(4.43) 

(4.44) 

(4.45) 

(4.46) 

(4.47) 

(4.48) 

(4.49) 

CH 3·CH 2 

Isoleucine* 

CH 3 

CH 3 

CH 3 

Leucine* 

NH 2·(CH 2) 4·CH(NH 2)·COOH 

Lysine* 

CH 3·S·(CH 2) 2·CH(NH 2)·COOH 

Methionine* 

Phenylalanine* 

HO·CH 2·CH(NH 2)·COOH 

Serine* 

CH 3·CH(OH)·CH(NH 2)·COOH 

Threonine* 

Tryptophan* 

Tyrosine* 

Valine* 

CH·CH 2·CH(NH 2)·COOH 

COOH 

N 

H 

Proline* 

CH·CH(NH 2)·COOH 


COOH 

N 

H 

Pipecolinic acid 

(= piperidine- 

2-carboxylic acid) 

HO CH 2·CH(NH 2)·COOH 

CH 3 

CH 3 

N 

H 


CH·CH(NH 2)·COOH 

Fig. 4.26 The formulae of most common amino acids and the imino acids pipecolinic acid, 

proline and hydroxyproline. 


It was believed that virtually all the soluble nitrogen-containing substances in wort, 

prepared by isothermal mashing at about 65 ëC (149 ëF), were preformed in malt. This is 

incorrect. The amounts of soluble nitrogen depend on the malt and the way that it is 

mashed. In mashes made at 65 ëC (149 ëF) about 50% of the total soluble nitrogen (TSN) 

and 30 50% of the free amino nitrogen (FAN) is formed by enzyme action during 

mashing. The mixture of malt enzymes involved in the hydrolytic breakdown of proteins 

is complex (Briggs, 1998; Enari, 1986; Enari et al., 1964; Burger and Schroeder, 1976; 

Mikola et al., 1972; Sopanen et al., 1980). The major endo-peptidases, or proteases attack 

polypeptide chains at various particular locations. Both insoluble (`bound') and soluble 

(`free') enzyme activities have been detected. The most important proteases are thioldependent, 

have pH optima in the range 3 6.5 and contribute about 90% of the 

proteolytic activity. Metalloproteases (pH optima 5 8.5) provide most of the remaining 

activity, but serine proteases and aspartate proteases are also present. 

About 42 soluble endopeptidases have been detected (Zhang and Jones, 1995a, b). In 

addition 4 5 amino-peptidases (pH optima 5.5 7.3, which attack peptide chains at their 

amino-termini) and at least four carboxypeptidases (pH optima 4 6, which attack peptide 

chains at the carboxyl termini) together with two alkaline peptidases (pH optima 8 10) 

are also present. In mashing, the proteases and the carboxypeptidases are the most 

important enzymes in generating soluble nitrogenous substances. There is normally an 

èxcess' of carboxypeptidase activity, and so the rate-limiting activity is due to limiting 

amounts of proteases. At the end of mashing there is always a substantial amount of 

protein remaining in the spent grains and so a lack of substrate is not what limits the 

generation of soluble nitrogen. During mashing ammonium ions are released as well as 

peptides and amino acids (Jones and Pierce, 1967; Pierce, 1982). Presumably the 

ammonium ions arise from the hydrolysis of glutamine (4.34) and asparagine (4.30) by 

amidases. Apparently transaminases are not active during mashing, but some glutamic 

acid (4.33) may be enzymically decarboxylated to give -amino butyric acid (4.27). With 

such a complex mixture of enzymes involved it is not surprising that alterations in 

mashing conditions can have dramatic effects on the patterns of nitrogenous substances 

present in the wort, including the proportions of amino acids. The reported temperature 

òptima' of permanently soluble nitrogen (PSN), after 15 min., 1h. and 3h. mashing, were 

61 ëC (141.8 ëF), 58 ëC (136.4 ëF) and 53 ëC (127.4 ëF) respectively while the 

corresponding values for formol-nitrogen were 59 ëC (138.2 ëF), 52 ëC (125.6 ëF) and 

50 ëC (122 ëF). 

Mash tun adjuncts are often used as sources of extract that will act as `nitrogen 

diluents', that is, wort prepared using these adjuncts will contain less soluble nitrogen 

than an all-malt wort having the same extract content. However, mash tun adjuncts 

contribute some nitrogenous substances to the wort. Wheat adjuncts contribute 

polypeptide material that favours foam formation and stability, as do preparations of 

raw barley. Raw barley (like raw wheat) contains proteins that inhibit proteases from malt 

and some microbial enzymes. Thus the addition of a barley adjunct to a mash can reduce 

the level of wort-soluble nitrogen to a disproportionate extent and in barley brewing it is 

necessary to ensure that a sufficient amounts of a protease is used to ensure that enough 

soluble nitrogen is generated during mashing. In one trial the amounts of -amino 

nitrogen in worts (as mg N/kg grist) were, with all malt, 949; with 16% malt replacement 

with the named adjunct, wheat, 763; rice, 821; maize, 832 (Jones, 1974). 

The proteins and polypeptides that survive into the beer contribute to the `body' and 

`mouth-feel' of the beer, its foaming characteristics, and its susceptibility to haze 

formation. The colour of the beer is influenced by Maillard reactions between the 


sugars and amino-compounds (including the amino acids) during the hop-boil, which 

give rise to coloured and flavoured substances. The proportions of the flavoured 

fermentation products made by yeast are dependent on the nitrogenous substances that 

are present. The rate of wort separation is reduced by the presence of inadequately 

degraded `gel proteins' in the mash contributing to the fine particles and the Oberteig 

whichimpedetheflowofthewortthroughthegoods.Inwortsfromall-maltmashesthe 

levels of amino acids are nearly always adequate for good yeast growth. However, in 

worts made using mash tun and/or copper adjuncts the FAN levels may fall below the 

100 140mg/litre level, which is regarded as the minimum needed for trouble-free 

fermentations. 

4.4.5 Nucleic acids and related substances 

Nucleic acids make up 0.2 0.3% of malt. About 70% of the nucleic acid is 

deoxyribonucleicacid, DNA(4.50)and30%ribonucleicacid,RNA(4.51).Theenzymic 

hydrolysis ofthenucleic acidsfirst givesriseto nucleotides(base sugar phosphate), 

then nucleosides (base sugar +inorganic phosphate) and finally free bases and sugars. 

In addition, the malt contains avariety of other nucleotides and derivatives including 

ATP (adenosine triphosphate, (4.53)), NAD + (nicotinamide adenine dinucleotide, 

(4.54)), UDPG (uridine diphosphate glucose, (4.55)) and so on. The nucleic acids are 

chains of alternating phosphate and sugar (ribose (4.12) or deoxyribose (4.52)) residues, 

eachsugarunitbeingsubstitutedwithapurineorapyrimidinebase(Fig.4.27).Together 

these substances contribute 8 9% to the total nitrogen content of malt. During mashing 

the degradation of the more complex materials appears to be nearly complete, the 

products in the wort being free bases (adenine (4.56), guanine (4.57), cytosine (4.58), 

uracil (4.59) and thymine (4.60)) or breakdown products such as allantoin (4.61), 

hypoxanthine (4.62) and xanthine (4.63). The related nucleosides (e.g. adenosine, 

guanosine) and deoxynucleosides, in which the bases are attached to ribose or 

deoxyribose, are also present, as are smaller amounts of nucleotides. (Briggs, 1998; 

Ziegler and Piendl, 1976). 

Several nucleases are present in malt and these are sufficiently stable to ensure the 

complete hydrolysis of the nucleic acids to nucleotides, molecules in which the 

nucleoside structures (base-sugar) are phosphorylated (base-sugar-phosphate) on the 

sugar residues in various positions. The phosphatase (nucleotidase) enzymes are also 

active as dephosphorylation of the nucleotides to nucleosides proceeds during mashing. 

The nucleosidases, which hydrolyse nucleosides to their constituent bases and sugars, are 

evidently more heat labile since highly kilned malts give relatively more nucleosides to 

free bases when mashed, compared to lightly kilned malt and mashing at higher 

temperatures gives worts richer in nucleosides. Some nucleic acid breakdown products 

are known to have flavour-enhancing properties, but the amounts reaching beer are so 

small that they can have only a marginal effect. Yeast probably uses the free bases to 

support growth in the initial stages of fermentation. 

4.4.6 Miscellaneous substances containing nitrogen 

A wide range of nitrogen-containing substances, besides proteins, peptides and amino 

acids, occur in malt and wort in widely differing amounts (Briggs,1998; Engan, 1981; 

MacWilliam, 1968; Fig. 4.28). Ammonia, as ammonium ions, and many amines, often 

formed by the decarboxylation of amino acids, are found in worts. Thus glycine (4.35) 


B 

H 

H 

B 

H 

H 

B 

H 

H 

O 

H2C–O··· x O 

H 

H 

O H2C–O– P O 

x O 

H 

H 

O H2C–O– P O 

x O 

H 

H 2C–O– P O 

NH 2 

OH 

C 

N C 

N 

CH 

HC C 

N N O 

OH 

H H 

H H 

OH 

OH 

HO·CH 2 

H 

H 

O 

OH H 

OH 

H 

H 

* 

Deoxyribose; β-Ddeoxyribofuranose 

(4.52) 

Deoxyribonucleic acid, DNA 

(x = H, Major bases, B: Adenine, guanine, cytosine, thymine) 

Ribonucleic acid, RNA 

(x = OH, Major bases, B; Adenine, guanine, cytosine, uracil) 

OH 

O O O 

CH 2·O·P·O·P·O·P·O – 

O – 

O – 

Adenosine triphosphate ATP 

(4.53) 

O – 

CH2OH H O 

H 

OH 

HO 

H OH 

NH 2 

O·R 

OH 

(4.54) 

OH 

(4.50) 

(4.51) 

O CH O O 

2·O·P·O·P·OCH2 O 

+ 

N 

H 

H 

H 

H 

O H 

H 

H 

H 

– O – 

N C 

C 

N 

CH 

N 

C 

HC 

N 

H 

O·P·O·P·O·CH2 O 

O 

O 

N 

H 

H 

H 

H 

OH 

H·N 

OH 

(4.55) Uridine diphosphate 

glucose, UDPG 

OH 

Nicotinamide adenine. 

dinucleotide, NAD + 

R = H 

NADP + , R = phosphate 

CO·NH 3 

Fig. 4.27 The structures of the nucleic acids, the purine and pyrimidine bases and the breakdown 

products allantoin, xanthine and hypoxanthine and of three chemically related cofactors. 


Fig. 4.27 Continued. 

gives rise to methylamine (4.64), alanine (4.24) gives ethylamine (4.65), valine (4.49) 

gives isobutylamine (4.66), phenylalanine (4.42) -phenylethylamine (4.67), tyrosine 

(4.48) tyramine (4.68), histidine (4.36) histamine (4.69), tryptophan (4.47) tryptamine 

(4.70), and proline (4.44) pyrrolidine (4.71). Similarly, the decarboxylation of arginine 

(4.28) yields agmatine (4.72), while ornithine (4.73) and lysine (4.40) give the diamines 

putrescine ((4.74); 1, 4-diaminobutane) and cadaverine ((4.75) 1, 5 diaminopentane), 

respectively. These last are precursors of spermine (4.76) and spermidine ((4.77) Briggs, 

1978). The origins of dimethylamine (4.78), trimethylamine (4.79) p-hydroxybenzylamine 

(4.80) and gramine (4.81), butylamine and amylamine are less obvious. The 

methylated derivatives of tyramine (4.68), N-methyl-tyramine (4.82), di-N-methyl 

tyramine ((4.83), hordenine) and the quaternary tri-N-methyl-tyramine ((4.84), candicine) 

are also present. 


Awide range of heterocyclic, N-containing substances also occurs in worts prepared 

using dark or roasted malts or adjuncts. Various other N-containing substances are also 

vitamins and/or yeast growth factors. Other substances, which should be present in only 

tiny amounts or be absent, include nitrosoamines and hydrocyanic (prussic) acid (Fig. 

4.28). Nitrosamines can arise when malt is kilned and oxides of nitrogen are present in 

the air. Using modern malting techniques these substances should be nearly absent. The 

compound originally attracting interest was N-nitrosodimethylamine (NDMA, (4.85)), 

which was formed on the surface of the malt, but other substances, such as Nnitrosoproline 

have subsequently been detected. Barley malts contain widely variable 

amounts of cyanogenic glycosides, such as epi-heterodendrin (4.86). The amounts 

formed during malting are strongly influenced by the barley variety. This compound can 

be hydrolysed to glucose and isobutyraldehyde cyanohydrin by the enzyme - 

glucosidase, then the cyanohydrin breaks down to isobutyraldehyde and hydrocyanic 

acid. Traces of this acid have been reported in beers, but the levels present are 

insignificant. This is not the case in distilleries, when the acid can give rise to urethane 

(ethyl carbamate, (4.87)). Some sorghum malts contain very large amounts of dhurrin 

(4.88), another cyanogenic glycoside, and in these cases there is a clear risk that 

significant levels of hydrocyanic acid (HCN) may reach beers made from them (Briggs, 

1998). As with barley the amount of cyanogenic glycoside present in the sorghum malt is 

strongly influenced by the variety. 

4.4.7 Vitamins and yeast growth factors 

Many of the growth factors needed by yeast contain nitrogen (Table 4.19; Fig. 4.29). In 

addition to the substances mentioned in this section brewer's yeast needs some sterols 

and unsaturated fatty acids for growth, after periods of anaerobic growth (Chapter 12). 

The quantities of vitamins reported in worts vary widely. In part these discrepancies 

probably represent real differences and in part are caused by difficulties with the 

bioassays used in the estimations. Many of these substances occur combined in various 

ways and these may or may not be broken down during mashing and may or may not be 

available to yeast. Thus folic acid (4.89) and related compounds occur in various forms, 

nicotinic acid occurs (as nicotinamide (4.90)) in the oxidation/reduction cofactors NAD + 

(4.54) and NADP + ,and myo-inositol (4.91) occurs combined with phosphate in phytic 

acid (4.156) and in some lipids (e.g. (4.122)). Riboflavin (4.92) occurs in the oxidation/ 

reduction co-factors flavin mononucleotide (FMN, (4.93)) and flavin adenine dinucleotide 

(FAD, (4.94)), while thiamine (4.95) occurs as thiamine pyrophosphate. 

Malt contains some fat-soluble vitamin precursors (Briggs, 1998), of which the 

tocopherols (vitamin E) might be significant, but it is unclear if any of these reach the 

hoppedwort.Thewatersolublevitaminsaresignificant(Table4.19).AlthoughvitaminC 

(ascorbic acid (4.96) and dehydroascorbic acid (4.97)) are present in green malt these 

materials are destroyed during kilning. Traces of vitamin B12 have sometimes been found 

in malt, but the significance is unclear. It is not known what alterations may occur to 

vitamins and their precursor substances during mashing. At first it seems strange that so 

many uncertainties surround these compounds. The reason is probably that the levels 

present in conventional worts rarely or never limit yeast growth and fermentation, and so 

there is no stimulus for investigating their origins and fates. However, problems with 

fermentations have been encountered with high-adjunct, barley brews and these have 

been overcome by the addition of yeast extracts to the worts. Various water soluble 

vitamins and growth factors are known to be present in these complex preparations and so 


CH 3·NH 2 

Methylamine 

(4.64) 

CH3·CH2·NH2 Ethylamine 

(4.65) 

(CH3) 2CH·CH2·NH2 Isobutylamine 

(4.66) 

HO CH 2·CH 2·NH 2 

N 

CH 2·CH 2·NH 2 

Phenylethylamine 

(4.67) 

Tyramine 

(4.68) 


NH 

Histamine 

(4.69) 


N 

H 

Tryptamine 

(4.70) 

N 

H 

Pyrrolidine 

(4.71) 

NH 2·(CH 2) 4NH·C 

Agmatine 

(4.72) 

NH2·(CH2) 3·CH(NH2)·COOH Ornithine 

(4.73) 

NH2·(CH2) 4·NH2 Putrescine 

(4.74) 

NH2·(CH2) 5·NH2 Cadaverine 

(4.75) 

NH2·(CH2) 3·NH·(CH2) 4·NH·(CH2) 3·NH2 Spermine 

(4.76) 

NH 

NH 2 

NH2·(CH2) 3·NH·(CH2) 4·NH2 Spermidine 

(4.77) 

HO 

CH 2·CH 2·N(CH 3) 2 

Hordenine; di-N-methyl tyramine 

HO 

(CH3) 2·NH 

Dimethylamine 

(4.78) 

Trimethylamine 

HO CH 2·NH 2 

p-Hydroxybenzylamine 

(4.80) 

N 

H 

Gramine 

(4.81) 

(4.83) 

+ 

CH2·CH2·N(CH3) 3 

Candicine; tri-N-methyl tyramine 

(4.84) 

OH – 

NDMA; N-Nitroso-dimethylamine 

(4.85) 

(4.86) 

NH2·COO·C2H5 Urethane (ethyl carbamate) 

(4.87) 

CH 2·N(CH 3) 2 

HO CH 2·CH 2·NH·CH 3 

N-Methyltyramine 

CH 3 

CH 3 

CH2OH CH(CH3) 2 

H O 

O C CN 

H 

OH H 

HO H 

H 

H OH 

(CH 3) 3N 

(4.79) 

(4.82) 

N·NO 

epi-Heterodendrin 

CH2OH H 

H O 

O H 

OH H 

HO OH 

C 

CN 

OH 

H OH 

Dhurrin 

(4.88) 

Fig. 4.28 Some amines and other nitrogenous substances found in wort. 


Table 4.19 Vitamins and yeast growth factors in wort (Briggs et al., 1981; MacWilliam, 1968). 

The amounts of these factors present in worts are likely to vary considerably with the different 

compositions of grists and with differing wort concentrations 

Growth factor or vitamin Reported concentrations 

Choline (4.101) 20±25 mg/100ml 

myo-Inositol (4.91) bound 1±6mg/100ml 

free 1.5±3.5 mg/100ml 

Thiamine ((4.95); aneurine; vitamin B1) 28±75(155) g/100ml 

Riboflavin ((4.92); vitamin B2) 33±90 g/100 ml 

Folic acid (4.89), and related substances 10±13 g/100 ml 

Nicotinic acid ((4.90); niacin) 0.8±1.8 mg/100 ml 

Pyridoxin ((4.100); pyridoxal, pyridoxamine, vit. B6) 59±105 g/100 ml 

Biotin (4.98) 0.8±1.2 g/100 ml 

Pantothenic acid (4.99) 48±98 g/100 ml 

these may be partly or wholly responsible for the improved fermentation performance of 

the yeast. 

4.4.8 Lipids in mashing 

Malt contains about 3.5% of lipids and is the major source of lipids in all-malt beers. 

Most adjuncts contain less lipid than malt, and specifications often specify the maximum 

that any batch may contain. They also contribute some lipid to the wort. Older reports 

often quoted low values for lipid contents, because only the non-polar materials were 

extracted by the methods then in use. Malt lipids are acomplex mixture containing 

hydrocarbons, fatty acid esters, sterol esters, esterified steryl glycosides, waxy esters, 

monoglycerides, diglycerides and triglycerides, free sterols, free fatty acids, long chain 

alcohols, phospholipids, glycolipids, carotenoids and tocopherols (Figs 4.30 and 4,31; 

Table 4.20; Anness, 1984; Briggs, 1978, 1998; MacWilliam, 1968; Morrison, 1978, 

1988). Brewing science has concentrated on relatively few of these groups of substances. 

Using the analysis of fatty acids, the lipid types present in afree wort (as mg fatty 

acid/litre) were reported to be: phospholipids+glycolipids 14.8, monoglycerides, 1.7; 

diglycerides, 2.8; triglycerides, 15.3; free fatty acids, 28.4; steryl esters, 1.0 and 

unknowns, 0.3 (Table 4.21). These analyses do not record lipids lacking fatty acids in 

their make-up. Triglycerides (4.102) predominate in malt, but free fatty acids (4.107± 

4.113) predominate in the wort. The amounts of lipid extracted into wort is increased by 

using better modified malts, finer grinding, higher mashing and sparging temperatures, 

thinner mash beds during wort separation, careless and excessive raking in the lauter tun, 

by using smaller proportions of adjuncts, and by running off faster, by `squeezing' the 

mash to recover residual extract, or by adopting other techniques to maximize extract 

recovery. Much of the lipid in amash is present as oil droplets spread among the grist 

particles. In general, more turbid worts carry more lipids and techniques are usually 

adopted to minimize turbidity and the amounts of lipid remaining in the wort, even 

though their presence can increase the fermentation rate (Chapter 12). Thus the wort is 

recirculated through the filter bed until it `runs bright', or it can be filtered through 

kieselguhr or centrifuged. The last two treatments are probably not much used on the 

production scale. 

More lipids are also removed at later stages of the brewing process, for instance, 

during wort boiling and clarification, but it is sound practice to obtain the sweet worts as 


H 2N 

H 3C 

H·N 

N 

H 3C 

H 3C 

OH 

N 

Riboflavine, vitamin B 2 

(4.92) 

CH 2OH 

(CHOH) 3 

CH 2 

N N 

N 

CH 2·NH CO·NH·CH·CH 2·CH 2·COOH 

Pteridine p-Aminobenzoic 

acid residue 

N 

O 

S 

NH 3 

N 

N·H 

N 

N 

(4.98) 

Pteroic acid 

Cl 

CH 2 

N 

Cl 

O 

Folic acid 

(4.89) 

NH 

O 

Vitamin B 1 (Thiamin, Aneurine) 

S 

Ascorbic acid 

(4.95) (4.96) 

(CH2) 4·COOH 

Biotin 

CH 3 

CH 2·CH 2·OH 

CH 3 

O 

COOH 

Glutamic acid residue 

Flavin mononucleotide; 

FMN or riboflavin phosphate 

O 

H·N 

O 

OH OH 

N 

CH 2·OH 

CH·OH 

(4.93) 

N 

Nicotinic acid (R = OH) 

Nicotinamide (R = NH 2) 

CH 3 

O N N CH3 CH 2 

H·C·OH 

H·C·OH 

H·C·OH 

2– 

CH2OPO3 HO·CH 2·C·CHOH·CO·NH·CH 2·CH 2·COOH 

CH 3 

Pantothenic acid 

(4.99) 

β-alanine residue 

O 

O 

O O 

(4.97) 

CH 2·OH 

CH·OH 

(4.90) 

Dehydroascorbic acid 

HO 

H 3C 

R 

N 

CO·R 

H·N 

O 

Pyridoxin (R = CH 3OH) 

Pyridoxal (R = CHO) 

Pyridoxamine (R = CH 2·NH 2) 

(4.100) 

N 

CH 2 

H·C·OH 

H·C·OH 

H·C·OH 

CH 2 

O 

– O P O 

O 

(4.94) 

OH OH 

H OH 

H 

OH 

H 

H 

HO H 

H OH 

myo-inositol 

(4.91) 

CH 3 

O N N CH3 NH 2 

– 

O O P 

O 

N N 

CH2 O 

N N 

H 

H 

H 

OH OH 

Flavin adenine dinucleotide, FAD 

CH 2OH 

(CH 3) 3·N·CH 2·CH 2·OH 

OH 

– 

Choline 

(4.101) 

Fig. 4.29 Some water-soluble vitamins and yeast growth factors that occur in wort and ascorbic and 

dehydroascorbic acids, which occur in green malt. 



R 2 ·CO·O·C·H 


Triglyceride (triacylglycerol) 

(R·CO·OH = fatty acid) 

HO 

H 2·C·O·CO·R 

HO·OH 

H 2·C·OH 

1-Monoglyceride 

(monoacylglycerol) 

CH 3 

CH3 CH3 β-sitosterol 

CH 3 


R 2 ·CO·O·C·H 

H 2·C·OH 

1,2-Diglyceride 

(diacylglycerol) 

H 2·CO·CO·R 1 

HO·CH 

H 2CO·CO·R 2 

1,3-Diglyceride 

(diacylglycerol) 

(4.102) (4.103) (4.104) (4.105) 

bright as possible. Less than 5% of malt lipids are extracted into the sweet wort, the 

amount depending on the equipment used and the way it is operated. The amounts of malt 

lipids in sweet worts (expressed as % of the lipid in the malt) prepared using a traditional 

mash tun, a lauter tun and a traditional mash filter were 0.3%, 1.0% and 4.5% 

respectively (Anness and Reid, 1985). In contrast, the new 2001 filter releases relatively 

little lipid into the wort (Letters, 1994). In another report the lipid contents of worts 

prepared using a deep bed mash tun, a lauter tun, a Strainmaster and a traditional mash 

filter were 10, 50, 150 and 400 mg/litre, respectively (Whitear et al., 1983). 

Lipids have a number of effects in brewing but there are disagreements about the 

details, probably because different criteria and lipid fractions have been used in the 

studies (Letters, 1992, 1994; Letters et al., 1986; Isherwood et al., 1977; Wainwright, 

1980). Unsaturated fatty acids and sterols (Fig. 4.30) have a beneficial effect on yeast, 

improving its fermentation performance, its viability and its resistance to high alcohol 

contents, such as occur in high-gravity brewing. These materials can only be produced by 

yeast in sufficient amounts under aerobic conditions. When turbid worts, which contain 

elevated amounts of lipids, reach the fermenter it is often found that fermentation is 

enhanced. It seems that, in the amounts found normally in beers lipids do not influence 

gushing. On the other hand the lysophospholipids (such as lysophosphatidyl choline, 

(4.23)) complexed with amylose in starch and the lysophospholipids together with the 

free fatty acids present in mashes can slow down starch gelatinization and impede 

amylolysis and the degradation of complexed dextrins. As the lipid-polysaccharide 

complexes do not give positive iodine tests these can be misleading when estimating the 

amounts of starch remaining in the spent grains. 

CH 3 

CH 3 

(4.106) 

Common fatty acids 

CH 3·(CH 2) 12·COOH Myristic acid (14:0) 

CH 3·(CH 2) 14·COOH Palmitic acid (16:0) 

CH 3·(CH 2) 5·CH =CH·(CH 2) 7·COOH Palmitoleic acid (16:1) 

CH 3·(CH 2) 16·COOH Stearic acid (18:0) 

CH 3·(CH 2) 7·CH =CH·(CH 2) 7·COOH Oleic acid (18:1) 

CH 3·(CH 2) 4·CH =CH·CH 2·CH==CH·(CH 2) 7·COOH Linoleic acid (18:2) 

CH 3·CH 2·CH =CH·CH 2·CH =CH·CH 2·CH =CH·(CH 2) 7·COOH Linolenic acid (18:3) 

(4.107) 

(4.108) 

(4.109) 

(4.110) 

(4.111) 

(4.112) 

(4.113) 

Fig. 4.30 Non-polar lipids and fatty acids, with -sitosterol as an example of a sterol. R.CO.O 

represents an esterified fatty acid residue. 


(4.114) 

(4.118) 

(4.120) 

CH 2OH 

HO O 

H 

OH H 

H H 

(4.116) 

H OH 

O 

CH 2 

Mono-β-D-galactosyl 

diglyceride 

R 2·CO·O·CH 

CH·O·CO·R 1 

CH 2·O·CO·R 2 

H 2C·O·CO·R 1 

H 2C·O· P 

CH 2OH 

OH O H 

H 

OH H 

H 

H OH 

Phosphatidyl inositol 

Digalactosyl diglyceride 

The addition of most preparations of lipids to beer reduces head formation and 

survival. The effects are complex, both because different mixtures of lipids have different 

effects and because lipid binding proteins (such as LTP1, a 10 kDa albumin) can survive 

into beer and `mask' the effects of the lipids. Indeed, LTP1 constitutes a major proportion 

of the protein in foam. At least in some combinations, mixtures of different groups of 

lipids act synergistically to destroy foam when added experimentally to beer. Possibly 

O 

HO 

CH 2 

O 

OH H 

H H 

H OH 

Phosphatidic acid Phosphatidylglycerol 


R2·CO·O·CH O 

+ 

H2C·O·P·O·CH2·CH2·N(CH3) 3 

O – 

O 


R 2·CO·O·CH O 


R2·CO·O·CH O 

H 2C·O·P·O·CH 2 

H 2C·O·P·O·CH 2 

O – H·C·OH 

O – H·C·OH O 

CH 2 

CH 2OH 

R 4·CO·O·CH 2 

CH 2·O·P·O·CH 2 



Phosphatidylcholine (lecithin) Diphosphatidylglycerol (cardiolipin) 



H 2CO·P·O·CH 2·CH 2·NH 2 

O – 



O – 

H·C·O·CO·R 3 

H 2·C·O·P·O·CH 2·CH·COOH 

Phosphatidylethanolamine (cephalin) Phosphatidylserine (cephalin) 

OH OH 

H 

H 

OH 

HO 

H 

H 

H OH 

O 

O P O CH 2 

O – 

(4.115) 

(4.117) 



O – 

(4.122) 

NH 2 

(4.119) 

(4.121) 

Fig. 4.31 Examples of polar phospholipids and glycolipids (see also 4.23), phosphate residue. 


Table 4.20 Analyses of the lipid classes in two barley malts (Anness, 1984). The results are 

expressed as the fatty acids in each fraction per unit dry weight, (mg fatty acids/g dry weight). This 

type of analysis does not reveal the presence of lipids that do not contain fatty acids (e.g. free 

sterols) 

Class of lipid Weeah malt Sonja malt 

PHOSPHOLIPIDS (4.116±4.122) 

GLYCOLIPIDS 

2.4 3.2 

Digalactosyl monoglyceride 0.9 1.1 

Digalactosyl diglyceride (4.115) 1.0 1.1 

Monogalactosyl monoglyceride 0.2 0.1 

Monogalactosyl diglyceride (4.114) 0.4 0.7 

NEUTRAL LIPIDS 

Acylsterylglycosides 0 0 

Monoglycerides (4.103) 0.3 0.3 

Free fatty acids (4.107±4.113) 2.0 2.2 

Diglycerides (4.104, 4.105) 1.0 1.2 

Triglycerides (4.102) 19.8 25.3 

Steryl esters 0.3 0.7 

Table 4.21 Examples of the major total fatty acids of the lipids found in the grist, the spent grains 

and the sweet wort from (A) a mash prepared in a mash tun and (B) a mash separated in a lauter tun 

(Anness and Reed, 1985) 

Sample Fatty acid composition (%) 

16:0 18:0 18:1 18:2 18:3 

(4.108) (4.110) (4.111) (4.112) (4.113) 

(A) 

Grist 21.2 0.8 9.7 58.4 9.9 

Spent grains 24.8 1.2 10.1 54.5 9.3 

Sweet wort 

(B) 

47.6 5.2 5.9 36.3 4.9 

Grist 20.8 1.0 11.3 57.9 8.9 

Spent grains 25.2 0.8 10.5 57.5 6.0 

Sweet wort 41.4 3.4 5.4 45.3 4.4 

lipid levels in beers are normally so low that their effects on foam are negligible. The 

majority view is that the lipids are important, and that it is the lipids in the last runnings 

from the mash that are responsible for their ability to reduce foam stability. 

Greatinterestattachestotheeffectsoflipidsonflavour.Theconcentrationsofthefree 

fattyacidsinbeerappeartobetoo lowtohavedirecteffects.However,differentlevels of 

free fatty acids can influence the production ofesters by yeast during fermentationand so 

alter the flavour of the beer produced. The greatest interest is in the effects of lipids on 

the flavour deterioration of beer during storage. During mashing some lipid seems to 

disappear because it is oxidized, by oxygen dissolved in the mash, to more polar 

substances, some of which reach the beer and, during storage, give rise to unsaturated 

aldehydes (such as trans-2 nonenal and trans-2, cis-6-nonadienal) which give the beer an 

unpleasant, cardboard like flavour. The chain of reactions is complicated (Fig. 4.32). 

Lipids are hydrolysed by lipases (lipid hydrolases) and esterases to free fatty acids, a 

major proportion of which is linoleic (4.112) and linolenic (4.113) acids, which are 


Glycerides, glycolipids, phospholipids 

H2O Lipases, hydrolases 

Many products, including 

OH 

Linoleic acid (18:2) 

Linolenic acid (18:3) 

(4.112) 

O2 Oxidations 

Lipoxidases, LOX 

(4.113) 

OOH 

OOH 

Aldehydes 

Aldehyde acids 

(4.123) 

O 

OH 

O 

OH 

OOH 

OOH 

Hydroperoxydi- and triene acids (various isomers) 

Chain cleavage 

Hydroperoxide 

lyase 

Isomerase Cyclase Isomerase 

α- and γ-Ketol 

fatty acids 

Cyclic 

acids 

unsaturated. Some of these acids may have been oxidized while still combined in the 

original lipid. Malt acrospires are rich in lipases and lipid degrading enzymes. Lipases are 

active to some extent during mashing. The unsaturated acids are partly oxidized by 

oxygen in the presence of lipoxidase enzymes (LOX, two isoenzymes occur), and perhaps 

peroxidase, giving rise to several unsaturated hydroperoxides (4.123). Some autoxidation 

may also occur, but it seems that enzyme-catalysed oxidation is the most important, and 

this may be reduced by reducing the mash pH from 5.5 to 5.0 (Kobayashi et al., 1993). 

The diene and triene systems of linoleic (4.112) and linolenic (4.113) acids are very 

readily oxidized. 

Under the influence of hydroperoxide isomerases (which are relatively heat stable) and 

other enzymes a complex mixture of substances is formed, including saturated and 

unsaturated aldehydes and aldehyde acids, ketols, cyclic compounds, epoxyhydroxyacids 

and trihydroxyfatty acids, including various trihydroxyoctadecenoic acid isomers (Fig. 

4.32). The aldehydes contribute to a cardboard flavour, but most of these substances are 

O 

OH 

Epoxy-hydroxyene 

fatty acids 

H2O 

Hydrase 

Trihydroxy unsaturated 

fatty acids 

Fig. 4.32 Possible stages in the oxidative breakdown of the major unsaturated fatty acids during 

mashing (after Briggs, 1978; Gardner, 1988). The number of possible products is very large indeed. 

It is thought that the unsaturated trihydroxy-fatty acids are the precursors of staling flavour 

compounds in beers. 


O 

O 

O 

OH 

OH

lost by evaporation during the hop-boil and the rest are reduced to the corresponding 

alcohols by the yeast. These alcohols are not oxidized during beer storage and they are 

not the source of the `staling aldehydes' in beer. It seems that these arise, by some 

unknown mechanism, from the various unsaturated trihydroxyacids which are not 

retained in the mash, are easily water soluble and which survive into the beer (MoÈller- 

Hergt et al., 1999). The trihydroxy fatty acids have foam-collapsing properties. 

Increasingly, ways to minimize oxidation during mashing are being sought, to minimize 

theformationofprecursorsofòff-flavours'.Minimumquantitiesoffattyacid-containing 

lipids are extracted from mashes made at 62 64ëC (143.6 147.2ëF). The amounts 

extracted from mashes made at 68ëC (154.4ëF) are roughly double those extracted from 

the cooler mashes (Forch and Runkel, 1974). 

4.4.9 Phenols 

Cereal grains contain complex mixtures of phenols and polyphenols. Barley grain and 

malts have been most studied, but sorghum phenols have also been studied, especially in 

dark-grained, `birdproof' cultivars. Barley and sorghum seem to be alone among the 

cereals in possessing polymeric flavanols. The phenols in barley vary widely in their 

complexity.Compoundssuch astyrosine(4.48),tyramine(4.68)andhordenine (4.83)are 

present as are arange of phenolic acids. These may be divided into two groups, the 

substituted benzoic acids and the substituted cinnamic acids (Fig. 4.33; Briggs, 1998; 

McMurrough et al., 1984; 1996). Of the benzoic acids vanillic acid (4.127) is the most 

abundant in wort, while of the cinnamic acids ferulic acid (4.131) is the most abundant in 

malt (Table 4.22). The acids occur free and in combination, apparently as esters (as in 

chlorogenic acid (4.133) and glycosides). Ferulic acid occurs free and attached to some 

arabinose residues in pentosans and it is released into wort during mashing, with an 

optimum temperature of about 45 ëC (113 ëF). During boiling or under the influence of 

some bacteria and wild yeasts some ferulic acid is decarboxylated to yield 4-vinyl 

guaiacol (4.134) a substance which in most beers confers an undesirable flavour. 

However, in wheat beers the presence of this substance is desirable. In some barleys, and 

their malts, coloured anthocyanin pigments (delphinidin (4.135), cyanidin (4.136) and 

perhaps pelargonidin (4.137)) occur, either free or as glycosides, but these substances 

seem to have no significance in brewing. 

Table 4.22 The concentrations of the major phenolic acids in an unboiled lager wort (McMurrough 

et al., 1984) 

Benzoic acid derivatives Concentration (mg/litre) 

Gallic acid (4.125) 0.1 

Protocatechuic acid (4.126) 0.5 

4-Hydroxybenzoic acid (4.124) 0.6 

Vanillic acid (4.127) 1.4 

Syringic acid (4.128) 0.6 

Cinnamic acid derivatives 

Caffeic acid (4.130) 0.1 

p-Coumaric acid (4.129) 0.6 

Ferulic acid (4.131) 1.3 

Sinapic acid (4.132) 0.4 

TOTAL 5.6 (mg/litre) 


(4.124) 

(4.125) 

(4.126) 

(4.127) 

(4.128) 

Benzoic acid series 

HO 

R1 

R2 

COOH 

p-Hydroxybenzoic acid 

(R1 = R2 = H) 

Gallic acid 

(R1 = R2 = OH) 

Protocatechuic acid 

(R1 = OH, R2 = H) 

Vanillic acid 

(R1 = OCH3, R2 = H) 

Syringic acid 

(R1 = R2 = OCH3) 

HO COOH 

Cinnamic acid series 

CH == CH·COOH 

Of more significance are the colourless flavan-3-ols ((+)-catechin (4.138), (-)epicatechin 

(4.139), (+)-gallocatechin (4.140) and epigallocatechin (4.141)) and related 

polymeric materials (Fig. 4.34). These four monomeric substances are not proanthocyandins. 

In contrast, polymeric flavan-3-ol materials give rise to anthocyanin pigments 

when heated in acidic butanol in air. At first this was thought to be because they were 

monomeric flavan-3,4-diols, and so they were called leucoanthocyanins. This is not 

correct and at present they are usually called anthocyanogens by brewers and 

proanthocyanidins by chemists. As analytical methods have improved so the great 

complexity of this group of substances has been recognized. In arecent study dimers (7), 

trimers (19), tetramers (23), and pentamers (7) were recognized (Whittle et al., 1999), 

and this total, of 56, may well increase as analytical methods are further refined. An 

example of aprodelphinidin pentamer (4.145) is shown in Fig. 4.35. In addition to the 

HO 

HO O·CO·CH == CH OH 

OH 

(4.133) Chlorogenic acid 

(3-Caffeoylquinic acid) 

(4.129) 

(4.130) 

(4.131) 

(4.132) 

OH 

R1 

R2 

p-Coumaric acid 

(R1 = R2 = H) 

Caffeic acid 

(R1 = OH, R2 = H) 

Ferulic acid 

(R1 = OCH3, R2 = H) 

Sinapic acid 

(R1 = R2 = OCH3) 

OH 

OCH3 

CH == CH2 

(4.134) Vinyl guaiacol 

Fig.4.33 Substitutedbenzoicandcinnamicacidsandsomerelatedcompoundsthatoccurinsweet 

wort. 


(4.138) 

(4.140) 

(4.142) 

(4.143) 

(4.144) 

HO 

HO 

HO 

OH 

OH (+) Catechin 

OH 

OH 

O 

OH 

OH 

OH 

O 

Cl 

O 

Barley procyanidin B3 

(Catechin – catechin) 

Prodelphinidin B3 

OH 

(+) Gallocatechin 

OH 

Anthocyanin pigments 

R1 

OH 

OH 

OH 

R2 

(4.139) 

(4.141) 

(R1 = R2 = OH) Delphinidin chloride (4.135) 

(R1 = H, R2 = OH) Cyanidin chloride (4.136) 

(R1 = R2 = H) Pelargonidin chloride (4.137) 

HO 

Sorghum 

Procyanidin B1 

(epicatechin – catechin) 

OH 

OH 

OH 

(gallocatechin – catechin) OH OH 

OH 

O OH 

HO 

OH 

HO 

HO 

HO 

OH 

HO 

OH 

HO 

OH 

OH 

O 

OH 

O 

OH 

O 

OH 

O 

OH 

O 

OH 

OH 

OH 

OH 

OH 

OH 

OH 

(–) Epicatechin 

OH 

OH 

O 

OH 

OH 

Epigallocatechin 

OH 

O 

Fig. 4.34 Flavanols, including some that occur in sweet worts. 


OH 

OH 

OH 

OH 

OH 

OH 

OH

HO 

OH 

HO 

O 

OH 

HO 

OH 

O 

OH 

HO 

OH 

OH 

OH 

OH 

O 

OH 

HO 

OH 

number and identity of the constituent monomers the structures are complicated by 

differences in stereochemistry, as illustrated by the dimers procyanidin B 1 (4.142), 

procyanidin B3 (4.143) and prodelphinidin B3 (4.144). Concentrations of some flavans in 

barley malts are catechin (4.138) 25 75 mg/kg; prodelphinidin B3 (4.144), 186 362 mg/ 

kg; procyanidin B3 (4.143), 130 276 mg/kg and 4 trimers, 336 671 mg/kg. (McMurrough 

and Delcour, 1994). 

Malt polyphenols are partly dissolved and partly destroyed during mashing with pale 

malts chiefly, it is thought, by oxidative reactions mostly catalysed by peroxidase, 

perhaps with contributions from catalase and polyphenol oxidase. The nature of the 

phenolic material in wort is influenced strongly by how strongly the malt has been kilned, 

the availability of oxygen during mashing and the mashing programme. Mashing or 

sparging at elevated temperatures extracts more polyphenol into wort. Polyphenols are 

largely destroyed in experimental mashes made with green malt, and mash aeration or 

additions of hydrogen peroxide are most effective at removing anthocyanogens if the 

OH 

O 

OH 

OH 

OH 

A-prodelphinidin pentamer 

(4.145) 

Fig. 4.35 The formula (ignoring the stereochemistry) of a pentameric proanthocyanidin 

(anthocyanogen) from wort (Whittle et al., 1999). The composition may be represented as 

g-g-c-c-c, where g and c are gallocatechin (4.140) and catechin (4.138) units respectively. 


OH 

OH 

OH 

O 

OH 

OH 

OH 

OH 

OH

malt hasbeen only lightly kilned. Additions of small amounts offormaldehyde tomashes 

reduces proanthocyanidin levelsgreatly,probablybylinkingthemtoproteinsinthemash 

(Macey, 1970; Macey et al., 1966). Similar results are obtained by mashing with malts 

prepared by steeping or re-steeping barley or washing sorghum in dilute solutions of 

formaldehyde (Briggs, 1998). All these treatments were tested or used to reduce 

proanthocyanidin levels in beers, giving them greater resistance to haze formation, but 

their use has been generally discontinued. During wort run-off the mixture of phenolics 

altersincompositionandthiscanbeofsignificance(WoofandPierce,1966).IfthepHof 

the sparge is allowed to increase more phenols are extracted into the wort, which is 

undesirable. 

When phenols are oxidized they polymerize and give rise to red-brown coloured 

substances, called phlobaphenes. Some phenols have tanning properties (that is they bind 

to proteins and may precipitate them), and all of them are potentially involved in beer 

flavour. However, reports on their significance conflict. Phenolics are credited with 

altering the astringency, the mouth-feel and the after-taste of beers. However, the 

proanthocyanidins are the major `tannins' in malt and it is believed that tannins are 

responsible for astringency, yet beers made with malts from mutant barleys and hop 

extracts, both lacking proanthocyanidins, are said to have the same flavour as beers made 

from `conventional' barleys. Also the removal of proanthocyanidins from beer by 

adsorption on to PVPP (insoluble polyvinyl polypyrrolidone) is reported to have little 

effect on flavour. Some phenols have antioxidant effects, that is, they block the oxidative 

reactions of other substances, at least sometimes by destroying free radical intermediates. 

The significance of malt phenol's antioxidant properties in brewing is unclear. Many 

oftheo-diphenolsshouldreadilyoxidizetoo-quinones(Briggs,1998),whichwillreadily 

reactwithmanyothersubstances.Theinteractionsduringbrewingbetweenphenolicsand 

proteinsarecertainlyimportant.Maltsmadefrommanybirdproofsorghumsaresorichin 

phenolic tannins that during mashing many of the enzymes are inhibited and conversion 

is insufficient. In mashes, and indeed in wort, there is apartition of proanthocyanidins 

between being free in solution, being bound to soluble proteins and being bound to 

insoluble proteins. The associations may be reversible, as in chill hazes, or may be 

irreversible, as in permanent hazes. More phenolics are extracted from well-modified 

malts during mashing because, it is supposed, less insoluble protein remains in the grist 

andsoasmallerproportionofphenolicsarebound.Hopsalsocontributepolyphenolicsto 

wort. 

4.4.10 Miscellaneous acids 

Worts contain awide range of aliphatic acids, in addition to fatty acids (Fig. 4.30), mostly 

in small amounts. These include several saturated and unsaturated, low molecular weight 

fatty acids (C6 C10), mesaconic/laevulinic acid, pyruvic acid (4.146) -ketoglutaric acid 

(4.147), fumaric acid (4.148), succinic acid (4.149), lactic acid (4.150), oxalic acid (4.151), 

malic acid(4.152), citric acid (4.153), kojic acid (4.154)and gluconicacid (4.155).Most of 

these are well-known intermediary metabolites (Fig. 4.36; MacWilliam, 1968; Moll, 1991). 

The significance of these substances is usually unclear, although they must contribute to the 

pH buffering capacity of the wort and most can probably be metabolized by yeast. 

Two of these materials are of particular interest. Lactic acid ((4.150) a mixture of the 

D- and L-isomers) arises from the malt and the microbes on its surface. Under some 

conditions excessive production results in malts that give too acid mashes. Sometimes 

lactic acid is deliberately added to mashes (as acidified wort, or as acid malt) to reduce 


CH 3·CO·COOH 

Pyruvic acid 

(4.146) 

HOOC·(CH 2) 2·CO·COOH 

α-Ketoglutaric acid 

(4.147) 

HOOC H 

C 

C 

H COOH 

Fumaric acid 

(4.148) 

CH 2·COOH 

CH 2·COOH 

Succinic acid 

(4.149) 

CH 3·CH(OH)·COOH 

Lactic acid 

(4.150) 

COOH 

COOH 

Oxalic acid 

(4.151) 

HO 

HO·CH·COOH 

CH 2·COOH 

Malic acid 

(4.152) 

CH 2·COOH 

HO·C·COOH 

CH 2·COOH 

Citric acid 

(4.153) 

O 

O 

Kojic acid 

(4.154) 

COOH 

H·C·OH 

HO·C·H 

H·C·OH 

H·C·OH 

CH 2OH 

(4.155) 

CH 2OH 

D-Gluconic acid 

Fig. 4.36 Formulae of some of the organic acids present in worts. 


Table 4.23 The concentrations of some inorganic ions in various sweet worts (A, MaÈndl, 1974; B, 

Rudin, 1974; C, Moll, 1991).The all-malt worts in column A were adjusted to a concentration of 

12%. The worts in column B were made with and without adjuncts 

Ionic species Concentrations (mg/litre) 

A B C 

Potassium, K + 

522 310±770 300±700 

Sodium, Na + 

Calcium, Ca 

25 11±112 10±100 

2+ 

Magnesium, Mg 

39 40±62 40±100 

2+ 

130 ± 100±150 

Copper, Cu 2+ 

Iron, Fe 

0.08 0.12±0.13 0.1±0.2 

3+ (ferric) 

Manganese, Mn 

0.08 0.23±0.37 0.1±0.3 

2+ (manganous) 0.13 ± 0.1±0.2 

Zinc, Zn 2+ 

0.12 0.07±0.16 0.1±0.3 

the mash pH in acontrolled fashion. Worts also contain oxalic acid (4.151). This, the 

simplest dicarboxylic acid, has astrong affinity for calcium ions and readily forms 

crystalline precipitates of calcium oxalate. If formed late in the brewing process this can 

give rise to oxalate haze and the crystals can form nuclei for the release of carbon 

dioxide, so potentiating gushing (over-foaming) in beer. During fermentation calcium 

oxalate can be deposited on the walls of the fermentation vessel as beer stone. Malted 

barleyscancontain5.6 22.8mgoxalicacid/100gdrymatterwhileformaltedwheatsthe 

values may be 22.1 50.3mg/100g(Narziss et al., 1986). Presumably much of this is 

precipitated in the mash when the liquor contains asufficient concentration of calcium 

ions. Since oxalate is potentially toxic it is desirable to reduce its concentration to alow 

level. 

4.4.11 Inorganic ions in sweet wort 

Some 1.5 2.0% of wort solids are materials which, after evaporation and combustion, 

remain in the ash. Of these the inorganic ions originate in the brewing liquor (Chapter 3) 

andinthe grist materials,both the malt (around 2 3%ash) andthe adjuncts(1 3%ash). 

Small amounts of ions (copper, iron, nickel, tin, zinc, etc.) may be picked up from the 

brewing plant. Later contributions may come from the copper adjuncts and hops. The 

proportions of the most important ions vary (Table 4.23). Sulphur may be present as 

sulphate (200 400mg/litre), in amounts that vary widely with the nature of the brewing 

liquor. Sulphuralsooccursinorganic combinationsintheamino acidsmethionine(4.41), 

cysteine (4.31) and cystine (4.32), the tri-peptide glutathione ( -glutamyl-cysteinylglycine), 

coenzyme A, mercaptans, as well as hydrogen sulphide and sulphite ions. 

Chloride ions (40±500mg/litre) and phosphate ions (500±900mg/litre) also occur 

(Briggs, 1998; Lee, 1990; MacWilliam, 1968; Moll, 1991). 

Little phosphate comes from the brewing liquor, except where phosphoric acid or acid 

phosphate salts have been used for pH adjustment. Most comes from the grist and may 

originate from nucleic acids, nucleotides, phospholipids (Fig. 4.31), and especially myoinositol 

hexaphosphate (phytic acid, 4.156), of which there may be 0.6±0.9% in the dry 

malt (Lee, 1990). Phytic acid is a strong chelating agent, and binds copper, iron and zinc 

ions as well as calcium ions. This material undergoes some hydrolysis during mashing, 

inorganic phosphate being removed sequentially from successively lower phosphate 

esters until free myo-inositol (4.91) is released. The extent of phytate hydrolysis is 

dependent on the level of the enzyme (or enzymes) with phytase activity remaining in the 


malt after kilning and the temperature programme used in mashing. Lightly kilned malts 

may contain about aquarter of the enzyme originally present, while ale malts may 

contain little active enzyme. The pH optimum of the enzyme is about 5, and the optimum 

temperatureis45 50ëC(113 122ëF).Theinteractionsbetweenphosphatesandcalcium 

ions contribute to the desirable fall in pH which occurs during mashing (Chapter 3). 

Usually about two-thirds of the phosphate in sweet wort is inorganic phosphate. 

It seems that usually wort provides all the necessary inorganic ions that yeast requires 

for sound fermentations, with the occasional exception of zinc (Donhauser, 1986; Jacobsen, 

1986; Lie et al., 1980). Because not all the ash components are extracted during mashing 

the ionic composition of the sweet wort is not the simple sum of the ions initially present in 

the grist and the mashing liquor (Table 4.24). Less than 5% of the zinc (or copper or iron) 

present in the grist is dissolved during mashing, and the proportions dissolved can be very 

variable. Of the zinc present in the wort only a proportion is available to yeast, presumably 

because the remainder is chelated or otherwise bound to other substances. Consequently 

simple analyses of wort zinc contents are unreliable for predicting zinc deficiency. 

Probably a 12% wort should contain at least 0.08 and preferably 0.1 0.2 mg Zn/litre to 

ensure a good fermentation. Where permitted, traces of a soluble zinc salt may be added to 

the wort. Where this is not permitted the use of well-modified malts and carefully acidified 

mashes reduce problems of zinc deficiency, as does the use of mashing equipment with 

metal components from which traces of zinc can dissolve. 

4.5 Mashing and beer flavour 

H 

O 

H2O3P PO 3H 2 PO 3H 2 

O O PO 2H 2 

H H 

OPO2H3 H 

H 

Phytic acid, myo-inositol hexaphosphate 

(4.156) 

Malts and other components of the grist influence beer flavour. However, the 

relationships are extremely complicated. During mashing flavoured substances are 

extracted into the wort. Some will be destroyed or partly or wholly lost during the hopboil, 

while other `flavour precursors' will be converted into flavoursome substances, and 

others will reach the fermenter unchanged. The yeast may then metabolize many of these 

O 

H 

O 

PO3H2 Table 4.24 The ionic compositions (mg/litre) of a brewing water, a sweet wort made with it and 

the beer (Rudin, 1974) 

Liquid Ca 2+ 

Mg 2+ 

Na + 

K + 

Cl SO 2 4 PO 3 4 HCO 2 3 

Brewing liquor 169 36 55 6 247 205 ± 165 

Wort (12 ëPlato) 165 127 101 550 450 338 846 ± 

Beer (12 ëPlato, 

original gravity) 168 113 110 440 420 330 520 ± 


substances altering their flavours. Some malt flavour substances, such as vanillin and - 

phenyl ethanol, partly occur combined as -glycosides. During mashing they can be 

released by hydrolysis, catalysed by -glucosidase. In addition, oxidative changes 

occurring to the lipids during mashing can give rise to precursors of staling flavours. The 

`redox state' of the beer has an influence on the rate of flavour deterioration and haze 

formation (Bamforth et al., 1993, 1997; Briggs, 1998; Moir, 1989; Van Den Berg and 

Van Eerde, 1982). It can be advantageous to exclude oxygen from the mash and 

subsequently, during the production of hopped wort. 

Higher levels of ànti-oxidant' substances in the beer retard deterioration processes. 

These agents, which originate in the grist, seem to work in different ways. They compete 

for oxygen by being oxidized themselves, or they inhibit enzymes catalysing oxidations, 

and/or they `scavenge' free radicals. Anti-oxidants include sulphite and bisulphite ions, 

polyphenols and reductones, which are ene-diol substances resembling ascorbic acid 

(4.96), which are formed during Maillard reactions. These compounds are found in dark 

malts, which have long been known to have flavour-stabilizing properties. Many 

hundreds of potentially active flavour substances are derived from malts or adjuncts and 

include aldehydes, ketones, amines, thiols and other sulphur-containing substances, 

heterocyclic oxygen-, nitrogen- and sulphur-containing substances and phenols. Sparging 

may extract unwanted flavours. Moderating sparge temperatures, for example up to 75ëC 

(167ëF), and keeping the pH low, e.g., to 5.5, improves beer flavour and stability. 

Asubstance that has attracted particular interest is dimethyl sulphide (DMS, (4.158) 

Fig.4.37;Dickenson,1983).InsomeEuropeanlagersappreciablelevelsofthissubstance 

are desirable while in some other beers its absence is preferred. The precursor of this 

highly volatile material is S-methyl methionine (SMM, 4.157), asulphonium compound 

formed by the metabolic methylation of methionine (4.41) in the malt. This substance is 

heatlabileandsowillonlysurviveinmaltifthisislightlykilned.Someisdecomposedto 

DMS and homoserine (4.159) and the DMS produced is mostly lost with the kilning air. 

Some is oxidized to the less volatile dimethyl sulphoxide (DMSO, 4.160). More SMM is 

decomposed during the hop-boil. Surviving DMS, SMM and small amounts of DMSO 

reach the fermenter. Yeast may reduce the DMSO to DMS. Thus the level of DMS 

present in abeer depends on the malt used and the details of the production process. 

CH 3 

CH 3 

S·CH 2·CH 2·CH·COOH 

NH 2 

+ OH 

CH 3 

CH 3 

S-Methyl methionine (SMM) (4.158) Dimethyl 

(4.157) 

sulphide (DMS) 

Oxidation Reduction 

CH 3 

CH 3 

S + HO·CH 2·CH 2·CH·COOH 

S==O 

Dimethyl sulphoxide (DMSO) 

(4.160) 

NH 2 

Homoserine 

(4.159) 

Fig. 4.37 The formation of dimethylsulphide ((4.122), DMS) from S-methyl methionine ((4.121), 

SMM, the DMS precursor, DMS-P) and the interconversions of DMS and dimethylsulphoxide 

((4.123), DMSO). 



At the end of every mash abrewer has aload of spent grains or draff, which must be 

disposed of. Of the original grist some 17 22% of the original dry matter remains, 

about 18 20kg fresh weight/hl beer produced. This material is wet and, depending on 

the mashing system employed, contains up to 80% water. Liquid from this wet 

material is turbid and represents an effluent that, because of its high BOD, is costly to 

dispose of if directed to drain (Chapter 3). Obtaining representative samples of the 

spent grains is difficult, because of inhomogeneities in the filter bed, but they should 

be investigated to evaluate the mashing process. It is desirable to determine their 

content of residual extract (which is inversely related to extract recovery in the wort), 

and residual starch which, ideally, would have been converted during mashing. In 

addition inspection of the solids shows how well milling has been carried out and to 

what extent husk material has remained intact. The wet grains may be squeezed to 

remove some liquor (minimizing drainage and handling problems) and the squeezed 

draff may be dried with hot air. Drying is unusual as it is costly (Chapter 3). Often the 

squeeze liquor is returned, with or without treatment, to the process stream to recover 

the extract. 

The composition of the spent grains depends on the grist used, the mashing 

programme and the recovery of the extract (Table 4.25). Relative to the malt, spent grains 

are enriched in protein, fibre, ash and lipid. Their major use is as a valuable feed for 

ruminants. They may be fed directly or after ensilage. As they deteriorate rapidly draff 

should be removed from the brewery as quickly as possible, to prevent their becoming a 

source of undesirable organisms. It has been suggested that they should be treated with a 

preservative, such as propionic acid, to minimize decomposition. Other actual or 

proposed uses for spent grains are inclusion in mushroom compost, and use as a substrate 

for the cultivation micro-organisms, for example filamentous fungi for feeding pigs. The 

draff might be fed to the pigs directly, after the mechanical removal of the fibre. After 

composting the spent grains might be used for turf management or as a horticultural soil 

conditioner. It has also been proposed that the grains be digested with enzyme 

preparations to produce extra `wort' to use in the brewery or as a culture medium for 

other microbes. 

Table 4.25 The composition and nutritive value (for sheep) ranges of brewers' wet spent grains 

(Briggs et al., 1981). The values in parentheses are the digestible amounts. Metabolizable energy ˆ 

digestible energy 0.81 

Mean Range 

Dry matter (%) 26.3 24.4±30.0 

Crude protein (%) 23.4 (18.5) 18.4±26.2 (13.9±21.3) 

Crude fibre (%) 17.6 (7.9) 15.5±20.4 (6.6±10.2) 

Ether extract (%) 7.7 (7.7) 6.1±9.9 (5.6±9.2) 

Total ash (%) 4.1 3.6±4.5 

Digestible energy (mJ/kg dry wt.) 13.8 13.0±14.8 

Gross energy (mJ/kg dry wt.) 21.4 21.1±21.8 

DOMD* in vivo (%) 59.4 55.2±64.3 

DOMD* in vitro (%) 48.6 44.8±51.5 

* DOMD ˆ Digestibility of organic matter (dry). 



ALBINI, P. A., BRIGGS, D. E. and WADESON, A. (1987) J. Inst. Brewing, 93, 97. 

ALLEN, W. G. (1987) Brewer's Digest, 62 (3), 18. 

ANDERSON, I. W. (1993) Proc. 24th Congr. Eur. Brew. Conv., Oslo, 193. 

ANNESS, B. J. (1984) J. Inst. Brewing, 90, 315. 

ANNESS, B. J. and REED, R. J. R. (1985) J. Inst. Brewing, 91, 82, 313. 

BAJOMO, M. F. and YOUNG, T. W. (1993) J. Inst. Brewing, 99, 153. 

BAMFORTH, C. W., MULLER, R. E. and WALKER, M. D. (1993) J. Amer. Soc. Brew. Chem., 51, 79. 

BAMFORTH, C. W., MOORE, J., MCKILLOP, D., WILLIAMSON, G. and KROON, P. A. (1997) Proc. 26th Congr. 

Eur. Brew. Conv., Maastricht, 75. 

BARRETT, J., CLAPPERTON, J. F., DIVERS, D. M. and RENNIE, H. (1973) J. Inst. Brewing, 79, 407. 

BARRETT, J., BATHGATE, G. N. and CLAPPERTON, J. F. (1975) J. Inst. Brewing, 81, 31. 

BATHGATE, G. N. (1974) Eur. Brew. Conv. Monograph-I. E. B. C.-Wortsymposium, Zeist, 198. 

BATHGATE, G. N. and DALGLEISH, C. E. (1975) Proc. Amer. Soc. Brew. Chem., 33, 32. 

BATHGATE, G. N. and PALMER, G. H. (1972) Die StaÈrke, 24 (10), 336. 

BOURNE, D. T. (1998) Proc. 5th Aviemore Conf. Malting, Brewing and Distilling. (Campbell, I. ed.). 

London, The Inst. of Brewing, p. 237. 



BOURNE, D. T., POWLESLAND, T. and WHEELER, R. E. (1982) J. Inst. Brewing, 88, 371. 

BRENNER, M. W. (1972) M. M. B. A. Tech. Quart., 9, 12. 

BRIGGS, D. E. (1978) Barley. London. Chapman and Hall, 612 pp. 

BRIGGS, D. E. (1992) in Barley: genetics, biochemistry, molecular biology, and biotechnology. (Shewry, P. 

R. ed.) p. 361. Wallingford. C.A.B. International. 

BRIGGS, D. E. (1998) Malts and Malting. London. Blackie Academic, 796 pp. 

BRIGGS, D. E., HOUGH, J. S., STEVENS, R. and YOUNG, T. W. (1981) Malting and Brewing Science (2nd edn), 

Vol. 1. Malt & Sweet Wort. London, Chapman and Hall, 387 pp. 

BRYCE, J. H., KIM, N. J., ZAMMIT, R., MCCAFFERTY, C. A., PINHEIRO, M. G. M. and KENDALL, G. L. (1995) 

Proc. 25th Congr. Eur. Brew. Conv., Brussels, p. 285. 

BRYCE, J. H., COOPER, D. and STEWART, G. G. (1997) Proc. 26th Congr. Eur. Brew. Conv., Maastricht, 

p. 357. 

BUÈ HLER, T. N., MCKECHNIE, M. F. and WAKEMAN, R. J. (1996) Food and Bioproducts Processing, 74 (no. 

C4), 207. 

BURGER, W. C. and SCHROEDER, R. L. (1976) J. Amer. Soc. Brew. Chem., 34, 133; 138. 

BUTTIMER, E. T. and BRIGGS, D. E. (2000) J. Inst. Brewing, 106, 71, 83. 

BUTTON, A. H. and PALMER, J. R. (1974) J. Inst. Brewing, 80, 206. 

COMRIE, A. A. D. (1967) J. Inst. Brewing 74, 335. 

COULTATE, T. P. (2002) Food: the chemistry of its components. Cambridge. The Royal Society of 

Chemistry, 432 pp. 

CRABB, D. and BATHGATE, G. N. (1973) J. Inst. Brewing, 79, 519. 

DEBYSER, W., DERDELINCKX, G. and DELCOUR, J. A. (1997a) J. Cereal Sci., 26, 67. 

DEBYSER, W., DERDELINCKX, G. and DELCOUR, J. A. (1997b) J. Amer. Soc. Brew. Chem., 55, 153. 

DE CLERCK (1957) A Textbook of Brewing, Vol. 1. (Transl. Barton-wright, K.) London, Chapman and 

Hall, p. 587. 

DEMUYAKAR, B., OHTA, Y., NAKATANI, K., FUKUI, N. and KANAGAWA, K. (1994) J. Amer. Soc. Brew. 

Chem., 52, 111. 

DICKENSON, C. J. (1983) J. Inst. Brewing, 89, 41. 

DIXON, M. and WEBB, E. (1958) Enzymes. London. Longmans, Green & Co., 784 pp. 

DONHAUSER, S. (1986) Eur. Brew. Conv. Monograph-XI. E. B. C. ± Symposium on Wort Production, 

Maffliers, p. 176. 

ELIASSON, A.-C. and TATHAM, A. (2001) in Cereals and Cereal Products ± Chemistry and Technology. 

(Dendy, D. A. V. and Dobraszczyk, B. J. eds) Gaithersburg. Aspen Publishers, p. 68. 

ENARI, T. M. (1974) Eur. Brew. Conv. Monograph-I. E. B. C.-Wortsymposium, Zeist, p. 73. 

ENARI, T. M. (1986) Cerevisia, 11 (1), 19. 

ENARI, T. M., MIKOLA, J. and LINKO, M. (1964) J. Inst. Brewing, 70, 405. 

ENEVOLDSEN, B. S. (1974) Eur. Brew. Conv. Monograph-I. E.B.C.-Wortsymposium, Zeist, p. 158. 

ENEVOLDSEN, B. S. (1975) Proc. 15th Congr. Eur. Brew. Conv., Nice, p. 683. 

ENGAN, S. (1981) Brewing Science, Vol. 2 (Pollock, J. R. A. ed.). London, Academic Press, p. 93. 

EVANS, E., MA, Y., EGLINTON, J., LANGRIDGE, P., LOGUE, S. and BARR, A. (2002) Proc. Congr. Inst. Brew. 

(Asia Pacific Sect.)., Adelaide, CD No. 12. 

FINCHER, G. B. (1992) in Barley: Genetics, Biochemistry, Molecular Biology and Biotechnology (Shewry, 

P. R., ed.). Wallingford. C. A. B.-International, p. 413. 

FORAGE, A. J. and LETTERS, R. (1986) Proc. 19th Conv. Inst. Brewing (Australia and New Zealand Sect.), 

Hobart, p. 79. 


FORCH, M. and RUNKEL, U.-D. (1974) Eur. Brew. Conv. Monograph-I. E. B. C.-Wortsymposium, Zeist, 

p. 269. 

FORREST, I. S., DICKSON, J. E. and SEATON, J. C. (1985) Proc. 20th Congr. Eur. Brew.Conv., Helsinki, 

p. 363. 

GARDNER, H. W. (1988) in Advanc. Cereal Sci. and Technol. IX (Pomeranz, Y. ed.) St. Paul. A. A. C. C. 

p. 161. 

GEIGER, K. (1972) M. B. A. A. Tech. Quart. 9, 195. 

GJERTSEN, P. and HARTLEV, P. (1980) Eur. Brew. Conv. Monograph-VI. E. B. C.-Symposium on the 

relationship between malt and beer, Helsinki, p. 186. 

HALFORD, M. H. and BLAKE, G. G. (1972) Proc. 12th Conf. Inst. Brewing (Australia and New Zealand 

Sect.), Perth, p. 85. 

HALL, R. D., HARRIS, G. and MACWILLIAM, I. C. (1956) J. Inst. Brewing, 62, 232. 

HALLGREN, L. (1995) in Sorghum and Millets; Chemistry and Technology. (Dendy, D. A. V. ed.). St. Paul. 

A. A. C. C., p. 283. 

HAN, J. Y. and SCHWARZ, P. B. (1996) J. Amer. Soc. Brew. Chem., 54, 216. 

HARRIS, G. and MACWILLIAM, I. C. (1961) J. Inst. Brewing, 67, 144, 154. 

HIND, H. L. (1950) Brewing Science and Practice Vol II. Brewing Processes. London. Chapman and Hall, 

pp. 507±1020. 

HOME, S., PIETILAÈ , K. and SJOÈ HOLM, K. (1993) J. Amer. Soc. Brew. Chem., 51, 108. 

HOME, S., LAUREN, A. and AUTIO, K. (1994) Proc. 23rd Conf. Inst. Brewing, (Asia Pacific Sect.), Sydney, 

p. 52. 

HOPKINS, R. H. and KRAUSE, B. (1947) Biochemistry Applied to Malting and Brewing. London. Allen & 

Unwin, 342 pp. 

HUDSON, J. R. (1975) Eur. Brew, Conv. Monograph-II. E. B. C. Barley and Malting Symposium, Zeist, 

p. 271. 

HUG, H. and PFENNINGER, H. B. (1979) Proc. 17th Congr. Eur. Brew. Conv., Berlin (West), p. 355. 

HUITE, N. J. and WESTERMANN, D. H. (1974) Proc. 14th Conv. Inst. Brewing (Australia and New Zealand 

Sect.), Brisbane, p. 133. 

HUMBERSTONE, F. J. and BRIGGS, D. E. (1998) Proc. 5th Aviemore Conf. Malting, Brewing and Distilling 

(Campbell, I. ed.) London. The Institute of Brewing, p. 246. 

IMBERTY, A., BULEON, A. and VINH TRAN (1991) StaÈrke, 43 (10), 375. 

ISHERWOOD, N. D., KIRBY, W., WHEELER, R. E. and JONES, M. (1977) Proc. 16th Congr. Eur. Brew. Conv., 

Amsterdam, p. 457. 

JACOBSEN, T. (1986) Eur. Brew. Conv. Monograph-XI. E. B. C.-Symposium on Wort Production, 

Maffliers. p. 196. 

JONES, M. (1974) Eur. Brew. Conv. Monograph-I. E. B. C.-Wortsymposium, Zeist, p. 94. 

JONES, M. and PIERCE, J. (1967) J. Inst. Brewing, 73, 342. 

KIENINGER, H. (1969) Proc. 12th Congr. Eur. Brew. Conv., Interlaken, p. 139. 

KIENINGER, H. (1972) Brauwelt, 112, 1535; 1793. 

KOBAYASHI, N., KANEDA, H., KANO, Y. and KOSHINO, S. (1993) Proc. 24th Congr. Eur. Brew. Conv., Oslo, 

p. 405. 

KROTTENTHALER, M., ZUÈ RCHER, J., SCHNEIDER, J., BACK, W. and WEISSER, H. (1999) Proc. 27th Congr. 


KUNZE, W. (1996) Technology Brewing and Malting. (Wainwright, T. transl.). Berlin. VLB, pp. 726. 

LAING, H. and TAYLOR, D. G. (1984) Proc. 18th Conv. Inst. Brewing (Australia and New Zealand Sect.), 

Adelaide, p. 109. 

LEE, W. J. (1990) J. Amer. Soc. Brew. Chem., 48, 62. 

LETTERS, R. (1992) Ferment, 5, 268. 

LETTERS, R. (1994) Proc. 4th Aviemore Conf. Malting, Brewing and Distilling. (Campbell, I. and Priest, 

F. G. eds). London, the Inst. of Brewing, p. 128. 

LETTERS, R. (1995a) Proc. 5th Sci. Tech. Conv. Inst. Brewing (Central and Southern African Sect.), 

Victoria Falls, p. 115. 

LETTERS, R. (1995b) Ferment, 8, 301. 

LETTERS, R., HURLEY, J. C. and HORAN, H. (1986) Eur. Brew. Conv. Monograph-XI. E. B. C.Symposium on 

Wort Production, Maffliers, p. 250. 

LEWIS, M. J. and OH, S. S. (1985) M.B.A.A. Tech. Quart., 22, 108. 

LIE, S., HAGE, T. and JACOBSEN, T. (1980) Eur. Brew. Conv. Monograph-VI. E. B. C. Symposium on the 


LISBJERG, S. and NIELSEN, H. (1991) Proc. 3rd Sci. Tech. Conv. Inst. Brewing (Central and Southern 

African Section), Victoria Falls, p. 157. 

LITTLE, B. T. (1994) Ferment, 7, 163. 

LUCHSINGER, W. W., ENGLISH, H. and KNEEN, E. (1958) Proc. Ann. Mtg. Amer. Soc. Brew. Chem., p. 40. 

LUÈ ERS, H., KRAUSS, G., HARTMANN, O. and VOGT, H. (1934) Wochensch. f. Brau. 51 (46), 361. 

MACEY, A. (1970) Proc. 11th Conv. Inst. Brewing (Australia and New Zealand sect.), Hobart, p. 117. 

MACEY, A., STOWELL, K. C. and WHITE, H. B. (1966) J. Inst. Brewing, 72, 29. 


MACFADDEN, D. P. (1989) Proc 2nd Sci. Tech. Conv. Inst. Brewing (Central and Southern African Sect.), 

Johannesburg, p. 306. 

MACGREGOR, A. W. (1990) Proc. 3rd Aviemore Conf. On Malting, Brewing and Distilling. (Campbell, I. 

ed.). London. Inst. Brewing, p. 10. 

MACGREGOR, A. W. and YIN, X. S. (1990) J. Amer. Soc. Brew. Chem., 48, 82. 

MACWILLIAM, I. C. (1968) J. Inst. Brewing, 74, 38. 



MACWILLIAM, I. C., HUDSON, J. R. and WHITEAR, A. L. (1963) J. Inst. Brewing, 69, 303. 

MAÈ NDL, A. (1974) Brauwiss. 27, 177. 

MARTIN, P. A. (1978) Brewer's Guard., Aug., p. 29. 

MARTIN, P. A. (1979) J. Inst. Brewing, 85, 290. 

MAULE, D. R. (1986) Eur. Brew. Conv. Monograph-XI. E. B. C.-Symposium on Wort Production, 

Maffliers, p. 72. 

MAULE, D. R. and CRABB, D. (1980) Eur. Brew. Conv. Monograph-VI. E. B. C.-Symposium on the 


MCMURROUGH, I. and DELCOUR, J. A. (1994) Ferment, 7, 175. 

MCMURROUGH, I., ROCHE, G. P. and CLEARY, K. G. (1984) J. Inst. Brewing, 90, 181. 

MCMURROUGH, I., MADIGAN, D., DONNELLY, D., HURLEY, J., DOYLE, H., HENNIGAN, G. and MCNULTY, N. 

(1996) J. Inst. Brewing, 102, 327. 

MEDDINGS, P. J. and POTTER, O. E. (1971) J. Inst. Brewing, 77, 246. 

MIKOLA, J., PIETILAÈ , K. and ENARI, T.-M. (1972) J. Inst. Brewing, 78, 384. 

MOIR, M. (1989) Brewer's Guard., Sept., p. 64. 

MOLL, M. (1991) Beers and Coolers. (Wainwright, T. transl.) Andover. Intercept, p. 495. 

MOLL, M., FLAYEUX, R., VINH THAT and MARTIN, J. (1974) Eur. Brew. Conv. Monograph-I. 

Wortsymposium, Zeist, p. 41. 

MOLL, M., FLAYEUX, R., MATHIEU, D. and PHAN TAN LUU, R. (1982) J. Inst. Brewing, 88, 139. 

MOÈ LLER-HERGT, G., WACKERBAUER, K., TRESSL, R., GARBE, L.-A. and ZUFULL, C. (1999) Proc. 27th Congr. 


MORRAYE, C. (1938) Pet. J. Brasseur., 46, 464. 

MORRISON, W. R. (1978) in Advanc. Cereal Sci. and Technol. II (Pomeranz, Y. ed.) p. 221. 

MORRISON, W. R. (1988) J. Cereal Sci., 8, 1. 

MULLER, R. F. (1989) Proc. 22nd Congr. Eur. Brew. Conv., Zurich, p. 283. 

MULLER, R. (1991) J. Inst. Brewing, 97, 85, 93. 

MULLER, R. (1995) J. Amer. Soc. Brew. Chem., 53, 53. 

MUTS, G. C. J. (1991) Proc. 3rd Sci. Tech. Conv. Inst. Brewing, (Central & Southern African Section), 


MUTS, G. C. J. and PESMAN, L. (1986) Eur. Brew. Conv. Monograph-XI. E. B. C.-Symposium on Wort 

Production, Maffliers, p. 25. 

MUTS, G. C. J., KAKEBEEKE, M. G., PESMAN, L. C. and VAN DEN BERG, R. (1984) Proc. 18th Conv. Inst. 

Brewing (Australia and New Zealand Section), Adelaide, p. 115. 

NARZISS, L. (1977) Brauwelt, 117 (37), 1420. 

NARZISS, L. (1978) Proc. 15th Congr. Inst. Brewing (Australia and New Zealand Sect.), Christchurch, 

p. 35. 

NARZISS, L. (1991) Brauwelt Internat., 4, 284. 

NARZISS, L. (1992a) Die Bierbrauerei, 2 Band. Die Technologie der WuÈrzebereitung (7th edn). Stuttgart. 

Ferdinand Enke, p. 402. 

NARZISS, L. (1992b) Brauwelt, (23), 1072. 

NARZISS, L., REICHENEDER, E. and IWAN, H.-J. (1986) Monatss. f. Brauwiss., 39 (1), 4. 

NARZISS, L., MIEDANER, H. and NITZCHE, F. (1990) Monatss. f. Brauwiss., 3, 96. 

OLIVER-DAUMEN, B. (1988) Brauwelt Internat., (III), 256, 370. 

PIERCE, J. S. (1980) Eur. Brew. Conv. Monograph-VI. E. B. C.-Symposium on the relationship between 

malt and beer, Helsinki, p. 179. 

PIERCE, J. S. (1982) J. Inst. Brewing, 88, 228. 

POLLOCK, J. R. A. and POOL, A. A. (1968) Proc. Ann. Mtg. Amer. Soc. Brew. Chem., p. 33. 

PRECHTL, C. (1967) M. B. A. A. Tech. Quart., 4, 98. 

RENNIE, H. and BALL, K. (1979) J. Inst. Brewing, 85, 247. 

RUDIN, A. D. (1974) Eur. Brew. Conv. Monograph-I. E. B. C.-Wortsymposium, Zeist, p. 239. 

SCHILD, D. (1936) Wochenschr. Brau., 53, 345, 353. 

SCHUR, F., PFENNINGER, H. B. and NARZISS, L. (1973) Proc. 14th Congr. Eur. Brew, Conv., Salzburg, 

p. 149. 

SCHUR, F., PFENNINGER, H. B. and NARZISS, L. (1975) Proc. 15th Congr. Eur. Brew. Conv., Nice, p. 191. 

SCHWARZ, P. B. and HAN, J.-Y. (1995) J. Amer. Soc. Brew. Chem., 53, 157. 

SCOTT, R. W. (1972) J. Inst. Brewing, 78, 179, 411. 

SCOTT, P. M. (1996) J. A.O.A.C. International, 79, 875. 


SISSONS, M. J. (1996) J. Amer. Soc. Brew. Chem., 54, 19. 

SJOÈ HOLM, K., MACRI, L. J. and MACGREGOR, A. W. (1995) Proc. 25th Congr. Eur. Brew. Conv. Brussels, 

p. 277. 

SOPANEN, T., TAKKINEN, P., MIKOLA, J. and ENARI, T.-M. (1980) J. Inst. Brewing, 86, 211. 

STARS, A. C., SOUTH, J. B. and SMITH, N. A. (1993) Proc. 24th Congr. Eur. Brew. Conv., Oslo, p. 103. 

STENHOLM, K. and HOME, S. (1999) J. Inst. Brewing, 105, 205. 

STENHOLM, K., HOME, S., PIETILaÈ, K., MACRI, L. J. and MACGREGOR, A. W. (1996) Proc. 24th Conv. Inst. 

Brewing, (Asia Pacific Section), Singapore, p. 142. 

STONE, B. A. (1996) in Cereal Grain Quality (Henry, R. J. and Kettlewell, P. S. eds). London, Chapman & 

Hall, pp. 251, 288. 

SUN, Z. and HENSON, C. A. (1990) Plant Physiol., 80, 310. 

TAYLOR, D. G. (1981) Brew. Distill. Internat., 11 (4), 35, 42. 

TAYLOR, L. (1974) Eur. Brew. Conv. Monograph-I. E. B. C.-Wortsymposium, Zeist, p. 208. 

TENG, J., STUBITS, M. and LIN, E. (1983) Proc. 19th Congr. Eur. Brew. Conv., London, p. 47. 

TESTER, R. F. (1997) in Starch Structure and Functionality (Frazier, P. J., Richmond, P. and Donald, A. M. 

eds). Cambridge. Royal Society of Chemistry, p. 163. 

VAN DEN BERG, R. and VAN EERDE, P. (1982) Proc. 17th Conv. Inst. Brewing (Australia and New Zealand 

Sect.), Perth, p. 70. 

VOSE, J. R. (1979) M. B. A. A. Tech. Quart., 16, 186. 

WACKERBAUER, K., ZUFALL, C. and HOÈ LSCHER, K. (1993) Brauwelt Internat., 11, 107. 

WAINWRIGHT, T. (1980) Eur. Brew. Conv. Monograph-I. E. B. C.-Symposium on the relationship between 

malt and beer, Helsinki, p. 118. 

WALKER, M. D., BOURNE, D. T. and WENN, R. V. (1997) Proc. 26th Congr. Eur. Brew. Conv., Maastricht, 

p. 191. 

WEBSTER, R. (1978) Brewer's Guard., 107 (7), 51, 56. 

WHITEAR, A. L., MAULE, D. R. and SHARPE, F. R. (1983) Proc. 19th Congr. Eur. Brew. Conv., London, p. 81. 

WHITTLE, N., ELDRIDGE, H., BARTLEY, J. and ORGAN, G. (1999) J. Inst. Brewing, 105, 89. 

WIEG, A. J. (1987) in Brewing Science 3 (Pollock, J. R. A., ed.). London. Academic Press, p. 533. 

WINDISCH, W., KOLBACH, P. and SCHILD, E. (1932) Wochensch. f. Brau., XLIX, (37, 38), 289, 298. 

WOOF, J. B. and PIERCE, J. S. (1966) J. Inst. Brewing, 72, 40. 

WRIGHT, H. E. (1892) A Handybook for Brewers. London. Crosby Lockwood, 516 pp. 

YAMADA, K. and YOSHIDA, T. (1976) Rept. Res. Lab. Kirin Brewery Co., no. 19, 25. 

ZHANG, N. and JONES, B. L. (1995a) J. Cereal Sci., 21, 145. 

ZHANG N. and JONES, B. L. (1995b) J. Cereal Sci., 22, 147. 

ZIEGLER, L. and PIENDL, A. (1976) M. B. A. A. Tech. Quart., 13, 177. 


5 

The preparation of grists 

5.1 Intake, handling and storage of raw materials 

Maltsandmashtunadjunctsmaybedeliveredbyroad,railor,morerarely,water.Forlarge 

breweries deliveries are in bulk, but for smaller ones deliveries may be in sacks varying in 

weight from about 25kg. to 1t. The largest sacks are handled with fork-lift trucks. Lorries 

generally carry loads of 20±25t, while American railway wagons carry about 68±82tof 

malt. Some small breweries receive their malt ready ground. Before aload is accepted it 

shouldbecheckedtoensurethatitisofthecorrectquality.Visualinspectionandrapidtests 

onmaltshouldindicatethatthecolour,moisturecontent,nitrogen(protein)content,flavour, 

aroma, friability, homogeneity and range of corn sizes are correct and that the load is free 

from insects and the corns are not unduly damaged (Briggs, 1998; Kunze, 1996; Narziss, 

1992; Nicol and Andrews, 1996; Rehberger and Luther, 1994; Sugden et al., 1999). 

Different types of malt and adjuncts must not be allowed to mix, either during handling 

or subsequent storage. Each load is weighed, for example, by weighing each lorry on a 

weighbridge before and after unloading, and during movements within the brewery. While 

mechanical handling can be used for malts and many adjuncts, flours must be moved using 

pneumatic equipment. Many different arrangements of equipment are used in dry-goods 

handling (Fig. 5.1). In the UK lorries usually discharge by tipping their loads into a 

reception pit which is sheltered from the weather and is aspirated to remove dust. 

Handling is with equipment that damages the material being moved as little as 

possible and so causes minimal losses and generates least dust. Usually the machines 

used are belt and bucket elevators, and screw, drag, chain-and-flight or en masse 

conveyors (Briggs, 1998). All these should be aspirated to remove dust, which is usually 

collected at a central point. Initially, malt is often roughly screened (sieved) to remove 

coarse and fine impurities, and is passed over magnetic separators (of fixed or revolving 

magnet types) to remove fragments of metal. Sometimes the malt is separated from 

`heavy contaminants' by passing it through a transverse airflow, which deflects the 

comparatively light malt while allowing denser objects to continue falling downward, to 

be collected separately. The removal of metal items (`tramp iron') and subsequent destoning 

are necessary to reduce wear on conveying equipment and the brewery mills and 

to reduce the risks of sparks which can lead to fires or explosions. 


Blower 

Malt 

silo 

Adjunct 

silo 

Rotary seal 

Malt 

bins 

Screen 

Rejected 

material 


Elevator 

Magnet 

Dust 

Fan 

Screened 

malt 

Dust 

Adjunct blowline 

Waste 

Filter 

unit 

Rotary 

seal 

Tip 

weigher 

De-stoner 

Buffer bin 

Steam 

Fig. 5.1 An example of a dry goods handling system. 

Waste 

Adjunct 

grist case 

Fan 

Cyclone 

Conditioning 

screw 

Mill 

Grist conveyor 

To cereal 

cooker or 

mashing vessel 

Waste 

To mash 

vessel 

Air 

Filter 

unit 

Malt 

grist 

case

In British breweries malt, as now delivered, often meets such tight specifications that 

preliminary screening is not needed. Grist materials, and especially malts, are 

hygroscopic and efforts are made during handling and storage to prevent them picking 

up moisture by contact with water or from the air and so becoming `slack'. Pneumatic 

conveyors may function on positive or negative (`suction') pressures. In positive-pressure 

systems a blower forces air along fixed or flexible pipework and malt or other material is 

introduced into the airstream from hoppers, via air-tight, rotating valves. The material is 

swept along and is recovered from the airstream in expansion chambers and cyclones. 

The material drops out from the slower airflow into a receiving hopper. Dust is separated 

from the air, using cyclones and textile filters, before escaping from the building. Modern 

filters are self-cleaning and occasional backflows of air dislodge the dust caked onto the 

filter sleeves, allowing it to fall into a reception hopper, from which it is removed through 

a valve. Negative-pressure systems work under suction so that material is sucked into the 

system. This is convenient for unloading rail wagons and barges and is also advantageous 

in that dust does not leak and escape. This type of conveyor can form part of the main 

dust collection system. The ùsed' air passes through cyclones and dust filters before it 

reaches the pump that provides the suction. Some grain-cleaning equipment re-circulates 

air, preventing the dust escaping and collecting it in the machine. 

Handling dry goods always generates dust. This is heavily infested with microbes and 

their spores and, mixed with air in a range of proportions, is highly explosive. It is 

estimated that 0.4±1.4% of malt delivered to a brewery is damaged and converted into dust. 

Intake hoppers, conveyors, elevators, stores, screens, de-stoners and dry mills should all be 

ventilated and the aspirating air directed to the dust-collection system. The dust is usually 

mixed with the spent grains and sold. Less usually it may be discarded or destroyed. Dust 

from malt and adjuncts is rich in starch and may be added into mashes. Dust, either in or 

around equipment such as mills and conveyors, should not be allowed to accumulate as it 

becomes damp and microbes and insects multiply on it. Dust must be kept away from the 

brewing plant as the microbes in it can initiate beer spoilage. In addition it constitutes a 

health risk, giving rise to skin allergies and lung infections. Dust explosions can be highly 

destructive and life-threatening. Rigid rules of behaviour and the use of well-designed 

equipment must be used to minimize the risks. Dust containment and removal are essential. 

Explosions may be initiated by sparks or flames and so stones and tramp iron are removed 

from malt and adjuncts, smoking and the use of flames (e.g. welding torches) is forbidden 

and all equipment is earthed or `grounded' to discharge static electricity safely and avoid 

sparking. Sensors are mounted on bearings to detect increases in temperature, which 

indicate poor lubrication, wear and the risk of initiating a dust explosion. 

Overloaded equipment is automatically closed down. Silos have dust-sealed lighting and 

silos and ducts are fitted with physically weak explosion relief hatches that will burst on a 

sudden pressure rise and will allow the explosion to vent in a comparatively safe way. This 

reduces the force being transmitted through the ducting and so reduces the chance of 

secondary explosions and other damage remote from the initial explosion site. Some pieces 

of equipment, such as dry mills, are fitted with devices which, when triggered, release an 

inert gas and quench explosion flames. The three most common sites of explosions in 

breweries, in decreasing order, are bucket elevators > silos > mills. As with all equipment, 

the dust-collection system should be inspected regularly, and be cleaned and maintained. 

Malts and adjuncts are stored so that they do not become damp and `slack' and they 

are protected from birds, rats, mice and insects. Storage is costly, and so it is desirable to 

have as small stocks as are considered prudent. In practice, breweries may carry stocks 

sufficient to allow from three days' to three weeks' production. In smaller or older 


eweries themaltsandadjunctsmaybestored insacks,traditionallyinaroomabovethe 

mill room. Usually malt is stored in bulk in bins or silos. These are usually of steel or 

concrete,withsmoothinternalsurfaceswithconicalbaseshavingavalveat thebottomto 

allow metered unloading, under gravity, into a conveyor. Different materials have 

different angles of repose and they need different valley angles in the tapered or conical 

bases to allow them to flow freely (see Appendix). 

There is atendency for flowing, granular material to segregate by size and density. To 

avoid this stores should be unloaded using `bulk flow' systems that avoid funnel flow. 

This can be achieved using suitable vessel designs (Farnish, 2002). Flour is often slow to 

flow, anditmay benecessarytohave avibrating cone oramechanical device (suchasan 

arch-breaker) in the cone to encourage movement. Silos are loaded by conveyor from 

above and are equipped with hatches and manways to allow inspections and cleaning. 

Level-detecting devices determine the approximate levels of the contents and prevent 

over-filling. They should contain thermometers to detect any temperature rises that give 

warning of the onset of spoilage. It is good practice to use old stocks first and never to 

mix old and new stocks. Each storage container should be completely emptied in turn, 

and be cleaned and, if necessary, fumigated before being refilled. If grist materials must 

be treated to control insects this must be done using only fully trained staff, using an 

approved insecticide or fumigant at apermitted dose rate. 

Batchesofmaltsoradjunctsandgristsareweighedastheyaremovedaroundinthestore, 

or to or from the mills. The records of the weighings are used to check on losses, that the 

correctmixturesofmaterialsentereachmash,andthattheyieldofextractfromeachmashis 

acceptable. Electronic and mechanical weighers and load cells are in use and some are 

capable of high degrees of accuracy. In many weighers the malt, or other material, flows 

from above into a weighing container. When the chosen weight is approached the flow-rate 

is reduced and as the weight is reached the flow is stopped. The weighing container then 

empties from the base into a receiving buffer hopper. When it is empty the base closes and 

refilling is resumed. Weighing is recorded as the number of fillings that have taken place. 

Sometimes hoppers and other containers, such as grist cases, are mounted on load cells, 

which indicate the weight of the contents. Load cells can be mounted under conveyors to 

measure the weight of material continuously as it is being conveyed. While older weighers 

were perhaps accurate to 2% new machines, using load cells, can be good to 0.1%. 

Before the grist materials reach the mill they may be screened again, removing all 

items larger or smaller than malt corns. They will pass over a magnetic separator and they 

should pass through a de-stoner and/or a `heavy object separator' and be aspirated. In destoners 

a thin stream of malt flows slowly onto the upper end of an inclined screen. Air 

passes up through the screen at such a rate that the malt `floats' and flows downwards, 

being collected at the base. Heavier objects, such as stones, are not lifted and rest on the 

screen. This has a jerking motion that moves the stones up the screen and over the upper 

edge into a collector (Briggs, 1998). 

When mixtures of different grist materials are to be used they must be moved to the mills 

and the grist case, bypassing the mill in some cases (as when wheat flour is used), in the 

correct proportions and at the correct rates so that the correct grist, well mixed, is supplied 

when mashing in occurs. `Dry goods handling' is controlled from a central point, where the 

states of the stores, whether equipment is running and the states of processes are monitored 

and controlled often with a mimic display or, increasingly, through a computer. Each system 

has built-in safety and warning devices (Bhaduri, 1996). For example, when the flow of malt 

from a silo to the mill is switched off the valves, conveyors, elevators, etc., are switched off 

in sequence, beginning at the silo, so each unit is emptied of malt before it stops running. 


Start up is in the reverse order. For safety and effective operation all the staff must be well 

trained.Checkingandmaintenanceoftheequipmentmustbecarriedoutregularly.Safetyis 

also dependent on the layout of the plant and this and the way it is operated must meet high 

design standards and local regulations. 

5.2 The principles of milling 

The objective ofmilling istobreak up maltandadjuncts tosuchanextent thatthegreatest 

yield of extract is produced in the shortest time in the mashing equipment in use. With 

most wort separation systems it is desirable to keep malt husk as intact as possible, to help 

maintain an open filter bed that favours wort separation. No unbroken grains should 

survive milling. Only comparatively coarse grists can be used in mash tuns. If the grist is 

too fine (has too high aproportion of small particles) the wort will not separate from the 

spent grains. Finer grists can be used in lauter tuns and still finer grists in mash filters. A 

`best practical' grist, with an optimal mixture of coarse, fine and very fine particles, is 

determined for each mashing system. Optimal milling is not simple to achieve. Mill 

settings need to be altered for adjuncts or malts differing in their degrees of modification 

(badly modified malts should be milled to afiner grist, well modified malts readily 

shatter), and quite small differences in moisture contents alter the effectiveness of milling 

using one mill setting (Crescenzi, 1987). Sometimes different mills are used for malt and 

for adjuncts, or for pale malts and coloured malts. In other cases adjuncts, such as torrified 

barley or wheat, flaked cereals, and roasted barley, may be milled mixed with the malt(s). 

The term `grinding', for milling, comes from the days when malt was ground between 

millstones, and the ground material is still termed `grist'. This should be distinguished 

from grain particles, present in grists, which are termed `grits'. Milling systems may use 

rollermilling,impactmillingwithdisc-orhammer-mills,andwetmilling.Properly,none 

of these modern mills `grinds' the malt. Roller milling (dry or conditioned), giving 

comparatively dry grists, is the most usual. These grists are evaluated by inspection, by 

sieve analyses and sometimes in laboratory mashes. A grist should be uniform in 

appearance, be free from taints and insects or other contaminants and should not contain 

any whole grains or large grain pieces that indicate that part of the grist has not been 

effectively milled. 

Sieve analyses should be used regularly to check that amill's performance is not 

drifting. The most commonly used sets of sieves are the Pfungstadt plansifter (EBC), the 

MEBAK sieves and the sieves of the ASBC (Table 5.1). Asample of grist is loaded onto 

the top of aset of horizontal sieves, which is shaken mechanically for afixed period of 

time. The percentage, by weight, of the grist retained on each, successively finer sieve is 

then determined. By convention the sieve fractions are given names (Table 5.1), but the 

fractions are not `pure' so, for example, the `husk' fraction contains tissues from the rest 

of the malt grains and husk tissue occurs in the other fractions. Furthermore, the grist 

fractions from amalt milled in different ways are likely to have different compositions. 

The results in Table 5.2 show that starch is present in the `husk' fractions proving 

contamination with endosperm material while the distribution of fibre in all the fractions 

suggests that at least traces of husk occur in all of them. The variable extract yields and 

qualities obtained when different fractions of agrist are mashed has often been noticed 

(Hind, 1950; Narziss, 1992; Stubits et al., 1986; Table 5.3). It has been suggested that 

different beers could be produced by combining the fractions in different proportions in 

the mash (e.g. Isoe et al., 1991), or that beers should be made with husk-depleted grists or 


Table 5.1 Details of the sieves used to characterize grists. The names given to the fractions do 

NOT indicate that these are morphologically pure grain fractions. The composition of the fractions 

changes with the type of milling employed. The American Society of Brewing Chemists (ASBC) 

and the European Brewery Convention (EBC) (Pfungstadt) sieves are significantly different and the 

MEBAK sieves are different again (see footnote). All dimensions are in millimetres 

ASBC EBC 

Sieve Mesh Fraction Sieve Wire Mesh Fraction 

number width name (Pfungstadt) thickness width name 

(mm) number (mm) (mm) 

10 2.000 Husk 1 (16) 0.31 1.270 Husk 

14 1.410 Husk 2 (20) 0.26 1.010 Coarse grits 

18 1.000 Husk 3 (36) 0.15 0.547 Fine grits I 

30 0.590 Coarse grits 4 (85) 0.07 0.253 Fine grits II 

60 0.250 Fine grits 5 (140) 0.04 0.152 Flour 

100 0.149 Flour Tray (through) ± ± Fine flour 

Pan (through) ± Fine flour 

The MEBAK screens I, II, III, IV and V have mesh widths (mm) of 1.25, 1.00, 0.5, 0.25 and 0.125. 

Table 5.2 The analyses of ASBC sieve fractions of Larker malt milled with a 5-roller BuÈhler- 

Miag malt mill (data of Stubits et al., 1986). Grist samples (100 g) were mashed with a liquor/grist 

ratio of 3/1. The mash programme was 45 ëC/67 min., then a temperature rise of 1 ëC/min. to 69 ëC, 

a hold for 35 min., then a rise to 72 ëC 

Sieve Starch Protein Fibre Wort Viscosity 

number (%) (%) (%) (ëP) (relative) 

10 50.1 12.0 7.1 16.6 2.07 

14 56.1 13.8 4.1 19.1 2.60 

18 63.4 11.6 3.5 20.9 2.62 

30 54.9 13.4 3.7 20.4 2.50 

60 47.1 15.4 5.2 19.8 2.44 

100 49.4 17.7 4.3 20.1 2.41 

Pan 69.9 11.8 1.9 21.7 2.54 

Table 5.3 Analyses of a malt grist and the Pfungstadt sieve fractions 1±6 of that grist (data of 

Narziss and Krauss, via Narziss, 1992) 

Analyses Entire Husks Coarse Fine Fine Flour Fine 

grist (1) grits (2) grits I (3) grits II (4) (5) flour (6) 

Proportions (%) 100 27.6 15.3 22.9 13.2 6.6 14.4 

Extract (%)* 80.2 64.4 79.5 87.9 84.2 83.3 96.8 


time (min.) 9 8 9 8 9 10 12 

Fermentation 

limit (%) 80.9 77.3 78.0 82.0 82.5 80.9 83.2 

Protein (%)* 11.1 12.4 11.9 10.6 11.4 13.4 7.6 

TSN (mg/100 g)* 711 584 681 705 847 854 526 

Viscosity (mPa.s, 

8.6% wort) 1.515 1.534 1.463 1.481 1.443 1.467 1.407 

D P (ëW.-K) 302 225 323 361 347 327 250 

-Amylase (ASBC) 40 32 44 51 48 47 36 

Tannoids (mg PVP/ 

100 g)* 22 12 32 20 21 24 12 

Colour (ëEBC) 3.3 4.7 2.8 2.5 2.8 2.8 1.3 

TSN, total soluble nitrogen. D P, diastatic power. PVP, polyvinyl pyrrolidone. * Dry weight basis. 


Table 5.4 EBC/Pfungstadt sieve analyses (% in each fraction) of all-malt grists suitable for 

different mashing systems (Kunze, 1996; Narziss, 1992; Sugden et al., 1999) 

Sieve Mash tun Lauter tun Mash filter (conventional) Thin-bed filter 

a b a b c a b 

1 27 18 18±25 11 8±12 10±12 1 1 

2 9 8 < 10 4 3±5 4±6 4 2 

3 24 35 35 16 15±25 < 15 9 15 

4 18 21 21 43 35±45 > 43 26 29 

5 14 7 7 10 8±11 > 12 19 24 

Tray 8 11 < 15 16 12±18 < 16 41 29 

the fractions should be mashed separately in different ways, to obtain abeer with a 

`refined' flavour (Krottenthaler et al., 1999). While this is technically possible this 

approach has been little used, probably because of the extra cost of the fractionation 

procedures and the problem of using or disposing of the `husk' fraction. 

Examples of all-malt grists regarded as being suitable for use with different mashing 

systems are given in Table 5.4. In moving from the mash tun grist across to the grist for the 

thin-bed filter the grists become finer. Itis seen thatthere are small differences between the 

proposed grists. This is expected as `best' grists are determined by brewers using trial and 

error, employing different mills and equipment. Grists prepared by wet milling are not 

suitable for particle size analyses. In these cases grist quality is judged indirectly by noting 

the appearance and performance of mashes, particularly extract yield, run-off rate, wort 

clarity, the appearance of the spent grains and so on. The same criteria are also applied to 

mashes made with dry milled grists, which are also subjected to sieve analysis. 

Roll (or roller) mills are commonly used in breweries. The rollers work in pairs. The 

malt grains are delivered to the first pair of rolls by afeed roll that determines the feed 

rateandisintendedtodelivereachcornènd-on' totheworkingrolls.Each cornisdrawn 

between the rolls and is crushed, sheared (if the rolls are rotating at different speeds) and 

cut if the rolls are fluted (grooved). The theoretical capacity of aroll mill, Q(m 3 /h), is 

given as Qˆ60.s.NL10 9 ,where sˆrotational speed (rpm); Nˆthe gap between the 

rollers or `nip'(mm); Lˆthe length of the working surfaces of the rolls (mm). In fact the 

practical working capacity is 10±30% of the theoretical capacity (Sugden et al., 1999). 

Working rates of different mills (in kg/h/mm roll length) are given as two-roll mills, 1.5± 

2.5; four-roll mills, 2±6 and six-roll mills, 1.5±10. 

Theactionofcrushingrolls,withcentresA1andA2,onaparticle,centre B,isillustrated 

in Fig. 5.2. Acomponent, e, of the force ttends to draw the particle between the rolls. The 

force t depends on the force r and the coefficient of friction between the surfaces of the 

particle and the roll, , so t ˆ r. The force components e and m are opposed and so unless 

e > m the particle will not be pulled between the rolls and be crushed. Thus r cos > r 

sin and so > tan which must, therefore, be less than the coefficient of friction. Often 

a typical value for is 16 ë. The angle OEF, the angle of nip, equals 2 There is a definite 

relationship, cos ˆ radius of the roll + half the gap between the rolls)/(radius of the roll + 

radius of the particle). If is 16 ë, then cos ˆ 0.961. If the angle of nip is too large 

(because the rollers are too small) and/or the rolls are rotating too quickly, particles `ride 

the rolls', so they are not drawn between them and they are not crushed. 

The peripheral roll speeds in brewery mills are often 2.4±4 metres/sec (8±13 ft./sec.). As a 

particle moves between the rollers so it deforms and, if it is brittle, its structure fails and it 

breaks up (Sugden et al., 1999). Rolls may move at different speeds, for example the faster 

may rotate at 1.25 the speed of the slower. Consequently a particle passing between them 


m 

A 1 

r 

α 

n 

will be torn by shear as well as being crushed. As well as that the grooves, or `fluting' milled 

in spirals on the surfaces of the rollers can be arranged not only to increase the coefficient of 

friction between the particles and the rolls, but also to cut the particles (Kunze, 1996). 

Roll mills differ in their complexity. They have to cope with materials, particularly malt 

corns, that differ in their ranges of widths and so the nips, the gaps between the rolls, must be 

adjustable. In addition corns and other materials break up to give particles of a range of sizes. 

Some of these, the fine grits and flour, need no more milling while larger materials can, with 

advantage, be broken up more. In more complex mills this is achieved by sieving the material 

coming from a pair of rolls on cylindrical or flat-bed screens. The fine fractions leave the mill 

directly while coarser fractions are directed between other pairs of rolls and are broken up 

further. In the cylindrical screens the grist may be disrupted by spinning `beaters'. In the flatbed, 

oscillating screens rubber balls move about and oscillate vibrating the sieving surfaces 

and keeping them clear of blockages. Dry mills deliver their grist to a receiving hopper or 

`grist case', so several hours working time may be available to prepare a batch of grist. In 

contrast, with `wet' mills the grist is mashed in as it is produced and the mashing-in period 

must not be too long. Consequently wet mills operate at faster rates than dry mills. 

5.3 Laboratory mills 

O 

f 

e α t 

D 

C 

α α 

E 

B 

Fig. 5.2 The action of two crushing rolls, rotating at equal speeds, on a particle such as a malt 

corn (see the text). 

Many kinds of mills are used in laboratories, but for standard analyses particular mills, 

operated in closely defined ways, are employed. This is because the analyses are used as 

bases of commercial transactions as well as standards by which the value of malts to a 

brewery can be judged and so different laboratories must obtain analytical results that are in 

close agreement. During the last century the `standard' mills have changed. Initially handcranked 

Boby or Seck mills, each with one pair of small rolls, were used and, in time, these 

were powered with electric motors. These mills were not able to give sufficiently 

reproducible results and so, at least for fine grinds, some used cone mills. The EBC 

introduced a Casella mill in which the malt was shattered by the blades on a spinning rotor 

and the grist was collected after passing through a sieve, different sieves being used for the 

coarse and fine grinds. In a comparatively short time this mill was replaced by another, a 

BuÈhler-Miag disc mill, which is in use at present by both the EBC and the IGB. In this 

sophisticated machine the sample to be ground is fed into the central area between two 


F 

A 2

discs, one static, the other rotating fast at afixed speed. The grist emerges from the gap at 

theperipheryandiscollected. The fineness ofgrindisdeterminedbythedistancebetween 

thediscs,whichfortheIoBmethodsare0.7mmforthecoarsegrindand0.2mmforthefine 

grind. It is apparent that this mill differs in its method of working from almost all 

commercial malt mills used in brewing. This `disadvantage' is more apparent that real, 

since the results it gives are highly reproducible and, as with every analytical system, each 

brewer has to establish the relationships between the laboratory analyses and the results 

obtained with his, possibly unique, brewing system. 

5.4 Dry roller milling 

Alarge number of types ofroller millshave been, or are,in use (Kunze, 1996; Narziss, 1992; 

Sugdenet al.,1999). `Dry' millshave some characteristics in common. Thefeed roll delivers 

themalttothefirstpairofcrushingrolls,acrosstheirfullwidth,atacontrolledrate.Therolls 

aredesignedtodeliverthecornsèndon'tofavourtheirbeingcrushedalongtheirlengthwith 

theminimumdegreeofhuskbreakage.Therollsareoftenabout250mm(9.84in.)indiameter 

(inwetmillstheyareoftenlarger).Asnotedtherollsareusuallyflutedandmayrunatdifferent 

speeds. Bothmay be drivenbutsometimes onlyone ispowered, the`follower' being dragged 

bythefriction between thegristandthemoving,poweredroll. Millrollsmayoperate at 250± 

500rpm. The roll length increases with machine capacity, to amaximum of about 1500mm 

(59.06in.). The rolls are spring loaded, and so the gaps between them must be checked while 

theyareworkingunderload.Thisisachievedbypassingsoftmetalwirebetweentheworking 

rolls,thenmeasuringthethicknessofthesquashedmetalwithamicrometerscrewgauge.Lead 

wire was originally used, but now aluminium, copper or lead-free solder wires are employed, 

toavoidanychanceofleadcontamination.Therollgapsandthealignmentoftherollsmustbe 

checked regularly, to ensure that they are parallel. In addition the rolls must be checked for 

wear. The grist coming from each section of the roll pairs must be inspected and checked by 

sieveanalysis.Thistestwilldetectwhenthecentralportionsoftherollsarebadlywornbutthe 

endregionsarenot.Wherescreensareusedthesetoomustbecheckedforintegrity.Afteruse 

the mills should be cleaned. At no time should insect infestations be present. 

The simplest mills are single-pass, two-roll dry mills. These are relatively slow 

working and are inflexible, being suitable only for well-modified malts, special malts or 

rice. They are used only in small units, such as pub breweries. Gaps used are 0.6±1.0mm 

(0.024±0.039in.) and working capacities are 1.5±2.5kg/h/mm roll length (a working rate 

of 1kg/h/mm roll length is almost exactly 60lb/h/in. roll length). It seems that three-roll 

mills (with or without screens) are no longer in use. The arrangement of the three rolls 

was like the grouping found in five-roll mills (see below; Sugden et al., 1999). 

Four-roll(two-high)millsarecommoninsmaller,traditionalalebreweries(Figs5.3and 

5.4). They are robust, relatively inexpensive and well suited for milling well-modified 

malts with a small range of corn sizes. Working rates are 2±6 kg/h/mm, the gaps between 

the upper rolls are 1.3±1.9 mm (0.051±0.075 in.) and between the lower rolls 0.3±1.0 mm 

(0.012±0.039 in.). In the more simple types the cracked grist from the first rolls piles up on 

a `dam', `shelf' or èxplosion preventer', where it can smother any sparks and so prevent an 

explosion (Fig. 5.3). It then spills over and falls into a beater chamber, where the crushed 

malt is broken up and flour and grits are partly separated from the husk before passing 

through the second pair of rolls, over a second explosion preventer and out to the grist case. 

More sophisticated types of four-roll mills use screens to separate the grist into fractions of 

which only the coarse grits are crushed further by passage between the second pair of rolls 


Fixed roll (fluted) 

Inspection door 

Sample valve 

Handle operating 

valve for sample 

for top rails 

Inspection door 

Fixed roll (plain) 

Sample shoot 

Feed inspection 

door 

Sample 

tube 

Feed valve 

Explosion 

preventer 

Feed roll 

Fixed roll 

(fluted) 

Explosion 

preventer 

Beaters 

Loose roll 

(plain) 

Fig. 5.3 Asimple four-roll mill (after Hind, 1950). Note that in the second pair of rolls one is 

driven (`fixed') while the other is not powered (`loose'). 

(Fig.5.4).Thisavoidsmoredamagetothehuskmaterialandsofavoursagoodwortrun-off 

rate. The screens may be semi-cylindrical and equipped with revolving beaters, as in Fig. 

5.4, or they may be oscillating, flat-bed screens. The larger the screen areas the better the 

separations achieved. The arrangements of the screens can be varied so that, for example, 

flourandgritsarere-milledbuthuskisnotor,alternatively,gritsandhuskarere-milledbut 

flour and fine grits are not (Sugden et al., 1999). 

So-called eight-roll mills were really two four-roll mills working in parallel. The malt 

was pre-screened to divide it into bold (plump) and thin corns and these were milled 

separately, each stream passing through afour-roll mill adjusted to deal with the particular 

cornsizes.Five-rollmillsarenowuncommon(Fig.5.5),buttheyarelessexpensivethatthe 

six-roll mills. The malt is crushed between the first pair of rolls (say 1and 2) and the grist 

fromthisoperationisscreened.Thehuskfractioniscrushedfurther,(betweenrolls2and3) 

and the products are separated on asecond set of screens which direct the flour and husks 

(withgritsattached)toleavethemill,andthegritstopassthroughthethirdpairofrolls(say 

4and5).Atthesametimetheflourfromthefirstpairofrollsisguidedoutofthemillandthe 

grits join those from the second pair of rolls and are broken up further by the third pair. 

Six-roll mills are widely used in larger breweries as they are flexible in their operation 

and are able to mill malts with differing degrees of modification and some adjuncts. 

Several arrangements of the rolls and screens are used (Sugden et al., 1999; Wilkinson, 

2001). In one type malt from the feed roll is crushed by the first pair of rolls and then by 

the second set of rolls (so at this stage the treatment resembles that given by simple fourroll 

mills). The grist from the second pair of rolls falls onto aset of screens, which divide 


1st pair of rolls 

Cylindrical 

screen with 

revolving 

beater 

Fine grits 

and flour 

Feed roll 

Malt 

To grist case 

Feed valve 

Anti-explosion device 

Anti-explosion 

device 

Coarse 

grits 

2nd pair 

of rolls 

Fig.5.4 Afour-rollmillusingcylindricalscreenstofractionatethegristfromthefirstpairofrolls 

and to direct the coarse grits to the second pair of rolls (after Hind, 1950). Similar results are 

obtained in mills with oscillating, flat-bed screens. 

it into size fractions. Several different arrangements are then possible. For example, the 

huskandflourfractionsleavethemillwhilethegritsarebrokenmorebypassingbetween 

the third pair of rolls. In one particular system the grist from the first and second rolls is 

divided into two equal streams, which pass to two identical sets of tandem flat screens. 

This arrangement is compact and allows the use of large screen areas (and so more 

efficient fractionation of the grist) and, because the sets of screens are set to oscillate ìn 

opposition', the vibration of the mill is reduced. 

The arrangement of amore common pattern of six-roll mill is shown in Fig. 5.6. Sets 

of screens operate between the first and second and second and third pairs of rolls. The 

coarse material from the first stage pass to the second set of rolls, the grits go to the third 

pair of rolls and the flour leaves the mill. The husk and flour from the second rolls leave 

the mill while the grits pass to the third pair. These mills, appropriately adjusted and with 

the correctly fluted rollers, can be used to prepare grists that are suitable for use in lauter 

tuns or the older types of mash filters. The malt delivered to these mills may be dry or 

`conditioned'(see below). Suggested gaps (mm) for the first, second and third pairs of 

rolls are, (a) for dry malts and cereals using alauter tun, 1.6±2.0, 0.7±1. and 0.2±0.4; (b) 

for conditioned malts using alauter tun, 1.4±1.9, 0.5±1.0 and 0.2±0.4, while for dry malt 

and cereals using an older type of mash filter; (c) the values are 1.0±1.4, 0.4±0.6 and 0.1± 

0.3 (1mm ˆ0.03937in.). The working rates of six-roll mills are 2±10kg/h/mm roll 

length for lauter tun grists and 1.5±8kg/h/mm for mash filter grists. 


Feed-roll 

Pair of 

rollers 

crushing 

the husk 

fraction 

Second 

set of 

shaking 

screens 

5.5 Impact mills 

Pair of rollers, carrying 

out first crushing 

Malt 

Husk and attached materials 

Grits 

Flour 

Flour Grits Flour 

Pair of 

rollers 

crushing 

grits 

Husks 

with grits 

attached 

First 

set of 

shaking 

screens 

Fig. 5.5 The working principles of a five-roll mill (various sources). 

Feed roll 

First pair of 

crushing rolls 

First set of 

shaking 

screens 

Malt 

inlet 

Flour 

Grits Flour 

Husks and Pair of rollers 

attached for reducing 

grits coarse grits 

Grits retained by 

second screen 

Coarse materials 

retained by top screen 

Fig. 5.6 A scheme of a six-roll mill (various sources). 

Pair of rollers 

crushing 

coarse fraction 

Second set 

of shaking 

screens 

Disc and pin `dry' mills are often used on the experimental or pilot brewery scale and 

they are used in some small breweries (Biche et al., 1999). In these mills the material to 


e ground is delivered to the gap at the centre, between two discs the faces of which are 

roughened or covered with short projections. One disc may be static and the other rotate 

or both may spin, but in opposite directions. The gap between the discs may be adjusted. 

As the material moves from the centre to the periphery it is ground. Usually the grist is 

carried away in an airflow, but there has been interest in disc milling under water, when 

milling and mashing in occur together (see Section 5.9). 

For years hammer milling has been used to prepare some raw grain adjuncts. It was also 

used experimentally and in alimited number of breweries for making finely ground grists 

from which, after mashing, worts were separated by centrifugation or with rotating vacuum 

filters(Briggsetal.,1981;RehbergerandLuther,1994).However,theintroductionofhighpressurefiltersandtheMeura2001filterhasencouragedtheuseofhammermillingforgrist 

preparation because in these devices the survival of large fragments of husk is irrelevant in 

collecting the wort. Hammer mills differ in detail, but the principles of operation are the 

same (Fig. 5.7). Malt, sometimes mixed with adjuncts, is fed at apre-determined rate 

through arotary valve or afeed roll, into the milling chamber, which is strongly ventilated. 

The chamber may be mounted vertically or horizontally. The malt must be scrupulously 

cleanedtoremovestones,piecesofmetaland`heavyobjects'.Themillingchambercontains 

aspinning rotor (e.g. turning at 1500rpm) on which are mounted freely swinging pieces of 

metal, the beaters or `hammers', which travel at about 100m/s. The inertia of the rotors is 

such that they may take 20±30min. to stop. 

Air 

inlet 

Inlet to 

milling 

chamber 

Hammers 

Buffer 

hopper 

Airflow carrying 

grist 

Rotary valve/ 

feed roll 

Slide 

Rotor 

Screen 

Anti-vibration 

mounting 

Fig. 5.7 The layout of atype of hammer mill (various sources). 


The impacts of the hammers on the malt smash it up and this process continues until the 

fragments escape through the semicircular screen that makes up part of the wall of the 

chamber. Sometimes the inner wall of the chamber also carries short projections against 

which the moving malt can impact. The screens may have mesh widths of 0.5, 1.0mm or 

even 2±4mm. The powdered grist is carried out of the mill in the airflow, which transports 

it to the grist case, which is equipped with explosion vents. In some mills the rotors are 

reversible to obtain even wear on the hammers. These mills are comparatively inexpensive, 

but the wear is substantial and so maintenance must be regular and easy to carry out and 

screens and hammers needregular replacement. Such mills mustbe set up onanti-vibration 

mountings, and housed in acoustically insulated chambers, to muffle the noise. 

5.6 Conditioned dry milling 

In mashing and the subsequent separation of the wort from the spent grains, in mash tuns 

or lauter tuns, it is desirable that husk materials should be as nearly intact as possible, but 

withdrymilling this is difficulttoachieve while comminuting the endosperm tissue toan 

adequate extent. The problem is largely due to the extreme brittleness of the dry husk. 

The ideal arrangement for preparing most grists would be to be able to mill malt with 

damp, flexible husks but with dry, brittle interiors. To adegree this is achieved by 

`conditioning' malt by briefly dampening it, wetting the husk with water or steam, before 

it reaches aconventional `dry' mill (Fig. 5.1). The intention is to mill the treated malt 

while the husk is damp and flexible but before any moisture reaches the endosperm and 

reduces its brittleness. A conditioning screw consists of a screw or paddle conveyor 

working in a heated casing. The moving stream of warm malt may be exposed to lowpressure 

steam, at 0.5 bar, for 30±60 s. Then, after a 90±120 s equilibration time, the malt 

is delivered to a six-roll mill. To dry the equipment and minimize corrosion the steam is 

turned off 5 min. before the last of the malt passes through. Alternatively, and with a 

lower risk of enzyme inactivation, water at 30 ëC (86 ëF) or, at least, < 40 ëC (104 ëF), is 

sprayed onto the malt and then, after a one-minute equilibration period, the malt enters 

the mill. The moisture content of the husk is increased by 1.5±1.7%. 

The dampened husk is more flexible and survives milling better and the volume of the 

husk sieve fraction is increased by 20±30%. Dry milled grist volumes are 500±700 ml/100 g 

while the volumes of conditioned milled grists are 700±1000 ml/100 g. The volumes of the 

spent grains are also increased. The mill gaps need to be set closer to obtain the best extract 

from conditioned malt. The run-off rate is increased by conditioning and lauter tuns may be 

loaded more deeply using a mash made with conditioned malt as the bed density is reduced 

and its porosity is increased (Narziss, 1992; Stoscheck, 1988; Sugden, et al., 1999; 

Wilkinson, 2001). It is also said that conditioning gives a better yield of extract, better 

attenuation and faster saccharification. To prevent clogging dust must be completely 

removed from the malt before it reaches the conditioning screw and this and the adjacent 

pieces of equipment must be cleaned regularly. 

5.7 Spray steep roller milling 

Spray steeping malt before it is milled is a comparatively recent innovation. In spray 

steeping mills the malt is held dry in a hopper from which it is fed, at a controlled rate, into a 

spray chamber or conditioning shaft or chute, so designed that the malt moves through with a 


Mashing liquor 

Steep-conditioning 

chamber 

CIP spray 

Sprays for 

mashing liquor 

Malt inlet 

Malt hopper 

CIP spray 

`plug flow', that is `first in' is `first out' (Fig. 5.8). Here it is sprayed with warm/hot water for 

a short time, some quoted conditions being 60±80 ëC (140±176 ëF) for 45±60 s, or 50±70 ëC 

(122±158 ëF) for 60±120 s. (Herrmann, et al., 1998; Kunze, 1996; Langenhan, 1992; 

Wilkinson, 2001). Surplus water is collected from the spray chamber and is re-circulated. 

Water usage is 30±80 litres/100 kg malt. Generally the mill has two large, stainless steel rolls 

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.), 

but four-roll versions are also used. If the malt is hard the mill `senses' the increased 

workload, so the machine automatically runs more slowly, exposing the malt to a longer 

period of wetting. The moisture content of the husk is increased to 18±22%. 

Immediately after milling the mashing water is added to the grist and the mixture is 

pumped to the mash vessel. Lactic acid can be added at this stage to adjust the pH of the 

mash. Because this method of milling and mashing allows air to be mixed into the mash 

there is a chance that unwanted oxidations may occur. This can be prevented by 

excluding air, by filling the mill chamber with carbon dioxide or nitrogen gas. Husk 

survives milling well, and endosperm tissue is adequately broken up. The entire milling/ 

mashing process is complete in 20±30 minutes. In an alternative arrangement the malt is 

CIP 

Feed roll 

CIP spray 

Surplus water 

Feed roll 

Crushing rolls 

Mash 

Mash pump 

Fig. 5.8 A diagram of a mill using spray-steep conditioning (various sources). CIP, fittings of the 

`cleaning in place' system. 


metered into the mill by a rotating, segmented wheel (where it is wetted) which replaces 

the conditioning shaft (Kunze, 1996). This system has all the advantages of conditioned 

milling, that is, husk survival is enhanced, the grist and spent grains occupy a large 

volume so their porosity is increased giving shortened lauter times and good yields of 

extract. Between uses these mills must be thoroughly cleaned. They are equipped with 

CIP systems. The performance of these mills is relatively inflexible (Wilkinson, 2003). 

5.8 Steep conditioning 

In another, older type of wet mill the malt, with or without some raw cereal, is held in a 

hopper in which it is steeped (Fig. 5.9; Healy and Armitt, 1980; Kunze, 1996; Meisel, 

1997; Narziss, 1992; Sugden et al., 1999). The liquor used in the steep is usually at 30± 

50 ëC (86±122 ëF) and steeping lasts from 10 to 30 minutes. The steep liquor may be 

recirculated. After steeping the water is drained away, before the malt is milled. 

However, moisture takes time to penetrate the corns and so, inside the corns, the first malt 

to reach the mill is less moist than the last, so treatment is uneven. The malt has to be 

dust-free to prevent clogging. The steeped malt reaches a moisture content of about 30%, 

so the contents are partly softened. During milling they are gently squeezed flat by the 

large rolls (400 mm, 15.75 in. diameter; 440 rpm, gap 0.30±0.45 mm, approx. 0.012± 

0.018 in.) of the mill, squeezing out some of their contents. 

A mill may have two or four rolls. As with spray steeping the husks remain intact but 

hard ends in the malt are probably not adequately disrupted. Wet milling gives rise to a 

large volume of spent grains, indicating that mashes are òpen' and that the beds have 

high porosity. The control of the moisture content of the malt is less exact than that 

Spray-head 

for steep 

Malt inlet 

Malt steeping 

hopper 

Slide 

Water Feed roll 

To mash 

vessel 

Steep water 

outlet 

Drain 

CIP ring 

Overflow 

Crushing rolls 

Mashing liquor 

Mash-mixing chamber 

Homogenizer 

Mash pump 

Fig. 5.9 A mill using steep conditioning (various sources). 


achieved with spray steeping and the mill must be cleaned very well, initially with hot 

mashing liquor, to prevent microbes multiplying. The steep water may be discarded, in 

which case some extract is lost and effluent is produced or it may be added to the 

mashing liquor. This may carry aflavour penalty. Mashing liquor is added to the mixture 

immediately after milling and after mixing the mash is transferred to the mash vessel 

using amash pump. Advantages claimed for this system include increased brewhouse 

yield, faster wort separation in the lauter tun and smoother tasting beers as well as dust 

suppression (Stauffer, 1974). On the other hand the system is inflexible and has other 

inherent disadvantages (Wilkinson, 2001). Spray steeping, with its ability to break up dry 

and brittle endosperm while keeping husk tissue intact, is now preferred. 

5.9 Milling under water 

Experimentally, other types of wet milling are being investigated. It has been found that 

by steeping malt and by grinding it under water in adisc mill, afine grist suitable for use 

in aMeura 2001 mash filter can be rapidly produced (Biche et al., 1999; De Brackeleire 

et al., 2000; Wilkinson, 2003). Milling and mashing in occur together. The milling discs 

(often 600mm, 23.6in. diameter) rotate at 1275rpm. The gap between the discs is 

variable, 0.35±0.55mm (0.0138±0.0217in.) often being used. The smaller the gap 

betweentheplatesthefinerthegrindachieved,butwithagreaterpowerconsumption.To 

prevent oxidation it is desirable to de-gas the liquor and displace the air from the malt 

using an inert gas. The system is said to be slightly superior to hammer milling and to 

give agrist with very good filterability and ahigh yield of extract. 

Another proposal is to obtain an exceptionally fine grind by breaking up malt and 

adjuncts in a`dispersion chamber' (Menger et al., 2000a,b). In this device the mixture of 

malt and water passes through aseries of spinning, short, slotted rotors and stators which 

disrupt it by shear and probably impacts. Thus milling and mashing in are carried out 

together. Again, it is desirable to de-aerate the liquor to minimize oxidative changes. 

Within limits it is possible to increase the fineness of the grist by using more disrupting 

units in the series, set to achieve increasing degrees of disaggregation. 

5.10 Grist cases 

With wet mills, milling is coincident with mashing in and so the entire operation must be 

completed in acomparatively short time, say 20±30 minutes. However, this is not the 

case with dry milling as the grist can be accumulated and stored, at least for ashort 

period, ready for mashing in, in aseparate operation. Dry grists can be evaluated using 

standard sieves (Tables 5.1, 5.4). Grists can have low bulk densities; for example, 100kg 

can occupy 3 hl (approx 20.8 lb./ft. 3 ). Thus a dry milled grist may occupy 2.6 hl/kg, while 

a conditioned malt grist might occupy 3.2 hl/ kg (Narziss, 1992). Thus grist cases, the 

containers in which grists are stored, have large volumes and must contain enough grist 

for a mash. Usually they are made of mild steel, and are designed to contain dust and 

exclude steam and damp air, while being able to release their load under gravity, at the 

required rate. Sometimes the cases are fed from a mill and, by a proportional feeder, from 

a store of an adjunct such as wheat flour, that needs to be mixed uniformly into the grist. 

The grist may be directed to the mashing in system through a chute or by way of a gentle 

belt conveyor. Unless the grist is intended for a new type of mash filter abrasion caused 


y rough conveying or handling should be avoided so that the grist is not broken down 

further. All the components of a grist should be well mixed, and so any vibrations or 

conveying that favour the segregation of the components of the grist, for example, into 

layers of husk and fines, must be avoided. Commonly a grist case is above one mashing 

vessel, which it serves. However, sometimes a grist case can be equipped with chutes or 

conveyors that allow it to serve more than one mashing vessel, while other grist cases can 

be moved so that they can serve more than one vessel and several containers, each 

containing enough grist for one mash, may be grouped together in one structure. 

In two old processes, that have fallen out of use, the grist in the case could be heated by 

steam-heated pipes placed in the grist case. In `retorrification' the grist was warmed to the 

mashing-in temperature. Thus the temperature of the mashing liquor did not need to be 

greatly above that of the grist to achieve the correct initial heat, and so thermal inactivation 

of enzymes, caused by local overheating, was minimized on mashing in and more uniform 

mashing temperatures were achieved when the ambient temperature fluctuated. 

When grists were subjected to àromatization' they were heated to about 130 ëC 

(266 ëF) for 15 minutes to increase the colour and aroma. Inevitably this must have caused 

some enzyme inactivation. 


BHADURI, R. (1996) Brew. Distill. Internat., Mar., 18. 

BICHE, J., HARMEGNIES, F. and TIGEL, R. (1999) Proc. 27th Congr. Eur. Brew. Conv., Cannes, p. 593. 

BRIGGS, D. E. (1998) Malts and Malting. London. Blackie Academic and Professional, 796 pp. 

BRIGGS, D. E., HOUGH, J. S., STEVENS, R. and YOUNG, T. W. (1981) Malting and Brewing Science (2nd edn). 

Vol. I. Malt and Sweet Wort. London. Chapman and Hall, 387 pp. 

CRESCENZI, A. M. (1987) J. Inst. Brewing, 93, 193. 

DE BRACKELEIRE, C., HARMEGNIES, F., TIGEL, R. and MENDES, J. P. (2000) Brauwelt Internat., 18 (5), 372. 

FARNISH, R. (2002) Brewers' Guard., Feb., p. 26. 

HEALY, P. and ARMITT, J. D. G. (1980) Proc. 16th Conv. Inst Brewing (Australia and New Zealand 

Section), Sydney, p. 91. 

HERRMANN, H., KANTELBERG, B. and WIESNER, R. (1998) Brauwelt Internat., 16 (1), 44. 

HIND, H. L. (1950) Brewing Science and Practice, Vol. II. Brewing Processes. London, Chapman and Hall, 

pp. 507±1020. 

ISOE, A., KANAGAWA, K., ONO, M., NAKATANI, K. and NISHIGAKI, M. (1991) Proc. 23rd Congr. Eur. Brew. 

Conv., Lisbon, p. 697. 

KROTTENTHALER, M., ZUÈ RCHER, J., SCHNEIDER, J., BACK, W. and WEISSER, H. (1999) Proc. 27th Congr. 


KUNZE, W. (1996) Technology Brewing and Malting. (Wainwright, T. Transl.) Berlin, VLB, 726 pp. 

LANGENHAN, R. (1992) Brew. Distill. Internat., 23 (5), 16. 

MEISEL, D. (1997) A Practical Guide to Good Lager Brewing Practice. Inst. Brewing, (Central and 

Southern African Sect.). 

MENGER, H.-J., SALZGEBER, G. and PIEPER, H. J. (2000a) Brauwelt Internat., 18 (1), 54. 

MENGER, H.-J., MIROLL, F., FORCH, M., BIURRUN, R., SCHWILL-MIEDANER, A., HERRMAN, J. and RAPP, T. 

(2000b) Brauwelt Internat., 18 (2), 120. 

NARZISS, L. (1992) Die Bierbrauerei.(7nt Auflage) Bd. II. Die Technologie der WuÈrzebereitung. Stuttgart. 

Ferdinand Enke Verlag, 402 pp. 

NICOL, S. O. and ANDREWS, J. M. H. (1996) Ferment, 9 (3), 145. 

REHBERGER, A. J. and LUTHER, G. E. (1994) in Handbook of Brewing (Hardwick, W. A. ed.). New York, 

Marcel Dekker Inc., p. 247. 

STAUFFER, J. (1974) MBAA Tech. Quart., 11, (4), 240. 

STOSCHECK, W. (1988) MBAA Tech. Quart., 25 (2), 108. 

STUBITS, M., TENG, J. and PEREIRA, J. (1986) J. Amer. Soc. Brew. Chem., 44 (1), 12. 

SUGDEN, T. D., WEBB, C., BYRNE, H., VAN WAESBERGHE, J. and WULFF, T. (1999) Milling. E. B. C. 

Handbook of Good Practice. NuÈrnberg. Hans Carl, 102 pp. 

WILKINSON, R. (2001) Brewers' Guard., 130, (4), 29. 

WILKINSON, R. (2003) Brewers' Guard., 132, (1), 26. 


6 

Mashing technology 


Mashing is the process by which sweet wort is prepared. It involves `mashing in', the 

mixing of the milled grist and the brewing liquor at the correct temperature and in the 

correct proportions to obtain the mash. After aperiod, with or without temperature 

changes, during which the necessary biochemical changes occur, the liquid `sweet wort', 

which contains the extract, is separated from the residual solids, the `spent grains' or 

`draff'. Some extract remains in the draff, and as much of this as possible is recovered by 

`sparging', washing the grains with hot brewing liquor. 

In traditional brewing, as practised in homes or small inns, hot water was placed in a 

wooden tub or tun, and the grist (malt that had been ground between millstones) was 

mixed and mashed in by stirring or rowing with arake, paddle or òar' (Fig. 6.1). No 

reliable means of measuring temperatures was available. In one method, which gave rise 

to the `classical' British infusion system, the water temperature was guessed to be 

suitable by feel or by how clearly the brewer's face was reflected in the water. After a 

period a basket was pushed into the mash and wort that seeped into it was ladled into a 

receiver, in readiness for boiling with hops or other flavouring herbs. When wort 

recovery became difficult more hot water was mixed into the mash (re-mashing) and 

another, weaker wort was recovered. This was repeated until the worts were too weak to 

be worth collecting. In later times wort was collected from mashes using primitive mash 

tuns, in which the wort drained from the mash through a perforated `strainer' in the base 

of the tun. The structures of old (approx. 200 years), relatively small British country 

house breweries are documented (Sambrook, 1996). 

In an alternative method, which gave rise to the classical mainland European 

decoction mashing system, traditionally used for brewing lager beers, the mash was made 

with slightly warm water. At intervals a `decoction' was carried out, that is, a proportion 

of the mash, perhaps one-third, was withdrawn and slowly raised to boiling in the copper 

that would later be used for boiling the wort. The hot mash was then transferred back to 

the `main mash', and was mixed in. In this way the temperature of the whole mash was 

increased. Repeated decoctions increased the mash temperature in steps, an approach that 


minimized the risk of overheating and premature total enzyme destruction. In more 

modern variations of this system the temperatures achieved (now exactly controlled) are 

optimal for various enzyme-catalysed processes in the mash and allow comparatively 

under-modified malts to be mashed successfully (Chapter 4). Raw cereal adjuncts may 

conveniently be cooked in decoction vessels. In contrast, the traditional infusion system 

of mashing requires well-modified malts and not more than about 20% of unmalted 

adjuncts (Chapter 4). 

The equipment used in breweries has been progressively refined and there has been a 

convergence in the practices of ale and lager brewers. The motives for these alterations 

are chiefly economic. It is often desirable to maximize productivity (`throughput', the 

numberofbrewscompletedevery24h)andtorecoverasmuchextractasiseconomically 

worthwhile from agiven grist (Chapter 18). It is also necessary to reproducibly recover a 

certain volume of wort having exactly the characteristics needed to make aparticular 

beer. At the same time energy and water usage must be minimized and so must the 

production of effluents. At the present time there are breweries operating with many 

different kinds of `traditional' and `modern' equipment. The more common types will be 

described. Sometimes old types of plant have been retained, despite some inconvenience 

or poorer efficiency, because a newer system has not been able to produce abeer 

matching that produced by the old system. In the case of smaller breweries older kinds of 

equipment may be retained because of its simplicity, or because replacement is not 

economic. While older equipment is often made of attractive polished wood and copper, 

in newer equipment these materials have been largely replaced by the cheaper, more 

deterioration-resistant stainless steels (Chapter 10). 

6.2 Mashing in 

Fig. 6.1 Awooden mashing rake or oar. 

Mashing in, the process of mixing the mashing liquor and the grist, is critical. The 

proportions of liquor to grist and the temperature of the mixture must both be correct and 

the grist must be evenly mixed in the mash, with no clumping or `balling', which reduces 

extract recovery, and no segregation of the grist components. In traditional infusion 

mashing,whereitisnoteasytoadjustthemashtemperature,thevalueinitiallyattainedis 


Slide valves 

Hot liquor 

Drive 

Grist case 

‘Worm’ or 

screw 

Beater 

rods 

Swivel mount 

To mash 

Fig. 6.2 Steel's mashing machine, used in traditional British infusion mashing. In more modern 

versions the rate of grist flow is regulated mechanically and the slide valves prevent steam from the 

mash travelling into the grist case. 

critical. Historically, grist was poured into attemperated water in the mash tun or `tub' 

and was mixed in by `rowing', alaborious process that often resulted in inadequate 

mixing and balls of unmixed grist. Mechanical means of mixing mashes were necessary. 

Mash tun rakes were invented by Matterface in 1807 (see below; Sykes and Ling, 1907). 

This kind of equipment remained widely in use in the UK until the 1960s. It is now rare. 

Several èxternal' mashing machines were invented and, in various forms, some of these 

are still in use. 

The Steel's masher, introduced in 1853, consists of ahorizontal tube about 46cm 

(18in.) in diameter and was once typical of ale breweries (Fig. 6.2). Grist from the gristcaseisdeliveredtooneendviaaverticaltubeatacontrolledrate.Slidevalvesareusedto 

prevent vapour from the mashrising intothe grist case.Thegrist issprayedwith mashing 

liquor and the wetted material is driven along the tube by ascrew conveyor and is then 

mixed by aseries of short beater-rods mounted on the same shaft as the conveyor screw. 

The mashed material is then dropped from aspout into the vessel below. This device is 

capableofmixingthethickmashestypical oftraditionalinfusionmashing(e.g.1.6±3.2hl 

liquor/100kg grist; 9.9±19.9imp. brl/ton). With this system it is impossible to prevent 

oxygen uptake. Sometimes the masher is mounted so that it can be swivelled sideways 

and so can deliver into either of two vessels. 

Mash hydrators, or `pre-mashers', are designed for making the thinner mashes used in 

decoction mashing or temperature-programmed infusion mashing (3.3±5hl liquor/100kg 

grist; 20.5±31.0imp. brl/ton), which must be stirred and pumped between vessels. In 

general, light beers are made with more dilute mashes than those used for dark beers. In 

mash hydrators grist, falling down a tube at a controlled rate, meets a spray of 

attemperated water flowing at acontrolled rate from aperforated central tube or from a 

surrounding casing (Figs 6.3, 6.4). In other increasingly common devices the water is 

injected tangentially creating avortex, which rapidly mixes with the grist (Fig. 6.5). This 

device contains dust and minimizes oxygen pick up. It is desirable that the mash from the 

hydratordoesnotfallintothemash,butisdirected ontothesideofthemashingvessel, or 

is pumped into the base, so that it slides down to the mash with the minimum uptake of 


Thermometer 

Liquor 

inlet 

Slide 

oxygen (Wilkinson, 2003). None of these devices has moving parts to wear out. Asprayball 

should always be mounted above the liquor inlet to facilitate `cleaning in-place', 

CIP. The grist from the hydrator is not uniformly hydrated. Hydration is completed 

during stirring in the mashing vessel. 

In addition to the methods mentioned, wet milling coincides with mashing in and so 

with this approach no grist case is involved (Chapter 5). In some plants the material from 

the pre-masher enters an inclined disc mixing vessel. This consists of a cylinder with 

rounded ends mounted on its side and having a longitudinal shaft mounted about a third 

Grist 

Window 

To mashing vessel 

Fig. 6.3 Premasher or Maitland grist hydrator, of a traditional mainland European type, used in 

conjunction with decoction mashing (after Narziss, 1992). Note that the grist falls past a perforated 

tube, which delivers the mashing liquor. The mixture pushes open a counter-balanced door and falls 

into the vessel. 

Grist 

Mash 

Water 

Fig. 6.4 An alternative type of mash hydrator (after Rehberger and Luther, 1994). 


Grist 

Mash 

of the way up from the bottom. The shaft carries aseries of discs inclined at angles to the 

shaft which rotates. The discs mix the mash very gently and thoroughly (Kunze, 1996). 

In most plants it is usual for the pre-hydrated grist to be transferred into amashing 

vessel, where mixing occurs. To minimize oxidative changes in the mash or wort, the 

mash may be gently pumped into the vessel at the side or up through the base, so 

minimizing turbulence and the uptake of oxygen. Degassing the mashing liquor, flushing 

the grist with an inert gas such as nitrogen or carbon dioxide and filling the base of the 

vessel with an inert gas to displace air and so limit oxygen pick-up may also be used 

(Yamaguchi et al., 1997). 

In many breweries the older names for critical temperatures are retained (Hind, 1940). 

Thus liquor heat and striking heat are terms for the temperature of the mashing liquor in 

the hot water tank and at the mashing machine respectively. The initial heat is the 

temperature of the freshly mixed mash, sparge heat is the temperature of the sparging 

liquor and the tap heat is the temperature of the wort as it is drawn off. In infusion 

mashing in modern mash tuns the opportunities for adjusting the mash temperature are 

limited and it is essential to achieve the correct initial heat (temperature) when mashing 

in. In modern plant this may be achieved automatically by varying the temperature of the 

mashing liquor sothat the correct initial heat isattained. In older plant this isachieved by 

askilled operator's judgement, guided initially by calculation and later by experience. 

When malt is mixed with water heat is generated and this slaking heat or heat of 

hydration is less for malts with higher moisture contents (Table 6.1). So, for example, a 

malt with amoisture content of 2%, mashed in aparticular way, may give atemperature 

rise of about 4.8ëC (8.6ëF), while amalt with amoisture content of 6% would give a 

temperature rise of 2.6ëC (4.6ëF) when mashed in the same way. The initial heat of the 

mash can be calculated from the formula 

I=[St +RT/S+R] +[0.5H/S+R] 

Water 

Fig. 6.5 Avortex mash mixer (after Rehberger and Luther, 1994). The grist falls into aswirling 

tube of liquor from that tangentially injected into the outer chamber. With these, and other prehydrators, 

it is desirable to have a mixing chamber through which the grist must pass before it 

leaves the unit. 


Table 6.1 The specific heats and slaking heats of malt at different moisture contents. Other malts 

may have slightly different values. Data recalculated, by interpolation, from the data of (a) Brown 

(1910) and (b) Hopkins and Carter (1933) 

Moisture Specific Slaking heat at 65 ëC (150 ëF) 

(%) heat (a) g.cal/ëF g.cal/ëC 

where Sˆspecific heat of the malt, tˆthe temperature of the malt, Rˆthe weight of 

water, relative to the unit weight of the malt, Tis the temperature of the water, His the 

slaking heat of the malt expressed in the correct units and Iis the initial temperature of 

the mash. Others prefer to use H(rather than 0.5H) and make allowances for heat losses, 

determined by trial and error. The final temperature of amash warmed by ùnderletting', 

that is the addition of hot liquor to the mash, can be calculated from the formula 

Final temperature =[M(S+R) +QT] /(S+R+Q) 

where Mˆtemperature of the mash at the time of underletting, Qis the quantity of water 

used in the underlet, Tˆtemperature of the underlet liquor. The other symbols are as 

used before. These calculations can be used only for guidance. They cannot give exact 

results because no allowance is made for heat losses from the system, and these will vary 

with the temperature of the brewhouse. 

6.3 The mash tun 

(a) (b) (a) (b) 

0 0.38 28.0 33.4 15.6 18.6 

1 0.38 24.7 28.7 13.7 15.9 

2 0.39 21.5 24.9 11.9 13.8 

4 0.40 15.8 18.8 8.8 10.4 

6 0.41 11.5 14.2 6.4 7.9 

8 0.42 8.5 11.7 4.7 6.5 

The mash tun (`kieve' in Ireland) is the simplest device for mashing and preparing sweet 

wort.Mashconversionandwortseparationfromthespentgrainstakeplaceinonevessel, 

and so mash tuns should be distinguished from the mash mixing and incubation vessels 

which are used to carry out the mash conversion step only, wort collection being carried 

out in aseparate device, usually alauter tun or amash filter. 

6.3.1 Construction 

Mash tuns are circular in cross-section and vary greatly in size, but they are usually 2.0± 

2.5m(approx. 6±8ft.) in depth. Originally the tuns were of wood and, for many years 

mostwereequipped withrakingmachinery(Fig.6.6). Wood ishard tocleanandhasonly 

alimited life. Increasingly, mash tuns were made of iron or copper or, more recently, 

stainless steel. The properties of some materials used in the construction of brewing 

vesselsareoutlinedinChapter10.Theyareinsulatedatthebaseandaroundthesidesand 

areoftencladwithwood.Originallythetunswereopen,butnowtheyareusuallycovered 

with metal domes, equipped with inspection ports and lights. 

Covering atun slows heat loss and prevents the water vapour from the mash spreading 

through the brewing room (Fig. 6.7). Sometimes the vapours are carried away in apipe 


Rake 

Axis, carrying rakes, 

which rotates and 

moves around the tun 

False, gun-metal 

plates 

True base of 

tun (wood) 

Spent grains 

discharge 

Rotating 

sparge arm 

Driving 

machinery 

Hot 

liquor 

Vertical 

rotating drive 

Wooden 

wall of tun 

Fig. 6.6 An old pattern of mash tun as used in about 1880 (various sources). Some few of this 

general type are still in use. 

fitted to the dome. The true base of the tun is covered by a`false bottom' or deck. This 

consists of interlocking slotted (or drilled in some small tuns) metal plates of gun-metal 

or stainless steel or of stainless steel wedge wire (Figs 6.8, 6.9) mounted on short legs 

about 5±7.6cm (2±3in.) above the true base. The plates interlock in aunique pattern and 

can be lifted for cleaning or repairs. The slots give afree area of about 10±12% of the 

false bottom, while wedge wire gives the same or ahigher value (up to 22%). The slots 

are typically 0.7±1.0mm (0.028±0.039in.) wide at the top and widen out below to 

facilitate cleaning and reduce the chances of fragments of grist wedging in the spaces. In 

the past all cleaning was by hand, involving the regular lifting of the plates of the false 

bottom and later re-assembling them, but now sprays may be fitted to allow CIP both 

above and below the deck. One, or more, large holes, fitted with valves, are in the false 

bottom. These are opened at the end of mashing to allow the removal of the spent grains 

and some rinse water. 

In the past, grain removal was by hand but now, except in small breweries where 

manual removal is still used, grains are swept out of the discharge ports by horizontally 

rotating arms. The grains are collected and transferred by screw conveyor or compressed 

air to the collection silo for removal to farms or animal compounders. An alternative 

procedure of slurrying the grains with water and pumping the mixture toacollection tank 

has been discontinued because wetting the spent grains reduces their value and the use of 

more water and the production of effluent with ahigh BOD increases costs. Few English 

breweries have retained rotating rakes. The machinery is clumsy and it is suspected of 

allowing channelling during sparging, with aconsequent reduction in extract recovery. 

On the other hand rakes facilitate mixing the mash with underlet water, allowing 

controlled increases in temperature. 

Some few mash tuns have been equipped with knives for `cutting' the mash and for 

discharging the grains, as is done in many lauter tuns (Section 6.5). Originally drained 


Device for 

adjusting 

hydrostatic 

head 

Wort to 

underback 

Grist case 

Steel’s masher 

Run-off pipe 


Cork 

insulation 

on true base 

Hot liquor 

to sparge Bearings for 

sparge arm 

Motor for 

driving grains 

discharge gear 

Sliding inspection 

hatches 

Fig. 6.7 A current pattern of mash tun (after Briggs et al., 1981). 

Grains 

discharge 

pipe 

Sparge arm 

Wood and cork 

insulation 

Mash 

Floor 

Grains 

discharge 

arm 

False bottom

1 in. 1 cm Scale 

Fig.6.8 Slotsinplatesthatformthefalsebottomsofmashtuns(afterBriggsetal.,1981).(a)Plan 

view of the upper surface. (b) Vertical section across the slots. (c) Plan view of the lower surface. 

(d) Avertical section, at right-angles to (b), along two slots. 

Fig. 6.9 Apiece of wedge wire (after various sources). 

grains were remashed withfreshhot watertorecoverentrainedextract. Now`sparging' is 

employed. Hot water is sprinkled onto the surface of the mash from centrally mounted, 

rotating perforated tubes, the sparge arms. These may be mechanically driven, but more 

usually they are driven by their reaction to the streams of water coming from the 

perforations, which are at an angle to the vertical (Fig. 6.7). Because it is desirable to 

apply the sparge liquor equally to all areas of the mash the perforations are more widely 

spaced towards the centre of the tun and more closely spaced towards the periphery. 

During wort collection the liquid is driven down through the bed of grain by the pressure 

difference between the liquid at the top of the mash and the pressure below the false 

bottom. This pressure difference is measured. To control this pressure difference, and 

hence the `suction' on the grain bed and its compression, various devices, which act as 

weirs, may be used including aswinging, inverted U-tube (a Valentine tube), or adevice 

equivalent to it (Fig. 6.7). Alternatively the rate of collection can be regulated with a 

pump, and the rate of sparging can be set to equal the rate of run-off using two matched 

pumps. Excessive pressure can force the goods down onto the plates and cause blockages 

or so compress the bed that the resistance to the flow of the liquid through the bed is high 

and run-off is slowed. The result is aset mash. 


(a) 

(b) 

(c) 

(d)

To obtain an even run-off and sparge it is desirable to have multiple wort collection 

pipes,connectedtothetruebottomofthetun,distributedinanevenpattern.Theobjectis 

to have each pipe draining the same area beneath the deck, say 1.9±2.3m 2 (20±25ft. 2 ).In 

the traditional arrangement these discharge the wort via taps and swan-necked tubes into 

an open collecting trough. This allows the clarity, the gravity and the rate of flow of the 

wort from each tap to be checked. By manipulating the taps the rate of flow can be 

regulated and balanced. Differences in the quality of the wort from different taps 

indicates problems with the quality of the grain bed or the sparge. 

6.3.2 Mash tun operations 

Before use the clean mash tun is preheated to the mashing temperature with steam or hot 

water.Alltapsareclosed, thetemperatureofthemashingliquor ischecked andsoarethe 

contents of the grist case. Some liquor is admitted into the base of the tun to drive air out 

from beneath the false bottom and to cover the plates with alayer of water 2±5cm 

(roughly1±2in.)deep.ThiscushionsthemashasitisdroppedinfromtheSteel'sorother 

premasher, helps the mash to spread across the false base and reduces the chance of the 

slots being blocked. Mashing in may take 20±30min. In modern mash tuns the mash 

entrains air and most of the goods float (Harris, 1968, 1971). In older tuns the mash was 

raked for afew minutes to mix the contents. The depth of the mash is typically 0.91± 

1.5m(approx. 3±5ft.), but depths of up to 2.74m(9ft.) have been used. One advantage 

of mash tuns is that they can be used with mashes of widely varying volumes. 

The mash then begins its stand, normally aperiod of 1.25±2.5h, to allow all necessary 

biochemical processes to take place. The temperature of the mash during the stand is not 

normally altered, but this is acomparatively recent development. Formerly moderate 

upward temperature adjustments were usual (Chapter 4; Hind, 1940; Sykes and Ling, 

1907).Theapplicationofheattoathickmashthroughaheaterinthewallsandbaseofan 

unstirredmashtunisnotefficient.Atvarioustimesheatincreaseswerebroughtaboutwith 

steam-heated coils beneath the false bottom or by direct steam injection, by cycling wort 

fromthebaseofthetuntothetopofthemashviaanexternalheater,bya`steamplough'(a 

device heated by steam or water which rotated with the machinery near to the base of the 

tun) or by underletting (sometimes with simultaneous sparging). With underletting, the 

mostusualmethod,thecalculatedvolumeofhotliquor,atanappropriatetemperature,was 

slowly let into the base of the tun, beneath the plates and raising the mash bed without 

disruptingit,thenthemashwasmixed,soitstemperaturewasincreasedanditwasdiluted. 

Sometimes wort run off becomes slow or even ceases. This is liable to occur if poor 

quality malt has been used, if the grist has been milled too finely, if the grist contains a 

high proportion of adjuncts, or if the operator has tried to draw off wort too quickly and 

has pulled the mash down onto the plates and caused the bed to become compressed. 

Such a`set mash' may be cleared by using an underlet to lift the goods off the plates. Set 

mashes caused by poor quality grist components are made less likely by adding an 

enzyme preparation containing -glucanase, cellulase and pentosanase activities. 

Drainage from the mash is dependent on it having an open structure and this, in turn, 

requires that the malt husk fraction be damaged as little as possible during milling 

(Chapter 5). In the past some brewers used aproportion of `husky' malt or even added 

chopped straw or oat husks to òpen-out' the mash and aid drainage. The addition of such 

materials is undesirable as unwanted flavours can be conferred to the beer. 

When the stand, with or without an underlet, is completed the first wort is run off. The 

first runnings may be a little turbid, and these are usually pumped back to the top of the 


mashandrecirculated.Whenthewortisperfectlyclear(i.e.is`runningbright')theflowis 

diverted and the liquid is either collected in aholding vessel, an underback, where it is 

maintained at 71±82ëC (160±180ëF) to prevent the multiplication of contaminating 

microbes, or it is transferred directly to the copper. When the `first worts' have been 

collected and the mash has settled, but before the surface of the grains has become dry, 

sparging begins, the hot water (usually at about 78ëC, 172.4ëF) being sprinkled on from 

therotatingspargearms.Thespargingratematchestherateofwortrunoff,andtheliquor 

displaces the wort downward and through the plates and leaches extract from the grist 

particles.Thehightemperaturefacilitatesextractrecoverybecauseitreducestheviscosity 

ofthewortandsofacilitatesrunoffanditacceleratestheleachingofextractfromthegrist. 

Afterthefirstwortshavebeencollectedthespecificgravityofthecollectedwortbegins 

to fall. The decline continues until achosen gravity (often SG 1003±1005, 0.78±1.3ëP) is 

reached,whencollectionceases.Thegoodsarethenallowedtodrain.Thedrainingsmaybe 

senttowaste,withaconsequentlossofextractandatthecostofaneffluentcharge,orthey 

may be stored hot for ashort period and then be mixed into asubsequent mash, when the 

extractisrecovered.However,thequalityoftheextractinthedrainingsisinferiortothatin 

themainwort.Spargingoftenlasts 4±5h.The sparge liquorshouldnotcontain `temporary 

hardness' and should contain an adequate level of calcium ions (Chapter 4). At higher pH 

values more tannins and undesirable materials are extracted, leading to reductions in beer 

quality. About 5.35hl of sparge liquor may be used for 100kg malt (33imp. brl/ton) so in 

allatotalofabout8.1hlofliquorareusedforeach100kgmalt(50.3imp.brl/ton).Someof 

this liquor, perhaps 2%, remains in the spent grains. The spent grains have amoisture 

content of about 80%. These are discharged and the tun is cleaned before the next mash. 

The entire cycle time with amash tun is often 5±9h, but in one famous brewery the time 

was 18hand mashing in began at midnight. So mash tuns generate clear, high-quality ale 

worts which, because the mashes are thick, can produce high-gravity worts but by modern 

standards their turn-round times are slow. Oxygen exclusion is not feasible but for 

traditional ales it is not necessary. The thick bed of grains allows the production of very 

brightwortsbutalsoensuresthattherunofftimeiscomparativelylong.Extractrecoveries 

(relative to the laboratory extract) of 98% have been claimed, and 96±97% is usual but in 

some small breweries the recovery may be as little as 85%. Other disadvantages of mash 

tuns are their inflexibility with regard to mashing temperatures, their requirement for a 

coarsely ground grist (with aconsequent reduction in extract recovery), the need for well 

modified malts and the difficulty in using wet, cooked adjuncts. 

6.4 Mashing vessels for decoction, double mashing and 

temperature-programmed infusion mashing systems 

In these mashing systems mashes are relatively thin (3.3±5hl liquor/100kg grist; 20.48± 

31imp. brl/ton) and so they can be stirred and pumped between vessels. While, in 

principle, the types of vessels have remained constant for aconsiderable time there have 

been significant improvements in design and in the designs of the wort separation 

devices, the lauter tuns and mash filters, used in conjunction with them. 

6.4.1 Decoction and double mashing 

Inthetraditionaldecoctionmashingsystemgristismashedinusingahydratorandfallsinto 

amash-mixingvessel(Fig.6.10).Tominimizeoxygenpick-upthemashmayberundown 


Registering 

thermometer 

Propeller 

Worm 

gear 

Throttle 

valve 

Mashing 

valve 

Pipe for 

ground malt 

Pre-masher 

Fig. 6.10 A section of an older pattern of a decoction mash-mixing vessel (after Hind, 1940). 

thesideofthevesselorpumpedintothebase.AtthisstagepHadjustmentmaybeachieved 

bytheadditionofafood-grademineralacidorbiologically preparedlacticacid.Acidmalt 

may have been included in the grist to achieve the same effect. Mashes are stirred, usually 

by apropeller in the base of the circular vessel. However, vessels with rectangular cross 

sections have been used, the contents being mixed by stirrers mounted on long shafts 

extending downwards from above. The older vessels are often of copper, have adouble 

walled steam heating jacket and are well insulated. Heating may also be by internal steam 

heated coils or direct steam injection. Low-pressure steam (15 bar, 171ëC, 340ëF) is 

preferred to superheated water as the heating agent (Wilkinson, 2003). The vessels are 

coveredwithcopperdomes thatcarry centrallyplacedflues thatcarry awaysteam.Ineach 

decoction some of the mash is pumped to asmaller vessel, (the mash cooker, copper or 

kettle) where it is heated to boiling and then, after aboil, is transferred back to the mash 

mixing vessel (Chapter 4). These vessels can also be used for cooking adjuncts. 

Mash cookers are similar in construction to mash-mixing vessels, except that often 

stirrers are more powerful and the vessels have relatively greater steam heated areas, 

since their contents must be heated to boiling. An internally mounted heater may 

supplement the heating supplied by a steam jacket. In some small breweries the mash 

cooker also acts as the hop-boiling copper (kettle). Mashing vessels have specified 

`duties', for example they must be capable of heating the vessel contents linearly at 0.5 or 

1 ëC/min. Some decoction mashing breweries, and all breweries using double mashing, 


have separate cookers for cooking raw cereal grits. Some may be heated under pressure, 

and the higher temperatures achieved are adistinct advantage if rice grits are being used. 

Early cookers were often horizontally mounted, cylindrical vessels closed by rounded 

ends, heated by asteam jacket and equipped with ahelical ribbon stirrer supported by 

arms extending from acentral rotating shaft (Scott, 1967). Newer cookers are generally 

vertically mounted cylinders closed with arounded base and cover, heated by asteam 

jacket and, sometimes, direct steam injection. 

6.4.2 Temperature-programmed infusion mashing 

Temperature-programmed infusion mashing is less costly than decoction mashing in 

terms of number of vessels and energy used, decoction mashing needing 20±50% more 

energy (steam). Estimates of typical energy use levels are, for programmed infusion 

mashes 8.5MJ/hl wort, for single decoction mashing 11.0MJ/hl and for double decoction 

mashing, 11.6MJ/hl. Mashing in amash mixing vessel (with or without aprogrammed 

rise in temperature) and separating the wort in alauter tun or mash filter is amore rapid 

way of obtaining wort than using amash tun. The greater speed of processing is achieved 

because in two-vessel systems asecond mash may be in preparation while the first is 

being lautered or filtered. It is not surprising that brewing practices in mainland Europe 

and the UK are converging on the use of infusion mash mixing vessels and lauter tuns or 

mash filters. In addition there has been aconvergence in vessel designs, incorporating 

refinements that confer increased heating efficiency, ease of use, reduction in shear by 

stirring, flexibility, and reductions in oxygen uptake. 

Modern mash mixing vessels, mash cookers, cereal cookers and temperatureprogrammed 

mash mixing vessels are very similar. They often have higher height/width 

ratiosthanoldervesselsandcommonlyhavetwoorthreeheatingzones,allowingdifferent 

volumesofmashtobeheatedefficiently(BarnesandAndrews,1998;BuÈhleretal.,1995; 

Herrmann, 1998; Kunze, 1996; McFarlane 1993; Wilkinson, 2001, 2003; Wilkinson and 

Andrews, 1996; Figs 6.11, 6.12). Probably in all newer vessels stainless steel is the only 

metal in contact with the mash. Steam jackets were all double walled and later the walls 

were`dimpled'.Nowsteamisoftensuppliedtotheheatingsurfacesinsemi-circularpipes 

welded onto the heat exchange surfaces (cf. Chapter 10). The turbulence in the steam in 

the pipes improves its heat transfer efficiency and so the heating performance and, in the 

event of sudden cooling or other loss of pressure, the heating system is not so liable to 

collapse under vacuum as are double-walled units. Steam heating is applied to the base of a 

vessel and to the sides. The side and base heating is applied in zones that are operated 

separately so only the zone(s) covered with mash are heated. This arrangement allows 

mashes having different volumes to be processed in one vessel. Sometimes, when direct 

steam injection is used for heating, the tangential injection of steam helps to mix the mash. 

Burning on a heat exchange surface cannot occur but steam condensation slightly dilutes 

the mash, local òverheating', with enzyme destruction, must be a risk and the steam must 

be `pure' and carry no odorous or other contaminants. 

Stirring a mash or transferring it between vessels can create shear which damages the 

grist, breaking up the particles, extracting more -glucan and perhaps altering the structure 

of the particles and making them more gelatinous. One consequence of shear-induced 

damage is that wort separation is slowed. Modern stirrers create less shear than the older 

types in which propellers with relatively small diameters turned rapidly. Newer stirrers are 

larger, extend across about 85% of the vessel diameter, and move more slowly. They cause 

less shear and create a minimal surface vortex, mix the mash well, and sweep the mash 


(a) 

Condensate 

Condensate 

(b) 

Pre-masher 

Liquor 

Condensate 

Grist 

CIP 

Vapour 

stack 

over the heat exchange surfaces causing local turbulence so optimizing heat exchange. 

Scaling and local overheating are minimized and so there is little consequent enzyme 

destruction and burning on to the heat exchange surfaces (Figs 6.11, 6.12). In some cases 

the need for baffles (used to prevent the mash rotating with the stirrer, but inevitably 

causingshear)is overcomebymountingthestirreroff-centre(say5ëfrom thevertical)in 

the asymmetrical, dished base of the vessel (Scott, 1967; Wilkinson and Andrews, 1996). 

Shear during the transfer of mash between vessels in minimized by using gentle 

gravity feed where possible or carefully rated, slow-running pumps with wide `throats' 

and wide piping, avoiding bends where possible using gentle curved bends where 

CIP 

Outlet 

Agitator drive 

Inspection 

port 

Rotation 

tip speed 

3.8 m/s 

maximum 

Maximum 

level 

Steam 

Steam 

Steam 

Fig. 6.11 (a) Asection of modern mash-mixing vessel. Note the inclined base (after Wilkinson 

and Andrews, 1996). (b) A plan view of the interior of the base of a mash-mixing vessel, showing 

the offset stirrer. 


Fig. 6.12 An alternative pattern of mash-mixing vessel (after Narziss and others, 1992). 

essential,andminimizingthelengthsofpipework.Whileolderstirrerscreatedsubstantial 

vortices, which drew air down into the mash, and so favoured oxidations, new stirrers do 

not, and often they run at ahigh speed only during mashing in and during heating periods 

but at aslower speed at other times. Freshly made mash is often delivered to the base of 

the receiving vessel and rises in the vessel with little turbulence, rather than being 

dropped from above. Both of these refinements reduce oxygen uptake and so reduce 

oxidationinthemash.Itislesscostlytoincreasethetemperatureofamashbyaddinghot 

water rather than by heating with steam (Sommer, 1986). Thus at the lower temperature 

thecooler,thickermashfavoursproteolysiswhileinthehotter,thinnermash,obtainedby 

the addition of hot water, amylolysis is favoured. It is amusing that this `new' proposal 

follows the older British practice of underletting with hot water in raked mash tuns. New 

mashing vessels are now all equipped with the plumbing needed for automated cleaning 

(CIP). CIP has usually been retro-fitted to older equipment. The steam flues of some 

mashing vessels, especially mash coppers, may be equipped with economizers to recover 

heat, as is the case with hop coppers (Chapter 10). 

6.5 Lauter tuns 

Lauter tuns have been used for many years to separate worts from decoction mashes. 

Now they are also used with temperature-programmed infusion mashing and double 

mashing systems. In the UK, in a few instances, lauter tuns have been operated as mash 

tuns. In recent years the need to accelerate brewing processes, to minimize oxygen uptake 


and to improve extract recoveries while obtaining spent grains with lower moisture 

contents, and to produce equipment to compete with mash filters (Section 6.6) has led to 

many refinements in the designs and methods of operation of lauter tuns and in 

improvements in their performances (Andrews and Wilkinson, 1996; De Clerck, 1957; 

Herrmann et al., 1990; Kunze, 1996; Lenz, 1989; Narziss, 1992; Wilkinson, 1993, 2003). 

In general layout lauter tuns resemble mash tuns and are usually circular in cross-section 

(Fig. 6.13), although rectangular lauter tuns have been used. They are covered with 

domes connected to chimneys that carry away steam, and they are heavily insulated. 

There are significant differences between mash tuns and lauter tuns. In particular the 

gristsusedinthemashesthatarefilteredinlautersaremorefinelyground,themashesare 

more dilute, `thinner', and because more rapid wort filtration is desired the bed depth is 

comparatively shallow, so it provides aless good filter bed. For agiven capacity, alauter 

tun may have a50% greater diameter than amash tun. The bed depths used are very 

variable, for example, 35cm (13.8in.), and 46cm (18.1in.). Maximum bed loadings may 

beinthe range339±153kg/m 2 for cycles allowing6±12 brews/24h(Wilkinson,2003). In 

the older system mash was dropped into the lauter tun from above through apipe curved 

to minimize the impact on the plates and the layer of water covering them. In modern 

practice the mash is allowed to enter gently through one or more ports either in the sides 

of the tun or up through the base, to reduce turbulence and the entrainment of air. 

Possiblythemashmaybeloadedunderablanketofcarbondioxidegastoexclude airand 

atop-pressure of carbon dioxide can accelerate wort run off (Stippler et al., 1994). 

During transfer to the lauter shear should be minimized, for example, by using aslowrunning 

pump with an open-throated impeller, giving atransfer velocity of less than 

1.3m/s (4.26ft./s). More entry ports enable loading to be carried out faster, so saving 

time. The loading that may be used varies with the way in which the grist has been 

milled. For dry milled grist the loading on the plates may be 160±175kg/m 2 , for 

conditioned and milled grist 170±210kg/m 2 ,while for steep-conditioned and milled grist 

the values are 200±280 or even 310kg/m 2 (Kunze, 1996; Lenz, 1989). 

The false bottoms were originally made of brass or gun-metal plates, but now are of 

stainless steel or are substituted by stainless steel, profiled wedge wire (Figs 6.8, 6.9). 

These plates may be lifted for cleaning, a laborious process that, as with mash tuns, had to 

be carried out after every run but now, with modern CIP installations, may need to be 

carried out only weekly or even less frequently. While some plates may have drilled 

circular perforations, most have slots typically 30±40 mm (about 1.2±1.6 in.) long and 

0.6±0.7 mm (0.024±0.028 in.) wide. As with mash tun plates, these slots widen out below. 

There may be 2500±3000 slots/m 2 of plate area. Single milled plates may have a free area 

of 6±8%, double milled plates up to 12% and wedge wire `plates' of 18% and even up to 

25%. The plates provide little resistance to the flow of wort, relative to the resistance 

provided by the bed of grain, and run off rates are the same with plates of 12 or 18% free 

area (Lenz, 1989). Run off times may be 1±2 h. 

The true bottom of a lauter tun may be flat or consist of a series of depressions or 

`valleys' that assist the drainage of wort into the collecting tubes while reducing the 

under-deck volume, perhaps by 30%. This reduces the volume of hot water needed to 

cover the plates, the amount of washing water required and hence the volume of effluent. 

The depth of the under-deck space is usually about 20 mm (0.70 in.). The deeper this 

space the more water is used in driving out the air, covering the plates and heating the 

unit, and so the more the initial wort is diluted. The wort collection tubes in the true base 

may be spaced 1/1.2±1.5 m 2 . The tubes are now usually joined into the base through 

conical extensions that reduce the local flow rate across the under-deck space and èven 


Insulation 

Spent 

grains 

discharge 

Spent grains 

receiver 

Mash 

supply 

pipe 

Supports 

CIP 

Sparging 

spray 

Wort 

collection 

tank 

Vapour 

stack 

Pillar supporting 

rakes 

Motor and machinery 

for rotating, raising and 

lowering the lauter arms 

Lighting 

Wort pump 

Lauter arms with 

attached ‘zig-zag’ 

knives 

Supporting pillar 

Inspection 

port 

Run-off pipes 

Wort 

collection 

line 

Fig. 6.13 A section of a modern lauter tun equipped with fixed sparging nozzles and zig-zag, chevron-shaped knives, 

the outer ones of which are the double-shoe type (after Herrmann et al., 1990). 


Spent grain 

removals 

beam, 

lowered 

position

Perforated false bottom 

Under-deck space 

True base 

Wort collection tube, 

with a conical inlet 

Fig. 6.14 Adiagram of aconical entry to awort collection tube in the base of alauter tun (after 

various sources). 

out' the suction through the false bottom and on to the bed of grains above (Fig. 6.14). In 

the past, as with traditional mash tuns, each pipe carried the wort down to atap with a 

`swan neck' that discharged it into acollecting trough or `grant'. The flow of the wort 

was regulated by hand and the clarity and gravity was checked. In newer tuns the wort is 

usually collected in ring mains or directly into asingle vessel. These arrangements avoid 

exposure to the air and are necessary for automated operation. 

The mashing machinery consists of rotating beams that carry downwardly directed 

knives and, often, asweep arm or`plough' for removing spent grains. The machinery can 

be raised or lowered and this is usually automatically controlled. Knives may have 

`winglets' which project from the sides. As the knives cut through the mash they lift it 

slightlybecause the wingletsare angledtothe horizontal (Fig.6.15a), looseningthemash 

and assisting run off. The bottom of each knife terminates in a`shoe' which, like the 

`winglets' lifts the mash. In some tuns the knives may be rotated through 45±90ë so that, 

as the rotating machinery is lowered into the spent grains at the end of amash, these are 

pushed into the freshly opened discharge ports. In some lauter tuns the direction of 

rotation of the machinery can be reversed and this automaticallyalters the attitudes of the 

knives from their cutting to their grain discharging angles. Because of their greater 

efficiency discharge is increasingly by sweep arms that are lowered into the grain bed in 

front of the knives (Fig. 6.16). Greater efficiency of grains removal results in less 

pollution of cleaning water and so smaller effluent charges (Barnes, 2000). 

Several other types of knives are in use. Rather than being straight, some have `zigzag' 

blades that cut longer channels in the mash but which, because they are not simple 

vertical slots, are less likely to encourage channelling with consequent reductions in 

extract recovery. Another development is the `double shoe' knife (Fig. 6.15b). Neither of 

thesetypes liftthe mash.Inall cases the knives arespacedalongthe carrier beamssothat 

they travel between the tracks of the other knives, the most efficient arrangement for 

raking all parts of the grain bed (Fig. 6.17). Knives are arranged so that each cuts the 

same area of the mash bed. With large lauter tuns this requires special arrangements of 

the arms supporting the knives (Figure 6.17). Inevitably the further knives are positioned 

from the centre of the tun the faster they must move through the mash. Sparging liquor 

may be supplied from rotating overhead sprays, either free or mounted on the beams of 

the raking machinery, or from fixed overhead ducts. The equipment is washed with hot 


(b) 

(a) 

Upright 

Double shoe 

Fig. 6.15 A lauter tun knife equipped with `winglets' to lift the mash (after Andrews and 

Wilkinson, 1996). Note the plastic base to the `shoe' which minimizes the risk of damage to the 

false bottom of the tun. (b) A pattern of double shoe lauter tun knife with a `zig-zag', chevronshaped 

blade (Lenz, 1989). 

water at the end of each run, the false bottom being cleaned with high-pressure jets and 

the under-deck space being washed with fixed sprays. CIP with strong cleaning solutions 

may be needed only once a week. 

A novel lauter tun has an annular filtration area, the central region being occupied with 

a cylinder, up through which the shaft carrying the mashing machinery passes (Putman, 

2002; Wasmuht et al., 2002). It is argued that extract recovery from the central region of 

a conventional tun is inadequate, but since extract recoveries of 99.5% have been 

achieved in such tuns, this claim may be doubted. Many details of this tun have been 

optimized and are claimed to favour the rapid collection of high-gravity worts. 

Lauter tuns are increasingly being automated, but whether operated manually or 

automatically the stages of an operation cycle are the same. The first operation is to warm 

the vessel and flood the under-deck space with liquor at the mashing-off and sparging 


(a) 

Plough bar in rest position 

Plough bar in graining-out position 

(b) 

Grain outlet valves 

Following sweep arm 

Spent grains sweep arm 

Fig. 6.16 (a) A lauter tun with the grain discharge plough bar in the raised (rest) and lowered 

(graining-out) working positions (after Barnes, 2000). (b) A plan view of another pattern of grain 

plough with a trailing, second sweep arm (after Herrmann et al., 1990). 

temperature (often 75±78 ëC; 167±172.4 ëF) to drive out the air and to create a shallow 

layer of water over the deck (1.2±2.5 cm; 0.5±1 in.) to help spread the mash. The mash is 

delivered from the mashing vessel, preferably to the sides or up from below, onto the 

false bottom. In the past loading was from above. The mash must be uniformly well 

mixed. The delivery period is `wasted time' and so is shortened as far as possible, for 

example, by using several large delivery ports. The freshly loaded mash may be raked a 


(a) (b) 

Fig. 6.17 Possible arrangements of knives on the rotating arms in (a) asmall lauter tun and (b) a 

large lauter tun. The knives rotate clockwise and are so spaced along the bars that they track 

between the slots made by the knives that precede them (Andrews and Wilkinson, 1996). 

few times to help ensure even loading and then it is usually (but not invariably) allowed 

to stand for 5±30min., when some settling occurs. The first layer to form is the shallow 

Unterteig, about 1cm (0.39in.) deep, consisting of comparatively large and dense grits, 

which are often under-modified, starch-rich fragments of malt. Then the bulk of the 

grains settles as the Hauptteig, and finally athin layer of very fine particles, the Oberteig 

isformed. This last layerprovides adisproportionately large part of the resistancetowort 

flow and which is first raked or `knived' to accelerate wort collection. Above the settled 

layer of goods alayer of wort forms. In the past some wort was collected directly from 

this upper layer, which could usefully be filtered through a250 min-line filter to 

remove floating particles (Berthold, 2000). This process is now rarely used. 

The stirred and water-logged mash does not float (in contrast to an infusion mash in a 

mashtun).Thefirstwortiswithdrawnand,becauseitcarriesfinelydividedmaterialsfrom 

beneath the false bottom, it is turbid and so is re-circulated to the top of the mash where it 

is discharged, preferably under the surface of the liquid to reduce the pick-up of oxygen. 

The grain bed acts as afilter. When the wort becomes clear, after 5±10min., the flow is 

diverted to thecopper (kettle) or acollection vessel. Inmodern practice, designed tospeed 

up lautering, wort re-circulation may begin when about 50% of the mash has been 

transferred. When a filter layer of spent grains is established and the wort is clear 

collection begins. The first worts are collected until the surface of the grains appears, then 

sparging begins. If wort run-off becomes too slow and/or the pressure drop across the bed 

becomes too high raking will be used, beginning with shallow cuts into the top of the bed, 

through the Oberteig, but with progressively deeper cuts as time passes. If the wort 

cloudiness increases too much the depth of cutting is reduced. Probably, the continuous 

applicationofspargeliquorisusualbutintermittentapplications(twoorthreeapplications 

of known volumes) may be more efficient at recovering extract. If the mash sets and runoff 

ceases an underlet and/or sparge may be used, and the mash may be `re-slurried' using 

the mashing machinery. Sparging may take 2±3h. but is usually less (Table 6.2). 

After wort collection, which is stopped when the wort gravity has fallen to some 

chosen value, the grains are allowed to drain for about five minutes, to reduce their 

moisture content. The discharge ports are then opened and the grains are discharged in 

ten minutes. or less. Efficient and rapid discharge is required, so several large discharge 

ports may be provided. The drainings and vessel washings contain heat, are rich in 

suspended solids and have ahigh COD so they are costly to dispose of as effluent. 

Furthermore, they contain valuable extract although in dilute solution and of alow 


Table 6.2 An example comparison between a mash tun, a lauter tun and a modern mash filter 

(Geering, 1996) 

Mash Lauter tun 2001 Mash 

tun (modern, 1996) filter 

Mashing rate (hl/100 kg) 2.8 3.0 2.9 

Sparging rate (hl/100 kg) 4.2 3.8 2.4 

Total liquor/grist (hl/100 kg) 7.0 6.8 5.3 

Filtration area* (m 2 ) 50 90 708 

Bed depth* (m) 0.9±1.2 (2 max.) 0.3±0.5 0.03±0.06 

Bed loading* (kg/m 2 ) 400 160±220 28 

Capacity range (% normal) 60 to +10 50 to +20 20 to +10 

Initial wort haze (ëEBC) 10 > 20 > 20 

Average wort haze(ëEBC) 4 5±8 < 2 

First wort gravity (ëSacch) 80 94 100 

Copper gravity (ëSacch) 61 67 70 

Lipid content (ppm) 2 18 17 

Solids (Imhoff cone, ml/l) < 8 < 12 < 5 

Polyphenols (ppm) 165 180 195 

Extract recovery (% lab. extract) 96±97 98±99 102 

Moisture draff (%) 81 76 64 

* For 20 t mashes. 

Mash tun cycle: mash in, 20 min.; stand, 75 min.; run off, 185±330 min.; drain and unload draff, 20 min. Total 

time, 300±440 min. Lauter tun cycle: underlet, 3 min.; fill, 11 min.; re-circulate first wort, 4 min.; collect first 

wort, 41 min.; second wort, 74 min.; last wort, 10 min.; weak worts, 16 min.; drain, 8 min.; grains out, 25 min. 

Total, 192 min. 2001 Mash filter cycle time: filling, 4 min.; filtration, 20 min.; pre-compression, 3 min.; 

sparging, 46 min.; compression, 8 min.; drain and grains out, 34 min. Total, 115 min. 

quality, being relatively poorly fermentable and rich in polyphenols and lipids. This 

liquid, together with lauter tun rinsings, can be stored hot for short periods and used in the 

mashing liquor for a subsequent mash, sometimes after a treatment with active charcoal 

or PVPP. The quantities of solids in the drainings and rinsings should be reduced by 

ensuring that the grain discharge is thorough, so fewer solids remain to rinse away. It has 

been suggested that the liquid should be filtered through a 30 m screen. The possibility 

of further purification through a 100 nm pore-size, cross-flow membrane filter with 

consequent reductions in the polyphenols and the virtual elimination of lipids and 

suspended solids also exists (Barnes, 2000). 

The operation of lauter tuns is often fully automated. The flow of wort and its 

cumulative volume, sparge liquor temperature, volume and flow, wort turbidity (haziness), 

the pressure differences across the grain bed (between the top of the vessel contents and the 

under-deck space), and between the under-deck space and the end of the collection pipe or 

the central collection vessel, and the positions of the knifing machinery (rotating speed, or 

stationary, height, direction of movement, knife setting, setting of grain discharge plough) 

are all determined and the measurements are fed to a computer. Increases in the cross-bed 

pressure instigate deeper cutting into the grain bed while increases in wort turbidity initiate 

reductions in the depth of cut. Because lautering is often the slowest process stage in wort 

production, there is a constant pressure to reduce wort separation times. The number of 

cycles that could be achieved with a lauter tun in 24 hours has risen from six to ten, with 12 

being routinely achieved in some instances and even 15 cycles/24 h being claimed. 

However, for these high rates of use to be achieved low bed loadings must be used. 

These rates are dependent on the design of the tun, the nature and grind of the grist 

being used, as well as the manner in which the mash is carried out and the way in which 


the lauter tun is operated. Operating more quickly gives more turbid worts and reduces 

extract recoveries. The use of inert gases, nitrogen or carbon dioxide, to minimize 

oxidative deterioration and reduce the oxidation of gel proteins, which contribute to the 

flow-resistance provided by the Oberteig, has been combined, at least experimentally, 

with the use of gas pressure to accelerate wort run off (Lee et al., 1997; Stippler et al., 

1994, 1995; Stippler and Johnstone, 1994). This requires a lauter tun to be fitted with 

valves and seals to contain the pressure and to prevent an excessive pressure increase or 

the formation of a vacuum. Pressure-lautering may also give better extract recoveries and 

spent grains containing less moisture. 

6.6 The Strainmaster 

The Strainmaster, or Nooter tun, was developed as a compact and exceptionally rapid 

lautering device in the late 1950s (Irvine, 1985; Narziss, 1992; Vermeylen, 1962). Each 

unit consisted of a cylindrical or rectangular hopper-bottomed tank, with discharge doors, 

`bomb doors', in the base. Each tank had seven layers of stainless steel strainer tubes, 

with pear-shaped cross-sections in the lower half of the vessel. The tubes (130 mm 

(5.25 in.) high and maximal width 50 mm (about 2 in.)) were perforated with vertical 

slots, 0.63 mm (0.025 in.) wide and 12.7 mm (0.5 in.) long. Each layer of tubes was 

connected by a main to a separate wort pump then to separate `swan neck' tubes, which 

discharged into a grant (Fig. 6.18). The tun was used with a finely milled grist made into 

a thin mash (2.5 hl/100 kg). After preheating the vessel a mash was delivered through two 

inlets, which allowed fast and even loading. As each layer of tubes was covered by the 

mash the appropriate pump was switched on and cloudy wort was drawn off and recirculated 

until the grains formed a filter layer on the slots. As soon as the wort ran bright 

it was directed to the copper. When the mash transfer was complete the continued wort 

withdrawal caused the volume of the mash to contract. When the volume had declined to 

Grant 

Vorlauf* 

pump 

Vorlauf* line 

Sewer line Wort to kettle 

* Vorlauf = first wort (re-circulation) 

Strainmaster 

Vapour vent 

Mash-transfer 

pump 

Mash 

Sparge water pump 

Sparge water 

Spent-grain gates 

Spent grain to spent-grain receiving tank 

Five wort pumps 

Five wort draw-off lines 

Fig. 6.18 A Stainmaster lauter tun. The patent is owned by Anheuser-Busch (after Briggs et al., 

1981). 


apre-selected value sparging began automatically. When wort collection was complete 

cold water might be added to the top of the grains, the discharge doors were opened and 

the spent grains fell and were rinsed into acollecting vessel.Various performance figures 

were quoted. For example, pre-heating the vessel and transferring the mash (6.8t) 30± 

40min.; run off, 90min.; grains out, 5min.; flush, 10min., total time 135±145min. 

Others claimed turn round times of 105min. at adate (1985) when alauter tun cycle 

might be 180min. The performance of this exceptional device, which created an 

important benchmark for rapid wort recovery, has now been matched or overtaken. 

Problems associated with its use included the production of exceptionally wet grains 

(moisture contents over 80%) which necessitated dewatering with presses or vibrating 

screens. The press liquor was used to make subsequent mashes. Extract recoveries were 

generallylow,rarelyexceeding96%.Wortswerecloudyandeffluentloadingswerehigh. 

The system was not suitable for making high-gravity worts. 

6.7 Mash filters 

The use of low-pressure mash filter presses was suggested as long ago as 1819, but the 

first practical experiments were not carried out until 1874, by Galland, and the first 

commercialpresswasmadebyMeurain1890(Dixon,1977).Sincethattimemashfilters 

have undergone many refinements and, since various types are in use, examples will be 

described in turn. Filters compete with lauter tuns and so comparisons between these 

devices will be made. All filters retain the spent grains with filter cloths. Depending on 

the cloth used the grist can be moderately, finely or very finely ground, permitting better 

extractrecovery.Becausethefilterlayersofgristarethinwortcollectionisrapid(despite 

the resistance of the fine grist to liquid flow), and so wort collection times are short. 

Traditional filters consist of two types of units (plates and frames), made of iron or 

stainless steel, suspended alternately along aframe and pressed together when in use 

(Figs 6.19, 6.20. De Clerck, 1957; Dixon, 1977; Irvine, 1985; Kunze, 1996; Narziss, 

1992). The plates from which the wort is collected were originally of grooved iron but 

these have been replaced by lighter grid plates or folded-sheet plates. The plates have 

filter cloths draped over them, which provide the support for and retain the mash (Fig. 

6.20). In the past cotton cloths were used. These have been replaced by monofilament 

polypropylene cloths, which last longer (five times or more) and are easier to clean. 

Thorough cleaning with caustic solutions may be needed only every 800 brews. In 

contrast, cotton cloths needed cleaning after every brew but gave marginally less cloudy 

worts. The other units, the frames, are hollow and contain the mash. When the stacks of 

alternating units, which may be 1.5 m square (4.92 ft. square), are pressed together, often 

by a hydraulic ram, the gaskets seal between them so that a series of alternating mashcontaining 

and wort-collection chambers are formed and the holes around the edges of 

the units are joined to form channels for the mash, the sparge liquor and the wort. 

Filters may contain around nine tonnes of grist and consist of stacks of 10±60 units. 

The bed thickness is chosen with reference to the grist to be used, so the finer the grind 

the narrower the bed. Values of 4±6 cm (1.57±2.36 in.) and even up to 10 cm (3.94 in.) 

have been used. The end plates are different and seal off channels or provide connections 

to the mash-delivery and sparge lines. For mash filters to work well, they need to be 

completely filled and so they should be used with one invariable volume of mash. To 

reduce the capacity of a filter it is necessary to introduce blanking plates, which is 

inconvenient and laborious. These older units are made of metal and the heat losses to the 


Wort 

Water 

Flow of 

wort 


Flow of water 

I II III 

Wort pan 

To copper Grains conveyor 

Fig. 6.19 A schematic longitudinal section through an older pattern of mash filter (Hind, 1940).

Sparge 

Hollow 

frame 

Cloth-covered 

grid 

Hollow 

frame 

Filter cloth 

covering grid 

(a) 

Top sparge 

channel 

Bottom 

sparge channel 

Hollow frame 

(c) 

Mash entry 

Channel for 

mash entry 

Air 

channel 

Wort 

channel 

surroundings are high, the filters are uncomfortably hot to work near and must be 

mounted in a well-ventilated area. Filters occupy a smaller floor area than lauter tuns of 

similar capacities. To discharge the spent grains the plates and frames are separated 

automatically and in turn, when the spent grains fall from the frames and into a collecting 

trough from which they are moved by a screw conveyor. The opening and grain 

discharging operations are relatively slow. To halve the time taken `double filters' are 

used, in which the filters are opened simultaneously from both ends. In operation a filter 

is preheated to about 80 ëC (176 ëF) with steam or hot water and is checked for leaks. The 

filter is emptied and well-mixed mash is pumped from the mashing vessel and is loaded 

into the frames, from above in older designs, but from below and with minimum 

turbulence in newer types, to minimize oxygen uptake. 

Grid 

(b) 

Wort 

Spargings 

(d) (e) 

Bottom sparge in 

Spargings out 

Fig. 6.20 The arrangement of the essential components of a traditional mash filter (Briggs et al., 

1981). (a) A vertical section showing how the filter cloths hang over the grids and alternate with the 

hollow frames, which will contain mash. (b) A face-view of a grid. (c) A face-view of a frame. 

(d) A stage when the filter is just full of mash and the first worts are escaping (solid lines). The next 

stage with sparging into odd numbered grids and collecting from even numbered grids (dashed 

lines). (e) Last stage in sparging. 


Pre-filling of filters with an inert gas to displace oxygen has been considered. The air 

or inert gas displaced by the mash is vented. When the frames are full of mash the gas 

vents are closed while mash is still being transferred into the frames. The taps or outlets 

from the plates are opened and the wort, which has escaped through the filter cloths, is recirculated 

to the mashing vessel until it is clear and then is collected. In older mash filters 

the flow of wort from each plate was regulated through individual taps leading to swan 

necks, which delivered the wort into a wort collection trough or grant. 

In modern units the wort is collected into a common main and air is excluded. After 

the first wort has been collected the sparge liquor, which may have been rendered 

oxygen-free, is admitted into the tops of alternate plates, displacing the wort downwards. 

When all the wort has been driven out, the wort outlets from these plates are closed while 

those of the remaining plates are opened. Now the sparge liquor passes from the first 

plates across through the mash in the frames and out from the second plates. When 

sparging has been sufficient the last of the sparge liquor is driven through the spent grains 

by a top pressure provided by compressed air, which also drives out the last of the 

spargings. The wort emerges at about 75 ëC (167 ëF). Extract recovery can be as high as 

99.5% of the laboratory extract, the first worts are strong (SG 1085±1090) and the final, 

average values are 1044±1048. Thus filters are useful for high-gravity brewing. The spent 

grains have moisture contents in the range of 60±80%, usually around 70±75%. The turn 

around times of these filters is fast enough to allow 8 brews/24h. Sometimes the sequence 

of production runs must be interrupted for cleaning, which involves rinsing with hot 

water and treatments with 1±2% caustic soda containing a wetting agent and a chelating 

agent, such as EDTA, if polypropylene sheets are used. 

A novel development is the recessed chamber-plate filter (Nguyen, 1996). These units 

are made of reinforced polypropylene, and a filter press consists of only this single type of 

unit, together with appropriate end plates, supported on a frame. Each recessed chamber 

plate, typically 2.1 m (6.89 ft.) square, consists of a single unit with a recess each side into 

which the filter cloth fits and is retained by an Ò-ring' (Fig. 6.21). Unlike other filters, the 

cloth does not project from below the units. When the units are assembled the space 

between the filter cloths is the mash chamber while the spaces between the cloths and the 

transverse partitions of the plates make up the wort collection chambers and the sparge 

inlets and outlets. The plates are opened in the usual way to discharge the spent grains 

which separate from the two filter cloths. The use of an air purge at the end of sparging 

gives relatively dry spent grains, with < 74% moisture. The worts are concentrated, extract 

recoveries equal laboratory yields and the cycle time is less that 120 min., permitting 12 

cycles/24 h. Like other newer filters made primarily of polypropylene, heat losses are less 

Recessed chamber plate 

Filter cloth 

Fig. 6.21 Two recessed chamber plates, seen in vertical section (after Nguyen, 1996). The space 

between the cloths is the mash chamber. The wort is collected in the spaces between the cloths and 

the plate partitions. 


ii i ii i ii 

(a) (b) 

Fig. 6.22 Working diagram of ahigh-pressure mash filter (Waesburghe, van, 1979, 1989). 

(a) Vertical section of part of a filter series in which vertically grooved, stainless steel plates (ii); 

(b) alternate with the filter pockets (i) which contains the mash. The light arrows indicate the mash 

loading position. The spent grains were discharged by unsealing the bases of the pockets which are 

sealed by pressure against the groove on the metal plates. The bold arrows indicate the direction of 

compression used on the filter assembly. 

than those from metal filters, giving more pleasant working conditions and allowing the 

use of more aggressive cleaning agents without risk of corrosion. 

Anotherdevice,introducedin1978andwhichenjoyedlimitedsuccess,istheHP(highpressure) 

filter (Waesberghe, van, 1979, 1991; Irvine, 1985; Nguyen, 1996). This filter 

consisted of aset of polyester or polypropylene filter pockets, which could be opened or 

closed at the base, alternating with metal plates vertically grooved on one side and with a 

recessinvolvedwithbagclosureandsealingontheother(Fig.6.22).Thebagswerefilled 

fromaboveand,afterashortre-circulationperiodforclarification,thewortwascollected. 

Because the grists were finely milled the initial worts were bright. The mash was sparged 

withwateraddedtothetopsofthepockets,thenthepocketsweregentlycompressedfrom 

7±8cm (2.76±3.15in.) thickness and finally squeezed more strongly to 3±4cm (1.18± 

1.57in.). The pockets were then opened and the comparatively dry grains (50±60% 

moisture)droppedout.Extractrecoveriesweregood,thewortswereconcentratedandthe 

turn-round time was rapid, about 70min. These filters did not remain in favour, possibly 

because of cleaning problems, turbidity in the later `squeezed' worts, difficulties in 

obtaining even sparging and the complexity of the equipment (Nguyen, 1996). 

The newer generation of filters may be termed membrane compression filters. The 

first of these is the Meura 2001 filter (Eyben et al., 1989; Hermia and Rahier, 1991a,b, 

1992; Jones, 1992; Mieleniewski, 1999). The filter is made of alternating chamber 

modules with membranes and plates, which are covered with filter cloths. Compression 

and gaskets seal between the units and create the necessary channels for the mash, wort, 

spargings, etc. The units are made chiefly of reinforced polypropylene and are 2by 1.8m 

(6.56 by 5.91ft.). The membrane chamber module consists ofathin, grooved plate which 

supports two elastic membranes or `diaphragms' (of polypropylene and rubber) on each 

side that can be inflated with compressed air. The space between amembrane and afilter 

cloth, which holds the mash, is about 4cm (1.57in.) wide. The mash is made with afine, 

dry hammer-milled or awet disc-milled grist. A60-plate filter can accommodate a10.5t 

mash, divided into 120 beds and resting against 60 double filtration cloths. The stages of 

operation are as follows (Fig. 6.23). The filter is pre-warmed and then mash is pumped 

into the chambers from below, while the air is vented. This upward delivery minimizes 

oxygen uptake. When the chambers are full the vents are closed, the wort outlets are 

opened, wort collection begins and mash delivery is continued until all the mash is in the 

chambers. During this process afilter-layer of solids quickly builds up on the filter cloth 

and the wort becomes bright so quickly that re-circulation is rarely needed. The 


ii

membranes are then inflated with compressed air, (0.5±0.6 bar; 7±9psi), and these 

squeeze the mash, displacing the solids across the chamber and compressing them so that 

they adhere to the filter cloths and strong wort is squeezed out. Squeezing improves the 

homogeneity of the bed, as well as making it thinner, and so improves the efficiency of 

sparging.Beforesqueezingabout50%oftheextractisretainedinthegrist.Afterthisfirst 

compression the membranes are relaxed and the spaces, which appear between them and 

the grist layers, are filled with sparge liquor (degassed for preference), usually at about 

78ëC (172.4ëF). 

When sparging is complete the liquor inlets are closed and the mash is compressed a 

second time, by inflating the membranes with compressed air (0.7bar; 10psi), and the 

last spargings are squeezed out. The plates and frames are automatically separated in turn 

and the spent grains fall into areceiving trough with aconveyor. Higher air pressures in 

the last compression give grains with smaller moisture contents, but extremely low 

moisture contents making the grains difficult to handle and values of 60±70% are 

probably usual. These filters have rapid turn-round times, allowing 12 cycles/24h, the 

worts are strong and extract recovery is exceptional, often 101 and even 103% of the 

laboratory value (Table 6.2). The fatty acid contents of these worts are very low, 10± 

24mg/l for Meura 2001 worts, 27mg/l for lauter tun worts and even 143mg/l for filter 

press worts. It has been claimed that the loading can be reduced to 70% of normal, but 

practical difficulties have been noted even at loadings reduced to 85% of the standard 

value.Thisdiscrepancymaybeduetodifferencesinthegristsused.Thisfiltercopeswell 

with awide variety of grists, including those containing mostly unmalted sorghum or 

roasted and flaked barley. The first and total worts are unusually concentrated, e.g., 

14.5ëP (SG 1059) and 11.8ëP (SG 1048) respectively, and so are suited to high-gravity 

brewing. Filters are cleaned regularly by rinsing and more completely, say weekly, using 

caustic cleaning agents and possibly hydrogen peroxide. As the filter cloths become dirty 

the resistance to wort flow increases. As with other polypropylene filters the heat loss is 

much less than with the older, metal filter presses. 

The other membrane filter to be described is the MK 15/20, intended for 5±12.5t 

mashes (Karstens, 1996; Kunze, 1996; Michel, 1993). Smaller versions have other 

reference numbers. The filters consist of alternating membrane and chamber plates, 

mostly made of polypropylene. Filter cloths are firmly fastened over both sides of both 

kinds of plates (Fig. 6.24). When the units are assembled and pressed together the edges 

of the cloths seal (replacing gaskets) and create mash chambers bounded on both sides by 

filter cloths, so wort can escape from both sides of the mash. The sequence of operations 

is indicated in Fig. 6.24. The mash cycle time is about 120min. 

6.8 The choice of mashing and wort separation systems 

The choice of mashing system depends on many factors. For the production of traditional 

beers it may be imperative to continue using near-isothermal mashing in a mash tun or 

decoction mashing to retain the character of a product. In the same way, a particular 

fermentation system may have to be retained. For small traditional breweries making ales 

a mash tun, with its simple construction and low maintenance requirements, serves well if 

not more that two to three brews ervery 24 hours are required. For larger breweries, 

making large volumes of single beers, the economic advantages of using all the brewing 

equipment for as much of the week as possible, and making as many brews every 24 

hours as possible are very great. Because both conversion and wort separation occur in 


(a) 

(c) 

Air 

Mash 

1st wort 

1st wort 


(d) 

(b) 

Mash 

1st wort 

Sparge liquors 

Spargings

(e) 

the single vessel in amash tun and the depth of the bed of grist is large, the turn round 

time is slow. Some decoction mashing schedules are slow (Chapter 4), but with 

temperature-programmed infusion mashing with an all-malt grist conversion should be 

completeintwohours,orless,andwortseparationmaybecompletedsufficientlyquickly 

to readily allow 8±10 (and possibly 12) brews/24hwith alauter tun and 10±12 brews/ 

24h. with anew mash filter. 

Both lauter tuns and mash filters can produce high-quality worts (Table 6.2). Amash 

filter using hammer milled grist will give abetter extract recovery, amore concentrated 

wort, drier spent grains, use less water and generate less effluent, but is likely to be more 

costly than alauter tun of equivalent capacity. If the brewery is making avariety of beers 

that necessitate the use of different brew lengths then the greater flexibility of the lauter 

tun becomes decisively important. Furthermore, the use of acomparatively coarsely 

ground grist, suited to alauter tun, may be needed to retain the character of aparticular 

beer. Lauter tuns are likely to be less expensive than newer types of mash filters and 

require less maintenance and fewer replacements. The operations of both are now usually 

automated. Flexibility may be increased by having mash conversion vessels of different 

Air 

Sparging 

Fig. 6.23 Stages in the operation of aMeura 2001 membrane mash filter (various sources). (a) 

The mash being loaded into the frames from below the membrane and the filter cloth. Air is being 

expelled at the top. (b) All the air has been expelled, the vents have been closed, mash is still being 

pumped into the filter and the first wort is escaping through the filter cloth and is being collected. 

(c) All the mash has been transferred into the filter, the inlet lines are closed and compressed air has 

been directed behind the membranes, pushing them out and causing them to squeeze the mash, 

releasing entrained first wort. (d) Sparge liquor has been admitted between the deflated membrane 

and the compressed mash and the first spargings are being collected. (e) The membranes have been 

inflated for the second time, to squeeze the spargings from the draff. In the next stage the filter is 

opened and the spent grains fall into a collection trough and are conveyed away. 


A 

Sp 

A 

Wort 1 Wort 1 

(a) (b) (c) 

L 

capacities and the choice of using either one of two wort separation units (Mieleniewski, 

1999). Alternatively, when slow mash conversions occur, for example because relatively 

high concentrations of slow-converting mash adjuncts are used, two mash conversion 

vessels may alternately deliver mash to a single wort separation unit. It is apparent that 

for optimal working all the components used in the preparation of wort, raw materials 

handling, milling and grist handling, mashing/conversion, sweet wort separation, hopboiling, 

trub separation, wort cooling and fermentation must have the appropriate 

capacities and working rates to avoid `bottlenecks' and permit fast working at periods of 

maximum demand. 

6.9 Other methods of wort separation and mashing 

Sp 

Various unusual methods of wort separation have been tested (Briggs et al., 1981; Lotz et 

al., 1997; Dixon, 1977; BeÂndek et al., 1991; Darling, 1968; Schneider et al., 2001; 

L 

Sparge 

(d) (e) (f) 

Fig. 6.24 The operational stages of an MK 15/20 filter (various sources). (a) Mash is added at the 

base and rises between the two filter cloths, and air is displaced, A. (b) All the air has been 

displaced, the vents are closed and mash entry continues. Layers of grain build up on the filter 

cloths and escaping wort is collected. (c) Compressed air, A, inflates the membranes and 

compresses the mash. The displaced wort is collected. (d) Sparge liquor, L, is admitted and the 

spargings, Sp, are collected. (e) Possible sparging between the two filter cloths, but in the reverse 

direction. (f) At the end of sparging the grist is squeezed again and the displaced spargings are 

collected. At the end of this sequence the filter is opened and the draff falls into a collecting trough. 


A 

A

SchoÈffel and Deublein, 1980). With the proposed dynamic disc mash filtration technique 

the rotating disc drives the mash tangentially to the filtration surface and the wort is 

membrane filtered. In the Pablo system the wort is separated from the mash with two, 

doublestagedecantingcentrifuges,intheReitersystemarotaryvacuumfilterisusedand 

gives a high extract recovery. Other devices tried include vibrating screen filters, 

vibrating membrane filters, horizontal and inclined belt filters, with or without suction, 

cross-flow filters, Archimedean screws working in slotted casings and cyclones. Many of 

these methods were intended to be used with fractionated malt grists lacking husk 

materials so that relatively little draff would remain after the conversion process. Some 

were used for short periods but many did not pass beyond the experimental stage. 

During the period 1955 to 1975 there was intense interest in continuous brewing, with 

its advantages of relatively compact, continuously operating equipment with its steady 

demand for services and its predicted uniform product quality. Large production plants 

were constructed and used but many problems were encountered that led to them being 

replacedbybatchproductionunits. Atthepresenttimecontinuousfermentationiscarried 

out on an industrial scale by only one company in New Zealand (Chapter 14) and this is 

not linked to continuous wort production. However, there is a revival of interest in 

continuous fermentation and conditioning using immobilized yeasts and this may lead 

back to an interest in continuous mashing, since processing is most advantageous when 

all stages are continuous (Briggs et al., 1981; Darling, 1968). Perhaps the most promising 

continuous mashing system was the rotary table filter proposed by APV (Fig. 6.25). In 

this equipment the mash travelled, in sequence in plug flow, through stainless steel tubes 

Hot 

water 

Mash 

mixer 

Steam 

Hot water 

Flake Roast Malt 

Flake weigh 

feeder 

Mash 

filter 

Empty 

wash 

Prefill 

fill 

Malt weigh 

feeder 

Mill 

Converter tank 

Recycle 

Recycle 

Recycle 

Re-circulation 

Re-circulation 

Hot water 

returned 

Wort out 

Hot water in 

Condensate 

Fig. 6.25 A diagram of the A. P. V. continuous mashing and lautering system (Briggs et al., 1981). 


held in `converter tank(s)' at the chosen mashing temperature(s). The converted mash 

was loaded into one of eight mash buckets, each of which resembled asmall mash tun. 

For two hours atun moved around acentral support, occupying each of 16 positions in 

turn, for15 min. The positions were used for filling, wort recycling, collection and 

sparging, spent grain discharge and cleaning, then return to the filling position. 


The undesirable characteristics of mash and spent grain drainings as effluents, with their 

high COD, BOD and suspended solids values, have been noted (Chapter 3), as has the 

nutritional valueofdraffinanimal feeds(Chapter4).Itisbestforthe brewerandtheuser 

that the draff is reasonably dry and certainly not seeping liquid. Drier draff reflects better 

extract recovery for the brewer and smaller transport costs and storage problems for the 

user. The wet state of the draff from Strainmasters was a major factor in replacing them 

with other equipment. Conversely, the low moisture content of the draff from mash filters 

is a factor in their favour. The production of dry draff becomes even more important 

where local conditions dictate that it must be thoroughly dried. Drying is an expensive 

process. Spent grains are usually moved by helical screw or compressed air conveyors to 

storage bins which typically discharge their contents directly into lorries. This area of the 

brewery must be regularly cleaned, since many microbes multiply on wet grains and 

drainings, so these represent potential sources of contamination and product spoilage. 

6.11 Theory of wort separation 

The theories applied to wort separation have been considered and have been of value in 

designing new equipment (lauter tuns and mash filters) but they apply to ìdeal' situations 

while, in practice, the conditions are often far from perfect (Briggs et al., 1981; Dixon, 

1977; Harris, 1968, 1971; Hermia and Rahie, 1992; Royston, 1966; Wilkinson, 2001). 

The modified Darcy's equation (originally derived from the passage of water through 

beds of sand) may be written: 

V = K A P/L 

where V is the rate of liquid flow through the bed of particles, A is the area of the bed, 

P is the pressure difference across the bed, K is the permeability of the bed, L is the 

length of the path through the bed, or the bed depth, and is the viscosity of the liquid. K, 

the mash bed permeability, ˆ 3 de 2 /180(1 ) 2 , where ˆ bed porosity (wort volume/ 

mash volume) and de ˆ effective particle diameter (sum of the weight fraction/particle 

diameter). In a mash the viscosity will depend on the concentration and nature of the wort 

and the temperature. The use of adjuncts that give viscous worts creates run-off problems. 

The porosity of the bed is more important than that of the support, the false bottom of a 

mash or lauter tun or the clean filter cloth of a mash filter, and is proportional to the mean 

diameter squared of the grist particles. Thus the finer the grind of the grist, the smaller the 

particles and the greater the resistance to flow. 

The particle shape is also important. The faster flow of lauter tuns relative to mash 

tuns is achieved (despite the finer grists used) by reducing L, the bed depth, and this 

process is carried further in mash filters. While increasing the pressure can increase the 

rate of wort flow this approach can be counter-productive since the mash beds are far 


from ideal and can be compressed to such an extent that flow is reduced or even ceases 

and a`set mash' is created. As wort separation proceeds so the bed of grains builds up to 

amaximum thickness, L, and may then decrease as the bed contracts or is compressed. 

After the first wort is collected and sparging begins wort concentration and hence its 

viscosity, falls. Similar conclusions can be drawn from modifications of Poiseuille's 

equation, derived to describe the flow of liquids through bundles of equal sized 

capillaries, in which the diameter of the capillaries is replaced by the mean diameter of 

the pores in the bed (or the related value of `voidage'). The flow through the bed is 

proportional to the square of the mean diameter of the pores. 

As sparge liquor moves into the bed of grist it displaces the wort from between the 

particles and leaches extract from within them. Efficient leaching is essential if high 

yields of extract are to be obtained. The solid phase mass transfer coefficient, K/D/d 

whereDisthediffusioncoefficientanddisthediameteroftheparticle.Thus,thesmaller 

the particle the greater the surface/volume ratio and the shorter the distances from points 

within the particle to the surface and the faster extract can be leached from it. Diffusion 

occurs faster at higher temperatures but sparging temperatures are limited to about 78ëC 

(172.4ëF)byotherconsiderations.Leachingtakestimeandspargingtoorapidlyresultsin 

inadequate leaching and areduced recovery of extract. Since leaching is favoured by a 

finely ground grist while flow rate is favoured by larger particles, in practice, a 

compromise grist particle size must be sought for each type of wort separation 

equipment. The more brews/day that are required the lower the loading on alauter tun 

mustbe.Forexample,fortotalcycletimesof240,180and120min.alautertun,usingall 

malt grists, had loadings (kg/m 2 )of 339, 246 and 153 respectively (Wilkinson, 2001). 

During the movement of sparge water through the mash bed there is aprogressive 

leaching of soluble substances from within the particles and a gradient of wort 

concentration is established, which increases downwards through the bed. The theoretical 

stages are, firstly, the diffusion of extract from the interiors to the surfaces of the grist 

particles and, secondly, the movement of the extract into the liquid between the particles 

and its removal with this flowing liquid. Leaching has been approximately described by 

equations based on the number of theoretical washing stages involved (Table 6.3). More 

washing stages are required to minimize losses in the preparation of strong worts. Greater 

retention of liquor in the spent grains leads to greater losses of extract and the need for 

more extensive washing. The `squeezes' applied in operating membrane mash filters 

reduce, firstly, the amount of strong wort and, secondly, the amount of spargings in the 

draff and so enhance extract recovery. As grain beds become more free running so the 

washing efficiency decreases and there is a tendency for extract recovery to decline. It 

follows that increasing the rate of sparging carries with it the risk of a significant fall in 

extract recovery. As previously noted with some plant, like the Strainmaster, the losses of 

extract in wet spent grains can be so high that it is desirable to recover the extract from 

the grain pressings, with the extra cost and effort involved and the risk of reducing 

product quality. 

Another process that occurs during wort separation is the filtration of the fine particles 

from the wort. It is desirable that the wort be as bright as possible. The nature and 

abundance of these particles depend on many factors. The large bed depths used in mash 

tuns ensure that, combined with wort re-circulation, very clear worts can be obtained. 

Generally lauter tuns give less clear worts, in part because of the need to `knife' or `rake' 

the bed to maintain an acceptable rate of run off. Older, `classical' mash filters gave rise 

to even more hazy worts but the newer, membrane compression mash filters give rise to 

exceptionally clear worts. This sequence seems to be the result of a several conflicting 


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) 

For a wort of SG 1050 (12.4 ëP) For a wort of SG 1100 (23.7 ëP) 

Spent grains 

moisture First 

Washing stages 

First 

Washing stages 

contents (%) worts 1 2 3 4 5 worts 1 2 3 4 5 

87 22 6 1.5 0.5 0.1 0.1 40 18 11 8 6 4 

75 12 2 0.4 0.1 0.1 0.1 25 11 6 4 3 2 

60 7 0.7 0.1 0.1 0.1 0.1 15 5 3 1.5 0.8 0.5 


Extraction (%) 

100 

80 

60 

40 

20 

traits, that thicker beds make better filters, more fines are formed with finer grinding and 

exceptionally finely ground (hammer milled) grists are able to retain very fine suspended 

materials, presumably because the mean pore sizes through the beds are very small. The 

haziness of worts is also influenced by the grist composition and is probably increased by 

attempts to collect worts too fast, at rates that exceed the optima of the different pieces of 

equipment. 

Figure 6.26 illustrates the progressions of recovery of extract during wort run off from 

several types of separation devices. The rates are Strainmaster > lauter tun > mash tuns. 

Thus, as predicted, the larger the filtration area /unit mash the faster extract collection. 

Wort recovery from the mash tun with medium depth and an all-malt grist is complete 

(97% extract recovery) in 255 min., while the values with the deeper tuns, with malt and 

malt and maize grits, (98% extract recovery) are 330 and 400 min. The performance (true 

filtration efficiency) of the Stainmaster and the lauter tuns was not as great as predicted 

on theoretical grounds, possibly because, at least in part, the mash beds were compressed 

(Harris, 1971). 


1 2 3 4 5 6 

0 0 1 2 3 4 5 6 7 8 

Run-off time (h) 

Fig. 6.26 The patterns of the cumulative extract recovery against wort filtration time for various 

systems (Harris, 1968). (1) A Strainmaster lautering vessel. (2) A lauter tun loaded with a decoction 

mash. (3) A lauter tun loaded with a transferred infusion mash. (4) A mash tun with a medium depth 

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 

deep bed of grist containing both malt and maize grits, which slow wort separation. 

ANDREWS, J. M. H. and WILKINSON, N. R. (1996) Ferment, 9 (5), 257. 

BARNES, Z. C. (2000) Ferment, 13 (1), 27. 

BARNES Z. C. and ANDREWS, J. M. H. (1998) Proc. 5th Aviemore Conf., Malting, Brewing and Distilling 

(Campbell, I. ed.). London, The Institute of Brewing, p. 127. 

BEÂ NDEK, G., HORVAÂ TH, I. and ULLMANN, P. (1991) Proc. 23rd Congr. Eur. Brew. Conv., Lisbon, p. 633. 

BERTHOLD, U. (2000) Proc. 26th Conv. Inst. of Brewing (Asia Pacific Section), Singapore, p. 52. 

BRIGGS, D. E., HOUGH, J. S., STEVENS, R. and YOUNG, T. W. (1981) Malting and Brewing Science (2nd edn), 

Vol. I Malt and Sweet Wort. London, Chapman and Hall, 387 pp. 

BROWN, H. T. (1910) J. Inst. Brewing, 16, 112. 

BUÈ HLER, T. M., MATZNER, G. and MCKECKNIE, M. T. (1995) Proc. 25th Congr. Eur. Brew. Conv., Brussels, 

p. 293. 

DARLING, R. O. (1968) Brewers' Guard., Apr., p. 147. 


DE CLERCK, J. (1957) A Textbook of Brewing, vol. 1. (Trans. Barton-Wright, K.). London, Chapman and 

Hall, 587 pp. 

DIXON, I. (1977) Brewers' Guard., 106 (10), 43. 

EYBEN, D., HERMIA, J., MEURENS, J., RAHIER, G. and TIGEL, R. (1989). Proc. 22nd Congr. Eur. Brew. Conv., 

Zurich, p. 275. 

GEERING, P. (1996) Brewers' Guard., 125 (10), 32. 

HARRIS, J. O. (1968) J. Inst. Brewing, 74, 500. 

HARRIS, J. O. (1971) in Modern Brewing Technology (Findlay, W. P. K. ed.). London, MacMillan Press, p. 1. 

HERRMANN, H. (1998) Ferment, 11 (1), 36. 

HERRMANN, H., KANTELBERG, B. and LENZ, B. (1990) Brauwelt Internat., (1), 22. 

HERMIA, J. and RAHIER, G. (1991a) Louvain Brew. Lett., 4 (3/4), 30. 

HERMIA, J. and RAHIER, G. (1991b) Louvain Brew. Lett., 4 (2), 24. 

HERMIA J. and RAHIER, G. (1992) Ferment, 5 (4), 280. 

HIND, H. L. (1940) Brewing Science and Practice. Vol. II. Brewing Processes. London, Chapman and Hall, 

pp. 507±1020. 

HOPKINS, R. H. and CARTER, W. A. (1933) J. Inst. Brewing, 39, 59. 

IRVINE, J. A. (1985) The Brewer, Feb., p. 46. 

JONES, I. R. (1992) Brewers' Guard., 121 (10), 21. 

KARSTENS, W. (1996) Brauwelt Internat., 14 (4), 340. 

KUNZE, W. (1996) Technology Brewing and Malting. (Wainwright, T., transl.). Berlin, VLB, pp. 726. 

LEE, B. W., KIM, I. K. and SHIN, J. Y. (1997) MBAA Tech. Quart., 34 (2), 75. 

LENZ, B. (1989) Proc. 2nd Sci. Tech. Conv. Inst. of Brewing (Central and Southern African Sect.), 


LOTZ, M., SCHNEIDER, J., WEISSER, H., KROTTENTHALER, M. and BACK, W. (1997) Proc. 26th Congr. Eur. 

Brew. Conv., Maastricht, p. 299. 

MCFARLANE, I. K. (1993) Ferment, 6 (3), 177. 

MICHEL, R. (1993) Brauwelt Internat., (2), 123. 

MIELENIEWSKI, A. (1999) Brew. Distill. Internat., 30 (3), 12. 

NARZISS, L. (1992) Die Bierbrauerei, Bd. II. Die Technologie der WuÈrzebereitung, (7th edn). Stuttgart, 

Ferdinand Enke, 402 pp. 

NGUYEN, M. T. (1996) Ferment, 9 (6), 329. 

PUTMAN, R. (2002) Brewer Internat., 2 (10), 4. 

REHBERGER, A. J. and LUTHER, G. E. (1994) in Handbook of Brewing, (Hardwick, W. A. ed.). New York, 

Marcel Dekker, p. 247. 

ROYSTON, M. G. (1966) J. Inst. Brewing, 72, 351. 

SAMBROOK, P. (1996) Country House Brewing in England, 1500±1900. London, The Hambledon Press, 

p. 311. 

SCHNEIDER, J., KROTTENTHALER, M., BACK, W. and WEISSER, H. (2001) Proc. 28th Congr. Eur. Brew. 

Conv., Brussels, CD, paper 22. 

SCHOÈ FFEL, F. and DEUBLEIN, D. (1980) Brauwissenschaft, 33 (10, 11), 263, 304. 

SCOTT, P. MCM. (1967) Brewer's Guild J., p. 339. 

SOMMER, G. (1986) Brauwelt Internat., (1), 23. 

STIPPLER, K. and JOHNSTONE, J. (1994) Brew. Distill. Internat., Apr., 25, 26. 

STIPPLER, K., WASMUHT, K., PRITSCHER, R. and KEIM, N. (1994) Proc. 23rd Conv. Inst. of Brewing (Asia 

Pacific Sect.), Sydney, p. 56. 

STIPPLER, K., WASMUHT, H., PRITSCHER, R. and KEIM, N. (1995) MBAA Tech Quart., 32 (1), 1. 

SYKES, W. J. and LING, A. R. (1907) The Principles and Practice of Brewing (3rd edn) London, Charles 

Griffin and Co., 588 pp. 

VERMEYLEN, J. (1962) TraiteÂe de la Fabrication du Malt et de la BieÁre. 2. Gand. Assoc. Royale des 

Anciens EleÁves del'Institut SupeÂrieur des Fermentations, pp. 751±1624. 

WAESBERGHE, J. W. M. VAN (1979) Brew. Distill. Internat., 9 (9), 54. 

WAESBERGHE, J. W. M. VAN (1989) MBAA Tech. Quart., 17 (2), 66. 

WAESBERGHE, I. R. J. VAN (1991) Proc. 3rd Sci. Tech. Conv. Inst. of Brewing (Central and Southern 

African Section), Victoria Falls, p. 96. 

WASMUHT, K., STIPPLER, K. and WEINZIERL, M. (2002) Brauwelt Internat., 20 (5), 286. 

WILKINSON, N. R. (1993) Brew. Distill. Internat., May, 24, 12. 

WILKINSON, N. R. (2001) Brewers' Guard., 130 (4), 29; (5), 22. 

WILKINSON, N. R. (2003) Brewers' Guard., 132 (2), 24; (3), 20. 

WILKINSON, N. R. and ANDREWS, J. M. K. (1996) Ferment, 9 (4), 215. 

YAMAGUCHI, I., UEDA, T., UJIHARA, S., YAMADA, M. and FUKUSHIMA, S. (1997) Proc. 26th Congr. Eur. 

Brew. Conv., Maastricht, p. 257. 


7 

Hops 


Medieval ale (unhopped beer) rapidly went sour and turned into malt vinegar. Many 

herbs were used in attempts to prolong the shelf-life of such ale (Johnstone, 1997; Behre, 

1999) but only the hop, Humulus lupulus L., is used in large-scale brewing today 

although some microbreweries use other herbs. Detailed information about hops is found 

in abook by Neve (1991) and an earlier book by Burgess (1964). Abooklet, Hops and 

Hop Picking, by Filmer (1982) gives many illustrations. The European Brewery 

Convention has published aManualofGood Practice±Hops andHop Products (Benitez 

et al., 1997) and the Proceedings of two symposia on Hops (European Brewery 

Convention,1987and1994).Moir(2000)providedamillenniumreview.Hopsaregrown 

throughout the world, as illustrated in The Hop Atlas (Barth et al., 1994), solely to meet 

the requirements of the brewing industry (Table 7.1); the amounts used by herbalists and 

for hop pillows are negligible. Hops of commerce are the dried cones of the female plant 

but today much of the crop is processed into pellets and extracts. In Europe the yield of 

hops is usually expressed in zenters (1 zenterˆ50kgˆ110.23lb.) but in the USA, the 

yield is usually expressed in pounds. 

Although hops were probably used first for their preservative value, they introduced 

bitterness and apleasant flavour, which was liked, and which is the reason for their 

continued use. These flavours were found to originate mainly in the resins and essential 

oils found in the lupulin glands of the hop. The chemistry of the hop resins and essential 

oils is discussed in detail in Chapter 8but it is useful to note here that from the brewing 

standpoint the most important hop resins are the alpha-acids ( -acids) sometimes referred 

to as àlpha'. In conventional brewing hops are boiled with sweet wort for 1 1 

2 

2 hours 

during which time the -acids go into solution and are isomerized into the iso- -acids, 

the main bitter principles of beer. In open coppers the bulk of the essential oil constituents 

are vaporized during this period of boiling so brewers may add a portion of choice 

àroma' hops late in the boil to replace this loss. Alternatively, dry hops may be added to 

beer, either in cask or conditioning tanks, to introduce hop aroma ± a process known as 

dry hopping. 


Table 7.1 World production of hops and alpha acids ± 2002 

Country Hop acreage Production Alpha production 

(hectares) (metric tonnes) (metric tonnes) 

Australia 862 2384 316.9 

Austria 225 296 20.6 

Belgium 250 438 43.0 

Bulgaria 239 303 28.9 

China 5642 14167 1423.1 

Czech Republic 5968 6200 10.8 

Germany 18354 31500 2951.0 

France 814 1550 43.8 

New Zealand 406 884 95.3 

Poland 2197 1800 76.0 

Portugal 37 57 5.5 

Russia 862 440 56.6 

Slovakia 350 350 12.2 

Slovenia 1816 2200 151.0 

South Africa 500 965 117.5 

Spain 730 133 13.3 

UK ± England 1896 2653 242.0 

Ukraine 1778 1000 48.9 

USA 11864 25815 3140.0 

Yugoslavia 493 616 35.1 

2002 Totals 55058 93455 8810.0 

Source: November 2002 IHGC Report 

7.2 Botany 

Three species of Humulus are known, H. lupulus, H. japonicus and H. yunnanensis 

(Neve, 1991). H. lupulus is indigenous to and is cultivated in much of the Northern 

hemisphere between 35ë and 55ëNbut it is also cultivated in the Southern hemisphere in 

Australia, New Zealand and South Africa. H. japonicus is widespread in China and Japan 

but it lacks lupulin glands and therefore brewing value; it is sometimes grown as an 

ornamental garden plant. Little is known about H. yunnanensis from southern China. 

Humulus and Cannabis are the only two genera in the family Cannabinaceae; some 

authorities classify them in the Moraceae. Cannabis is represented only by C. sativa L. 

(Indian hemp, hashish, marijuana). There are some chemical similarities between hops 

and hashish but the resins of the two species are distinct; those of hops provide the bitter 

principles of beer while those of Cannabis include the psychotomimetic drug, 

tetrahydrocannabinol. Cannabis and Humulus spp. have been grafted on to each other 

but the characteristic resins do not cross the grafts. 

The hop is aperennial climbing plant; the aerial part dies off in the autumn but the 

rootstockstaysinthesoil, sometimesformanyyears.Theplantneeds asupportup which 

to grow. In the wild, hops are found in hedgerows but for cultivation they are trained up 

strings attached to permanent wirework. In the spring the stem tissue in the upper part of 

the rootstock produces numerous buds from which many shoots develop. The farmer 

selects the strongest shoots and trains them clockwise up the strings. As the bines climb, 

young flowering shoots develop in the leaf axils ±the so-called `pin' stage -which then 

form the young female inflorescence with papillated stigmas ±the `burr' (Fig. 7.1). From 

this the strobiles or hop cones develop. The cones consist of a central strig with bracts and 


(a) 

(b) 

(e) 

(e) 

Fruit 

(seed) 

(g) 

(c) 

1 cm 

Bract 

scar 

Lupulin 

glands 

bracteoles attached. Most of the lupulin glands are formed at the base of the bracteoles 

but they are readily detached and adhere to the bracts, strig and seed (Fig. 7.1). A few 

lupulin glands are found on the undersides of hop leaves but not enough to make these 

useful for brewing. The lupulin glands can contain as much as 57% of -acids and the 

sum of the ( ‡ )-acids is equal to 75 6% of the weight of the gland. The ratio / can 

range from 0 to about 4. The amount of resin/gland is fairly constant; the `high-alpha' 

varieties (see later) contain many more glands than the `low-alpha' varieties. It is 

predicted that the maximum lupulin content/cone that could be obtained by breeding is 

about 32% w/w which corresponds to a ( ‡ ) content of about 23% of the cone (Likens 

et al., 1978) New varieties with over 16% -acid have been bred. 

The hop is dioecious, male and female flowers are produced on different plants. Male 

flowers have five sepals and five anthers but since the flowers drop off after flowering 

any brewing value is lost. However, the male flowers produce pollen which can be 

carried long distances by the wind so any female plant in the vicinity will be fertilized 

and produce seeds at the base of the bracteoles. Despite many demonstrations that 

excellent lager beers can be produced with seeded hops, lager brewers do not like seeds 

so most varieties are grown `seedless'. In Europe this means that dried hops contain less 

than 2% w/w of seed; in the USA the limit is 3%. It was shown in England, as long ago as 

(f) 

(d) 

1 cm 

0.5 cm 0.1 cm 

Stipular 

bract 

Bracteole 

Fig. 7.1 Hop (Humulus lupulus L.) (a) young shoot; (b) male flowers; (c) `pin', young flowering 

shoot developing in the leaf axils; (d) `burr', young female inflorescence with papillated stigmas; 

(e) part of axis (`strig') of cone; (f) single mature hop cone; (g) bracteole with seed and lupulin 

gland; and (h) lupulin gland (After Burgess (1964) and Neve (1991) by kind permission of Kluwer 

Academic Publishers). 


(f) 

(h)

1908, that the yield/acre was higher if the hops were fertilized so English growers were 

encouraged to plant male hops in their gardens and by now many are wild in the 

hedgerows. Fertilized hops may contain as much as 25% w/w of seed. In contrast, in 

Germany, except in breeding stations, male hops must be removed by law. In the USA 

most varieties are grown seedless but in some areas males are planted. For example, a 

commercial yield of the variety Fuggle could not be obtained in the absence of male hops 

The bulk of the English hop crop contains seeds in excess of the European limit but in 

isolated areas male hops have been removed and the crop grown seedless. The hop plant 

is usually diploid with 20 chromosomes but triploid plants have been bred which are very 

infertile and have a low level of seeds even when pollinated. 

The growth of hops is strongly influenced by the amount of daylight. They require at 

least 13 hours of daylight for vegetative growth to occur; with shorter periods the plant 

becomes dormant. The plant must produce 20±25 nodes before being ready to flower 

when the days shorten. However, flowering will be inhibited if the days are too long. In 

South Africa, where the daylength is marginally too short, artificial light has been used to 

delay flowering and so improve the yield. Hops have also been grown in Kenya but here 

artificial light is necessary to allow vegetative growth as well as to delay flowering. A 

period of illumination in the middle of the night is probably more efficient than delaying 

nightfall. 

7.3 Cultivation 

Setting up a hop garden or yard requires considerable capital and the necessary wirework 

is not readily adaptable to any other crop so, once a garden is established, farmers will 

continue to grow hops even if it is barely profitable. The wirework must support the 

weight of the crop, probably in adverse weather conditions, so the corner posts must be at 

least 15 cm (6 in.) in diameter but the intermediate posts, every third or fourth hill, need 

not be so robust. Wooden posts are usually used but sometimes concrete or steel posts are 

more readily available. Before the advent of mechanical picking the wirework in English 

gardens was 3.75 4.25 m (12.5 14 ft.) high. At harvest the bine was cut down from the 

overhead wirework and the hops were picked by hand in the garden into coarse woven 

sacks called pokes. Hop picking in Kent was the annual holiday of many from the East 

End of London; in the West Midlands the pickers came from Birmingham. As the bine 

withered the nutrients therein returned to the rootstock. This was not possible with 

mechanical picking when the bine was cut down from the wirework and above the 

rootstock and transported in a trailer to a static picking machine Then, higher wirework 

was advantageous as in America (4.0 5.5 m, 13 18 ft.) and continental Europe 

(6.0 7.0 m, 20 23 ft.). 

In height-of-wirework trials at Wye College, Kent, the maximum yield of most 

varieties was with 5 m (16 ft.) wirework but some vigorous varieties showed increased 

yields at 5.5 m (18 ft.). As well as the height of the wirework, the layout and stringing 

patterns in hop gardens can show considerable regional variations. Nowadays the rows of 

plants are usually 2.8 3.2 m apart to allow tractors to pass freely. In Germany and 

continental Europe there is often a spacing of 1.5 m between plants or hills with two 

wires/hill. In the USA the spacing between rows and plants is about 2.25 m with two 

strings to each plant. In England practice differs between Kent and the West Midlands. In 

Kent the spacing between rows and plants within rows is 2 m but with the umbrella 

system of stringing there are four strings/hill. In Worcester the distance between rows 


may be as much as 3 m but within the rows the plants are only 1 m apart with two strings/ 

hill. One-year old plants will be used to lay out a new garden. 

Apart from breeding new varieties, hops are propagated vegetatively from `setts'. 

Three techniques are used to produce setts: (i) hardwood `strap cuts' are taken from the 

base of the rootstock in winter and allowed to root, (ii) layering, and (iii) mist 

propagation. For layering, a bine is allowed to grow until it is taller than the distance to 

the next hill. It is then taken down, laid along the ground, covered with soil, and the top of 

the bine trained up the next hill. In the autumn the hardwood bine is uncovered and cut 

up, each piece with a node, and planted out in nursery beds. With mist propagation, 

young growing shoots are taken and cut up each piece with two nodes and two leaves, 

rooted in sterilized peat at 21 ëC under an intermittent mist-like spray controlled by an 

èlectronic leaf'. Under these conditions the plant remains turgid and rooting takes 10±14 

days. After hardening off the plants are transferred to a nursery garden. It is obviously 

most important that only healthy plant material should be used to lay out a new garden. It 

should be free both of viruses and viroids. In order to reduce the risk of infection, 

commercial hop propagators are usually sited well away from the main hop growing 

areas. For example, in England, they are in East Anglia, well away from Kent and the 

West Midlands. 

When a garden is laid out a strong hook or peg is put into the ground near each hill to 

facilitate stringing. In England this is carried out by a man on the ground with a long pole 

and a continuous length of coir string. In bygone days stringing was carried out by men 

on stilts. In continental Europe cut lengths of soft wire are used. In the USA and Australia 

cut lengths of string are used and attached to the upper wire by men in a tractor-drawn 

tower. As might be expected the largest yield/plant was obtained with the widest spacing 

but the highest yield/hectare was obtained with closer spacing. The angle of slope of the 

string also influences the yield. In England the highest yields were obtained with a slope 

of 65ë but in Germany with a slope of 72±78ë. 

Traditionally, hop growing involved intensive cultivation but today, with the high cost 

of labour, `non-cultivation' is becoming increasingly popular. This involves controlling 

the weeds with herbicides (Paraquat in the autumn and Simazine in the spring). It was 

found that non-cultivation did less damage to the soil structure. As mentioned above, 

when the bines are about 0.5 m (18 in.) long they are trained clockwise up a string; 

usually two or three bines/string. Where practical, complete or partial self-training may 

be employed. Excess shoots may be cut off or removed with a chemical defoliant. The 

latter technique is also used to remove leaves and laterals from the lowest 1±3 m (3±6 ft.) 

of the bine. This helps to discourage mildew and red spider mites. The defoliants used 

include tar oil with sodium monochloroacetate. Paraquat and Diquat can be used when 

the bines reach the top wire but may damage younger plants. 

The hop is a deep-rooted plant which requires a good depth of soil, the pH of which 

should be kept above pH 6.5 by liming. The luxuriant growth of the plant makes heavy 

demands on soil nutrients which must be replaced. However, the present view is that 

earlier fertilizer regimens, for example, up to 225 kg N/ha, were excessive. It is 

recommended that, wherever possible, fertilizer treatments should be based on soil 

analyses and for nitrogen should not exceed 135 kg/ha. In Germany it is recommended 

that fertilizer treatments should be calculated from the amount of nutrient removed with 

the crop. The soil nitrate level should also be monitored. Hops contain up to 1.2% w/w of 

nitrate, which could significantly influence the level of nitrates in beer. Low levels of 

nitrates are desirable in beer because they can be reduced to nitrites, which can react with 

primary and secondary amines to produce carcinogenic N-nitrosoamines. It is also 


ecommended that phosphate and potash fertilizer treatments should be based on soil 

analyses and not exceed 300kg P2O5/ha and 450kg K2O/ha. With high residual levels no 

further treatment may be necessary. Soil analysis may also indicate the need for 

additional magnesium, 30±100kg/ha, to prevent the plants becoming chlorotic. Trace 

element deficiencies due to alack of zinc and/or boron have been observed in hops. 

In England and western Europe the water requirements of the hop crop are usually 

supplied by natural rainfall but elsewhere irrigation is often necessary. In the USA the 

crop requires 400±500mm (18±20in.) of rain in the Willamette valley, Oregon and 

760mm (30in.) in the Yakima valley, Washington State, which is supplied by furrow 

irrigation. Elsewhere, in Australia and in the Backa area of Serbia, overhead sprinkler 

systems are used. It is claimed that irrigation produces better hops, of more even quality, 

than natural rainfall. 

As the hop ripens in the Northern hemisphere, the first traces of resin can be detected 

in early August, the -acids afew days before the -acids, and resin synthesis is almost 

complete by the end of the month. De Keukeleire et al. (2003) have measured the 

formation and accumulation of -acids, -acids, desmethylxanthohumol and xanthohumol 

during flowering. Essential oil synthesis starts later and in some varieties resin 

synthesis may be complete before essential oil synthesis starts. Oxygenated compounds 

and sesquiterpenes are formed first but as the hop ripens the synthesis of myrcene 

becomes quantitativelythe most important process.Different varieties mature at different 

rates so agrower may choose amixture of early and late maturing varieties to spread 

picking over three to four weeks in September. 

The characteristics by which agrower decides that the crop is ready for picking are 

(Burgess, 1964): 

1. The bracts and bracteoles close towards the axis of the cone giving it acompact 

form. 

2. Thefullgrowthoftheterminalbracteole,whenseeded,causesittoprotrudefromthe 

top of the cone. 

3. The bracts and bracteoles become firm and slightly resilient. They rustle when 

squeezed in the hand and are rather easily detached from the axis. 

4. Thecolourofthebracteolesand,toalesserextent,thebractschangestoayellowishgreen. 

5. The contents of the seed become firm. The fruit coat (pericarp) becomes brittle and 

of apurplish colour. 

6. The lupulin glands are completely filled with resins. 

7. The aroma of the hop is fully developed. 

Hops should be picked as soon as possible after they become ripe; overripe cones tend to 

open and become more fragile and thus may be easily shattered by the wind, birds, or 

during picking. In all cases hops should be picked within ten days of ripening. 

As mentioned above, the bines are cut down from the top of the wirework and 1.2± 

1.5m(4±5ft.) from the ground and laid on atrailer for transport to astatic picking 

machine (Fig. 7.2). Here the bine is attached to atrackway and, depending on the design, 

enters the machine either horizontally or vertically. The hops and leaves are stripped from 

the bine by numerous moving wire hooks and then passed over various screens to 

separate the hop cones from unwanted debris. According to EEC regulations certified 

hops must not contain more than 6% of leaf and more than 3% of waste. The waste from 

the picking machine may be composted but should be burnt if there is any risk of disease. 

Mobile picking machines that can pick the hops in the garden have been designed but for 


Bine 

conveyor 

track 

Comb for 

disentangling 

bine 

Cones removed 

by flailing rotors 

Chopper for 

cutting up 

stripped bine 

Conveyor for 

cones, leaves 

and sprays 

removed by 

flails 


Revolving 

rollers for 

separating 

leaves 

Spray 

conveyor 

chain 

Cones 

removed 

from 

sprays 

Loose petals 

separated from 

leaves here and 

may be conveyed 

to final hop harvest 

Leaves 

separated 

from cones 

Main waste 

lateral 

conveyor 

Fig. 7.2 Hop-picking machine (Courtesy of Bruff Mfg. Co. Ltd.) 

Screener to 

separate small 

leaves, pieces 

of stem from 

the hops 

Lateral waste 

conveyor 

Supply of 

cones 

to bags 

Bagging 

off point

high wirework they are large, heavy and clumsy and have not been widely used. 

However, mobile picking machines have been successfully used with dwarf hops (see 

pages 249±50). Green hops contain about 80% w/w moisture and must be dried as soon as 

possible after picking. While waiting to be put on the kiln the hops must not be allowed to 

`sweat' as this will seriously reduce the quality of the crop. Although the vast majority of 

the crop is dried, some brewers in hop growing areas make seasonal brews using green or 

partially dried hops (feathered at 40% moisture) which have exaggerated hop flavours. 

7.4 Drying 

Most hops are dried on the farm to a final moisture level of about 10% w/w or less. 

However, in Germany the farmer dries his hops to about 14% moisture and then sends 

them, loosely packed, to a merchant. The merchants sorts and blends the hops and then 

completes the drying to the specified level. In a traditional English circular oast house the 

hops were spread on a horsehair cloth on a slatted floor 4.5 m (12±16 ft.) above a charcoal 

or anthracite fire. The drying floor is equipped with two sets of doors, the green hops are 

loaded on one side and the dried hops removed onto a conditioning floor on the other. 

Only natural draught, produced by tapering the roof to a cowl, was available and this was 

less than 0.1 m/s (19 ft/min) and only capable of drying a shallow (20±27 cm, 8±12 in.) 

bed of hops. 

Modern kilns are more likely to be rectangular, the air heated directly or indirectly with 

an oil burner and blown, or sucked, through the bed of hops with a powerful fan (Fig. 7.3). 

If an oil burner heats the air directly it is necessary to ensure complete combustion so the 

hops are not contaminated with unburnt oil. To avoid this a heat exchanger or stove may be 

used to heat the air. The air speed and temperature have to be carefully controlled 

throughout the drying as -acids are destroyed with increasing air temperature. When the 

hops are first put on the kiln with c. 80% moisture, evaporation cools them below the air 

temperature but as the hops dry they approach the air temperature. If the air speed is too 

low, the air may become saturated with moisture in the lower layers of the bed and then 

deposit moisture on the hops in the upper layer. This `reek' results in serious discoloration 

of the hops which will result in a poor hand evaluation. On the other hand, air speeds of 

STORE FOR BAGS 

OF GREEN HOPS 

Loading door 

Slide to retain hops 

Open slatted floor 

10 feet 

KILN 

Hops 

Thermometer 

bulb 

Sulphur 

pan Fan 

Louvres 

COOLING ROOM 

Unloading door 

Press 

Slide to retain hops 

Thermometer dial 

Combustion Pocket 

chamber 

Oil burner 

Fig. 7.3 Modern oast house (Burgess, 1964) 


Door 

Door

above 0.3 m/s (60 ft./min) are likely to blow hops off the bed (sometimes a wire mesh is 

used to retain them). In practice hops are usually dried with air between 60±80 ëC (140± 

176 ëF) and at a speed of 0.2±0.3 m/s (40±60 ft./min). 

For the most efficient use of fuel, the air leaving the bed of hops will be almost 

saturated with moisture. This is easily achieved when the hops are first put on the kiln but 

the removal of the last few percent of moisture is much less efficient. This does not 

matter where fuel oil is cheap, as in the USA where deep layers (up to 1 m, 40 in.) of hops 

are dried on large floors (e.g. 13.4 m 13.4 m, 32 ft. 32 ft.) at 60±65 ëC (140±150 ëF). 

In Europe, where fuel is more expensive, multi-layered kilns have been built with 

movable floors like venetian blinds. Green hops are put onto the top floor and when dried 

hops are removed from the bottom floor, the partially dried hops are dropped to a lower 

level and the top floor reloaded. Diagrams of such kilns are given by Neve (1991). 

However, such kilns are expensive to construct and a cheaper alternative with a single 

floor is to recirculate warm air towards the end of drying. In another method green hops 

are collected in large bins with wire mesh bottoms and the bins moved over hot air ducts 

at different temperatures. Continuous hop dryers have been built in Europe and again, 

diagrams are given by Neve (1991). 

It used to be normal English practice to burn sulphur in the kiln during the first 30± 

45 min. of drying. This caused the hops to assume a uniform bright yellow colour which 

was highly rated on hand evaluation. However, sulphuring has been shown to destroy - 

acids so the practice has been generally discontinued in England. In Germany merchants 

often burn sulphur in the final drying of the hops and in the USA some hops, but not all, 

are sulphured. The experience of the oast man in determining when the drying is 

complete is important. The remaining moisture is unevenly distributed making sampling 

for analysis difficult. Within the bed of hops there is more moisture in the upper layers 

than in the lower and in the cone most moisture is found in the strig. Accordingly the 

hops are removed from the kiln on to a conditioning floor where they are covered with 

cloths and the moisture allowed to equilibrate for several hours. 

In England hops are traditionally packed in jute/polypropylene sacks called pockets. 

They are usually about 2.1 m (7 ft.) long and 0.6 m (2 ft.) in diameter and hold about 76 kg 

(170 lb.) of hops. The empty pocket is suspended through a hole in the cooling room floor 

beneath a press with a circular foot, usually operated by an electric motor. The base of the 

suspended pocket is supported further by a strong canvas webbing belt. The cooled hops 

are pushed into the pocket with a canvas shovel called a scuppet. When the pocket is full of 

loosely packed hops the press is operated and, after compression, more hops are added. 

The process is repeated until the pocket is tightly packed with a density of 137±145 kg/m 3 

(8.5±9 lb./ft. 3 ). The pocket is then supported on the webbing belt while it is sewn up. In the 

USA, and increasingly elsewhere, hops are packed in rectangular bales measuring 137 

51 76 cm (4 ft. 6 in. 1 ft. 8 in. 2 ft. 6 in.). The baling press is essentially a steel box 

with detachable sides in which a ram operates. The box is lined with hessian and after the 

final filling a hessian cloth is placed on top. While the hops are compressed by the ram, the 

sides are removed and the hessian cloths sewed together. Bales are more expensive to 

produce than pockets but the hops are more densely packed (180 kg/m 3 , 13 lb./ft. 3 ) so they 

can be transported more efficiently. However, pockets can be rolled whereas bales have to 

be lifted. Before the advent of pellets bales were sometimes compressed to half their 

normal size for export. Most of the lupulin glands were ruptured by this treatment. 

Hops deteriorate on storage, in some cases significantly, before the next season's crop 

makes up 100% of the hop grist (70±100 weeks). The deterioration can be slowed either 

by cold storage at 0.20 ëC (33 ëF) and/or by storage in an inert atmosphere. For pockets or 


ales cold storage is the most practical but pellets are usually packed in an inert 

atmosphere. The pattern of deterioration shows alag period, which appears to be a 

varietal characteristic, probably due to the amount of antioxidants present (Lemusieau et 

al., 2001), followed by aperiod when the loss of both -and -acids can be fitted to 

either zero-order or first-order rate equations (Green, 1978). The chemical changes that 

take place during storage and the methods of chemical analysis are discussed in Chapter 

8. Samples for chemical analysis are taken with a`cork borer' sampler 70±80mm in 

diameter and 200±250mm long with a serrated edge at the cutting end. For hand 

evaluation asquare sample approximately 10cm 10cm 10cm (4in. 4in. 4in.) 

is cut from the middle of the pocket. From this sample the expert will verify that it is true 

to type and variety, assess the wholeness of the cones on the face and cut side of the 

sample and assess the aroma by rubbing the cones between the palms and inhaling. The 

stickiness of the sample gives an indication of the resin content. The expert will also 

determine whether any cone damage or discoloration is due to superficial mechanical 

damage or wind bruising as opposed to pests and diseases. In particular he will look for 

aphid and/or red spider mite infestation, powdery mildew, downy mildew or other 

diseases. Further, he will assess that the leaf and strig content is below the EEC limit of 

6%, that the moisture content is below the EEC limit of 12% and, from the brightness or 

dullness of the sample, determine if the hops had been picked under wet conditions or if 

the air flow in the oast had not been properly controlled. English hops are graded on the 

basis of hand evaluation. For alpha/aroma varieties (see pages 249±52) and high alpha 

varieties there are two grades: grade Ihops attract the base contract price but there is a 

deductionforgradeIIhops.Foraromahopsthereisalsoa`choicest'gradethatreceivesa 

premium above the base contract price. 

7.5 Hop products 

Whole hops are abulky, sticky product not suited to automated delivery into the copper. 

They only contain about 20% of useful brewing materials which are concentrated in the 

lupulinglands.Anyconcentrationoftheseactiveprincipleswillreducetransportandcold 

storage costs and give aproduct easier to store in an inert atmosphere. The main hop 

products are listed in Table 7.2. The non-isomerized products, added to the copper, are 

utilized (see page 271) no better than whole hops. Hop powders, although much denser 

than whole hops, are still sticky and unsuitable for automated dosing. They are not 

commercially available but are converted into pellets. 

7.5.1 Hop pellets 

For the preparation of hop pellets it may be necessary to dry the hops further so they 

contain 8±10% moisture. They are then cooled to 30 ëC and crushed in a hammer mill. 

For Type 90 pellets (normal hop pellets), the resulting powder (1 5 mm) is homogenized 

in an orbital screw mixer and then pelleted in a ring or horizontal die. Friction within the 

die will raise the temperature which should not be allowed to exceed 55 ëC. In the die the 

lupulin glands will be crushed so it is important that the pellets should be cooled and 

packed as soon as possible. The pellets are usually packed in metallized polyethylene 

laminate foils (0.1 0.15 mm thick) either under a vacuum (hard packs) or under an inert 

gas (nitrogen or carbon dioxide) at atmospheric pressure (soft packs). The packs are 

usually protected in cardboard boxes and are ideally stored at 1 5 ëC. Pack sizes range 


Table 7.2 Hop products 

Non-isomerized 

Double-compressed whole hops 

Hop pellets Type 90 


Stabilized hop pellets 

Solvent extracts: hexane 

ethanol 

liquid carbon dioxide 

supercritical carbon dioxide 

Isomerized hop products 

Isomerised hop pellets 

Isomerized kettle extract 

Isomerized hop extracts for post-fermentation bittering 

Reduced isomerized hop extracts: Dihydro- (rho)-iso- -acids 

Tetrahydroiso- -acids 

Hexahydroiso- -acids 

Hop oil products 


Oil-rich hop extract 

Pure hop oil: steam distillation 

molecular distillation 

Hop oil emulsions 

Fractionated hop oil 

Dry hop essences 

Late hop essences ± spicy, floral, estery and citrussy 

Miscellaneous 

Base hop fraction 

Purified beta fraction 

(European Brewery Convention Manual of Good Practice ± Hops and Hop Products, 1997) 

from 2 150 kg or may contain a specific weight of -acid so that the whole contents of a 

pack may be added to the copper without further weighing. The pellets are approximately 

6 mm 10 15 mm with a bulk density of 480 550 kg/m 3 (cf. whole hops at c. 140 kg/ 

m 3 ). Such pellets are called Type 90 because roughly 90 g of pellets are obtained from 

100 g of hops but 98% of the -acids (dry weight) are recovered. At the start of the 21st 

century over 50% of the hop crop was processed into pellets. At least one large hop farm 

in the USA processes the hop crop directly into pellets. 

For lupulin-enriched hop pellets (Type 45) the hop powder from the hammer mill is 

sieved at 30 ëC. The material that passes through a 0.3 mm sieve contains the lupulin 

glands and represents about half of the weight of hops used; the larger particles go to 

waste. The bulk density of the Type 45 pellets is similar to that of the Type 90 pellets but 

the Type 45 pellets contain twice as much of the brewing principles. By further sieving it 

is possible to obtain a fraction composed almost entirely of lupulin glands but this is not 

necessary for brewing purposes. 

Stabilized hop pellets are prepared by mixing up to 2% by weight of magnesium (or 

calcium) oxide with the hop powder before pelletization. In the die the -acids are 

converted into their salts which are more stable than the free acids. During storage the - 

acid salts may isomerise into iso- -acid salts which are better utilized than -acid salts. If 

stabilized hop pellets (in soft packs) are kept at 45±55 ëC for 10±14 days, the 

isomerization is complete and isomerized hop pellets are formed. 


7.5.2 Hop extracts 

Many different solvents have been used to extract the brewing principles from hops but 

brewers have become increasingly worried about the possibility of solvent residues in 

their beer so only hexane (b.p. 69ëC), ethanol (b.p. 78ëC), and carbon dioxide are still 

used and of these carbon dioxide is the most important. Although hexane is widely used 

in the food industry, the only hop extract plant using hexane is scheduled to close. 

Hexane is anon-polar solvent which will only extract soft resins and hop oil constituents 

from hops. In contrast, ethanol is miscible with water so an ethanol extract will also 

contain hard resins and polyphenols. After removal of the ethanol amixture of resins and 

ahot water extract is obtained but now only the resin extract is normally used as the hot 

water extract is rich in nitrates. 

At room temperature carbon dioxide is agas so it can only be used as asolvent under 

pressure. The phase diagram (Fig. 7.4) shows that both liquid carbon dioxide (below 

31ëC and 73 bar abs) and supercritical CO2 (above 31ëC and 73 bar abs) can be used for 

extraction. Liquid carbon dioxide has been used in England and Australia but elsewhere 

supercriticalcarbondioxide hasbeen thechoice.Despite claims madetothecontrary,the 

composition of the two extracts is very similar. The liquid CO2 extract is pale yellow and 

may contain fewer hard resins and polar bitter substances than the supercritical CO2 

extracts which are yellow to green in colour. Both are free of polyphenols and nitrates 

and contain few pesticide residues. Extraction with both liquid and supercritical CO 2is a 

batch process. Figure 7.5 is adiagram of aliquid CO 2 extraction plant. The extractor is 

charged with remilled hop pellets and, typically, carbon dioxide at 60±65 bar and 5±15 ëC 

is passed through the powdered pellets. The carbon dioxide is then evaporated from the 

extract at 40±50 ëC, condensed and returned to the extraction cycle. For supercritical CO2 

extraction the cooler in Fig 7.5 is replaced by a heat exchanger. Liquid CO2 at 60±70 bar 

is raised to the extraction pressure (200±250 bar) by a pump and the supercritical liquid 

raised to 40±60 ëC by a heat exchanger. After extraction the pressure on the extract is 

reduced to 60±80 bar before evaporation. The CO 2 is recovered, condensed and recycled 

as with liquid CO 2 extraction. Such CO 2 extracts can be used in the copper or as the raw 

material for isomerized extracts or molecular distilled oils. 

Pressure (bar abs) 

140 

120 

100 

80 

60 

40 

20 

Solid 

phase 

Triple point 

5.18 bar abs 

–56.6°C 

Liquid phase 

Liquid extraction 

Vapour 

phase 

Supercritical 

phase 

Supercritical 

extraction 

Critical point 

73 bar abs 

31°C 

0 

–100 –50 0 


50 100 

Fig. 7.4 Pressure temperature equilibria for carbon dioxide (Benitez et al., 1997). 


Extractor 1 

Cooler 

Pump 

Extractor 2 

Vapour phase 

carbon dioxide 

Condenser 

Extract 

Liquid phase carbon dioxide 

Evaporator 

Separator 

Liquid carbon 

dioxide buffer 

Fig. 7.5 Liquid carbon dioxide extraction with two extractors (Benitez et al., 1997). 

Hop extracts are usually packed in cans with food-grade linings. The cans can contain 

0.5±5.0 litres or aspecified weight of -acids. The use of glucose syrup or ahot water 

extract of hops to dilute resin extracts is now thought undesirable; such diluted extracts 

have amuch shorter shelf-life. For use, holes are punched in the cans which are then 

suspended (in abasket) in boiling wort. On alarger scale the extract is warmed to 

increase its mobility and then pumped into the copper (carbon dioxide extracts are less 

viscous than those prepared with hexane or ethanol). Isomerized and reduced hop 

extracts, designed for post-fermentation addition, are discussed in Chapter 8. 

7.5.3 Hop oils 

By definition, essential oils are volatile in steam so they are prepared by steam 

distillation. Usually minced hops are boiled with water and the condensate is collected in 

atrap which retains the oil and allows the aqueous phase to return to the boiler ±a 

process known as cohobation (the same technique is used in the analytical determination 

of hop oil). Such hop oil preparations have been available for many years, and some 

brewers used them in place of dry hopping, but they do not retain the true aroma of the 

hop and beers so treated could usually be distinguished from those which had been dry 

hopped. It is likely that some constituents are damaged at 100ëC and any constituents 

slightly soluble in water would be washed out in the trap. 

By steam distillation under reduced pressure (0.008mm Hg) at 25ëC, Pickett et al. 

(1977) obtained hop oil emulsions that were comparatively stable and imparted asound 

hop character to beer. However, hop oil constituents are soluble in liquid and supercritical 

CO2 with no risk of thermal degradation. The hop oil constituents are more soluble in 

liquid CO2 than the resins so extraction of a column of milled hops with 10±15% of the 

liquid CO2 needed for complete extraction will recover most of the hop oil with little of 

the resins. Similar preparations can be obtained with supercritical CO 2 but in practice it is 

better to aim for complete extraction. The pressure on the extract is then reduced to 100± 

120 bar when the hop acids are precipitated; the hop oil constituents stay in solution and 

can be recovered by evaporation. Hop oils can also be recovered from CO2 extracts by 

molecular distillation at low pressure (0.001 mm Hg). Such oils prepared from a single 

cultivar of hops retain the characteristic aroma of that cultivar and can be used, with a 


food grade emulsifier, to produce hop oil emulsions (normally with an oil content of 

0.25% v/v). The production of dry hop essences is discussed in Chapter 8. 

7.6 Pests and diseases 

Today many consumers are prepared to pay a premium for foodstuffs produced 

òrganically', that is without the use of agricultural chemicals. To meet this market 

some brewers are producing òrganic' beers from òrganic' barley and hops. However, 

hops are susceptible to attack by many pests and diseases so most hop growers need to 

use agrochemicals to produce acommercial crop. The agrochemicals which can be used 

are licensed by national or international bodies who also set amaximum allowable 

residue (MRL, mg/kg) for the chemical in the final product. To complicate matters the 

infecting organisms may develop resistance to the agent used so new agents have to be 

developed and approved periodically. Further, different strains of adisease may be 

susceptible to different agents. In general, more pests and diseases are found in the 

long-established hop-growing areas of Europe and North America than in the Southern 

hemisphere. 

7.6.1 Damson-hop aphid (Phorodon humuli Schrank) 

The most serious pest in the Northern hemisphere is the damson-hop aphid (Fig. 7.6) 

which overwinters as shiny black eggs in the bark of Prunus spp. (damson, sloe or 

plum). In early April wingless female insects hatch out and give birth to live young 

which rapidly multiply. After several generations winged females (alatae) arise which 

migrate to the hop when the flight threshold temperature of 13 ëC (55 ëF) is reached in 

late May or early June. The migrating alatae may not colonize the first hop plant on 

which they land and may show a varietal preference. However, they feed and 

reproduce mainly on leaves on the top of the bine. They insert their long stylets into 

the phloem strands of the leaves to feed which weakens the leaves and causes 

defoliation. The most serious situation is when the alatae enter the cones. In addition 

the honeydew that the aphids produce supports the growth of sooty moulds, which 

lower the value of the hops on hand evaluation. More heavily infected cones will turn 

brown and limp and will probably shatter during machine picking. Those hops that 

survive will not find a ready market. Any aphids that survive until September± 

October, when the light period falls below 13.5 h/day, form winged females which 

Fig. 7.6 Damson-hop aphid (Phorodon humuli Schr.). 


(a) Insecticides/Acaricides 

R1 

R2 

F3C 

C C 

H 

H 

H3C CH3 

(7.1) R1 = R2 = Cl, R3 = H 

(7.2) R1 = R2 = Cl, R3 = F 

(7.3) R1 = R2 = Br, R3 = H 

CO2 C 

H 

(7.4) R1 = CF3, R2 = Cl, R3 = H 

Cl 

H3C 

H3C 

H3C CH3 

C C 

H 

H 

H3C CH3 

N 

CO2 

CO2 

CN 

C 

H 

CN 

H 

CH2 

migrate back to Prunus spp. The males follow later and the fertilized females lay the 

eggs which overwinter. 

Control includes removing Prunus sp. from nearby hedgerows but the alatae can fly 

considerable distances. It is important that the grower knows when migration onto the 

hop starts so that he can maximize his control measures. Usually this involves spraying 

with suitable insecticides such as the synthetic pyrethoids. Cyfluthrin, Cypermethrin, 

Deltamethrin, Fenpropathrin, Lambda-cyhalothrin, or Imidacloprid (Fig. 7.7). For many 

years organo-phosphorus insecticides were used but in Europe most aphids have 

developed resistance to these agents but they may still be effective elsewhere. As 

mentioned, most brewers are not happy with the risk of agrochemical residues in their 

R3 

Cypermethrin 

Cyfluthrin 

Deltamethrin 

(7.5) Fenpropathrin 

H 

(7.6) Bifenthrin 

Cl CH2 N 

O 

Lambda-Cyhalothrin 

O 

CH3 

N·NO2 

(7.7) Imidacloprid (Admire) 

Fig. 7.7 (a) Insecticides/acaricides, (b) herbicides and (c) fungicides accepted for use on hops 

(British Beer and Pub Association, Technical Circular No. 376, 2003). 


NH


raw materials, so there is considerable interest in the biological control of aphids using, 

for example, ladybirds (Coccinellidae) but such control is never complete. Hops with 

resistance to aphids have now been bred and are being evaluated. The ASBC describe 

methods for estimating the number of aphids in a sample of hops. 



7.6.2 (Red) Spider Mite (Tetranchus urticae Koch) 

The spider mite is widespread and flourishes in hot dry conditions. The bright red females 

overwinter in the soil, under leaves or in cracks in hop poles. In the spring they climb the 

bines and suck sap from epidermal and sub-epidermal cells. They lay small translucent 

eggs and the mites that emerge are greenish-yellow with black markings (two-spotted 


mite).Thefirstsignofmiteattackisasilveryspecklingofthehopleaves.Smallnumbers 

of mites do not do much damage but severe infestation may result in loss of crop. When 

hops were first treated with organo-phosphorous insecticides there was good control of 

both aphids and mites but later both species developed resistance. The acaricides in use 

today include Bifenthrin and Propargite. 

7.6.3 Other pests 

Atleast40insectspecieshave beenfound living onthehopbut,with theexceptionofthe 

species discussed above, they do not usually cause serious damage. Other pests include 

the dagger nematode (Xiphinema diversicaudatum), which may be the vector for the 

Arabis Mosaic Virus (AMV), the hop-root eelworm (Heterodera humuli), clay-coloured 

weevils (Otiorrhynchus singularis), the rosy rustic moth (Hydroecia micacea), the flea 

beetle (Psylliodes attenuata Koch), earwigs (Forficula auricularis), wireworms 

(Agricotes spp.), and slugs (Agriolimax reticulatus and Arion hortensis) (Neve, 1991). 

7.6.4 Downy Mildew (Pseudoperonospora humuli (Miyabe and Tak.) G. W. 

Wilson) 

Worldwide, downy mildew is probably the most serious disease of hops. It was first 

observed in Japan in 1905, in the USA in 1909, and in England and Europe by 1920. It is 

found in most hop-growing areas of Europe and North America but, by applying strict 

quarantine precautions, has been kept out of Australia, New Zealand, and South Africa. 

The fungus survives the winter in the rootstock of the hop plant. In spring, when the hop 

produces numerous shoots, it migrates to infect some of the shoots so that they do not 

develop and produce stunted basal spikes which are the characteristic symptom of the 

disease (Fig. 7.8). On the undersides of the leaves of these spikes numerous black spores 

(conidia) develop which readily infect other young leaves (Fig. 7.9). Under wet 

conditions the conidia germinate on the leaves and produce motile zoospores which enter 

the plant through open stomata. Such infection causes black angular spots on the leaves. 

Infection of the growing tip of the bine will cause extension to cease and the formation of 

a basal spike. If the zoospores infect the flower or `burr' no cone will develop. If they 

enter a developing cone serious losses will occur; some of the bracts and bracteoles will 

turn brown giving the cones a variegated appearance and lower the value of any crop. 

Control involves the early removal of all basal spikes and lower leaves with infecting 

conidia. If this is done by hand all infected material should be burnt but by spraying with 

defoliating chemicals (anthracene oils, Diquat, Paraquat, or sodium monochloroacetate) 

any spikes will be killed in situ. Before downy mildew appeared on hops, copper 

fungicides, such as Bordeaux mixture, were widely used to control mildew on grapes, and 

such treatments were readily applied to hops. Later, copper oxychloride was found to be 

less phytotoxic than Bordeaux mixture. In the 1950s organic dithiocarbamates such as 

Zineb were favoured but the Pesticide Safety Directorate has revoked the approval of 

such chemicals. The most effective fungicides in use today are Metalaxyl (Ridomil) and 

Fosetyl-aluminium (Aliette). These can be used as foliar sprays or as a soil drench when 

they are capable of eliminating the fungus from the rootstock. However, this latter 

method of application has been discontinued in England in fear that the fungus was 

developing resistance to Metalaxyl. Metalaxyl is applied as a foliar spray in admixture 

with copper oxychloride. Other approved fungicides include Chlorothalonil, Fenpropimorph 

(Corbel), Myclobutanil, and Peconazole (Topas). Despite the fact that Metalaxyl 


Healthy 

shoot 

Earth 

Basal 

spike 

Fig. 7.8 Hop plant infected with downy mildew (Pseudoperonospora humuli). Left: healthy 

shoot. Right: basal spike. 

50 μm 

Fig. 7.9 Downy mildew (Pseudoperonospora humuli). Sporangiophore with sporangia (Burgess, 

1964). 

can eliminate the fungus from the rootstock, it is good practice to dig up and burn any 

rootstock that has been infected and replace it with healthy material. Varieties with 

resistance to downy mildew are being bred. 

7.6.5 Powdery mildew (Sphaerotheca macularis (DC.) burr) 

Mould, white mould, red mould or powdery mildew are synonyms for this fungal disease 

that appears as white pustules on the leaves. Towards the end of the season red 

overwintering spores (perithecia) are formed. The disease has little effect on the 


vegetative growth of the host but serious losses occur if the cones are infected. The 

disease is probably more troublesome when non-cultivation is practised and plant debris 

infected with perithecia is left on the ground. Control involves treatment with sulphur, 

before burr, Triadimefon (Bayleton) and Bupirimate (Nimrod). Triforine (Saprol), which 

may reduce the yield, is usually reserved for serious infections. 

7.6.6 Verticillium Wilt (Verticillium albo-atrum Reinke and Berth) 

This disease, first detected in the Weald of Kent in 1924, is serious in England and 

Germany. It was thought, at first, to exist in two forms: mild (`fluctuating') and severe 

(`progressive') but the present view is that there is a continuum of strains giving infection 

ranging from very mild to very severe. There is no simple test, chemical or biological, to 

distinguish between the strains except by their behaviour towards a susceptible cultivar, 

e.g., Fuggle. Such tests, whether outside or in growth chambers, are of necessity slow. The 

fungus occurs in the soil either as spores (condia) or as mycelium on infected debris (Fig. 

7.10). It enters the hop roots and the dark coffee-coloured mycelium spreads through the 

500 μm 

10 μm 

10 μm 

(c) 

(d) 

(b) 

(a) 

50 μm 

Fig. 7.10 Verticillium wilt (Verticillium albo-atrum). (a) conidiophore; conidial heads and 

conidia; (b) groups of conidiophore; (c) `dark' mycelium; (d) conidia (Burgess, 1964). 


vascular system of the plant to the leaves. The infected leaves develop yellow patches and 

black necrotic areas between the veins giving the so-called `tiger stripe' effect before they 

fall off. Another symptom is that the lower 1.2±1.5 m (4±5 ft.) of an infected bine becomes 

swollen and may become detached from the rootstock. The wood of the infected bine 

shows brown areas due to the mycelia. In `fluctuating' wilt the brown colour is restricted 

to the centre of the bine and the hill will probably survive with little or no increase in wilt 

the following year. In severe `progressive' wilt the symptoms appear earlier in the season 

and are readily transferred to other plants in the garden especially in the direction of 

cultivation (such observations helped to popularize non-cultivation). Infected plant debris 

is readily carried on the wind, on boots and machinery to spread the disease. 

At present no form of wilt responds to chemical agents so the only control measures 

are hygiene and the breeding of wilt-tolerant varieties. The Progressive Wilt of Hops 

Order (1947) attempted to restrict the disease to the Weald of Kent where wilt-tolerant 

varieties could be planted. Elsewhere, in èradication' areas, only susceptible varieties 

could be grown and any outbreak of the disease had to be notified, the infected plants and 

those nearby had to be grubbed and all infected materials burnt. The infected area was 

then fenced off and kept under grass for several years. Wilt-tolerant varieties must not be 

grown in èradication' areas as they may be symptomless carriers. These strict measures 

managed to keep the disease in check in East Kent and Hampshire and, for a time, the 

West Midlands. Later, wilt spread rapidly through the West Midlands so wilt-tolerant 

varieties may now be planted there on clean land. The wilt-tolerant varieties are discussed 

below. Other fungal diseases which infect hops from time to time include Fusarium 

canker (Fusarium sambucinum Fuckel), black root rot (Phytophthora citricola Sawaba), 

grey mould (Botrytis cinerea Pers.), Alternaria alternata, black mould (Cladosporum sp.) 

and armillaria root rot (Armillia mellea mellea (Fr.) Quel.). 

7.6.7 Virus diseases 

In the past, virus diseases caused considerable damage to hops. Nettlehead was perhaps 

the most serious, the symptoms of which were well known before viruses had been 

recognized. In 1966 Bock identified arabis mosaic virus (AMV) and necrotic ringspot 

virus (NRSV) in hops. All hop plants with nettlehead contained AMV but not all plants 

infected with AMV developed nettlehead. It was found that a satellite of low molecular 

weight nucleic acid (SNA) was necessary as well as AMV to produce nettlehead. It was 

also found that the soil-borne dagger nematode (Xiphinema diversicaudatum) was a 

vector for AMV. Land infected with nematodes should be fumigated with dichloropropene 

and left fallow for two years before replanting with hops. 

Hop mosaic virus (HMV) is a carlavirus transmitted by aphids feeding on one plant 

and then moving to another. In English Goldings it causes severe stunting, the bines fall 

away from the strings, and the leaves show translucent banding along the veins and curl 

downwards. Most other varieties are symptomless carriers of the virus and there is no 

evidence that HMV causes any reduction in yield with them. Nevertheless, Goldings 

should not be planted with, or adjacent to, other varieties. Other carlaviruses are the hop 

latent virus and the American hop latent virus. The Prunus necrotic ringspot virus 

(NRSV) rarely produces recognizable symptoms in hops but laboratory tests showed that 

virtually 100% of the older varieties were infected. Comparison of yields from virus-free 

and infected plants showed that the infected plants contained 30% less -acid. 

Enzyme-linked immunosorbent assay (ELISA) tests have greatly facilitated the 

detection and characterization of viruses in hops. Virus can be eliminated from hop plants 


y a combination of heat treatment and meristem-tip culture. Multiplication of viruses is 

inhibited by maintaining the plants at high temperature and if the tip of the apical shoot is 

dissected out of a plant and grown under sterile conditions, the regenerated plants are 

usually virus-free. By such techniques virus free clones of most varieties are available to 

hop propagators so most new planting material, with A+ certification, is virus free. 

Viroids consist of a small piece of RNA without a protein coat so they cannot be 

detected serologically. At least two viroids have been detected in hops. The hop stunt 

viroid has been found only in Japan but the hop latent viroid was found to be widespread 

in Germany and England; the English variety Omega was particularly susceptible. 

However, virus- and viroid-free planting material is now available for most cultivars. 

7.7 Hop varieties 

The first cultivators of hops selected the most vigorous wild plants they could find. If 

their neighbours' plants gave better yields they would seek planting material from them 

so that, in time, the same cultivar would be grown in a given area. As transport improved 

these areas grew larger. Indeed, German hops were named after the area where they grew, 

e.g., Hallertau, Tettnang, Spalt and Hersbrucker. Early growers only had small kilns or 

oasts to dry their crop so they were interested in having early and late varieties to 

lengthen the harvest period, e.g., Hallertau mittelfruÈh, Hersbrucker spaÈt. 

An early rhyme suggests that: 

Hops, Reformation, Bays and Beer 

Came to England in One Bad Year 

and that year was 1524 when Flemish immigrants brought hops to Kent. However, the 

finding of hop residues in a boat found in Graveney Marsh and dated c. AD 949 suggests 

they were known earlier (Wilson, 1975). Many different selections were grown in 

England. According to Burgess (1964) the Hops Marketing Board classified Amos's 

Early Bird (selected in 1887), Bramling (before 1865), Cobbs (1881), Eastwell Golding 

(before 1889), Petham Golding, Rodmersham Golding (1880), Mathon (1901), and 

Canterbury Goldings as Goldings and Tutsham and Whitbread's Golding Variety (WGV 

or 1147) as Golding varieties. Before genetic fingerprinting it was not possible to confirm 

how similar these varieties were. British brewers preferred Goldings but they were 

susceptible to mosaic viruses and the variety introduced by Richard Fuggle in 1875, 

which bears his name, was not. So Fuggles became popular with growers and, to a lesser 

extent, with brewers and in 1949 made up 78% of the English hop acreage. Goldings 

contain slightly more -acid than Fuggles and have a better aroma; Fuggles was used in 

the copper and Goldings for dry hopping. Brewers were usually prepared to pay a 

premium for Goldings. 

The early settlers introduced hop growing along the eastern seaboard of North 

America but from 1900 it moved to the North-West States of Washington, Oregon, 

California and Idaho. The main variety grown was Clusters (Early and Late) which 

contained more resin than Goldings or Fuggle but had a strong aroma described as 

`blackcurrant' or `tom-cat' which was disliked by most British brewers. Nevertheless 

some British brewers imported Clusters, and Salmon at Wye College set out to breed 

hops with more resin but English aromas. From an open pollinated wild hop from 

Manitoba (BB1) he obtained Brewer's Gold and Bullion which gave excellent yields and 

were richer in -acids than other varieties then available. A cross between a male 


seedling of Brewer's Gold and a Canterbury Golding gave Northern Brewer. These three 

new varieties were planted worldwide, in particular Bullion was grown in Oregon and 

Brewer's Gold in Germany. 

Later, Northern Brewer became important in Germany because of its resistance to the 

German strain of Verticillium wilt and by 1978 these three varieties accounted for 47% of 

the German hop acreage. They were also grown in Belgium, Bulgaria, Spain and what 

was East Germany. They were less popular in Britain because of their Àmerican' aroma. 

The Germans also maintained that their traditional varieties had better aromas than the 

new varieties and this led to the classification of hops as either Àroma' hops or `Bitter' 

hops. The latter usually had higher -acid contents and were put into the copper at the 

beginning of the boil when any unpleasant volatiles would be evaporated. The Àroma' 

hops were reserved for late addition or dry hopping and usually commanded a higher 

price than `Bitter' hops. However, there is no reason why a high -acid hop should not 

have a good aroma. Salmon bred many more new varieties of hops, some of which were 

grown commercially for a time encouraged by the Association of Growers of New 

Varieties of Hops (1944±1998). They include Concord, Brewers' Standby, Early Choice, 

Copper Hop, Quality, College Cluster, Brewers' Favourite, Sunshine, Malling 

Midseason, and Norton Court Golding. These, and many only assigned code numbers, 

served as the parents of the next generation of new varieties. 

The devastation caused by Verticillium Wilt in Kent prompted an urgent need for wilttolerant 

varieties. From among Salmon's seedlings Keyworth's Early, Keyworth's 

Midseason and Bramling Cross were found to be wilt tolerant but they were never very 

popular. The Whitbread Golding Variety was also wilt tolerant and was more acceptable. 

However, conservative brewers wanted a wilt-tolerant replacement for Fuggle and 

Density, Defender, and Janus were introduced in 1959 to meet this need. Again they were 

not very popular but Progress and Alliance, introduced in 1966, were more acceptable 

and Progress is still grown today. However, by this time there had been many brewery 

amalgamations and the accountants had more say in the boardrooms. They knew that - 

acids, which could be measured, produced bitterness so they wanted high-alpha hops to 

produce bitterness as cheaply as possible. Aroma, which could not be easily measured, 

was much more subjective. Wye Northdown and Wye Challenger, introduced in 1971, 

had higher levels of -acid and good aroma but were susceptible to wilt. Wye Target, 

however, was wilt tolerant, immune to powdery mildew, and contained up to 12% -acid. 

In the early 1990s it accounted for almost 50% of the English hop acreage. Wye Viking 

and Wye Saxon were not successful; Yeoman was popular at first but had declined by 

1998 as had Zenith and Omega. By the end of the 20th century most hop-growing 

countries were breeding `super-alpha' hops with more than 15% -acid. Wye's 

contribution was Phoenix and Admiral, released in 1995. `Super-alpha' hops were wanted 

because it is more economic to process them than those with less -acid and 60% of the 

world hop crop is processed today. However, unless the world beer market increases 

dramatically, there will be no increased demand for -acid so the growth of high-alpha 

hops will reduce the amount and area of hops required further. 

Most commercial hops are grown on wirework up to 7 m (23 ft.) high which is 

expensive to set up and maintain. Towards the end of the 20th century there was 

considerable interest in growing hops on a low trellis, not more than 3 m (10 ft.) high. The 

hops grow up plastic netting to produce a hedge which, besides being cheaper to set up, 

simplifies cultivation, spraying, and harvesting. Machines have been developed which 

straddle the rows and, in particular, pesticides can be applied in a much more controlled 

manner resulting in savings and much less environmental damage. In addition, crop 


otation is easier with the cheaper support systems. Conventional varieties can be grown 

on these low trellises but special dwarf varieties, bred at Wye, are more successful. Dwarf 

hops contain a special gene which produces short internodal distances so the plants are 

half the height of conventional plants. The dwarf varieties released so far include First 

Gold (an aroma variety), Herald (high-alpha), Pioneer (a semi-dwarf dual purpose hop) 

and Pilot. The yields/hectare from these low trellises are comparable with those from 

standard wirework. 

In Germany traditional varieties were grown in five areas: Hallertau (85% of the 

German hop area), Spalt, Hersbruck, Tettnang and Jura. Since 1992 Jura has been 

incorporated into Hallertau and Elbe-Saale is a collective name given to the hop-growing 

areas of the former German Democratic Republic (Barth, 1999). Later, in addition to the 

traditional varieties, foreign varieties, Brewer's Gold, Northern Brewer and Record (from 

Belgium), with higher levels of -acid were also grown. Breeding was originally started 

at HuÈll, in 1926, to produce varieties resistant to downy mildew which retained the 

traditional aromas. The outbreak of wilt in the 1950s destroyed much of the major variety 

Hallertau mittelfruÈh but Northern Brewer and Hersbrucker spaÈt showed resistance to the 

German strain of Verticillium. Breeding continued to produce three new varieties: HuÈller 

Bitterer (commonly known as HuÈller), Hallertauer Gold, and Perle. HuÈller and Perle were 

resistant to both downy mildew and wilt and had higher levels of -acid than the 

traditional varieties. Hallertauer Gold was not resistant to wilt but both it and Perle were 

judged to have the same number of aroma fineness points as the traditional varieties. 

Taurus and Magnum are the latest German super-alpha varieties. The Saaz (Zatec) hop 

grown in Czechoslovakia is renowned for its fine aroma. The low yields it produces have 

been improved by clonal selection rather than breeding. The Saaz hop is thought to be 

related to the German varieties Tettnang and Spalter and the Japanese Shinshuwase. Hops 

were imported from England to what was Yugoslavia and the Savinga (Styrian) Golding, 

imported by some brewers into England, is identical with a seedless Fuggle. Yugoslavian 

hop breeding produced the `Super Styrian'; the high-alpha hops Atlas, Apolon, Ahil, and 

Aurora. In Slovenia Blisk, Bobek and Buket have been produced and in the Backa region 

three more high-alpha hops were raised from Northern Brewer: Neoplanta, Vojvodina 

and Dunav. 

In the United States also the varieties grown have changed in the last twenty years. 

Eighty per cent of the US hops are grown in the Yakima valley Washington State, where 

Clusters (Early and Late) were the main variety and the standard kettle hop. The English 

varieties Fuggle, Bullion, and Brewers' Gold were also grown; the last two of these 

varieties were sometimes classified together as Ènglish'. Clonal selections were made 

from Clusters and Talisman was released in 1968. Cascade, the first aroma hop from the 

US breeding programme, was released in 1972 and Willamette and Columbia in 1976. 

These last are triploid seedless aroma hops and Willamette still accounted for 14% of the 

US hop acreage in 2002. In the same year the high-alpha varieties Galena, Nugget, 

Columbus-Tomahawk, Zeus, Millenium and YCR-5 (Warrior) made up over half of the 

American hop area. Clusters accounted for only 3% but Eroica, Olympic and the English 

varieties had almost disappeared. 

Hops are grown in Australia in Victoria and Tasmania and in New Zealand around 

Nelson on South Island. Being geographically isolated the hops produced in these 

countries are free of most of the pests and diseases found in the Northern hemisphere. 

Only the two-spotted mite (Tetranychus urticae Koch) and the red spider mite 

(Panonychus ulmi) occasionally give trouble. Thus, these countries can produce òrganic' 

hops, without the use of pesticides, more easily than anywhere else in the world. Such 


hops from New Zealand are termed `Bio-Gro'. The major variety grown in Australia was 

Pride of Ringwood, bred by Nash from Salmon's Pride of Kent, which at one time 

occupied 90% of the Australian hop acreage. It is now being replaced by atriploid Super 

Pride and two super-alpha hops Opal and Victoria. Aroma hops occupy only 4% of the 

Australian hop area. In contrast, they occupy almost 50% of the New Zealand acreage 

(Inglis,1999).NearlyallthehopsgrowninNewZealandaretriploids,includingthedualpurpose 

hops Green Bullet, Sticklebract, Super Alpha, Southern Cross and Pacific Gem. 

The European aroma hops do not grow well in New Zealand but from Hallertau 

mittelfruÈhtwo new aroma varieties more suited to New Zealand conditions have been 

bred, NZ Hallertau Aroma and Pacific Hallertau. In Japan the traditional variety 

Shinshuwase was replaced by Kirin No. 2which, in turn, is being replaced by the superalpha 

varieties Toyomidori, Kitamidori and Eastern Gold. 

The characteristics of the main varieties are collected in Table 7.3. These data were 

collected from many sources and may not be strictly comparable. No doubt in twenty 

years time different varieties will be grown. Plant breeders continue to seek new varieties 

with increased resin content, increased disease resistance and better yields. Other goals 

relate to resin quality. Some brewers think that humulone gives a better bitter flavour than 

cohumulone so require hops in which the proportion of cohumulone in the a-acids is as 

low as possible. Some cultivars deteriorate on storage more rapidly than others and, since 

all hops cannot be processed immediately after harvest, good storage stability is 

desirable. Although many new varieties show resistance to fungal diseases most growers 

still have to use pesticides to control aphids and mites. New breeding programmes have 

produced aphid-resistant hops that can be grown òrganically'. The hop is a long-day 

plant which grows between 30 and 55ë of latitude; when grown nearer the equator 

artificial illumination is necessary. Countries which grow hops under these conditions are 

trying to breed varieties adapted to the shorter day length. 

The unambiguous characterization of hop cultivars is difficult although methods based 

on morphology and chemical analysis usually give good indications. However, methods 

based on DNA analysis reflect the genotype of the cultivar irrespective of the stage of 

plant development, the environmental or disease status. Briefly a sample of hop DNA is 

subjected to a random amplified polymorphic DNA (RAPD) process using arbitrary 10 

mer primers. The products are then separated by electrophoresis followed by staining, 

ultraviolet visualization and DNA sequencing. Using this method Murakami (2000) 

produced a dendrogram, based on genetic distance, which resolved most of the common 

varieties into six clusters; most of the high-alpha hops were in the first cluster. 


Table 7.3 Properties of principal hop cultivars a 

Cultivar -acid (%) = Cohumulone Oil (%) Humulene/ Comments 

ratio (%) Caryophyllene 

ratio 

Australia 

Opal 13.0 3.2 30 1.5 2.5 

Pride of Ringwood b 

9.0±11.0 1.7 33 2.0 0.1 

Super Pride 13.9 2.4 27 1.0 0.4 

Topaz 11.5 2.0 40 1.0 0.2 

Victoria 11.0±14.0 2.1 38 1.1 1.6 

Belgium 

Record 5.5±8.5 1.0 30 1.8 2.5 

Czech Republic 

Bor 6.5±11 1.8 25 1.5 3.2 Dual 

Premiant 7.0±11.0 1.9 22 1.5 3.0 Dual 

Saaz 3.0±4.5 0.9 26 0.4 3.5 

Sladek 4.0±8.0 1.0 28 1.5 2.4 Aroma 

England c 

Admiral 13.5±16.2 2.6±3.2 26±32 1.0±1.7 Wilt tolerant 

Bramling Cross 6.0±7.8 2.4±3.1 26±31 0.7±1.0 2.2 Wilt tolerant 

Brewers Gold 5.5±8.5 1.9 38 1.5 2.3 

Bullion 6.0±9.0 1.9 36 3.2 1.5 

First Gold (Dwarf) 5.6±8.7 2.4±3.2 29±34 0.7±1.4 3.2 Wilt tolerant 

Fuggle 3.0±5.6 1.5±2.2 29±30 0.7±1.1 3.3 

Goldings 4.4±6.7 2.1±2.6 26±32 0.8±1.0 3.5 

Herald (Dwarf) 11.0±13.0 2.4 37 1.0±2.2 2.4 Wilt resistant 

Northern Brewer 6.5±10.0 2.0 23 2.0 2.8 

Phoenix 12.0±15.0 2.1±2.6 24±28 1.2±2.5 Wilt resistant 

Progress 6.0±7.5 2.8±3.3 27±36 0.5±0.8 3.3 Wilt tolerant 

Whitbread Golding 

Variety (WGV) 5.4±7.7 2.3±3.0 32±43 0.8±1.2 3.5 Wilt tolerant 

Wye Challenger 6.5±8.5 1.8±2.1 20±25 1.0±1.5 3.1 

Wye Northdown 6.8±9.6 1.5±2.2 24±29 1.2±2.2 2.7 

Wye Target 9.9±12.6 2.2±2.8 35±39 1.2±1.4 2.4 Wilt resistant 

France 

Strisselspalt 2.0±5.0 1.0 24 0.7 2.4 

Germany d 

Hallertau mittelfruÈh 3.5±5.0 1.0 21 0.6±1.0 3.7 Aroma 

Hallertau tradition 5.0±7.0 1.2 26 1.2±1.4 3.8 Aroma 

Hersbrucker spaÈt 3.0±5.0 0.9 25 0.6±1.0 2.5 Aroma 

HuÈller (Bitterer) 5.0±7.0 1.2 30 1.0±1.3 1.9 

Magnum 12.0±14.0 2.6 26 1.6±2.1 3.6 Bitter 

Perle 6.0±8.5 1.6 29 1.0±1.3 2.6 Aroma 

Spalter 4.0±5.0 1.1 24 0.6±1.0 3.3 

Spalter 4.0±5.5 1.3 24 0.6±1.0 2.0 Aroma 

Taurus 12.0±15.0 3.0 24 1.2±1.5 3.7 Bitter 

Tettnang 3.5±5.0 1.0 25 0.4±1.0 2.7 Aroma 

New Zealand e 

Green Bullet 12.5±13.5 1.8 42 0.8 3.2 

NZ Hallertau aroma 8.5 1.3 29 1.25 2.0 

Pacific Gem 14.0±16.0 1.8 41 1.5 3.2 

Pacific Hallertau 6.0 1.0 26 1.26 2.9 

Southern Cross 11.0±12.0 1.9 27 1.2 3.4 

Stricklebract 13.5±14.5 1.7 38 1.0 3.2 

Super Alpha 12.5±13.5 1.5 38 1.5 3.2 


Table 7.3 (Continued). 

Cultivar -acid (%) = Cohumulone Oil (%) Humulene/ Comments 

ratio (%) Caryophyllene 

ratio 

Poland f 

Limbus 5.3 ± 36 1.7 2.2 Aroma 

Lubekski 4.0 ± 31.8 1.2 5.1 Aroma 

Lublin 3.5±4.5 1.3 27 1.0 3.7 

Marynka 11.1 ± 25 2.3 5.0 Bitter 

Oktawia 10.6 ± 34 1.6 4.4 Bitter 

Sybilla 7.3 ± 34 1.7 3.8 Bitter 

Zbyszko 8.5 ± 26 1.0 2.5 Bitter 

Slovenia 

Ahil 8.0±10.0 2.2 27 1.0 2.4 

Apolon 8.0±10.0 2.2 27 1.0 2.5 

Atlas 8.0±10.0 2.2 31 0.8 2.3 

Aurora 8.5±10.5 2.1 25 1.0 2.9 

Bobek 5.7 1.1 33 2.3 3.2 

Blisk 6.0 2.0 37 1.9 2.4 

Buket 8.2 2.3 24 2.7 2.9 

Cekin 5.6 2.2 27 1.3 3.3 

Celeia 5.4 2.1 28 2.1 2.4 

Cerera 5.2 1.7 30 2.0 2.7 

Cicero 8.7 2.9 28 1.5 2.9 

Savinja (Styrian) 

Golding 4.5±6.0 2.0 28 0.8 3.1 

South Africa g 

Outeniqua 12.0±13.5 2.8 29 1.6 3.0 Bitter 

Southern Brewer 9.0±10.5 2.4 39 1.5 2.1 Bitter 

Southern Promise 9.5±11.5 2.3 21 0.7 2.4 Dual purpose 

Southern Star 12.0±15.5 2.8 31 1.6 1.5 Bitter 

United States h 

Cascade 4.5±7.0 1.0 37 1.2 2.7 

Chelan 14.5 1.4 35 1.5 1.2 

Chinook 12.0±14.0 3.9 32 2.0 2. 

Cluster 5.5±8.5 1.4 36±42 0.4±0.8 2.5 

Columbus/Tomohawk/ 

Zeus (CTZ) 14±18 3.1 29±34 2.0±3.5 1.7 

Galena 11.0±13.0 1.6 44 0.9±1.4 2.0 

Horizon 13.6 2.2 19 1.9 1.6 Dual 

Millenium 15.5 3.2 30 2.0 2.4 

Mount Hood 5.0±8.0 1.1 23 1.1 2.5 

Nugget 12.0±14.0 3.3 27 2.0 2.2 

Willamette 5.0±7.0 1.6 33 1.2 2.9 

YCR-5 (Warrior) 14.5±16.5 2.6 24±26 1.0±2.0 1.9 

Humulus lupulus var. 

neomexicanus 3.1 1.4 65 0.6±0.8 0.73 

a 

From Neve (1991) and Darby (personal communication). Neve also gives data for Omega and Yeoman (GB), 

Orion (D) and Aquila, Banner, Olympic and Talisman (USA). 

b 

Leggett (2004). 

c 

National Hop Association of England (2000). 

d 

Barth (1999). 

e 

Inglis (1999). 

f 

Brudzynski and Baranowski (2003). 

g 

Brits and Linsley-Noakes (2001). 

h 

www.usahops 



BARTH, S. J. (1999) Ferment, 12 (5), 40. 

BARTH, H. J., KLINKE, C. and SCHMIDT, C. (1994) The Hop Atlas ± The History and Geography of the 

Cultivated Plant. Barth, NuÈrnberg, 383 pp. 

BEHRE, K.-E. (1999) Vegetation History and Archaeobotany, 8, 35±48. 

BENITEZ, J. L., FORSTER, A., DE KEUKELEIRE, D., MOIR, M., SHARPE, F. R., VERHAGEN, L. C. and WESTWOOD, 

K. T. (1997). European Brewery Convention ± Manual of Good Practice: Hops and Hop Products, 

pp. xiv + 186. Verlag Hans Carl, NuÈrnberg. 

BRITS, G. and LINSLEY-NOAKES, G. C. (2001) Proc. 8th. Conv. Inst. Brewing, Africa Section, Sun City, 

South Africa, p. 176. 

BRUDZYNSKI, A. and BARANOWSKI, K. (2003) J. Inst. Brewing, 109, 154. 

BURGESS, A. H. (1964) Hops: Botany, Cultivation and Utilization. Leonard Hill, London, pp. xx + 300. 

DE KEUKELEIRE, J., OMMS, G., HEYERICK, A., ROLDAN-RUIZ, I., VAN BOCKSTAELE, E. and DE KEUKELEIRE, 

D. (2003) J. Agric. Food Chem., 51, 4436. 

EUROPEAN BREWERY CONVENTION (1987) Monograph XIII. Symposium on Hops, Weihenstephan, pp. xii 

+ 286. 

EUROPEAN BREWERY CONVENTION (1994). Monograph XXI. Symposium on Hops, Zoeterwoude, 

pp. xviii + 300. 

EUROPEAN BREWERY CONVENTION (1997) see Benitez et al. 

FILMER, R. (1982) Hops and Hop Picking, Shire Publications, Princes Risborough, 80 pp. 

GREEN, C. P. (1978) J. Inst. Brewing, 84, 312. 

INGLIS, T. (1999). Ferment, 12 (5), 19±28. 

JOHNSTONE, D. I. H. (1997) Ferment, 10, 325±329. 

LEGGETT, G. (2004) Brewer Intern., 4 (2), 42. 

LEMUSIEAU, G., LIEÂ GEOIS, C. and COLLIN, S. (2001) Food Chemistry, 72, 413. 

LIKENS, S. T., NICKERSON, G. B., HAVOULD, A. and ZIMMERMAN, C. E. (1978) Crop Sciences, 18, 380±386. 

MOIR, M (2000) J. Amer. Soc. Brew. Chem., 58, 131±146. 

MURAKAMI, A. (2000) J. Inst. Brewing, 106, 157±161. 

NATIONAL HOP ASSOCIATION OF ENGLAND (2000). The Hop Guide, 22 pp. 

NEVE, R. A. (1991) Hops. Chapman & Hall, London, xii + 266 pp. 

PICKETT, J. A., COATES, J. and SHARPE, F. R. (1977) Proc. 16th Congr. Eur. Brew. Conv., Amsterdam, 

p. 123; J. Inst. Brewing, 83, 302. 

WILSON, D. G. (1975) New Phytologist, 75, 627±648. 


8 

The chemistry of hop constituents 


Freshly picked hop cones contain about 80% moisture and rapidly go mouldy if not dried 

on a kiln or oast. Commercial hops contain c. 10% moisture. Moisture in hops is usually 

measured as the loss on drying at 105±107 ëC for 1 hour (Analytica-EBC, 1998). For green 

hops a longer period of drying (four hours) is necessary. Some essential oil may be lost 

during oven drying. Alternative methods for estimating moisture in hops include drying in a 

vacuum desiccator or azeotropic distillation (Dean and Stark method). Hop analyses are 

usually reported às is' but may occasionally be given with reference to dry matter. 

Most of the brewing value of the hop is found in the resins and essential oils which are 

only slightly soluble in water. However, Goldstein et al. (1999) showed that hops contain 

20±25% of water-soluble constituents which dissolve directly in the boiling wort. This 

fraction will include carbohydrates, amino acids, proteins, polyphenols, and inorganic 

salts. MacWilliam (1953) showed that hops contain c. 2% of sugars mainly fructose, 

glucose and raffinose; hops also contain 1±2% of pectin. Goldstein et al. (1999) drew 

attention to the presence of glycosides in the water-soluble fraction of hops. Most 

organisms transport water-insoluble substances by conjugating them with a sugar, usually 

glucose, to produce a water-soluble glucoside. -Sitosteryl glucoside was found in hops 

as long ago as 1913 (Power et al., 1913) and many polyphenols are found as glycosides in 

hops (see later). Goldstein et al. (1999) also found that many volatile constituents of hops 

are also present in the water-soluble fraction bound as glycosides. 

Hops contain 2.0±3.5% of nitrogen equivalent to 12.5±21.7% of protein. About 0.5% 

of the nitrogen, equivalent to 3.1% of protein, is soluble in water. Hops contain c. 0.1% of 

amino acids and dried hops yield about 8% of ash (inorganic matter). Lipids ± oils, fats, 

and waxes ± are, like the hop resins, insoluble in water. Hop seeds contain up to 32% of 

triglycerides but they are not usually dispersed from intact seeds during wort boiling. Hop 

wax is derived from the cuticle of cones and leaves and is a mixture of long chain 

hydrocarbons (C29 predominates), alcohols, acids and esters together with -sitosterol. 

Although these materials are only slightly soluble in water they may be included with the 

resins in solvent extracts of hops. 


The chemistry of hop constituents was reviewed by Stevens (1967) and Moir (2000) 

provided amillenium review. Other reviews have concentrated on individual classes of 

hop compound. 

8.2 Hop resins 


A book, Chemistry and Analysis of Hop and Beer Bitter Acids (Verzele and De 

Keukeleire, 1991) provides adetaileddiscussionofmostofthereactions discussedinthis 

section. Later information is found in the European Brewery Convention Symposium on 

Hops, Zoeterwoude (1994) and the European Brewery Convention: Manual of Good 

Practice-Hops and Hop Products (Benitez et al., 1997). The Nomenclature Committee of 

the Hops Liaison Committee (1969) made recommendations defining A, non-specific 

fractions and B, specific compounds and mixtures of specific compounds. 

A. Non-specific fractions 

1. Total resins The part of the hop constituents that is characterized by solubility both 

in cold methanol and diethyl ether (mainly hard resins, uncharacterized soft resins, 

-acids and -acids). The requirement that the total resins should be soluble in cold 

methanol is designed to exclude hop wax which will slowly crystallize from cold 

methanol. 

2. Total softresins The fraction ofthetotal resins thatischaracterizedbysolubilityin 

hexane (mainly -acids, -acids and uncharacterized soft resins). 

3. Hard resins The fraction of the total resins that is characterized by insolubility in 

hexane. It is calculated as the difference between total resins and total soft resins. 

4. -Fraction The total soft resins minus the -acids. 

5. Uncharacterized soft resins That portion of the total soft resins that has not been 

characterized as specific compounds. 

B. Specific compounds and mixtures of specific compounds 

1. The -acids (8.1). These are mainly humulone (8.1a), cohumulone (8.1b) and 

adhumulone (8.1c). 

2. The -acids (8.2). These are mainly lupulone (8.2a), colupulone (8.2b) and 

adlupulone (8.2c). 

3±8.Chemical descriptions of the -and -acid analogues are given in Table 8.1. 

9. The iso- -acids (8.40). These are mainly isohumulone, isocohumulone and isoadhumulone. 

10. Isohumulone (8.40a). The mixture of cis- and trans-isohumulone. Similarly, 

isocohumulone (8.40b) refers to a mixture of cis- and trans-isocohumulone and 

isoadhumulone (8.40c) to a mixture of cis- and trans-isoadhumulone, 

11. cis-Isohumulone (8.43). The iso- -acid with the empirical formula C21H30O5. It is 

an oil with the higher partition coefficient in a phase system of a hydrocarbon and a 

buffer, and contains an isovaleryl side chain. Cis- means that the 3-methyl-2-butenyl 

side chain and the tertiary hydroxyl group are on the same side of the ring. 

12. trans-Isohumulone (8.44). The iso- -acid with the empirical formula C21H30O5, 

with a m.p. 72 ëC and the lower partition coefficient in a phase system of a 

hydrocarbon and a buffer, and contains an isovaleryl side chain. Trans-means that 


the 3-methyl-2-butenyl side chain and the tertiary hydroxyl group are on opposite 

sides of the ring. 

13. cis-Isocohumulone. As in 11 but with reference to C20H28O5 and an isobutyryl 

(R ˆ Pr i ) side chain. 

14. trans-Isocohumulone. As in 12 but with reference to C 20H 28O 5 and an isobutyryl 

side chain. 

15. cis-Isoadhumulone. As in 11 but with reference to a 2-methylbutyryl 

(R ˆ CHMeEt) side chain. 

16. trans-Isoadhumulone. As in 12 but with reference to a 2-methylbutyryl side chain. 

17. Allo-Iso- -acids (8.41). These are isomers of the iso- -acids having a shifted 

double bond in the isohexenoyl side chain (i.e. 4-methyl-2-pentenoyl). Of each alloiso- 

-acid there is a cis- and a transform. The following specific names are 

therefore proposed: cis-allo-isohumulone, trans-allo-isohumulone, cis-allo-isocohumulone, 

trans-allo-isocohumulone, cis-alloisoadhumulone and trans-allo-isoadhumulone. 

18. Hulupones (8.85). These consist of hulupone, cohulupone and adhulupone. 

19. Hulupone (8.85a). Has the empirical formula C20H28O4. It is 2,2-di[3-methyl-2butenyl]-5-isovaleryl-1,2,4-cyclopentanetrione 

and is formed from lupulone. 

20. Cohulupone (8.85b). As in 19 but with reference to C19H26O4 and a 5-isobutyryl 

side chain. 

21. Adhulupone (8.85c). As in 19 but with reference to C 20H 28O 4 and a 5-[2methylbutyryl] 

side chain. 

22. Humulinic acids (8.3). These consist of the cis-and trans- forms of humulinic acid, 

cohumulinic acid and adhumulinic acid. 

23. cis-Humulinic acid. Has the empirical formula C15H22O4 with m.p. 68 ëC and the 

higher partition coefficient in a phase system of a hydrocarbon and a buffer. Cismeans 

that the 3-methyl-2-butenyl side chain and the alcoholic hydroxyl are on the 

same side of the ring. 

24. trans-Humulinic acid. Has the empirical formula C 15H 22O 4 with m.p. 95 ëC and the 

lower partition coefficient in a phase system of a hydrocarbon and a buffer. Transmeans 

that the 3-methyl-2-butenyl side chain and the alcoholic ring hydroxyl group 

are on opposite sides of the ring. Similar considerations will apply to ciscohumulinic 

acid, trans-cohumulinic acid, cis-adhumulinic acid and trans-adhumulinic 

acid. 

The sub-committee also recommended that where the intention is to refer to `beer bitter 

substances', usually an incompletely known mixture, this phrase should be used. Using 

these definitions the bulk of the brewing and bittering value of the hop is found in the 

total soft resins and, in particular, in the -acids. Only traces of the -acids survive into 

beer; they are transformed during wort boiling into the iso- -acids which are the major 

bittering principles of beer. The importance of the allo-iso- -acids is still debatable. The 

-acids are too insoluble in water to contribute to beer flavour themselves but they can be 

oxidized into hulupones which are bitter and are minor bittering principles in some beers. 

Hydrolysis of the -acids and the iso- -acids gives a mixture of humulinic acids which 

are not bitter. The humulinic acids are not normally found in beer but they may be present 

in some isomerized hop extracts and beers brewed therefrom. 

The -acids can be separated from the total soft resins by their ability to form an 

insoluble lead salt (chelate?) with lead(II) acetate in methanol; the -fraction may be 

recovered from the mother liquors. The -acids can be regenerated from the lead salts by 


suspending them in methanol and adding either sulphuric acid or hydrogen sulphide gas. 

After removal of the inorganic matter by filtration, evaporation of the solvent leaves a 

mixture of -acids from which humulone may slowly crystallize. Humulone (8.1a) was 

the only -acid known until 1953 when Rigby and Bethune isolated the analogues 

cohumulone(8.1b)andadhumulone(8.1c)bycountercurrentdistribution.Traces ofother 

analogues have been found (Table 8.1). Later, Hermans-Lockkerbol and Verpoorte 

(1994) used centrifugal partition chromatography to separate the -acids. Most hop 

varieties contain about 10% of adhumulone in their -acids but the proportion of 

cohumulone appears to be a varietal characteristic (Table 7.3). The -acids form 

crystalline 1:1 complexes with 1,2-diaminobenzene (o-phenylenediamine). Repeated 

recrystallization concentrates the humulone complex with respect to those of the other - 

acids. Decomposition of the yellow complex with 2N-hydrochloric acid followed by 

recrystallization from cyclohexane at 20ëC gives humulone, m.p. 63ëC. Most of the 

chemistry of the -acids has been carried out on humulone purified in this way but 

Simpson (1993a) found that asample prepared in this manner still contained 8% of 

cohumulone and 1% of adhumulone. 

The -acids (8.2) may crystallize from the -fraction obtained after the lead salts have 

been precipitated. In practice it may be better to dilute the methanol solution with brine 

and extract the -fraction into light petroleum. Alternatively, asolution of ahop extract 

inhexanemaybeextractedfirstwithdisodiumcarbonate,toremovethestronger -acids, 

and then with sodium hydroxide to recover the -acids. From the mixture of -acids 

Lermerisolatedlupulone(8.2a)in1863.Muchlater,inthe1950s,itwasfoundthatthe - 

acidwhichcrystallizedfromEnglishhopextractswascolupulone(8.2b)Thus,likethe - 

acids, the -acids are amixture of analogues. They are too sensitive to aerial oxidation to 

be separated by countercurrent distribution. They can be separated by HPLC, but before 

this technique was available, the proportions of the -acid analogues was found by 

converting them to tetrahydro- -acids (8.9, Fig 8.1) which were stable during 

countercurrent distribution. It was found that the -acids were always richer in 

colupulone than the -acids were in cohumulone. Indeed aregression equation was 

obtained: 

%Colupulone in -acids =0.943 (% cohumulone in -acids) +20.2 

The structures of humulone (8.1a) and lupulone (8.2a), except in minor detail, were 

worked out by Wollmer and Wieland (Fig. 8.1). Both are acylphloroglucinols substituted 

with 3-methyl-2-butenyl- (dimethylallyl- or isoprenyl-) groups, three in lupulone and two 

in humulone, which also has a tertiary hydroxyl group. Hydrolysis of humulone, 

C21H30O5, gives humulinic acid (8.3a), C15H22O4, and 4-methyl-3-pentenoic acid 

(isohexenoic acid, (8.13)), C6H10O2, which accounts for all the carbon atoms, but 

isobutyraldehyde (8.14) is also formed. Hydrogenation of humulinic acid gave 

dihydrohumulinic acid (8.4a) which by Clemmersen reduction gave 1, 3-di-isopentylcyclopentane 

(8.11) establishing the five-membered ring in humulinic acid. Mild 

hydrogenation of humulone gave the tetrahydro-derivative (8.13a), showing two double 

bonds, but with palladium chloride hydrogenolysis occurs giving humuloquinol (8.5a), 

C16H24O5, which is readily oxidized to humuloquinone (8.6a). Hydrolysis of 

humuloquinone gives isohumulinic acid (8.7a) also obtained from dihydrohumlinic acid 

(8.4a) by oxidation with bismuth oxide. Similarly, mild hydrogenation of lupulone (8.2a) 

gives a hexahydro-derivative but hydrogenolysis gives tetrahydrodeoxyhumulone (8.8a) 

which, when shaken in air or oxygen with lead acetate solution, gives the lead salt of 

tetrahydrohumulone (8.9a) linking the - and -acids. 


Table 8.1 Analogues of the - and -acids 

-acids -acids 

Acyl side chain (R) Name Formula m.p. (ëC) [ ] 24 

0 pKa Name Formula m.p. (ëC) 

a -CO.CH 2.CH(CH 3) 2 Humulone C 21H 30O 5 64.5ë 211ë 5.5 Lupulone C 26H 38O 4 92ë 

isovaleryl 

b -CO.CH(CH 3) 2 Cohumulone C 20H 28O 5 oil 208.5ë 4.7 Colupulone C 25H 36O 4 93 94ë 

isobutyryl 

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 

2-methylbutyryl 

d -CO.CH2.CH3 Posthumulone a 

propionyl 

e -CO.CH2.CH2.CH(CH3)2 Prehumulone b 

4-methylpentanoyl 

f -CO.(CH2)4.CH3 Adprehumulone c 

hexanoyl 

C19H26O5 oil ± ± ± d 

C22H32O5 oil 172ë ± ± e 

C22H32O5 ± ± ± ± e 

C24H34O4 

C27H40O4 

C27H40O4 

g -CO.CH2.CH2.CH(CH3).CH2.CH3 ± C23H34O5 ± ± ± ± C28H42O4 91ë 

4-methylhexanoyl 

a Verzele (1958). 

b Rillaers and Verzele (1962). 

c Smith et al. (1998). 

d Riedl et al. (1956). 

e Riedl (1954). 


101ë 

91ë 

90ë


Fig. 8.1 Reactions of - and -acids.

In both humulone (8.1a) and lupulone (8.2a) carbon atom-6 in the ring is sp3 

hybridized; the other atoms in the ring are sp2 hybridized. In humulone carbon atom-6 is 

substituted by four different groups and so it is chiral and humulone is optically active. 

The natural -acids are laevorotatory and their specific angular rotations are given in 

Table 8.1. Using the sequence rules of Cahn et al. (1951, 1956 and 1966) the absolute 

stereochemistry of the -acids is R(ectus) so natural humulone is R-( )-humulone. 

Synthetic racemic humulone is an equimolecular mixture of R-( )-humulone and 

unnatural S-(+)-humulone and, as such, will not rotate the plane of polarized light. Fairly 

mild chemical treatment will convert R-( )-humulone into racemic RS-( )-humulone. 

( )-Humulone has not been resolved into its enantiomers but this should be possible. 

In lupulone (8.2a) and the other -acids carbon atom-6 is substituted with two 3methyl-2-butenyl- 

(isoprenyl-) groups so it will not be chiral and will not rotate the plane 

of polarized light. However, in adlupulone (8.2c) there is achiral centre at C-2 of the 2methylbutyryl-side 

chain so natural adlupulone should be optically active. It follows that 

natural adhumulone (8.1c) has two chiral centres so four enantiomers are possible. By 

analogy with L-isoleucine, the 2-methylbutyryl-side chain in natural adhumulone and 

adlupulone probably has the S-configuration. 

It is difficult to write a single structure for most of the hop resins because they exhibit 

keto-enol tautomerism where ketones exist in equilibrium with the related enol: 

With an isolated ketone, as in acetone (8.15), the equilibrium lies well on the ketone side: 

However, -dicarbonyl compounds such as acetylacetone (pentane-2, 4-dione, (8.16)) 

exist largely in the enol form which is stabilized by hydrogen bonding. 

Even more tautomers can be drawn for -tricarbonyl compounds such as triacetylmethane 

(8.17) 

(CH 3.CO) 3.CH 

(8.17) triacetylmethane 

Most of the hop resins contain -di- or -tricarbonyl systems which are enolized. In most 

cases the individual tautomers cannot be isolated but it is usually possible to estimate the 

proportions of the individual tautomers by proton magnetic resonance spectroscopy 

(PMR) or other physical methods. Irrespective of the main component of a tautomeric 


mixture it can react both as a ketone and as an enol. It is an equilibrium mixture so that if 

one component is removed by reaction it will be regenerated according to the 

equilibrium. Thus, for example, phloroglucinol, the parent of the hop resins, exists almost 

entirely in the trienol form (8.18) and can form a triacetate (8.20) but it can react as 

cyclohexane-1, 3, 5-trione (8.19) to produce a tri-oxime (8.21). 

O·Ac 

AcO O·Ac 

(8.20) 

phloroglucinol 

triacetate 

OH 

HO OH 

(8.18) 

phloroglucinol 

O 

O O 

(8.19) 

cyclohexane- 

1,3,5-trione 

HO·N 

N·OH 

(8.21) 

cyclohexane- 

1,3,5-trioxime 

N·OH 

PMR measurements show that humulone exists principally as a dienol. Of the possible 

tautomeric structures (8.1.1±4), structure (8.1.1) is thought to represent the major 

tautomer; structures (8.1.2 ) and (8.1.4) were excluded on the basis of optical rotatory 

dispersion measurements and (8.1.1) was preferred over (8.1.3) by comparison with 

model compounds (De Keukeleire and Verzele, 1970). 

OH 

O 

OH OH 

R' CO·R R' CO·R R' CO·R R' 

HO 

O 

HO OH 

O 

OH 

O O 

HO R' HO R' HO R' HO R' 

(8.1.1) (8.1.2) (8.1.3) (8.1.4) 

Similarly, PMR measurements suggest that lupulone exists as a mixture of two tautomers 

(8.2.1) and (8.2.2) in the ratio 7:3 (Collins et al., 1971). 

HO 

H 

O O 

O 

R HO 

(8.2.1) (8.2.2) 

Consideration of the tautomers of acetylacetone (8.16) shows that the enol hydrogen atom 

is bound to different oxygen atoms in the two cyclic forms. This hydrogen can dissociate 

leaving an enolate ion stabilized by resonance. Thus, -di- and -tri-carbonyl compounds 

are acids and these functions provide the acidity of the hop resins. The strength of acids 

can be compared on the pKa scale where 

pKa= log 10Ka 


O 

R 

O 

OH 

C 

O 

H 

R

and Ka is the dissociation constant of the acid. On this scale completely dissociated 

mineral acids have negative values while carboxylic acids have values such as: methanoic 

(formic) acid, pKa 3.77, ethanoic (acetic) acid, pKa 4.76, propanoic acid, pKa 4.88, and 

benzoic acid, pKa 4.20. Triacetylmethane (8.17), pKa 5.81, is a stronger acid than 

acetylacetone (8.16) pKa 8.13, while phenol, pKa 10.0, is weaker still. Phloroglucinol 

(8.18) has three phenolic hydroxy groups; the pKas of the first two are 7.97 and 9.23. 

pKa measurements should be made in dilute aqueous solution but, because of their 

limited solubility, early estimates of the pKas of the hop resins were made in aqueous 

methanol solutions giving pKa values for humulone, 5.5, cohumulone, 4.7, and adhumulone, 

5.7. In aqueous solutions Simpson and Smith (1992) found equilibrium pKa values for the 

most acidic functions were: humulone, 5.0, colupulone (8.2b), 6.1, trans-isohumulone 

(8.44), 3.1; and trans-humulinic acid (8.3a), 2.7. Thus, the - and -acids are weaker than 

carboxylic acids, such as acetic acid, but their isomerization products, the iso- -acids, are 

stronger. It should be recalled that when the pH of the medium equals the pKa of the acid 

50% of the acid will be present as the anion and 50% undissociated. 

The solubilities of humulone, lupulone and (trans-)humulinic acid were measured by 

Spetsig (1955) (Fig. 8.2). As expected, the solubilities increase with temperature and increasing 

pH; in each case the anion is more soluble than the undissociated acid. For example, in boiling 

wort at pH 5.0, about 200 mg/litre of humulone will dissolve but, if no transformation takes 

place, most of this will be precipitated from conditioned beer (pH 4.0) at 0 ëC. 

mg/l 

5000 

1000 

500 

100 

50 

10 

5 

1 

100° 

40° 

25° 

0° 

100° 

40° 

25° 

0° 

Humulinic 

acid 

100° 

40° 

25° 

0° 

Humulone 

Lupulone 

2 4 6 8 10 

pH 

Fig. 8.2 Solubilities of humulinic acid, humulone and lupulone (Spetsig, 1955). 


The structure of the - and -acids have been verified by synthesis. Acylation of 

phloroglucinol (8.18) gives the parent phloracylphenone (8.22); phlorisovalerophenone 

(8.22a) for humulone and lupulone and phlorisobutyrophenone (8.22b) for cohumulone 

and colupulone, etc. 

This can be alkylated with 3-methyl-2-butenyl bromide (8.24) (isoprene hydrobromide, 

dimethylallyl bromide), which is prepared by the 1,4-addition of hydrogen bromide to 

isoprene (8.23). 

Alkylation of phloracylphenone (8.22) can give a mixture of one mono- (8.25), two di- 

(8.26 and 8.27), one tri- (8.28) and one tetra- (8.29) isoprenyl derivative. 


The tri-substitued derivatives are the -acids (8.2), the di-substituted derivatives (8.26) 

the deoxy- -acids; these and the monosubstituted derivatives (8.25) have been found in 

hops but neither the di-substituted derivative (8.27) nor the tetra-substituted derivative 

(8.29, lupones) have been found to occur naturally. With one molecule of base and 3methyl-2-butenyl 

bromide the mono-substituted derivatives (8.25) can be obtained in 

good yield and in liquid ammonia the -acids (8.28) are obtained in up to 70% yield 

(Collins et al., 1971) but the synthesis of the the deoxy- -acids (8.26), and thus the - 

acidsismoredifficult.Intheoriginalsynthesisof( )-humulone(Riedl,1951)theoverall 

yield was only 5.7%. Here the deoxyhumulone was oxidized to the lead salt of ( )humulone 

with oxygen in the presence of lead (II) acetate. 

Slightly better yields of deoxy- -acids were obtained using the weakly basic ionexchange 

resin DeAcidite H-IP (OH form) (Collins and Laws, 1973) and by using 3methyl-buten-3-ol 

with boron trifluoride-etherate as the alkylating agent (Collins and 

Shannon, 1973). Deoxy- -acids (8.26) have also been prepared by the irradiation of - 

acids with ultraviolet light (Fernandez, 1967). Better yields of -acids from deoxy- - 

acids are found when the autoxidation is carried out in the presence of triethyl phosphite 

(Sigg-Grutter and Wild, 1974). Without this reducing agent the intermediate hydroperoxide 

was isolated. Even with the optimal yields reported (Pfenninger et al.,1975) 

synthetic racemic -acids are unlikely ever to be as cheap as the naturally produced 

enantiomers from the hop. 

8.2.2 Biosynthesis of the hop resins 

The biosynthesis of the hop resin (Fig. 8.3) within the plant is thought to follow asimilar 

route to the chemical synthesis. When CH3. 14 CO2Na was injected into aripening hop 

plant the labelling of the radioactive humulone suggested that the phloroglucinol nucleus 

was made up from three acetate units. The acyl side chains were derived from amino 

acids or intermediates in their biosynthesis. Thus humulone and lupulone come from a 

leucine metabolite, cohumulone and colupulone from avaline metabolite and the adanalogues 

from isoleucine (with aromatic amino acids the prenylflavanoids are formed, 

see Section 8.2.5). Transamination and decarboxylation of the amino acids leads to the 

Coenzyme Aesters of isovaleric, isobutyric and 2-methylbutyric acids (8.30). These are 

thought to react with three molecules of malonyl Coenzyme A (8.31) to give the 

polyketide (8.32) which with an enzyme similar to chalcone synthetase forms the 

phloracylphenone (8.22) (Fung et al., 1997). This with dimethylallyl pyrophosphate 

(8.34) and with the enzyme(s) prenyltransferase(s) forms the mono-prenyl derivative 

(8.25), the deoxy- -acids (8.26) and, probably, the -acids. Oxidation of the deoxy- - 

acids gives the required -acids (8.1). 

Sincethe1950sithasbeenthoughtthatdimethylallylpyrophosphate(8.34),theparent 

of the isoprenoids, terpenoids and steroids, was formed via mevalonic acid (8.35) (see 

Fig. 8.4 Pathway A) and that this was the only route to these compounds. Now, an 

alternative route, via 1-deoxyxyulose-5-phosphate (8.37), has been found (Eisenreich et 

al., 1998. Fig. 8.4 Pathway B). Goese et al., 1999) have shown that the isoprenyl- side 

chains of humulone are formed by this latter route and that the biosynthesis of humulone 

goes through asymmetrical intermediate. According to the new pathway, glyceraldehyde 

3-phosphate, produced from glucose by the normal Embden-Meyerhof-Parnas pathway 

(Fig. 12.7) condenses with àctive acetaldehyde' (8.36), produced from pyruvate and 

thiamine, to give D-1-deoxyxylulose 5-phosphate (8.37). This re-arranges, via 2-Cmethylerythrose 

(8.38) and 2-C-D-erythritol 4-phosphate (8.39) to isopentenyl pyropho- 


R·CO·CO2H 

R·CO·S·CoA (8.30) 

+ 3HO2C·CH2·CO·S·CoA (8.31) 

R·CO·CH2·CO·CH2·CO·CH2·CO·S·CoA (8.32) 

OH 

CO·R 

HO OH 

(8.22) 

OH 

CO·R 

HO OH 

HO 

– _ 

R·CH·CO2 

· + 

NH3 

(8.25) 

OH 

CO·R 

OH 

H2C 

H3C 

C·CH2·CH2·O 

O 

·· 

P 

· 

OH 

O 

O 

·· 

P 

· 

OH 

OH 

isopentenyl pyrophosphate (8.33) 

H2C 

H3C 

C CH·CH2·O 

O 

·· 

P 

· 

OH 

O 

O 

·· 

P 

· 

OH 

OH 

γ-Dimethylallyl pyrophosphate (8.34) 

sphate (8.33). This with an isomerase is converted into dimethylallyl pyrophosphate 

(8.34) which is thought to be the biological isoprenylating agent. Isopentenyl 

pyrophosphate (8.33) and dimethylallyl pyrophosphate (8.34) can condense together to 

form, first, geranyl pyrophosphate (8.88), the parent of the monoterpenes, and then 

farnesyl pyrophosphate (8.105), the parent of the sesquiterpenes, both of which are 

important constituents of the essential oil (see later). From the limited data available 

(Eisenreich et al., 1998) it appears that both the mevalonic acid (8.35) and the deoxyxyulose 

pathways (Fig. 8.4) are found in most higher plants. Steroids are mainly formed 

HO 

HO 

deoxy-α-acids (8.26) α-Acids (8.1) 

Fig. 8.3 Biosynthesis of the -acids. 


OH 

CO·R 

O

Pathway A 

CH3 

CH3 

O 

SCoA 

O 

SCoA 

O O 

acetyl CoA acetoacetyl CoA 

Pathway B 

CO2H 

CH3 

O 

–CO2 

HO 

by the mevalonate pathway but isoprene and the essential oil constituents arrive by the 

deoxy-xyulose pathway. 

8.2.3 Analysis of the hop resins 

Procedures for the estimation of the total resins, total soft resins, and hard resins, by 

difference, (see definitions above) in hops and hop products are given in Analytica-EBC 

but since it was found that the -acids are the most important brewing principles few 

brewers bother to measure the total and soft resin contents. The -acids were originally 

estimated gravimetrically as their lead salts but since the precipitate is soluble in excess 

of the lead acetate reagent, trials had to be made so that only the correct amount of 

CH3 

SCoA 

NADPH 

OH 

3 ATP 

–CO2 

O O 

– Pi 

HO 

H3C 

N 

S 

mevalonate 

(8.35) 

+ 

CHO 

OH 

CH2OX 

pyruvate (8.36) GAP 

glyceraldehyde 

3-phosphate 

HO 

O OH 

OX 

2C-methylerythrose 

(8.38) 

HO 

NAD(P)H 

HO OH 

O 

HO 

SCoA 

O O 

4 

3 

5 

O 

HMG CoA 

2 

1 

IPP isopentenyl 

pyrophosphate (8.33) 

H 

X = H:2C-methyl-D-erythritol 

X = phosphate: 

2C-methyl-D-erythritol 

4-phosphate 

(8.39) 

OX 

O 

CH3 

O 

H 

OH 

CH2OX 

OPP 

SCoA 

X = H:D-1-deoxyxylulose 

X = phosphate: 

D-1-deoxyxylulose 

5-phosphate 

(8.37) 

IPP 

isopentenyl 

pyrophosphate 

(8.33) 

OPP 

Fig. 8.4 Isoprenoid biosynthesis: (A) via the mevalonate pathway, (B) via the glyceraldehyde 

3-phosphate/pyruvate pathway (Rohmer, 1998). 


Conductance 

Lead acetate (ml) 

Fig. 8.5 Conductometric titration of -acids. 

reagent was used making it alengthy process. It was found that if the conductivity of a 

methanolic solution of the -acids was measured during titration with methanolic lead 

acetatesolutionitdidnotincreaseuntiltherewasanexcessofthereagent.Thusifregular 

aliquots (0.20ml) of the methanolic lead acetate reagent (2 or 4%) are added and the 

conductivitymeasuredaftereachaddition,agraphcanbedrawnwheretheintersectionof 

the two straight portions provides the endpoint (Fig. 8.5). The absolute value of the 

conductivity is not needed and the lead acetate reagent can be standardized by asimilar 

titration against 0.100 N sulphuric acid. The shape of the graph is different in the 

presence of other solvents, so, in the approved method, pyridine (1ml) is added to the 

titration. The conductivity may be plotted against the volume of the reagent on Cartesian 

coordinate paper (Fig. 8.5) or the resistance may be plotted directly on to reciprocal ruled 

paper. Since the reaction of lead acetate is not specific for -acids, the result is expressed 

as the Lead Conductance Value (LCV). However, with fresh hops the LCV is very 

similar to the -acid content but, as hops age on storage, oxidation products are formed 

which may react with lead acetate. 

The -acids are the only hop resins which show significant optical activity ([ ]D 20 

237ë in hexane) so they can be estimated in the soft resins by polarimetry but this 

method has not been officially adopted. 

Alderton et al. (1954) measured the light absorption of humulone and lupulone under 

both acid and alkaline conditions (Fig. 8.6). In particular, they measured in alkaline 

solution the Absorbance (A) at 275nm ( min for both humulone and lupulone), 325nm 

( maxforhumulone)and355nm( maxforlupulone)andproducedregressionequationsto 

determine both humulone and lupulone. This method was adopted by the ABSC when: 

-acids, %ˆd … 51:56A355 ‡73:79A325 19:07A275† 

-acids, %ˆd …55:57A355 47:59A325 ‡5:10A275† 

where dis the dilution factor. 

These regression equations contain large multiplying factors so the procedure requires 

ahigh degree of precision in instrument calibration and the purity of the solvents. This 

was the first method to give values for the % -acids. The assumption that the solution of 

hop resins is abinary mixture of -and -acids with constant background absorption is 

probably true with fresh hops but not with deteriorated samples. Indeed, Likens et al. 

(1970) using this method proposed that the ratio 


Specific absorption coefficient (l/g cm) 

50 

25 

Humulone 

complex, 

acid 

Humulone 

complex, 

alkaline 

0 

200 250 300 350 400 

Wavelength (nm) 

A275=A325 ˆHop Storage Index (HSI) 

Lupulone, 

alkaline 

Lupulone, 

acid 

Fig. 8.6 Absorption spectra of lupulone and humulone complex in acidic (0.002N) and alkaline 

(0.002 N) methanol (Alderton et al., 1954). Copyright (1954) American Chemical Society. 

was found to increase from c. 0.24in fresh hops to c. 2.5in completely oxidized lupulin. 

The measurement has been adopted by the ASBC. For more precise analysis some 

method of separation is necessary before measurement. Methods based on Dowex and 

Sephadexion-exchange resinshave nowbeenarchivedandreplacedbyaninternationally 

agreed HPLC method using a250 4mm, 5 mRP18 Nucleosil C19 column. The 

chromatogram usually shows four peaks: cohumulone, humulone + adhumulone, 

colupulone and lupulone + adlupulone. Other systems will resolve humulone and 

adhumulonebutitremainsdifficult.Verzele andDe Keukeleire(1991) describeHPLC of 

an ethanolic extract of hops using two coupled columns with diode array detection when 

sixty peaks could be recognized including 10 -acids, 10 iso- -acids, and 11 deoxy- - 

acids. Theoccurrence ofiso- -acids (2.7%)inanethanolicextract ofhopsisnoteworthy. 

8.2.4 Isomerization of the -acids 

As long ago as 1925 it was suggested that the hydrolysis of humulone (8.1) to humulinic 

acid (8.3) proceeded via an intermediate (8.40, Fig. 8.7). Later this structure (8.40) was 

given to the bitter-tasting oil obtained by boiling humulone with N/15 (0.067M) sodium 

hydroxide solution for three minutes (Windisch et al., 1927). It was originally called 

`Resin A' but later the name ìsohumulone' was adopted for the parent ofthe iso- -acids. 

The iso- -acids are much more soluble in water (c. 120mg/l) than the -acids (3mg/l) 

and according to Peacock (1998) nine times more bitter. To account for the formation of 

isobutyraldehyde (8.14), asecond pathway leading to 8.14 and `Resin B' (4-acetylhumulinic 

acid, 8.42) was proposed. Later the products of this second pathway were 

thought to be formed via allo-iso- -acids (8.41). However, only 4.5% of allo-iso- -acids 

were formed inboiling wort but better yields were found at pH9.0. Less than 1mg/l were 

found in beer. So the iso- -acids, isohumulone, isocohumulone and isoadhumulone 

(8.40) are the major bittering principles in beer. 


HO 

HO 

OH 

CO·R 

O 

HO 

They are usually estimated by the light absorption of a solvent extract. In the 

internationally agreed method degassed, acidified beer (10 ml) is extracted with isooctane 

(2, 2, 4-trimethylpentane) (20 ml) and, after centrifugation, the absorbance of the 

isooctane layer is read at 275 nm in a 1 cm cell against an isooctane blank. This is a 

modification of the method of Moltke and Meilgaard (1955) who gave a regression 

equation to present results as ìsohumulones' mg/l. In order to avoid assumptions about 

the chemical nature of the bittering principles in beer, the EBC Analysis Committee 

simplified the regression equation and gave the results in Bitterness Units (BU): 

Bitterness Units (BU) = 50 absorbance 

CO·R CO·R 

CO CO 

HO OH 

(8.1) humulone (8.40) isohumulone (8.41) allo -isohumulone 

H3C 

H3C 

HO 

CO·R CO·R 

With worldwide adoption they are sometimes called International Bitterness Units (IBU). 

Most commercial beers contain 10±50 BU but exceptional beers with up to 100 BU have 

been reported (Glaser, 2002). Rigby and Bethune (1955) estimated the iso- -acids in an 

isooctane extract of beer by measuring the light absorption at 255 nm after dilution with 

alkaline methanol. Beers produced by conventional wort boiling contain only traces of - 

acids or humulinic acids which would interfere with the above two methods. However, 

worts, isomerized hop extracts and beers brewed therefrom may contain such impurities 

so accurate values for the iso- -acids can be obtained only after a suitable separation 

technique. Originally countercurrent distribution was the method of choice but HPLC is 

more likely to be used today. 

O 

O 

(8.3) humulinic acid 

OH 

OH HO 

OH 

H CO·CH3 

O 

O 

(8.42) 4-acetylhumulinic acid 

+ + 

H3C 

C CH·CH2·CO2H CH·CHO 

H3C 

(8.13) isohexenoic acid (8.14) isobutyraldehyde 

Fig. 8.7 Isomerization and hydrolysis of humulone (R ˆ Bu i without stereochemistry). 


Having established that the -acids are isomerized into iso- -acids during wort 

boiling and that the iso- -acids are the main bittering principles in beer, hop utilization 

can be defined as: 

%Hop utilization ˆ 

Amount of iso- -acids in beer 

Amount of -acids in hops used 

In conventional wort boiling only about 50% of the -acids available in the hops go into 

solution and further losses on to the break and on to the yeast during fermentation occur 

so that the overall utilization into beer will seldom exceed 40% and may be as low as 

10%.Inwortboilinghigherutilizationoccurswithweakwortsandlowlevelsofhopping. 

Indeed, the solubility of humulone (and other -acids) is alimiting factor in utilization 

(Fig. 8.2). Only 50±60% of pure humulone was isomerized during a1.5hboil while with 

the same amount of -acids in hops 65±75% were utilized in the same period. Pure 

humulone in boiling wort forms an oily layer or droplets of minimal surface area and the 

utilization is improved when the resin is spread over an inert surface such as hops, break, 

or even Celite. Utilization is also improved at higher pH values but only small variations 

are possible in wort. However, if the -acids are isolated from the hops, isomerized at 

higher pH values and then added to the beer after fermentation much better utilization is 

obtained. 

Many brewers now use such isomerized extracts. Some brewers will use low levels of 

hops in the copper, with improved utilization, and then achieve the desired bitterness by 

addition of isomerized extracts after fermentation, others will depend entirely on postfermentation 

additions. The isomerized extract must not contain any -acids as they 

would be precipitated in the beer necessitating a further filtration with loss of iso- - 

acids. Many patents exist for the preparation of isomerized extracts. For example, the - 

acids may be extracted from a solvent (ethanol or CO2) extract of hops using either 

disodium or dipotassium carbonate and the carbonate solution boiled to effect 

isomerization. The weaker -acids will not dissolve in the carbonate solutions. 

Alternatively, both - and -acids may be extracted with alkali hydroxides and the 

solution then saturated with CO 2 gas to lower the pH and precipitate the -acids. Care 

must be taken that the -acids are not boiled with alkali hydroxides as hydrolysis to 

humulinic acids will occur. The iso- -acid solution is best injected into a beer main 

because addition to conditioning tanks may cause local supersaturation and precipitation 

of the iso- -acids which will only slowly redissolve. The calcium and magnesium salts 

of the -acids can be isomerized by heating at 70 ëC for two hours. The iso- -acid salts 

formed are finely ground (particles < 10 m) and added to conditioning tanks when at 

least 24 hours are necessary to achieve solution and 85% utilization. The -acids 

recovered from the above processes may be added to the copper with or without 

deliberate oxidation. The official methods of analysis give two HPLC methods to 

determine the iso- -acids in isomerized hop extracts. Both allow for the separation of 

isocohumulone, isohumulone and isoadhumulone; one uses 4-methylbenzophenone as 

internal standard, the other -phenylchalcone. The gross structure of the iso- -acids has 

been confirmed by synthesis (Ashurst and Laws, 1966, 1967) but the low yield means 

that this is not a practical route to synthetic iso- -acids. 

The structures assigned to isohumulone (8.40) and humulinic acid (8.3) both contain 

two chiral centres so each should exist as two pairs of enantiomers. However, since 

natural (R)( )-humulone is a single enantiomer only two diasteroisomeric forms are 

found, the cis- and trans-isomers (see definitions above). The enantiomers of these 

compounds would be obtained from unnatural (S)(+)-humulone. Countercurrent 


100

distribution of the bittering substances in beer showed three peaks corresponding to 

isocohumulone, isohumulone and isoadhumulone but each peak was broader than that 

calculated for apure compound and two theoretical curves could be fitted under each 

observed peak. The same broadened pattern was exhibited by isohumulone obtained by 

boiling humulone in N/15 sodium hydroxide for three minutes, 0.1N-disodium carbonate 

for 30min. or in apH5.0 buffer solution. The proton magnetic resonance spectra of these 

isohumulone preparations were in agreement with them being a mixture of two 

stereoisomers (Burton et al., 1964). Eventually the two stereoisomers of isohumulone 

were separated by reversed-phase partition chromatography (Spetsig, 1964), countercurrent 

distribution after 2000 transfers (Alderweireldt et al., 1965), partition 

chromatography on silica gel (Clarke and Hildebrand, 1965) and, later, thin layer 

chromatography (Aitken et al., 1970) and shown to be the cis- (8.43) and trans- (8.44) 

-isomers as expected. At the same time it was found that irradiation ofhumulone at either 

365 or 254nm gave pure crystalline trans-isohumulone (8.44 photoisohumulone) (Clarke 

andHildebrand,1965(seealsoSharpeandOrmrod,1991)),sothisisomerismorereadily 

available pure than the oily cis-isomer. 

Chemically, the isomerization of humulone is atype of benzilic acid or acyloin 

rearrangement and the mechanism is given in Fig. 8.8. The isomerization follows firstorder 

kinetics in buffer solutions of constant pH but falls off in wort probably due to the 

lowering of the pH. The two isomers of isohumulone are not readily converted into each 

other but, since the isomerization of humulone is reversible, they can be interconverted 

via humulone. When isohumulone is heated alone, in wort, in abuffer solution pH4.5, or 

in 0.1N-disodium carbonate, 10±15% of humulone is formed. The same percentage of 

humulone is found when isohumulone is shaken in atwo phase system of isooctane and a 

pH5.0 buffer solution so it follows that pure isohumulone cannot be isolated by 

countercurrent distribution. Koller (1969) found that the isomerization of the -acids was 

catalysed by divalent ions, especially calcium and magnesium, and that the iso- -acids 

prepared in this way were free of humulinic acids. As mentioned above, the calcium and 

magnesium salts of the iso- -acids can be used as bittering agents and they are formed in 

the preparation of isomerized pellets (Chapter 7). 

The separation and analysis of the six iso- -acids, as in isomerized extracts and beer, 

remains difficult; HPLC is the method of choice. Thornton et al. (1993) found that the 

trans-iso- -acids in ethyl acetate formed insoluble salts with dicyclohexylamine leaving 

the cis-isomers in solution. Thus amixture of six iso- -acids is converted into two 

mixtures of three which are easier to resolve by HPLC (Hughes, 1996). Using these 

techniques Hughes et al. (1997) were able to show that the amount of cis- and trans-iso- 

-acids formed depends on the reaction conditions employed (Table 8.2) and that the cisand 

trans-isomers behave differently. During wort boiling with different hop products 

(Type 90 pellets from Wye Target and a liquid CO2 extract from Galena) the proportion 

of isocohumulone present is at a maximum after 30 min.; the utilization of humulone and 

adhumulone is slower. This may be due to the fact that in wort cohumulone (pKa 4.7) is 

more ionized than humulone (pKa 5.5). 

The cis/trans-ratio of the iso- -acids (68:32) appears to be constant during wort boiling 

and it is generally thought that this ratio represents the thermodynamic equilibrium, i.e., the 

cis-isomers are energetically more likely to be formed during wort boiling. During 

fermentation the iso- -acids are lost by adsorption on to the yeast head. The less polar iso- 

-acids, isohumulone and isoadhumulone, interact more strongly with the yeast cells 

causing an enrichment of isocohumulone in the beer. Earlier workers had found that 

cohumulone was utilized into beer better than the other -acids but the above work 


HO 

HO 

H 

HO 

O 

(8.1a) 

H 

O O 

O 

O O 

H 

O 

H O 

B – 

O – O 

H O 

HO 

O – 

suggests that this is due to the preferential removal of the other iso- -acids. The importance 

of the proportion of isocohumulone in the beer iso- -acids on the flavour of beer is still a 

subject for debate. Rigby (1972) said that isocohumulone had a harsh bitter flavour and, in 

the days of boiling hops in wort, brewers preferred hops with a low proportion of 

cohumulone in their -acids. This harsh flavour has not been commented upon by later 

O 

H + 

O 

O 

O O 

OH 

H 

HO 

O 

OH 

(8.43a) cis (8.44a) trans 

Fig. 8.8 Mechanism for the isomerization of humulone (De Keukeleire and Verzele, 1971). 

Copyright (1971) with permission from Elsevier. 

Table 8.2 Typical ratios of cis/trans-iso- -acids under a range of isomerization conditions 

(Hughes et al., 1997) 

Means of isomerization cis-isomers trans-isomers Reference 

(%) (%) 

Magnesium oxide 80 20 Hughes (unpublished) 

Wort boiling 68 32 Verzele and De Keukeleire (1991) 

Aqueous alkali 55 45 Koller (1969) 

Light 0 100 Clarke and Hildebrand (1965) 


H +

workers. Of the four major iso- -acids found in beer cis-isohumulone was found to be the 

most bitter. (c. 1.82 times more bitter than trans-isohumulone), trans-isocohumulone was 

the least bitter (c. 0.74 that of trans-isohumulone) and there was little difference in 

bitterness between cis-isocohumulone and trans-isohumulone (Hughes and Simpson, 

1996). Hughes (2000) has reviewed the significance of the iso- -acids for beer quality. 

It is well known that beer deteriorates on storage, slowly at 0 ëC and in cool cellars, 

but more rapidly in centrally heated houses and supermarkets. Forced ageing at 37 ë or 

40 ëC is used to forecast beer stability. During the storage of beer there is a loss of iso- - 

acids probably due to autoxidation and free radical reactions accelerated by iron ions and 

hydrogen peroxide (Kaneda et al., 1992). It follows that the level of antioxidants in the 

beer will moderate the rate of deterioration. However, De Cooman et al. (2000) found 

that both in lagers and top-fermented beers the trans-iso- -acids deteriorated at a much 

faster rate than the cis-isomers. During the first 15 months of storage the loss of iso- - 

acids was mainly due to decomposition of the trans-isomers. In the trans-iso- -acids the 

double bonds in the 3-methyl-2-butenyl- and 4-methyl-3-pentenoyl-side chains lie close 

together and it is suggested that this increases their susceptibility to autoxidation. As 

mentioned earlier, the - and -acids also undergo autoxidation. The -acids, with 

geminal-isoprenyl groups, are too sensitive to autoxidation to be separated by 

countercurrent distribution. The unsaturated side-chains of the hop resins are 

undoubtedly the site of attack at the carbon next to the double bond and the 

hydroperoxides formed could break down to isobutyraldehyde (8.14) and acetone (Fig. 

H3C 

H3C 

H3C 

H3C 

H3C 

H3C 

C 

C C 

CH·C 

C 

H 

O 

H 

C 

H 

• 

CH·R 

H R 

H3C 

H3C 

O·O·H 

C C 

–H• 

O2 

H3C 

H3C 

H 

CH2·R 

H3C 

H3C 

H·O·O 

H3C 

H3C 

•C C 

C C 

H 

C O + O C 

(8.14) isobutyraldehyde (8.15) acetone 

CH·R 

H 

CH·R 

H 

CH2·R 

Fig. 8.9 Proposed autooxidation of 3-Methylbut-2-enyl side chains. 


HO 

HO 

OH 

CO·R 

O 

HO 

HO 

(8.1) α-Acids (8.45) dihydro-α-acids (8.51) dehydro-α-acids 

HO 

CO 

O 

HO 

HO 

OH 

CO·R 

(8.9) tetrahydro-α-acids (8.52) 

(8.40) iso-α-acids (8.47) tetrahydoiso-α-acids 

O 

HO 

H·C·OH 

(8.49) ρ-Iso-α-acids 

CO·R 

OH 

CO·R 

O 

HO 

CO 

8.9).Thisroutetoisobutyraldehyde isprobablymoreimportantthanthesecondpathway 

shown in Fig. 8.7. 

Obviously if these unsaturated side-chains are reduced (Fig. 8.10), the saturated 

products will be much less sensitive to autoxidation. Hydrogenation of humulone (8.1a) 

in the presence of platinum (IV) oxide give first the dihydrohumulone (8.45a) and then 

tetrahydrohumulone (8.9a). The alternative dihydrohumulone (8.46a) does not appear to 

O 

O 

OH 

HO 

H·C·OH 

CO·R CO·R 

O 

O 

HO 

OH 

O 

HO 

CO·R 

OH 

CO·R 

H3C 

H3C 

CO·R 

C CH·CH2·SH 

(8.48) 3-Methyl-2-butenyl 

mercaptan (thiol) 

(8.50) hexahydroiso-α-acids (8.46) 

Fig. 8.10 Reduced iso- -acids. 


O 

OH 

HO 

HO 

O 

OH 

CO·R 

O

e formed. As mentioned, hydrogenation of humulone in the presence of palladium 

chloride leads to hydrogenolysis loss of an isoprenyl side-chain (Fig. 8.1). Similarly 

hydrogenolysis of the -acids (8.2) gives tetrahydrodeoxy- -acids (8.8) which can be 

oxidized to tetrahydro- -acids (8.9). The tetrahydro- -acids were resolved by countercurrent 

distribution and these reactions were used to determine the analogues present in 

the -acidswhenitwasfoundthatthe -acidsarealwaysricherintheco-componentthan 

the -acids. So tetrahydro- -acids, obtained from the -acids will always be richer in 

cohumulone thanthoseobtainedbyhydrogenationofthe -acids. Inaddition,tetrahydro- 

-acids obtained by hydrogenation of -acids will be optically active, while those from 

-acids will be racemic. 

Both the dihydro- -acids and the tetrahydro- -acids can be isomerized to the 

corresponding iso- -acids. In particular, tetrahydroiso- -acids (8.47) are light stable, 

more bitter than the unsaturated iso- -acids and more potent foam stabilizers (Baker, 

1990). Thetetrahydroiso- -acids areapprovedby the FDAfor useasbitteringagentsand 

many brewers use them, with or without iso- -acids, for post-fermentation bittering. The 

tetrahydroiso- -acids can be prepared from -acids either by hydrogenation followed by 

isomerization or by isomerization followed by reduction. The latter route is preferred and 

the two stages can be combined to give an efficient one-step preparation (Hay and 

Homiski,1991).Numerouspatentsalsodescribethepreparationoftetrahydroiso- -acids. 

Analysis of beer bittered with amixture of iso- -acids and tetrahydroiso- -acids showed 

the expected 12 peaks on the HPLC chromatogram. After 12 months storage at 25ëC, 

only the trans-iso- -acids had deteriorated significantly; the tetrahydroiso- -acids were 

stable (De Cooman et al., 2000). It is noteworthy that these workers developed aHPLC 

system to resolve the 12 peaks. 

Beforetheuseoftetrahydroiso- -acids(8.47)asbitteringagentsitwasfoundthatbeer 

stored in clear glass bottles and exposed to sunlight developed an unpleasant skunky, 

sunstruck flavour found to be due to 3-methyl-2-butenyl mercaptan (thiol) (8.48). It was 

envisagedthatphotolysisoftheiso- -acidseitherproduceda3-methyl-2-butenyl-radical 

directly or produced a4-methyl-3-pentenoyl- radical which decarbonylated to the 3methyl-2-butenyl- 

radical. This radical then scavenges athiol group from any available 

sulphur amino acid orprotein (see also Heyerick et al., 2003). It was found that reduction 

of the iso- -acids with sodium borohydride produced abittering agent insensitive to 

light, theso-called rho-( )-iso- -acids (8.49).Borohydride reduction attacks the carbonyl 

group of the 4-methyl-3-pentenoyl side chain of the iso- -acids forming asecondary 

alcohol and anew chiral centre. Each iso- -acid will form two -iso- -acids and all four 

isomers of -isohumulone have been separated and characterized. They are less bitter 

than normalisohumulonebutnosignificantdifference wasnoticed between thebitterness 

of the individual isomers. Sodium borohydride reduction of tetrahydroiso- -acids (8.47) 

produces hexahydroiso- -acids (8.50). Again each steriosomer of the tetrahydroisohumulone 

will produce two isomeric hexahydrohumulones. The hexahydroiso- -acids are 

light stable, more bitter than conventional iso- -acids but less bitter than the 

tetrahydroiso- -acids and lead to an unnaturally dense foam. The properties of these 

semi-synthetic bittering agents are compared in Table 8.3. The dicyclohexylamine salts 

of the iso- -acids , the -iso- -acids, and the hexahydroiso- -acids have been prepared 

as standards for HPLC analysis (Maye et al., 1999); the tetrahydroiso- -acids may be 

used directly after recrystallization. 

The dihydrohumulone (8.45a) has been found in hop extracts, especially those 

containing the water-soluble fraction (Moir and Smith, 1995). The level decreases during 

storage and the formation is catalysed by the monovalent sodium and potassium ions. It 


Table 8.3 Semi-synthetic bittering agents (Marriott, 1999) 

Product Relative bitterness Relative foam enhancement Light stable 

at equivalent bitterness 

Iso- -acids 1.0 x No 

Rho ( )-iso- -acids 0.65 x Yes 

Tetrahydroiso- -acids 1.7 xxx Yes 

Hexahydroiso- -acids 1.1 xxxx* Yes 

* Foam unnaturally dense 

was found that at low temperatures, e.g., 3ëC, adisproportionation reaction occurs: two 

moleculesofhumulonegive onemoleculeofdihydrohumulone(8.45a)andonemolecule 

of dehydrohumulone (8.51a). At higher temperatures, e.g., 35ëC, the isomerization of the 

-acids predominates. Dehydrohumulone (8.48) is avery reactive compound which may 

either cyclize to apyran (8.52) or polymerize. 

8.2.5 Hard resins and prenylflavonoids 

By definition the hard resin is soluble in methanol and diethyl ether but insoluble in 

hexane. As hops age during storage the level of soft resins falls and that of the hard 

resins increases. Thus we must distinguish between `native' hard resins, present in 

fresh hops, and those formed by oxidation during storage. Many of the native hard 

resins are made up of prenylflavonoids which are deposited with the soft resins and 

essential oils in the lupulin glands. Hop flavonoids have been reviewed by Stevens et 

al. (1998) including, as well as the prenylflavonoids, flavonoid glycosides, condensed 

tannins and other polyphenols which are largely soluble in water. The biosynthesis of 

the flavonoids (Fig. 8.11) is similar to that of the hop resins. The aromatic amino acids 

phenylalanine and tyrosine (8.53) lose ammonia to give, for example, p-coumaric acid 

(8.54), the Coenzyme Aester of which condenses with three molecules of malonyl 

Coenzyme Ato form achalcone (8.55, chalconaringenin) which can cyclize to a 

flavanone (8.56, naringenin). Dehydrogenation of flavanones can give flavones while 

dehydrogenation and hydroxylation leads to flavanols, e.g., kaempferol (8.57) and 

quercetin (8.58). 

The major component of the prenylflavonoids in hops (Fig. 8.12) is the chalcone 

xanthohumol, first isolated in 1913 although the structure, 6'-O-methyl-3'-prenylchalconaringenin 

(8.59), was not worked out until the 1960s. It is accompanied by smaller 

amounts of the related flavanone isoxanthohumol (8.65, humulol, 5-O-methyl-8prenylnaringenin). 

This is racemic and it is thought that the lupulin glands lack the 

enzyme chalcone isomerase which converts chalcones to flavanones. Accordingly, all the 

racemic naringenin derivatives isolated from hops and beer are thought to be artefacts. 

Optically active isoxanthohumol has been isolated from Sophora angustifolia. Hops 

contain 0.1±0.8% of xanthohumol, the level falling on storage. Part of a growth of 

Eastwell Golding hops was freeze dried when the level of xanthohumol was 0.86%. 

When the same hops were kilned, with or without sulphur, the level fell to 0.31%. The 

level of isoxanthohumol appeared to be the same in all three samples. Stevens et al. 

(1997) found 8 prenylflavonoids and 3'-geranylnaringenin (8.61) in hops. Xanthohumol 

accounted for 82±89% of this fraction, desmethylxanthohumol (8.60) 2±3%, dehydrocycloxanthohumol 

(8.63) 2±4%, and dehydrocycloxanthohumol hydrate (8.64) 3±5%. 

The level of isoxanthohumol (8.65) was only 1±2% but during the brewing process nearly 

all of the chalcones are cyclized to flavanones. Xanthohumol can only cyclize to 


HO 

OH 

– 

HO CH2·CH·CO2 

O 

+ 

NH3 

HO OH 

OH 

–NH3 

HO CH CH·CO2H 

(8.53) tyrosine (8.54) p-Coumaric acid 

OR 

O 

(8.57) kaempferol (R = H) 

HO 

isoxanthohumol but desmethylxanthohumol (8.60) gives a mixture of 6- (8.66) and 8prenylnaringenin 

(8.67). Similarly, 3'-geranylchalconaringenin (8.61) can give two 

products. 6-Geranylnaringenin (8.68) has been found in beer but not in hops so it is 

presumably formed from (8.61) during the brewing process. 

During the brewing process 69% of the available xanthohumol was in solution after 

the whirlpool and 13% was recovered from the spent hops. However, further losses on to 

the trub and the yeast mean that only about 30% of the available xanthohumol ends up in 

the beer (as isoxanthohumol). Thus the level of xanthohumol in 11 beers was 0.002± 

0.69 mg/l while that of isoxanthohumol was 0.04±3.44 mg/l. Similarly, only 11% of the 

available desmethylxanthohumol was found in beer as a mixture of 6- and 8prenylnaringenin 

(Stevens et al., 1999). The prenylflavonoids show interesting biological 

activities in vitro including antiproliferative effects on cancer cells, anti-carcinogenic 

C 

CH 

O 

(8.55) chalconaringenin 

OH 

OH 

O 

O 

(8.56) naringenin 

CH OH 

HO 

OH 

+ 

3 HO2CH2·CO·S·CoA (8.31) 

OH 

Fig. 8.11 Biosynthesis of flavonoids. 


O 

O 

OR 

(8.58) quercetin (R = H) 

OH 

OH

R2 

HO 4' 

5' 

O·R3 

6' 

1' 

R1 3' 

2' 

OH O 

R1 R2 R3 

Prenyl H Me xanthohumol (8.59) 

Prenyl H H desmethylxanthohumol (8.60) 

Geranyl H H 3'-Geranylchalconaringenin (8.61) 

Prenyl Prenyl Me 5'-Prenylxanthohumol (8.62) 

HO 

O 

O·Me 

(8.63) dehydrocycloxanthohumol 

OH 

4 

(8.64) dehydrocycloxanthohumol hydrate 

HO 

R2 

O 

6 

7 

5 

OH 

R3 

8 

OH 

O 

O 

A C 

OR1 

O 

O·Me 

R1 R2 R3 

Me H Prenyl isoxanthohumol (8.65) 

H Prenyl H 6-Prenylnaringenin (8.66) 

H H Prenyl 8-Prenylnaringenin (8.67) 

H Geranyl H 6-Geranylnaringenin (8.68) 

effects, effects on lipid metabolism, oestrogenic and antimicrobial activity (Stevens, et 

al., 1998). Whether the concentrations in beer are sufficient to influence the consumer is 

yet to be determined. For example, the highest concentration of 8-prenylnaringenin, 

reported to be a phytoestrogen, found in beer was 19.8 M/litre. 

O 

2 

3 

OH 

Fig. 8.12 Prenylflavonoids in hops. 


B 

OH 

OH

8.2.6 Oxidation of hop resins 

As mentioned above hops deteriorate on storage, largely by oxidation; the soft resins 

decrease, the hard resins increase. The rate of deterioration can be followed by the Hop 

Storage Index (HSI) (p. 269). Nikerson and Likens (1979) found the regression equation 

%… ‡ † lost = 110 log (HSI/0.25) 

and that there was a linear relationship between the %( + ) lost, determined from the 

HSI, and the formation of hard resin. The rate of deterioration depends very much on the 

variety which appears to determine the length of the lag phase before deterioration starts. 

Thereafter the loss of both - and -acids can be fitted to either zero or first order kinetic 

equations (Green, 1978). To minimize such deterioration brewers used to keep their hops 

in cold store (0.20 ëC) which was expensive for such a bulky crop. Whitear (1965, 1966) 

questioned the need for this with copper hops since he found the bittering value of aged 

hops did not decline as rapidly as the analyses indicated (Fig. 8.13). However, on a 

commercial scale hops lost as much as 15±20% of the brewing value of the original - 

acids over two years at ambient temperature. From these and other trials, it was suggested 

that the level of -acids at harvest was the best guide to the amount of hops to be used in 

the copper. Accordingly, the bulk of the English hop crop was analysed within a month of 

harvest (see J. Inst. Brewing, 1969±1988). The production of hop pellets greatly reduced 

the volume of the crop and the saving in cold storage space helped to offset the cost of 

pelletization. Further, the pellets could be stored in an inert atmosphere more easily than 

bales or pockets. 

Old hops develop a cheesy aroma due to isovaleric, isobutyric and 2-methylbutyric 

acids produced by oxidative cleavage of the acyl side chains of the resins. Fresh hops 

contain 1±3% of volatile acids; after three years storage this increased to 20%. As 

mentioned earlier photolysis of colupulone (8.2b) causes loss of a 3-methyl-2-butenyl 

(isoprenyl) side chain to give deoxycohumulone (8.26b). Similarly, mild acid hydrolysis, 

wort boiling or atmospheric degradation of the -acids causes loss of isoprenyl side 

chains giving eventually the phloracylphenone (8.22). The isoprenyl side chains so 

displaced form mainly 2-methyl-3-buten-2-ol (8.69) which is absent from green hops but 

found after kilning and increases in stored hops. It is thought to be responsible for the 

soporific effects of hops and their use in hop pillows. Other volatile compounds which 

develop during hop storage include 2-methyl-2-butene, isoprene (8.23), 3-methyl-2- 

α-Acid content 

A 

Time 

B Bittering value 

α-Acids 

Lead acetate 

values 

Fig. 8.13 Schematic diagram of the changes in resin content and bittering value of hops during 

storage (Whitear, 1965, 1966). 


uten-1-ol, 5,5-dimethyl-(5H)-2-furanone (8.70), acetone, methyl isopropyl ketone, and 

methyl isobutyl ketone. The last two compounds are thought to be relics of the acyl side 

chains. 

As mentioned above, isoprenyl side chains are very sensitive to autoxidation but the 

ring system can also be attacked and such reactions can be studied in saturated 

compounds produced by hydrogenation. Many compounds have been characterized by 

deliberate oxidation of individual hop resins (Fig. 8.14) but it is often difficult to find 

these oxidation products in stored hops or beers brewed therefrom. For example, 

oxidation of humulone (8.1a) with organic peroxides in the presence of base gives transhumulinone 

(8.71), pKa 2.8, which is intensely bitter, but, over 50 years, its presence in 

stored hops has not been established unequivocally. In contrast, oxidation of humulone 

with lead tetra-acetate gives tricyclodehydroisohumulone (8.72, TCD), which has been 

found in stored hops, in amounts up to 0.3%, and in beer (4 ppm). The bitterness of TCD 

is reported to be 70% that of trans-isohumulone (8.44) and it may contribute as much as 

5% of the total bitterness of beer. Oxidation of humulone with monoperphthalic acid 

gives (8.73) which may account for 1±5% of the hard resin. This compound (8.73) may 

bean intermediate inthe formationof(8.74)obtainedby the autoxidationofhumulone in 

hexane. (8.74) has abitter taste and is more water soluble than most hop resins but the 

bitterness is lost on boiling. Oxidation of humulone with m-chloroperbenzoic acid gives 

(8.75) but the bitterness and importance of this compound is not reported. Boiling 

humulone in aqueous buffer solutions gives, in addition to isohumulone, acomplex 

mixture of products from which (8.76) has been isolated. Later, (8.76a) and (8.76b) were 

detected in beer but only in trace amounts. When oxygen is bubbled through aboiling 

solution of humulone containing Celite aseries of abeo-isohumulones (8.77±8.83 Fig. 

8.15) are formed. They are reported to be present in hops and beer; they are only slightly 

bitter but display strong foam-stabilizing activity. 

Probably the most important oxidation products of the -acids (8.2) (Fig. 8.16) are the 

hulupones (8.85) obtained by autoxidation. The hydroperoxide (8.84) has been proposed 

as an intermediate but has not been isolated. However, the saturated hydroperoxide, 

derived from (8.84), has been isolated from the autoxidation of hexahydrocolupulone. 

The hulupones (8.85) are reported to be twice as bitter as the iso- -acids; they are not 

found in green hops but accumulate during storage when concentrations of 3% have been 

reported. However, they are formed when hops are macerated in air in a blender so the 

analytical results are probably high. They are also formed in wort boiling and survive into 

beer (several ppm). They may be more important in stout brewing when hops are boiled 

with wort more than once. After the first boil, the resins that have not dissolved will be 

spread over the surface of the spent hops making them accessible to autoxidation between 

boils. Probably most claims for better utilization of the -fraction involve oxidation of 

the -acids into hulupones. In the laboratory, oxygenation of the -acids in the presence 

of sodium sulphite or oxidation with sodium dioxypersulphate, gives better yields of 

hulupones. Hulupone (8.85a), pKa 2.6, has been synthesized by the akylation of 

dehydrohumulinic acid (8.10a) with 3-methyl-2-butenyl bromide (8.24). With regard to 

bitterness of hulupones, Verzele and De Keukeleire (1991, p. 377) report that beers 

bittered only with pure hulupone (100 mg/l) were undrinkable but beers with 100 IBU 

derived from iso- -acids are also likely to be unacceptable. 

Autoxidation of hulupones in boiling ethanol gives hulupinic acid (8.86) which lacks 

bitterness but has been found in the hard resin of old hops (0.05%). Apart from the 

hulupones, autoxidation of the -acids gives complex mixtures of products. Verzele and 

De Keukeleire (1991) devote 86 pages to oxidation products of the -acids and describe 


H 3C 

H 3C 

C CH 

OH 

CH 2 

(8.69) 2-Methyl-3-buten-2-ol (8.70) 5,5-Dimethyl-(5H )-furan-2-one 

HO 

(8.71) trans-Humulinone (8.72) tricyclodehydroisohumulone (TCD) 

HO 

OH HO 

O 

HO 

O 

OH 

CO·R 

(8.73) (8.74) 

O 

CO 

CO 

O 

OH 

OH 

O 

CO·R 

OH 

CO·R 

(8.75) (8.76) 

over 40 products. Probably most of these occur in trace amounts in old hops and beers 

brewed therefrom but their importance has yet to be determined. 

Although used today principally for their bittering properties, hops were originally 

used for their preservative value. Simpson (1993b) has studied the effects of the hop bitter 

acids on lactic acid bacteria. He found that the hop bitter acids act as mobile carrier 

O 

HO 

O 

CO 

O 

HO 

OH 

Fig. 8.14 Oxidation products of the -acids. 


O 

O 

O 

O 

OH 

O 

OH 

CO·R 

CO·R 

OH 

OH 

CO·R 

OH

O 

ionophores and inhibit the growth of beer spoilage organisms by dissipating the 

transmembrane pH gradient. 

8.3 Hop oil 

O 

O 

O 

O 

O 

(8.77) (8.78) 

(8.81) 

O 

OH 

CO·R 

CO·R 

O 

(8.83) 

CO·R 


Hops produce up to 3% of essential oil which is responsible for the pleasant hoppy 

aroma of beer. It is produced in the lupulin glands along with the resins, mostly after 

resin synthesis is finished. It is largely from the aroma that the grower judges that the 

O 

HO 

OH 

O 

O 

O 

(8.79) (8.80) 

O 

O 

O 

OH 

O 

CO·R 

OH 

O 

O 

O 

O 

O 

O 

O 

(8.82) 

O 

OH 

OH 

O 

OH 

CO·R 

CO·R 

CO·R 

Fig. 8.15 Abeo-iso- -acids (Verzele and De Keukeleire, 1991). 


HO 

OH 

CO·R CO·R 

O 

HO O OH 

(8.2) β-Acids (8.84) 


O O 

Fig. 8.16 Oxidation products of the -acids. 

O 

CO·R 

O O 

(8.85) hulupones 

HO OH 

O O 

(8.86) hulupinic acid

hop is ripe and ready to pick. The composition of the essential oil depends on genetic 

(cultivar) and cultural factors but since most commercial hops are picked at an 

equivalent degree of ripeness varietal factors will dominate. Nevertheless, in any one 

variety, seedless hops will produce more essential oil than seeded hops. For reviews on 

hop oil see Sharpe and Laws (1981), Stevens (1987), Moir (1994), Deinzer and Yang 

(1994) and Siebert (1994). 

By definition, essential oils are volatile in steam, so most essential oil will be lost 

when hops are boiled in wort in a copper open to the atmosphere. To add hop aroma to 

their beers brewers either add a portion of choice hops towards the end of the boil (most 

lager brewers) or add dry hops to the beer either in cask or conditioning tank (premium 

ales). The hops used for late and dry hopping are chosen for their choice aroma, which 

may be transferred directly to the beer. Brewers are usually prepared to pay a premium 

for such choice àroma' hops. The level of -acids is immaterial since the majority will 

not be isomerized. The amount of essential oil in a sample of hops is usually measured by 

steam distillation, usually with cohobation whereby the oil is retained in a trap and the 

denser aqueous phase returns to the boiler together with any water-soluble constituents. 

Commercial hop oil, prepared in this way, does not have the true aroma of the hops from 

which it was prepared. Essential oil constituents are soluble in most organic solvents and 

CO2 so will be present in most solvent extracts of hops. The most volatile constituents 

may be removed with the solvent but this is less likely with CO 2 extracts. Hop oils 

obtained from CO 2 extracts by molecular distillation (ambient temperature and less than 

0.001 mm Hg) smell more like the original hops than steam distilled oils. Such oils may 

contain water-soluble constituents which would be washed out of steam distilled oils 

obtained by cohobation. It is also known that some essential oil constituents are altered 

during steam distillation. Before the advent of CO2 extracts some brewers used 

commercial steam-distilled oils to dry hop their beers but most people could distinguish 

such beers from those dry hopped normally. Today CO2 extracts, or fractions derived 

therefrom, are used to impart hop aromas to beers. Brewers are interested in which 

constituents of the essential oil influence the flavour of beer and, further, whether the 

composition of the essential oil can aid the identification of hop cultivars. 

The essential oil of hops is a complex mixture of well over 300 compounds, but it can 

be separated into two fractions by chromatography on silica gel. The fraction eluted with 

light petroleum consists of hydrocarbons while that subsequently eluted with diethyl 

ether consists of oxygen-containing compounds such as alcohols, acids, esters, and 

carbonyl compounds. This latter fraction may also contain traces of sulphur-containing 

compounds. 

The contribution of the individual constituents towards the overall aroma of hop oil 

can be assessed by the use of the Flavour Unit (FU) defined as: 

‰Concentration of the flavour compoundŠ 

Flavour Unit (FU) = 

‰Sensory threshold of the flavour compoundŠ 

The FU values depend on how, and in what medium, the sensory threshold was measured 

but if the concentration of a compound does not exceed the sensory threshold (FU < 1) it 

will not have a large influence on the overall flavour. Compounds providing 1±2 FU will 

be detectable by the assessor and compounds providing more than 2 FU are likely to be 

dominant flavours. For example, beers contain 10±60 mg/l of iso- -acids, the sensory 

detection level of which is 5±6 mg/l, so the iso- -acids will provide 2±12 FU and the 

bitterness will be perceived by most drinkers (Baxter and Hughes, 2001). Individual hop 


oil constituents will be discussed later but, in general, the sensory detection levels of 

sulphur compounds are much lower than those of the corresponding oxygen compounds 

which, in turn, are lower than those of the hydrocarbons. 

8.3.2 Hydrocarbons 

The hydrocarbon fraction may account for 50±80% of the essential oil and the major 

components, found in most varieties, are the monoterpene myrcene (8.89) and the 

sesquiterpenes -caryophyllene (8.136) and humulene (8.111) which were characterized 

by classical means. Farnesene (8.106), which was first isolated from Saaz (Zatec) hops, 

was found tobe present in some cultivars but not in others. For example, it was not found 

in the oils of Hallertau, Goldings, Fuggle or Cluster hops. From later plant breeding 

studies it is thought that the presence of farnesene is asex-linked character controlled by 

a single pair of genes with presence dominant to absence. Similarly, on gas 

chromatograms of the hydrocarbons of some varieties, several compounds are eluted 

after humulene. Two of these were later identified as -(8.108) and -selinene (8.110). 

The presence or absence of selinene is also controlled by asingle pair of genes but with 

incomplete dominance. Traces of many other terpenes and sesquiterpenes have been 

detected in hop oil. 

As the hop ripens, trace of oxygenated compounds of the essential oil appear first, 

then the cyclic sesquiterpenes -caryophyllene and humulene, and finally the 

monoterpene myrcene is formed. The percentage of myrcene probably reflects the 

ripeness of the cones but the humulene/caryophyllene ratio is usually constant and a 

varietal characteristic (Table 7.3). The selinene/caryophyllene ratio also appears to be a 

varietal characteristic. 

The biosynthesis of the terpenoid compounds in the essential oil uses the same 

building blocks as required for the isoprenyl side chains of the hop resins (Fig. 8.3). 

Dimethylallyl pyrophosphate (8.34) condenses with amolecule of isopentenyl pyrophosphate 

(8.33) to give geranyl pyrophosphate (8.88), the parent of the monoterpenes and 

the source of the side chain in the chalcone (8.61) and the related flavanone (8.68). With 

another molecule of isopentenyl pyrophosphate, geranyl pyrophosphate forms farnesyl 

pyrophosphate (8.105), the parent of the sesquiterpenes. Elimination of pyrophosphoric 

acid from geranyl pyrophosphate (8.88) gives the major monoterpene myrcene (8.89). 

This and other monoterpene relationships are shown in Fig. 8.17. Similarly, elimination 

of pyrophosphoric acid from farnesyl pyrophosphate (8.105) gives ( -)farnesene (8.106) 

(Fig. 8.18). Cyclization of trans, trans-farnesyl pyrophosphate can give two monocations 

(8.107) and (8.109); 8.107 can lose aproton to give humulene (8.111) while (8.109) can 

give -(8.108) and -selinene (8.110). Although -caryophyllene (8.136) nearly always 

occurs with humulene, it is thought to be formed from atrans, cis-farnesyl cation. An 

alternative deprotonation of 8.109 can lead to germacrene B(8.112), germacrene D 

(8.113) and bicyclogermacrene (8.114). Cyclization of germacrene Bcan give selina- 

3,7(11)-diene (8.115) and selina-4(15),7(11)-diene (8.116). Germacrene D(Fig. 8.19) is 

readily converted into amixture of -(8.117) and -muurolene (8.120) and -(8.118) 

and -cadinene(8.121).Allthesesesquiterpenestogetherwithcopaene(8.119)havebeen 

identified in hop oil. 

Tressl et al. (1993) found at least 15 tricyclic sequiterpenes in the essential oil of the 

German variety Hersbrucker spaÈt. These hydrocarbons are probably derived from 

bicyclogermacrene (8.114, Fig. 8.20) and include viridiflorene (8.126), alloaromadendrene 

(8.127), aromadendrene (8.128) and -gurjenene (8.129). Traces of the diterpenes 


Fig. 8.17 Monoterpene relationships in hop oil. 

m- and p-camphorene have been found in hop oil but these are thought to be artefacts 

formed by a Diels-Alder reaction between two molecules of myrcene. Few, if any, of 

these hydrocarbons survive wort boiling but traces may be found in late and dry hopped 

beers. 


(8.99) 8(9)-Menthene (8.100) p-Cymene 

(8.101) α-Pinene (8.102) β-Pinene (8.103) camphene 

Fig. 8.17 Continued 

Guadagni et al. (1966) found the following sensory thresholds: myrcene, 13; 

caryophyllene, 64; and humulene, 120 ppb. Thus, in the hydrocarbon fraction myrcene is 

by far the most potent odorant. In an oil from Brewers' Gold hops, myrcene (63%) 

accounted for 58% of the odour units. Similarly, Steinhaus and Schieberle (2000) found 

that myrcene was the most potent odorant in Spalter Select hops. 

8.3.3 Oxygen-containing components 

The oxygenated fraction of hop oil is even more complex than the hydrocarbon 

fraction but these polar constituents are more likely to survive into beer. Probably the 

first oxygenated component of hop oil to be characterized was undecan-2-one (methyl 

nonyl ketone, luparone), which is now known to be accompanied by other methyl 

ketones. Sharpe and Laws (1981) report 60 aldehyes or ketones, 70 esters, 50 alcohols, 

25 acids, 30 oxygen heterocyclic compounds and 30 sulphur-containing compounds in 

hop oil. Among the esters both straight chain and branched chain fatty acids and 

alcohols are involved. For example, the methyl esters of hexanoic, octanoic, decanoic, 

4-decenoic and 4, 8-decadienoic acids are probably by-products of fatty acid 

biosynthesis. Branched chain compounds such as 2-methylbutyl isobutyrate presumably 

arise from pathways to the carbon skeletons of amino acids. Whether these hop 

esters survive into beer is difficult to determine as similar products are formed during 

fermentation. 

During fermentation methyl 4-decenoate and methyl 4, 8-decadienoate undergo 

transesterification to produce the corresponding ethyl esters in beer. Probably other esters 

behave similarly but brewing yeasts are known to produce esterases. When methyl 

heptanoate was added to a fermentation only 35% of the parent acid was recovered with 

little methyl or ethyl heptanoate. Nevertheless, transesterification of geranyl pyrophosphate 

(8.88) is likely to be the source of geranyl acetate, propionate and isobutyrate 

which give a floral note to hop oil. Mild (enzymatic) hydrolysis of these esters gives the 

primary alcohol geraniol (8.90, R ˆ H) but under acid conditions the tertiary alcohol 

linalol (linalool, 8.87, Fig. 8.17) is formed. cis,trans-Isomerization of geraniol gives nerol 

(8.93, R ˆ H) and neryl esters have been found in hop oil. Oxidation of geraniol and 

nerol gives citral, a mixture of the aldehydes geranial (8.91) and neral (8.94) which on 

further oxidation give geranic acid (8.92 and 8.95), the methyl ester of which occurs in 

hop oil. 


OH 

O· P · P 

(8.104) nerolidol (8.105) farnesyl pyrophosphate 

+ 

+ 

O· P · P 

(8.106) β-Farnesene (8.107) (8.108) α-Selinene 

H 

(8.109) (8.110) β-Selinene (8.111) humulene 

(8.112) germacrene B (8.113) germacrene D (8.114) bicyclogermacrene 

H H 

(8.115) selina-3,7(11)-diene (8.116) selina-4(15),7(11)-diene 

Fig. 8.18 Sesquiterpene relationships in hop oil. 

Reduction of the 2, 3-double bond in either geraniol or nerol gives citronellol (8.96). 

Nerol readily cyclizes to give -terpineol (8.97) present in hop oil and beer. Dehydration 

of -terpineol gives limonene (8.98), the major hydrocarbon of citrus oils but also present 

in hop oil. Limonene is probably the precursor of the bicyclic monoterpenes such as - 

(9.101) and -pinene (8.102) and camphene (8.103). Limonene can also disproportionate 

into p-cymene (8.100) and 8(9)-menthene (8.99). Similarly, mild (enzymatic) hydrolysis 

of farnesyl pyrophosphate (8.105) can give farnesol while acid hydrolysis gives nerolidol 

(8.104); both have been found in hops and beer. For these oxygenated compounds 


H

H 

H 

H 

H 

(8.113) germacrene D 

Guadagni et al. (1966) found the following odour theshold values: undecan-2-one, 7; 

linalol, 6; methyl heptanoate, 4; methyl 4-decenoate, 3; methyl 4, 8-decadienoate, 10; 2methylbutyl 

isobutyrate, 14; geranyl acetate, 9; geranyl propionate, 10; and geranyl 

isobutyrate, 13 ppb. Thus, most of these odorants are more potent than myrcene but their 

concentration is usually much lower. For example, in the sample of Brewers' Gold oil 

mentioned above, methyl 4-decenoate contributed the largest percentage (3.0%) of the 

total odour units from the oxygenated fraction. 

During hop storage the proportion of hydrocarbons in the essential oil decreases and 

that of the oxygenated components increases. The increase in volatile acids (3 ! 20%) 

after three years storage at 0 ëC was mentioned earlier but these acids will be formed by 

oxidation of both resins and essential oil components. Carbon-carbon double bonds react 

with oxygen to form epoxides (oxiranes); in vitro peracids are the usual reagent. These 

three-membered rings are readily opened to give diols which may undergo further 

reactions. Monoterpene epoxides have not been isolated from hop oils but may be 

H 

H H 

(8.117) α-Muurolene (8.118) δ-Cadinene (8.119) copaene 

(8.120) γ-Muurolene (8.121) γ-Cadinene (8.122) α-Cadinene 

H 

H 

OH OH 

H 

H 

H 

H 

H 

H 

(8.123) δ-Cadinol (8.124) α-Cadinol (8.125) T-Cadinol 

Fig. 8.19 Sesquiterpene relationships: Germacrene D and the Cadinenes. 


H 

OH

H 

(8.126) 

viridiflorene 

H 

H 

OH 

–H + 

H 

H 

+ 

(8.114) bicyclogermacrene 

–H + 

O 

H 

H 

H + 

–H + 

(8.127) 

alloaromadendrene 

HO 

H 

intermediates in, for example, the autoxidation of myrcene which leads to linalol, 

geraniol, nerol, citral (8.91 and 8.94) together with the cyclic limonene. In contrast, 

sesquiterpene epoxides are found in hop oil and their concentration increases during hop 

storage. -Caryophyllene (8.136) (Fig. 8.21) can theoretically form two mono-epoxides 

but only one (8.137) has been found in hop oil. Some workers, but not others, have found 

H + 

H 

H 

H 

H H 

(8.128) 

aromadendrene 

H 

H 

OH 

–H + 

(8.129) 

α-Gurjunene 

(8.130) alloaromadendrene epoxide (8.131) aromadendrene epoxide 

(8.132) ledol (8.133) viridiflorol (8.134) globulol (8.135) epiglobulol 

Fig. 8.20 Sesquiterpenoids from Bicyclogermacrene. 


+ 

O 

H 

H 

OH

H H 

H O 

H H H HO 

(8.136) caryophyllene (8.137) caryophyllene epoxide (8.138) caryolan-1-ol 

Fig. 8.21 Reactions of Caryophyllene. 

it in beer. It undergoes hydrolysis to give at least six products (Deinzer and Yang, 1994). 

Similarly, caryolan-1-ol (8.138), obtained by acid treatment of caryophyllene, has been 

found in beer by some workers but not others. 

Humulene (8.111) (Fig. 8.22) can form three mono-epoxides (8.139), (8.140) and 

(8.141). Of these humulene epoxide I(8.139) is most resistant to hydrolysis and most 

likelytobefoundinbeer.HumuleneepoxideII(8.140)andhumuleneepoxide III(8.141) 

can be interconverted and hydrolysis of either gives amixture of at least 30 products; 

humulol (8.146) and humulenol II (8.147) predominate in beer. Although only traces of 

humulene diepoxides are found in fresh hops the amount increases during storage. 

Theoreticallytherearesixpairsofenantiomersofhumulenediepoxidebutonlyfivewere 

formed by treatment of humulene with m-chloroperbenzoic acid: humulene diepoxides 

A-E(8.142±8.145).HumulenediepoxideAisthemostabundantisomerandgivesatleast 

ten products on hydrolysis. There is considerable evidence that sesquiterpene oxidation 

products, and/or hydrolysis products therefrom, contribute to the hoppy aroma of beer 

although the actual compound(s) responsible have not been identified. Further, many 

lager brewers regard the aroma produced by the traditional European varieties such as 

Hallertauer mittelfruÈh,HersbruckerandTettnang±theso-called`noble'aroma±superior 

to that produced by other varieties. These hops produce high levels of sesquiterpene 

oxidation products and some brewers store their àroma' hops for aperiod to facilitate 

this oxidation. Humuladieneone (8.148) has been found in beer and associated with a 

hoppy aroma (Shimazu et al., 1974). It is not acommon oxidation product of humulene 

but it may be formed from humulenol II. 

The minor sesquiterpenes will react with oxygen in asimilar manner. -Selinene 

(8.110, Fig. 8.23) can form two monoepoxides but only one (8.149) has been found in 

hop oil together with the related tertiary alcohol, selin-11-en-4-ol (8.151). Reduction of 

the other epoxide (8.150) would give -eudesmol (8.153) which with -(8.154) and - 

eudesmol (9.155) is found in hop oil. In asimilar manner selina-3,7(11)-diene (8.115) 

and selina-4(15),7(11)-diene (8.116) would give juniper camphor (8.152) found in hop 

oil. Analogous reactions in the cadinene series (Fig. 8.19) lead to -(8.124), -(8.123) 

and T-cadinol (8.125) also found in hop oil and beer. The tricyclic sesquiterpenes found 

in Hersbrucker spaÈt hops (Fig. 8.20) can also form epoxides. Reduction of 

alloaromadendrene epoxide (8.130) gives ledol (8.132) and viridiflorol (8.133) while 

aromadendrene epoxide (8.131) gives globol (8.134) and epiglobol (8.135). Traces of 

these compounds probably occur in beers brewed with these hops. 

Various oxygen heterocyclic compounds found in hop oil and beer are shown in Fig. 

8.24. Compounds (8.156±8.160) were first characterized in Japanese hops but later found 

in German Spalter hops where the concentration was found to increase on storage. They 

are reported to contribute aflowery note to hop aroma. Various furans such as 5,5- 


H

(8.139) 

humuleme epoxide I 

O 

O 

O O 

O 

(8.142) 

humulene diepoxide A 

O 

(8.111) 

humulene 

(8.140) 

humulene epoxide II 

O 

O O 

O 

O OH OH 

(8.145) 

humulene diepoxide D & E 

(8.143) 

humulene diepoxide B 

(8.146) 

humulol 

O 

(8.148) 

humuladienone 

Fig. 8.22 Reactions of Humulene. 


(8.141) 

humulene epoxide III 

(8.144) 

humulene diepoxide C 

(8.147) 

humulenol II

H 

H O H 

H 

(8.110) β-Selinene (8.149) β-Selinene epoxide (8.116) selina-4(15),7(11)-diene 

O 

HO 

H H 

HO 

(8.150) (8.151) selina-11-en-4-ol (8.152) juniper camphor 

H OH H OH 

OH 

(8.153) β-Eudesmol (8.154) α-Eudesmol (8.155) γ-Eudesmol 

Fig. 8.23 Reactions of Selinene. 

dimethyl-(5H)-furan-2-one (8.70), 2-hexyl-5-methylfuran, dendrolasin, perillene and 

compoundssuchas(8.163,RˆPr i ,Bu i orCHMeEt)havebeenfoundinhopoilandbeer 

(Moir, 1994). Probably more important for the overall hop aroma are the traces of - 

ionone (8.164), -damascenone (8.165) and cis-jasmone (8.166) reported. These have 

verylowthresholdvalues(seeTable8.6onpage299);itisreportedthatsomeindividuals 

can smell as little as 50 fg (10 15 g) of -damascenone. 


O 

HO 

O 

O 

O 

(8.156) (8.157) (8.158) 

O 

(8.159) hop ether (8.160) karahana ether (8.161) rose oxide 

(8.162) linalol oxide (8.163) 

O 

O O 

Fig. 8.24 Oxygen heterocyclic compounds in hop oil. 


Only traces of sulphur-containing compounds are found in hop oil but many have very low 

flavour thresholds (Table 8.4). They can be detected by gas chromatography using either a 

flame photometric or a Sievers' chemiluminescence detector. Hops in the field may be 

treated with elemental sulphur to control mildew and it used to be common practice to burn 

sulphur in the oast. Both of these treatments can influence the spectrum of sulphur-containing 

compounds in the oil. For example, the sesquiterpenes caryophyllene and humulene can react 

with elemental sulphur under mild conditions to give episulphides such as (8.137), (8.139) 

and (8.140) where sulphur replaces oxygen. The level of these compounds is higher in oils 

that have been steam distilled at 100 ëC than in those obtained by vacuum distillation at 25 ëC. 

Thus more of these compounds will be introduced into beer by late hopping than by dry 

hopping. Myrcene also reacts with sulphur but less readily than the sesquiterpenes. However, 

with a suitable activator, a mixture of at least ten sulphur-containing compounds was formed 

of which (8.167) is the major component and this has been found in hop oil. 


O 

O 

O 

O 

R

Table 8.4 Sulphur compounds in the essential oil of hops 

Name Structure Flavour threshold 

(ppb) 

Methanethiol CH 3SH 0.02 

Dimethyl sulphide CH 3.S.CH 3 7.5 

Dimethyl disulphide CH 3.S.S.CH 3 7.5 

Dimethyl trisulphide CH3.S.S.S.CH3 0.1 

(2, 3, 4-Trithiapentane) 

2, 3, 5-Trithiahexane CH3.S.S.CH2.S.CH3 ± 

Dimethyl tetrasulphide CH3.S.S.S.S.CH3 0.2 

(2, 3, 4, 5-Tetrathiahexane) 

S-Methyl 2-methylpropanethioate (CH 3) 2 CH.CO.S.CH 3 40 (5) 

S-Methyl 2-methylbutanethioate CH 3CH 2CH(CH 3)CO.S.CH 3 1 

S-Methyl 3-methylbutanethioate (CH 3) 2CH.CH 2.CO.S.CH 3 50 

S-Methyl pentanethioate CH3.(CH2)3.CO.S.CH3 10 

S-Methyl 4-methylpentanethioate (CH3)2CH.CH2CH2CO.S.CH3 15 

S-Methyl hexanethioate CH3(CH2)4CO.S.CH3 1 

S-Methyl heptanethioate CH3(CH2)5CO.S.CH3 ± 

S-Methylthiomethyl CH3CH2CH(CH3)CO.S.CH2S.CH3 1 

2-Methylbutanethioate 

S-Methylthiomethyl (CH 3) 2CH.CH 2CO.S.CH 2S.CH 3 ± 

3-Methylbutanethioate 

3-Methylthiophene (8.168) 500 

3-(4-Methylpent-3-enyl)thiophene (8.169) ± 

4-(4-Methylpent-3-enyl)-3, 6- (8.167) 10 

dihydro-1, 2-dithiine 

4, 5-Epithiocaryophyllene (8.137) with S in place of O 200 

1, 2-Epithiohumulene (8.139) with S in place of O 1800 

4, 5-Epithiohumulene (8.141) with S in place of O 1500 

2-Methyl-5-thiahex-2-ene (CH 3) 2C ˆ CH.CH 2.S.CH 3 0.2 

Methylthiohumulene C15H24.S.CH3 ± 

Polysulphides also occur in hop oil. Dimethyl trisulphide (2, 3, 4-trithiapentane) is 

found only in hops that have not been treated with sulphur (dioxide) on the kiln. It is 

formed during steam distillation at 100 ëC from (S)-methylcysteine sulphoxide 

(CH3SO.CH2CH(NH2)CO2H) which is destroyed when sulphur is burnt on the kiln but 

slowly regenerates during storage. Dimethyl tetrasulphide and 2, 3, 5-trithiahexane have 

also been found in hop oil. These polysulphides have cooked vegetable, onion-like, 

rubbery sulphur aroma with low thresholds. 

Thioesters are also present in hop oil (> 1000 ppm), the concentration of which does 

not appear to be influenced by treatment of the hops with sulphur or sulphur dioxide. 

They appear to be formed by the action of heat so only low levels will be introduced 

into beer by dry hopping. Few of the sulphur compounds discussed survive wort 

boiling but late addition of hops introduces traces of these compounds, including 

thioesters, into wort. During fermentation dimethyl trisulphide and some thioesters are 

lost but some sulphur volatiles survive into beer, in particular, S-methyl 2methylbutylthioate. 

This ester and S-methyl hexanethiolate are the major thioesters 

introduced into beer by dry hopping. In the sample of Brewers' Gold hops discussed 

above Guadagni et al. (1966) found that methyl thiohexanoate contributed 4.8% of the 

total odour units. 


8.3.5 Most potent odorants in hop oil 

The pioneering work of Guadagni et al. (1966), discussed above, suggests that 

myrcene, methyl thiohexanoate, methyl 4-decenoate, caryophyllene and humulene are 

the most potent odourants in Brewers' Gold hops. Using a gas chromatographyolfactometry 

(GC-O) technique called Òsme' Sanchez et al. (1992) found that linalol 

(8.87), neral (8.94) and humulene epoxide III (8.141) were the most potent odorants in 

Hallertau mittelfruÈh and Mount Hood hops. Steinhaus and Schieberle (2000) used 

another GC-O technique ± Aroma Extract Dilution Analysis (AEDA) ± to examine an 

extract of Spalter Select hops. First, 36 areas were identified on the gas chromatogram 

by olfactory analysis. The extract was then diluted with an equal volume of diethyl 

ether and the analysis repeated. Further dilutions were made until no aroma was 

detected. The Flavour Dilution value (FD) is the highest dilution at which the aroma 

can be detected. The results of this study with dried and undried Spalter Select hops is 

given in Table 8.5 together with the results of a headspace analysis of the dried hops. 

Twenty-three odorants (FD > 4) were identified. In green undried hops (Z)-3-hexenal 

and linalol were the most potent odorants. Much of the (Z)-3-hexenal was lost during 

drying after which the most potent odorant was trans-4, 5-epoxy-(E)-2-decenal (8.171), 

probably formed by thermal cleavage of 12, 13-epoxy-9-hydroperoxy-11-octadecenoates 

(8.170). 

Table 8.5 Most odour-active compounds in hops cv. Spalter Select (Flavour Dilution values) 

(Steinhaus and Schieberle, 2000) 

Undried Dried Headspace* 

1a. Ethyl 2-methylpropionate 128 128 32 

2. Methyl 2-methylpropionate 256 128 16 

3a. (Z)-3-hexenal 2048 16 

3b. Hexanal 16 

4. Ethyl 2-methylbutanoate 32 16 16 

9. Propyl 2-methylbutanoate 16 64 8 

Dimethyl trisulphide 16 

11. 1-Octen-3-one 32 32 1 

12. (Z)-1, 5-Octadien-3-one 32 32 2 

13. Myrcene 512 1024 256 

Octanal 8 8 1 

17. Phenylacetaldehyde 16 

20. Linalol 2048 2048 256 

21. Nonanal 64 218 

14. (E, Z)-2, 6-Nonadienal 16 4 1 

23a. 1, 3(E), 5(Z)-Undecatriene 128 

23b. 1, 3(E), 5(Z),9-Undecatetrene 128 

15. 4-Ethenyl-2-methoxyphenol 32 32 

trans-4, 5-Epoxy-(E)-2-decenal 512 4096 

33. Humulene 16 8 

37. Butanoic acid 32 

38. 2-Methylbutanoic acid 64 

3-Methylbutanoic acid 

39. Pentanoic acid 4 

* Relative FD. Unidentified compounds (FD < 32) omitted. 


CH3·(CH2)4·CH 

O 

CH3·(CH2)4·CH 

C CH·CH2·CH·(CH2)7·CO2R 

O 

Δ 

OOH 

(8.170) 12,13-epoxy-9-hydroperoxy-11-octadecenoate 

CH·CH CH·CHO 

(8.171) trans-4,5-epoxy-(E)-2-decenal 

After this fatty acid oxidation product, linalol and myrcene were the most potent 

odorants in dried hops followed by ethyl and methyl 2-methylpropionate, (Z)-1,5octadien-3-one, 

nonanal, 1,3(E),5(Z)-undecatriene and 1,3(E),5(Z),9-undecatetraene. In 

the headspace of the dried hops myrcene and linalol were again the most potent odorants. 

The sesquiterpene epoxides are probably not sufficiently volatile to be found in these 

studies. It is noteworthy that none of the odorants characterized in these studies was 

classed as `hoppy'. This suggests that the typical hop aroma is probably due to a 

synergistic effect. However, it is possible there is an, as yet unknown, highly potent 

`hoppy' odorant (Siebert, 1994). 

8.3.6 Hop oil constituents in beer 

Quantitatively, any contribution that the hop makes to beer volatiles will be dwarfed by 

ethanol and other volatile products of fermentation. Also, during fermentation, carbonyl 

compoundsandpossiblyepoxidesmaybereducedtoalcohols.Thefirstdetailedstudyof 

hop volatiles in aGerman Pilsener beer was by Tressl et al. (1978) who characterized 47 

compounds in beer which were known constituents of Spalter hops. These included 28 

terpenoids andsesquiterpenoids (Table 8.6). Also included in Table 8.6 are the threshold 

values, collected from various sources, from which it is possible to judge which 

compounds make an important contribution to the overall flavour. However, it is often 

difficult to obtain some of these odorants sufficiently pure to obtain accurate threshold 

values (Siebert, 1994). From Table 8.6 it appears that hop ether (8.159), karahana ether 

(8.160), linalol and, perhaps, -ionone and -damascenone make important contributions 

to the hop aroma in beer. Humulene epoxide I may contribute but humulene 

epoxide II, humulol and humuleneol II do not produce sufficient flavour units. However, 

the hydrolysis mixture from humulene epoxide II, with a cedar, lime, banana character, 

may influence the overall flavour (Deinzer and Yang, 1994). Goiris et al. (2002) 

obtained an oxygenated sesquiterpene fraction which when added to beer after 

fermentation produced spicy or herbal flavour notes reminiscent of typical `noble' hop 

aroma. 

Peacock et al. (1980) examined pilot brews made with Hallertau mittelfruÈh and 

Washington Cluster hops and found that the Hallertau brew contained higher levels of - 

terpineol, humulene epoxide I, humulol, T-cadinol (8.125), -eudesmol (8.154) and 

humulenol II than the Cluster brew. Similarly, in an American commercial beer, brewed 

mainly with Hallertau mittelfruÈh hops, caryolan-1-ol (8.138), humulene epoxide I, - 

cadinol (8.123) and -eudesmol were found but these compounds were not detected in 

beers brewed with Cascade hops. Indeed Cascade hops introduced a floral hop aroma into 


Table 8.6 Terpenoids and sesquiterpenoids characterized in beer (Tressl et al., 1978) 

Compound Approx. conc. (ppb) Threshold value (ppb) 

(8.153) 5 ± 

(8.152) 10 ± 

(8.151) 50 ± 

Hop ether (8.159) 35 5 (in water) 

Karahana ether (8.160) 60 5 (in water) 

trans-Linalol oxide (8.157) 20 ± 

Humuladienone (8.148) 10 100 

Caryophyllene epoxide (8.139) 18 ± 

Humulene epoxide I (8.137) 125 10 (in water) 

Humulene epoxide II (8.140) 40 450 

Linalol (8.87) 470 27 

-Fenchyl alcohol 40 ± 

Terpinen-4-ol 15 

-Terpineol (8.97) 40 2000 

Citronellol (8.96) 10 ± 

Geraniol (8.90) 5 36 

Caryolan-1-ol (8.138) 25 ± 

Nerolidol (8.104) 25 ± 

Juneol 5 ± 

Epi-Cubenol 20 ± 

Caryophyllenol 5 ± 

T-Cadinol (8.125) 45 ± 

Humulol (8.146) 220 2000 

-Cadinol (8.123) 35 ± 

Humulenol II (8.147) 1150 2500 

-Ionone (8.164) 1 0.007 

-Damascenone (8.165) Tr 0.002 

cis-Jasmone (8.166) 10 ± 

All the above compounds were found in Spalter hops. 

beer (Peacock et al., 1981). Cascade hops contain high levels of linalol, geraniol and 

geranyl isobutyrate and produce beers in which the level of linalol and, in particular, 

geraniol exceed the threshold value (the threshold value of geranyl isobutyrate, 450 ppb, 

was not exceeded). All the hops examined contained linalol and Clusters, Talisman and 

Shin-shu-wase also contained high levels of geraniol. Geraniol was not found in 

Hallertau, Hersbrucker and Perle hops. 

Moir et al. (1983) compared the hop oil constituents in copper hopped, late hopped 

and dry hopped beers. They found 40±75ppb of hop oil constituents in copper hopped 

beer, the major constituents being humulene-8,9-epoxide (8.139) and -humulen-1-ol. 

In late hopped lagers they found, in addition, linalol, methyl geranate, 2-nonanol and 2undecanol. 

In dry hopped ales they found myrcene, linalol, 2-undecanone, and the 

esters isobutyl isobutyrate, isoamyl isobutyrate, isoamyl isovalerate and methyl 4decenoate. 

NickersonandVanEngel(1992),followingFosterandNickerson(1985),regardedthe 

22 components listed in Table 8.7 as being important for the hop aroma of beer. It will be 

noted that in addition to the oxidation products and the floral-estery compounds they 

included anumber of citrus-piney compounds. They measured the concentration of these 

compounds individually and as agroup in hops (nl/g) and in beer ( l/l). They regarded 

thesumoftheamountofoilcomponentsinthethreegroupsasHopAromaUnits.Having 


Table 8.7 Classification of hop aroma constituents (Nickerson and Van Engel, 1992) 

Oxidation products Floral compounds Citrus-piney compounds 

(Caryolan-1-ol) Geraniol -Cadinene 

Caryophyllene oxide Geranyl acetate -Cadinene 

(Humulene diepoxide A) Geranyl isobutyrate (Citral) 

Humulene diepoxide B Linalol Limonene 

(Humulene diepoxide C) (Limonene-10-ol) 

Humulene epoxide I -Muurolene 

Humulene epoxide II (Nerol) 

Humulene epoxide III -Selinene 

Humuleneol II 

Humulol 

Compounds not usually detected in steam-distilled oil from fresh hops are in parentheses. 

determined the Hop Aroma Units needed to produce a desirable beer, Van Engel and 

Nickerson (1992) used this concept to calculate the late hop addition required to produce 

beers of consistent hoppiness. They found that during hop storage the Aroma Units and 

the amount of - and -acids varied independantly of each other. Therefore, the 

bitterness was determined by the addition of bitter hops to the kettle after making 

allowance for the -acids in the aroma hops. 

Lermusieau et al. (2001) used aroma extract dilution analysis (AEDA) to compare 

three beers; one brewed without hops, one brewed with Saaz hops and one brewed 

with Challenger hops (added seven minutes before the end of the 75 min. boil). Fortyfive 

odours were detected in the unhopped beer with an additional 15 in the beers 

treated with Saaz hops and 16 in the beers treated with Challenger hops. Those with an 

FD >16 in the hopped beers included dimethyl trisulphide, -nonalactone and 

components suspected to be N-methylmercaptoacetamide, -damascenone and ethyl 

cinnamate. An unknown compound in hopped beers (initially thought to be 3mercaptobutan-2-ol) 

gave the highest FD values and was also present in the hops used. 

Saaz pellets were readily distinguishable from Challenger pellets by AEDA; 

Challenger pellets were richer in sulphur compounds notably dimethyl disulphide 

and diethyl disulphide. 

8.3.7 Post fermentation aroma products 

Products added to bright beer must be completely soluble to achieve 100% utilization 

and a reproducible flavour. Further filtration would remove both bitter and aroma 

products indiscriminately. Addition of whole hop oils and products containing them 

such as emulsions or oil rich fractions are likely to produce a haze. In hop oil, whether 

produced by steam distillation or molecular distillation of a CO2 extract, the 

hydrocarbon fraction is much less soluble, and contains less FU/g, than the 

oxygenated fraction. Thus, starting with a molecularly distilled CO2 extract, which 

retains the aroma of the parent hop, by liquid/liquid separation a sesquiterpene-less oil 

can be obtained with the true aroma of the hop. A 1% solution of this sesquiterpeneless 

oil in ethanol is known as a Dry Hop Essence which can be added to bright beer 

(250 g/l) to impart a hoppy aroma. Such dry hop essences can be prepared from a 

single cultivar of hops giving more control of the hop aroma produced (Marriott, 

1999). 


Table 8.8 Flavour descriptors associated with late hop essences (Marriott, 1999) 

Spicy-LHE Floral-LHE Estery-LHE Citrussy-LHE 

Sandalwood Geranium Pear skin Lemon 

Oakmoss Neroli Pineapple Grapefruit 

Astringent Rose Fruity Bergamot 

Synergist Lavender Sweet Lime 

It is generally held that late hopping and dry hopping produce different aromas in 

beers; late-hopped beers are said to have hoppy, fruity-citrus, floral and fragrant 

flavours. By column chromatography of aCO2 extract Westwood and Daoud (1985) 

produced aLate Hop Essence, which when added (150±200ppb) to bright beer could 

not be distinguished from alate-hopped beer. This essence contained at least 130 

compounds including linalol (6±20%), humulenol, linalol oxide, caryophyllene oxide, 

hop ether and humulene epoxide(s). Further fractionation (Gardner, 1994) led to four 

late hop essences (Table 8.8). The `spicy' fraction is composed of mono- and sesquiterpene 

alcohols, the `floral' fraction is principally ketones, epoxides and esters, the 

èstery' fraction is composed exclusively of methyl esters of C6 to C10 branched and 

straight chain fatty acids, and the `citrussy' fraction is acomplex mixture of terpene 

alcohols, ketones, and C 5to C 8aliphatic alcohols (Marriott, 1999). These essences 

are typically used at 50±100 g/l and are supplied as 1% solutions in ethanol. 

8.4 Hop polyphenols (tannins) 

Hops contain 0.8±1.5% polyphenols; the amount and composition of this fraction 

depends on the cultivar and growing history of the hop examined. By HPLC Forster 

et al., (1995, 1996) found over 100 compounds in the polyphenol fraction of hops. 

Many of these constituents are also found in malt including the hydroxycinnamic 

acids, the hydroxybenzoic acids and chlorogenic acid (4.133); gallic acid (4.125) has 

long been known as aconstituent of hops. Jerumanis (1985) found 817±2821mg/kg 

of (+)-catechin (4.138) in hops and the related epi-catechin (4.139) has been reported. 

Jerumanis also found 428±1472mg/kg of procyanidin B-3 (4.143) and 287±875mg/ 

kg of procyanidin C-2 (8.172) in hops; the highest concentration being found in Saaz 

hops. Acid treatment of these procyanidins will liberate the red pigment cyanidin 

(4.136) but similar treatment of hop polyphenols also liberates delphinidin (4.135) 

indicating that prodelphinidin compounds are present. Indeed the cyanidin: 

delphinidin ratio (1.2±6.2) may be avarietal characteristic (McMurrough, 1981). In 

beer proanthocyanidins will slowly react with the proteins present to form anonbiological 

haze which will limit the shelf-life of bottled beers. Brewers use various 

treatments to remove proanthocyanidins and increase the shelf-life of their beers 

(Chapter 15). Normally, malt contributes more of these compounds to beer than hops 

but the breeding and use of proanthocyanidin-free barley varieties such as Caminant 

and Clarity leave hops as the major source of these materials. Beers brewed with a 

proanthocyanidin-free malt and bittered with apure hop resin extract are said to lack 

character. 


HO 

HO 

HO 

HO 

HO 

OH 

OH 

OH 

OH 

O 

OH 

OH 

OH 

HO 

(8.172) procyanidin C3 (trim α C8-4, βC8-4, γC) 

HO 

O 

HO 

O 

O 

O·R 

(8.173) 

It was mentioned above that the lupulin glands appear to lack the enzyme chalcone 

cyclase but this must be present elsewhere in the hop since the flavanols kaempferol 

(8.57) and quercetin (8.58) are found in the polyphenol fraction mainly as their 

glycosides. Fourteen glycosides have been characterized in hops including quercitrin 

(8.58, Rˆrhamnosyl, R' ˆH), isoquercitrin (8.58, Rˆ -D-glucosyl, R' ˆH), rutin 

(8.58, Rˆ -L-rhamnosido-6- -D-glucosyl) and astragalin (8.173, Rˆ -D-glucosyl). 

Recently the 3-O- (6'-O-malonylglucosides) of kaempferol and quercetin have also been 

identified. It is noteworthy that phloroglucinol (8.18) and phlorisobutyrophenone -Dglucoside, 

putativeintermediatesinthebiosynthesisofthehopresins, have beenfound in 

the hop polyphenols. 

8.5 Chemical identification of hop cultivars 

Hop cultivars can now be determined by DNA typing but this technique is not widely 

available so interest remains in identification by chemical means. From the resins the 

percentage of cohumulone in the -acids (and colupulone in the -acids) is avarietal 

factor (Table 7.3) but more information can be obtained from essential oil analyses. By 

inspection of the gas chromatograms of the essential oil, or oils obtained by molecular 

distillation of CO2 extracts, experienced analysts can usually suggest the cultivar from 

which the oil came. Numerically the humulene/caryophyllene ratio (Table 7.3) and 

perhaps the selinene/caryophyllene ratio are varietal characteristics. Stenroos and Siebert 

(1984) applied pattern recognition and multivariate analysis techniques to hop oil 


O 

OH 

OH 

OH

chromatograms and found that SIMCA and Stepwise Discriminant Analysis were most 

successful in classifying the major hop varieties. Perpete et al. (1998) examined the oils 

obtained from pellets of twelve varieties and developed a flow chart to distinguish 

between them and Lermusieau and Collin (2001) extended the work to aged samples. De 

Cooman et al. (1998) studied only three varieties : Saaz, Nugget and Wye Target, but by 

three different techniques. The cohumulone ratio only distinguished Saaz from the other 

two varieties but essential oil analyses separated all three varieties. They also used 

flavonoid analyses, the glycosides of quercetin and kaempferol, which readily 

distinguished Wye Target from the others. The application of these techniques to more 

varieties is promised. 


AITKEN, R. A., BRUCE, A., HARRIS, J. O. and SEATON, J. C. (1970) J. Inst. Brewing, 76, 29. 

ALDERTON, G., BAILEY, G. F., LEWIS, J. C. and STITT, F. (1954) Analyt. Chem., 26, 983. 

ALDERWEIRELDT, F., ANTEUNIS, M., DIERCKENS, J. and VERZELE, M. (1965) Bull. Soc. Chim. Belg., 74, 29. 

ASHURST, P. R. and LAWS, D. (1966) J. Chem. Soc. C., 1615. 

ASHURST, P. R. and LAWS, D. (1967) J. Inst. Brewing, 73, 535. 

BAKER, C. W. (1990) Tech. Quart. MBAA, 27, 33. 

BAXTER, E. D. and HUGHES, P. S. (2001) Beer: Quality, Safety and Nutritional Aspects. Royal Society of 

Chemistry, Cambridge, pp. xiv + 138. 

BENITEZ, J. L., FOSTER, A., DE KEUKELEIRE, D., MOIR, M., SHARPE, F. R., VERHAGEN, L. C. and WESTWOOD, K. 

T. (1997) European Brewery Convention Manual of Good Practice, Hops and Hop Products, 

pp. xiv + 186. Verlag Hans Carl, NuÈrnberg. 

BURTON, J. S., STEVENS, R. and ELVIDGE, J. A. (1964) J. Inst. Brewing, 70, 345. 

CAHN, R. S., INGOLD, C. and PRELOG, V. (1951) J. Chem. Soc., 612. 

CAHN, R. S., INGOLD, C. and PRELOG, V. (1956) Experientia, 12, 81. 

CAHN, R. S., INGOLD, C. and PRELOG, V. (1966) Angewandte Chemie, 78, 413. 

CLARKE, B. J. and HILDEBRAND, R. F. (1965) J. Inst. Brewing, 71, 26. 

COLLINS, E. and SHANNON, P. V. R. (1973) J. Chem. Soc. Perkin I, 419. 

COLLINS, M. and LAWS, D. R. J. (1973) J. Chem. Soc. Perkin I, 2013. 

COLLINS, M., LAWS, D. R. J., MCGUINESS, J. D. and ELVIDGE, J. A. (1971) J. Chem. Soc., 3814. 

DE COOMAN, L., EVERAERT, E. and DE KEUKELEIRE, D. (1998) Phytochem. Anal., 9, 145. 

DE COOMAN, L., AERTS, G., OVERMEIRE, H. and DE KEUKLEIRE, D. (2000) J. Inst. Brewing, 106, 169. 

DEINZER, M. and YANG, X. (1994) EBC Monograph XXII Symposium on Hops, Zoeterwoude, p. 181. 

DE KEUKELEIRE, D. and VERZELE, M. (1970) Tetrahedron, 26, 385. 

DE KEUKELEIRE, D. and VERZELE, M. (1971) Tetrahedron, 27, 4939. 

EISENREICH, W., SCHWARZ, M. CARTAYRADE, A., ARIGONI, D., ZENK, M. H. and BACHER, A. (1998) 

Chemistry & Biology, 5, R221. 

EUROPEAN BREWERY CONVENTION. (1994) Monograph XXII. Symposium on Hops, Zoeterwoude, xviii + 

300 pp. 

EUROPEAN BREWERY CONVENTION (1997) see Benitez et al. 

FERNANDEZ, C. M. (1967) Chem. Commun., 1212 

FORSTER, A., BECK, B. and SCHMIDT, R. (1995) Proc. 25th Congr. Eur. Brew. Conv. Brussels, p. 143 

FORSTER, A., BECK, B. and SCHMIDT, R. (1996) Proc. 24th Conv. Asia-Pacific Section Institute of Brewing, 

Singapore, p. 243. 

FOSTER, R. T. and NICKERSON, G. B. (1985) J. Amer. Soc. Brew. Chem., 43, 127. 

FUNG, S. Y., ZUURBIER, K. W. M., PANIEGO, N. B., SCHEFFER, J. J. C. and VERPORTE, R. (1997) Proc. 26th 

Congr. Eur. Brew. Conv. Maastricht, 1997, 215. 

GARDNER, D. S. J. (1994) in EBC Symposium on Hops, Zoeterwoude, p. 114. 

GLASER, G. (2002) Modern Brewery Age, 53 (12), 6, 32. 

GOESE, M., KAMMHUBER, K., BACHER, A., ZENK, M. H. and EISENREICH, W. (1999) Eur. J. Biochem., 263, 

447. 

GOIRIS, K., DE RIDDER, M., DE ROUCK, G., BOEYKENS, A., VAN OPSTALE, F., AERTS, G., DE COOMAN, L. and 

DE KEUKELEIRE, D. (2002) J. Inst Brew., 108, 86. 

GOLDSTEIN, H., TING, P., NAVARRO, A. and RYDER, P. (1999) Proc. 27th Congr. Eur. Brew. Conv., Cannes, 

p. 53. 

GREEN, C. P. (1978) J. Inst. Brewing, 84, 312. 

GUADAGNI, D. G., BUTTERY, R. G. and HARRIS, J. (1966) J. Sci. Food Agric., 17, 142. 


HAY, B. A. and HOMISKI, J. W. (1991) J. Agric. Food Chem., 39, 1732. 

HERMANS-LOCKKERBOL, A. C. J. and VERPORTE, R. (1994) J. Chromatography A, 664, 45. 

HEYERICK, A., ZHAO, Y., SANDRA, P., HUVAERE, K., ROELENS, F. and DE KEUKELEIRE, D. (2003) Photochem. 

Photobiol. Sci., 2, 306. 

HUGHES, P. S. (1996) J. Chromatography A, 731, 327. 

HUGHES, P. S. (2000) J. Inst. Brewing, 106, 271. 

HUGHES, P. S. and SIMPSON, W. J. (1996) J. Amer. Soc. Brew. Chem., 54, 234. 

HUGHES, P. S., MENNNEER, I. D., WALTERS, M. T. and MARINOVA, G. (1997) Proc. 26th Congr. Eur. Brew. 

Conv., Maastricht, p. 231. 

JERUMANIS, J. (1985) J. Inst. Brewing, 91, 250. 

KANEDA, H., KANO, Y., KOSHINO, S. and OHYA-NISHIGUCHI (1992) J. Agric. Food Chem., 40, 2102. 

KOLLER, H. (1969) J. Inst. Brewing, 75, 175. 

LERMUSIEAU, G. and COLLIN, S. (2001) J. Am. Soc. Brew. Chem., 59, 39. 

LERMUSIEAU, G., BULENS, M. and COLLIN, S. (2001) J. Agric. Food Chem., 49, 3867. 

LIKENS, S. T., NICKERSEN, G. B. and ZIMMERMANN, C. E. (1970) Proc. Am. Soc. Brew. Chem., p. 68. 

MACWILLIAM, I. C. (1953) J. Inst. Brewing, 59, 142, 480. 

MARRIOTT, R. (1999) Proc. 7th Conv. Inst. Brewing, Africa Section, Nairobi, p. 140. 

MAYE, J. P., MULQUEEN, S., WEIS, S. XU, J. and PRIEST, M. (1999) J. Amer. Soc. Brew. Chem., 57, 55. 

MCMURROUGH, I. (1981) J. Chromatography, 218, 683. 

MOIR, M. (1994) EBC Monograph XXII ± Symposium on Hops, Zoeterwoude, p. 165. 

MOIR, M. (2000) J. Amer. Soc. Brew. Chem., 58, 131. 

MOIR, M. and SMITH, R. J. (1995) Proc. 20th Congr. Eur. Brew. Conv., Brussels, p. 125. 

MOIR, M., SEATON, J. C. and SUGGERT, A. (1983) Proc. 19th Congr. Eur. Brew. Conv. London, p. 63. 

MOLTKE, A. B. and MEILGAARD, M. (1955) Brygmesteren, 12, 65. 

NICKERSON, G. B. and LIKENS, S. T. (1979) J. Amer. Soc. Brew. Chem., 37, 184. 

NICKERSON, G. B. and VAN ENGEL, E. L. (1992) J. Amer. Soc. Brew. Chem., 50, 77. 

NOMENCLATURE SUB-COMMITTEE: HOPS LIAISON COMMITTEE (1969) J. Inst. Brewing, 75, 340. 

PEACOCK, V. (1998). Tech. Quart. MBAA, 35 (1), 4. 

PEACOCK, V. E., DEINZER, M. L., MCGILL, L. A. and WROLSTAD, D. E. (1980) J. Agric. Food Chem., 28, 774. 

PEACOCK, V. E., DEINZER, M. L., LIKENS, S. T., NICKERSON, G. B. and MCGILL, L. A. (1981) J. Agric. Food 

Chem., 29, 1265. 

PERPEÁ TE, P., MEÂ LOTTE, L., DUPIRE, S. and COLLINS, S. (1998) J. Am. Soc. Brew. Chem., 56, 104. 

PFENNINGER, H. B., SCHUR, F., VATERLAUS, B. P., SIGG, T. and WILD, J. (1975) Proc. 15th Congr. Eur. Brew. 

Conv., Nice, p. 159. 

POWER, F. B., TUTIN, F. and ROGERSON, H. (1913) J. Chem. Soc., 103, 1267. 

RIEDL, W. (1951) Brauwissenschaft, 81. 

RIEDL, W. (1954) Ann., 585, 38. 

RIEDL, W., NICKL, J., RISSE, K. H. and MITTELDORF, R. (1956) Ber., 89, 1849. 

RIGBY, F. L. (1972) Proc. Ann. Meet. Amer. Soc. Brew. Chem., 46. 

RIGBY, F. L. and BETHUNE, J. L. (1955) J. Inst. Brewing, 61, 325. 

RILLAERS, G. and VERZELE, M. (1962) Bull. Soc. Chim. Belg., 71, 438. 

ROHMER, M. (1998) Progress in Drug Research, 50, 136±154. 

SANCHEZ, N. B., LEDERER, C. L., NICKERSON, G. B., LIBBY, M. L. and MCDANIEL, M. R. (1992) Food Science 

and Human Nutrition Charambous, G. (ed.) Elsevier, Amsterdam, p. 370. 

SHARPE, F. R. and LAWS, D. R. J. (1981) J. Inst. Brewing, 87, 96. 

SHARPE, F. R. and ORMROD, H. (1991) J. Inst. Brewing, 97, 31. 

SHIMAZU, T., HASHIMOTO, N. and KUROIWA, Y. (1974). Proc. Amer. Soc. Brew. Chem., 33, 7. 

SIEBERT, K. J. (1994) EBC Monograph XXII. Symposium on Hops, Zoeterwoude pp. 198, 215. 

SIGG-GRUTTER, T. and WILD, J. (1974) Swiss Patent, 1458 343. 

SIMPSON, W. J. (1993a) J. Inst. Brewing, 99, 317. 

SIMPSON, W. J. (1993b) J. Inst. Brewing, 99, 405. 

SIMPSON, W. J. and SMITH, A. R. W. (1992) J. Applied Bacteriology, 72, 327. 

SMITH, R. J., DAVIDSON, D. and WILSON, R. J. H. (1998) J. Amer. Soc. Brew. Chem., 56, 52. 

SPETSIG, L. O. (1955) Acta Chem. Scand., 9, 1421. 

SPETSIG, L. O. (1964) J. Inst. Brewing, 70, 440. 

STEINHAUS, M. and SCHIEBERLE, P. (2000) J. Agric. Food Chem., 48, 1776. 

STENROOS, L. E. and SIEBERT, K. J. (1984) J. Am. Soc. Brew. Chem., 42, 54. 

STEVENS, J. F., IVANCIC, M., HSU, V. L. and DEINZER, M. (1997) Phytochemistry, 44, 1575. 

STEVENS, J. F., MIRANDA, C. L., BUHLER, D. R. and DEINZER, M. L. (1998) J. Amer. Soc. Brew. Chem., 56, 

136. 

STEVENS, J. F., TAYLOR, A. W., CLAWSON, J. E. and DEINZER, M. L. (1999) J. Agric. Food Chem., 47, 2421. 

STEVENS, R. (1967) Chem. Rev., 67, 19. 

STEVENS, R. (ed.) (1987) An Introduction to Brewing Science and Technology. Series II Volume I. Hops. 

Institute of Brewing, London, pp. vi + 118. 


THORNTON, H. A., KULANDAI, J., BOND, M., JONTEF, M. P., HAWTHORNE, D. B. and KAVENAGH, T. E. (1993) J. 

Inst. Brewing, 99, 473. 

TRESSL, R., FRIESE, L., FENDESACK, F. and KOPPLER, H. (1978) J. Agric. Food Chem., 26, 1422. 

TRESSL, R., ENGEL, K.-H., KOSSA, M. and KOPPLER, H. (1983) J. Agric. Food Chem., 31, 892. 

VAN ENGEL, E. L. and NICKERSON, G. B. (1992) J. Amer. Soc. Brew. Chem., 50, 82. 

VERZELE, M. (1958) Bull. Soc. Chim. Belg., 67, 278. 

VERZELE, M. and DE KEUKELEIRE, D. (1991) Chemistry and Analysis of Hop and Beer Bitter Acids. 

Elsevier, Amsterdam, xx + 417 pp. 

WESTWOOD, K. T. and DAOUD, I. S. (1985) Proc. 20th Congr. Eur. Brew. Conv., Helsinki, p. 579. 

WHITEAR, A. L. (1965) Proc. 10th Congr. Eur. Brew. Conv., Stockholm, p. 405. 

WHITEAR, A. L. (1966) J. Inst. Brewing, 72, 177. 

WINDISCH, W., KOLBACH, P. and SCHLEICHER R. (1927) Wochschr. Brau., 44, 453. 


9 

Chemistry of wort boiling 


Wort boiling may be regarded as the turning point in the brewing of beer; it is omitted in 

distilling and vinegar brewing. At its simplest, in Bavarian practice, the all-malt wort is 

boiled with hops for 1±2hat atmospheric pressure. Elsewhere, the sweet wort may be 

produced from amixed cereal grist and additional carbohydrate, either brewing sugars or 

wort syrups, may be added to the copper. The boiling vessels were originally made from 

copper so are often called coppers although today they are more likely to be made of 

stainlesssteel.IntheUSAthewortboilingvesselsarecalledkettles.Aspointedoutinthe 

last chapter, the addition of whole hops to the copper is likely to be replaced by hop 

pellets or extract. Alternatively, part or all of the hop grist to be added to the copper may 

be replaced by post-fermentation isomerized bittering agents and hop essences. After the 

wort has been boiled with whole hops, the copper is cast into a vessel with a slotted base 

called a hop back. A bed of spent hops is formed through which the wort is circulated 

until it runs bright, when it is run off, through a heat exchanger, into the fermentation 

vessel. Hop pellets and extracts do not provide a bed of spent hops for filtration and such 

worts are usually clarified in awhirlpool (see Chapter 10). 

The EBC have published a Manual of Good Practice ± Wort Boiling and Clarification 

(Denk et al., 2000) in which they list the principal changes that occur during wort boiling 

as: 

1. Inactivation of malt enzymes 

2. Sterilization of the wort 

3. Extraction and isomerization of compounds derived from hops 

4. Coagulation of protein material in the wort 

5. Formation of protein/polyphenol complexes 

6. Formation of flavour and colour complexes 

7. Formation of reducing substances to give the wort reducing potential, which is 

thought to protect the wort from oxidation later in the process 

8. Fall in wort pH 

9. Concentration of wort gravity through evaporation of water 


10. Evaporation of volatile compounds in wort derived from mashing 

11. Evaporation of volatile compounds in wort derived from hops. 

Other reviews have been provided by Miedaner (1986) and O'Rourke (1999, 2002). Of 

the changes discussed above, the extraction and isomerization of compounds derived 

from hops has been discussed in Chapter 8. 

Wort boiling requires alot of energy (24±54MJ/hl) and the increase in energy prices 

over the last few decades has led to many changes in plant design and practice to 

conserve energy (Chapter 10). However, any changes in wort boiling practice must be 

carefully monitored to ensure they donot affect the quality and flavour of the final beer. 

The rate of all chemical reactions increases with arise in temperature. Boiling under 

pressure increases the temperature and so shortens the time necessary to complete the 

changes that occur during wort boiling. Low-pressure boiling (c. 1bar of counterpressure) 

will raise the temperature to 104±110ëC but not shorten the reaction times 

appreciably. However, boiling plants with temperatures of 118±122ëC will require a 

holding time of only 8±10min. and those with atemperature range of 130±140ëC will 

require shorter times with acorresponding saving in energy consumption (see Chapter 

10). 

9.2 Carbohydrates 

During mashing (Chapter 4), malt and adjunct polysaccharides, mainly starch, are 

enzymatically broken down to simpler units (Table 4.15). Carbohydrates are responsible 

for 90% of the extract, 68±75% of which is usually fermentable by yeast. Whole hops 

and pellets only contribute 0.15% of the total carbohydrates in hopped wort and most 

hop extracts, prepared with solvents, will contain no carbohydrate. The carbohydrates 

undergo little change during wort boiling so the carbohydrate composition of hopped 

wort (Table 9.1) is very similar to that of sweet wort. The rate of enzyme-catalysed 

reactions also increases with arise in temperature but so does the rate of enzyme 

deactivation. Therefore the temperature òptimum' for most enzyme reactions is a 

compromise between the two processes. For example, in mashing, the `protein rest' at 

50ëC is close to the òptimum' temperature for proteases, peptidases and -glucanase 

whereas amylases show higher òptima' at 60±70ëC. Above this temperature the rate of 

enzyme deactivation dominates so that by 100ëC enzyme activity has ceased and the 

composition of the wort is fixed. In whisky manufacture, where the wort is not boiled, 

enzyme-catalysed reactions continue longer and few dextrins survive into the wash. 

Similarly, few micro-organisms will survive temperatures of 100ëC. The exceptions are 

thermophilic bacteria, mainly Bacillus sp., which form spores that survive wort boiling 

(see Chapter 17). However, standard beer is a poor growth medium for these 

organisms.. 

9.3 Nitrogenous constituents 


As discussed in Chapter 4, malt contains arange of nitrogenous constituents: proteins, 

peptides, and amino acids (Section 4.5.1), nucleic acids and related substances (Section 

4.5.2), miscellaneous substances (Section 4.5.3), and vitamins (Section 4.6.1). Many of 


Table 9.1 Carbohydrate composition of worts (results expressed in g/100 ml wort) (MacWilliam, 1968) 

a, b 

Origin (and ref.) Danish 

Canadian c 

Canadian d 

Canadian d 

German e 

Type of wort Lager Lager Lager Grain lager Lager Pale ale Pale ale 

OG 1043.0 1054.0 1048.0 1046.5 1048.5 1040.0 1040.0 

Sugar 

Fructose 0.21 0.15 0.13 0.10 0.39 0.33 0.97 

Glucose 0.91 1.03 0.87 0.50 1.47 1.00 

Sucrose 0.23 0.42 0.35 0.10 0.46 0.53 0.60 

Maltose 5.24 6.04 5.57 5.50 5.78 3.89 3.91 

Maltotriose 1.28 1.77 1.66 1.30 1.46 1.14 1.30 

Total ferm. sugar 7.87 9.41 8.58 7.50 9.56 6.89 6.78 

Maltotetraose 0.26 0.72 0.54 1.27 0.20 0.53 

Higher sugars 2.13* 2.68 2.52 2.94 2.32 1.95 

Total dextrins 2.39 3.40 3.06 4.21 2.52 2.48 

Total sugars 10.26 12.81 11.64 11.71 9.41 9.26 

Sugars (% total extract) 91.1 

Fermentability 76.7 73.7 73.7 64.1 73.3 73.2 

* The contents of maltopentanose, maltohexaose and maltoheptaose in this wort were 0.13, 0.19 and 0.18 g/100 ml wort, respectively. 

(a) Gjertsen (1953) 

(b) Gjertsen (1955) 

(c) McFarlane and Held (1953) 

(d) Latimer et al. (1966) 

(e) Kleber et al. (1963) 

(f) Harris et al. (1951) 

(g) Harris et al. (1954) 


British f 

British g

these compounds will be modified during mashing and may undergo further changes 

during wort boiling. Whole hops and pellets will also contain small amounts of these 

nitrogenous products. 

9.3.2 Proteins 

Proteins are macromolecules which may be classified according to their solubility, size 

and charge (molecular weight and isoelectric point), or function, i.e. storage proteins, 

structural proteins, and proteins that bind other molecules: hormones, immunoproteins, 

carrier proteins and enzymes. The backbone of all proteins is the polypeptide chain built 

up of amino acid residues which is partially broken down by enzymes during mashing. 

However, after the enzymes are deactivated, further cleavage of the polypeptide chain 

during wort boiling is unlikely as several hours treatment with 6N-hydrochloric acid at 

100ëC is necessary for complete hydrolysis of proteins and peptides. The primary 

structure of aprotein is the sequence of amino acids in the polypeptide chain. These 

amino acids (Fig. 4.24) contain other functional groups which can interact to stabilize the 

three-dimensional `tertiary' structure of the protein necessary for its biological activity. 

Of these cross-linking reactions, the formation of covalent disulphide bridges is very 

important. Oxidation of the thiol, -SH, group in cysteine residues gives a disulphide 

bridge, -S-S-, as in cystine residues. This reaction is reversible depending on the redox 

status of the system. In contrast, the formation of a covalent link between two phenolic 

tyrosyl residues is not readily reversible. Aspartic and glutamic acid residues in a 

polypeptide chain have a free carboxylic acid group (pKa c. 3.5) which can form salts 

(ionic bonds) with basic groups found in the side chains of lysine (pKa c. 10), arginine 

(pKa 12±13), or histidine (pKa 5.5±7.5). 

Hydrogen bonds are also important; hydrogen bonds between the C=O and the 

-NH- of the peptide bonds can lead to `secondary' structures such as the -helix and 

the -pleated sheet. Hydroxyl groups in serine and tyrosine can also form hydrogen 

bonds, and hydrogen bonds with water will aid the solubility of proteins. In contrast, 

the side chains of valine, leucine, isoleucine and phenylalanine will produce 

hydrophobic areas in the protein structure; such areas may be in the core of the 

protein structure. Proteins denature on heating, examples are curdled milk and boiled 

eggs. As the temperature rises increasing violent thermal motion disrupts the tertiary 

structure of the protein, necessary for enzyme activity, and hydrophobic groups will 

come to the surface of the structure. Here they may interact with other hydrophobic 

groups to reduce the solubility of the protein and it coagulates. In wort, the protein is 

precipitated as the hot break or trub. 

The removal of some high molecular weight protein is one of the objects of wort 

boiling. Insufficient coagulation and the removal of such proteins may effect exchange 

processes between yeast cells and the surrounding medium (membrane blocking) leading 

to an insufficient pH drop in the fermentation. The excess protein may not then be 

eliminated during the fermentation and lead to clarification problems and harsh bitterness 

in the final beer. Further, proteins surviving into the final beer may react, on storage, with 

polyphenols to form a non-biological haze which will shorten the shelf-life of the beer. 

Nevertheless, some proteins are necessary in beer to produce acceptable head retention 

and mouth-feel. 

The coagulation of proteins is strongly influenced by the pH of the wort and is most 

successful at the isoelectric point of the individual proteins, when the number of positive 

and negative charges is equal. During wort boiling the pH will drop by 0.1±0.2 pH units 


to about 5.0. This may be due to the addition of hop bitter acids, the formation of acidic 

Maillard products (see later), the precipitation of alkaline phosphates 

3Ca 2‡ ‡2HPO 2 4 ˆCa3…PO4† 2 ‡2H ‡ 

or the reaction of polypeptides with calcium, liberating protons. It is found that above 

pH5.0 the amount of nitrogen precipitated during atwo-hour boil is fairly constant but 

less is precipitated from more acidic solutions. The amount of hot break formed also 

depends on the length and vigour of the boil. Even after boiling for three hours more 

nitrogen is precipitated if the boil is continued for afurther three hours. Miedaner (1986) 

gives the level of coagulable nitrogen in unboiled wort as 35±70ppm which is reduced to 

15±25ppm after boiling with a recommended level of 15±18ppm. However, the 

measurement of coagulable nitrogen is difficult. It is usually measured as the difference 

between the total soluble nitrogen (TSN), determined by Kjeldahl analysis of an aliquot 

of the hot water extract, and the permanently soluble nitrogen (PSN), remaining after 

boiling. It is this last analysis which is not very reproducible. 

Worts held at 98±100ëC without boiling or agitation remain turbid which explains the 

importance of the vigour of the boil. The intensity of wort circulation depends on copper 

design (Chapter 10) and evaporation rate. In classical wort boiling at atmospheric 

pressure, evaporation of 8±10% of the wort volume/h was recommended, i.e., up to 20% 

of the wort volume over two hours. This requires alot of expensive energy and to-day 

modern kettles operate with a60min. boil with 5±8% evaporation (O'Rourke, 1999). 

Reed and Jordan (1991) claim that evaporation in excess of 2±3% can be replaced by 

suitable agitation. In their laboratory apparatus the fastest clarification was found with 

stirringat125rpm.Insufficientorexcessiveagitationincreasesclarificationtime.Itisthe 

increase in floc size that is responsible for rapid sedimentation and excess agitation will 

have ashearing effect on the flocs. The hot break or trub is largely composed of protein. 

It was long thought that polyphenols were also involved but protein/polyphenol 

complexes, based on hydrogen bonds, are not stable at 100ëC. However, protein/ 

polyphenol complexes become increasingly stable below 80ëC and are found in the cold 

break, but this accounts for only 2.5% of the material precipitated (Crompton and 

Hegarty, 1991). 

Proteins can be separated according to molecular size by elution from Sephadex gels. 

These experiments (Table 9.2) suggest that more high-molecular weight compounds are 

precipitated during wort boiling than smaller molecules. Nevertheless, some malt 

proteins survive into beer (Sections 4.5.1 and 19.1.5). Proteolysis during mashing 

produces arange of amino acids (Table 9.3). Whole hops and pellets will add small 

amounts of these compounds during wort boiling. Small amounts of nitrogen may be lost 

as volatile compounds during wort boiling and some may be incorporated into 

melanoidins but the amino acid spectrum of sweet and boiled wort is very similar. The 

aminoacidspectrumofbeerisverydifferentasmanyofthesecompoundsaretakenupas 

yeast nutrients during fermentation. 

Table 9.2 Effect of boiling on the MW distribution of wort proteins (Guenther and Stutler, 1965) 

Mol. Wt < 5000 5±10,000 10±50,000 50±100,000 > 100,000 

Boiled for 95 min. 0.0175 0.0125 0.0040 0.0010 0 

Not boiled 0.0336 0.0185 0.0101 0.0023 0.0028 


Table 9.3 Free amino acids in wort and beer (mg/100 ml) (Sandegren et al., 1954) 

Nitrogen and amino acids Wort Hopped wort Beer Beer 

refermented 

Total nitrogen 88.0 84.8 62.6 47.0 

Low molecular nitrogen alcohol soluble 63.4 69.5 50.7 35.1 

Total -amino nitrogen 42.7 38.0 21.0 13.0 

Alcohol soluble -amino nitrogen 37.6 30.8 18.2 2.5 

Alanine 9.8 10.2 7.7 1.8 

-Amino butyric acid 8.3 7.9 9.6 2.5 

Arginine 13.8 5.9 3.0 0.6 

Aspartic acid 7.0 9.8 1.6 1.0 

Glutamic acid 6.4 3.3 0.8 0.7 

Glycine 2.3 2.6 2.1 1.3 

Histidine 5.7 3.8 2.8 0.2 

Isoleucine 6.2 6.5 2.1 0.3 

Leucine 18.1 17.5 4.7 0.7 

Lysine 14.9 10.7 2.2 0.5 

Phenylalanine 13.7 14.0 4.4 0.6 

Proline (imino acid) 45.7 48.3 31.8 33.3 

Threonine 5.9 7.3 0.3 0.3 

Tyrosine 10.6 9.3 5.9 1.1 

Valine 11.9 16.0 6.8 0.4 

Serine + Asparagine mM in 100 ml 168.6 171.8 7.9 5.6 

Ammonia 2.4 2.4 1.7 1.0 

9.4 Carbohydrate-nitrogenous constituent interactions 

Sweetwortboiledwithorwithouthopsincreasesincolour.Thisisdueto`non-enzymatic 

browning', a reaction between amines, or amino acids, and carbonyl compounds, 

especially reducing sugars. It is often named after its discoverer Louis-Camille Maillard 

(Ikan, 1996, Fayle and Gerrard, 2002). The pigments formed by non-enzymatic browning 

are called melanoidins and the ultimate product is caramel. Non-enzymatic browning 

should be distinguished from the action of polyphenoloxidase on substrates such as 

tyrosine, which produces the brown or black hair and skin pigments called melanins. 

In the brewing process the Maillard reaction occurs when malt is kilned and continues 

during wort boiling. Obviously, dark and crystal malts will contain more melanoidins 

than pale ale orlagermalts.It was estimatedfor an American beer (presumably pale) that 

about one-third of the colour was formed during kilning and the other two-thirds during 

wort boiling but the ratio will be different using dark malts. In addition to these pigments 

the Maillard reaction produces many volatile compounds, some of which have very low 

flavour thresholds and can influence the flavour of beer. As well as aliphatic compounds, 

oxygen, nitrogen and sulphur heterocyclic compounds are formed (Mottram, 1994). 

There are many reviews of the Maillard reaction (Hodge, 1953; Reynolds, 1963,1965; 

Nursten,1980;WallerandFeather,1983; Parlimentetal.,1994;Ikan,1996;andFayleand 

Gerrard,2002).ThebasicchemistryisshowninFig.9.1.Theamine,oraminoacid,addsto 

the reducing group of the reducing sugar, in the aldehyde form (9.1), to give a product which 

is dehydrated to a Schiff's base (9.2). This rearranges to an Amadori compound (an Nsubstituted 

1-amino-1-deoxy-2-ketose, (9.3) in the keto form, (9.4) in the enol form). 

Amadori compounds characterized in malt include those from: Fru-Ala, Fru-Gly, Fru-Val, 

Fru-Leu, Fru-Ile, Fru-Ser, Fru-Thr, Fru-Pyr, Fru-Asp, Fru-Glu, Fru- -aminobutyric acid, 

and Fru-Pro (Eichner et al., 1994). Fru-Pyr, fructose-pyrrolidonecarboxylic acid, is formed 



Fig. 9.1 The chemistry of non-enzymatic browning.

fromFru-GluandFru-Glutamineduringworkup.Darkmaltscontainhigherconcentrations 

of Amadori products but they cannot be found in malts heated above 200ëC. Considerable 

degradation of Amadori products occurs during mashing and wort boiling, but apparently 

not during fermentation, and some products survive into beer. Amadori compounds can 

decompose by two routes: 1,2-enolization occurs at low pH values and 2,3-enolization at 

higherpHvalues.Eichneretal.,(1994)studiedthemodelsystemFru-Glyduringtwohours 

inacitratebuffersolution(pH3.0)at90ëCandfoundthatthemajordecompositionwasvia 

3-deoxyglucosone (9.6, R ˆ CH2OH) to 5-hydroxymethylfurfural (HMF, 9.8, R ˆ 

CH2OH). The proportion of 3-deoxyglucosone reaches amaximum after 15hafter which 

nomoreaccumulates(formationˆdecomposition)buttheconcentrationofHMFcontinues 

to increase throughout the reaction. Indeed, the level of HMF can be used to follow the 

Maillard reaction. 

In beer acorrelation between the stale flavour and the HMF content has been reported 

(Shimizu et al., 2001). Increased, and possibly unacceptable, levels of HMF and other 

Maillard products are formed during high temperature wort boiling (Fig. 9.2). It is 

recommended (Miedaner, 1986) that the temperature should not exceed 140ëC and the 

holding time should not exceed 2.5min. at 140ëC and 3min. at 130ëC. In the same way 

pentose sugars will produce furfural (9.8, RˆH, 9.17). Related heterocyclic compounds 

found in roasted barley (Harding et al., 1978), malt (Tressl et al., 1977) and beer 

(Harding et al., 1977; Tressl et al., 1977) are shown in Fig. 9.3±9.7 and in Tables 9.4 and 

9.5.Thelevelof2-acetylfuran(9.25),2-acetylthiophene(9.27),furfurylalcohol(9.20),5methylfurfural 

(9.22), and 5-methylthiophenecarboxaldehyde (9.24)) is higher in ales 

than in lagers probably due to the higher kilning temperatures used in the preparation of 

themalts. Comparedtopalealemalt,crystal maltcontainsenhancedlevelsofmost ofthe 

heterocyclic compounds discussed. 

2,3-Enolization of Amadori products gives rise to 1-deoxyosones (9.9). With hexose 

sugars, e.g. (9.9, RˆCH2OH), dehydration can give maltol (3-hydroxy-2-methyl-4Hpyran-4-one, 

9.11), isomaltol (1-(3-hydroxy-2-furanyl)ethanone, 9.30), and 2,5dimethyl-4-hydroxy-3(2H)-furanone 

(furaneol, 9.12, RˆCH 3). All these compounds 

have sweet caramel flavours and maltol (9.11) and furfuryl mercaptan (9.29) occur in 

beerabove the flavour threshold concentrations.Compounds,such as9.10,which contain 

HMF (mg/l) 

50 

40 

30 

20 

10 

0 0 60 120 180 240 

Heating time (s) 

150°C 

140°C 

130°C 

120°C 

Heating time (s) 

10 4 

10 3 

10 2 

τ H = f〈σ, [‘HMF’]〉 

HMF 

25 

20 

15 mg/l 

120 130 140 150 


Fig. 9.2 Hydroxymethylyfurfural production during high temperature boiling (Miedaner, 1986). 


X = O 

X = NH 

X = S 

X 

CHO 

(9.17) Furfural 

(9.18) 2-Formylpyrrole 

(9.19) 2-Formylthiophene 


X 

(9.20) Furfuryl alcohol 

— 

(9.21) 2-Hydroxymethylthiophene 

CH2OH H3C CHO 

X 

(9.22) 5-Methylfurfural 

(9.23) 2-Formyl-5-methylpyrrole 

(9.24) 2-Formyl-5-methylthiophene 

O O 

O 

CO2H CH2SH CO·CH3 

(9.28) Furoic acid (9.29) Furfuryl mercaptan 

(9.30) Isomaltol 

Fig. 9.3 Heterocyclic compounds in roasted barley, wort and beer. 

CO·CH3 

X 

(9.25) 2-Acetylfuran 

(9.26) 2-Acetylpyrrole 

(9.27) 2-Acetylthiophene 

OH

R O 

C 

R' 

R 

C 

C 

O 

C 

R' O 

NH2 

+ 

H2N 

R N R' 

R' N R 

R 

R' 

+ R"·CH 

NH2 

COOH 

α-Diketone α-Amino acid 

α-Aminoketone 

N 

N 

[ – O] 

R' 

R 

Alkylpyrazine 

O 

+ R"·CHO 

Strecker 

aldehyde 

C 

C 

R' 

R 

+ H2O 

the -C(OH) ˆC(OH).CÔ grouping are called reductones and combine with oxygen to 

maintain the oxidation/reduction (redox) balance of asystem. Another example of a 

reductone is ascorbic acid (vitamin C). Fragmentation of the deoxyosone intermediates 

can lead to the -dicarbonyl compounds, pyruvaldehyde (9.13) and 2,3-butanedione 

(diacetyl, 9.14) and the related ketols hydroxyacetone (9.15) 3-hydroxy-2-butanone 

(acetoin, 9.16). 

The deoxyosones and other -dicarbonyl compounds can react with amino acids 

according to the Strecker reaction (Fig. 9.4) to give aldehydes, with one carbon less than 

the amino acids, carbon dioxide, and an -amino ketone. Most of the Strecker aldehydes 

(Table 9.6) have potent aromas but those that survive wort boiling are likely to be 

reduced to the corresponding alcohols during fermentation. The amino ketones formed in 

the Strecker reaction can condense together to form, after oxidation, pyrazines (9.34). 

The pyrazines found in roasted barley malt and beer are listed in Table 9.6. As illustrated 

in Fig. 9.4, the amino acid cysteine breaks down in the Strecker reaction to liberate 

O 

R" 

C 

O 

H 

H 

R" 

N 

O 

N 

O 

– CO2 

C 

C 

C 

R 

C R' 

R 

R' 

For cysteine (R" = HS·CH ) 

CH3CHO + NH3 + H2S 

Fig. 9.4 Strecker reaction of -amino acids and formation of alkylpyrazines. 


R 

R' 

+ 

C 

C 

O 

O 

2–

N 

H 

H 

COOH 

(4.44) Proline α-Diketone 

N 

H 

N N 

CO·CH3 

(9.36) Pyrrolidine (9.37) 1-Pyrroline (9.38) 2-Acetyl-1-pyrroline 

N 

H 

N 

CO·CH3 

(9.40) 2-Acetyl-1,4,5,6tetrahydropyridine 

CO·CH3 

+ 

hydrogen sulphide, which is probably incorporated into the sulphur heterocyclic 

compounds formed during wort boiling. The imino acid proline is not strictly an amino 

acid but is the major àmino acid' in wort (Table 9.3). In the Strecker reaction it cannot 

form a Strecker aldehyde and an -amino ketone but it reacts with -dicarbonyl 

compounds to form nitrogen heterocyclic compounds such as 1-pyrroline (9.37), 

pyrrolidine (9.36), 1-acetonyl-2-pyrroline (9.39), and 2-acetyl-1,4,5,6-tetrahydropyridine 

(9.40) (Fig. 9.5). 

Roberts and Acree (1994) studied the Glu-Pro Malliard reaction by GC-Olfactometry. 

The seven most potent products, characterized by olfactometry, were: diacetyl (9.14, 

%`Charm', 0.5), 2-acetyl-1-pyrroline(9.38, 19), 2-acetyl-1,4,5,6-tetrahydropyridine 

(9.40, 12), 2-acetylpyridine (9.42, 19), 2-acetyl-3,4,5,6-tetrahydropyridine (9.41, 63), 

furaneol (9.12, RˆCH3, 3.9) and 5-acetyl-2,3-1H-pyrrolizine (9.43, 0.3). According to 

O 

O 

C 

C 

N 

O 

(9.39) 1-Acetonyl-2-pyrroline 

CO·CH3 CO·CH3 

N N 

(9.41) 2-Acetyl-3,4,5,6tetrahydropyridine 

O 

(9.42) 2-Acetylpyridine 

(9.43) 5-Acetyl-2,3-1H-pyrrolizine (9.44) Maltoxacine 

Fig. 9.5 Products of the Strecker reaction with proline. 


N 

O

SH 

NH2 

(9.45) 

Cysteamine 

+ H 

O 

O 

(9.13) 

Pyruvaldehyde 

Flavour profile score 

1.0 

0.8 

0.6 

0.4 

0.2 

N 

H 

S 

CO·CH3 

(9.46) 

2-Acetylthiazolidine 

N 

the mass yield this last compound was the major product of the reaction. Maltoxazine 

(9.44) was also formed in considerable amounts but had little or no aroma. Details of 

Maillard reactions between most of the amino acids (Table 9.3) and the common sugars 

can be found in the literature. 

The presence of thiazoles in roasted barley, malt and beer is recorded in Table 9.4. Of 

particular interest is 2-acetyl-2-thiazoline (9.47) and 2-acetylthiazole (9.48) which have 

popcorn flavours with low thresholds. The former compound is thought to be formed 

(Fig. 9.6) by condensation of cysteamine (9.45) (formed by decarboxylation of cysteine) 

and 2-oxopropanal (9.13, pyruvaldehyde) to give 2-acetylthiazolidine (9.46) which, by 

dehydrogenation, gives (9.47) and probably (9.48). A Maillard reaction between 

cysteamine and fructose produces N-(2-mercaptoethyl)-1,3-thiazolidine (9.49) which has 

an intense popcorn-like odour with an extremely low threshold (0.005ng/l in air) (Engel 

andSchieberle,2002).Thiazolesmayalsobeformedbydegradationofthiamine(vitamin 

B1, 4.68). 

As well as these Maillard reaction products, malt and hops will contribute volatile 

compounds to wort which, if not partially removed, will lead to unacceptable beers. 

Normally the excess of these compounds is lost by evaporation during wort boiling. As 

– 2H 

N 

S 

CO·CH3 

(9.47) 

2-Acetyl-2-thiazoline 

SH 

(9.49) N-(2-Mercaptoethyl)-1,3-thiazolidine 

S 

Fig. 9.6 Formation of thiazoles. 

Sulphidic 

Grassy/grainy 

0 0 1 2 3 4 5 

Evaporation (%) 

Fruity 

– 2H 

N 

S 

(9.48) 

2-Acetylthiazole 

Fig. 9.7 Effect of evaporation on some important flavour characters (O'Rourke, 1999). 


CO·CH3

Table 9.4 Derivatives of pyrrole, thiazole and pyridine in roasted barley, malt and beer (see text 

for references) 

(9.31) Pyrrole (9.32) Thiazole (9.33) Pyridine 

Unsubstituted Unsubstituted Unsubstituted 

2-Methyl 4-Methyl Methyl 

5-Methyl 

Dimethyl 

1-Acetyl 

2-Acetyl (9.26) 2-Acetyl 2-Acetyl 

3-Acetyl 

2-Formyl-1-methyl 

2-Formyl-5-methyl 

1-Ethyl-2-formyl 

1-Furfuryl 

5-Hydroxyethyl-4methyl 

Table 9.5 Pyrazines in roasted barley, malt and beer (for references see text) 

(9.34) Pyrazine (9.35) Cyclopentapyrazine 

Pyrazine Concentration (ppb) 

Malt Beer 

1. Methylpyrazine 280 70 

2. 2, 5-Dimethylpyrazine 130 110 



5. Ethylpyrazine 140 10 

6. 2-Ethyl-6-methylpyrazine 80 }35 

7. 2-Ethyl-5-methylpyrazine 40 } 

8. 2-Ethyl-3-methylpyrazine 80 + 

9. Trimethylpyrazine 320 20 

10. 2-Ethyl-3, 6-dimethylpyrazine 10 20 

11. 2-Ethyl-3, 5-dimethylpyrazine 30 10 

12. 2-Ethyl-5, 6-dimethylpyrazine 10 + 

13. Tetramethylpyrazine 110 + 

14. 6, 7-Dihydro-5H-cyclopentapyrazine + 10 

15. 5-Methyl-6, 7-dihydro-5H-cyclopentapyrazine 20 15 

16. 2-Methyl-6, 7-dihydro-5H-cyclopentapyrazine 20 10 

17. 5-Methylcyclopentapyrazine 10 + 

18. 2-Furfurylpyrazine + 25 

19. 2-(2'-Furfuryl)methylpyrazine 10 

20. 2-(2'-Furfuryl)dimethylpyrazine + 

+ Detected but not quantified 

Not detected 


Table 9.6 Aldehydes produced in the Strecker reaction (after Ho, 1996) 

Amino acid Aldehyde Odour properties Flavour 

threshold 

(ppm) a 

Alanine Acetaldehyde Pungent, ethereal, green, sweet 10 

CH3.CHO 

Valine Isobutyraldehyde Extremely diffusive, pungent, green, (1.0) 

(CH3)2CH.CHO in extreme dilution almost pleasant, 

fruity, banana-like 

Leucine Isovaleraldehyde Very powerful acrid-pungent. In (0.6) 

(CH 3) 2CH.CH 2CHO extreme dilution fruity, rather pleasant 

Isoleucine 2-Methylbutanal Powerful, but in extreme dilution 1.25 

CH3CH2CH(CH3)CHO almost fruity ± `fermented' with a 

peculiar note resembling that of 

roasted cocoa or coffee 

Methionine Methional Powerful onion-meat-like, potato-like (0.25) 

CH3S.CH2CH2CHO 

Phenylalanine Phenylacetaldehyde Very powerful, pungent, floral and sweet (1.6) 

C6H5CH2CHO 

a Meilgaard, 1975 

shown in Fig. 9.7 grassy/grainy and sulphidic aromas are greatly reduced with only 2% 

evaporation. An important malt-derived volatile is dimethyl sulphide (DMS, 4.112), the 

flavour threshold of which is 40±60ppb but some all-malt lagers with 100ppb DMS are 

found acceptable. Above this level DMS gives asweetcorn flavour. DMS is produced by 

thermal decomposition of S-methylmethionine (Fig. 4.34), the half-life of which is 

reported to be 35min. at 100ëC. DMS formed by kilning and wort boiling will be rapidly 

lost by evaporation but S-methylmethionine will continue to break down during wort 

cooling and the DMS formed then will persist into beer. To minimize such DMS 

formation it is recommended (O'Rourke, 1999, 2002) to use malts with low Smethylmethionine 

contents and to extend the wort boiling time to decompose the 

majority of the precursor and drive off the DMS. Worts from high-temperature wort 

boiling systems contain negligible amounts of DMS and its precursors. It is also 

recommended to minimize the whirlpool stand time and to use quick wort cooling to 

reduce the time that the wort is held hot. 

When wort is boiled with whole hops or pellets, the majority of the hop oil 

constituents will be lost during a60±90min. boil in an open copper. If late hop character 

is required aportion (up to 20%) of the hop gristmay be added, as choice aroma hops, 5± 

15min. before the end of the boil. Early attempts at high temperature wort boiling, with 

insufficient venting, produced worts with unacceptable levels of hop oils. Excess 

Maillard volatile products must also be evaporated. Figure 9.8 shows the amounts of 

various heterocyclic compounds in the vapour condensate during wort boiling. Of 

particular interest is 2-acetylthiazole, which has a flavour threshold of 10 ppb in beer, and 

must be reduced if not to cause an off-flavour. 

9.4.1 Melanoidins 

The volatile products of the Maillard reaction have been studied in more detail than the 

melanoidin pigments. These are obviously heterogeneous depending on the sugars and 

amino acids involved, their ratios and the pH and temperature of the reaction (Ikan, 1996). 


μg per 6l condensate 

300 

200 

100 

2-Ac-thiazole 

2-Ac-pyridine 

2-Ac-pyrrole 

2-Mc-pyrazine 

0 0 20 40 60 80 100 120 

Boiling time (min.) 

Fig. 9.8 N-Heterocyclic compounds in the evaporation condensate during wort boiling 

(Miedaner, 1986). 

At neutral or slightly acidic pH, osones, furfurals, and probably other heterocyclic 

compoundsare involved inthe formation of melanoidins, which are highmolecular weight 

compounds (10,000±30,000 daltons by ultracentrifugation; 20,000 by gel filtration). 

Isoelectric focusing electrophoresis of the melanoidins from aGlu-Gly reaction showed 20 

bands while those from aXyl-Gly reaction showed 16 bands. Spectroscopic studies of the 

products, with and without isotopic labelling, and ozonolysis has led to the suggestion that 

the melanoidins contain the repeating structure shown in Fig. 9.9. The melanoidins, which 

may resemble the humic acids in soil, show many interesting biological reactions which 

may influence the brewing process (Ikan, 1996). For example, they show antimicrobial 

activity, especially towards enteric bacteria; they inhibit trypsin and have dietary fibre-like 

action; they show mutagenic action; they react with metal ions and show antioxidative 

effects. In particular, melanoidins scavenge hydroxyl radicals, hydrogen peroxide and 

superoxides (Hayase, 1996) and have an influence on beer foam. 

9.4.2 Caramel 

TheFAO/WHOandtheEECScientific Committeeforfoodhaveacceptedfourclassesof 

caramel (Thornton, 1989): 

ClassI. Caramelpreparedbythecontrolledheattreatmentofcarbohydrateswithor 

without the presence of food quality alkali or acid. 

Class II. Caramel prepared by the controlled heat treatment of carbohydrates with 

caustic sulphites. 

Class III. Caramel prepared by the controlled heat treatment of carbohydrates with 

ammonia. 

Class IV. Caramel prepared by the controlled heat treatment of carbohydrates with 

ammonium and sulphite containing compounds. 


CH O CH O 

CH R 

C N R C N R 

HC N CH R 

CH2 

HC OH 

CH 

HC OH 

O 

CH C 

C OH CH 

N CH2 

C 

R 

N CH2 R 

HC 

R' 

OH HC 

R' 

OH 

CH 

R' 

C 

HC 

OH 

OH 

O 

CH2 

C O 

C 

CH 

N 

R' CH C OH 

R – NH2 = amine 

R' C OH 

R' = H or CH2OH R' 

^ 

n 

Fig. 9.9 Possible repeating units of melanoidins and their precursors 

(after Kato and Tsuchida, 1981). 

Class Icaramels are used inspiritsandliqueurs, Class IIcaramels invermouths andother 

aperitifs, and Class IV caramels in soft drinks. Class III caramels, ammonia caramels, are 

used in bakery and meat products and in brewing. Ammonia caramel (E 150 (c)) is used 

for colouring/colour adjustmentofbeer andistheonlycolouring matter permittedinbeer 

in the UK. These electropositive caramels are prepared from glucose syrups, with high 

dextroseequivalents,and0.880ammonia.Theyareallowedtoreacttogetherforatleasta 

week at ambient temperature, then at 90ëC overnight, and finally at 120ëC for about 

three hours. Heating must be carefully controlled to maintain abalance between colour 

and viscosity. At the appropiate time the mixture is cooled to 80ëC, softened water added 

and the product blended as required. 

The product contains about 25% extract and 32,000±48,000 EBC colour units (see 

Chapter 19). The UK Food Advisory Committee has recommended that there should be a 

limit for caramel in beer of 5000 mg/kg. Other committees have established an acceptable 

daily intake of ammonia caramel of 200 mg/kg bodyweight for man. The EEC 

specification states that not more than 50% of the colour should be bound by DEAE 

cellulose but more than 50% should be bound by phosphoryl cellulose. Caramel contains 

0.7±3.3% total nitrogen but not more than 0.3% ammoniacal nitrogen and not more than 

0.2% total sulphur. The EEC also proposes limits for two possible by-products of caramel 

production (Fig. 9.10): not more than 250 mg/kg of 4-methylimidazole (9.50) and not 

more than 10 mg/kg of 2-acetyl-4-tetrahydroxybutylimidazole (9.51). There are the usual 

limits for heavy metals, etc. Most commercial caramels will comply with these limits. 

Class IV electronegative caramels cannot be used in beer as they react with the 

electropositive finings. Caramel can be added to the copper but is usually added with 

primings to make minor adjustments to the colour of beer. However, the fundamental 

colour of beer will be determined by the choice of malt and adjuncts added to the copper. 

H3C 

N 

H 

N HO·CH2·CH(OH)·CH(OH)·CH(OH) N 

N 

H 

CO·CH3 

(9.50) 4-Methylimidazole (9.51) 2-Acetyl-4-tetrahydroxybutylimidazole 

Fig. 9.10 Imidazoles limited in a caramel. 


9.5 Protein-polyphenol (tannin) interactions 

The polyphenols in malt (Section 4.4.9) and hops (Section 8.4) have been discussed earlier 

(see pp. 158±60; 302 for structures). The phenolic acids (hydroxybenzoic acids and 

hydroxycinnamic acids) and the hop flavonols (usually as glycosides) readily dissolve 

during mashing and wort boiling and most survive into beer. McMurrough and Delcour 

(1994) regard the flavanoids (proanthocyanidins) as the most important polyphenols in 

wort and classify them as: simple flavanols ((+)-catechin, ( )-epicatechin, dimeric and 

trimeric flavanoids); polymeric flavanols (òxidation' products of the simple flavanoids, 

and complex flavanols (in water-soluble association with polypeptides). 

Polyphenols form strong hydrogen bonds with polypeptides and these complexes are 

comparatively inert. As already mentioned they dissociate above 80ëC and the 

polypeptide can be displaced with urea, H2N.CO.NH2. They are non-tanning and their 

presence in beers does not necessarily lead to haze formation, even in beers stored in 

bottles with excessive headspace air. Simple flavanoids are very susceptible to free 

radical reactions initiated either by oxygen or plant oxidase/peroxidase enzymes. These 

reactions can lead to cross linking, an increase in molecular size, the formation of aredbrown 

colouration and the formation of polymeric flavanols which are strongly tanning. 

The tanning properties of the simple flavanoids increase with increases in the molecular 

weight. 

Wort boiling causes abig change in its phenolic make-up. In sweet wort the dimeric 

flavanols (prodelphinidin B3 and procyanidin B3, 18mg/l) and (+)-catechin (6mg/l) 

predominate.Afterboilingwithhopsthelevelofdimericandtrimericflavanolsfalls(to6 

and 5mg/l respectively) while that of (+)-catechin and ( )-epicatechin increases (to 

10mg/l). The ( )-epicatechin is derived not only from hops but also by epimerization of 

(+)-catechin. Derdelinckx and Jerumanis (1987) also showed that proanthocyanidins 

depolymerize during brewing. McMurrough and Delacour (1994) conclude that it is the 

simple flavanols that are the haze precursors in wort. 

Similarly, Whittle et al. (1999) studied the polyphenols in barley and beer. They 

identified over 50 flavanols in barley, including seven pentamers, but in beer only 24 

flavanols were found including (+)-catechin and ( )-epicatechin. No tetramers and 

pentamers survived into beer but it was not established whether they were destroyed in 

mashing or in wort boiling. However, these flavanols do not appear to be necessary for 

protein coagulation. Beers were made with aproanthocyanidin-free malt (Galant) and a 

regularmalt (Triumph)and bittered either with Saaz hops oratannin-freehexane extract. 

All four brews showed similar levels of coagulatable nitrogen after a60 minute boil. 

After alonger boil (90 min.) the coagulatable nitrogen was reduced further in every case 

(Delacour et al., 1988). Techniques to remove protein-polyphenol complexes from beer 

are discussed in Chapter 15. 

9.6 Copper finings and trub formation 

The importance of removing some of the protein from sweet wort and the difficulties that 

can arise later if this is not accomplished have been mentioned and is discussed further in 

Chapter 15. With a view to improving trub formation many brewers add electronegative 

finings to the copper at or near the end of the boil (4±8 g/hl). These copper finings are 

usually Irish moss, the dried red marine algae Chondrus cripus (plus some Gigartina 

stellata) or the purified polysaccharide therefrom, -(kappa)-carrageenan. This 


Addition rate 

Sediment 

Haze 

Fig. 9.11 The effect of copper fining rate on performance (Leather, 1998). 

polysaccharide is made up of achain of galactose and occasionally anhydrogalactose, 

units linked alternately 1±3 and 1±4. Some of the free hydroxyl groups are esterified with 

sulphate groups providing the negative charge. Instead of Irish moss, some brewers add 

silica gel. These copper finings should be distinguished from the proteinaceous isinglass 

finings usedtoclarify beereither incask orconditioningtank(Chapter 15).Nevertheless, 

the use of copper finings reduces the non-microbiological particles (NMP) in the final 

beer (Leather et al., 1996). 

Under optimum boiling conditions, e.g., avigorous rolling boil under atmospheric 

pressure at 102ëC for at least an hour, the hot break is formed as large flocs which can be 

removedinthehopbackorwhirlpool.Whenthewortisboiledwithwholehopsorpellets 

the trub adheres to the hop debris. In contrast, with inefficient wort boiling, or in aplant 

producing excessive shear, the trub may separate as fine flocs, which remain in 

suspension. Similarly, the cold break consists of very fine particles that are slow to form 

and settle and consequently may survive fermentation and be carried over into the beer 

(Chapter 10). Leather (1998) lists eleven factors which affect copper fining performance. 

As the dose rate increases, wort clarity improves and the amount of sediment also 

increases (Fig. 9.11). The optimum fining rate is that which produces the best wort clarity 

with the minimum volume of sediment. In practice this produces a beer containing 

approximately 10 6 non-microbiological particles/ml in each of the three size fractions, 

< 2 m, 2±10 m, and > 10 m. The optimum fining rate is found by experiment. Wort 

clarity is assessed on an arbitary scale from A (brilliant) to F (cloudy) or can be measured 

by the absorbance at 600 nm against a membrane filtered wort. 

Copper finings have no significant effect on hot wort clarity, their main effect being 

the production of bright cold wort. They are added to hot wort since -carrageenan does 

Table 9.7 Typical composition of hot and cold break (Moll and de Blauwe, 1994) 

Hot break Cold break 

Particle size mm 30±80 0.5±1.0 

Typical wet weight g/hl 150±400 5±30 

Moisture % 73±85 70±80 

Proteinaceous matter degree 40±70 45±75 

Bittering substances % 10±20 n/a 

Polyphenols % 5±10 10±30 

Carbohydrates % 4±8 20±30 

Ash % 3±5 2±3 

Fats % 1±2 n/a 


not dissolve below 60ëC. The copper finings should be added early enough in the boil so 

that they all dissolve but late enough so that they are not significantly degraded. For the 

same reason long stands in the whirlpool should be avoided. Awort of pH of c. 5.0 is 

required for efficient fining and worts below pH 4.5 often fail to fine. Adifference of 0.3 

pH units can make adifference between optimum (A) clarity and poor (D) clarity. Malt 

variety and quality also influences copper fining performance. As well as the influence 

malt has on the pH of the wort, the amount of cold break protein (the amount of cold 

break that forms naturally without the addition of copper finings) correlates positively 

with copper fining performance. 

Analyses of hot and cold break are given in Table 9.7. It will be noted that in terms of 

mass the cold break is less than 20% of the hot break and that both are largely made up of 

proteinaceous materials. The cold break is richer in polyphenols and carbohydrates. 

Although hydrogen bonds between polyphenols and polypeptides are largely dissociated 

above 80 ëC, some malt and hop polyphenols are precipitated with the hot break but, 

according to McMurrough and Delacour (1994), `their role is more passive than active'. 

Ninety per cent of the lipids in the copper are deposited with the trub and spent hops 

(Anness and Reed, 1985). In addition some fatty acids may be lost by steam 

volatilization; with 12% evaporation 0.9% of the lipids in wort were lost by this route. 


ANNESS, B. J. and REED, R. J. R. (1985) J. Inst. Brewing, 91, 82. 

CROMPTON, I. E. and HEGARTY, P. K. (1991) Proc. 23rd Congr. Eur. Brew. Conv., Lisbon, p. 625. 

DELCOUR, J. A., VANHAMEL, S., MOERMAN, E. and VANCRAENENBROECK, R. (1988) J. Inst. Brewing, 96, 

371. 

DENK, V., FELGENTRAEGER, H. G. W., FLAD, W., LENEOL, M., MICHEL, R., MIEDANER, H., STIPPLER, K., 

HENSEL, H., NARZISS, L. and O'ROURKE, T. (2000) European Brewery Convention ± Manual of Good 

Practice, Wort Boiling and Clarification, pp. xvi + 176. Fachverlag Hans Carl, NuÈrnberg. 

DERDELINCKX, G. and JERUMANIS, J. (1987) Proc. 21st Congr. Eur. Brew. Conv., Madrid, p. 577. 

EICHNER, K., REUTTER, M. and WITTMANN, R. (1994) in Parliment, T. H., Morello, M. J. and McGorrin, R. 

J. (eds) Thermally Generated Flavors ± Maillard, Microwave and Extrusion Processes. ACS 

Symposium Series No. 543, American Chemical Society, Washington DC, p. 42. 

ENGEL, W. and SCHIEBERLE, P. (2002) J. Agric. Food Chem., 50, 5391. 

EUROPEAN BREWERY CONVENTION See Denk et al. 

FAYLE, S. E. and GERRARD, J. A. (2002) The Maillard Reaction, Royal Society of Chemistry, London, xiv 

+ 120 pp. 

GJERTSEN, P. (1953) J. Inst. Brewing, 59, 296. 

GJERTSEN, P. (1955) Proc. 5th Congr. Eur. Brew. Conv., Baden-Baden, p. 37. 

GUENTHER, K. R. and STUTLER, J. R. (1965) Proc. Annu. Meet. Amer. Soc. Brew. Chem., p. 30. 

HARDING, R. J., NURSTEN, H. E. and WREN, J. J. (1977) J. Sci. Food Agric., 28, 225. 

HARDING, R. J., WREN, J. J. and NURSTEN, H. E. (1978) J. Inst. Brewing, 84, 31. 

HARRIS, G., BARTON-WRIGHT, E. C. and CURTIS, N. (1951) J. Inst. Brewing, 57, 264. 

HARRIS, G., HALL, R. D. and MACWILLIAM, I. C. (1954) J. Inst. Brewing, 60, 464. 

HAYASE, F. (1996) in Ikan, R. (ed.) The Maillard Reaction. John Wiley, Chichester, p. 89. 

HO, C.-T. (1996) in Ikan, R. (ed.) The Maillard Reaction. John Wiley, Chichester, p. 27. 

HODGE, J. E. (1953) J. Agric. Food Chem., 1, 928. 

IKAN, R. (1996) The Maillard Reaction. John Wiley, Chichester, p. 228. 

KATO, H. and TSUCHIDA, H. (1981) Prog. Food Nutr. Sci., 5, 147. 

KLEBER, W., SCHMID, P. and SEYFARTH, I. (1963) Brauwissenschaft, 16, 1. 

LATIMER, R. A., LAKSHMINARAYANAN, K., QUITTENTON, R. C. and DENNIS, G. E. (1966) Proc Conv. Inst. 

Brew. Australian Section, p. 111. 

LEATHER, R. V. (1998) J. Inst. Brewing, 104, 9. 

LEATHER, R. V., WARD, I. L., MORSON, B. T. and DALE, C. J. (1996), Ferment, 9, 31. 

MCFARLANE, W. D. and HELD, H. R. (1953) Proc. 4th Congr. Eur. Brew. Conv., Nice, p. 110. 

MCMURROUGH, I. and DELCOUR, J. A. (1994) Ferment, 7, 175. 


MEILGAARD, M. C. (1975) Tech. Quart. MBAA, 12, 151. 


MIEDANER, H. (1986) J. Inst. Brewing, 92, 330. 

MOLL, M. and DE BLAUWE, J. J. (1994) Beers and Coolers, (trans. by Wainwright, T.), Intercept, Andover. 

495 pp. 

MOTTRAM, D. S. (1994) in Parliment, T. H., Morello, M. J. and McGorrin, R. J. Thermally Generated 

Flavors ± Maillard, Microwave and Extrusion Processes. ACS Symposium Series No. 543. 

American Chemical Society, Washington DC, p. 104. 

NURSTEN, H. E. (1980) Food Chem., 6, 263. 

O'ROURKE, T. (1999) Brewers Guardian, 128 (8), 34; 128 (9), 38. 

O'ROURKE, T. (2002) The Brewer International, 2 (2), 17. 

PARLIMENT, T. H., MORELLO, M. J. and MCGORRIN, R. J. (eds) (1994) Thermally Generated Flavors ± 

Maillard, Microwave and Extrusion Processes. ACS Symposium Series No. 543. American 

Chemical Society, Washington DC, pp. x + 492. 

REED, R. J. R. and JORDAN, G. (1991) Proc. 23rd Congr. Eur. Brew. Conv., Lisbon, p. 673. 

REYNOLDS, T. M. (1963) Advances in Food Research, 12, 1. 

REYNOLDS, T. M. (1965) Advances in Food Research, 14, 167. 

ROBERTS, D. D. and ACREE, T. E. (1994) in Parliment, T. H., Morello, M. J. and McGorrin, R. J. Thermally 

Generated Flavors ± Maillard, Microwave and Extrusion Processes. ACS Symposium Series No. 

543. American Chemical Society, Washington DC, p. 71. 

SANDEGREN, E., ENEBO, L., GUTHENBERG, H. and LJUNGDAHL, L. (1954) Proc. Annu. Meet. Amer. Soc. 

Brew. Chem., p. 63. 

SHIMIZU, C., NAKAMURA, Y., MIYAL, K., ARAKI, K., TAKASHIO, W. and SHINOTSUKA, K. (2001) J. Amer. Soc. 

Brew. Chem., 49, 51. 

THORNTON, J (1989) Brewing and Distilling International, 20 (10), 36. 

TRESSL, R., RENNER, R., KOSSA, T. and KOPPLER, H. (1977) Proc. 16th Congr. Eur. Brew. Conv., 


WALLER, G. R. and FEATHER, M. S. (1983) The Maillard Reaction in Food and Nutrition. ACS Symposium 

Series No. 215. American Chemical Society, Washington DC. 



10 

Wort boiling, clarification, cooling and 

aeration 


Sweet wort is boiled with hops in acopper (kettle, hop-boiler). Sometimes, if the copper 

is not immediately available, the wort is held in an intermediate vessel, or underback, 

before it is boiled. It is held hot (75±80ëC; 167±176ëF) to minimize the risk of microbial 

infection.This`hold'shouldnotbeprolonged assomethermophilicorganisms, including 

some that can reduce nitrate ions to nitrite ions, can continue to grow and the wort will 

darken and its flavour will alter. Avariety of hop preparations may be used instead of 

whole hop cones (Chapter 7), necessitating the use of different types of equipment to 

clarify the wort after the boil. The changes that occur during the boil are complex 

(Chapter 9) but at the end of the boil spent hops (whole cones or fragments) and flocks of 

precipitated material, the hot break or trub, should be suspended in a perfectly clear, or 

`bright' wort, the colour of which will have increased during the boil. The trub and spent 

hops are separated from the wort, which is then cooled and aerated or oxygenated, usually 

while being transferred to a fermenter, where it is pitched (inoculated) with yeast. 

For many years the hop-boil, which usually lasted for 1.5±2 h but sometimes longer, 

was regarded as a simple process, the only variations being the duration of the boiling 

period, the choice of hops, the hopping rate and whether the hops were added at the start 

of the boil, in the middle or near the end (late hopping, when aroma hops are added). 

However, the need to reduce the cost of boiling has resulted in the testing of different 

technologies for saving energy and the difficulties encountered have emphasized the 

complexity of the boiling process and the necessity of balancing the changes that occur. 

There is a range of objectives to be met by boiling (Hough et al., 1982; Miedaner and 

Narziss, 1986; Narziss, 1993). The first is to evaporate water and so concentrate the wort. 

Traditionally, an evaporation rate of 10% or more of the collected wort volume/h was 

usual. The cost of evaporating so much water is high, because the consumption of energy 

is high, and so it is now usual to minimize evaporation. The boil also removes unwanted 

volatile substances. Simply reducing the evaporation rate or shortening boiling times 

gives flavour and other problems, but newer designs of boilers reduce or eliminate this 

difficulty, by favouring the evaporation of the unwanted volatile substances, and boils of 


60±90min., with evaporation rates of as little as 4%/h can be used in appropriate 

equipment. The boil also sterilizes the wort, or at least destroys the `vegetative' forms of 

microbes probably in the first 10±15min. Spores may survive this process. From the boil 

onwards wort is handled under nearly aseptic conditions. 

The changes that occur during boiling include the dispersion of the hop resins and oil, 

the isomerization of some of the -acids, the conversion of dimethyl sulphide precursor 

(SMM; DMSP; 4.157) to DMS (4.158), the formation of new flavour and aroma 

compounds, (largely through Maillard reactions, which also give rise to coloured 

compounds and so adarkening of the wort), the denaturation and inactivation of residual 

enzymes carried forward from the mash and the denaturation and coagulation of the 

proteins that, combined with polyphenols, form the trub (which may weigh 20±70gdry 

wt./hl Chapter 9). The more vigorous and prolonged the boil the more complete the 

removal of coagulable proteins and the more resistant the beer to the formation of nonbiological 

haze, but it will have less good foaming properties. The inactivation of the 

enzymes stabilizes the composition of the wort in one sense, but of course the high 

temperature maintains the on-going chemical reactions. The wort needs to be sufficiently 

agitated to cause the denatured proteins to coagulate and form flocks. This process may 

be assisted by the addition of copper finings, mainly the negatively charged, sulphated 

polysaccharide -carrageenan, Ìrish moss', added at 4±8 g/hl (Chapter 9). Instead some 

breweries use silica gel. 

While vigorous mixing is essential, shear forces must be minimized to prevent the 

flocks being disrupted. Local òverheating' at a heating surface should be avoided as the 

level of coagulable protein is unduly reduced and it can remove proteins that stabilize 

foam. Local overheating, as was common with older, direct-fired coppers, can also cause 

`burn-on' and caramelization of wort sugars and copper adjuncts, which gives rise to 

(usually unwanted) flavours and cleaning difficulties with the copper. The boil is also 

used to `refine' the flavour of the beer by evaporating unwanted volatile flavour and 

aroma compounds. These are more volatile that water and so, under the correct 

conditions, they can be evaporated to the desired extent at relatively low water 

evaporation rates. In the past it was often assumed that, during boiling, some oxidation of 

the wort by entrained air was desirable. The newer view is that this is not so, as oxidation 

darkens the wort, reduces flavour stability and favours haze formation in beers. The 

importance of oxidation varies with the type of beer being made. 

Other substances, besides hop preparations and copper finings, that may be added to 

wort in the copper, include mineral or biologically prepared acids (for pH adjustment), 

salts, tannins, malt extracts, sugars and syrups. The addition of acids, such as lactic acid, 

reduces the pH of the wort as does the addition of calcium salts. Calcium ions, displacing 

hydrogen ions from phosphates and other molecules, giving salts that precipitate during 

boiling, reducing the pH by 0.1±0.2 units. The reduction of wort pH during the boil 

reduces the colour, gives beer with a `cleaner' flavour, and a better `break' formation, but 

hop utilization is reduced unless pre-isomerized preparations are used. Where zincdeficiency 

may occur in the fermenters, small amounts of zinc chloride (0.1±0.2 mg/l) 

may be added to the wort. Gallotannins are sometimes added in calculated amounts to 

precipitate `haze-sensitive' proteins with the trub, so stabilizing the beer. Other 

substances tested include active carbon, nylon powder, PVPP powder, kieselguhr and 

bentonite, but these are probably only rarely or never used in this way now. 

In small-scale brewing malt extract may be dissolved to form wort, which is boiled 

with hops in the same operation. In larger-scale brewing sugars and syrups (copper 

adjuncts derived from cereals or starch and so differing in whether or not they contain 


nitrogenous andothersubstances(Chapter2))maybeaddedtoincreasetheconcentration 

of the wort and/or adjust its fermentability. The material must be added with care to 

ensure that it is dissolved and thoroughly dispersed in the wort. Failure can result in 

syrupy deposits settling onto the heating surfaces, where they are caramelized and burnt. 

Copper adjuncts are often used to increase the concentration of worts for use in highgravity 

brewing. For example if one part of asyrup, SG 1150 (36.4ëPlato) is added to 

nineparts ofawortofSG1040(10ëPlato) the final, mixed wort willhave anSGofabout 

1051 (12.6ëPlato). Hop boiling is normally abatch process. Attempts have been made to 

develop continuous hop-boiling to allow all stages of beer production to run 

continuously, but only with limited success (e.g. Hall and Fricker, 1966; Hough et al., 

1982; see below). 

10.2 The principles of heating wort 

Early wort boilers were made of cast iron, but by the early 20th century most were made 

of copper. While copper has advantages (Table 10.1), more recently boilers have been 

made of stainless steel, but the vessels are still often called `coppers'. Copper is easily 

worked, has ahigh thermal conductivity (approx. 380W/mKat 100ëC; 212ëF) and 

vessels made from it have an attractive appearance (Hancock and Andrews, 1996; 

Royston, 1971; Wilkinson, 1991a). It can catalyse oxidative reactions, it increases wort 

colour during boiling and is said to remove sulphur-containing substances from wort, 

(perhaps by catalysing the oxidation of thiol groups), and so improves beer flavour. It 

also favours nucleate boiling because its surface is relatively wettable. Stainless steel, 

generally an austenitic grade such as 304 or 316, is less expensive but is not so easily 

worked, and so vessels are often of simpler shapes than those made of copper. It is more 

resistant to dilute acids or strongly caustic cleaning agents, such as 2±4% caustic soda, 

and because of its greater strength vessels made from it can be thinner (e.g. 1.6mm) than 

those from copper, so its lower thermal conductivity (approx. 167W/mKat 100ëC; 

212ëF),isnotveryimportant atheatexchangesurfaces.Thelesserwettabilityofstainless 

steel does not favour nucleated boiling or tolerance of high heat flux densities, so the 

maximum temperatures used at heat exchange surfaces are lower than those which may 

be used with copper. 

Earlycopperswere directly heated bywood orcoalfires,andhadcalorificefficiencies 

ofabout 40±50%. Solid fuelheating isnow rare,but some kettles are heated directly with 

gas or oil-fuelled flames, with calorific efficiencies of about 70%. The heated area must 

becovered bywortbefore heatisappliedtopreventcaramelizationandcharring.Thefire 

orflamesmustbeextinguishedbeforethewortiswithdrawn.Directheatingoftenimparts 

characteristic flavours to beers. As breweries became larger, some used many relatively 

Table 10.1 Some properties of copper and stainless steel (Hough et al., 1982) 

Property Copper Stainless steel 

Density (kg/m 3 ) 8930 7930 

Specific heat (J/kg K) 385 510 

Thermal conductivity (W/m K) 385 150 

Yield stress (MN/m 2 ) 75 230 

Heat flux (kW/m 2 ) 80* 60 y 

* For a conventional, jacketed kettle. 

y For flat, stainless steel panels. 


small coppers so that as wort was collected it could be boiled in the shortest possible 

time. This reduced the chances of microbial infection and the operation was flexible. 

When steam heating became established some breweries continued to use several small 

coppers toretaintheseadvantagesandevenoutthedemandforsteam,whileothersbegan 

touseoneortwolargecoppers,whichhave lowercapitalandmaintenancecostsbuthave 

alarge steam demand while the wort is being heated from its collection temperature to 

boiling. One advantage of pre-heating wort on its way to the copper is that this sudden 

steam demand is reduced. Some breweries use hot water, often at 145±170ëC (293± 

338ëF), to heat the coppers. To keep the water liquid (prevent it boiling) it is held under 

substantialpressure (around 17bar).Thehot water systems, which are expensive, need to 

be well insulated and need substantial outgoing and return pipework between the heat 

sink and the furnace(s). Because the water contains asubstantial amount of heat sudden 

demands are more easily met than by steam. For heavily used, flat-sided coppers water 

heated systems have been preferred. Coppers have been boiled experimentally using 

microwave heating (Herrmann, 1999). 

Steam is now the most usual heating agent in breweries. Steam, raised in aboiler 

house, is used dry and saturated at arelatively low pressure, about 4bar, at about 148ëC 

(298ëF). (As the pressure is increased so is the boiling point of water and the temperature 

ofthesteam;AppendixA.11).At aheatexchangesurface thesteamcondenses, givingup 

its latent heat which, in acopper, heats the wort. Asudden large demand for heat can 

cause the condensation of agreat deal of steam and so asudden drop in pressure which 

may cause vessels to collapse. Thus steam-heated systems must be fitted with pressure 

and vacuum release valves as well as condensate traps, pipes to return the condensed 

water to the boiler, strainers and pressure reducing valves. Delivering steam at alower 

pressure reduces the temperature of aheating surface. Steam is used in acopper with 90± 

95% efficiency, but steam raising is less efficient and the overall efficiency is 65±70% 

(Hough et al., 1982). 

In modern installations wort is nearly always heated while flowing upwards in pipes 

mounted in steam jackets. The conduction of heat under ideal conditions, at asteady state, 

isdescribedbyFourier'sLaw,qˆkA T=X,whereqistherateofheattransmission,kis 

the thermal conductivity of the material, Ais the cross-sectional area at right angles to the 

heat flow, Tis the temperature difference between the steam and the liquid to be heated 

and Xis the thickness of the material across which the heat is flowing. However, heat 

transfer from steam to wort is better described by amore empirical formulation because 

the system is quite complex and changes with time as scale and baked-on materials 

(fouling) accumulate on the heat transfer surfaces, impeding heat flow. As wort at boiling 

temperature flows into the base of aheating tube it flows in as asingle, liquid phase and 

theflowshouldbefastenoughtobecometurbulent(Fig.10.1).Asitgainsheatitbeginsto 

boil and bubbles form, creating a second, vapour phase and reducing the density of the 

wort/steam bubble mixture in the tube. At first the bubbles form on and separate from the 

wall of the tube, giving saturated, nucleate boiling. However, if the temperature of the 

walls is too high the liquid may become separated from the wall by a film of steam. This 

film boiling is undesirable as heat transfer is impaired and solids can be deposited and 

baked onto the heating surface, causing fouling. To avoid film boiling the temperature of 

the steam is limited by limiting the pressure to about 3 bar with stainless steel and 5 bar 

with copper. This difference is because the greater surface wettability of the copper 

favours the separation of steam bubbles from the heated surface and so favours nucleated 

boiling (Andrews, 1992; Hancock and Andrews, 1996). 

The transfer of heat from steam to wort is less simple than might be supposed. The 


Rising film 

Saturated 

nucleate 

boiling; 

turbulent, 

two-phase 

flow 

Submerged 

boiling 

Turbulent, 

single-phase 

flow 

Two-phase flow 

Single-phase flow 

Annular flow, 

with entrainment 

Churn flow 

Slug flow 

Bubbly flow 

Tube wall 

Steam or hot 

water 

Fig. 10.1 Diagram of the stages of boiling in wort rising up a steam- or water-heated tube (after 

Wilkinson, 1991a, b). As the wort rises so boiling becomes more vigorous, the flow becomes `twophase' 

as steam vapour bubbles appear and come to occupy an increasing proportion of the volume. 

temperature in the steam, at a particular pressure, is supported by convection. However, 

the surface of the wall is covered by a laminar film of condensed water and this, in turn, 

may be separated from the metal wall of the heat exchanger by a layer of scale deposited 

from the steam. On the wort side of the wall there may be a deposit of baked-on organic 

material and outside this fouling, between it and the moving wort, there is a laminar, 

stationary wort film. The stationary film of condensate, the scale, the metal wall, the 

fouling and the stationary layer of wort must all conduct heat and all provide resistance to 

heat flow from the steam to the wort. Heat distribution in the wort flowing up the tube 

occurs by forced convection. To obtain good mixing and heat transfer to all the liquid the 

wort must have a turbulent flow, which will partly disrupt the stationary layer. With 

turbulent flow heat transfer in a liquid is proportional to the (velocity of flow) 0.8 . With the 

passage of time the scale and the fouling increase and so, in order to maintain wort 

boiling, the temperature of the steam side must be increased by increasing the steam 

pressure. 

The heat flow in a wort heater, Q, can be quantified as Q ˆ U A T, where U is the 

overall heat transfer coefficient of the system, A is the area at right-angles to the heat flow 

and T is the temperature difference between the bulk of the steam and the body of the 


Table 10.2 Typical heat transfer coefficients in wort heaters (kW/m 2 K) (Hough et al., 1982) 

Steam side (low pressure) 5±6 

High pressure, hot water side 2.5 

Wort side (clean) 1.6 

Wort side (dirty) 1.2 

Vessel wall (copper; k/x) 19.3 

Vessel wall (stainless steel; k/x) 7.5 

Overall heat transfer coefficient for a clean, stainless steel vessel 1.0 

As above, but for a dirty kettle 0.5 

wort. UA is called the heat flux, q, and is given in kW/m 2 . The upper limit of the heat 

flux, using stainless steel, is about 95, but in practice this is limited to 50 to reduce the 

risks of fouling. The steam condensate layer, the steam-side scale, the metal wall, the 

wort-side fouling and the wort stationary layer all contribute to U. With increased fouling 

or scaling U decreases. Thus this factor in the equation is empirical, it changes with time 

and must be determined by experiment (Table 10.2). In practical calculations allowance 

must be made for the fact that heat transfer is into a tube and not across a plane surface. 

At some point the heat exchanger must be cleaned to remove the scale and the fouling, 

the heat transfer coefficient, U, is increased and so processing can be resumed using 

lower pressure steam. 

Boiling is accompanied by evaporation and hence an increasing concentration of the 

wort. One ton of steam can evaporate about 15 m 3 of water from wort. The evaporation 

rate is monitored and as it tends to fall in successive brews, due to the fouling and scaling 

in the heat exchanger, the steam pressure/temperature is increased to maintain the 

evaporation rate. By indirectly monitoring the rate of fouling cleaning can be timed to be 

carried out when it is really needed. The evaporation rate may be monitored by 

· automatically determining the level of the boiling wort from the falling pressure at the 

base of the vessel, using a pressure transducer, or by measuring the depth of the liquid 

· measuring steam utilization or condensate production 

· automatic sampling and determination of the wort density (SG) 

· determining the density indirectly, in the base of the vessel at the inlet to an external 

calandria heater, by measuring the velocity of ultrasound generated by a transducer. 

This device is said to be accurate to within +/ 0.1% SG (Forrest et al., 1993). 

With older patterns of coppers cleaning was carried out every 6±12 brews, when the 

heat transfer rate might have fallen by as much as 25%. Reasons for infrequent cleaning 

included (i) the difficulty of finding time in a busy brewing schedule; (ii) the costs of 

cleaning including the costs of detergents, hot water, fitting and dismantling the 

equipment and the disposal of the effluent; (iii) oversizing the heat exchange surface and 

(iv) the reduced need for evaporation as weak worts were recycled to the following brew. 

In newer external wort heaters, (EWH), with rapid, turbulent flow through the heatexchange 

tubes, some scouring of the tube surfaces occurs and the large heating areas 

allow the use of lower wall temperatures and so cleanings can be less frequent, for 

example having 30 brews between cleanings. 

If the pressure on wort is increased the boiling point rises, trub formation and the 

isomerization of -acids are accelerated, as are colour formation, DMS formation and 

other changes. Thus by boiling at elevated temperatures boiling times can be reduced and 

energy (steam consumption) can be saved. On the other hand the increased removal of 

protein results in beers with poorer foaming characteristics. It is said that for each 4 ëC 


(7.8ëF) rise in temperature the boiling time can be halved (Chapter 9). However, as the 

different processes in the wort have different temperature coefficients, the changes in the 

wort will alter at different relative rates at different temperatures and so the nature of the 

wort will alter. Some problems encountered in boiling at elevated temperatures are 

discussed later. Generally boiling wort at, or near, normal atmospheric pressure gives the 

most acceptable results. 

Steam-heated vessels may have jackets or welded semicircular pipes to carry the 

steam (Fig. 15.2). Sudden steam demands can induce adrop in pressure that can cause 

jackets to collapse. Pipe heating is more resistant to this kind of damage. Steam is 

distributed around a brewery in well-insulated mains at a pressure above that needed at 

the various sites. Valves reduce the pressure at each site to an appropriate extent to 

achieve the desired temperature. As heat is given up the steam condenses and the 

condensate is collected and returned to the boiler or sent to drain, without allowing steam 

to escape. Each heating zone must have a pressure gauge and a thermometer and, for 

safety's sake, a vacuum valve and a pressure release valve. In addition there must be 

bleed valves to allow the escape of air when the system is first filled with steam (Kunze, 

1996). 

10.3 Types of coppers 

The increasing sizes of brewing vessels and the need to reduce costs (by increasing 

heating efficiency) and to reduce the emission of vapours (to meet anti-pollution 

legislation) have led to the development of a variety of new types of coppers (hop-boilers, 

kettles). In several cases these, while technically successful in saving energy, produced 

worts with unacceptable or unusual characteristics. In consequence the use of some 

vessels has been discontinued while in other cases coppers found unsuitable by some 

brewers have been retained in use by others, perhaps because of the different 

characteristics of the beers being produced (Andrews, 1992; Andrews and Axcell 

(private communication); Clarke and Kerr, 1991; Hackensellner, 1999; Herrmann, 

1998a,b; Hind, 1940; Kunze, 1996; Miedaner, 1986; Narziss, 1986a, 1992, 1993; 

Ormrod, 1986; Rehberger and Luther, 1994; Schwill-Miedaner and Miedaner, 2002; 

Vermeylen, 1962; Wilkinson, 1985, 1991a, b). 

The oldest coppers were made of iron with cylindrical sides and rounded bases and 

were open to the atmosphere. Often these were replaced by copper vessels of the same 

type. The copper was mounted in a brick housing with a furnace for burning solid fuel at 

the base and a flue that wound round the side of the copper to a chimney. At least one 

such copper is still in use. The copper must be filled with wort before the furnace is fired 

and the fire must be drawn before the copper is emptied to prevent the heated area 

becoming too hot, causing wort to burn on. Sufficient space must remain above the fill 

level to contain the boiling, frothing wort. Such open coppers release steam into the 

surroundings, creating unpleasant working conditions, condensation and drip-back that 

leads to deterioration of the building and cleaning difficulties. Whole hop cones are 

added by hand, in weighed amounts. As kettles became larger and more complex in shape 

they were increasingly made from copper. Usually they were covered with a dome 

provided with a chimney to carry steam outside the building and an inspection and access 

opening, which could be closed with sliding doors. The base was usually hemispherical 

but sometimes it was domed upwards, to encourage better circulation of the wort. There 

might be a mechanical stirrer, driven by a shaft from above. Sometimes stirrers, 


Steam 

Air escape 

Condenser 

Steps for 

cleaning 

copper 

Propeller 

Steam inlet 

Condenser 

pipe 

Motor 

Wort 

gauge 

Propeller 

drive 

Floor 

Floor 

Fig. 10.2 An old pattern of copper having a rounded base and heated by a symmetrical, external 

steam jacket (after Hind, 1940). 

`rummagers', carried a series of loops of chain that swept the bottom of the vessel, 

dispersing deposited materials. It was appreciated that a vigorous `rolling boil' was 

desirable and that there needed to be an evaporation rate of 8±10% or even 15% of the 

original wort volume/h. As the boil often lasted 2±2.5 h there was a substantial reduction 

in the volume of the wort and hence an increase in its concentration. While this allowed 

the use of large volumes of sparge liquor (and hence a good extract recovery from the 

mash) it was time consuming and costly because of the large amounts of fuel needed to 

generate the heat needed to evaporate the water. Direct heating with solid fuels, or oil, or 

gas has become unusual. 

As copper sizes increased it became apparent that simply heating the base, for example 

with steam (Fig. 10.2), was inadequate because the heating surface area to wort volume 

ratio decreased with increasing vessel size and it was undesirable to overheat the heating 

surface. Some brewers used a number of small coppers and filled them and brought them to 

the boil in sequence, saving time and avoiding having to accumulate the wort in a large 

copper or underback, with a consequent risk of microbial infection. A partial solution was 

to have the heating area of the copper divided into zones. As the first, lower zone was 


Floor 

Fascia panel 

Manhole and inspection 

port 

Valve for emptying 

copper 

Floor 

Steam jackets 

asymmetrically placed 

Heating surfaces 

Wort discharge 

Fig. 10.3 Acopper with arounded base and asymmetrically placed steam jackets 


covered with wort the heating could be switched on. Wort continued to enter the copper 

and when the second and third zones were covered steam to these was also switched on, in 

turn. Asecond problem was that the even application of heat to the base of acopper with a 

rounded base gave apoor wort circulation and inadequate mixing. To overcome this some 

coppers were fitted with mechanically driven impellers (Fig. 10.2) while others were fitted 

with asymmetric heating panels (Fig. 10.3). The wort adjacent to the heating surface 

expanded and became less dense, consequently it was driven upwards by the cooler, more 

dense wort that took its place. With asymmetric heating astrong circulation, with good 

mixing and asteady upflow of wort across the heating surface, could be achieved. 

A so-called `high-efficiency' copper became popular in Europe, and provided a 

reference for performance for anumber of years (Fig. 10.4; Table, 10.3; Schwill-Miedaner 

andMiedaner,2002;Narziss, 1992).With thisdesignsteamheatingwasappliedtothebase 

of the copper and to the centrally placed, truncated cone, increasing the heating surface 

area/wort volume ratio and inducing aconvective upward flow of wort in the centre of the 

vessel and acompensatory flow down the outside. The copper was fitted with an agitator, 

which assisted the flow and mixing. Another approach was to supplement or replace the 

surface heating panels with internal heaters. Sometimes these were coils of tubes, but these 

did not encourage good circulation in the wort and were difficult to clean. Other patterns, 

such as `star heaters', which were mounted centrally inthe bases of vessels andencouraged 

an upward, convective flow of acolumn of wort, were better (Fig. 10.5). 

Adifferent approach was to make flat-sided kettles of stainless steel, which were 

rectangular in plan (Fig. 10.6). Experience showed that wort circulation and mixing was 

inadequate, despite the asymmetrical disposition of the heaters, and flow had to be 

assisted by impellers. The headspace was inadequate to accommodate the foam that was 

generated and the foam had to be broken by downwardly directed jets of air. This was 

undesirable both from the point of view of favouring undesirable oxidations in the wort 

and reducing the chances of effective heat recovery by diluting the steam and vapour 

from the boil with air (Section 10.7). 


Table 10.3 The characteristics of some conventional coppers and some other boiling systems 

(data of Schwill-Miedaner and Miedaner, 2002) 

System of boiling Temperature* Boiling time Evaporation y 

(ëC) (ëF) (minutes) (%) 

`High performance' copper 100 212 120±150 12±16 

Internal/external boilers, with 102±103 215.6±217.4 60±80 about 8 

back pressure 

Low-pressure boiler 103±104 217.4±219.2 55±65 6±7 

Dynamic low pressure boiling 103±104 217.4±219.2 45±50 4.5±5 

High temperature wort boiling 130±140 266±284 2.5±3 6±8 

Thin film heating 100 212 35±40 4±4.7 

* Temperature at the exit from the heater. 

+ Percentage of the original volume collected. 

Stack 

Manhole and inspection port 

Valve for discharging wort 

Floor 

Agitator Steam jacket in dimple 

Agitator shaft 


Steam jacket on base of copper 

Fig. 10.4 A `high-efficiency' boiler having steam heating jackets on the base and on the sides of 

the central, truncated cone (Hough et al., 1982). 

Fig. 10.5 A `star' steam-heater, designed to have a large surface area while remaining compact, 

and be suitable for mounting in the base of a copper (Hough et al., 1982). 


(a) Entry for 

loading 

hops 

Heating 

surface 

Vapour 

stack 

Drive motor 

Inspection 

window 

Impeller Support 

Heating 

surfaces 

Heating 

surfaces 

Fig. 10.6 (a) Aflat-sided, stainless steel copper (rectangular in plan) with an asymmetric heating 

surface of semicircular tubes, and an impeller to assist mixing the wort. This design maintained the 

same size of end wall, but provided vessels of differing capacity by altering the length of the vessel 

(after Hough et al., 1982). (b) and (c) Cross-sections of alternative patterns of flat-sided kettles 

(after Rehberger and Luther, 1994). 

The increasing costs of fuels and the desire to reduce manpower and increase 

efficiency have driven the development of new wort boiling systems. Objectives, which 

are inter-connected, include the reduction in the use of primary energy, (involving the 

shortening of boiling times and reducing evaporation rates), the recovery and re-use of 

heat from the copper vapours (and the wort coolers), while avoiding the over-production 

of warm water, achieving the correct degrees of protein coagulation and removal of 

volatile substances, the avoidance of excessive colour generation or off-flavours and the 

maintenance or enhancement of the quality of the wort and the beer made from it. The 

equipment should be easy to clean and maintain and should not require cleaning too 

frequently. At least in large breweries cleaning should be fully automatic (CIP; cleaning 

in place) and not require the direct use of manpower. 

Newer coppers usually employ internal or external heaters in which the wort passes 

upwards through tubes surrounded by asteam-heated chamber. Figure 10.7 shows an 

internal heater in which the wort flows upwards through the heating unit into a 

constricting tube (sometimes called aVenturi tube), emerges above the level of the wort 

and strikes adeflector plate which directs it back as aspray onto the surface of the wort. 

In some instances two plates are used to direct the spray to the edge and to the midpoint 

of the radius of the copper. The sprays ensure agood circulation of the wort, they serve 

to `beat-back' foam and, by breaking the wort-stream into small droplets, they create a 

large surface area from which volatiles can evaporate into the vapour stream, which 

leaves the copper via the chimney. Being immersed in the wort internal, heaters are 

efficient but their size is restricted by the geometry of the vessel. This means that the 

heating surfaces must be heated to arelatively high temperature to obtain the necessary 

heat flux to obtain avigorous boil and this, in turn, leads to faster fouling and more 

frequent cleaning (sometimes as often as every six brews). By restricting the flow of 

wort in the Venturi tube asmall back-pressure can be achieved, raising the boiling point 


(b) 

(c)

Tubes carrying 

wort through the 

steam chamber 

Inflow of wort 

Steam 

inlet 

Deflector cap 

Upflow 

of wort 

Wort 

discharge 

Spray head 

Base of 

copper 

Steam and 

condensate outlet 

Wort outlet 

Fig. 10.7 An internal heater and `fountain' for coppers in which the wort moves up the steamheated 

tubes by convection, its flow is constricted, then it emerges from beneath the wort level and 

strikes the deflector plate which directs it down onto the wort surface (Various sources). 

of the wort to 102±103ëC (215.6±217.4ëF). In part the size limitation for internal heaters 

may be offset by increasing the depth of the copper under the heater which, in 

consequence, can be relatively larger (Fig. 10.8). 

If the wort has not been pre-heated it arrives at the copper at mashing and sparging 

temperatures, about 75ëC (167ëF), and must be heated to boiling in the copper. With 

simple internal tubular heaters it is sometimes found that violent pulsations occur until 

the pre-heating period is complete (Stippler et al., 1997). This phenomenon resembles 

`boiling with bumping', familiar to chemists. The static, cool wort in the tubes is heated 

to boiling and then wort and vapour escape violently to be replaced with more cool wort. 

The process is repeated until the bulk of the wort is nearly boiling, when asteady stream, 

driven by convection, flows upwards through the heater. Mechanically driven impellers 

have been installed below some heaters to drive the wort upward through the heater 

during the heating phase and so avoid the pulsations. An alternative approach is to pump 

wortfromtwoseparatesitesatthebaseofthecopper,usingthecastingpump,anddeliver 

it into the base of the internal heater, creating aforced upward flow over the heating 

surfaces (Hackensellner, 1999). When the wort is boiling steadily good mixing occurs. 

All efficient modern coppers are `closed', and so the steam is not diluted with air once 

boiling is established and has driven the air from the system, and so is available for heat 


ecovery. Internal heaters are less easily cleaned than external heaters. Boils of 60± 

80min., with total evaporations of around 8% are often suitable. External tube-and-shell, 

wort heaters (EWH), previously called `calandria' heaters are often preferred (Andrews, 

1992; Andrews and Axcell (private communication); Hancock and Andrews, 1996). 

Their size and geometry is not restricted by the dimensions of the kettle. Consequently 

the heat-exchange surface area can be relatively large, allowing lower wall temperatures, 

so less fouling is likely. External heaters are easily cleaned and may be retro-fitted to 

existingcoppers. Inanearly version, wort wasdrawn fromthe baseofthecopperthrough 

an axial flow pump and then upwards through atube-and-shell `calandria' steam heater 

and back into the vessel into the base of aVenturi `fountain' and was directed back onto 

the surface of the wort by aspreader (Fig. 10.9a). The axial flow pump remained running 

until the wort was boiling, then it was shut off and circulation in the system was 

maintained by the `thermosyphon' effect, driven by the reduced density of the wort 

boiling in the external heater being displaced upwards by the cooler, denser wort in the 

body of the vessel. 

The original heaters were comparatively short (1±1.2m; 3.28±3.94ft.) and had wide 

wort-carrying tubes (c. 75mm; 2.95in. diameter) to accommodate the passage of whole 

hops. If milled or milled and pelleted hops are used longer (2.5±4.5m; 8.2±14.76ft.) and 

narrower tubes (25±65mm; 1.00±2.59in.) may be used having greater heat exchange 

efficiency. Avessel to boil 1000hl (611brl) of wort, and equipped with an external 

heater,might be5.8m(19.03ft.)indiameter andthe straight sidesbe4m(13.12ft.)high. 

External wort heaters have been developed in two directions. In one the wort is pumped 

through the heater continuously during heating up and during the boil. In some cases 

these heaters may be mounted horizontally and/or the tubes may be bent into adistorted 

`S' shape, so that they traverse the heating steam chamber three times. At the exit of the 

heater arestriction valve may provide back-pressure raising the temperature of the wort 

Vent 

Inspection port 

Deflector cap 

Heater 

Sump 

Steam inlet, 

condensate 

outlet 

Fig. 10.8 Adiagram of acopper having an internal heater, mounted over and extending partly 

into, a `cup' or `sump'. The extra depth permits the heater to be larger than would otherwise be the 

case, and so it has a large heat transfer surface (after Michel, 1991). 


Fig. 10.9 (a) A copper heated by a comparatively small external shell-and-tube calandria heater 

(Hough et al., 1982). This is an older design in which the pump initiated wort circulation but which 

relied on the circulation being driven by the thermosyphon effect once boiling was established. (b) In 

newer designs the wort flow is directed around the pump in a bypass loop when boiling is occurring. 

Wort from the EWH may be discharged above or at the wort surface. (Courtesy of Briggs of Burton, 

plc.). Alternatively, the wort may be discharged through a `fountain', as in the older pattern. 

in the heater to 103±104 ëC (217.4±219.2 ëF), allowing a shortened boil of 60±70 min. 

Although higher temperatures can be attained (e.g. 110 ëC; 230 ëF) they are generally not 

used because wort quality can be impaired (Narziss, 1993; Narziss et al., 1992). The 

release of pressure at the valve favours the rapid production of small vapour bubbles, 


whichareefficientatcarryingunwantedvolatilesawayintothevapourstream.Usingthis 

system the heater must be pressure resistant, the pump must run continually, with 

implications for maintenance and running costs, and the shear provided in the pump and 

the restriction valve tends to break up the flocs of trub (hot break), complicating the 

subsequent clarification of the hot wort. 

Other external heaters (EWH) rely on the thermosyphon effect to drive the circulation 

of the wort once the system has reached boiling (Andrews, 1992; Andrews and Axcell 

(private communication); Hancock and Andrews, 1996; Wilkinson, 1991a,b; Fig. 10.9b). 

Whilethe wortis beingheatedtoboilingitisrecirculatedthroughtheEWH byapumpin 

abypass loop. Once boiling is established, and can be driven by the thermosyphon effect, 

the pump is bypassed, the wort flow is directed to asimple loop of obstruction-free 

pipework from the base of the copper into the base of the, vertically mounted, EWH. The 

absenceofpumpingsavespower,andwearonthepumpandavoidstheliquidshearinthe 

pump. The large sizes of EWHs, that can be used, allow low heating temperatures to be 

applied and hence much slower fouling. For example, with 7% evaporation/h, aheat 

transfer of 4.63kW/hl and the same heat transfer coefficient then atypical internal heater 

might have aworking area of 0.08m 2 /hl while an external heater could have 0.20m 2 /hl. 

The steam/wort temperature differences would be 37ëC and 15ëC (66.5ëF and 27ëF) 

respectively (Andrews and Axcell, private communication). Wort passes through these 

heaters at the rate of 6±10 or even 12 vessel volumes/h. The scouring effect of the wort 

and the low wall temperatures needed to maintain the correct heat flux minimize fouling 

and so the necessary cleaning frequency may be less than once in 30 brews. 

The temperature of the wort in the heater (not in the body of the kettle) may reach 

105ëC(221ëF).Thewortemergesfromthereturnpipeandisinjectedtangentially,bythe 

force of the flow, either into the wort or, preferably, just above its surface. In each case 

the wort is driven to rotate in the kettle, giving good mixing, and in the second case the 

vigorous breakout of the vapour bubbles, initiated by nuclear boiling, efficiently 

evaporates a proportion of the volatiles. Evidently this arrangement is suitable for 

combined kettle/whirlpool vessels (Section 10.8). The high efficiency of volatiles 

removal, (only 4% evaporation may be needed), reduces the need for lengthy boiling and 

soshorter,lessenergy-costlyboiling,ispossible.This,inturn,meanslessprimaryenergy 

is needed, less fuel is used and so carbon dioxide emissions are reduced. In afew small 

breweries plate and frame heat exchangers are used as EWHs. 

Arecent wort-boiling `Merlin' system uses athin film boiling and evaporation unit 

(Fig.10.10;Schwill-MiedanerandMiedaner,2002;Stippler,2000).Inthissystemwortis 

pumped from a whirlpool vessel (Section 10.8) and is discharged onto the apex of a metal 

cone, the upper third and lower two-thirds of which are independently heated by steam. 

The cone is enclosed in a vessel with a vapour-escape chimney. The wort flows 

downwards and boils in a thin, turbulent film over the heating surface and is then 

collected in a circumferential gulley from which it flows back to the whirlpool tank. This 

arrangement ensures a good evaporation of volatiles. The returning wort enters the tank at 

two locations, at the centre and tangentially at the periphery, where its entry drives the 

rotation of the vessel's contents. The two points of entry favour mixing. The heating 

phase, during which hops are added, may be for 35±40 min., with both zones of the 

heating cone at 130 ëC (266 ëF). The boil, which lasts about 40±60 min., takes place with 

only the lower zone being heated, to c. 120 ëC (248 ëF). During the heating-up period the 

wort is pumped at a rate of about 4±6 vessel volumes/h, while during the boil the rate is 4 

volumes/h. At the end of the boil the heater and pump are turned off and then, after the 

whirlpool rest (which may last as little as 10 min. because initial trub separation occurred 


Steam 

Valve 

Valve 

during the boil; Section 10.8), the clarified wort is pumped out over the heating cone to 

the cooler. At this stage only the lower part of the cone is heated to 116±120ëC (c. 241± 

248ëF). This `volatiles stripping' stage reduces the level of DMS very substantially. The 

total evaporation is about 4.5% of the initial wort volume, of which 1.5±2.5% is removed 

during the boil, the remainder during stripping. Thus this boiling system is energy 

efficient. 

10.4 The addition of hops 

Wort flow to top of cone 

Central 

wort entry 

Hop-dosing 

device 

Pump 

Pump 

Collection and 

whirlpool tank 

Thin film evaporator/heater 

Spreading cone 

Upper heating zone 

Lower heating zone 

Wort collection gully 

Pump 

Tangential 

wort entry 

Water 

Heated 

water 

Cooled 

wort 

Fig. 10.10 Adiagram of a`Merlin' wort boiling system in which the wort is collected in a 

whirlpool vessel and is heated, boiled and volatiles are stripped by pumping it in a thin film over a 

steam-heated cone in a separate vessel (after various sources, including Anton Steinecker 

literature). 

In older and/or smaller breweries it is normal to add weighed amounts of hops (whole 

hops, milled hops, pelleted milled hops or hop extracts) to the coppers by hand (Chapter 

9). Traditionally, English beers were hopped at rates of 0.5±3.5lb./imp. brl (0.139± 

0.970kg/hl; Hind, 1940; Chapter 9). Manual additions are hazardous as the wort may 

suddenly boil up and over, and late additions, as with aroma hops, mean that air is 

admitted to the copper, which interferes with heat recovery from the vapour. There seem 

to be no good ways of automating additions of whole hops, but means of adding hop 

powders, pelleted hop powders and hop extracts have been automated (Anon., 1994; 

Benitez et al., 1997; Boyes, 1993; Kollnberger, 1986, 1987; Kunze, 1996; Langenhan, 

1995). Additions may be regulated by weight or by -acid content and several additions 

may be made at different stages of the boil. Hops may arrive at the brewery in bales, 

pockets, foil-lined bags or boxes. Extracts come in cans or in 80- or 200-litre mild steel 

disposable drums or in 1000-litre, returnable stainless steel drums. 

The hops or hop preparations must be stored cool and dry. Often the store is not 

adjacent to the copper(s). Thus the hops must be unpacked (containing any dust that is 

generated), be transferred to the brewhouse and weighed amounts be delivered to each 


ew at the correct stage. Unpacking can be laborious and, if automated, only one or two 

types of packing can be handled conveniently. In one system boxes of pellets are 

conveyed forwardon aroller conveyor bygravity and, when the depth ofpelletsin afeed 

hopper has fallen to a predetermined extent, a box is opened, by two saws that 

automatically remove the ends. The empty box is discarded, its place being taken by the 

next full box, while the pellets fall into the hopper. Discharge from the hopper may be 

through a vibrating feeder into a weigher that determines the dose. Pellets can be 

transferred by mechanical or pneumatic means, provided these are not too damaging, or 

they can be broken up and be pumped after being slurried with wort or water. Pellets can 

become `sticky' and conveyers become contaminated and need regular cleaning. Pellets 

may beweighed into aconveyorand delivered directly into the copper, precautions being 

taken to prevent steam wetting the conveying system. 

With pressurized coppers the dose passes through the first valve, which is then closed, 

the pressure is equalized, then the second valve is opened and the hops are discharged 

into the vessel. In adifferent system each batch of hops or hop extract is weighed into a 

smallpressurevesseland,attheappropriatetimethehop`dose'isflushedintothecopper 

with hot water or hot wort. If different types of hops are to be added at different stages of 

the boil each batch must be in adifferent vessel. In yet another system hop pellets or 

extracts are slurried in hot water or wort and aliquots are pumped into coppers as 

required. Hop extracts may be added by suspending punctured tins in baskets immersed 

in the boiling wort. Alternatively, drums of extract may be pre-heated to not more than 

45±50ëC (113±122ëF) for 24h to reduce the viscosity, then the extract can be added to 

thewortflowingintothecopperviaameteringpump,ormaybeaddedwithaflushofhot 

wort or water. Hop oils may be added in at the end of the boil and it has been proposed 

that hop oils in the vapour condensate could be collected and added back in the same 

way. 

10.5 Pressurized hop-boiling systems 

The boiling point of water or wort depends on the pressure, (Appendix A.11), and so 

varies with changes in the atmospheric pressure. In breweries situated at high altitudes it 

is necessary to be able to seal and pressurize the coppers to elevate the boiling 

temperature to at least about 100ëC (212ëF). It has long been realized that by boiling 

wort at elevated temperatures hop -acid utilization is increased and potentially wort 

boiling can be shortened with savings in time and energy (Chapter 9). The acceptability 

of this approach is disputed. Recently, three types of elevated temperature/pressure 

boiling systems have been advocated, mainly in continental Europe. These have been 

called `low-pressure boiling', `dynamic low-pressure boiling' and `continuous, highpressure 

boiling'. 

10.5.1 Low-pressure boiling 

Low-pressure boiling involves increasing the òverpressure' in the copper (the pressure 

above atmospheric) to a relatively small extent, giving absolute pressures of up to 2 bar. 

This can be carried out in sealed and strengthened coppers with internal or external 

heaters. Temperatures of up to 110 ëC (230 ëF) have been used, but lower temperatures 

seem to be preferred, at pressures of 1.5±1.9 bar (Herrmann, 1985; Narziss, 1986a, b). A 

problem with low-pressure boiling is an inadequate evaporation of volatiles from the 


wort. In aparticular schedule wort is boiled for ten minutes at atmospheric pressure, 

when 1.3% evaporation occurs. The copper is sealed and the pressure rises, over ten 

minutes, until 106ëC (222.8ëF) is reached and 0.5% evaporation has occurred. This 

temperature is maintained, during boiling, for 18 minutes, when a further 1.3% 

evaporation takes place. Over 15 minutes the pressure is reduced until the temperature 

has declined to 100ëC (212ëF), allowing evaporation of 1.7%. Finally, the wort is boiled 

at atmospheric pressure for ten minutes, achieving afurther 1.3% evaporation, giving 

6.1% (6±7%) evaporation in total. This reduces boiling time by 20±30%, relative to 

unpressurized controls, and is convenient for energy recovery from the vapour (Section 

10.7). 

10.5.2 Dynamic, low-pressure boiling 

Dynamic, low-pressure boiling is designed to achieve a more rapid evaporation of 

volatiles and so asmaller, less costly total evaporation rate. Where the evaporation in a 

`traditional' boil is 8±10%/h, and in alow-pressure boil is 6±7% in total, that achieved in 

adynamic, low-pressure boil may be 4±5% in total (Kantelberg et al., 2000). In this 

system the pressure in the copper is allowed to increase, e.g., to 1.17 bar, and then is 

allowed tofall,e.g.,to1.05bar,with 6±7pressurechanges/h,givingtemperaturechanges 

between 101±102 and 104±105ëC (213.8±215.6 and 219.2±221ëF). Each time the 

pressure is released there is amassive release of small bubbles throughout the wort, 

which carry volatiles to the surface. The reduction in unwanted volatiles is good, with a 

total evaporation of 4.5±4.8%. 

10.5.3 Continuous, high-pressure boiling 

Continuous,high-temperaturewortboilinghasbeentriedinvarioussystems,withlimited 

success (Chantrell, 1983, 1984; Grasman and van Eerde, 1986; Rehberger and Luther, 

1994). The most successful attempts involved heating at 130±140ëC (266±284ëF) for 

150±180s. Savings in steam utilization of 60±65% were claimed (Fig. 10.11). Wort from 

a collection vessel at around 72 ëC (161.6 ëF), to which hop extracts are added, is 

successively heated in three heat exchangers to 90 ëC (194 ëF), 106 ëC (222.8 ëF) and 

140 ëC (284 ëF) using vapour from the expansion tanks in the first two heaters and live 

steam in the third as the heating agents. The wort is maintained for three minutes at 

140 ëC (284 ëF) in a holding tube, then it is cooled in two stages in expansion vessels 

which successively reduce the temperature to 120 ëC (248 ëF) and 100 ëC (212 ëF). The 

vapours, which flash off during boiling only in the flash-tanks, remove volatiles and 

transfer their heat to the incoming wort stream through the first two heat exchangers. The 

wort is then sent to the whirlpool/clarifier. The system is energy-efficient, but wort is 

darkened and good beer qualities have not always been achieved. The fouling of the 

equipment at the high temperatures used necessitates frequent cleaning using strong 

caustic solutions and hydrogen peroxide. 

10.6 The control of volatile substances in wort 

It is apparent, from the points made above, that the removal of unwanted flavour and 

aroma volatiles is essential and that this may be achieved, in the appropriate equipment, 

without the traditional high evaporation rates. In many cases the removal of volatiles 


Heat 

exchangers 

From wort 

collection tanks 

Condensate 

Vapour, 100°C 


Live steam 

6 bar 

90°C 106°C 

HE 1 HE 2 HE 3 

Condensate 

140°C 

Holding 

tubes 

120°C 

Expansion, 

evaporation 

vessels 

achieved during the copper-boil is not sufficient because, as with DMS, more are formed 

during the whirlpool rest or during the hold in the settling tank, periods during which the 

hot wort is clarified. Boiling wort most readily gives up volatiles when it is finely 

dispersedasinacopperwithaVenturitubeandspreaderorwhenitisdischargedfroman 

external heater in atwo-phase mixture of liquid and vapour bubbles above the surface, or 

if the pressure on the boil is reduced causing the creation and escape of many small 

bubbles. Evaporationfrom coppershas alsobeen improved,andthe boilingtimereduced, 

at least experimentally, by sparging the boiling wort with an inert gas, nitrogen, 

introduced at the base of the copper (Mitani et al., 1999). Maule and Clark (1985) 

demonstrated that boiling might be avoided by simmering the wort at 93ëC (199.4ëF), 

removing sufficient protein by treatment with silica hydrogel and removing volatiles by 

spraying the wort into an evaporation chamber held under apartial vacuum. Volatiles 

removal also occurs in the flash evaporation chambers used in continuous, hightemperature 

boiling (Fig. 10.11), and is of value in treating worts being transferred from 

the clarification vessel to the cooler (Lustig et al., 1997). 

Other approaches have been used to remove unwanted volatiles, generated in the hot 

wortduringclarificationrests.Thiscanbeachievedbyheatingathinlayerofwortduring 

the transfer from the whirlpool tank to the cooler (Stippler, 2000; Fig. 10.10). An 

alternative approach is to `steam-strip' the wort (Bonachelli et al., 2001; Braekeleirs and 

Bauduin, 2001; Seldeslachts et al., 1997; Fig. 10.12). In the stripper the wort is preheated 

to boiling and is then sprayed into the top of acolumn packed with rings and 

saddles giving ahigh surface area of 50±500m 2 /m 3 .As the wort percolates downwards, 

asathinfilmoverthecolumnpacking,itmeetsaslowcounter-flowoflivesteammoving 

upwards. This adjustable process carries away volatiles to acondenser and 0.5±2% 

evaporation occurs. Heat may be recovered, as hot water, from the condenser. The 

stripped wort moves directly to the cooler. 

By cooling the wort, e.g., to 89ëC (192.2ëF), during its transfer from the copper to the 

whirlpool tank, the reactions proceeding in the wort in the whirlpool are usefully slowed 

EV 

1 

EV 

2 

100°C 

Wort to whirlpool 

separator 

Fig.10.11 Aschemeofahigh-temperaturewortboilingsystem(afterChantrell,1983;Clarkeand 

Kerr, 1991). The vapours from the expansion vessels (EV 1 and 2) heat the wort in the first two heat 

exchangers (HE 1 and 2). Heat is provided to the third heat exchanger, (HE 3), raising the 

temperature of the wort to 140 ëC (284 ëF), by live steam. This temperature is held for three min. 

during the passage of the wort through the holding tube. The reduction in temperature and 

evaporation occurs in two controlled steps, when the wort is boiled in the expansion vessels and 

unwanted volatiles are removed. 


Steam 

Wort preheater 

Column packing 

Condensate 

Wort 

Wort receiver 

Wort distribution 

system 

and fewer volatiles are formed. Cooling may be achieved using cold water and a plate 

heat exchanger (producing warm water), or by expansion in a vacuum chamber (Lustig et 

al., 1997; Krottenthaler and Back, 2001; Krottenthaler et al., 2001). 

10.7 Energy conservation and the hop-boil 

Heated water 

Vapour 

condenser 

Supporting grid 

Vapour condensate 

Cooling water 

Steam to injector 

Stripped wort 

Fig. 10.12 A steam stripper for removing volatiles from wort (after Bonachelli et al., 2001; 

Seldeslachts et al., 1997). 

Energy and warm water production and consumption link all parts of the brewing process 

(Manger, 1998; Schu, 1995; Unterstein, 1992). As far as possible `waste' heat, that is heat 

other than that derived directly from primary heating, must be conserved and utilized. 

This implies `good housekeeping', well-insulated vessels, no steam or other leaks, 

automatic controls where possible, efficient designs of plant and brewing operations, and 

well-trained staff. Thus, besides steam and other `direct' heating, heat as vapour or hot 

water is produced in mash- (decoction) and adjunct-cooking, in wort boiling and in wort 

cooling. Hot water can be or is used in pasteurizers, in pre-warming lauter tuns or other 

vessels, in mashing in, in sparging, in preheating wort before it reaches the copper (to 

avoid a sudden high demand for steam), in copper boiling, in CIP and other plant cleaning 

and in cleaning cans, kegs, casks and bottles as well as space heating, in preheating boiler 

feed and even heating anaerobic waste digestion plant. This section is particularly 

concerned with saving heat during hop-boiling, but it is unrealistic to consider this in 

isolation from the rest of the brewing process (Boer, de 1991; Clarke and Kerr, 1991; 

Fohr and Meyer-Pittroff, 1998; Herrmann, 1998a, b; Kunze, 1996; Lenz et al., 1991; 

Miedaner, 1986; Narziss, 1993; Reed, 1992). Heat recovery from boiler vapours requires 

that the system is purged and air-free. The practice of letting air into the copper, to reduce 

fobbing and to increase the draught up the stack to encourage evaporation, as well as 

encouraging wort oxidation, blocks heat recovery (Hough et al., 1982). 

Heat recovery from the vapours generated during hop-boiling has attracted much 

attention. In one, relatively simple, system vapour from pairs of boiling coppers was fed 

into a main and was condensed initially with jet condensers but subsequently, and with 

much greater efficiency, by cold or warm water sprays (Morris, 1987). A rise in 

temperature in the main, caused by the arrival of vapour from the boiling coppers, 


automatically triggered the first of a pair of sprays to come on. If the temperature of the 

main downstream from the first spray rose, indicating incomplete vapour condensation, a 

second, back-up spray came into operation. A key factor was the use of decarbonated 

water in the sprays, to avoid the deposition of scale in the pipes. Heat recovery in the 

resultant hot water was > 95%. Cold water at about 7 ëC (45 ëF) was used initially in the 

sprays and water with vapour condensate at about 90 ëC (195 ëF) was produced. Later 

warm water was used in the sprays. The small amounts of organic materials present in the 

hot water had no unwanted effects. Only a small amount of cooled vapour was discharged 

outside the building, and so the escape of organic materials and odours was negligible. 

Vapour condensers, which are widely used, produce hot water by heating cooling 

water, while the vapours from boiling wort (or mash) are condensed. The hot condensate 

may be cooled further in a heat exchanger before being used as a first cleaning rinse or 

being directed to waste. The warmed water produced when cooling the condensate may 

be used in various ways, including being the cooling feed to the vapour condenser, where 

it is heated to a substantially higher temperature. Copper condensers, or economizers, are 

not new; one was illustrated in 1875 (Narziss, 1986b). They may be used with internally 

or externally heated coppers, boiling at either ambient or elevated pressures. They are 

invariably used in thermal vapour recompression systems (see below). A system in which 

hot water from a condenser is used, in conjunction with a hot water, ènergy storage tank' 

to heat a mash-mixing vessel is shown in Fig. 10.13. The temperature of the heated water 

is regulated by how the system is operated but is generally between 80 and 98 ëC (174.2 

and 208.4 ëF). In principle, it is desirable to use `waste' heat as it becomes available. 

However, as brewing schedules often cannot be arranged to achieve this it is necessary to 

have well-insulated hot water storage tanks to act as ènergy stores'. These are arranged 

in various ways, for example, single, tall and relatively narrow tanks with thermal 

gradients (Fig. 10.13) or linked pairs of tanks may be used (Vollhals, 1994). 

Wort collection 

vessel 

Mash mixing 

vessel 

72°C 

From 

lauter 

tun 

Wort 

preheater 

95°C 

Heated 

water 

Cool 

water 

99°C ~80°C 78°C 99°C 

To whirlpool 

Kettle vapour 

condenser 

78°C 

Condensate 

cooler 

Waste 

Deflector 

Internal 

boiler 

99°C 

99°C 

78°C 

Energy 

storage 

tank 

Fig. 10.13 An arrangement where vapour from a kettle is condensed and heat recovered in hot 

water is stored in a temperature gradient tank and is partly used to heat the wall jacket of a mash 

mixing vessel and the wort from the lauter tun, before it enters the copper (after Herrmann, 1998b). 


Valve 

Vapour 

Copper 


Wort, 106°C 

Wort 

spreader 

Wort return, 100°C Pump 

Mechanical 

vapour compressor 

Compressed 

vapour, 120°C 

Live steam supply 

Calandria reactor 

Condensate 

Heated water, 85°C 

Pump 

Cooling water, 15°C 

Condensate 

cooler Cooled 

condensate, 

30°C, to waste 

Fig. 10.14 Outline of amechanical vapour recompression (MVR) system (after Fortuin, 1995). 

During boiling the wort, passing through the external wort heater, is heated by the compressed 

vapours from the copper, supplemented with live steam as needed. Condensate from the wort heater 

is cooled, water being heated in the process. 

Wort cannot be boiled with water vapour at 100ëC (212ëF). However, if the vapour is 

compresseditstemperaturerisesand,provided thattheheatexchange areaisadequate,as 

can most easily be arranged with external copper heaters, boiling can be supported by the 

heat in the compressed vapour. In amechanical vapour recompression system (MVR), 

once all the air has been displaced from aclosed copper, acompressor (screw-, turbo- or 

rotary piston-compressor; often aRootes compressor) is switched on. The vapour is 

compressed to have an overpressure of 0.2±0.7 bar, and atemperature of 108±112ëC 

(226.4±233.6ëF). This vapour is returned to the copper heater and, by condensing and 

giving up its latent heat in the normal way, maintains boiling, with supplementary 

additions of live steam if necessary (Fig. 10.14). This system uses electrical energy to 

drive the compressor. Compressors can be expensive and noisy and need regular 

maintenance. On the other hand steam savings of 60% are claimed and, in contrast to the 

TVR system (see below), the only hot water produced is from the heater condensate heat 

exchanger/cooler. 

The thermal vapour recompression systems (TVR) have the same objectives as the 

MVR systems, but vapour recompression is achieved using asteam jet compressor, an 

important consequence of which is that substantial amounts of hot water are generated. 

The economics of this system largely depend on whether this hot water is needed. In the 

compressor ajet of live steam, regulated by aneedle valve, enters achamber where it 

draws in vapour from the copper stack and mixes with it and carries it forward to an 

expansion chamber where it slows down and the pressure rises (Fig. 10.15). Values vary, 

but the live steam must have an adequate pressure, at least 8bar up to 24 bar. The 

compressed vapour has an overpressure of 0.3±0.4 bar and atemperature of 106±110ëC 

(222.8±230ëF). Agood proportion, preferably the major part, of the vapour (55±74%) is 

compressedandused tosupporttheboilinthecopperwhilethe remainderisdiverted toa 

condenser which heats cooling water to 80±95ëC (176±203ëF; Fig. 10.16). 

The vapour condensate joins the condensate from the wort heater and passes through a 

cooler, which generates warm water, and cool condensate, which may be used or be 

diverted to waste (Dymond and Djurslev, 1994; Fohr and Meyer-Pittroff, 1998). The 


Valve 

actuator 

Vapours 

Live steam (providing 

motive power) 

Steam jet with 

regulator 

Steam jet 

Mixed vapours and steam 

pressure chamber 

Fig. 10.15 The principle of a steam-jet vapour compressor. A regulated jet of live steam sucks 

vapours into the nozzle. As the cross-sectional area of the tubing widens the stream of mixed steam 

and vapour slows down and the pressure rises, e.g., to 1.2±1.7 bar (104.8±115.2 ëC; approx. 

221±239 ëF). 

Heated 

water, 

80–95°C 

Vapour 

condenser 

Vapour 

condensate 

(hot) 

Cool 

water 

Cool water 

Heated water 

25–46% 

vapour 

Wort 

return 

Vapour 

Copper, 100°C, 

0.02 bar 

Throttle 

55–74% 

vapour 

Wort 

circulation 

pump 

Steam jet 

vapour 

compressor 

101–106°C 

External 

105–115°C 

calandria 

heater 

Condensate 

Mixed condensate (hot) 

Vapour 

condensate 

cooler 

Cooled, mixed 

condensate, 

20–30°C 

Effluent 

Motive steam, 

7–24 bar 

Mixed vapours 

and steam, 

1.2–1.7 bar 

Bypass 

Valve 

Fig. 10.16 A thermal vapour recompression (TVR) system (after Fohr and Meyer-Pittroff, 1998). 

The wort is boiled, under pressure, by the compressed mixture of vapour and steam from the steam 

jet vapour compressor. Vapour from the copper that is not compressed is condensed and used to 

heat water in the process. Mixed condensates from the vapour condenser and the heater are cooled 

and sent to waste. Hot water is also obtained from this operation. 


warm water from the condensate cooler may be used as the cooling water in the vapour 

condenser. Such systems use the vapour with 40±50% efficiency and produce hot water 

in amounts of 0.2±1.5 hl/hl of beer produced. TVR systems may be suitable for smaller 

and medium sized breweries. The investment in the steam jet compressor is relatively 

small, little maintenance is required and the compressor should run quietly. The system 

must be purged and air-free before it can operate. It must be free from leaks. 

10.8 Hot wort clarification 

At the end of the boil the wort should be absolutely clear (`bright') but contain, 

suspended in it, the remains of hops and flocs of trub or hot break. If whole hops are 

employed then the spent hops will probably weigh 0.7±1.4 kg/hl (2.4±5.0 lb./imp. brl) wet 

weight and will be associated with a significant amount of wort. The trub will be in the 

region of 0.21±0.28 kg/hl (0.75±1.0 lb./imp. brl) wet weight and will contain 80±85% 

water (Hough et al., 1982). Hot break contains roughly 50±60% crude protein, 20±30% 

tannin, 15±20% resins and 2±3% ash (dry wt., Andrews, 1992). Significant quantities of 

lipids are also present. Flocs of trub may reach 5±10 mm in diameter, but these can easily 

be disrupted, e.g., by pumping, into particles of 20±80 m diameter and a greater 

exposure to shear will reduce these to particles of 0.5±1.5 m. The hot break should be 

removed from the wort as thoroughly as possible, and this is most easily achieved with 

large particles. Consequently boiled wort should be handled gently and shear should be 

avoided to minimize damage to the trub. 

The methods used to separate spent whole hop cones and the remains of milled or 

milled and pelleted hops are different. In older, small breweries the remains of the hops 

were removed from the wort, when it was cast from the copper, by straining it through a 

cloth bag or a sieve. In larger breweries the hops were sieved out of the wort using a 

Montejus or hop jack (Fig. 10.17). Some trub was retained with the spent hops, which 


Sparge ring 

Wort in 

Mesh screen 

Agitator 

Hops discharge 

Fig. 10.17 A Montejus, or hop jack, used for sieving whole hops from wort when it is discharged 

from the copper (Hough et al., 1982). 


Spray ball 

for in-place 

cleaning 

Vent stack 

Spray plate 

Sparge arms 

Insulation 

Hot water for sparge 

Spray ball 

Wort main 

Water main 

Floor 

Slotted plates 

Underletting 

Spent hops 

discharge 

pipe Pump 

Wort 

Wort to 

cooling section 

recirculation 

pipe 

Fig. 10.18 Atraditional British hop back, for separating whole hops from cast wort (Hough et al., 

1982). 

were sparged with hot water from fixed sparge-rings, to recover entrained extract. The 

remaining trub settled in the coolships, when these were in use. The spent hops were 

discharged mechanically. Amore refined device, used for separating hops, is the hop 

back (Fig. 10.18). In these the wortcontaining the remains ofwhole hops is spread over a 

false bottom in avessel that, in later versions, is enclosed. The wort is withdrawn from 

below the plates, having slots of about 1.55mm (0.06in.) occupying 25±30% of the base 

and,atfirst,isreturnedtothetopofthebedofhopsthatisformed.Thisfiltersoffmostof 

the hot break. When the wort is clear (`running bright') it is directed to the cooler. The 

process is comparatively slow. The vessel, which in modern designs is equipped for CIP, 

superficially resembles amashtun, andindeedin some small breweriesthe mashtuns are 

used as hop backs. 

To work well the bed of spent hops should be 30±60cm (c. 1±2ft.) deep; 15cm (6in.) 

is regarded as the minimum depth. As hopping rates have been reduced and the use of 

milled and pelleted hops has become commonplace, hop backs have become less 

common. Sometimes aroma hops are added to the hop back, to impart more `hoppy 

character' to abeer. From the hop back, the clarified wort istransferred, either directly or 

after brief storage in abuffer tank, to the cooling system. To recover the wort retained in 

the spent hops these are sparged with hot water (typically 8l/kg spent hops; about 

0.8gal./lb.). Finally, the drained hops are removed. Spent hops may be disposed of with 

the spent grains. Cattle accept the grains/hops mixture. Cleaning may take place after 

every brew or after several brews. 

Hop separators, (Fig. 10.19), drain the hops over astrainer, carry them forward and 

then sparge them and compress them with either ascrew or abelt conveyor press, 

squeezing out residual liquid, before the spent hops are discharged. Separators do not 

separate most of the trub from the wort, and so trub removal has to be achieved in a 

separate operation. 

Increasingly hop powders, pelleted powders and extracts are used, and with these hop 

backs and separators cannot be employed. In these cases the wort may be clarified by hot 

filtration, by centrifugation, by sedimentation in acoolship or in amore modern settling 

tank. But the most commonly used device in newer breweries is the whirlpool tank or the 


1 

3 

2 

2 

2 

12 

Fig. 10.19 Ahop separator (Hough et al., 1982, courtesy of A. Ziemann, GmbH). 1. Valve;. 2. 

Level control electrodes; 3. Level controller; 4. Helical screw conveyor; 5. Sieve/strainer; 6. 

Compression chamber, where the hops are squeezed. 7. Fasteners; 8. Wort receiver under the 

second strainer; 9. Three-way valve (wort return and sparge water); 10. Return pipe; 11. Wort 

discharge; 12. Clean-out valve. Spent hops are discharged at the top. 

combined copper (or kettle)/whirlpool.Filtration of the hot wort may be carried out using 

vertical, horizontal or candle filters employing kieselguhr or Perlite filter aids. Filtration 

gives exceptionally clear worts, but wort losses can be significant, it is difficult to 

maintain sterility and there are the costs of purchasing the filter aids and disposing of the 

used aid and sludge. Coolships are now rarely or never used although their shallow liquid 

layers allowedeffectivesedimentationand wort clarification. Wortlossesandthe risksof 

microbiological contamination were high. 

Thelogical successorstocoolshipsare sedimentationtanks(Fig. 10.20; Haecht, vanet 

al., 1990; Rehberger and Luther, 1994; Kunze, 1996; Vermeylen, 1962; Versteegh, 

1989). These may have flat or conical bases and the sides of the tanks, which are 

generallycylindrical, may ormay not be cooled by water inwall jackets. If cooling water 

is used it is itself heated and is retained for use. The vessels are enclosed and have 

vapour-escape stacks. Usually each vessel is filled from above (although in modern 

vessels bottom filling, to minimize oxygen pick-up and wort oxidation, would be 

expected) and is allowed to stand for about one hour, while solids in suspension 

progressively sediment. Wort is drawn from the top of the liquid, which clarifies first, 

using ahinged tube supported by afloat. The turbidity of the wort that is collected is 

continuously monitored and if the value rises too much collection is slowed or stopped. 

The trub and hop fragments form asloppy, `turbid wort' that is collected and processed 

separately. There has been aproposal to cool the edge of the surface of the resting wort 

by evaporation, to encourage the downward flow of the wort by the vessel wall, across 

thebaseofthevessel,whereprecipitatedmaterialsaccumulate,andupwardsinthecentre 

(Versteegh, 1989). The result claimed is aconical deposit of material in the middle of the 

flat base of the vessel analogous to that obtained in whirlpool vessels (see below). 

Sometimes the hot wort is clarified by centrifugation in adecanter or awort-clarifying 

centrifuge. These devices may also be used for wort recovery from wet trub samples. It is 

desirable, particularly if the wort is heavily hopped, to pre-treat it by passage through a 

rotary brush strainer and ahydrocyclone, to remove some of the hop solids and abrasive 


11 

10 

4 

9 

5 

8 

6 

7

Air vent 

Lower wort 

collection tube 

Trub 

pump 

Valves 

Float 

Wort inlet 

Valve 

Trub outlet 

Valve 

Valve 

Hinged wort 

collection tube, 

with float 

Wort 

pump 

Fig. 10.20 An empty hot wort settling tank (after Rehberger and Luther, 1994). After aperiod of 

settling clear wort is drawn from the top, through the floating, hinged arm. At the end of the settling 

period the last of the clear wort is collected from the lower collection point and then the trub, mixed 

with wort, is also drawn off. 

materials, before it enters the centrifuge. Modern centrifuges discharge collected solids 

automatically, they give good wort recoveries (losses of

(a) 

Taylor eddy 

(b) 

Rotation 

Centrifugal 

force 

Trub cone 

Planetary 

eddy 

Wort outlet 

Surface of wort 

Desired pattern of 

wort circulation 

Torus eddy 

Level for inserting 

lattice or rings 

Fig. 10.21 Currents in whirlpool tanks. (a) The ideal flow pattern in a whirlpool. (b) Undesirable 

eddy currents that can occur and which interfere with trub separation. The most troublesome are the 

torus eddies, which may be checked by the insertion of a lattice or rings at the level indicated (after 

Denk, 1991). 

In the centre of the tank the liquid rises, leaving solids deposited. The vessel is usually 

filled in 10±15 minutes and after 20±30 minutes the wort has cleared and the solids are 

deposited as a cone in the centre of the base of the vessel, surrounded by a clear, annular 

space. Wort is typically withdrawn from three exit points, often about half-way down the 

vessel, three-quarters of the way down and in the base of the vessel towards the side, 

where the clear zone should be. This allows the wort to be drawn off at successively 

lower levels as it clears. Its turbidity is generally monitored automatically. One or more 

of the upper take-off points may withdraw wort tangentially to support the rotary motion. 

Sometimes new whirlpools do not work in a satisfactory way and must be modified. 

Empirical and theoretical studies have improved the situation. Problems arise quite 

separately from the design of the whirlpool and/or how it is operated and from problems 

with the trub caused by factors in the brewing process, which operate before the wort 

enters the whirlpool. In attempts to improve whirlpool performance a large number of 

types of vessel bases have been tried, including flat and level or slightly inclined (with or 


without central `cups', or sumps of various sizes or a circumferential gully), inverted 

cones and conical bases (Andrews, 1988; Andrews and Axcell (private communication); 

Denk, 1991, 1994, 1998; Wilkinson, 1991 a, b). 

One problem is that secondary eddies in the wort can disrupt the desired flow pattern 

and impede the sedimentation of the trub and hop residues (Fig. 10.21b). Two solutions 

have been proposed; first the installation of an annular, horizontal metal grid or, secondly, 

a set of 4±5 circular metal rings 25±60 cm above the base of the vessel. In each case these 

devices impede the flows of the secondary torus eddies and greatly improve the deposition 

of the trub. These devices are installed only if experience shows that they are needed. 

Sumps interfere with the desired flow pattern and encourage the collection of as much as 

2% of the wort volume as trub wort, which must be processed separately to recover the 

entrained extract. The most popular configurations are now vessels with either flat bases, 

slightly inclined to the horizontal, to encourage drainage of the wort to the exit point at the 

outer edge of the clear zone, which surrounds the trub cone, or shallow, conical bases, with 

angles of about 30ë (Fig. 10.22). In the first case, the cone remaining after the vessel has 

been drained should be dry. Subsequently it must be dislodged and dispersed, for example, 

with pulses of water from rotating, electrically driven jets, so that it drains to the lowest 

wort outlet. This cleaning should not be long delayed as the exposed trub cone can set hard 

in air. In the other case the trub, still with some wort, is withdrawn from the apex of the 

cone, and the wort is recovered by centrifugation. 

Problems with the trub cone may be caused at earlier stages in the brewing process by 

using an abnormal grist, by having inadequate trub flocculation in the boil or by breaking 

up the trub through shearing the wort during pumping, by using a high-gravity wort 

which, because of its high density `buoys-up' the trub, and so on. Attempts to improve 

cone formation have included the exclusion of air, additions of copper finings, nylon or 

PVPP powders, kieselguhr, bentonite and KMS (potassium metabisulphite). As the wort 

is withdrawn the trub cone is exposed and so is no longer supported by the buoyancy 

provided by surrounding wort. It then tends to slump and spread outwards and even break 

Working level 

Trub 

20° 

Vapour 

stack 

1:40 slope (1.5°) 

Trub jetting 

machine 

Tangential 

inlet 

Outlet 1, 

at 50% 

Outlet 2, 

at 10% 

Outlet 3, 

at trub 

Trub outlet 

(a) (b) 

20° 

Working level 

120° 

Vapour 

stack 

Outlet 3, 

at VT 

Tangential 

inlet 

Outlet 1, 

at 50% 

Outlet 2 

at 10% 

Fig. 10.22 Two patterns of kettle-whirlpools (Courtesy of Briggs of Burton, plc.). In each case 

boiling wort, driven by the thermosyphon effect generated in an external wort heater, is discharged 

tangentially above the surface of the wort. Volatiles readily escape with the vapour, and the wort 

rotates in the vessel driven by the incoming wort. (a) In this flat-based pattern direct wort recovery 

is favoured. The trub cone is drained and then, after wort recovery, is driven out with water jets. (b) 

In this pattern, with a conical base, trub is recovered as a slurry in wort, which is normally 

recovered. This design favours high wort clarity. 


,

up, so it may reach the lowest wort collection point. To counteract this, wort withdrawal 

is slowed when the trub cone is just exposed so that the trub is slowly drained and is 

exposed more slowly to the air, and the cone retains its shape better. 

Kettle-whirlpools are of various types. Some are equipped with internal heaters and 

wortiswithdrawn andpumpedback viaaninjectionentrytoinducerotationintheliquid. 

This design requires the continuous use of apump, with the equipment and maintenance 

costs and disrupting effects on the trub that this entails. The internal heater clutters the 

centre of the vessel. Abetter design uses an external heater that circulates the boiling 

wort using the thermosyphon effect. The boiling wort is returned to the top of the vessel 

at or above the wort surface and, being delivered through atangential entry, drives the 

rotation of the liquid (Fig. 10.22). It can be operationally more convenient to have two 

copper-whirlpool vessels rather than a single copper and a single, separate whirlpool. 

The wet trub produced by hot wort clarification contains, before washing, about 80% 

of wort. If this is sent to drain its high BOD and large amount of suspended solids give 

rise to high effluent charges. In addition extract is wasted. Disposal with the spent grains 

also wastes extract. If the brewing schedule permits it, extract is most easily recovered by 

transferring the trub to the mashing vessel or lauter tun making the next brew. Sometimes 

clear wort has been recovered from trub by filtration or by centrifugation (Hansen et al., 

1990). It is less costly to centrifuge the trub from a whirlpool than to omit the whirlpool 

and centrifuge all the wort. Cloudy wort has been obtained by vibrating screen filtration 

of trub, (Fig. 10.23), and this treatment has been used to pre-treat trub to provide the feed 

for complete clarification by cross-flow filtration (Maule et al., 1989; Visscher et al., 

1991). The cross-flow filtration was carried out with ceramic filter units with notional 

0.2 m pores. Their use increased the brewhouse yield by 1%. However, the units were 

easily fouled, reducing the filtration rate and extract recovery. To minimize this problem 

the re-circulating trub/wort was diluted in the later stages of treatment. 

After the boil wort may remain in the whirlpool for up to an hour before being cooled. 

During this warm period chemical changes continue. Some, like the isomerization of the 

Mesh screen 

Clarified 

material 

out 

Eccentric 

weights 

Feed material 

in 

Solid 

material 

out 

Motor 

Solids 

out 

(a) 

(b) 

Fig. 10.23 (a) A vertical section of a vibrating screen filter and (b) the distribution and movement 

of solids on the filter surface as seen from above (Button et al., 1977). 


-acids, are desirable but others, such as the generation of some volatile Maillard 

reaction products and DMS, may not be. The proposal to slow these changes by cooling 

the wort alittle as it enters the whirlpool has already been noted. Another approach, 

already noted, is to `strip' the wort of volatiles as it leaves the whirlpool, just before it is 

delivered to the cooling system. This has been achieved in different ways, for example, 

by spraying the wort into avacuum chamber, by thin-film evaporation or by steamstripping. 

Similar effects occur in the expansion/cooling chambers in high-temperature, 

continuous wort boiling (Section 10.5). 

10.9 Wort cooling 

After clarification the hot wort must be cooled to the temperature at which it is pitched 

(inoculated) with yeast. Traditionally this is about 15±22ëC (59±71.6ëF) for ales and 6± 

12ëC (42.8±53.6ëF) for lagers, but other temperatures are used. The cooling should be 

carriedoutrapidlyandunderasepticconditionstostopchemicalreactionscontinuingand 

to minimize chances of growth of any contaminating microbes. As the wort cools it 

becomeshazy asacoldbreakforms. Thismayormaynotberemoved(Section10.10).In 

addition the wort must be charged with oxygen to an appropriate level (Section 10.11; 

Chapters 11 and 12). In modern breweries the heat from the hot wort is partly recovered 

in hot water. 

The oldest coolers were the open, horizontal coolships, (Section 10.8), in which the 

wort was held in rectangular hardwood, copper, iron or aluminium trays in ashallow 

layer of, say, 15±25cm (c. 6±10in.), for as long as 12 hours, to cool, to pick up oxygen 

from the air and to deposit suspended materials, including hop fragments, hot break and 

some cold break. Ideally, the room holding the coolship had acurved ceiling designed to 

preventcondensation drippingintothewortand,inlatertimes,wasventilated withsterile 

air (Hind, 1940; Sykes and Ling, 1907; Vermeylen, 1962). In practice coolships were 

difficult to keep clean and free from microbiological contamination, they took up much 

space and wort losses were high. In some cases cooling was accelerated by passing cold 

water through pipes placed on the base of the coolship, an arrangement that accelerated 

cooling, but with the added disadvantage of making the equipment awkward to clean. 

The introduction of vertical coolers was a substantial advance, since they were 

compact, easier to clean, and cooled the wort comparatively rapidly. The wort was 

introduced to atrough at the top, from which it trickled down as athin film over aseries 

ofhorizontal tubeswithin which cold waterflowedupwards, inacounter-flow. Thetubes 

mightbeofvariouscross-sectionsorhaveprojectionsonthemdesignedtodirecttheflow 

of the wort and to increase the area of the heat-transfer surfaces. At the base of the cooler 

the wort was collected in atrough and piped to afermenter. The wort was aerated in this 

process, but the cold break was not removed. 

Sometimeswortiscooled,inclosedshellandtubecoolers,bypassingitdownthrough 

stainless steel tubes, which are successively cooled by air in the upper part, water in the 

central part and refrigerant in the lower part. Sterile air is injected into the base and rises 

upthetubesinacounterflowtothewort.Aunitabletocool450hl/h(275brl/h)wouldbe 

about 7m(23ft.) tall and 2m(6ft. 6in.) wide (Hough et al., 1982). By far the most 

common coolers are plate heat exchangers (paraflows), which are compact, efficient and 

versatile. Being enclosed microbes are excluded, and cleaning and sterilization are 

simple. In these coolers numerous stainless steel plates are suspended vertically from a 

strong metal frame and are clamped together (Fig. 10.24). The numbers of plates can be 


(a) 

Hot wort 

in (93°C) 

(b) 

Water out (68°C) Water in (16°C) Glycol out Glycol in 

(–1°C) 

Cool wort 

out (10°C) 

Frame supporting plates 

and holding them together 

Fig. 10.24 (a) The principles of a plate and frame heat exchanger, indicating the movement of the 

liquid to be cooled and the counterflow of the cooling refrigerant (Hough et al., 1982). (b) The 

flows in a two-stage cooling system (after Hough, 1985). In the first stage wort is cooled by a 

counterflow of water, which is heated in the process. In the second stage the wort is cooled further 

by a chilled glycol refrigerant. 

varied and, as a result, so can the exchangers' cooling capacity. The plates are indented 

with complex patterns and have four holes in the corners. The edges of the plates and the 

holes are rimmed with rubber gaskets so that when the plates are pressed together the 

gaskets seal, and a series of ducts and channels are formed. In operation the wort flows 

through one channel, across the stack, between two plates and out to another channel 

while cold water or refrigerant flows in the opposite direction (`counter-current') in 

alternating gaps between the plates. The indented patterns on the plates ensure turbulent 

flow and efficient heat exchange. 

The coolants can be supplied in various ways. For example, water should be the first 

so that it can be heated, to 70 ëC (158 ëF) or more, and used around the brewery. Part way 

through the cooler, cooling may be supplied by chilled water or a glycol, ammonia or an 

alcohol refrigerant. The arrangement used will be decided, in part, by the final wort 


temperature required. However, if acoolant other than water is used care is needed to 

regularly check the integrity of the plates and seals, so that there is no possibility that a 

coolant can leak and contaminate the wort. Coolers must be regularly cleaned, to remove 

scaleand fouling, and must be checked for leaks. The plates are mounted toallow ease of 

separation and replacement. Sometimes sterile air or oxygen is injected into the wort in 

the cooler to take advantage of the turbulent flow that encourages the gas to dissolve. 

10.10 The cold break 

As the wort cools it becomes cloudy as the cold break or trub separates from solution. 

This material contains about 50% protein, 15±25% polyphenols and 20±30% of wort 

carbohydrates (see Chapter 9). Unlike the hot break this material does not flocculate, and 

occurs as small particles,

Cold break (mg/ml) 

300 

250 

200 

150 

100 

50 

0 

10.11 Wort aeration/oxygenation 

50 68 86 104 122 140 158 176 194 

Temperature (°F) 

0 10 20 30 40 50 60 70 80 90 


Fig. 10.25 The amounts of cold trub (break) formed in aparticular wort at various temperatures 


In the initial stage of `fermentation' the freshly pitched yeast needs to be in wort that 

containsdissolvedoxygen.Theconcentrationofoxygenrequirediscritical,anddependson 

the wort, for example, the availability of sterols and unsaturated fatty acids, and the variety 

and history of the yeast (Chapters 11 and 12). In the past it seems that saturating wort with 

oxygen from air (approx. 21% O 2)was sufficient, but now saturation with pure oxygen is 

often required. It is surprisingly difficult to dissolve oxygen quickly in aqueous solutions. 

At equilibrium the amount of gas dissolved at achosen temperature is proportional to the 

partial pressure of the gas above the liquid. Although solubilities are often reported at a 

standard atmospheric pressure, (equal to that at the base of acolumn of mercury 760mm 

(29.92in.) high at 0ëC (32ëF), at sea level at 45ë latitude), decreasing or increasing this 

pressure will proportionally change the equilibrium concentration of the gas in solution. 

Atmospheric pressure is always fluctuating. As the temperature increases, so the amount of 

oxygen in solution, in equilibrium with air or pure oxygen, declines (Table 10.4). 

Dissolving substances (salts, sugars, etc.) in water reduces the amount of gas that can 

be dissolved. Wort is astrong solution of amixture of substances and so the solubility of 

oxygen in wort is less than in pure water (Table 10.4), and the stronger, more 

concentrated the wort the less oxygen will dissolve in it under afixed set of conditions. 

Wort may be injected with air or oxygen at the inlet to the cooler, part-way through the 

cooler (Section 10.9), or after the cooling process is complete. Adding the gas to hot or 

warm wort allows the oxidation of wort components, causing flavour changes and 

darkening, which are usually undesirable. However, adding the gas to the hot wort means 

that the sterility of the gas is not so critical as microbes will be killed at the elevated 

temperatures. Air or oxygen added to cooled wort must have been sterilized, either by 

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 

temperatures and from pure oxygen gas, all at a standard atmospheric pressure (data of Krauss, 

1967). Compare with Appendix A.12. 

Temperature From air From oxygen 

(ëC) (ëF) Water Wort (12%) Water 

0 32 14.5 11.6 69 

3 37.4 ± ± 64 

5 41 12.7 10.4 61 

8 46.4 ± ± 56 

10 50 11.2 9.3 54 

15 59 10.0 8.3 48 

20 68 9.9 7.4 ± 

The rate of dissolution of a gas in liquid depends on the pressure, the thoroughness of 

mixing and the area of the gas/liquid interface. Consequently, it is most efficient to direct 

a stream of very fine gas bubbles, under pressure, into a turbulent flow of wort. Devices 

used for aeration/oxygenation include sintered metal or ceramic candles, which release 

clouds of very fine bubbles into the base of a vessel or into a flowing stream of wort and 

are strong but which are difficult to keep clean, and centrifugal mixers which are very 

efficient but are expensive (Kunze, 1996). Devices based on the Venturi tube principle, 

which aerate/oxygenate cooled wort while flowing to a fermenter, are common. The 

flowing wort comes to a restriction in the pipework, which causes it to accelerate and the 

pressure to drop. At this point fine bubbles of air or oxygen are introduced into the liquid 

either from a fine nozzle, discharging into the stream of wort, or from fine perforations or 

sintered material in the tube wall. The clouds of bubbles are carried forward into the next 

section where the pipe abruptly expands, the flow slows and so the pressure rises and the 

flow becomes turbulent, conditions which favour the rapid solution of the gas. Sometimes 

the pipework downstream from the gas injection point contains inserts of metal `tapes' or 

net-like units that act as mixers, creating turbulence in the wort flow. 

In modern plant the dissolution of added oxygen is nearly complete. Oxygenation is 

automatically controlled, the rate of supply of gas is continuously monitored, for 

example, by measuring the flow rate or the decline in weight of gas cylinders. 

Alternatively, mass flow meters may be used. Sometimes oxygen is added to the first part 

of the wort flow and yeast is added to the second part. The level of dissolved oxygen, 

(really the equilibrium partial pressure), is usually monitored continuously by a 

membrane-covered oxygen electrode or cell. 


ANDREWS, J. M. H. (1988) Ferment, 1 (3), 47. 

ANDREWS, J. M. H. (1992) Proc. 22nd Conv. Inst. of Brewing (Australia and New Zealand Section), 

Melbourne, p. 65. 

ANDREWS, J. M. H. and AXCELL, B. C. Private communication. 

ANON. (1994) J. Inst. Brewing, 100, 130. 

BENITEZ, J. L., FORSTER, A., DE KEUKELEIRE, D., MOIR, M., SHARPE, F. R., VERHAGEN, L. C. and WESTWOOD, 

K. T. (1997) EBC Manual of Good Practice: Hops and Hop Products. 185 pp. NuÈrnberg, GetraÈnke- 

Fachverlag Hans Carl. 

BOER, F. P. DE (1991) Eur. Brew. Conv. Monograph-XVIII. EBC Symposium; Wort boiling and 

clarification. Strasbourg, p. 193. 

BONACCHELLI, B., HARMEGNIES, F. and GIL, R. T. (2001) Proc. 28th Congr. Eur. Brew. Conv., Budapest 

(CD, paper 24; p. 235). 


BOYES, R. (1993) Brewers' Guard., 122 (3), 23. 

BRAEKELEIRS, R. and BAUDUIN, C. L. (2001) Proc. 8th Brew. Conv., Inst. of Brewing (Central and 

Southern African Section), Sun City, p. 36. 

BUTTON, A. H., STACEY, A. J. and TAYLOR, B. (1977) Proc. 16th Congr. Eur. Brew. Conv., Amsterdam, p. 377. 

CHANTRELL, N. S. (1983) Proc. 19th Congr. Eur. Brew. Conv., London, p. 89. 

CHANTRELL, N. S. (1984) MBAA Tech. Quart., 21 (4), 166. 

CLARKE, M. H. M. and KERR, R. A. (1991) Eur. Brew. Conv. Monograph-XVIII. EBC Symposium: Wort 

boiling and clarification, Strasbourg, p. 139. 

COUTTS, M. W., RICKETTS, J. and SELKIRK, R. C. (1955) Proc. Conv. Master Brewer's Assoc. Amer., p. 20. 

CROMPTON, I. E. and HEGARTY, P. K. (1991) Proc. 23rd Congr. Eur. Brew. Conv., Lisbon, p. 625. 

DENK, V. (1991) Eur. Brew. Conv. Monograph-XVIII. EBC Symposium: Wort boiling and clarification. 

Strasbourg, p. 155. 

DENK, V. (1994) Ferment, 7 (5), 299. 

DENK, V. (1998) Brauwelt Internat., 16 (1), 31. 

DICKEL, T., KROTTENTHALER, M. and BACK, W. (2002) Brauwelt Internat., 20 (1), 23. 

DYMOND, G. and DJURSLEV, O. (1994) Brew. Distill. Internat., 25 (4), 16. 

FOHR, M. and MEYER-PITTROFF, R. (1998) Brauwelt Internat., 16 (4), 304. 

FORREST, I. S., SKRGATIC, D., COKER, I. A. J. and HEAP, J. (1993) Proc. 24th Congr. Eur. Brew. Conv., Oslo, 

p. 493. 

FORTUIN, B. (1995) The Brewer, 81, 443. 

GRASMAN, R. and VAN EERDE, P. (1986) Proc. 19th Conv. Inst. Brewing (Australia and New Zealand 

Section), Hobart, p. 161. 

HACKENSELLNER, T. (1999) Brauwelt Internat., 17 (6), 495. 

HAECHT, J.-L., VAN, DE BRACKELEIRE, C., DUFOUR, J.-P. and DEVREUX, A. (1990) Eur. Brew. Conv. 

Monograph-XVI. EBC Symposium, `Separation Processes', Leuven, p. 96. 

HALL, R. D. and FRICKER, R. (1966) Proc. 9th Conv. Inst. Brewing (Australia and New Zealand Section), 

Auckland, p. 45. 

HANCOCK, J. C. and ANDREWS, J. M. H. (1996) Ferment, 9 (6), 344. 

HANSEN, N. L., LUND, M. and OLSEN, N. O. (1990) Eur. Brew. Conv. Monograph-XVI. EBC Symposium 

`Separation Processes', Leuven, p. 84. 

HERRMANN, H. (1985) Brew. Distill. Internat., Mar., p. 32. 

HERRMANN, H. (1998a) The Brewer, 84 (1005), 333. 

HERRMANN, H. (1998b) Ferment, 11 (1), 36. 

HERRMANN, H. (1999) Ferment, 12, Feb/Mar., 36. 

HIND, H., L. (1940) Brewing Science and Practice, II. Brewing Processes. London, Chapman and Hall, 

pp. 507±1020. 

HOUGH, J. S. (1985) The Biotechnology of Malting and Brewing. Cambridge University Press. 169 pp. 

HOUGH, J. S., BRIGGS, D. E., STEVENS, R. and YOUNG, T. W. (1982) Malting and Brewing Science. II. 

Hopped Wort and Beer. London, Chapman and Hall, pp. 389±914. 

HUDSTON, H. (1969) MBAA Tech. Quart., 6 (1), 164. 

KANTELBERG, B., WIESNER, R., JOHN. L. and BREITSCHOPF, J. (2000) Brew. Distill. Internat., 31 (3), 16. 

KOLLNBERGER, P. (1986) MBAA Tech. Quart., 23, 126. 

KOLLNBERGER, P. (1987) Brauwelt, 127, 254. 

KRAUSS, G. (1967) Proc.11th Congr. Eur. Brew. Conv., Madrid, p. 35. 

KROTTENTHALER, M. and BACK, W. (2001) Brew. Distill. Internat., 32 (3), 14. 

KROTTENTHALER, M., HARTMANN, K. and BACK, W. (2001) Brauwelt Internat., 19 (6), 457. 

KUNZE, W. (1996) Technology Brewing and Malting. (Internat. edn, translated Wainwright, T.) Berlin, 

VLB, p. 254. 

LANGENHAN, R. (1995) J. Inst. Brewing, 101, 230. 

LENZ, B., LANGENHAN, R., HERRMANN, H., KANTELBERG, B. C. and FELGENTRAEGER, W. (1991) Proc. 3rd 

Sci. Tech. Conv. Inst. of Brewing, (Central and Southern African Section), Victoria Falls, p. 101. 

LUSTIG, S., KUNST, T. and HILL, P. (1997) Proc. 21st Congr. Eur. Brew. Conv., Maastricht, p. 341. 

MANGER, H.-J. (1998) Brauwelt Internat., 16 (4), 320. 

MAULE, D. R. and CLARK, B. E. (1985) Proc. 20th Congr. Eur. Brew. Conv., Helsinki, p. 379. 

MAULE, D. R., STEAD, J. R. and CLARK, B. E. (1989) Proc. 22nd Congr. Eur. Brew. Conv., Zurich, p. 393. 

MICHEL, R. A. (1991) Eur. Brew. Conv. Monograph-XVIII. EBC Symposium; `Wort boiling and 

clarification'. Strasbourg, p. 118. 

MIEDANER, H. (1986) J. Inst. Brewing, 92, 330. 

MIEDANER, H. and NARZISS, L. (1986) Eur. Brew. Conv. Monograph-XI. EBC Symposium; `Wort 

Production'. Maffliers, p. 80. 

MITANI, Y., SUZUKI, H., ABE, T., NOMURA, M. and SHINOTSUKA, K. (1999) Proc. 27th Congr. Eur. Brew. 

Conv., Cannes, p. 619. 

MORRIS, D. R. (1987) Brew. Distill. Internat., 17 (3), 22. 

NARZISS, L. (1986a) Eur. Brew. Conv. Monograph-XI. EBC Symposium; `Wort Production'. Maffliers, 

p. 98. 


NARZISS, L. (1986b) Brauwelt, 126 (32), 1419. 

NARZISS, L. (1992) Die Bierbrauerei. II. Die Technologie der WuÈrzebereitung. (7th edn). Stuttgart, 

Ferdinand Enke. 402 pp. 

NARZISS, L. (1993) Proc. 4th Sci. Tech. Conv. Inst. of Brewing, (Central and Southern African Section), 

Somerset West, p. 195. 

NARZISS, L., KIENINGER, H. and REICHENDER, E. (1971) Proc. 13th Congr. Eur. Brew. Conv., Estoril, 

p. 197. 

NARZISS, L., MIEDANER, H. and SCHNEIDER, F. (1992) Brauwelt Internat., IV, 346. 

ORMROD, I. H. L. (1986) J. Inst. Brewing, 92, 131. 

REED, R. J. R. (1992) Ferment, 5 (2), 125. 

REHBERGER, A. J. and LUTHER, G. E. (1994) in Handbook of Brewing (Hardwick, W. A. ed.), New York, 

Marcel Dekker, p. 247. 

ROYSTON, M. G. (1971) in Modern Brewery Technology. (Findley, W. P. K. ed.). London, Macmillan and 

Co., p. 60. 

SCHU, G. E. (1995, Oct). Brauwelt Internat., 13 (4), 316. 

SCHWILL-MIEDANER, A. and MIEDANER, H. (2002) Brauwelt Internat., 20 (1), 19. 

SELDESLACHTS, D., VAN DER EYNDE, E. and DEGELIN, L. (1997) Proc. 26th Congr. Eur. Brew. Conv., 

Maastricht, p. 323. 

STIPPLER, K. (2000) Ferment, 13 (1), 34. 

STIPPLER, K., WASMUHT, K. and GATTERMEYER, P. (1997) Brauwelt Internat., 15 (4), 358. 

SYKES, W. J. and LING, A. R. (1907) The Principles and Practice of Brewing. (3rd edn). London, Charles 

Griffin and Co., p. 496. 

UNTERSTEIN, K. (1992) Brauwelt Internat., 10 (1), 65. 

VERMEYLEN, J. (1962) TraiteÂ de la Fabrication du Malt et de la BieÁre. 2, pp. 861, 937. Gand. Assoc. 

Royale des Anciens EleÁves de l'Institute SupeÂrieur des Fermentations. 

VERSTEEGH, C. W. (1989) Proc. 22nd Congr. Eur. Brew. Conv., Zurich, p. 291. 

VISSCHER, H. J., TETTELAAR, M. E. and MARTENS, F. B. (1991) Proc. 23rd Congr. Eur. Brew. Conv., Lisbon, 

p. 649. 

VOLLHALS, B. (1994) MBAA Tech. Quart., 31 (1), 1. 

WILKINSON, N. R. (1985) Proc. 1st Sci. Tech. Conv. Inst. of Brewing (Central and Southern African 

Section), Johannesburg, p. 188. 

WILKINSON, N. R. (1991a) Eur. Brew. Conv. Monograph-XVIII. EBC Symposium; `Wort Boiling and 

Clarification'. Strasbourg, p. 100. 

WILKINSON, N. R. (1991b) Ferment, 4 (6), 388. 


11 

Yeast biology 

11.1 Historical note 

The abilities of Saccharomyces yeast to leaven dough in baking and generate ethanol for 

beverage production have been utilized by mankind for several millennia, albeit for most 

of that time, unwittingly. The consumption of alcoholic beverages is common to all 

civilizations. Production of the earliest drinks was probably serendipitous since natural 

sources of sugars are invariably contaminated with yeast. Metabolism of sugars by 

Saccharomyces yeasts results in the formation of ethanol and carbon dioxide even under 

aerobic conditions. The mind-altering effects of ethanolic beverages must have provided 

the primary motive for their continued consumption and a powerful impetus for empirical 

development of a controlled process of production. Coincidentally, such drinks would 

have been nutritious and a useful means of sanitizing potentially dangerous water supplies. 

Brewing of beer has its probable origins in the Middle East at some time between 6000 

and 8000 BC, where it apparently developed in tandem with organized agriculture 

(Corran, 1975). It seems that the use of cereals for baking and brewing developed 

simultaneously. Clearly, this must also have included the discovery of malting and the 

use of yeast for leavening of dough and fermentation. These activities became large-scale 

undertakings, for example, it has been reported that in ancient Mesopotamia 40% of 

cereals were cultivated purely for brewing (Corran, 1975). 

Brewing spread from the Middle East to become the dominant alcoholic beverage of 

northern Europe. Thus, Tacitus commented that beer was the common drink in Germany 

at the time of the Roman Empire (King, 1947). It was from old German that the word 

yeast arose, probably from gischen descriptive of the foaming or frothing during 

fermentation. Pliny described the use by the Gauls for leavening of dough of `foam' 

obtained from beer fermentations (King, 1947). It is implicit in this etymology that the 

importance of yeast to brewing and baking was appreciated in historical times although 

its vital nature was unrecognized. For example, in medieval England the yeast crop 

resulting from fermentation was known as godisgood (Forget, 1988). Nevertheless, the 

earliest manifestation of the German Reinheitsgebot or beer purity laws introduced to 

Bavaria in 1493 by Duke Albrecht IV stipulated the exclusive use of hops, malted barley 


and water for brewing. Yeast was not included as an ingredient since it was unknown 

(Narziss, 1984). 

The first description of individual yeast cells was published in 1680 by Antonie van 

Leeuwenhoek (Chapman, 1931). Using a primitive microscope he recorded the 

appearance of yeast flocs in fermenting wort and modelled the same in wax. He also 

noted the formation of gas bubbles but did not appreciate that this was a by-product of 

yeast metabolism. The realization of the biological nature of fermentation forms the 

basis of the development of modern biochemistry and microbiology. Early alchemists 

observed the vigorous gas formation that accompanied the metabolism of wort by yeast 

and coined the term fermentation from the Latin fevere, to boil. The alchemist view of 

fermentation was that this was a process of active separation. Thus, ethanol was present 

but could not be detected until the ìmpurities' yeast and carbon dioxide were removed 

(Florkin, 1972). 

Gay-Lussac in 1810 established the stoichiometry of ethanolic fermentation by 

demonstrating that two molecules each of ethanol and carbon dioxide devolved from each 

molecule of sugar consumed. He suggested that fermentation was initiated by exposure to 

oxygen since heated foodstuffs in sealed containers spoilt only when air was admitted. He 

also noted that spoilage could be stopped by further heating but failed to form the 

conclusion that there was a heat-labile component present. He suggested that heat 

changed the `ferment' to an inactive form in which it was impervious to the stimulating 

effects of oxygen. Yeast was not considered to play any part in fermentation on the basis 

of insolubility. 

Undoubtedly, early practitioners of brewing and baking appreciated the essential 

requirement for yeast in fermentation by dint of empirical observation. Later 

incarnations of the Reinheitsgebot included yeast as an ingredient in beer making 

based on the observations that addition to wort of some of the crop from a previous 

fermentation gave a more consistent process (Narziss, 1984). The role of yeast in 

fermentation and its vital nature was established independently by the Germans, 

Theodor Schwann (1837) and Friedriech Traugott KuÈtzing and a Frenchman, Charles 

Cagniard-Latour (1836). Both KuÈtzing and Cagniard-Latour used direct microscopic 

observation to describe yeast cells. The latter recorded yeast proliferation by the 

formation of buds and both deduced that the cells were living and had an active role in 

fermentation. 

Schwann's investigations were directed towards disproving the theory of spontaneous 

generation of life. He demonstrated that a meat infusion did not putrefy if heated. 

Furthermore, the onset of putrefaction subsequently brought about by the admission of air 

was prevented if the latter was first heated. He was able to demonstrate that heating of air 

did not change its essential character since it could still sustain the life of a frog. This 

disproved the earlier assertions of Gay-Lussac. Schwann also made microscopic 

observations of growing yeast cells, which he termed Zuckerpilz. 

Despite the observations of Schwann, the vital nature of yeast was not generally 

accepted until the work of Louis Pasteur was published (Anderson, 1995). Pasteur studied 

optically active molecules, which he considered were produced only by living organisms. 

He isolated optically active amyl alcohol from fermentations and therefore assumed that 

the process must be animate. Pasteur extended the work of Schwann and demonstrated 

that putrefaction of foodstuffs occurred via contamination with air-borne microorganisms. 

He showed that putrefaction did not occur in heated broth even with free 

access to air, provided that ingress of micro-organisms was prevented by the ingenious 

design of his `swan-necked' flasks. 


Pasteur made careful microscopic examination of beer fermentations, the results of 

which were published in his EÂtudes sur la bieÁre of 1876. He observed the growth of 

brewing yeast cells and demonstrated that these were responsible for fermentation. He 

also recorded the symptoms of several specific `diseases' of fermentation and identified 

the causative microbial contaminants associated with each. Thus, he was instrumental in 

pointing out the necessity of adopting the highest standards of hygiene for successful 

brewing. He designed equipment that fulfilled this requirement. Central to his 

recommendations was the realization that microbial growth on sugars or wort was via 

contamination and not by spontaneous generation. Furthermore, new microbial cells 

arose solely from similar parental cells. In order to reach these conclusions Pasteur 

developed methods for the sterilization of media and the preservation and propagation of 

microbial cultures. 

Pasteur recommended the use of microscopy as an aid in production brewing. Others 

have pointed out that this instrument had already been adopted by several brewers prior to 

Pasteur's visits (Anderson, 1995). In any case by the early twentieth century, JoÈrgensen in 

the third edition of his treatise entitled Micro-organisms and Fermentation, published in 

English translation in 1900 (JoÈrgensen, 1900) described methods for assessing yeast 

`health' by microscopic observation. Pasteur was comparatively indifferent to the 

nuances of cellular morphology of individual species. The ability to differentiate microorganisms 

was largely dependent on the development of methods for isolating and 

propagating pure cultures. This was difficult using the liquid cultures employed by 

Pasteur. 

The use of solid microbiological media on which pure cultures could be isolated from 

colonies derived from single cells was pioneered by the work of the medical 

bacteriologist, Robert Koch (1881). In 1883, Emil Hansen, working at the Carlsberg 

Foundation in Copenhagen, used similar techniques to isolate the first pure yeast culture. 

At the same time Hansen introduced the first modern apparatus for propagating yeast and 

the first pure culture of `Carlsberg Yeast Number 1' was used successfully in commercial 

brewing (Curtis, 1971). The availability of pure cultures provided the means of 

elucidating the yeast life cycle. Hitherto, yeast cells had been considered to be merely 

phases in the life cycles of other organisms such as moulds, bacteria or even algae (Rose 

and Harrison, 1971). An assistant of Hansen, SchioÈnning reported the occurrence of a 

sexual phase in the yeast life cycle. This was confirmed in the 1930s when éjvind Winge 

also working at the Carlsberg Foundation provided a full description of the yeast haploand 

diplophases. 

As early as 1897, BuÈchner demonstrated the formation of ethanol and carbon dioxide 

from sugar using a cell-free extract of yeast, thereby providing the foundation for the 

development of modern biochemistry. Yeast has been used as a convenient experimental 

organism in many subsequent investigations. The zymologist, A. H. Rose, proposed in 

the introduction to the second volume of the first edition of The Yeasts (Rose and 

Harrison, 1971) the initiation of `Project Y'. This suggested that yeast be used as a model 

eukaryotic organism in an integrated approach to the study of cell biology. This challenge 

has been taken up and the academic literature devoted to yeast in general and 

Saccharomyces cerevisiae in particular is now immense. Many of the discoveries in cell 

biology, physiology, biochemistry and genetics were made using yeast cells. Probably, S. 

cerevisiae is the most extensively studied cell. This has culminated in the sequencing of 

the entire genome of S. cerevisiae, the first species for which this has been accomplished 

(Goffeau et al., 1996). 


11.2 Taxonomy 

Taxonomy is the science of the classification of organisms. Using criteria such as 

morphology, life cycle, immunological properties, biochemical capabilities and genetic 

analysis, organisms are grouped into hierarchies of relatedness and difference. Systems of 

taxonomy indicate functional and evolutionary relationships between groups of 

organisms and they provide a framework for identifying unknown types. Taxonomy 

has practical importance in brewing. It allows the identification of proprietary yeast 

strains and the ability to distinguish these from contaminants such as wild yeasts. 

Each group is termed a taxon. In descending order of hierarchy the main taxonomic 

groups are kingdom, division, class, order, family, genus, species and strain. Of these, 

the last three are of most practical interest. A genus represents a group of organisms, 

which are closely related in evolutionary terms. Usually this is accompanied by 

structural and functional similarities. Organisms are grouped into species usually based 

on the ability to interbreed. In the case of yeast, many of which have no sexual cycle, this 

definition is of limited value. Where organisms are restricted to asexual reproduction, 

placement within a species has to be based on other criteria, the most reliable being that 

of similarity of genotype. Strains are clones derived from a single parental cell. For 

example, in the case of yeast a cell line propagated from a single colony of a pure 

culture. In this case, it as assumed that the colony was formed from asexual reproduction 

of a single parental cell. 

Organisms are described using a binomial nomenclature of genus name followed by 

species name. Both are Latin terms and by convention in typescript are italicized, the 

genus name being capitalized and the species name in lower case. After introducing the 

complete binomial name, the genus name may be abbreviated thereafter. Thus, brewing 

yeasts are classified as Saccharomyces cerevisiae (S. cerevisiae). Bionomial names may 

be descriptive of the appearance of the organism, its mode of growth, habitat or a Latin 

derivation of the name of the discoverer. The genus name, Saccharomyces translates as 

`sugar fungus', referring to the habitat in which the organism is usually to be found and 

the fact that it is a member of the Fungi. The species name cerevisiae derives from the 

Latin for beer as in ceres (ˆ grain) and vise (ˆ strength). 

Genus and species names are subject to stringent rules. In the case of yeast, these fall 

under the remit of the International Code of Botanical Nomenclature (Greuter et al., 

1996). Strain names are chosen by the discoverer and conventions are comparatively lax. 

Commonly, they are named using codes, which are combinations of letters and numbers. 

Many are proprietary strains owned and jealously guarded by individual companies. The 

ability to identify individual strains and maintain them as pure cultures is of the utmost 

importance to those who use yeasts in industrial applications. Many thousands of 

industrial strains of S. cerevisiae are in existence. In this regard, the finer points of 

taxonomy are of somewhat academic interest only. Commonly, systems of nomenclature 

are used that are now not recognized by `classical' taxonomists, however, they are 

retained because they are still of practical value. For example, see the discussion later in 

this section regarding the taxonomy of ale and lager brewing strains. 

`Yeast' and Saccharomyces are not synonymous terms. This is perhaps understandable, 

bearing in mind the economic importance of Saccharomyces strains in brewing 

and baking and numerous papers reporting work in which S. cerevisiae has been used as a 

type eukaryotic cell. In fact, approximately 100 genera of yeast encompassing 700 

species have been described (Kurtzman and Fell, 1998). Undoubtedly, many more remain 

unrecognized. Thus, some 70,000 species of fungi are currently recognized. It has been 


estimated that approximately 1.5 million species of fungi may exist (Hawksworth, 1991). 

Undoubtedly many of these will be classified within the yeast group. 

Kurtzman and Fell (1998) define yeasts as being fungi with vegetative states that 

reproduce by budding or fission resulting in growth that is predominantly in the form of 

single cells. Yeasts do not produce sexual states within or upon aspecialized fruiting 

body. This definition is relatively imprecise since many fungi are dimorphic. During 

certain phases in their life cycles, such fungi adopt ayeast-like unicellular form and at 

others they take on afilamentous hyphal habit and develop into amycelium. 

Brewing yeast strains are ascomycetous types classified within the genus 

Saccharomyces. The precise taxonomy of the fungi in general and the Saccharomyces 

in particular is still subject to debate and continual revision. Acurrent version is given in 

Table11.1.Atpresent,thegenusSaccharomycesisdividedinto14species.Thenamesof 

these species together with akey for their differentiation are shown in Table 11.2. The 

relationships between the 14 species assigned to the genus Saccharomyces have been the 

subject of intensive investigation. Based on homologies between ribosomal RNA and the 

mitochondrial genome it has been demonstrated that they can be arranged into smaller 

sub-groups (Vaughn-Martini and Martini, 1993; Piskur et al., 1998; Montrocher et al., 

1998). The Saccharomyces sensuo lato group includes S. dairensis, S. castelli, S. exiguus, 

S. servazzii, S. unisporus and possibly S. kluyveri. This grouping is considered to contain 

species that are relatively heterogeneous. 

The Saccharomyces sensuo stricto group contains S. cerevisiae, S. pastorianus, S. 

bayanus and S. paradoxus. As the nomenclature suggests these are much more closely 

related. Based on biochemical differences this group can be sub-divided into two subsets. 

S. cerevisiae and S. paradoxus can utilise melibiose, assimilate fructose via a 

facilitated transport mechanism and have a maximum growth temperature close to 37 ëC. 

Conversely, S. bayanus and S. pastorianus cannot utilize melibiose, have active transport 

systems for fructose uptake and are not capable of growth above 34 ëC. 

All yeast species in the Saccharomyces sensuo stricto group, with the exception of S. 

paradoxus, are exploited commercially for the production of ethanol or in baking. It is 

suggested that these commercial yeast strains actually arose via selective processes in 

industrial situations. S. paradoxus alone is known to occur in natural habitats, whereas the 

others are rarely found so (Vaughn-Martini and Martini, 1993). S. cerevisiae is used for 

baking as well as brewing and in winemaking. S. bayanus strains are utilized solely for 

enological purposes, whereas S. pastorianus includes those strains originally classified as 

S. carlsbergensis and used as bottom-fermenting lager yeasts. It is possible, therefore that 

S. bayanus arose because of its ability to ferment at the relatively low temperatures of 

wine fermentations and to withstand high ethanol concentrations. Similarly, S. 

pastorianus strains were selected for their ability to bottom-crop in low-temperature 

lager fermentations. 

The phylogenetic relationships between the Saccharomyces sensuo stricto highlight 

the undoubted differences between the individual species and support the arguments 

regarding their origins. Strains of the species S. pastorianus show a high degree of DNA 

homology with those of both S. cerevisiae and S. bayanus. Conversely, the strains of the 

latter two species show little similarity with each other. The genome of S. pastorianus 

strains is considerably bigger than that of S. cerevisiae and S. bayanus. It is proposed, 

therefore, that S. pastorianus is a hybrid species derived from S. cerevisiae and either S. 

bayanus or a closely related species, S. monacensis (Wolfe and Shields, 1997). The latter 

is now classified as S. bayanus but was originally considered to be S. pastorianus. 

Seemingly, the proportion of DNA contributed by S. bayanus to the S. pastorianus 


Table 11.1 Classification of Saccharomyces cerevisiae 

Taxon Name Comments 

Kingdom Fungi 

Phylum Ascomycotina Teliomorphic forms characterized by formation of ascospores enclosed within ascus 

Sub-phylum Saccharomycotina (syn. Hemiascomycotina) 

Class Saccharomycetes (syn. Hemiascomycetes) Single ascus not enclosed in ascocarp developing directly from zygotes 

Order Saccharomycetales (syn. Endomycetales) Yeast-like cells, rarely developing hyphae 

Family Saccharomycetaceae 

Genus Saccharomyces Globose, ellipsoidal or cylindroidal cells. Vegetative reproduction by multilateral budding. 

Pseudohyphae may be formed but hyphae are not septate. The vegetative form is predominantly 

diploid, or of higher ploidy. Diploid ascopores may be formed that are globose to short ellipsoidal 

with a smooth wall. There are usually 1±4 ascopores per ascus 

Type species S. cerevisiae 


Table 11.2 Key to the species of Saccharomyces (taken from Kurtzman and Fell, 1998) 

1 a. Maximum growth temperature above 30 ëC ! 3 

b. Growth absent above 30 ëC ! 2 

2 a. Sucrose, raffinose and trehalose fermented S. barnettii 

b. Sucrose, raffinose and trehalose not fermented S. rosini 

3 a. Ethylamine-HCl assimilated ! 4 

b. Ethylamine-HCl not assimilated ! 6 

4 a. Growth in the presence of 1000 ppm cycloheximide S. unisporus 

b. No growth in the presence of 1000 ppm cycloheximide ! 5 

5 a. Maltose, raffinose and ethanol assimilated S. kluyveri 

b. Maltose, raffinose and ethanol not assimilated S. spencerorum 

6 a. Maltose assimilated ! 7 

b. Maltose not assimilated ! 10 

7 a. Growth in vitamin-free medium S. bayanus 

b. No growth in vitamin-free medium ! 8 

8 a. D-mannitol assimilated, maximum growth temperature 37 ëC or greater S. paradoxus 

b. D-mannitol not assimilated, maximum growth temperature less than ! 9 

37 ëC or variable at 37 ëC 

9 a. Active transport mechanism for fructose present; maximum growth S. pastorianus 

temperature 34 ëC or lower 

b. Active transport mechanism for fructose not present; maximum S. cerevisiae 

growth temperature variable 

10 a. Sucrose, raffinose and trehalose fermented S. exiguus 

b. Sucrose, raffinose and trehalose not fermented ! 11 

11 a. Growth in the presence of 1000 ppm cycloheximide S. servazzii 

b. No growth in the presence of 1000 ppm cycloheximide ! 12 

12 a. D-ribose normally assimilated; 8±10 chromosomes 600±3000 kilobases S. castellii 

b. D-ribose not assimilated, mostly single highly refringent ascospores S. transvaalensis 

on acetate agar; 8 chromosomes 400±2200 kilobases 

c. D-ribose normally not assimilated, 7±9 chromosomes 750±3000 kilobases S. dairiensis 

hybrids is greater than that made by S. cerevisiae. Nevertheless, chromosomes identical 

to those from both parents have been found co-existing in strains of S. pastorianus 

(Tamai et al., 1998). 

The taxonomic history of lager brewing yeast strains is chequered. The original 

descriptor for lager strains, S. carlsbergensis changed when these were re-christened as S. 

uvarum. Later the position became more clouded when the latter were assigned to the 

species S. cerevisiae. As discussed already, bottom-fermenting lager yeasts are again 

considered distinct from S. cerevisiae brewing strains and have been given the name S. 

pastorianus. This confirms what brewers have already decided, based on experience and 

observation, that ale and lager yeasts are different (Quain, 1986). 

11.3 Yeast ecology 

Yeasts are predominantly saprophytes and are widely distributed in nature where they are 

found in both terrestrial and aquatic habitats (Phaff and Starmer, 1987). In nature, yeasts 


are primarily associated with higher plants, although the effects of rain and plant death 

inevitably means that soils act as areservoir in which they can survive and be passed on 

to other hosts. Yeasts are rarely plant pathogens, instead they are commonly found on 

damaged fruits, in flowers and in exudates associated with wounds. The transfer of yeasts 

between plants is most often accomplished by the intermediary of insect vectors. 

Natural populations of yeasts co-exist and compete with themselves and with other 

microbial species. Many yeast species are found in specialized plant habitats, which 

reflect their biochemical capabilities. Contrary to expectation, non-fermentative 

obligately aerobic yeast types are the most common. Typically, they occupy niches 

that provide a particular oxidizable substrate that they are capable of assimilating. 

Fermentativeyeastsareabletotakeadvantageofhabitatswherethereisasourceofsugar 

but no oxygen. Since such yeasts are facultative anaerobes the result of their own 

metabolic activity would be to remove oxygen from aerobic environments. They would 

then be able to continue to grow under conditions of anaerobiosis, whereas purely 

oxidative yeasts could not. In aquatic habitats containing asource of fermentable sugar 

the result would be that aerobic yeast would be restricted to the surface layers, possibly 

resulting in the formation of apellicle. The population of fermentative yeasts would be 

capable of growth throughout the body of the liquid. 

These observations are paralleled in brewing, for example, the growth of brewing 

yeast in afermenter and the effects of some spoilage organisms in beer exposed to air. 

Some of the associations of yeasts and other micro-organisms have elements of 

symbiosis. For example, acetic acid bacteria are often found growing on the ethanol 

produced as aresult of the metabolism of sugars by yeast. The bacteria remove ethanol, 

which can be toxic or even fatal to the yeast. The resultant acetic acid reduces the pH of 

the medium and inhibits the growth of other less acid-tolerant species. Some yeasts are 

able to occupy particular niches because of an ability to tolerate otherwise toxic products 

ofthegrowthofothermicro-organisms.Forexample,naturallyoccurringantibioticssuch 

as cycloheximide produced by Streptomyces griseus, inhibit the growth of many yeast 

species but not others. 

Other yeast types produce metabolites that are toxic to other potential yeast 

competitors. Some strains of Saccharomyces cerevisie and representatives of the genera 

Candida, Crytpococcus, Debaromyces, Hansenula, Kluyveromyces, Pichia and Torulopsis 

produce so-called killer factors, which are fatal to other susceptible yeast species 

(Young, 1987). The killer factors, several distinct types of which are produced by 

individual yeast genera, are protein or glycoprotein in nature and are coded for either by 

chromosomal genes or in some cases by the RNA genomes of mycoviruses. They appear 

to bind to chitin in the wall of susceptible cells and cause death by destroying the 

transmembrane potential. 

Yeasts are common contaminants of fruits and are potential spoilage organisms in 

extracted fruit juices, pureÂes and concentrates. For example, in one study (Arias et al., 

2002) using orange juices some 99 different yeast strains, representing 11 genera were 

isolated. Some yeast species are capable of growth in media with very low water activity. 

These so-called osmotolerant species (Section 12.3.1) can cause spoilage of products 

such as bulk sugar syrups, particularly if the storage conditions allow condensation to 

form on the surface of the liquid. 

Yeast strains used in industrial fermentations were originally isolated from nature. In 

rare cases, the natural microbial flora of the building in which the fermentation is 

conducted is used as the source of the inoculum, for example, in the production of 

traditional Belgian Lambic beer. In the modern brewery process, great care is taken to 


ensure that contamination of worts and beers with foreign `wild' yeast does not occur. 

Production strains of brewing yeasts are maintained as pure cultures and introduced into 

the brewery via asystem of propagation. Such strains have been selected based on their 

possession of particular sets of desirable properties. In consequence, as discussed in the 

previous section of this chapter, brewing strains appear to be distinct from those found in 

nature. 

11.4 Cellular composition 

Yeast cells contain approximately 80% water. Thus, inpressed brewers'yeast the ratio of 

wet to dry weight is roughly 5:1. Predictably, the most abundant element is carbon, 

which accounts for just under 50% of the dry weight. Other major elemental components 

are oxygen (30±35%), nitrogen (5%), hydrogen (5%) and phosphorus (1%). The total 

mineral content of yeast is approximately 5±10% of the cell dry weight. This fraction 

comprises amultitude of trace elements. The composition of some of these for three 

brewing and one bakers' yeast strain is shown in Table 11.3. The most abundant classes 

of macromolecules are proteins (40±45% cell dry weight), carbohydrates (30±35%), 

nucleic acids (6±8%) and lipids (4±5%). The precise composition of each class of 

macromolecules within agiven cell varies as afunction of physiological condition and 

phase in growth cycle. There is considerable variation between different yeast species. 

For these reasons, it is not possible to provide other than the generalized composition 

given above. The pathways resulting in the formation and dissimilation of the major 

classes of macromolecules are discussed in Chapter 12. 

Table 11.3 Trace elemental composition of yeast ( g dry wt. yeast) 

Bakers Brewers Brewers Brewers 

(Reed and Nagodarithana, 1991) (Eddy, 1958) (Eddy, 1958) (Eddy, 1958) 

Aluminium ± 3.0 2.0 1.0 

Calcium 0.75 ± ± ± 

Chromium 2.2 37.0 104.0 34.0 

Copper 8.0 ± ± ± 

Iron 20.0 17.0 104.0 25.0 

Lead ± 2.0 14.0 100.0 

Lithium 0.17 ± ± ± 

Magnesium 1.65 ± ± ± 

Manganese 8.0 4.0 5.0 11.0 

Molybdenum 0.04 0.1 0.04 2.7 

Nickel 3.0 3.0 4.0 3.0 

Phosphorus 13.0 ± ± ± 

Potassium 21.0 ± ± ± 

Selenium 5.0 ± ± ± 

Silicon 30.0 ± ± ± 

Sodium 0.12 ± ± ± 

Sulphur 3.9 ± ± ± 

Tin 3.0 3.0 > 100 3.0 

Vanadium 0.04 ± ± ± 

Zinc 170.0 ± ± ± 


11.5 Yeast morphology 

Individual yeast cells are not visible to the human naked eye and they become evident 

only when proliferation produces amass of many millions of cells. When this occurs, 

yeastcellstakeontheappearanceofsurfacepellicles,sedimentsorhazesonorwithinthe 

bodyofliquids.Saccharomyces yeast enmasse,isoff-white,grey orbeigeinappearance. 

Commonly, yeast biomass is stained by components of the growth medium adhering to 

theyeast cellwall.Brewingyeaststrainsthatarenon-flocculentincharacterformsmooth 

creamy slurries in beer. Such yeasts are termed `powdery'. Conversely, flocculent strains 

formslurrieswithadistinctgranularappearanceinwhichtheyeastreadilyseparatesfrom 

the beer. 

When ayeast cellistransferred toalaboratory nutrient mediumsolidifiedwith agar or 

gelatin subsequent proliferation results in the formation of aroughly circular mass of 

cells termed acolony.Thesize andshape of coloniesvaries with the yeastspecies, nature 

of the growth medium, the solidification agent and the conditions under which the plates 

are incubated. Providing these parameters are defined some individual yeast strains give 

rise to colonies that have acharacteristic morphology. This has been used as amethod of 

yeast differentiation (Section 13.9). 

The size and shape of cells and the patterns of vegetative propagation are 

characteristic of individual yeast species and may be used as aids to identification 

(Fig. 11.1). A description of cells of S. cerevisiae provided by Lodder (1970) is 

`spheroidal, subglobose, ovoid, ellipsoidal or cyclindrical to elongate, occurring singly or 

in pairs, occasionally in short chains or clusters'. The ratio of the long and short axes of 

cells of S. cerevisiae averages approximately 1.4:1, although some cells are longer and 

thinner than this. Most fall within the range of 2.5±4.5 m and 10.5±20 m along the 

short and long axes, respectively. Cell volumes range between 50 and 500 m 3 . 

Cell size is a characteristic of individual strains although there is considerable 

variation within strains depending on the phase of the growth cycle and cultural 

conditions. Thus, the mean cell size increases with increase in incubation temperature, 

within the normal range for any given strain. The cell wall of yeast is relatively flexible. 

This accommodates transient variations in cell volume in response to sudden shifts in 

osmotic pressure of the suspending medium. For example, when yeast slurries suspended 

in beer are transferred to wort there is an increase in cell size, which persists for the first 

few hours and precedes the onset of budding (Quain, 1988). As cells proceed through the 

growth cycle, there are changes in cell size and density. During the budding phase, the 

cell volume decreases by approximately a third. In the intervals between budding, there is 

an increase in cell density due to a loss of water (Baldwin and Kubitschek, 1984). Cell 

size of brewing yeast has been reported to decrease during the period when yeast is stored 

between cropping and re-pitching (Cahill et al., 1996). During 14 days storage at 4 ëC, the 

mean cell volume of an ale yeast fell from 302 to 244 m 3 and a lager strain from 208 to 

194 m 3 , ascribed to reduction in biomass due to glycogen turnover. 

The mean cell size increases with generational age. For any given strain there is a 

rough correlation between number of bud scars and mean cell size. Using a diploid lager 

yeast strain Barker and Smart (1996) reported that the mean cell volume of virgin 

daughter cells was approximately 150 m 3 . Cells, which had undergone around 20 rounds 

of budding and were approaching the end of their life spans, had mean cell volumes of 

approximately 850 m 3 . Brewing yeast cells tend to be bigger than haploid laboratory 

strains of S. cerevisiae. This is a consequence of the fact that the former tend to be 

polyploid/aneuploid. Galitski et al., (1999) measured the mean cell volume of haploid, 


(a) (b) (c) 

(d) (e) (f) 

Fig. 11.1 Drawings of vegetative cells of (a) Saccharomyces cerevisiae (multilateral budding); 

(b) Schizosaccharomyces pombe (vegetative reproduction by binary fission); (c) Nadsonia sp. 

(apiculate (lemon) shaped cells undergoing bipolar budding); (d) Sterigmatomyces halophilus 

(conidiospores borne on short stalks); (e) Trigonopsis variablis (triangular shaped cells buds borne 

in angles); (f) Oosporidium margaritiferum (multilateral budding in which cells remain attached 

and form long chains). 

diploid,triploidandtertraploidconstructsofastrainofS.cerevisiae.Thesewerefoundto 

be 72, 111, 152 and 289 m 3 ,respectively. 

11.6 Yeast cytology 

Yeast cell ultrastructure became amenable to detailed study with the advent of electron 

microscopy. Such techniques include scanning electron microscopy, which allows 

detailed examination of surface topology. Amore recent development, atomic force 

microscopy, performs asimilar function although in this case at the level of individual 

molecules (De Souza et al., 1996). Confocal microscopy provides three-dimensional 

imaging of cells by measuring fluorescence light intensity produced by alaser-scanning 

device (Bacallao and Stelzer, 1989). Transmission electron microscopy is suitable for 

producing images of very thin sections of cells. The study of individual cellular 

organelles was largely dependent on the development of techniques that allowed the 

controlled disruption of yeast cells and recovery of undamaged organelles (Lloyd and 

Cartledge, 1991). 

An idealized representation of asection through abudding yeast cell is shown in Fig. 

11.2. The diagram shows all the major organelles present that may occur in the cell. Not 

all of these are visible at any given time. In some non-brewing yeast genera, there is an 

external capsule, which is attached to the outside of the wall. The cell wall, plus capsule 

if present, plasma membrane and the intervening periplasmic space are collectively 

referred to as the cell envelope. Modifications to some organelles occur as the cell 

progresses through the cell cycle. Some changes occur in response to growth conditions. 


Plasma membrane 

Cytoplasm 

Tonoplast 

Ribosomes 

Volutin granule 

Mitochondrion 

Cell wall 

Endoplasmic reticulum 

Bud scar 

Vacuole 

Nucleolus 

Nucleus 

Mitochondrion 

migrating in 

emerging bud 

Periplast 

Golgi body 

Lipid granule 

Glycogen granule 

Birth scar 

Fig. 11.2 Diagrammatic representation of asection through atypical budding yeast cell. 

11.6.1 Cell wall 

The wall is the outermost layer of the cell. It is of rugged construction, typically between 

150 and 200nm in thickness and accounts for approximately 20% of the total cell dry 

weight (Smits et al., 1999). Cells that have undergone vegetative reproduction by 

budding bear characteristic circular bud scars. These mark the point on the wall at which 

the daughter cell was excised from the mother. The bud scar region is relatively rich in 

chitin.Thiscan beseen followingtreatmentwithfluorescent dyes suchascalcofluor.The 

point on the daughter cell wall, which corresponds to the bud scar on the mother cell, is 

termed the birth scar. Buds do not arise randomly on the cell surface but occur at specific 

locations. The patterns of bud scars are frequently characteristic of individual species. 

The process of budding is described in more detail in Section 11.7. 

The cell wall is approximately 90% carbohydrate, the remainder being protein. A 

diagrammatic representation of asection through the cell wall is shown in Figs 11.3±4. 

The most abundant carbohydrates are glucans, which make up around 30±50% of the 

total dry weight of the wall. Glucans are arranged into long fibrillar structures, which are 

joined together by -1,3 and -1,6 linkages. Most of the remaining cell wall 

carbohydrate is mannoprotein (Fig. 11.5). This is comprised of an inner core region of 

repeating -1,6linked mannose (4.8) residues with short side chains attached via -1,2 

and -1,3linkages. The inner core region is attached to an outer chain of 100±150 

mannose residues.Thisalsoconsistsofabackboneof -1,6linked mannose residue with 

side chains attached via -1,2linkages. The side chains are of mannobiose (M 2 1 M), 

mannotriose (M 2 1 M 2 1 M), mannotriose (M 2 1 M 3 1 M) and mannotetraose 

(M 2 1 M 2 1 M 3 1 M). The precise composition of these side chains varies between 

yeaststrains.Someofthesidechainscontainphosphodiesterlinkagesandtheseconferan 

overall negative charge to the cell envelope. The side chains of the mannose molecules 

are the sites of the receptors, which are implicated in yeast flocculation (Section 

11.6.1.1). The other end of the inner core region is attached to two N-acetyl glucosamine 

residues. One of these is attached to aprotein molecule via the carboxylic acid moiety of 

an aspartic acid residue. Attached to the protein via the hydroxyl groups of serine (4.45) 

and threonine (4.46) are short -1,2and -1,3chains of mannose residues. 


EXTERIOR 

Glycogen 

INTERIOR 

Chitin cross-linked 

to β-glucans 

β,1-3 glucans 

Mannoproteins 

covalently linked 

to glucans 

Mannoproteins 

anchored in 

membrane 

β,1-3 and β,1-6 

glucans 

Periplasmic space 

Periplasmic enzyme 

Plasma membrane 

Fig. 11.3 Diagrammatic representation of a section through the yeast cell envelope. 

(a) 

CH2OH 

H 

O 

H 

H 

OH HO 

HO OH 

H H 

H 

H 

HO OH 

NH·CO·CH3 

Chitin consists of a linear polymer of molecules of N-acetyl glucosamine linked by - 

1, 4 groups. It accounts for less than 5% of the dry weight of the wall in S. cerevisiae. 

Almost all is located within the bud scars although a small quantity is distributed 

throughout the rest of the wall. The fourth and also minor carbohydrate component of cell 

walls is glycogen. It is acid soluble and distinct from the alkali soluble pool, which 

functions as a reserve material. It has been demonstrated that the acid soluble fraction is 

structural and is linked to -, 1 3 glucans via the -1, 6 glucan side-chains (Arvindekar 

and Narayan, 2002). Most of the protein component of the wall is associated with 

mannose. This fraction confers immunological properties to the cell. Some of the proteins 

are surface receptors and others are enzymes. The latter are those responsible for cell wall 

biosynthesis and the initial metabolism of some nutrients (Fleet, 1991). 

The precise macromolecular structure of the cell wall remains uncertain. The glucan 

fibres are mainly located within the inner part of the wall. It is considered that the fraction 

of glucans attached by -1, 3 linkages form an interwoven network of fibrils responsible 

for conferring strength and flexibility to the wall. The -1, 6 linked glucans form a 

(b) 

CH2OH 

H 

O OH 

H H 

(11.1) Mannose, α-D-mannopyranose (11.2) β-D-N-Acetyl glucosamine 

Fig. 11.4 The structures of (a) -D-mannose and (b) -D-N-acetyl glucosamine. 


Asn 

Ser 

or ( thr ) 

Ser 

or ( thr ) 

Ser 

or ( thr ) 

Ser 

or ( thr ) 

1 4 

Gln Ac 

Protein 

1 M 

1 M 2 1 M 

1 M 2 1 M 2 

1 M 

M 1 2 

M 1 3 

M 1 

1 M 2 1 M 2 1 M 2 1 M 2 

Protein 

1 4 

Gln Ac 

1 6 

M3 

1 6 

M2 

M 1 3 

M 1 

M3 M2 M 

M 1 M 1 

(M2 M2 M2 M2)n 

P M 1 2 M 1 2 M 1 2 

M 1 3 M 1 3 M 1 3 

M 1 M 1 M 1 

Inner core Outer chain iso mannoses 

Fig. 11.5 Yeast phospho-mannoprotein. M. mannose; Gln-Ac, N-acetyl glucosamine; asn, 

asparagine; ser, serine; thr, threonine. 

connection between this fibrillar network and the mannoprotein, glycogen and chitin 

components. The mannoproteins are mainly situated towards the outside of the wall 

where they form a cross-linked layer covalently bonded to the glucans. The extent of 

cross-linking between the mannoproteins appears to determine the size of molecule that 

can pass through the wall. A second class of mannoproteins is attached to the plasma 

membrane and project across the periplasmic space through the glucan layer (Fleet, 

1991). These mannoproteins are implicated in flocculation and sexual agglutination. The 

fraction of chitin not located in bud scars is distributed throughout the body of the cell 

wall. Its function is unclear although it appears to have structural significance since 

mutants lacking it are sensitive to osmotic shock (Fleet, 1991). Chitin is the receptor for 

binding of killer toxin to yeast cells (Takita and Castilho-Valavacius, 1993). 

The cell wall has several functions. It forms a protective layer over the comparatively 

fragile plasma membrane. It has a degree of flexibility, which allows rapid fluctuations in 

cell volume in response to changes in the osmotic potential of the external medium. 

Conversely, it has sufficient mechanical strength to prevent lysis when cells are subject to 

hypo-osmotic shock. This rigidity is responsible for conferring characteristic shapes to 

individual cells. The generalized rigidity and targeted weakening, together with the 

motive force of turgor pressure provides the means for bud development. The cell wall is 

a repository for several enzymes and it ultimately limits the size of molecules that may 

pass into and out of the periplasm. The cell wall is important in determining interactions 

between cells and with the external medium. 

During fermentation some wort components, notably trub lipids and hop iso- -acids, 

bind to yeast cell walls. Hop iso- -acids dosage rates must be adjusted to allow for the 

proportion lost with the yeast crop. The impact of binding of trub components to yeast 


M 1 

M 1

cell walls is more difficult to assess. Potentially, the bound wort components will block 

receptor sites implicated in flocculation. Possibly, binding of trub components will 

change the overall cell surface charge and by implication affect fining behaviour. 

Isinglass finings are added to green beer to encourage yeast cells to aggregate together 

and forms flocs, which promotes sedimentation. The active component in isinglass is the 

positively charged protein collagen, which binds electrostatically to negatively charged 

yeast cells. Pre-loading of yeast cells with positively charged material interferes with the 

action of isinglass (Leather et al., 1997). 

The cell wall plays a role in determining whether or not yeast rises to the surface (top 

yeast) or sediments (bottom yeast) during wort fermentation. Differences in the two types 

of yeast can be demonstrated simply by placing a suspension of yeast in water in a test tube 

and shaking the mixture. Top yeasts form a surface pellicle at the interface between water 

and air. Bottom yeasts remain distributed throughout the water. Top fermenting yeasts form 

loose flocs during fermentation. These trap rising carbon dioxide bubbles and the mass of 

gas and yeast form a type of foam akin to a head on a beer. The cell walls of top fermenters 

are more hydrophobic than bottom yeasts. It is been demonstrated that hydrophobicity 

shows a negative correlation with the content of phosphate in the outer part of the cell wall 

(Mestdagh et al., 1990). Highly hydrophobic cells (low cell wall phosphate content) would 

tend to be attracted to bubbles and the surface of polar liquids such as fermenting wort. 

The phosphate content of the wall is the major determinant of another measure of the 

cell surface, that is zeta potential (Lawrence et al., 1989). This parameter is a measure of 

the surface charge of yeast cells. Apart from the phosphate content of the wall, it is also 

influenced by the pH of the surrounding medium. Zeta potential reportedly declines 

(becomes less negative) either during or at the end of fermentation (Iserentant, 1996). 

This would tend to reduce the electrostatic repulsion between individual cells and thereby 

favour the formation of flocs. 

11.6.1.1 Flocculation 

Flocculation is the reversible process by which some yeast cells adhere to each other to 

form aggregates. It is distinct from aggregates, which arise via budding and nonseparation 

of daughter cells. Flocculation is of enormous significance to brewing. The 

propensity of yeast to form flocs is an integral part of the process of separating the crop 

from green (immature) beer. Top fermenting types form flocs that rise to the surface of 

the fermenting vessel. The resultant yeast head can be removed by skimming or suction. 

Bottom fermenting yeast form flocs which settle into the base of the fermenting vessel, a 

process which is encouraged by chilling the green beer. The bottom crop can be removed 

from the fermenter before the beer is racked. The formation of flocs is an essential 

precursor of crop formation. Inadequate flocculation results in poor cropping such that 

there may be insufficient yeast for re-pitching and green beer with unacceptably high 

residual yeast counts. Conversely, if flocculation occurs too soon, fermentation may 

arrest because insufficient cells remain suspended in the fermenting wort. 

Flocculation is observed in strains from several genera. Conversely many strains, 

including several Saccharomyces brewing strains, do not flocculate to any great extent 

under any circumstances. Brewing strains possessing desirable flocculation characteristics 

will have been chosen by `natural selection' as being the most suitable for use with 

particular combinations of wort and fermenting vessels. Flocculation and flocculence are 

distinct. The latter is an inherent property that some strains possess. The former refers to 

the expression of flocculence in those strains capable of so doing. By inference, 

flocculation is not expressed under all conditions. Commonly, flocculation occurs only 


when sources of fermentable sugars are exhausted. It has been suggested (Iserentant, 

1996)thatundersuchstarvationconditionstheabilitytoformflocsmayrepresentastress 

response. Thus,flocs provide asheltered environment where thechanceofsurvival ofthe 

population is enhanced. Disaggregation of flocs occurs if the cells are again exposed to a 

source of fermentable sugar. In this case, the re-adoption of asingle cell mode affords 

unimpeded opportunity to utilize the supply of sugar. 

The precise mechanism by which flocculation occurs is controversial and there is no 

consensus that there is asingle mechanism that applies to all yeast strains. The onset of 

flocculation is observed in laboratory cultures that have just entered the stationary phase 

of growth. Similarly, in brewing, flocculation occurs towards the end of primary 

fermentation. Nevertheless, exponentially growing yeast cells can be made to flocculate 

providing they are removed from the growth medium and washed and suspended in 

water, supplemented with Ca 2+ ions. Cells must come into contact with each other for 

flocculation to occur, hence the surprising observation that flocculation and the vigour of 

mechanical agitation are positively correlated. Thus, in well-stirred systems there is a 

high probability that cells will contact each other and once formed, flocs are relatively 

stable structures. 

Flocculation occurs because of interactions between surface proteins on one cell and 

carbohydrate receptors on another cell (Miki et al., 1982). It has been demonstrated that 

flocculation can be inhibited by pre-treatment of cells with agents that block these 

interactions. This has allowed classification of yeasts based on the nature of agents that 

inhibit flocculation. The evidence suggests that the protein component is alectin since an 

excess of related lectins, such as Concanavilin A, abolishes flocculation. Similarly, 

simplesugars,whichalsobindtolectin-likeproteins,alsopreventorreverseflocculation. 

Three phenotypes have been recognized based on which sugars inhibit flocculation 

(Stratford and Assinder, 1991; Dengis et al., 1995). Flo1 types do not flocculate in the 

presence of mannose, whereas mannose, sucrose, glucose and maltose abolish 

flocculation of NewFlo types. The MI phenotype flocculates in the presence of both 

mannose and sucrose but not in the absence of ethanol (Table 11.4). The MI phenotype is 

totally distinct from the other two groupings. In these cells flocculation occurs via direct 

(non-lectin like) protein ±protein interaction. These strains are top-fermenters and have 

highly hydrophobic cell envelopes. Possibly the latter promotes both the formation of 

flocs and encourages formation of ayeast head. 

Flo1 and NewFlo types all use interactions between lectin-like proteins and cell 

surface mannans. The groups differ in the nature of the lectins. These can be 

differentiated based on differences in patterns of proteolytic digestion and response to 

pH. Synthesis of particular lectins is dependent on the possession of the relevant genes. 

Geneticdifferencesunderpintheflocculationphenotypicclassification.BothNewFloand 

Flo1 phenotypes use common carbohydrate receptors. Based on studies with mutants 

Stratford (1992) has demonstrated that the receptors are the side chains of the outer 

mannose chain of cell wall mannoproteins (Fig. 11.5). NewFlo and Flo1 types have an 

obligate requirement for Ca 2+ ions for flocculation to occur. Absence of this ion, or the 

presence of chelating agents prevent flocculation. The role of calcium is probably that of 

ensuring that the lectin-like protein is in the correct configuration for binding to the 

mannose receptors. 

The stability of flocs is proportional to the number of interactions between individual 

cells and the number of potential binding sites, both protein and mannose per unit area of 

cell wall will be influential. In addition, spatial considerations must play a role in 

allowing cells to pack together and make the intimate contacts necessary for interactions 


Table 11.4 Classification of yeast flocculence phenotypes (Straford, 1994) 

Class Phenotype Inhibitors Requirement Comments 

for Ca 2+ 

Flo 1 Heavily flocculent Mannose Yes Mannose absent from 

throughout fermentation wort 

New Flo Most brewing strains Mannose, sucrose Yes Flocculation at end of 

glucose, maltose primary fermentation 

MI Heavily flocculent and Flocculation not No Cells require presence 

top-fermenting inhibited by sugars of ethanol for 

flocculation to occur 

tooccur.Thissuggeststhatelectrostaticandhydrophobiceffectsmightalsobeimportant. 

In brewing, the onset of flocculation observed at the end of primary fermentation is 

explainedbecausetheinhibitoryeffectofsugarsisalleviatedbytheirexhaustionfromthe 

wort. Similarly, cells disaggregate at the start of fermentation when yeast suspended in 

beer is pitched into fresh sugar-containing wort. 

The genetics of flocculation is complex and still not fully characterized (for areview 

see Jin and Speers, 1998). Anumber of genes have been identified, termed FLO, which 

are reportedly implicated in flocculation. It is assumed that some of these encode for the 

lectin-like proteins. The possession of genes producing different lectin-like proteins 

presumably underpins the NewFlo and Flo1 phenotypes. Strains that do not possess any 

of these genes are not flocculent under any circumstances. Thus, there is evidence that a 

gene, termed FLO1 encodes for acell surface protein, which has been implicated in 

flocculation. Transfer of this gene from aflocculent yeast strain to anon-floccculent type 

is accompanied by the acquisition of aflocculent phenotype (Teunissen and Steensma, 

1995). 

11.6.2 The periplasm 

Theperiplasm is thespace betweenthe cellwall and theplasma membrane. Although not 

an organelle as such it is avoid that must be traversed by all in-coming nutrients and outgoing 

metabolic by-products. It is the location of several enzymes including invertase, 

acid phosphatase, melibiase and various binding proteins. These are produced 

extracellularly, that is exterior to the plasma membrane, but they are retained by virtue 

of being too large to pass through the cell wall. The periplasm is not acontinuous void 

sincesomecomponents ofthe cellwall areanchoredintheplasma membrane (Fig.11.3). 

Possibly (Arnold, 1991) the retention of enzymes that are capable of hydrolysing 

otherwise non-assimilable nutrients affords a competitive advantage compared to 

organisms that excrete free extracellular enzymes. The protein content of the periplasm 

may be sufficiently high to make the fluid gel-like. Thus, the periplasm may also function 

as a protective layer between the cell wall and plasma membrane. 

11.6.3 The plasma membrane 

Plasma membranes enclose the cytoplasm and form the inner barrier between the cell 

wall and periplasm. The yeast plasma membrane resembles that of other eukaryotic cells. 

It consists of roughly equal quantities of lipid and protein. The proteins are diverse and 

mainly functional as opposed to structural. They include the enzymes responsible for cell 


wall synthesis, those catalysing the uptake of nutrients, the ATPase responsible for 

maintaining the proton motive force and possibly receptors of cellular signalling systems. 

Membrane lipids are mainly the phospholipids, phosphatidyinositol (4.122), phosphatidylserine 

(4.121), phosphatidylcholine (4.118) and phosphatidylethanolamine (4.120). In 

addition, smaller quantities of sterols are present, the most abundant being ergosterol. 

The precise architecture of the plasma membrane remains uncertain. The fluid mosaic 

model (Singer and Nicholson, 1972) describes membranes as being bilayers of 

phosplipids in which the hydrophobic groups are turned inwards to face each other. 

The hydrophilic moieties are turned outwards and are in contact with the aqueous 

environment in the periplasm and cytoplasm, respectively. Proteins and sterols are 

located within the phospholipid bilayer. Sterols have polar hydroxy groups and a 

hydrophobic skeleton. This allows them to orientate themselves in a perpendicular plane 

between the hydrophobic chains of the phospholipids. 

The plasma membrane forms the barrier between the cytoplasm and the external 

environment. It prevents free diffusion of solutes and provides a support in which specific 

carrier proteins catalyse the selective uptake and excretion of metabolites. It provides a 

framework whose structure allows the generation of proton and ion gradients necessary 

for the generation of energy that drives many uptake reactions. An essential facet of 

cellular function is the ability to detect and respond to external stimuli. Receptors of 

cellular signalling systems are conveniently situated within the plasma membrane. The 

membrane provides a site where enzymes involved in various cellular synthetic pathways 

can be located in a manner that favours spatial arrangement and function. 

11.6.4 The cytoplasm 

The cytoplasm is that portion of the cell enclosed by the plasmalemma and excluding 

other membrane bound organelles. It is an aqueous colloidal liquid containing a multitude 

of metabolites. The cytoplasm of yeast is particularly rich in RNA. It is acidic, typically 

about pH 5.2, although localized metabolic activity may produce micro-environments of 

greater or lesser acidity. The enzymes of several major metabolic pathways are located 

within the cytoplasm, for example, glycolysis and fatty acid synthesis. Enzymes such as 

those of glycolysis are described as being `soluble' because in cell-free extracts the 

enzymes can be found in the supernatant when all membranous material has been 

removed by centrifugation. In fact, many of these enzymes are not randomly dispersed 

throughout the cytoplasm. Instead, they are arranged in spatial configurations that aid 

ordered activity, possibly in loose associations with intracellular membranes. 

The cytoplasm contains a number of inclusions. Glycogen accumulates under 

appropriate conditions and appears in the cytoplasm as small granules that can be stained 

purple with iodine. Lipid particles become visible in the cytoplasm during aerobic growth 

when there is a plentiful supply of carbon. The particles apparently contain a hydrophobic 

core of triacylglycerol and steryl esters surrounded by a membrane consisting of 

phospholipid and protein (Leber et al., 1994). Probably lipid granules are temporary 

storage structures from which sterols may be transported to growing membranes and 

triacylglycerols withdrawn in times of need. 

Ribosomes are cytoplasmic organelles that contain high concentrations of RNA and 

some protein. Their role is to assemble proteins from àctivated' amino acids sequenced 

in response to the code present in molecules of messenger RNA. Ribosomes are found 

throughout the cytoplasm either borne freely or often associated with the outer 

membranes of mitochondria, the endoplasmic reticulum and the outer nuclear envelope. 


Commonly, several ribosomes are associated together joined by a strand of messenger 

RNA in structures termed polysomes. 

11.6.5 Vacuoles and intracellular membrane systems 

Yeasts, like other eukaryotes, contain extensive and dynamic internal membranous 

systems. These provide a method for partitioning metabolic pathways and pools of 

metabolites. They are also involved in the transport of metabolites both within the cell 

and to and from the plasma membrane. The most visible intracellular membranous 

system is the vacuolar system. Vacuoles are bodies bound by a membrane, the tonoplast. 

Their size and number fluctuates with physiological condition and stage in the cell cycle. 

When cells are growing in a balanced medium, as is the case during active primary 

fermentation, they may not be visible. Extensive vacuolation is associated with stress in 

yeast, especially starvation. Commonly large vacuoles become apparent in late 

fermentation or in stored pitching yeast. 

Vacuoles serve as temporary metabolite stores and provide the cell with a mechanism 

for controlling the concentration of metabolites in other cellular compartments. They are 

the site for the catabolism of macromolecules such as proteins. Vacuoles contain several 

proteinases, hence high concentrations of amino acids, especially basic types are also 

present. Thus, vacuoles function as a repository for nitrogen-containing metabolites. 

Their role as sites for proteolysis is consistent with the observation that they are most 

visible during starvation. The tonoplast, or vacuolar membrane, contains several amino 

acid transporters and is the site of a membrane-bound proton translocating ATP-ase. The 

latter is reportedly responsible for vacuolar acidification, which is an essential part of 

protein sorting (Klionsky et al., 1992). 

Vacuoles store quantities of inorganic phosphate, in the form of linear polymers of 

polyphosphate linked by high-energy phospho-anhydride bonds. Reportedly, polyphosphate 

is associated with S-adenosyl-L-methionine in vacuolar structures, termed volutin 

granules. These may be involved in the sequestration of basic amino acids (Schwenke, 

1991). 

The Golgi body comprises a dynamic series of stacked membranes and vesicles. It is 

part of the yeast secretory system and forms a link between the endoplasmic reticulum, 

the tonoplast and the plasma membrane. The endoplasmic reticulum consists of a 

branching network of membrane bound tubules. The Golgi body and the lumen of the 

endoplasmic reticulum are sites where proteins are sorted, modified and possibly 

assembled into complexes whilst being directed towards a chosen destination. Proteins, 

which are synthesized by ribosomes, may be transported across the membrane and into 

the endoplasmic reticulum. From here, they are directed towards the Golgi body via 

vesicles. In the Golgi body proteins are sorted and directed towards the sites where they 

are required. This may be within other intracellular organelles although most are sent to 

growing membranes. Many proteins, as evidenced by the numbers of visible vesicles, are 

sent to the area of membrane around the growing bud tip. Incorrectly folded proteins can 

be encapsulated in vesicles and returned to the endoplasmic reticulum. There they are 

repaired or degraded. 

Some proteins, encapsulated in vesicles, are transported across the plasma membrane and 

secreted into the periplasm in a process termed exocytosis. A reverse process, termed 

endocytosis in which proteins are imported into the cell also occurs. The cytoplasm also 

contains a system of microtubules and microfilaments. These constitute the cytoskeleton and 

are involved in the spatial organization of the cell especially during meiosis and mitosis. 


Two cytoplasmic membrane-bound bodies, peroxisomes and glyoxysomes are 

associated with particular physiological states. Peroxisomes contain catalase and 

oxidases required for the metabolism of specific carbon sources such as hydrocarbons 

and methanol. In glucose-grown cells of many yeast strains, including Saccharomyces 

spp, peroxisomes are barely evident. Transfer to amedium containing acarbon source 

suchasmethanolresultsintherapidbiogenesisofperoxisomesinthosestrainscapableof 

utilizing such substrates. Glyoxysomes are similar to peroxisomes but are rich in the 

enzymes of the glyoxylate cycle. Neither peroxisomes nor glyoxysomes are of relevance 

to brewing yeast under the conditions experienced during fermentation. 

11.6.6 Mitochondria 

Mitochondria are cytoplasmic organelles whose appearance and structure is much 

influenced by physiological state (Visser et al., 1995). Their principal role is energy 

generation via oxidative phosphorylation. They are most evident in cells growing 

aerobically under derepressed conditions (see Chapter 12). During oxidative derepressed 

growth, the mitochondrial volume (chondriome) accounts for some 12% of the total cell 

volume in S. cerevisiae. During anaerobic growth or under repressing conditions the 

chondriome is much reduced. Yeasts typically have a single or small number of large 

multi-branched mitochondria. Mitochondria have two membranes separated by an 

intermembrane space. The internal membrane has projections (cristae) that project into 

the internal matrix. Mitochondria contain a self-replicating genome, which codes for 

around 5% of all mitochondrial proteins. The remaining mitochondrial proteins are 

encoded by the nuclear genome and therefore biogenesis of these organelles requires coordinated 

expression of both sets of genes. 

Mitochondria are the site of the electron transport chain and oxidative phosphorylation 

resulting in the generation of ATP. In addition, there are enzyme systems for transporting 

many metabolites both into and out of the mitochondria, not least ATP. Several other 

enzyme systems unrelated to energy transduction are also located within this organelle 

including those of the oxidative tricarboxylic acid cycle, pyruvate dehydrogenase 

complex, several amino acid biosynthetic enzymes, the Mn-linked superoxide dismutase 

and possibly some of the sterol biosynthetic pathway. Under the repressing and anaerobic 

conditions of brewery fermentations, mitochondria do not contain the enzymes systems 

associated with oxidative metabolism. The genes coding for the proteins of the electron 

transport chain and several enzymes of the tricarboxylic acid are not expressed. The 

organelles remain in a partially undifferentiated state, termed promitochondria. These are 

small, difficult to visualize and internally few or no cristae are present. The chondriome 

is reduced to approximately 3±4% of the total cell volume (Stevens, 1977). 

Mitochondria do not have an energy-generating role during fermentation. Nevertheless, 

their presence is essential for normal fermentation behaviour because of the other 

metabolic functions they perform. O'Connor-Cox et al. (1996) have reviewed the 

functions of mitochondria, considered essential for normal fermentation performance. 

These include expression of flocculation, amino acid and diacetyl metabolism, sterol 

biosynthesis and physiological adaptation to stress. 

11.6.7 The nucleus 

The nucleus contains the greater part of the genetic material of the cell. It is roughly 

spherical, 1±2 micron ( m) in diameter and enclosed by a double membrane. The 


membrane is not continuous but contains several pores. The organelle is visible, with 

some difficulty, under the light microscope. In stationary phase cells it is often closely 

associated with vacuoles. The internal structure and appearance of the nucleus changes 

throughout the cell cycle. During interphase, acrescent shaped nucleolus is visible, 

associated with the nuclear membrane. The nucleolus is the site of rRNA transcription, 

some of the initial stages of mRNA processing and the assembly of ribosomal sub-units. 

The nuclear DNA of ahaploid cell is distributed between 16 linear chromosomes. The 

smallest chromosome (I) is 230kb in length, the largest (IV) 1532kb. The nuclei of most 

strains contain up to 100 copies of a2 mplasmid. This is acircular molecule of DNA 

and accounts for approximately 1% of the total nuclear DNA. 

Chromosomal DNA and proteins, both histones and non-histones are arranged 

together to form acomplex termed chromatin, so named because of its property of 

staining with basic dyes. In chromatin, the double helical DNA molecule is bound to a 

core of histone. Individual building blocks of DNA and histone are termed nucleosomes. 

Nucleosome units are coiled and condensed in hierarchical levels to form asupercoiled 

macromolecule. The protein component allows selective transcription of the genes on the 

chromosome. In comparison with higher eukaryotes, the chromatin molecules are loosely 

wound. This supports the view that the small genome is highly transcribed (Williamson, 

1991). 

The terminal regions of chromosomes are termed telomeres. These are involved in 

protecting the ends of chromosomes from degradation, maintaining the structural 

integrity of chromosomes, assisting in replication of DNA at the terminal region and 

possibly in the attachment of chromosomes to the nuclear membrane during meiosis. 

Chromosomes also contain regions termed centromeres. These structures consist of DNA 

andproteinandarethepartofthechromosomethatisattachedtothe spindlebodyduring 

mitotic division. Now the DNA component alone is called the centromere. The 

centromere is required to assemble alarge protein complex, termed the kinetochore. This 

structuremediatestheattachmentofchromosomestospindlemicrotubulesduringmitosis 

and ensures that each daughter cell receives acomplete set of chromosomes. Errors in 

chromosome segregation lead to alterations in chromosome copy number in daughter 

cells. This condition is termed aneuploidy. Interestingly, this is the common condition in 

the case of the genome of brewing yeast strains (Section 11.8.2). 

The spindle body acts as ascaffold, which mediates nuclear division and migration 

and segregation of daughter chromosomes during mitosis. It is adynamic structure that 

undergoes considerable changes during cellular budding. A complete spindle body 

consistsoftwospindlepolebodiesandanumberofmicrotubules(Fig.11.6)Spindlepole 

bodies are electron dense plugs, approximately 0.15 m in diameter, which are located 

within the nuclear membrane. Short cytoplasmic (astral) microtubules project from a 

dense amorphous layer situated just above the spindle pole body. Two, or three, further 

types of microtubule are attached to a second amorphous dense area situated between the 

spindle pole body and the inner surface of the nucleus. Short discontinuous microtubules 

project a short distance into the nucleus. These are of two types. Firstly those that interact 

with the arms of chromosomes and secondly, those which interact with the kinetochore. 

Continuous microtubules form a direct link across the diameter of the nucleus between 

adjacent spindle pole bodies. All microtubules are approximately 20 nm in diameter. 

They all consist of repeating units of two proteins termed - and -tubulin. 


Nuclear 

membrane 

Continuous 

microtubules 

11.7 Yeast cell cycle 

Spindle 

pole body 

Cytoplasmic 

microtubules 

Nuclear pore 

Chromatin 

Discontinuous 

microtubules 

Fig. 11.6 Diagrammatic representation of asection through the nucleus of ayeast cell in early G2 

phase. 

Proliferation of unicellular organisms involves coordination of the biochemical processes 

that together underpin growth of individual cells and those specific events that culminate 

in cellular multiplication. The combination of events that occur during the intervals 

between the separation of successive daughter cells is termed the cell cycle. It requires 

coordination of continuous events such as cellular growth with the discontinuous 

processes of DNA replication, mitosis and daughter cell excision. These processes in 

yeast have been subjected to intensive scrutiny. 

Progression through the cell cycle can be considered from three standpoints. Firstly, 

themorphologicalchangesthatoccurasamothercellgivesbirthtoadaughter.Secondly, 

the biochemical events that underpin the process of cellular proliferation. Thirdly, the 

molecular mechanisms that regulate the coordinated processes of cellular growth and 

multiplication. For convenience, the cell cycle is divided into anumber of phases (Fig. 

11.7). These are termed G1, which is the pre-synthetic gap phase; S, the synthetic phase 

during which DNA is replicated; G2, the post-synthetic gap phase, M, the mitotic phase 

and cytokinesis, the phase during which the daughter cell separates from the mother. 

During mitosis, the nucleus divides and duplicate pairs of chromosomes (chromatids) are 

sorted and segregated between the mother and daughter cell. In `classical' descriptions of 

mitosis, the process is delineated as aseries of distinct and named morphological states. 

In S. cerevisiae these steps are not clearly distinct. 

The gross morphology of mitotic cellular multiplication depends on the yeast type. In 

fission yeast, such as Schizosaccharomyces pombe, the mother cell divides to form two 

equal sized progeny. In the case of budding yeast such as S. cerevisiae, division is 

asymmetrical, the daughter cell being smaller than the mother. In some budding yeasts, 

the point where the bud forms is limited to specific points on the cell wall such as the 

poles. In others, including S. cerevisiae, buds may arise from any point on the cell wall. 

Nevertheless,budsdonotarisemorethanonceatthesamelocation.Thedeterminationof 

the precise site of bud initiation is aregulated process. Bud formation and separation is 

accompanied by nuclear division and segregation. This is adirectional process and thus, 

the cell must be polarized to ensure correct orientation of the dividing nucleus and the 


CYTOKINESIS 

Separation of 

mother and 

G1 

daughter Single spindle 

pole body 

Nuclear division 

ANAPHASE B 

Completion of spindle 

elongation in two phases 

(1 μm/min. and 0.3 μm/min.) 

START 

(Critical 

cell size 

achieved) 

developing bud. This is accomplished by the intermediary of the actin cytoskeleton and 

the spindle body. 

Bud emergence occurs in response to localized degradation of the cell wall. The site of 

bud initiation is marked by the development of a ring structure, containing filaments 

termed septins (Harold, 1995). This process is accompanied by the accumulation of 

cytoplasmic vesicles in the area directly below the developing bud. These contain 

enzymes and precursors of cell wall synthesis. The vesicles are transported to the site 

using the actin microtubules of the cytoskeleton. The motive force for vesicle transport is 

provided by a multi-subunit protein termed dynein. Some of the sub-units of dynein 

provide an anchor attaching the vesicle `cargo' to the microtubule. Other subunits provide 

the motor force to move the vesicle from the positive to the negative end of the 

microtubule. The process is energized by the dissipation of ATP. The vesicles convey 

their contents to the site of bud growth by fusing with the plasma membrane in a process 

of exocytosis. The partial degradation of the wall allows cell turgor pressure to push out 

the developing bud. 

S 

Bud emerges 

DNA replication 

Spindle pole body 

replicated and 

moves round 

nuclear envelope 

METAPHASE 

All chromatids 

attached and 

oscillating round 

ANAPHASE A 

equatorial 

Segregation of chromatids at (metaphase) plate 

metaphase plate. Spindle pole 

bodies begin to move apart 

PRE- 

ANAPHASE 

Bi-polar spindle at 

2μm stage aligned 

with bud 

PRO-METAPHASE 

Spindle captures chromatids 

by binding centromeres to 

discontinuous microtubules 

via kinetochore 

Fig. 11.7 Phases in the cell cycle of S. cerevisiae (including brewing yeast strains). 


Growth is restricted to the bud and there is no increase in the size of the mother cell 

during budding. Cell wall synthesis occurs outside the plasma membrane of the growing 

bud.Initially,cellwallgrowthisrestrictedtothetipofthedevelopingbud.Graduallycell 

wall synthesis spreads to the whole surface of the growing bud. At the end of mitosis, 

when the daughter nucleus and organelles such as mitochondria have divided and have 

migrated to the daughter cell, aseptum develops across the plane of the chitin ring. This 

process (cytokinesis) is completed when the daughter cell becomes detached from the 

mother.Theseptumonthemothercell iscomposedofchitinandconstitutesthebudscar. 

The corresponding point on the cell wall of the daughter cell is deficient in chitin and 

persists as the birth scar. 

Whilst bud emergence, growth and separation is occurring the nucleus and the 

enclosed chromosomes are duplicated and distributed between mother and daughter. In 

yeast, unlike mammalian cells, the nuclear membrane remains intact during the entire 

process. The morphological changesthatoccur duringreplicationofthe nucleusandtheir 

relation to bud formation and excision are illustrated in Fig. 11.7. Abrief description of 

the key events and associated nomenclature is also provided. The morphological changes 

associated with the cell cycle ofS. cerevisiae have been visualizedusing time-lapse highresolution 

digital enhanced differential interference contrast and multimode fluorescence 

microscopy (Shaw et al., 1998). These techniques provide direct visualization of the 

nucleus, the spindle and microtubules. Time lapse video footage of cells of S. cerevisiae 

progressing through the cell cycle may be found at http://www.molbiolcell.org. This 

elegant work shows how the discontinuous nuclear microtubules capture the duplicated 

chromosomes and segregate them at poles of the spindle body. The extension of the 

spindle body via elongation of the continuous microtubules is visualized. The manner in 

which cytoplasmic microtubules provide the motive power and orientation for movement 

of the dividing nucleus into the growing bud is demonstrated. 

Under the conditions used by these authors (Shaw et al., 1998) the average cell cycle 

time was 125 +/ 9 minutes. Of this, G1 and S together required 54 minutes to complete. 

Development of the short bipolar spindle (preanaphase) took a further 16 minutes. 

Capture and sorting of the daughter chromatids and elongation of the spindle to its fullest 

extent (anaphase A, B) required a further 30 minutes. The final division of the nucleus, 

migration of daughter nucleus into the bud and cytokinesis took 25 minutes. Molecular 

control of the cell cycle is predictably complex. Before the cell progresses into the S 

phase of DNA replication, it must achieve a critical size. The cell has mechanisms for 

sensing that an adequate supply of nutrients is available before it commits itself to 

replication. When these prerequisites are satisfied the cell cycle progresses through a 

critical checkpoint termed START. In S cerevisiae this is marked by the onset of bud 

emergence and DNA replication. If the cell is starved of an essential nutrient, it enters a 

resting state termed G0. The requirement to achieve a critical size explains why the 

average duration of the cell cycle is longer for daughter cells than their mothers. Thus, the 

former are smaller at birth and therefore require a longer G1 phase to achieve the critical 

size. A further checkpoint exists at the boundary between S and G2. At this point, there is 

a further critical cell size constraint, which must be satisfied before progression into 

mitosis occurs. In the case of S. cerevisiae, cells are usually sufficiently large at start to 

ensure that this second checkpoint is hidden. 

It can be shown to exist in the case of the fission yeast, Schizosaccharomyces pombe. 

This was demonstrated using the so-called wee mutants (Nurse, 1975). These cells have a 

gene, termed Wee 1, that is defective when the yeast is cultivated at high (restrictive) 

temperature. At the restrictive temperature, the cells undergo fission at a much smaller 


size than the wild type. Thus, the normal product of the Wee 1gene restricts entry into 

mitosis. The function of this second checkpoint is to allow the cell to ensure that 

chromosome duplication and segregation has proceeded without error. If necessary, it 

provides an opportunity for corrective action. This is vital since errors at this stage could 

have fatal consequences. Progression through the cell cycle requires activation and 

inactivation of enzyme systems via the intermediary of protein kinases and phosphatases. 

These regulatory cascade reactions are of the type described in Sections 12.2 and 12.5.8. 

Passage through the critical control points of the cell cycle is regulated by the controlled 

expression of genes that encode for catalytic proteins termed cyclins. At least 22 cyclins 

have been isolated. Cyclins interact with protein kinase cyclin-dependent kinases 

(CDKs). 

Several versions of these enzymes have been isolated from various cell types. In the 

case of S. cerevisiae, activation and deactivation of cdc28 (gene CDC28) regulates 

passage through the critical checkpoints of the cell cycle. Other cyclin-dependent kinases 

arealsoinvolved (Stark, 1999).Theprecisesequence ofeventsishighlycomplexandnot 

fully characterized. Nevertheless, the general principles are that there is asensing system 

by which the cell monitors that the conditions are favourable for the cell cycle to 

progress. These include parameters such as nutrient availability, achievement of the 

critical cell size and confirmation that chromosome replication and segregation has been 

achieved. If the checks are satisfied, appropriate cyclins are synthesized and the cyclindependent 

kinase (cdc28) is activated. At START, for example, phosphorylation by the 

activated cdc28 kinase (S phase promoting factor) stimulates passage into the Sphase by 

activation of transcription. Mitosis promoting factor (cdc28 activated by B cyclins) 

controls passage from G2 to M. 

Individual cyclins are synthesized and degraded at various points in the cell cycle. 

Thus, G1 cyclins are associated with the passage from G1 to S. G2 cyclins (cyclin B) 

guide the cell through the transitions from Sto G2 and G2 into mitosis. Successive 

families of cyclins are responsible not only for promoting entry into the next stage of the 

cell cycle but also for degrading the cyclins responsible for driving the previous stage. 

Degradation is accomplished using the poly-ubiquination system associated with the 

yeast proteosome (Section 12.8). 

11.7.1 Yeast sexual cycle 

Many yeast genera are capable of sexual reproduction. S. cerevisiae is aperfect member 

of the Ascomycetae and is included in this group. Wild type strains of S. cerevisiae are 

usually diploid. Under appropriate conditions, these yeasts can be induced to undergo 

meiotic division and produce ascospores that are borne in afruiting body, an ascus. 

Industrial strains of S. cerevisiae, including brewing strains, are typically aneuploid/ 

polyploid and do not normally have asexual cycle. From this standpoint, the sexual stage 

of Saccharomyces yeasts is not relevant to brewing. Nevertheless, the sexual cycle of 

yeasthasbeenusedwidely, asamethodforexploringthegenome.Forthisreason,abrief 

descriptionoftheimportantfeaturesisincludedhere.Thekeyfeaturesoftheyeastsexual 

cycle are shown in Fig. 11.8. 

Haploid strains capable of sexual conjugation are of either of two mating types, termed 

MATa and MAT . Mating occurs in response to exposure of cells of opposite mating 

type to pheromones, termed a and factors, respectively. The pheromones are short 

peptides and they bind to specific receptor sites on the surface of cells of opposite mating 

type. Binding of the appropriate pheromone causes cells to arrest in the G1 phase of the 


a/α 

Zygote 

a 

α 

A mating type 

haploid phase 

Diplophase 

a/α 

∝ mating type 

haploid phase 

HOMOTHALLIC 

spores fuse to 

give diploid 

Meiosis 

Asci with four 

ascospores 

HETEROTHALLIC 

Stable haploid phases 

formed providing cells 

of opposite mating 

type kept apart 

Fig. 11.8 Sexual life cycle of S. cerevisiae. Note that brewing strains of S. cerevisiae do not show 

a sexual cycle. 

cell cycle at START and thereby achieve cell cycle synchrony. Cells develop projections 

known as schmoos. At conjugation, the schmoos on adjacent cells become aligned, grow 

towards each other and eventually fusion occurs. This is mediated by expression of genes 

that encode for a specific group of surface agglutinins. Sexual conjugation is therefore a 

specific type of flocculation. During the fusion process, the plasma membranes of each 

cell become contiguous so the cytoplasm is shared. Nuclear fusion, or karyogamy to give 

a diploid cell follows cellular fusion. In a complete growth medium such diploid cells 

proliferate by budding and mitosis. This stage is known as diplophase. Under conditions 

of nutrient starvation or in the presence of a non-fermentable carbon source such as 

acetate or ethanol, but no nitrogen source, meiosis occurs and four haploid spores are 

formed. Spores can be induced to germinate by transfer to a rich growth medium. 

Two types of cell phenotype are possible depending on the possession or absence of a 

dominant allele termed HO. Heterothallic strains are HO and, providing the germinating 

ascospores are not brought into contact with a cell of opposite mating type, they will 

divide mitotically and produce stable haploid clones. Normally, ascospores are not 

liberated from the asci of S. cerevisiae, instead they fuse to produce diploid cells. 

Homothallic strains carry the dominant allele HO. Cells growing from spores, which 

bear the HO gene, are able to change mating type (Herskowitz et al., 1992). In budding 


α 

a 

α 

a

cells,theswitchinmatingtypeoccursonlyinmothersandnotdaughters.Oncethelatter 

havebecomemothers,theyalsoacquiretheability.Theswitchoccurswithanefficiency 

of approximately 60%. The result is that mating between siblings takes place and 

therefore, the haploid phase in homothallic strains is transient. Mating type is 

determined by expression of genes at the MAT locus on chromosome III. In addition, 

chromosome III has two `silent' loci, or cassettes, which are located to the right and left 

of the centromere. These carry mating information for both MATa (HMR, right hand of 

centromere) and MAT (HML, left hand of centromere). As the terminology suggests, 

genes at the silent loci are not normally expressed. However, in some cases the genes at 

the MAT locus are exchanged with those from the silent HML or or HMR loci. These 

silent genes may then be expressed. Depending on the original MAT locus (either 

MATa or MAT ) and the substitute, mating type can be changed. In this set of 

circumstances the cells of opposite mating type will rapidly fuse and regenerate diploid 

cells. 

The mating process is induced by asignal transduction pathway involving aprotein 

kinase cascade termed MAPK (Stark, 1999). The mating pheromones bind to cell surface 

receptors and in so doing activate atetrameric protein (G protein) whose sub-units 

dissociate in response to exchange of GDP for GTP. Sub-units of Gprotein in turn 

activate the formation of a second multi-subunit complex, which triggers the MAP 

protein kinase cascade. In turn this interacts with cyclins with aresultant arrest in the cell 

cycle at START. Other MAP kinase target sites mediate the reactions leading to schmoo 

formation and fusion. 

Intheprocessofmeiosis,alsoknownasreductivedivision,thediploidnucleusdivides 

twice to form four haploid nuclei. In the initial fusion, two haploid nuclei fuse to form a 

single diploid zygote containing two sets of chromosomes. In sporulation medium, the 

two sets of chromosomes come together as homologous pairs (the two corresponding 

chromosomes from the original haploid parents). The DNA is then replicated, during 

which there is the opportunity for crossing over of genes between pairs of chromosomes. 

In this way genetic material can be shared to produce daughter cells with characteristics 

inherited from either parent. Aprocess of separation then follows involving homologous 

pairs of chromosomes and then individual replicas. 

Eventually four sets of chromosomes are formed each enclosed in anucleus. The 

process of sporulation begins when spore coat materials are synthesized and deposited 

around the nucleus. The cell enclosing the spores develops into aspore-bearing body, the 

ascus. Spore formation is accompanied by the accumulation of the carbohydrate, 

trehalose (Dickinson, 1988). This material makes cells resistant to stresses such as 

desiccation by virtue of its ability to stabilize membranes (Section 12.5.7). 

11.8 Yeast genetics 

Genetics is the study of the relationships between the structure of the genotype and 

phenotypic expression. Genetic analyses provide a means of exploring the evolutionary 

and taxonomic relationships between individual strains. An understanding of the make-up 

of the genotype is a prerequisite for phenotypic modification. Thus, with knowledge of 

the nature of the genotype opportunities may present themselves by which undesirable 

characteristics can be deleted and desirable characteristics acquired. The comparatively 

rapid cell cycle of yeast, its ease of cultivation and relatively compact genome has made 

these organisms a common choice for the study of eukaryotic genetics, consequently, the 


scientific literature is enormous. Nevertheless, the majority of these studies employ 

haploid laboratory strains of S. cerevisiae. As discussed below, industrial strains of S. 

cerevisiae have amore complex genetic make-up. Aglossary of some terms used in 

genetics is given in Table 11.7 on pages 397±398. 

11.8.1 Methods of genetic analysis 

Genetic analyses are commonly performed using mutants. Exposure of cells to ionizing 

radiation or certain chemical compounds (mutagens) induces damage to the genome. 

Providing the treatment is not lethal, it is possible to obtain mutants. In other words, cells 

with an altered genotype with one or more defective genes. Selective media or selective 

cultural conditions can be used to identify and isolate mutants that are defective in the 

area of interest. 

The traditional method for determining the location and number of genes within the 

genome is by mating and meiotic recombination. Yeast is particularly suitable for this 

approach since the four ascospores are the result of a single meiotic event. Tetrad analysis 

allows mapping of genes relative to their centromeres and the drawing up of linkage maps. 

The latter indicate the relative distances between genes located on the same chromosome. 

Tetrad analysis is laborious. It involves mating of selected haploid strains to produce a 

hybrid diploid. After the induction of meiosis by transfer to sporulation medium, 

individual haploid ascospores are isolated by micromanipulation following enzymic 

digestion of the ascus wall. After induction of germination, the phenotype of the resultant 

haploid yeast lines can be assessed. Many aspects of the phenotypes of isolates can be 

assessed using the technique of replica plating. Here isolates are plated out onto a 

complete growth medium and incubated to allow the formation of colonies. An imprint of 

the colonies is made by pressing a piece of sterile velvet onto the plate. This is then used to 

inoculate plates of media that are selective for chosen phenotypic attributes. Comparison 

of patterns of growth on master and replica plates allows the assessment of phenotype. 

Tetrad analysis has allowed more than 90% of the genetic map of S. cerevisiae to be 

established (Cox, 1995). Although more modern techniques have superseded tetrad 

analysis as the method of choice for gene mapping it is still used to determine that a 

mutation has resulted from alteration of a single locus. Analysis of tetrads relies on the 

ratios of individual phenotypes to determine the relationships of the genes under 

investigation. Thus, a hybrid that is heterozygous for two markers, AB and ab has the 

potential to produce three classes of tetrad. These are: AB, AB, ab, ab (PD or parental 

ditypes); aB, aB, Ab, Ab (DPD or non-parental ditypes) and AB, Ab, ab, aB (T or 

tetratype). If the genes are unlinked, a random assortment will yield a ratio of 1 : 1 : 4 for 

PD, NPD and T, respectively. In the case of linked genes, there is an excess of PD to NPD 

tetrads. Where two genes are on different chromosomes and are linked to their respective 

centromeres the proportion of T type tetrads is reduced. 

The distance between individual genes on chromosomes influences the chances of 

crossovers during meiosis. Therefore, the frequency of occurrence of the different classes 

of tetrads can be used to determine the distance between genes. The distance between 

genes is measured in centiMorgans (cM). A centiMorgan is the unit on a genetic map 

equal to the distance along a chromosome that gives a recombination frequency of 1%. 

For map distances up to 35 cM the following equation may be used: 

cM ˆ 100 

2 

T ‡ 6NPD 

PD ‡ NPD ‡ T 


For greater distances, up to 75 cM, a correction has to be made and the following 

empirically derived calculation is used: 

cM …corr:† ˆ 

…80:7† …cM† …0:883† …cM†2 

83:3 cM 

The distance between a gene and its centromere can be determined by measuring the 

percentage of T tetrads using a marker gene that is known to be tightly linked to the 

centromere. In this instance, the following formula is used: 

cM ˆ 100 

2 

T 

PD ‡ NPD ‡ T 

The proportion of the genome that resides in the nuclear chromosomes is described as 

being Mendelian since recombination events proceed according to Mendelian rules of 

inheritance. Other elements of the genome such as that found in mitochondria can also be 

detected by tetrad analysis. For example, non-Mendelian inheritance is observed for any 

character that is coded for by non-chromosomal DNA. 

Identification of a gene associated with a particular mutation can be accomplished by 

genetic complementation studies. This involves producing a range of diploid crosses by 

mating a haploid bearing the uncharacterized mutation with a range of characterized 

haploid mutants with the same phenotype as the test strain. A control diploid is produced 

using the test mutant and a normal wild type strain. The normal wild type phenotype in 

the control heterozygote confirms that the unknown mutation is recessive. The 

heterozygous cross that produces the mutant phenotype must be double recessive and 

therefore contains the same mutation as the unknown. All other heterozygotes lose the 

original mutant phenotype indicating that the uncharacterized mutation was not present in 

any of the control test strains. 

The development of recombinant DNA technology has revolutionized the ease by 

which the yeast genome may be analysed and manipulated. Very powerful and precise 

techniques are now available that allow detailed analysis of the genome. Individual genes 

can be targeted, modified and transferred between cells to facilitate the study of the 

relationships between genotype and phenotype. It is now relatively simple to introduce 

foreign DNA into a cell to effect desirable changes in the phenotype. Detailed and rapid 

genetic analyses provide convenient and precise methods for strain identification ± socalled 

genetic fingerprinting. 

Many and varied techniques have been developed. They share some common themes. 

Unlike classical genetic techniques, DNA is extracted and manipulated in isolation. In 

order to work with complex mixtures of DNA it is necessary to have methods of 

separation, enrichment and identification. In a DNA molecule, each gene is encoded by a 

unique sequence of bases. In single stranded form, this sequence can be targeted 

providing a probe is available that contains the complementary sequence. Occasionally, 

the DNA sequence of interest may constitute a very small proportion of the whole. 

Methods are available, which allow specific fragments to be selected and duplicated in 

vitro. In recent years, much effort has been directed towards identifying the base 

sequences of entire genomes. The enormous magnitude of this task has necessitated the 

development of automatic sequencing equipment. The prize is the ability to identify any 

gene within the genome. The structure of the protein expressed by the gene can be 

manipulated. Artificial genes can be created in vitro. 

Electrophoretic characterization of mixtures of DNA is achieved by loading onto a 


permeable gel made from a material such as polyacrylamide or agarose. The gel is 

submersed in an electrolyte and an electric current is applied. DNA molecules are 

charged molecules and in the electrical field, they migrate between electrodes. The extent 

of migration is dependent on molecular size and charge. Several different electrophoretic 

techniques are used which allow separation of different ranges of DNA molecule. These 

may range from whole chromosomes to small DNA fragments a few bases in length. 

Whole chromosomes can be separated using pulse field electrophoresis. Chromosomes 

can be visualized by staining with ethidium bromide and viewing under ultraviolet light. 

This procedure is termed karyotyping. It is used as a method of strain identification and to 

probe chromosome stability. More typically, the extracted DNA is broken into fragments 

using enzymes termed restriction nucleases. The latter are a large family of bacterial 

enzymes that break DNA molecules at particular known recognition sequences. The fragments 

can be separated by gel electrophoresis. In the technique termed Southern blotting, the 

separated fragments of DNA are transferred to a nylon or nitrocellulose membrane. During 

the process, the double-stranded DNA molecules are denatured to give single strands on the 

nylon sheet. The presence of specific DNA sequences can be detected by treating the 

membranes with probes consisting of complementary strands of DNA attached to a 

visualizing system. The latter may be a radioactive label or a dye such as a fluorochrome. 

This procedure can be used as a method of strain identification in the technique termed 

restriction fragment length polymorphism (RFLP). Providing the complementary DNA 

probe is available it can be used to detect the presence of any gene within the test 

organism. The same approach can be used to probe mixtures of RNA molecules. In this 

case, termed Northern blotting, mRNA molecules are subject to electrophoresis. Specific 

RNA sequences are detected using a probe, which is a single strand of DNA. 

In many cases, it is necessary to enrich extracted DNA to increase the yield of 

particular genes of interest. This process is termed amplification and it can be 

accomplished in several ways. It can be performed in vitro using DNA polymerase, the 

enzyme that is responsible for DNA replication in the cell. The technique is known as 

PCR, from polymerase chain reaction. The extracted target DNA is denatured by heating 

and each single strand attached to a suitable primer. The latter is a short nucleotide 

sequence that targets the DNA of interest, such that any that is attached to the primer is 

duplicated. The entire duplication process takes just a few minutes. Repeated cycling 

allows multiple copies of the genes of interest to be made. In a short time, this 

amplification process results in several million copies of the original DNA being 

manufactured. PCR can be used to amplify the DNA fragments formed in the restriction 

fragment length polymorphism (RFLP) approach to strain identification. This combination 

of methods, termed amplified fragment length polymorphism (AFLP) increases the 

sensitivity of the genetic fingerprint. 

Exogenous DNA fragments can be introduced into a cell and thereafter inherited and 

expressed with the cell's own genome, a process termed transformation. DNA may be 

introduced directly by first converting the target cell into a sphaeroplast. This is 

accomplished by enzymic removal of the cell wall in the presence of an osmotic 

stabilizer. Incubation of sphaeroplasts and DNA on solid media containing osmotic 

stabilizers provide conditions under which transformation and re-growth of the cell wall 

can occur. Newer procedures obviate the need for removal of the cell wall. For example, 

cell walls can be made permeable to DNA by treatment with lithium acetate. 

Alternatively, cell membranes can be made temporarily permeable to DNA by 

application of an electrical field in the process of electroporation. Exogenous 

mitochondrial DNA can be introduced into the mitochondria of host cells by high- 


velocity bombardment. In this case, the DNA is contained within a tungsten microprojectile, 

which is fired into the mitochondrion using compressed air. 

The most usual method of introducing exogenous DNA into a cell is via the 

intermediary of a vector. These are small pieces of circular DNA termed plasmids. The 

DNA fragment of interest is integrated into the plasmid DNA. They have several 

advantages as transforming agents compared to the use of pure DNA. Several different 

plasmids are available which can be used to accomplish several different functions. These 

include insertion, deletion, alteration and expression of genes. All have marker genes so 

those host cells containing the plasmid can be selected. Commonly, they bear genes 

conferring resistance to selected antibiotics. Plasmids are usually of the shuttle type, 

referring to the fact that apart from being able to maintain themselves in yeast cells they 

have an origin of replication that allows generation of a high copy number in an 

alternative host such as E. coli. 

The ability to grow the plasmid in a bacterium provides a convenient method of gene 

amplification. When introduced into a yeast cell some plasmids integrate into a 

chromosome whilst others are able to replicate autonomously. These differences are 

utilized in different gene manipulations. For example, integrative types might be used 

where a very stable transformation is required. Conversely, autonomously replicating 

types may be used where very high copy numbers of the transforming gene are needed. 

Plasmids can be used to perform a number of tasks. They afford the most precise way of 

identifying a mutant gene. Thus, mutant cells are transformed with plasmids bearing a 

library of fragments of the wild type genome. Any cells with the wild type phenotype 

must contain a functional copy of the gene on the plasmid. The latter can be recovered 

and analysed for the presence of the appropriate gene. Similar procedures can be used to 

introduce a selected gene into a cell or to disrupt a chosen gene within a cell. 

The sequence of bases in DNA molecules can be determined. In order to deal with 

whole genomes, the enormous number of bases involved requires that processes are 

automated. Robotic samplers and automatic analysers are now available. Typically, the 

shotgun sequencing approach has been used. In this method, the genome is cut into a 

large number of random fragments. The sequence of each fragment is determined. Using 

the fast and relatively inexpensive computational power now available, the sequence of 

the whole can be deduced from those of the fragments. 

The entire genome of S. cerevisiae has now been sequenced and the DNA 

corresponding to each identified chromosomal gene is available in the form of a library. 

Pieces of apparatus termed arrayers have been developed that transfer, or print, individual 

polynucleotide samples corresponding to each gene onto a membrane. Each sample dot is 

approximately 100±200 m in size and arranged on a membrane in the form of a grid. The 

loaded grids hold the entire chromosomal genome. They are termed variously as DNA 

microarrays, DNA chips or gene chips. The oligonucleotides on the grids can be probed by 

hybridization with other samples of DNA. The probes are attached to fluorescent dyes and 

hybridization can be visualized by confocal microscopy. Using this system, the DNA 

probe mixture can be prepared by reverse transcriptase from the entire mRNA content of 

the strain under investigation. By testing on a DNA array containing the original DNA 

genome of the strain, the relative expression of the whole genome can be assessed. 

11.8.2 The yeast genome 

Haploid cells of S. cerevisiae contain 16 chromosomes, ranging in size from 200 to 

2200 kb. All have been sequenced and 1996 saw publication of the complete sequence of 


the entire genome of astrain of S. cerevisiae (Goffeau et al., 1996). The magnitude of this 

taskisreflectedbythefactthatinthecaseofS.cerevisiae,sometwelvemillionnucleotide 

bases were sequenced. Comparison of this sequence with those obtained from other cells 

and with knowledge of gene structure allows identification of sequences that encode for 

specific proteins. Such sequences are termed open reading frames (ORFs). Some 6,217 

potential open reading frames have been identified in the yeast genome (Mewes et al., 

1997).Ofthese6,217potentialproteins,approximatelyhalfhavebeenpositivelyidentified 

basedonknownsequences.Afurther20%haveaputativeidentificationandtheremaining 

30%,theso-calledòrphangenes'havenoknownfunction.Thepositivelyidentifiedgenes 

have been classified based on cellular function (Table 11.5). 

An attempt to gain an understanding of the function of orphan genes is being 

addressed by novel techniques. These involve the construction of mutants in which a 

single orphan gene is deleted. The phenotype is then examined to determine the 

metabolic consequences of the deletion. This approach has been termed reverse genetic 

analysis (Oliver, 1997). It is anticipated that this will culminate in the development of 

automatic analysis of the proteome in amanner analogous to the DNA array systems 

described in Section 11.8.1. 

Compared to many other cells, the yeast genome is very compact. Approximately 72% 

of chromosomal DNA codes for actual genes. The average size of yeast genes is 1.45 kb 

or 483 codons representing 40 to nearly 5,000 codons. Approximately 4% of yeast genes 

contain introns (non-coding regions). Genes are not evenly distributed throughout the 

chromosomal DNA, instead there are gene-rich clusters. In haploid strains, approximately 

half of the genes are duplicated. This has led to the suggestion that this species arose from 

the fusion of two ancestral diploid strains, each with eight chromosomes. The resultant 

tetraploid cell was reduced to a 16 chromosome diploid by deletion (Wolfe and Shields, 

1997). Yeast chromosomes contain additional DNA known as retrotransposons or Ty 

elements. Typically, laboratory haploid strains contain up to 30 copies of a number of 

different retrotransposons. These are related to retroviruses. 

Approximately 10% of the yeast genome is located in mitochondria. The 

mitochondrial genome encodes around 5% of mitochondrial proteins. The wild type 

mitochondrial phenotype is denoted + . Mutants termed ë lack all mitochondrial DNA. 

They are viable but lack many components of the electron transport chain and are 

therefore, respiratory deficient. Mutants that are ë produce small colonies on solid media 

and for this reason are termed petites. Petite mutation is relatively common in production 

brewing (Donnelly and Hurley, 1996). Strains of S. cerevisiae contain a variety of other 

Table 11.5 Yeast genome analysis of identified genes based on function (Mewes et al., 1997) 

Gene function Proportion of identified genome (%) 

Cellular organization and biogenesis 28 

Intracellular transport 5 

Transport facilitation 5 

Protein trafficking 7 

Protein synthesis 5 

Transcription 10 

Cell growth, division and DNA synthesis 14 

Energy transduction 3 

Metabolism 17 

Cell rescue 4 

Signal transduction 2 


genetic elements. Most contain 2 mplasmids, which apparently function solely for their 

own replication since deletion has no observable effect on the yeast phenotype. 

Commonly, the cytoplasm of yeast contains viral nucleic acid termed dsRNA. Several 

types are found of which the M-type encodes for killer toxin. 

Much of the work, which has resulted in the current level of understanding of the yeast 

genome, has been performed on laboratory haploid strains. The genomes of industrial 

strains are very different. The most notable difference is the fact that brewing yeast strains 

are polyploid or aneuploid. Commonly, three or four sets of chromosomes are present 

(triploid,tetraploid).Oftenthesetsofchromosomesarenotpresentinmatchedsets,rather, 

one or more chromosomes is present as an extra or one less copy (aneuploid) (Hammond, 

1996). Polyploidy has many consequences. The sequenced haploid laboratory strain of S. 

cerevisiae contains a single copy of the genes encoding for the maltose fermenting 

enzymes, whereas brewing strains typically contain more than ten copies (Meaden, 1996). 

Polyploid strains of S. cerevisiae may have been selected for in industrial processes 

since they may have very stable phenotypes. Thus, the chance of asingle point mutation 

having an effect on the phenotype is reduced where multiple copies of the gene are 

present on other chromosomes. In addition, it is possible that multiple copies of some 

genes and concomitant increased expression might confer aselective advantage. For 

example, it has been claimed that multiple copies of maltose utilizing genes produces a 

phenotype where maltose utilization rates are higher than comparable haploid strains 

(Hammond, 1996). 

Noorganism has astable genotype. Mutation and selection provide the mechanism for 

evolutionary change. Although polyploidy reduces the chances of phenotypic change via 

mutation, it is not made impossible. Brewing yeast, in common with other fungi, exhibits 

chromosomal instability. This can be observed by karyotyping (Section 11.8.1). The 

changes show up as alterations in chromosome length (polymorphism) or even 

occasionally chromosome deletion. The changes in chromosome length are areflection 

of DNA rearrangement and distribution between individual chromosomes (Zolan, 1995). 

In brewing yeast, it has long been recognized that flocculation properties of individual 

strains are liable to abrupt shifts during repeated cycles of fermentation, cropping and repitching 

(Gilliland, 1971). Similar observations have been made with alager yeast strain 

used in production brewing. Genetic fingerprinting of the non-flocculent and flocculent 

variants showed differences (Wightman et al., 1996). 

11.9 Strain improvement 

Industrial strains of yeast, which have been in use for any length of time, will have been 

selected using empirical criteria. In brewing, for example, individual strains are used 

because they possess amix of desirable characteristics. These characteristics ensure that 

for particular combinations of wort, vessel type and yeast strain produce the desired 

fermentation performance and beer quality. Greater economic pressures of modern largescale 

brewing and the desire to develop new products have provided the impetus to 

improve strains. Improvement is taken to mean to introduce characters into existing 

strains such that they acquire an altered phenotype. Some of the potential goals of strain 

improvement are summarized in Table 11.6. 

Traditional genetic approaches for strain improvement involve mating between 

selected parents and selection of new hybrid offspring containing desirable traits. 

Alternatively, attempts can be made to select for mutants within a population that have 


Table 11.6 Goals for yeast strain improvement 

Aim Strategy 

Increased yield of ethanol Introduction of amylases to allow fermentation of wort 

dextrins 

Increased attenuation rate Increased concentration of glycolytic enzymes 

Very high gravity fermentation Increased ethanol and osmotolerance 

High temperature fermentations Increased thermotolerance without alteration in 

production of beer flavour metabolites 

Reduced risk of contamination by wild Incorporation of killer factor 

yeast 

Abolishment of diacetyl stand Increased flux through valine/isoleucine biosynthetic 

pathway 

Incorporation of -acetolactate decarboxylase 

Improved control of production of beer Manipulation of pathways for production of esters, 

flavour metabolites H 2S, SO 2 

Reduced ethanol yield Block ethanol production for low/zero alcohol beer 

production 

Altered flocculation Manipulation of Flo genes 

desirable characteristics. The frequency of occurrence of mutants can be increased by 

treatment withmutagens.Brewingyeaststrainsdonotlendthemselvestoeitherapproach 

because their polyploid nature makes them relatively resistant to mutation. In addition, 

they usually lack asexual cycle and sporulate at best poorly. 

It is possible to produce hybrids using rare mating (Spencer and Spencer, 1977; 

Young, 1981). This procedure is used to produce hybrids of respiratory deficient brewing 

strains and respiratory sufficient mutant strains. Hybrids are selected by ability to grow 

on defined media deficient in the nutrient requirement of the mutant but containing an 

oxidative carbon source. Using this technique, auxotrophic diploids with either aa or 

matingalleleshavebeenhybridizedwith brewingstrains.Thehybridswerecapableof 

sporulating. Providing one of the parents contains the kar allele it is possible to produce 

heteroplasmons. The kar allele prevents nuclear fusion andtherefore these strains contain 

somatic elements of both parents with the nucleus of only one parent. This approach has 

been used to construct brewing strains containing killer factor. 

Protoplasts or sphaeroplasts are yeast cells which have had the cell wall removed by 

treatment with asuitable enzyme preparation. Providing the cells are suspended in an 

osmotically stabilized medium the sphaeroplasts do not lyse. In the presence of Ca 2+ ions 

and polyethylene glycol sphaeroplasts can be induced to fuse and form hybrids (Pesti and 

Ferenczy, 1982). When sphaeroplasts are incubated on asuitable solid medium cell wall 

regeneration occurs and stable hybrids are obtained. Hybrids can be selected by ensuring 

that both parental types are auxotrophic for different nutrients. Only hybrids will grow on 

amedium that is deficient in both nutrients. The procedure can be used to produce stable 

tetraploid from diploid parents. It has been largely superseded by more amenable 

recombinant DNA techniques. 

Strain improvement using recombinant DNA technology has the enormous advantage 

of precision. It is possible to introduce or modify asingle gene within the target genome. 

Several methods may be used to introduce exogenous DNA into a recipient cell 

producing aphenotype with new and desirable attributes. The methods used are those 

outlined in Section 11.8.2. Usually the method of choice is one in which the exogenous 

DNA is introduced in the form of a plasmid vector of the type that forms a stable 

integrant with the host chromosomal DNA. These techniques have been applied 


Table 11.7 Glossary of terms used in genetics 

Term Definition 

Allelle One of a pair of genes located at the same point on homologous 

(syn. allelomorph) chromosomes 

Aneuploid Polyploid strains lacking whole copy numbers of all genes 

Ascospore Haploid spores produced via meiosis in Ascomycetous fungi 

Ascus Spore-bearing body in Ascomycetous fungi 

Auxotroph Mutant with specific nutrient requirement for growth 

Bases Constituent molecules of RNA and DNA whose order form the basis of 

genetic code. DNA contains adenine and guanine (purines); thymine and 

cytosine (pyrimidines). In RNA uracil replaces thymine 

Biochip Array of 1000 or more oligonucleotides printed on a solid support. (syn. 

DNA chip, DNA microarray, gene chip, genome chip) used to identify genes 

present within a genome 

Cdc Cell division cycle mutants 

cDNA Complementary DNA produced by reverse transcription of mRNA 

Centromere Attachment site of chromosome to spindle body 

CHEF Electrophoresis technique ± contour clamped homogeneous electric field 

Chromosomes Structures in which duplex DNA molecules are located 

CLP Chromosome length polymorphism 

Codon Triplet of bases which together code for a single amino acid 

Differential Study of relative levels of all mRNA in cells in different physiological 

expression studies states (syn. transcriptiomics) 

Diploid Organism having two sets of chromsomes 

DNA Deoxyribonucleic acid ± macromolecule which forms the genetic code. It 

consists of a backbone of sugar and phosphate groups to which purine and 

pyrimidine bases are attached. Two complementary molecules are attached 

together to form a double helix 

LTR Long terminal repeats ± long repeating sequences of bases flanking 

transposons in genome 

Meiosis Process of reductive division in which diploid nucleus divides with a single 

replication of DNA to give four haploid spores 

Mitosis Division of nucleus and replication of DNA during asexual cell division 

Monosomic Diploid cell lacking one of a pair of chromosomes 

Northern blot Electrophoretic technique using RNA instead of DNA 

ORF Open reading frame, sequences of DNA flanked by start and termination 

codons that code for a single protein 

Orphan genes DNA sequences identified as genes but not having recognized function 

PCR Polymerase chain reaction 

Plasmid Small, frequently circular sequence of extra-chromosomal DNA 

Probe Fragment of DNA of known sequence labelled with radioactive phosphorus 

or fluorochrome used for detection of complementary DNA in mixture 

Proteome Whole set of proteins produced by genome 

RFLP Restriction fragment length polymorphism 




Rho Mutant lacking mitochondrial genome 

(syn. 

- 

) 

RNA Ribonucleic acid ± macromolecule consisting of a backbone of phosphate 

and sugar molecules to which purines and pyrimidines are attached. RNA 

exists in several functional forms: 

mRNA ± messenger RNA, single stranded form which is synthesized during 

transcription of DNA. Provides template for amino acid synthesis 

rRNA ± ribosomal RNA, ribonucleic acid component of ribosomes, which 

mediate the process of translation of tRNA to form protein molecules 

tRNA ± transfer RNA, family of molecules which bind individual amino 

acids and transfer them to the ribosomes as directed by the codons on the 

mRNA molecule 

snRNA ± small nuclear RNA molecules that mediate post-translational 

processing of mRNA 

Sequence Order of nucleotide bases within molecules of DNA or RNA that collectively 

form the genetic code 

SDS-PAGE Sodium dodecyl sulphate polyacrylamide gel electrophoresis. Gel 

electrophoresis separation technique for proteins. SDS causes dissociation 

of multimeric proteins 

Sporulation Spore formation. In S cerevisiae, ascospores are formed during meiotic 

division 

Southern blot Electrophoresis of DNA on gel followed by denaturation and transfer of 

single-stranded DNA to membrane. Separated DNA fragments analyses by 

complementary DNA probes 

SSCP Single stranded confirmational polymorphism ± separation technique similar 

to Southerm blotting differing in that it relies on the migrational 

characteristics of single-stranded DNA 

TAFE Electrophoretic technique ± transverse alternating field electrophoresis 

Telomere Sequences of bases at end of eukaryotic chromosomes important for ensuring 

correct replication of terminal segments of DNA molecules 

Tetraploid Cell possessing four sets of chromosomes 

Transcription Synthesis of molecule of mRNA using single-stranded DNA molecule as 

template 

Translation Synthesis of polypeptide in ribosome using mRNA as template 

Transposon Sequence of DNA that that is capable of movement within or between 

(transposable element) chromosomes (see Ty elements) 

Triploid Cell with three sets of chromosomes 

Trisomic Diploid cell containing three copies of one chromosome 

Ty elements DNA elements that are classified as long terminal repeat (LTR)-containing 

retrotransposons. These are retroviruses 

Western blot Technique similar to Southern and Northern blot for electrophoretic 

separation and detection of proteins 

Zygote Stage in sexual reproduction comprising the diploid cell formed from the 

fusion of two haploid cells of opposite mating type 


successfully to brewing yeast strains. Recombinant yeast strains have been constructed 

which meet many of the needs outlined in Table 11.6 (Hammond, 1995). 

No yeast strains produced by recombinant genetic modification are used in production 

brewing even though one has been given approval for use by the United Kingdom 

regulatory authorities (Hammond, 1995). At present, at least within Europe, there is an 

extreme reluctance to accept genetically modified foods. Understandably, in this climate, 

commercial brewers are unwilling to risk the use of such yeast strains. Whether this 

situation will be transient or long lasting is unclear. It has had the result that development 

work in this area has been largely discontinued. Recombinant DNA technology continues 

to afford the most fruitful approach probing the secrets of the genome. Undoubtedly, 

within the short term considerable advances will continue to be made. 


ANDERSON, R. G. (1995) 25th Cong. Eur. Brew. Conv., Brussels, 13. 

ARIAS, C. R., BURNS, J. K., FRIEDRICH, L. M., GOODRICH, R. M. and PARISH, M. E. (2002) Appl. Environ. 

Microbiol., 68, 1955. 

ARNOLD, W. N. (1991) Periplasmic space. InThe Yeasts, Vol. 4, 2nd edn, A. H. Rose and J. S. Harrison 

eds, pp. 279±295, Academic Press, London. 

ARVINDEKAR, A. U. and NARAYAN, B. P. (2002) Yeast, 19, 131. 

BACALLAO, R. and STELZER, E. H. K. (1989) Meth. Cell Biol., 31a, 454. 

BALDWIN, W. W. and KUBITSCHEK, H. E. (1984) J. Bacteriol., 158, 701. 

BARKER, M. G. and SMART, K. A. (1996) J. Amer. Soc. Brew. Chem., 54, 121. 

CAHILL, G., WALSH, P. K. and DONNELLY, D. (1996) J. Amer. Soc. Brew. Chem., 57, 72. 

CHAPMAN, A. C. (1931) J. Inst. Brewing, 37, 433. 

CORRAN, H. S. (1975) A History of Brewing. David & Charles, London. 

COX, B. S. (1995) `Genetic analysis in Saccharomyces cerevisiae and Schizosaccharomyces pombe'. In 

The Yeasts, Vol. 6, 2nd edn, A. E. Wheals, A. H. Rose and J. S. Harrison eds, pp. 7±67. 

CURTIS, N. S. (1971) Brewers' Guardian, September, 95. 

DE SOUZA PEREIRA, R., PARIZOTTO, N. A. and BARANAUSKAS, V. (1996) Appl. Biochem., Biotechnol., 59, 

135. 

DENGIS, P. B., NELISSEN, L. R. and ROUXHET, P. G. (1995) Appl. Environ. Microbiol., 103, 257. 

DICKINSON, J. R. (1988) Microbiol. Sci., 5, 121. 

DONNELLY, D. and HURLEY, J. (1996) Ferment, 9, 283. 

EDDY, A. A. (1958) Àspects of the chemical composition of yeast'. In, The Chemistry and Biology of 

Yeasts, A. H. Cook, ed., pp. 157±249, Academic Press, London. 

FLEET, G.H. (1991) `Cell Walls'. In The Yeasts, Vol. 4, 2nd edn, A. H. Rose and J. S. Harrison eds, pp. 

199±277, Academic Press, London. 

FLORKIN, M. (1972) À History of Biochemistry', in Comprehensive Biochemistry, 30, 129. 

FORGET, C. (1988) A Dictionary of Beer and Brewing, Brewers' Publications, Colorado, USA. 

GALITSKI, T., SALDANHA, A. I., STYLES, C. A., LANDER, E. S. and FINK, G. R. (1999) Science, 285, 251. 

GILLILAND, R. B. (1971) Brewers Guardian, Oct., 29. 

GOFFEAU, A., BARRELL, B. G. and BUSSEY, H. (1996) Science, 274, 546. 

GREUTER, W., BARRIE, F. R., BURDET, H. M., CHALONIER, W. G., DEMOULIN, V., HAWKSWORTH, D. I., 

JOÈ RGENSEN, P. M., NICOLSON, D. H., SILVA, P. C., TREHANNE, P. and MCNEILL, J. (1996) International 

Code of Botanical Nomenclature, Koelz Scientific Books, KoÈnigstein, Germany. 

HAMMOND, J. R. M. (1995). Yeast, 11, 1613. 

HAMMOND, J. R. M. (1996). Yeast Genetics, In, Brewing Microbiology, 2nd edn, F. G. Priest and I. 

Campbell, eds, pp. 43±82, Chapman and Hall, London. 

HAROLD, F. M. (1995) Microbiology, 141, 2765. 

HAWKSWORTH, D. L. (1991) Mycological Res., 95, 641. 

HERSKOWITZ, I., RINE, J. and STRATHERN, J. N. (1992) `Mating-type determination and mating-type 

interconversion in S. cerevisiae'. In, The Molecular Biology of the Yeast Saccharomyces cerevisiae, 

Vol. 2 Metabolism and Biosynthesis, J. N. Strathern, E. W. Jones and J. R. Broach eds, pp. 583±656. 

Cold Spring Harbor Laboratory, USA. 

ISERENTANT, D. (1996) Cerevisia, 21, 30. 

JIN, Y. L. and SPEERS, A. (1998) Food Res. Internat., 31, 421. 

JOÈ RGENSEN, A. (1900) Micro-organisms and Fermentation, 3rd edn. Translated by A. K. Miller and A. E. 

Lennholm, Macmillan, London. 


KING, F. A. (1947) Beer Has a History, Hutchinson's Scientific and Technical Publications, London. 

KLIONSKY, D. J., NELSON, H. and NELSON, N. (1992) J. Biol. Chem., 267, 3416. 

KURTZMAN, C. P. and FELL, J. W. (1998) The Yeasts., A Taxonomic Study, 4th edn, (C. P. Kurzman and J. 

W. Fell eds., Elsevier, Amsterdam, Netherlands. 

LAWRENCE, D. R., BOWEN, W. R., SHARPE, F. R. and VENTHAM, T. J. (1989) Proc. 22nd Cong. Eur. Brew. 

Conv., Zurich, 505. 

LEATHER, R. V., DALE, C. J. and MOROSON, B. T. (1997) J. Inst. Brewing., 103, 377. 

LEBER, R., ZINSER, E., ZELLNIG, G., PALTAUF, F. and DAUM, G. (1994) Yeast, 10, 1421. 

LLOYD, D. and CARTLEDGE, T. G. (1991). `Separation of yeast organelles', in The Yeasts, 2nd edn, vol 4, 

A. H. Rose and J. S. Harrison eds, pp. 121±174, Academic Press, London. 

LODDER, J. (1970) The Yeasts, a Taxonomic Study, 2nd edn, North Holland Publishing Company, 

Amsterdam, Netherlands. 

MEADEN, P. (1996) Ferment, 9, 213. 

MESTDAGH, M. M., ROUXET, P. G. and DUFOUR, J. P. (1990) Ferment, 3, 31. 

MEWES, H. W., ALBERMANN, K. and BARR, M. (1997) Nature, 387 (suppl), 7. 

MIKI, B. L. A., POON, N. H., JAMES, A. P. and SELIGNY, V.L. (1982) J. Bacteriol., 150, 878. 

MONTROCHER, R., VERNER, M.-C., BRIOLAY, J., GAUTIER, C. and MARMEISSE, R. (1998) Internat. J. 

Systematic Bacteriol., 48, 295. 

NARZISS, L. (1984) J. Inst. Brewing, 90, 351. 

NURSE, P. (1975) Nature, 256, 547. 

O'CONNOR-COX, E. S. C., LODOLO, E. J. and AXCELL, B. C. (1996) J. Inst. Brewing, 102, 19. 

OLIVER, S. G. (1997) Microbiology, 143, 1483. 

PESTI, M. and FERENCZY, L. (1982) J. Gen. Microbiol., 128, 123. 

PHAFF, H. J. and STARMER, W. T. (1987) `Yeasts associated with plants, insects and soil'. In The Yeasts 2nd 

edn, Vol. 1, A. H. Rose and J. S. Harrison eds, pp. 123±180, Academic Press, London. 

PISKUR, J., SMOLE, S., GROTH, C., PETERSEN, R. F. and PEDERSEN, M. B. (1998) Internat. J. Systematic 

Bacteriol., 48, 1015. 

QUAIN, D. E. (1986) J. Inst. Brewing., 92, 435. 

QUAIN, D. E. (1988) J. Inst. Brewing., 95, 315. 

REED, G. and NAGODARITHANA, T. W. (1991) Yeast Technology, Van Nostrand, USA. 

ROSE, A. H. and HARRISON, J. S. (1971) In The Yeasts, A. H. Rose and J. S. Harrison eds, 1st edn, pp. 1±2, 

Academic Press, London. 

SCHWENCKE, J. (1991) `Vacuoles, internal membranous systems and vesicles'. In The Yeasts, Vol. 4, 2nd 

edn, A. H. Rose and J. S. Harrison eds, pp. 347±432, Academic Press, London. 

SHAW, S. L., MADDOX, P., SKIBBENS, R. V., YEH, E., SALMON, E. D. and BLOOM, K. (1998) Molec. Biol. Cell, 

9, 1627. 

SINGER, S. J. and NICHOLSON, G. L. (1972). Science, 175, 720. 

SMITS, G. J., KAPTEYN, J. C., VAN DEN ENDE, H. and KLIS, F. M. (1999) Curr. Opinion in Microbiol., 2, 348. 

SPENCER, J. F. T. and SPENCER, D. M. (1977) J. Inst. Brewing, 83, 287. 

STARK, M. J. R. (1999). `Protein phosphorylation and dephosphorylation'. In, The Metabolism and 

Molecular Physiology of Saccharomyces cerevisiae J. R. Richardson and M. Schweizer eds, 

pp. 209±275. Taylor & Francis, London. 

STEVENS, B. J. (1977) Biol. of the Cell, 28, 37. 

STRATFORD, M. (1992) Yeast, 8, 635. 

STRATFORD, M. (1994) Yeast, 10, 1741. 

STRATFORD, M. and ASSINDER, S. (1991) Yeast, 7, 559. 

TAKITA, M. A. and CASTILHO-VALAVACIUS, B. (1993) Yeast, 9, 589. 

TAMAI, Y., MOMMA, T., YOSHIMOTO, H. and KANEKO, Y. (1998) Yeast, 14, 923. 

TEUNISSEN, A. W. R. H. and STEENSMA, H. Y. (1995). Yeast, 11, 1001. 

VAUGHN-MARTINI, A. and MARTINI, A. (1993) Systematics and Appl. Microbiol., 16, 113. 

VISSER, W., VAN SPONSEN, E.A., NANNINGA, N., PRONK, J. T., KEUNEN, J. G. and VAN DIJKEN, J. P. (1995) 

Antonie van Leeuwenhoek., 67, 243. 

WIGHTMAN, P., QUAIN, D. E. and MEADEN, P. (1996) Lett. Appl. Microbiol., 22, 85. 

WILLIAMSON, D. H. (1991) `Nucleus, Chromosomes and Plasmids'. In The Yeasts, Vol. 4, 2nd edn, A. H. 

Rose and J. S. Harrison eds, pp. 433±488, Academic Press, London. 

WOLFE, K. and SHIELDS, D. (1997) Nature, 387, 708. 

YOUNG, T.W. (1981) J. Inst. Brew., 87, 292±295. 

YOUNG, T.W. (1987) `Killer Yeasts'. In The Yeasts, vol. 2, 2nd edn, A.H. Rose and J.S. Harrison eds, pp. 

131±164, Academic Press, London. 

ZOLAN, M. E. (1995) Microbiol. Rev., 59, 686. 


12 

Metabolism of wort by yeast 


The biochemical reactions that occur during fermentation represent the cumulative 

effects of the growth of yeast on wort. The disappearance of nutrients and the formation 

of ethanol, carbon dioxide and the other metabolites, which together contribute to beer, 

are all by-products of yeast growth. Wort is complex and not completely characterized. 

Similarly, the reactions which occur during fermentation are not fully characterized. The 

biochemistry of S. cerevisiae has been subject to intensive study. However, the majority 

of these studies involved laboratory strains growing under aerobic conditions, often on 

defined media. The literature describing the biochemistry of the metabolism of wort by 

brewing yeasts during brewing fermentation is much smaller. Here is provided an 

overview of the metabolism of S. cerevisiae. Where possible the discussion is limited to 

the metabolism of wort by brewing strains of S. cerevisiae. Some aspects of the 

metabolism of non-brewing strains of yeast are also included where they have relevance 

as potential spoilage organisms. 

Brewing yeast strains are heterotrophic, facultative anaerobes. Acomparatively wide 

spectrum of organic molecules can be oxidized to both generate energy in the form of 

ATP and simultaneously provide carbon skeletons for anabolic reactions. Depending on 

the availability of oxygen and the concentration and source of carbohydrate, metabolism 

may be fully aerobic and oxidative, or fermentative. Thus, brewing yeasts have a 

relatively versatile metabolism and are able to adapt to avariety of conditions and 

furthermore, environmental triggers modulate their physiology. The biochemical events 

that occur during fermentation reflect the genotype of the yeast strain used and its 

phenotypic expression as influenced by the composition of the wort and the conditions 

established in the fermenting vessel. In order to obtain satisfactory fermentation 

performance and beer of adesired quality it is necessary to choose ayeast strain with a 

suitable genotype and manipulate the conditions to encourage appropriate metabolic 

behaviour. 

All brewing yeast strains have limited respiratory capacities and are subject to carbon 

catabolite repression (Gancedo, 1992; Section 12.5.8). Under the conditions encountered 


during fermentation, this phenomenon ensures that yeast physiology is repressed at all 

times and, therefore, the major products of sugar catabolism are ethanol and carbon 

dioxide. The initial concentration and spectrum of fermentable carbohydrates control the 

concentration of ethanol synthesized during fermentation. The range of carbohydrates 

that the yeast is able to ferment and the maximum concentration of ethanol that it can 

tolerate are genetically determined. These factors are criteria used in the selection of 

yeast strains used in brewing. 

Besides ethanol and carbon dioxide, amultitude of other minor products of yeast 

metabolism are formed duringfermentation. Many ofthesecontribute tobeerflavour and 

aroma.Theproductionofadesiredspectrumofflavourandaromacompoundsconstitutes 

another selection criterion for yeast strains, albeit with the caveat that fermentation 

conditions must be controlled properly to ensure that these minor metabolic by-products 

are synthesized in desired and consistent quantities. 

In the vast majority of breweries fermentation is abatch process. Usually, the pitching 

(inoculum) yeast is derived from a previous fermentation. This practice has many 

ramifications, some of which influence the metabolism of wort by yeast. Most notably, it 

provides a requirement to oxygenate wort. Molecular oxygen is necessary for the 

synthesis of sterols and unsaturated fatty acids. These are essential components of 

membranes and are aprerequisite for yeast growth under anaerobic conditions (Parks, 

1978; Section 12.6). Asingle dose of oxygen is supplied with the wort at the start of 

fermentation and this is used for the synthesis of sterol and unsaturated fatty acids. 

During the subsequent anaerobic phase of fermentation, the pre-formed pools of these 

metabolites, together with asmall quantity of lipid supplied with wort, are progressively 

diluted amongst daughter yeast cells. In the yeast crop obtained at the end of 

fermentation, sterol and unsaturated fatty acid levels are reduced to growth-limiting 

concentrations, hence, the need for oxygenation of wort in the next fermentation. 

At the end of fermentation, it is necessary to separate the yeast crop from the immature 

(`green') beer. This is facilitated by the design of the fermenting vessel and the flocculation 

characteristics of the yeast. Bottom-fermenting yeasts sediment to the base of the vessel at 

the end of fermentation, whereas top-fermenting types form acrop at the surface of the 

fermenting wort. Fermenter design must accommodate harvesting of each type of yeast. 

Typical high-gravity lager wort with aspecific gravity of 1.060 contains approximately 

150g/l fermentable sugar and 150mg/l free amino nitrogen. At the start of fermentation, the 

wort is oxygenated to achieve adissolved concentration within the range 15±25mg/l. It is 

pitched with yeast at arate of around 1gdry wt./l, equivalent to roughly 5gwet wt./l or 

12±15 10 6 cells/ml. Duringfermentation,theyeastconcentration increases aroundfivefold. 

Yeast growth is accompanied by the formation of roughly 45g/l ethanol and 42g/l 

carbon dioxide. The conversion of sugars to ethanol is about 85% of the theoretical. The 

shortfall represents the proportion ofwortsugars utilised for yeast biomassformation and 

other metabolites. The yield of carbon dioxide is marginally less than that of ethanol 

since some of the former is fixed by yeast in carboxylation reactions. Growth of yeast on 

wort is an exothermic process and it is necessary to apply cooling during fermentation to 

dissipate heat. The changes in some key parameters during a high-gravity lager 

fermentation of this type are shown in Fig. 12.1. 

The practice of serial fermentation, cropping and re-pitching introduces a requirement 

for yeast storage in the intervals between fermentations. The duration of the storage phase 

and the conditions employed influence yeast physiological condition. In particular, the 

intracellular concentrations of storage carbohydrates (glycogen and trehalose) and sterols 

and other lipids can be influenced. Variations in the concentrations of these metabolites 


(a) 

(b) 

Dissolved oxygen 

concentration (mg l –1 ) 

Yeast dry wt. (g l –1 ) 

(c) 

5 

Yeast dry wt. 

1.06 

4 

1.05 

3 

2 

Suspended 

cell count 

1.04 

1.03 

1.02 

1 

PG 

1.01 

0 

0 50 100 

Time (h) 

150 

1.00 

200 

60 

40 

20 

Free amino nitrogen 

(mg l –1 ) 

(d) 

Total higher alcohols 

(mg l –1 ) 

Dissolved 

oxygen 

Temp. 

(Ethanol) 

0 

0 50 100 

Time (h) 

150 

0 

200 

280 

240 

200 

160 

120 

80 

40 

FAN 

Pyruvate 

20 

15 

10 

5 


PG 

60 

40 

20 

0 

60 

40 

20 

0 

Suspended cell count 

(×10 6 ml –1 ) 

Fig. 12.1 Changes in present gravity (PG), yeast dry weight, suspended yeast cell count 

(a); ethanol concentration, temperature, dissolved oxygen concentration (b); free amino nitrogen 

concentration (FAN), pH, pyruvate concentration (c); total vicinal diketones (VDK), total higher 

alcohols, total esters (d) during the fermentation of a high-gravity (15 ëPlato) lager wort performed 

in a 1600 hl cyclindroconical fermenter. The fermentation was maintained at a temperature of 12 ëC. 

The process was terminated after approximately 150 h by cooling the green beer to 3 ëC. 

5.0 

4.8 

4.6 

4.4 

pH 

4.2 

0 

0 50 100 

Time (h) 

150 

4.0 

200 

80 

60 

H. alcohols 

Esters 

2.4 

2.0 

1.6 

40 

1.2 

0.8 

20 

VDK 

0.4 

0 

0 50 100 150 

0 

200 

Time (h) 


pH 

Total VDK (mg l –1 ) 

120 

80 

40 

0 

20 

15 

10 

5 

0 

Pyruvate (mg l –1 ) 

Total esters (mg l –1 ) 

(Ethanol) (g l –1 )

in pitching yeast are acause of inconsistencies in the extent of growth during subsequent 

fermentation. Ipso facto, the conditions under which pitching yeast is stored should be 

such that the opportunity for physiological variability is minimized. 

The physiological state of pitching yeast is an important determinant of fermentation 

performance. Compared with other fermentations, pitching rates are comparatively high 

and growth extents are modest. The pitched yeast plays an active role in subsequent 

fermentation and aproportion may persist through to the crop and be subject to further 

rounds of storage and re-pitching. In amodern brewery, it is usual to limit the number of 

serial fermentations. When the permitted number of generations is reached, the yeast is 

discarded and anew pure culture is introduced into the brewery (Chapter 11). This 

practice necessitates the use of apure yeast culture plant and asystem for maintaining 

and cultivating reference cultures in the laboratory. Afeature of this approach is that a 

number of yeast `lines' of varying generational age will be in use at any given time. 

The influence of serial cropping and re-pitching on the consistency of fermentation 

performance and beer composition is unclear. Yeast cells undergo aprocess of ageing, 

senescence and death. The ageing process in yeast is associated with agradual disruption 

of many metabolic processes (Jazwinski, 1999). Depending on the type of fermenter, 

serial fermentation can select for asub-population enriched with elderly cells. 

12.2 Yeast metabolism ±an overview 

Metabolism is the sum of all the chemical processes occurring in a cell. The 

manifestations of metabolism are the disappearance of nutrients from the medium and 

the appearance of by-products, of heat, the growth of individual cells and cell 

proliferation. These processes are accomplished by sequences of individual chemical 

reactions, which together form pathways. Each reaction is catalysed by functional 

proteins, termed enzymes. Metabolism is divided into two areas. Catabolism includes 

those pathways in which organic molecules are degraded with the liberation of energy. 

Anabolism is that part of metabolism in which the energy formed by catabolic pathways 

is utilized to fuel the synthetic reactions required for cellular growth and multiplication. 

Enzymes and other proteins are synthesized as aresult of the expression of individual 

genes (Chapter 11) 

Carbohydrates are the preferred sources of carbon and energy in yeast. The oxidation 

of carbohydrates liberates energy, furnishes reducing power and generates carbon 

intermediates. Some carbon intermediates, together with other non-carbohydrate 

nutrients, are utilized in anabolic metabolism to generate cellular biomass and byproducts. 

The energy is partly conserved in the form of the high-energy phosphate bonds 

of a group of metabolites, principally adensosine triphosphate (ATP, 4.53). Cleavage of 

these bonds liberates the stored energy and this is used to drive processes such as active 

transport and anabolic metabolism and to generate heat. 

Reducing power is transferred using the pyridine dinucleotide coenzymes, 

nicotinamide adenine dinucleotide (NAD + , 4.54) and to a lesser extent nicotinamide 

adenine dinucleotide phosphate (NADP + , 4.54). These compounds function as electron 

acceptors in reactions with oxidoreductase enzymes. In these reactions, two hydrogen 

atoms are removed from the substrate. One is released as a hydrogen ion and the second 

is transferred as a hydride ion to the nicotinamide portion of the coenzyme. The resultant 

reduced coenzyme (NAD(P)H) is then available to reduce a substrate and the oxidized 

form of the coenzyme is regenerated. 


Glucose 

Glycerol 

Triosephosphate 

NAD 

Lactate 

Pyruvate 

+ NADH + H + 

NADH + H + 

CH2 OH 

CH OH 

CH2 OH 

NADH 

Acetyl-CoA 

TCA cycle 

NAD + NADH + H + 

NADH + H + 

NAD + 

CH3 H 

C COOH 

CH3 CO COOH 

OH 

CH 3 

COO – 

Acetate 

NADH + H + 

NADH 

NAD + 

Acetaldehyde 

NAD + 

NADH 

Ethanol 

CH 3 CH2 OH 

Electron transport 

NADH + H + 

Amino acids 

α-Keto acids 

NADH + H + 

Higher alcohols 

α-Acetolactate 

Diacetyl 

NADH + H + 

Acetoin 

2,3-Butanediol 

An essential of cellular metabolism is the requirement to maintain redox balance. The 

oxidation reactions of carbohydrate dissimilation generate reduced pyridine nucleotides. 

The cell has afinite pool of pyridine nucleotides. In order to maintain activity of the 

glycolytic pathways the cell must ensure asupply of oxidized pyridine nucleotides. The 

ways in which this is accomplished do much to explain why certain products of 

metabolism accumulate (Fig. 12.2). Thus, in fully aerobic oxidative growth, NADH is 

reoxidized via the electron transport chain, which indirectly drives oxidative phosphorylation 

(Section 12.5.4). In this case the terminal electron acceptor is oxygen and water is 

formed. During fermentative growth, the oxidative pathways are inoperative and NAD + 

has to be regenerated by the reduction of acetaldehyde to ethanol (Section 12.5.5). 

Aproportion of the carbon metabolite flow through the carbohydrate dissimilatory 

pathways is diverted into anabolic biosynthetic pathways and so this material is not 

availabletotheterminalredox-balancingreactionsandothermechanismscomeintoplay. 

The most often quoted is the formation of glycerol, via glycerol phosphate from 

dihydroxyacetone phosphate, although the formation of this metabolite is also considered 

aresponsetoosmoticstress(Section12.3.1).Duringbreweryfermentations,severalother 

options are also possible. For example, the terminal reductive step in the formation of 

higher alcohols from aldehydes and ketones derived from amino acids and the reduction 

of vicinal diketones. Most of these products of fermentation are important contributors to 

beer flavour. 

Metabolism is highly regulated. Control is exerted by the regulation of enzyme and 

protein synthesis (gene level) and enzyme activity (phenotypic level). Spatial 

considerations are also important. Individual metabolic pathways are commonly localized 

in specific intracellular membrane-bound compartments. Transport of substrates to and 

from these cellular compartments can control the activity of these pathways. Enzymes 

within a given cellular compartment probably have specific spatial relationships with one 

another and this too has regulatory significance. 

CH 3 

NAD + 

CHO 

CH3 C C CH3 

O O 

CH 3 

O2 

H2O 

H 

C C 

OH O 

CH3 

CH 3 

NAD + 

CH 3 

NAD + 

NADH + H + 

NAD + 

H H 

C C CH3 

OH OH 

C 

CH3 C COOH 

O OH 

Fig. 12.2 Some redox-balancing reactions utilizing the coenzyme, nicotinamide adenine 

dinucleotide (NAD + ). For clarity many reactions in the sequences of reactions have been omitted. 


Control at the phenotypic level is achieved by varying enzyme concentration and 

activity. Enzyme activity can be modulated by metabolites or proteins that exert 

stimulatory or inhibitory effects. These metabolites can be substrates or products of the 

enzyme in question. Alternatively, metabolites that are neither substrates nor products 

can alter enzyme activity by bringing about structural alterations, a process termed 

allosteric modulation. 

Some genes, termed constitutive, are always expressed and their protein products are 

present in the cell under all conditions. Other genes are expressed only under certain 

conditions. The control of the expression of these inducible genes may be simple. For 

example, the presence of agiven nutrient in the medium, e.g., maltose, commonly 

induces the expression of the genes which encode for the carriers (permeases) required 

for its transport into the cell and the enzymes involved in its subsequent utilization. 

Conversely, the accumulation of anutrient that is the product of an anabolic pathway 

often causes repression of the gene's coding for the proteins required for its synthesis. 

Othercontrolmechanismsinfluencebothgeneexpressionandenzymeactivityinacoordinated 

fashion. These metabolic signal transduction systems have far-reaching effects 

on metabolism and are of fundamental importance in explaining why and how yeasts 

respond to different environmental stimuli (Thevelein, 1994). In these metabolic control 

systems one, or asmall number, of exogenous triggering chemical compounds in the 

medium interact with specific targets within yeast cells. This initial interaction sets in 

motion aseries of metabolic events in which several enzymes and transport permeases 

may be activated or inactivated. Simultaneously, the expression of several genes may be 

induced or repressed. Thus, exposure of yeast to asingle metabolite has the potential to 

alter several facets of its physiology. Conversely, a number of different exogenous 

triggers may activate a common pathway. Several signal transduction pathways are 

known tooperate inyeastcells and,nodoubt,morewillbeidentified inthe future. Often, 

there is overlap between signal transduction pathways. The regulation of individual 

pathways is organized in ahierarchical fashion such that activation and inhibition is 

ordered with respect to time. In this way, the cell organizes metabolic activity to suit any 

particular set of environmental conditions. 

Signal transduction pathways are commonly mediated by the phosphorylation or 

dephosphorylation of target proteins in response to the activities of specific protein 

kinases and phosphoprotein phosphatases (Stark, 1999). The attachment of aphosphate 

group to susceptible proteins leads to areversible structural modification in which the 

catalytic activity can be eliminated or enhanced. Often, the direction of flow through a 

metabolic pathway is regulated by pairs of enzymes whose activity is modulated by 

phosphorylation or dephosphorylation (Fig. 12.3). The importance of protein 

phosphorylation in the control of cellular metabolism is emphasized by the fact that 

some 120 protein kinases have been identified in S. cerevisiae, accounting for 2% of the 

genome. 

12.3 Yeast nutrition 

Some chemical components present in the wort or other medium surrounding yeast cells 

may be used as nutrients, some may be toxic or growth-suppressing, others may have no 

effect whatsoever. In some cases, the same component may be a nutrient at a low 

concentration or toxic at a higher concentration. Some substances are assimilated only 

under particular cultural conditions. Major classes of nutrients such as sources of carbon 


Protein kinase 

Degradation 

Protein kinase 

ATP 

ADP 

ADP 

ATP 

E1 

(inactive) 

E1P 

(active) 

E2P 

(inactive) 

E2 

(active) 

Phosphoprotein 

phosphatase 

Synthesis 

Phosphoprotein 

phosphatase 

Fig. 12.3 A representation of a system for the regulation of flow through a metabolic pathway, in 

the direction of biosynthesis or breakdown, using reversible phosphorylation and dephosphorylation. 

E 1 ˆ enzyme 1, E 1P ˆ Phosphorylated enzyme 1, E 2 ˆ enzyme 2, E 2P ˆ phosphorylated 

enzyme 2. 

and nitrogen are assimilated in an ordered fashion. Thus, where several sources of carbon 

and nitrogen are present yeast first utilizes those which are most readily assimilated. 

12.3.1 Water relations 

All yeasts require an aqueous medium for growth. The concentration of available water 

elicits different responses in individual strains. The available water, or water activity (aw), 

is an inverse function of the concentration of solutes present in the medium. Media with 

low water activities (high solute concentrations) are stressful and only a limited number 

of yeast strains can tolerate these conditions. It is difficult to distinguish between the 

effects due solely to water availability and those in which the nature and concentration of 

the solutes are the predominant influences. A solute may exert a specific biochemical 

influence on the physiology of the yeast. Neutral solutes exert a general osmotic stress, 

whereas charged species introduce an additional electrochemical factor. The way in 

which the stress is applied is also important. A very rapid change in the water activity of 

the medium has the greatest potential for damage. 

The multiplicity of mechanisms by which water activity may influence yeast 

physiology has resulted in a complicated and confusing system of nomenclature. Strains 

which can grow in media with low water activities and where this parameter alone is 

considered to be the major influence, are said to be xerotolerant. The term osmotolerance 

is used where the osmotic effect due to dissolved solutes is considered most important. 

Both of these terms describe strains that can tolerate extreme conditions but prefer a 

medium with a more moderate water activity. The descriptor, osmoduric is also used for 

such strains. Rarely, some strains grow best in media with a low water activity. These are 

classified as being osmophilic or xerophilic. 

Yeast strains, which can tolerate low water activity conditions, include Debaromyces 

hansenii, Hansenula anomola, Pichia ohmeri, Schizosaccharomyces pombe, Torulopsis 

candida, Zygosaccharomyces rouxii and Zygosaccharomyces bisporus (Rose and 

Harrison, 1995). In terms of brewing, all of these are `wild' yeast strains. 

Some marine strains of Metschnikowia have been described which are reportedly 

obligate osmophiles (Phaff et al., 1978). Z. rouxii is of importance in brewing since it is 

the most common cause of microbial spoilage of concentrated sugar syrups. S. cerevisiae 


Pi 

Pi

is not considered to be xerotolerant. This `deficiency' places an upper limit on the 

concentration of wort that can be used. However, S. cerevisiae can withstand 

dehydration. Yeast used for baking and wine making is routinely supplied in adried 

form and this technique has now been extended to brewing yeast (Fels et al., 1999). The 

dehydration process results in morphological changes. The cells take on ashrunken 

appearance and the plasma membranes develop deep invaginations (Rapoport et al., 

1995). These changes are reversed during rehydration. Survival rates are low and for 

successful drying the yeast musthave been cultivatedunder aerobic fed-batch conditions. 

These cultural conditions are associated with elevated intracellular concentrations of 

trehalose and sterols, both of which probably protect membrane integrity during 

dehydration (Sections 12.5.7 and 12.7.3). 

Yeast responds to changes in the water activity of the suspending medium by transient 

changes in cellular volume. This depends on acombination of the inherent elasticity of 

the cell wall and an uncontrolled adjustment in the volume of intracellular water, which 

occurs in response to alterations in external osmotic potential (Marechal et al., 1995). 

Following ahyperosmotic shock (suspension in amedium with an osmotic potential 

higher than intracellular contents), the cells contract then there is agradual recovery 

period during which the cellular volume increases and returns to the initial value. 

Osmotolerant strains are more resistant to uncontrolled water loss following 

hyperosmotic shock and better able to control cellular volume (Walker, 1998). 

Xerotolerant strains withstand the conditions of low water activity by altering 

intracellular conditions. This is accomplished by the synthesis of neutral polyols such as 

arabitol, mannitol, erythritol, sorbitol and especially glycerol. These are termed 

compatible solutes. The physiological adaptation to hyperosmotic shock is termed the 

osmostress response. Intracellular accumulation of compatible solutes reduces the 

difference in osmotic potential between the interior of the cell and the environment. The 

synthesis of compatible solutes allows the re-establishment of cell volume and aids the 

stabilization of enzymes and membrane proteins and phospholipids (Mager and Varela, 

1993). Xerotolerant strains are able to retain compatible solutes within the cell. Less 

xerotolerant types, including S. cerevisiae, lack this ability and some leakage occurs. S. 

cerevisiae also differs from more xerotolerant species in that following ahyperosmotic 

shock, the accumulation of glycerol is preceded by the formation of trehalose and 

accumulation of potassium and sodium ions (Singh and Norton, 1991). 

The release or retention of glycerol, in response to changes in osmotic potential, is 

regulated.InZ,rouxii,aspecificactiveglyceroltransporterhasbeenidentified(VanZyl 

et al., 1990). In S. cerevisiae, the presence of aplasma membrane channel has been 

described through which the passage of glycerol is controlled in response to the osmotic 

potential of the medium (Luyten et al., 1995). Glycerol accumulates in beer during 

fermentation, typically 1±2g/l being formed. The osmotic stress response in S. 

cerevisiae is mediated by aspecific sensing pathway termed the HOG (high osmolarity 

glycerol) signal transduction pathway. Two separate osmosensors located in the plasma 

membrane activate a common MAP (mitogen-activated protein) kinase signal 

transduction cascade. In response to the osmotic shock, ametabolic signal is passed 

through the kinase cascade, which culminates in the transcription of genes that encode 

enzymes including those responsible for glycerol synthesis (Maeda et al., 1994). The 

HOG signal transduction pathway activates genes, which contain an element termed 

STRE in their promoter regions. The STRE element is present in many genes associated 

with responses to stresses other than osmotic shock. This general stress response is 

discussed in Section 12.9. 


12.3.2 Sources of carbon 

Yeasts, as a group, can assimilate a comparatively wide range of organic carbon 

compounds. An assessment of the range of such compounds that are utilized by individual 

strains is used as adiagnostic criterion in yeast taxonomy (Barnett et al., 1990; Chapter 

11). The most commonly used carbon sources are carbohydrates, including mono-, di- and 

trisaccharides, higher dextrins and starches. Some species can utilize pentoses, although 

not brewing yeast strains. Growth may be oxidative and/or fermentative depending on the 

strain and the cultural conditions. Less usual carbon sources are used by certain species 

such as methanol and both aliphatic and aromatic hydrocarbons. 

Strains of S. cerevisiae utilize a limited repertoire of carbon sources for growth (Table 

12.1). Differences in the patterns of utilization are strain-specific. Ale strains lack the 

ability to utilize the disaccharide, melibiose. Lager strains can grow on melibiose because 

they have -D-galactosidase activity, which hydrolyzes it to galactose and glucose 

(Barnett, 1981). There are other differences between ale and lager strains. The latter 

utilize maltotriose more rapidly than ale strains (Stewart et al., 1995) and lager strains are 

more efficient at assimilating galactose. Lager strains utilize mixtures of galactose and 

maltose simultaneously, whereas ale strains assimilate maltose preferentially (Crumplen 

et al., 1993). The yeast, S. cerevisiae var. diastaticus can utilize dextrins, albeit with the 

chemically unrelated formation of styrene and 4-vinyl guaiacol which impart a phenolic 

medicinal taint to beer (Ryder et al., 1978). 

Many metabolic intermediates accumulate in the growth medium. These include 

pyruvate (4.146), acetaldehyde and several organic acids such as citric (4.153), malic and 

acetic. Some of these may be re-assimilated at a later stage in growth. The major products 

of fermentative growth are ethanol, glycerol, lactic acid and carbon dioxide. A proportion 

of the latter is fixed in a series of carboxylation reactions. Up to 5% of the carbon 

Table 12.1 Utilization of carbon sources by S. cerevisiae under aerobic and anaerobic conditions 

Carbon source Aerobic Anaerobic 

D-Glucose All strains All strains 

Cellobiose ± ± 

Ethanol Some strains ± 

D-Galactose Some strains ± 

D-Glucitol ± ± 

Glycerol Some strains ± 

Inulin ± ± 

DL-Lactate Some strains ± 

Lactose ± ± 

Maltose Some strains Some strains 

D-Mannitol Some strains ± 

Melezitose Some strains Some strains 

Melebiose Some strains Some strains 

Methanol ± ± 

Methyl- -D-glucopyranoside Some strains Some strains 

Raffinose Some strains Some strains 

L-Sorbose ± ± 

Starch Some strains Some strains 

Succinic acid Some strains ± 

Sucrose Some strains Some strains 

Trehalose Some strains Some strains 

D-Xylose ± ± 

Xylitol ± ± 


equirement of yeast is provided by carbon dioxide fixation (Oura et al., 1980). In the 

presence of oxygen, some strains can undergo ametabolic adaptation, termed the diauxic 

shift, and utilize ethanol, lactic acid and glycerol for oxidative growth. 

12.3.3 Sources of nitrogen 

Yeasts cannot assimilate gaseous nitrogen, however, simple inorganic sources such as 

ammonium salts may be readily utilized. Adiverse range of organic sources of nitrogen 

can be assimilated (Soumalainen and Oura, 1971) including amino acids, peptides, 

amines, pyrimidines and purines. Many of these, for example, amines, are utilized as a 

source of nitrogen only in the presence of additional sources of carbon and energy. The 

ability, or inability, to use a specific organic nitrogen source can have taxonomic 

significance. Saccharomyces yeasts cannot utilize nitrate or nitrite but readily assimilate 

ammonium ions. In natural media, such as brewers' wort, ammonium ions, amino acids, 

peptides, purines and pyrimidines provide most of the nitrogen. These yeasts strains do 

not produce extracellular proteases and therefore, proteins are not utilized. 

12.3.4 Sources of minerals 

Sulphur may be assimilated from both inorganic and organic sources. The latter include 

the sulphur-containing amino acids, methionine (4.41) and cysteine (4.31). In addition, 

glutathione may be assimilated. The preferred inorganic source of sulphur is sulphate but 

sulphite and thiosulphate can also be utilized. S. cerevisiae can reduce elemental sulphur 

to thiol ions in the periplasm and thereafter these are assimilated (Rose, 1987). The 

requirement for phosphorus is satisfied by the assimilation of inorganic phosphate ions. 

Elemental mineral ions are essential co-factors for numerous enzyme activities. Some 

have structural roles and others are necessary components of transport systems where 

they fulfil acharge-balancing role. The concentrations required for growth are small, 

typically less than 10 M(Jones and Greenfield, 1984). Essential mineral ions include 

B + , Ca 2+ , Co 2+ , Cu 2+ , Fe 3+ , K + , Mo 2+ , Mn 2+ , Mg 2+ , Ni 2+ and Zn 2+ . Many ions, 

particularly those of heavy metals, are toxic in excess. Metal ions such as Na + ,in high 

concentrations, exert asalt stress on yeast (Section 12.3.1). Precise requirements are 

strain specific and the combination and concentration of mineral ions available in the 

medium is important since synergistic and antagonistic interactions occur. Some metal 

ions may be tolerated by certain yeast strains but be growth inhibitory to others. For 

example, a commonly used test to differentiate between brewing and non-brewing `wild' 

yeast strains is based on the ability of the latter to grow in the presence of relatively high 

concentrations of copper ions. A few ions, notably Mg 2+ and K + , are required at higher 

concentrations, typically at the millimolar level. 

Brewers' malt wort supplies all the mineral nutritional requirements of yeast, with the 

possible exception of zinc. Zinc ions can be chelated by wort amino acids, proteins and 

phytate and a proportion of these may be removed as insoluble precipitates during the 

copper boil (Daveloose, 1987). For this reason, zinc supplements are commonly added to 

wort in the fermenter. 

12.3.5 Growth factors 

Growth factors are a diverse group of organic compounds which individual yeast strains 

are unable to synthesize and which are essential for growth. Their presence in the 


medium is obligatory. The group includes vitamins, some purines, pyrimidines, 

polyamines, nucleosides, nucleotides and certain lipids. Growth factors may be required 

as intermediates in certain essential metabolic pathways or have astructural role. More 

commonly, they are essential catalytic components of coenzymes as exemplified by 

vitamins. Concentrations required for growth are low, typically in the M range. 

Vitamins, or derivatives of them, are involved in many fundamental biochemical 

processes, for example, biotin (carboxylation reactions), thiamine (oxo-acid decarboxylation 

reactions), nicotinic acid (redox reactions), pyridoxine (transaminations), paminobenzoic 

acid (one-carbon transfer) and pantothenic acid (acetylation reactions) 

(Fig. 4.29). The requirement for vitamins and other growth factors varies widely between 

individual strains. Individual strain requirements have taxonomic significance (Kregervan 

Rij, 1984; Kurtzman and Fell, 1988). 

With regard to brewing strains of S. cerevisiae, all require biotin and pantothenic 

acid. Many strains require inositol and thiamine. In general, lager strains have agreater 

requirement for growth factors compared to ale strains. Most brewers' worts contain 

adequateconcentrationsofgrowthfactorsforallyeaststrains.Inrarecases,anutritional 

supplement, enriched in growth factors, is added to worts that are considered deficient. 

Such supplements, termed `yeast foods', are partially purified extracts of yeast, 

occasionally with added vitamins. Under anaerobic conditions Saccharomyces strains 

become auxotrophic (have aabsolute requirement for growth) for certain lipids. Thus, 

some sterols and unsaturated fatty acids, accumulated during prior aerobic growth or 

supplied in the medium, are essential components of membranes. In addition, 

unsaturated fatty acids appear to have roles as regulatory effector molecules. The 

syntheses of sterols and unsaturated fatty acids, de novo, require molecular oxygen 

(Section 12.7). 

12.4 Nutrient uptake 

Yeasts possess mechanisms to regulate the passage of nutrients from the external medium 

into the cell (Cartwright et al., 1989). The plasma membrane forms the principal semipermeable 

barrier through which all nutrients must pass. Cells have systems for sensing 

the nature and concentration of nutrients in the external medium. They have the ability to 

selectively assimilate individual compounds from complex mixtures in an ordered 

manner. Some nutrients can be transported into the cell against a concentration gradient. 

In addition to assimilating compounds from the external medium, some of the products of 

metabolism are excreted from the cell. 

Cartwright and co-workers (1989) differentiate between vectoral and scalar 

metabolism. The latter is that portion of metabolism in which molecules are subject to 

chemical modification with no appreciable movement. The former is defined as that 

portion of metabolism that involves the controlled physical movement of molecules. 

Some vectoral metabolism is intracellular, for example, the controlled transport of 

metabolites between intracellular compartments. Although it is convenient to consider 

these two facets of metabolism as separate entities, they are, of course, regulated and coordinated 

processes. 

A number of distinct mechanisms result in the passage of solutes across membranes. 

The simplest and slowest is free diffusion in which molecules traverse the membrane via 

the lipid components driven by a concentration gradient. Transport ceases when the 

solute concentration on each side of the membrane becomes equal. Other solute transport 


processes require the intervention of specific membrane proteins, termed transporters or 

permeases. Facilitated diffusive transporters regulate the passage of solutes in the 

direction of a concentration gradient. As in the case of free diffusion, transport ceases 

when the concentration gradient ceases to exist. 

The most prevalent type of transport system in yeast is that in which solutes passage 

is against a concentration gradient. The process is termed active transport since it 

requires metabolic energy. Movement of individual solutes is controlled by specific 

permeases, which may be inducible or constitutive. The energy is provided by the 

plasma membrane H + -ATPase (Serrano, 1989). The latter energizes the membrane by 

promoting the uni-directional movement of protons utilizing the energy released by the 

hydrolysis of ATP. The resultant proton motive force is a combination of the membrane 

potential and the proton gradient. Solute molecules usually travel across the membrane 

accompanied by protons, a mechanism termed proton symport transport. Occasionally, 

solute movement occurs against a reverse passage of protons, termed proton antiport 

transport. 

Commonly, yeast cells possess multiple carriers for the same nutrient, or class of 

nutrients. Frequently, these carriers have different affinities for the same substrate. Thus, 

within the same strain there may be both high and low affinity transporters for given 

substrates. Often the latter is constitutive and the former inducible. Presumably, the high 

affinity inducible systems represent an evolutionary mechanism, which confers a 

selective advantage where mixed populations of yeast and other micro-organisms are 

competing for small amounts of essential nutrients. 

Yeast membranes, including the plasma membrane and others enclosing intracellular 

organelles, possess channels which have been implicated in the transport of ions, water 

and some organic molecules, such as glycerol (Gustin et al., 1986; Luyten et al., 1995). 

The channels are proteinaceous in nature and are activated or deactivated by 

perturbations in membrane polarization. The predominant ion channel controls the 

efflux of K + . This phenomenon balances the influx of protons that accompany sugar 

uptake and is, therefore, necessary for intracellular charge homeostasis. Solute molecules 

may be transported enclosed in a vesicle, which arises from the plasma membrane. This 

process is termed pinocytosis. Its occurrence in yeast is disputed, although it may be 

implicated in the intracellular trafficking of macromolecules. 

The transport systems described so far facilitate the entry of solutes without chemical 

modification. Group translocation systems mediate the uptake of solutes and in so doing 

modify chemical structures. In yeast, glucose uptake is coupled to phosphorylation. This 

process may be associated with the glucose transporters (Lagunas, 1993). 

Before solutes pass through the plasma membrane and into the cell, they must 

negotiate the capsule, if present, the cell wall and the intervening periplasm. Yeast cell 

walls are freely permeable but their physical structure places an upper limit on the size of 

molecules that can pass through unimpeded. This feature is of benefit in that it allows 

retention of enzymes in the periplasm. The upper limit of cell wall porosity is in the 

region of 200±400 kDa (de Nobel et al., 1991). The comparatively wide range reflects 

changes in cell wall porosity that occur during different phases in the yeast cell growth 

cycle. 

12.4.1 Sugar uptake 

The uptake of sugars by Saccharomyces strains have been subject to the closest scrutiny 

as befits their role in industrial fermentations (Andre, 1995; Horak, 1997). Although 


g l –1 

g l –1 

10 

8 

6 

4 

2 

100 

80 

60 

40 

20 

Glucose 

0 

0 20 40 60 80 100 

Time (h) 

Maltose 

Maltotriose 

Fructose 

Sucrose 

0 

0 20 40 60 80 100 

Time (h) 

Fig. 12.4 Patterns of uptake of fermentable sugars from 10 ëPlato ale wort. The yeast was a top 

fermenting ale strain and the temperature was 18 ëC (C. A. Boulton, unpublished data). 

some sugars may enter the cell by free diffusion, uptake appears to be predominantly via 

active processes and is against aconcentration gradient. Control of sugar uptake is 

complex and highly regulated. In industrial fermentations, using feedstocks containing 

complex mixtures of several sugars, uptake into the cell limits the overall rate of ethanol 

formation. 

When presented with amixture of assimilable sugars, yeasts have mechanisms for 

selecting first, those which are most readily utilized. In the case of brewers' wort, the 

utilization of sugars is an ordered process (Fig. 12.4). Sucrose is hydrolysed by an 

invertase that is secreted into the periplasm. This results in atransient increase in the 

concentrations of fructose and glucose. Fructose and glucose are assimilated 

simultaneously. The predominant sugar, maltose is then taken up. When the maltose 

concentration falls to an undetectable level maltotriose is assimilated. Longer chain 

sugars are not utilized by brewing yeasts. 

Amultiplicity of hexose uptake systems has been identified in various yeast strains 

(Kruckenberg, 1996; Ozcan and Johnston, 1999). Glucose is the preferred substrate but 

frequentlyotherhexoses are also transported.In S.cerevisiae, atleast 19separate genes are 

responsible for the synthesis of hexose carriers. Two classes of hexose carrier are 

recognized,termedhighandlowaffinity.ThehighaffinitysystemstransportD-glucose,DfructoseandD-mannoseandareofthefacilitateddiffusionvariety.TheyhaveK 

mvaluesof 

approximately 1mM. In the presence of relatively high glucose concentrations (>0.1M) 

activityisrepressed(DoesandBisson,1989).Thisphenomenonispartofawidersystemof 

metabolic control, termed catabolite repression (Section 12.5.8). Activity of the high 

affinity carriers is associated with phosphorylation of the substrate via the glycolytic 

hexokinases and a glucokinase. Whether, or not, the phosphorylation involves interactions 

between kinases and the glucose carrier remains to be elucidated (Lagunas, 1993). 


The low affinity hexose carriers are constitutive and have Km values in the region of 

20mM. Their existence has been disputed since it has been claimed that low affinity 

glucose uptake could simply reflect passive diffusion. However, this suggestion has 

been refuted on the basis that uptake rates are 2±3 times higher than would be predicted 

for passive diffusion (Gamo et al., 1995). The reason(s) for the multiplicity of hexose 

carriers is not clear. Detailed genetic analysis is needed to determine under what 

conditions each is active. It has been demonstrated that some of the carriers are subject 

to nitrogen catabolite inactivation (Busteria and Lagunas, 1986; Section 12.8). This 

process occurs following the exhaustion from the medium of certain nitrogen sources 

and leads to inactivation of hexose and other sugar uptake systems via proteolysis of 

the carriers. Probably other global metabolic control mechanisms, such as the 

availability of oxygen and other nutrients, will also influence the activity of individual 

hexose carriers. 

Although the glucose transporters show activity towards other hexoses, S. cerevisiae 

also possesses a specific transport system for galactose. In common with the glucose 

uptake system, both constitutive low affinity and inducible high affinity galactose 

transporters exist. There are specific uptake systems for pentoses, in those yeast strains 

capable of utilizing them. In general, the activity of these is repressed by glucose. 

Sucrose assimilation is dependent on the production of invertase. Synthesis of this 

enzyme is dependent on the presence of a number of SUC genes (SUC1±SUC5 and 

SUC7). Yeast cells produce an intracellular constitutive invertase, whose function is 

unknown. A second invertase is secreted into the periplasm and this is responsible for the 

assimilation of extracellular sucrose. The enzyme is also active towards raffinose. The 

fructose and glucose produced by the hydrolysis of sucrose are both transported via the 

hexose carrier system. The periplasmic invertase is subject to glucose repression 

although, surprisingly, for maximum transcription of SUC2 to occur a low concentration 

(0.1%) of glucose is required (Ozcan et al., 1997). 

Specific disaccharide carriers mediate the uptake of maltose (4.4) and trehalose ( -Dglucopyranosyl 

(1, 1)- -D-glucopyranose). A constitutive low affinity transport system 

and a high affinity proton symport carrier accommodate trehalose uptake. The latter is 

subject to glucose repression. Derepressed yeast cells have high levels of intracellular 

trehalose. This suggests that under these conditions trehalose accumulation is against a 

concentration gradient. Probably the cell does this since there is evidence that trehalose 6phosphate 

is involved in the control of glycolysis (Thevelein and Hohmann, 1995). 

Maltose is the most abundant sugar in wort. Its uptake and utilization is controlled by a 

complex series of MAL genes. These occur at five unlinked homologous loci, not 

restricted to a single chromosome. Each locus consists of three genes, which encode for a 

maltose carrier, maltase and a post-transcriptional regulator of the carrier and maltase 

genes. Maltose utilization is repressed by glucose and requires the maltose for induction 

(Busteria and Lagunas, 1986). Maltose utilization is subject to nitrogen catabolite 

inactivation. Thus, under conditions of nitrogen exhaustion or in the presence of glucose, 

the maltose permease is irreversibly inactivated via the action of a protease. Maltose 

uptake is an energy-requiring proton symport process. Efflux of K + ensures 

electrochemical neutrality. The uptake system is of the high affinity variety. In brewing 

strains of S. cerevisiae, a second low affinity system has been identified (Crumplen et al., 

1996). Transport of maltotriose in S. cerevisiae is accomplished by a constitutive 

facilitated diffusion carrier. It has been suggested that the permease may be absent in 

some ale strains thus accounting for the observations that some of the latter are not able to 

utilize maltotriose (Stewart et al., 1995). 


12.4.2 Uptake of nitrogenous nutrients 

Yeasts possess transport systems for mediating the uptake of both inorganic and organic 

nitrogen sources. In wort, yeast is presented with acomplex mixture of nitrogen sources. 

As is the case with the utilization of carbon sources, uptake of nitrogenous nutrients is an 

ordered process. Thus, the presence in the medium of readily assimilable nitrogen sources 

represses the synthesis of the uptake systems and catabolic enzymes of other less readily 

utilized sources of this nutrient. This is termed nitrogen catabolite repression (Wiame et 

al., 1985). 

All yeasts can utilize ammonium ions and indeed this nutrient is usually utilized in 

preference to organic sources of nitrogen. However, in brewing yeast, during 

fermentation, several amino acids are utilized before ammonium ions. In S. cerevisiae 

high and low affinity carriers control the uptake of ammonium ions. The high affinity 

carrier is an active transport system and requires the presence of an oxidizable substrate 

for activity. S. cerevisiae does not utilize nitrate or nitrite. Individual yeast strains can 

assimilate awide range of organic sources of nitrogen. Most strains can utilize urea and 

in S. cerevisiae high and low affinity urea transporters occur. Anumber of transporters 

occur in yeast specific for one or small groups of amino acids. In addition, there is a 

general amino acid permease (GAP) with broad specificity. Some of the transporters, 

including GAP, are repressed by ammonium ions, asparagine and glutamine (Grenson, 

1992). In S. cerevisiae, 12 constitutive and four nitrogen-repressible amino acid carriers 

have been identified (Horak, 1986). The transporters can be of the high or low affinity 

type, uptake is active and involves proton symport. 

Maximum activity of GAP occurs only under conditions of nitrogen starvation. In this 

regard, GAP functions as anitrogen scavenger. Regulation of theother carriers is complex 

and dependent on the spectrum and concentrations of amino acids present in the medium. 

The presence of multiple carriers for amino acids affords the yeast an opportunity to order 

uptake in response to need. Based upon chemostat studies with S. cerevisiae growing 

underconditionsofnitrogenlimitation, ithasbeensuggested(Oliveraetal.,1993)thatthe 

specific permeases are involved in the uptake of amino acids destined for use in anabolic 

pathways.Conversely,thosepermeasessubjecttonitrogencataboliterepression,including 

GAP, mediate the uptake of amino acids used in catabolic pathways. 

In brewing yeasts growing on wort, the uptake of amino acids is an ordered process. 

Pierce (1987) divided amino acids into four classes, based on their order of assimilation 

from wort during fermentation (Table 12.2). Surprisingly, in view of its role in nitrogen 

catabolite repression, ammonia is not a member of the first group, although asparagine 

and glutamine are. The amino acids in classes A and B are required for anabolic 

metabolism, principally protein synthesis. They are taken up by those permeases that are 

not subject to nitrogen catabolite repression. Conversely, those in Class C are only taken 

up when the Class A amino acids have disappeared and nitrogen catabolite repression is 

relieved. Proline is the sole member of class D. This imino acid is not utilized during 

brewery fermentation since its oxidation requires a mitochondrial oxidase, which is 

repressed during fermentation (Wang and Brandriss, 1987). 

Short homopeptides may be taken up by yeast although not as readily as the free 

amino acids (Ingledew and Patterson, 1999). Ammonia inhibits the uptake of dipeptides. 

Peptides containing no more than five amino acid residues may be transported into the 

cell. The carrier in S. cerevisiae is reportedly of broad specificity, active and capable of 

transporting di- and tripeptides (Marder et al., 1977). S. cerevisiae strains are not capable 

of transporting oligopeptides and, since they do not produce exogenous proteases, are not 

able to utilize these nitrogen sources. 


Table 12.2 Amino acid classification based on the order of assimilation from wort during 

fermentation (Pierce, 1987). Amino acids are assimilated in the order A, B, C, D 

Class A Class B Class C Class D 

Arginine Histidine Alanine Proline 

Asparagine Isoleucine Ammonia 

Aspartate Leucine Glycine 

Glutamate Methionine Phenylalanine 

Glutamine Valine Tryptophan 

Lysine Tyrosine 

Serine 

Threonine 

Uptake of purines and pyrimidines is accommodated by both active proton symport 

systems and by facilitated diffusion carriers. The active systems reportedly mediate the 

uptake of adenine, adenosine, cytosine, guanine and hypoxanthine. Transport of uracil 

and uridine is apparently by facilitated diffusion (Cartwright et al., 1989). 

12.4.3 Lipid uptake 

Saccharomyces cerevisiae utilizes lipids such as fatty acids and sterols. These may be 

used for direct incorporation into cellular structures, as sources of metabolic 

intermediates for both catabolic and anabolic pathways or to fulfil roles in cellular 

signalling systems. At high concentrations, fatty acids are taken up by simple diffusion, a 

process aided by the lipophilic nature of the plasma membrane (van der Rest et al., 1995). 

In S. cerevisiae, the presence of a medium chain length fatty acid transporter has been 

inferred (Faergeman et al., 1997). In anaerobic brewery fermentations, brewing yeasts are 

auxotrophic for unsaturated fatty acids. These compounds must be synthesized during the 

aerobic phase of fermentation. Some of this requirement may be satisfied by direct uptake 

from wort. 

Yeast can assimilate exogenous sterols but only under anaerobic conditions when de 

novo synthesis is precluded. This phenomenon is termed aerobic sterol exclusion. Sterol 

uptake requires expression of the SUT1 and SUT2 genes (Bourdot and Karst, 1995). 

These are hypoxic genes, expressed only under anaerobic conditions. It is assumed that 

the product of their expression inhibits the transcription of the genes encoding the sterol 

transporter. 

12.4.4 Ion uptake 

Uptake of metal ions by yeast is a biphasic process. Firstly, ions are concentrated by 

attachment to the cell surface, a passive process termed biosorption. Suggested 

mechanisms for attachment to the cell wall include complexation, ion exchange, 

adsorption and precipitation (Blackwell et al., 1995). The process is independent of 

temperature, does not require metabolic energy or indeed viability. It has been suggested 

that it may serve as a protective strategy for the removal of potentially toxic ions. 

Secondly, ions are transported across the plasma membrane and into the cell by 

bioaccumulation. This is an active process involving proton symport and K + efflux. Once 

in the cell, metal ions are commonly compartmentalized in the vacuolar system. Metal 

ion uptake is tightly regulated since individual ions may be essential or toxic at low and 

high concentration, respectively. Different yeast species possess metal ion carriers of 


oth broad and narrow specificity. Metal ion transport is a dynamic process and 

synergistic and antagonistic effects are possible depending on the spectrum of ions 

present. Metal ion uptake and efflux occur reflecting the need of the cell to maintain an 

internal ionic balance and still supply ions where needed for proper enzyme function 

(Jones and Gadd, 1990). 

S. cerevisiae possesses a metal ion carrier with broad specificity, capable of 

transporting the divalent ions of calcium, cobalt, nickel, manganese, magnesium, 

strontium and zinc. The carrier has the highest affinity for Mg 2+ and uptake of this ion 

occurs simultaneously with phosphate. Uptake of Co 2+ ,Mn 2+ ,Fe 2+ and Zn 2+ ions in S. 

cerevisiae is mediated by distinct high and low affinity transporters. Activity of the 

individual carriers is dependent upon the availability of the ions and typically, the high 

affinity permeases have scavenging roles. Uptake of manganese is competitively 

inhibited by magnesium ions. The iron transporter also mediates the uptake of cadmium, 

cobalt and nickel. 

Zincuptakeisofparticularrelevancetobrewinginthatmalt worts maybedeficient in 

this ion (Daveloose, 1987; Chapter 4). It plays an essential role in the function of many 

enzymes, including alcohol dehydrogenase. S. cerevisiae maintains zinc homeostasis by 

the operation of carriers that mediate both uptake and efflux. Copper is an essential 

nutrient at low concentration but can be toxic at higher levels. The ability of certain 

strains of `wild' yeast to grow in the presence of copper at a concentration that brewing 

strains of S. cerevisiae cannot tolerate is of diagnostic importance. Copper uptake is 

mediated by a high affinity carrier expressed by the CTR 1 gene. The transporter may 

also be responsible for the uptake of ferrous ions. Copper homeostasis is achieved by 

regulated intracellular sequestration in the form of a metalothionin protein. In addition, 

transporters are present that function as mediators of copper ion efflux. 

Calcium ions have important roles in metabolic signalling systems. They are 

implicated in the control of the mating response (not applicable to brewing yeast), 

modulation of cellular growth and progress through the cell cycle. In consequence, 

intracellular calcium ion concentration is highly regulated (Youatt, 1993). An active 

proton antiport permease regulates calcium transport across both the plasma and vacuolar 

membranes. Calcium ions are sequestered in the cell by specific binding proteins such as 

calmodulin. 

Transport of potassium ions is used by yeast cells to maintain charge homeostasis. 

Frequently, efflux of potassium ions is used by yeast in conjunction with proton symport 

permeases, to maintain electrochemical neutrality. It is the most common intracellular ion 

in the yeast cytosol. The high intracellular concentration is maintained by several parallel 

transport systems. These include uptake via both active systems and plasma membrane 

pores. In S. cerevisiae, the active system is a proton antiporter. 

Phosphate uptake is dependent on the presence of a fermentable substrate and both 

high and low affinity transporters occur in S. cerevisiae. Under conditions of phosphate 

limitation, the PHO5 gene is induced. This encodes for a secreted acid phosphatase and 

this is involved in phosphate scavenging. In some yeasts, for example Zygosaccharomyces 

bailii, active carriers exist for the uptake of sulphate and sulphite. In brewing 

yeast, uptake of the latter is apparently via simple diffusion of sulphur dioxide, whereas 

several carriers mediate the uptake of sulphate. The presence of sulphite inhibits the 

utilization of sulphate. 

Movement of hydrogen ions into and out of the cell is of paramount importance in 

controlling the transport of other charged species and maintenance of intracellular pH. 

The trans-membrane proton gradient and proton-motive force is generated by 


H + -ATPase. The importance of the latter is indicated by the fact that it is the most 

abundant membrane protein (Serrano, 1989). 

12.4.5 Transport of the products of fermentation 

During fermentation, the transformation of wort into beer is accompanied by adecrease 

in pH. Much of this decrease is aresult of the proton antiport component of other uptake 

systems. Acidification of beer is also contributed to by the formation of carbonic acid 

derived from the carbon dioxide produced during fermentation. In addition, several 

organic acids, notably, lactic, citric, pyruvic, malic, acetic, formic, succinic and butyric 

acids are excreted from fermenting yeast cells. In the latter stages of fermentation, some 

of these compounds may be re-assimilated. Transport may be via simple diffusion or 

active uptake systems involving proton symport. 

Many products of yeast metabolism contribute to beer flavour (Chapter 23). These 

include ethanol, higher alcohols, esters, aldehydes and vicinal diketones. It follows that 

these metabolic by-products must be transported out of or released from the cell during 

fermentation. This subject has received little attention; probably non-concentrative 

diffusion is the mechanism used. 

12.5 Sugar metabolism 

The catabolism of sugars provides yeast with energy and carbon skeletons for anabolic 

pathways. This is an essential activity and in consequence, alarge proportion of total 

metabolism is devoted to it. Several distinct pathways are involved. The flux of carbon 

through individual pathways is influenced by yeast genotype and its phenotypic 

expression as brought about by the conditions to which they are exposed. 

12.5.1 Glycolysis 

Glycolysis, or the Embden-Myerhof-Parnas pathway, is the major sugar catabolic 

pathway in yeast. It operates under both aerobic and anaerobic conditions and is the route 

by which approximately 70% of exogenous hexose sugars are assimilated. The 

importance of this pathway is reflected in the fact that glycolytic enzymes comprise 

30±65% of the total soluble protein pool in S. cerevisiae. (Fraenkel, 1982). The pathway 

catalyses the conversion of one molecule of glucose into two molecules of pyruvate (Fig. 

12.5). The initial phosphorylation reaction, in which ATP is the phosphate donor, may be 

catalysed by one of three enzymes. Hexokinases 1and 2show activity towards both 

glucose and fructose and glucokinase with glucose, alone. All show activity towards 

mannose. All of the glycolytic reactions are reversible with the exceptions of the initial 

phosphorylation ofglucose,the phosphorylation offructose-6-phosphate toyieldfructose 

1,6bisphosphate and the dephosphorylation of phospho-enol-pyruvate to form pyruvate. 

Several of the steps are catalysed by multiple enzymes, as indicated. Glycolysis can 

operate in the reverse (gluconeogenic) direction. In this case, three additional enzymes, 

phospho-enol-pyruvate carboxykinase, fructose 1,6-bisphosphatase and glucose 6phosphate 

phosphatase catalyse the contra-flow of carbon past the irreversible steps of 

glycolysis. Other sugars feed into the glycolytic pathway as shown in Fig. 12.6. Like 

glucose, the utilization of these other sugars also involves reactions in which ATP is 

consumed and phosphorylated intermediates are formed. Some of the reactions use the 


HXK1 

HXK2 

GLK1 

PGI1 

PFK1 

PFK2 

CH2OH O 

H H 

H 

HO OH H OH 

H OH 

Dihydroxyacetone 

phosphate 

Oxaloacetate 

Glucose 

Glucose 6-phosphate 

CH2O P 

H 

O 

H 

H 

HO HO 

OH 

H OH 

Fructose 6-phosphate 

P OCH2 CH2OH O 

H HO 

H OH 

OH H 

CH2O P CO CH2OH Fructose 1,6-bisphosphate 

FBP1 

PCK1 

TPI1 

CH2O P CHOH COO P 

CH2O CHOH COO – 

P 

CH2OH COO – 

CHO P 

Glyceraldehyde 

3-phosphate 

1,3-Diphosphoglycerate 

3-Phosphoglycerate 

2-Phosphoglycerate 

Phosphoenolpyruvate 

Pyruvate 

coenzyme, uridine triphosphate (UTP), which gives rise to acarrier of sugar molecules 

(e.g. UDPG; 4.55). 

Glycolysis generates reducing power in the form of NADH. This is re-oxidized in 

redox balancing reactions (Section 12.2). During the conversion of glucose to fructose 

1, 6-bisphosphate, two molecules of ATP are consumed. In the later stages of glycolysis, 

four ATP molecules are generated in the reactions catalysed by phosphoglycerokinase 

ATP 

ADP 

ATP 

ADP 

P OCH2 CH2O P 

O 

H HO 

H OH 

OH H 

GPP 

FBP1 

CH2O P CHOH CHO 

COO – 

CH2 CO P 

CH 3 CO 

COO – 

NAD + 

NADH 

ADP 

ATP 

ADP 

ATP 

TDH1 

TDH2 

TDH3 

PGK1 

GPM1 

ENO1 

ENO2 

PYK1 

PYK2 

Fig. 12.5 The main glycolytic pathway. The genes and enzymes resposible for each step are: 

HXK1, HXK2, hexokinases; GLK1, glucokinase; PGI1, phosphoglucose isomerase; PFK1, PFK2, 

phosphofructokinase; FBP1, fructose 1, 6-bisphosphatase; FBA1, fructose 1, 6-bisphosphate 

aldolase; TPI1, triose phosphate isomerase; TDH1, TDH2, TDH3, glyceraldehyde 3-phosphate 

dehydrogenase; PGK1, phosphoglycerate kinase; GPM1, glycerophosphate mutase; ENO1, ENO2', 

enolase; PYK1, PYK2, pyruvate kinase. Three gluconeogenic enzymes are also shown: GPP, 

glucose 6-phosphate phosphatase; FBP1; fructose 1, 6-bisphosphatase; PCK1, phosphoenolpyruvate 

carboxykinase. 


Sucrose 

Maltose 

(4.4) 

Glucose 


Fructose 6-phosphate 

GLYCOLYSIS 

Fructose (4.3) 

Trehalose 

Trehalose 

phosphate 

Mannose 

phosphate 

Galactose (4.9) 

Galactose 

1-phosphate 

UDP-glucose 


Melibiose 

Raffinose 

Mannose 

(4.8) 

UDP-galactose 

Fig. 12.6 The mode of entry of various sugars into the glycolytic pathway. The numbers in 

parentheses indicate where the molecular structures may be found. 

and pyruvate kinase. Therefore, for every molecule of glucose catabolized there is anet 

gain of two molecules of ATP. This phenomenon is termed substrate level 

phosphorylation. It is the predominant mechanism used by yeast to generate energyrich 

compounds under fermentative conditions. Glycolysis is active in yeast under all 

conditions and the component enzymes are constitutive. During growth on sugars, the 

direction of carbon flow is from glucose to pyruvate. During growth on oxidative carbon 

sources such as ethanol, the glycolytic pathway is reversed and is used to generate 

intermediates for anabolism. In this respect, glycolysis/gluconeogenesis is an amphibolic 

pathway, which serves both anabolic and catabolic roles. Reverse glycolysis is part of 

gluconeogenic sugar generating metabolism (Section 12.5.6). 

Carbon flux through glycolysis is aregulated process and controls are exerted on both 

gene expression and enzyme activity. Transport of sugars into the cell, phosphorylation 

of glucose and regulation of the activities of phosphofructokinase and pyruvate kinase by 

metabolic effectors have all been implicated. Possibly uptake of sugars is the ratedetermining 

step in glycolysis since the maximum rates of transport are close to the 

maximum observed rates of glycolytic flux. The intracellular concentration of glucose is 

always lower than that in the external medium. Nitrogen starvation, which brings about a 

progressive decline in rates of glycolytic flux, affects rates of sugar transport but not the 

activities of the glycolytic enzymes. 

Modulation of the activity of hexokinases by trehalose 6-phosphate may be of 

regulatory significance. Trehalose 6-phosphate is a strong competitive inhibitor of 

hexokinase and this may be used to control the entry of glucose into glycolysis. 

Alternatively, or possibly as well as, the synthesis of trehalose (Section 12.5.7) may be 

used as a mechanism for controlling levels of phosphate. During exponential growth of S. 

cerevisiae on glucose, the maximum glycolytic flux is within the range 200± 

300 mol.hexose/min./g dry weight of yeast. Most of the glycolytic enzymes, under 

optimal conditions, are capable of greater activity than this, indicating that they are 

probably not rate-determining. Phosphofructokinase alone has a maximum activity close 

to the measured maximum rate of glycolytic flux, suggesting a possible regulatory role. 


Further evidence for this regulatory role is the fact that the activity of this enzyme is 

modulated by several effectors. Phosphofructokinase has an obligate requirement for 

Mg 2+ .NH4 + and K + .Binding of fructose 6-phosphate is positively co-operative. ATP 

inhibits activity and AMP is stimulatory. 

The concerted effect of these metabolites on the activity of phosphofructokinase 

possibly explains why in yeast, under some conditions, flux through glycolysis becomes 

oscillatory. The explanation for this also supports the rate-determining role of 

phosphofructokinase in glycolysis. The oscillatory behaviour can be induced in cellfree 

extracts by the addition of fructose 6-phosphate but not fructose 1,6-bisphosphate, 

indicating that phosphofructokinase is the site of the effect. Fructose 6-phosphate 

activatesphosphofructokinase,which results inadecline inthe concentration of ATPand 

concomitant increase in levels of ADP and AMP. In turn, this activates phosphoglycerate 

kinase and pyruvate kinase further down the glycolytic pathway. In addition, fructose 

1,6-bisphosphate activates pyruvate kinase. This removes ADP and AMP and increases 

the concentration of ATP via substrate level phosphorylation. High levels of ATP and 

reduced fructose 6-phosphate now combine to reduce the activity of phosphofructokinase. 

In consequence, high ATP concentration inhibits the glycolytic kinases, fructose 6phosphate 

accumulates and the cycle is triggered again. 

Modulation of the activity of phosphofructokinase by substrates and products of the 

reaction undoubtedly has significance. However, in terms of overall control of glycolytic 

rates, the effect of another metabolite, fructose 2,6-bisphosphate appears to be of greater 

importance. Fructose 2,6-bisphosphate (F2,6bP) is produced from fructose 6-phosphate 

and ATP by the action of 6-phosphofructo-2-kinase (6PF2K). In yeast, asecond enzyme, 

fructose 2,6-bisphosphatase (2,6bPase) degrades F2,6bP to fructose 6-phosphate and 

phosphate. Fructose 6-phosphate is avery potent activator of 6-phosphofructo-1-kinase, 

the major glycolytic enzyme. Furthermore, F2,6bP inhibits the activity of the 

gluconeogenic enzyme, fructose 1,6-bisphosphatase. 

The activities of 6PF2K and 2,6bPase are regulated by reversible phosphorylation. In 

the phosphorylated form, 6PF2K is activated and 2,6bPase is inhibited. Activity of the 

kinase responsible for the phosphorylation is itself controlled by cAMP in aglucosemediated 

regulatory cascade similar to that shown in Fig. 12.14 on page 437. In S. 

cerevisiae, avery close correlation exists between the intracellular concentration of 

F2,6bP and the production of ethanol. This suggests that F2,6bP is of predominant 

importance in controlling glycolytic flux. 

12.5.2 Hexose monophosphate (pentose phosphate) pathway 

The hexose monophosphate pathway is an alternative route to glycolysis for sugar 

metabolism (Fig. 12.7). The pathway is often referred to as ashunt since it diverts a 

proportion of glucose from the main glycolytic path and returns metabolites at the level 

of triose phosphate and fructose 6-phosphate. The first part of the pathway is 

irreversible and catalyses the oxidation of glucose 6-phosphate into a pentose 

phosphate. The oxidation generates reducing power in the form of NADPH+H + . This 

is utilized in anabolic reactions, which have a specific requirement for this pyridine 

nucleotide. 

The second part of the pathway involves a series of mostly reversible interconversions 

of pentose phosphates, hexose phosphates and triose phosphates. This provides a route for 

the assimilation of pentoses, in those yeast strains (not S. cerevisiae) capable of utilizing 

them. In addition, it provides precursors for the biosynthesis of some vitamins, purine and 


Glucose 

CH2O P 

H H H 

HO OH H OH 

H OH 

Glucose 

6-phosphate 

ZWF1 

6-Phosphogluconolactone 

NADP + NADPH + H + 

GLYCOLYSIS Xylulose 

5-phosphate 

P OCH2 CH2OH O 

H HO 

H OH 

OH H 

CH2O P 

H H 

HO OH 

H 

H OH 

CH 2OH C O HCOH HCOH CH2O P 

CHO HCOH CH2O P 

Ribuluose 5-phosphate 

RPE1 


3-phosphate 

Fructose 

6-phosphate 


3-phosphate 

TKL1 

TKL2 

6-Phosphogluconate 

GND1 NADP 

GND2 

CO2 

+ 

NADPH + H + 

Ribose 

5-phosphate 

TLK1 

TLK2 

Sedoheptulose 

7-phosphate 

TAL1 

Erythrose 

4-phosphate 

Xylulose 

5-phosphate 

pyrimidine nucleotides as well as the aromatic amino acids, phenylalanine, tryptophan 

and tyrosine. 

The proportion of carbon flow that is diverted through the hexose monophosphate 

shunt is dependent on the cellular requirements for anabolism. Since yeast growth during 

brewery fermentations is modest, it seems likely that this proportion is commensurately 

small. For yeast growing under fully oxidative conditions, where growth yields are 

relatively higher, the proportion of carbon diverted through the hexose monophosphate 

shunt will be higher. In addition, where the growth medium is relatively simple, anabolic 

requirements are increased, as is the need for NADPH. Bruinenberg et al. (1983) 

estimated that under aerobic conditions 2±3% of carbon flow had to pass through the 

hexose monophosphate shunt to fulfil the anabolic requirements of yeast cells growing on 

glucose. 

12.5.3 Tricarboxylic acid cycle 

In oxidative metabolism, some of the pyruvate derived from glycolysis is oxidized to 

acetyl Coenzyme A (acetyl-CoA), a reaction catalysed by the pyruvate dehydrogenase 

complex. The acetyl units are then completely oxidized to two molecules of carbon 

dioxide in a series of reactions variously termed the citric acid cycle, tricarboxylic acid 

cycle (TCA) or Krebs cycle. Coenzyme A, serves as a carrier of acyl groups in many 

enzymatic reactions. It contains the growth factor pantothenic acid. The latter possesses a 

terminal thiol group, which is capable of forming thioesters with acyl groups. Pyruvate 

O 

H2O 

– 

CH2O P CHOH HCOH HOCH HCOH COO 

CH 2OH C O HCOH HCOH CH2O P 

RKI1 

CHO HCOH HCOH HCOH CH 2O P 

CH 2OH C O HCOH (HCOH)3 CH2O P 

CHO HCOH HCOH CH 2O P 

Fig. 12.7 The hexose monophosphate pathway (pentose phosphate pathway). The genes and 

enzymes responsible for each step are: ZWF1, glucose 6-phosphate dehydrogenase; GND1, GND2, 

6-phosphogluconate dehydrogenase; RPE1, D-ribulose 5-phosphate; RKI1, D-ribose 5-phosphate 

ketoisomerase; TKI1, TKI2, transketolase; TAL1, transaldolase. 


dehydrogenase is a multi-enzyme complex, which is located in the mitochondrial matrix. 

It shares structural and functional similarities with glycine dehydrogenase, - 

ketoglutarate dehydrogenase and branched-chain -ketoacid dehydrogenase. All contain 

a common component, lipoamide dehyrogenase, which contains the reducible prosthetic 

group, flavin adenine dinucleotide (FAD). 

The conversion of pyruvate to acetyl-CoA is an oxidative decarboxylation reaction. 

Following the initial release of CO2, the coenzyme, thiamine pyrophosphate (TPP, 

vitamin B1) acts as a carrier of the resultant -hydroxyethyl group. The latter is reduced 

and the bound acetyl group is transferred to another coenzyme, lipoic acid, in a reaction 

catalysed by lipoate acetyl transferase. The acetyl group is then transferred to coenzyme 

A and acetyl CoA is released. The reduced lipoic acid coenzyme is re-oxidized by 

lipoamide dehydrogenase. Finally, the reduced FAD prosthetic group of the latter enzyme 

is re-oxidized by NAD + and NADH+H + is liberated. 

Acetyl-CoA enters the tricarboxylic acid cycle (TCA) in a reaction in which it 

condenses with oxaloacetate to form citrate (Fig. 12.8). For each turn of the cycle, two 

– 

OOC CH 

HC COO – 

CH 2COO – 

CH 2COO – 

Pyruvate 

CO2 + H2O 

NAD 

CoASH 

Pyruvate carboxylase ATP 

CO2 

+ 

NADPH + H + 

ADP + Pi 

NADPH + H + 

NAD + 

Malate 

H2O 

Fumarate 

SDH1 

FADH 

SDH2 

FAD SDH3 

SDH4 

Succinate 

CoASH 

CH 2COO – 

HO CH 

GTP 

CH 2 

CH 2CO S CoA 

COO – 

COO – 

GDP 

Succinyl-CoA CO2 

NADPH + H + 

NAD + 

FUM1 

COO – 

COO – 

CO 

CH2 Oxaloacetate 

MDH1 

MDH2 

MDH3 

CoASH 

LPD1 

KGD1 

KGD2 

α-Ketoglutarate 

CH 2COO – 

CH2 C O COO – 

Acetyl-CoA 

CIT1 

CIT2 

CIT3 

CoASH 

Citrate 

cis-aconitate 

H2O 

H2O 

Isocitrate 

NAD(P) 

Oxalo-succinate 

+ 

CH 

IDH1 

IDH2 

HOC H 

IDP1 

IDP2 

NAD(P)H 

CO2 

CH 3 

CO 

CH 3 CO S CoA 

ACO1 

COO – 

CH 2COO – 

HO C 

CH2COO – 

COO – 

CH2COO CH 

– 

COO – 

C O COO – 

LPD1 

PDB1 

LAT1 

CH 2COO – 

C COO – 

OH 

CH 2COO – 

Fig. 12.8 The tricarboxylic acid (TCA) cycle. The genes and enzymes responsible for each step 

are: LPD1, PDA1, PDB1, LAT1, component parts of pyruvate dehydrogenase complex; CIT1, 

CIT2, CIT3, citrate synthase; ACO1, aconitase; IDH1, IDH2, NAD-dependent isocitrate 

dehydrogenase; IDP1, IDP2, NADP-dependent isocitrate dehydrogenase; KGD1, KGD2, LPD1, 

component enzymes of -ketoglutarate dehydrogenase (LPD1 lipoamide dehydrogenase); SDH1, 

SDH2, SDH3, SDH4, succinate dehydrogenase; FUM1, fumarase; MDH1, MDH2, MDH3, malate 

dehydrogenase. 


COO – 

COO – 

COO –

molecules of CO2 are liberated and reducing power is generated in the form of 

NADH+H + and FADH. Aproportion of the reduced coenzymes is re-oxidized by the 

electron transport chain. The passage of electrons/H + down the chain to oxygen, where it 

is oxidized to water, is aprocess that generates large quantities of ATP (Section 12.5.4). 

Several of the steps in the TCA cycle are catalysed by isozymes, multiple forms of the 

same enzyme, each encoded by adistinct gene. This apparent redundancy is explainable 

in that, likeglycolysis, the TCA cycle is anamphibolicpathway that serves both anabolic 

and catabolic functions. Typically, individual isozymes fulfil different roles in cellular 

metabolism. Thus, some function in the oxidative cycle and generate reducing power for 

the electron transport chain. Others are part of another pathway, the glyoxylate cycle, 

which has arole in gluconeogenesis (Section 12.5.6). 

SomeTCA cycleisozymeshavebiosynthetic rolesandprovide precursorsused forthe 

formation of amino acids such as aspartate and glutamate. Other TCA cycle enzymes 

contribute to anaplerotic pathways. These are metabolic sequences which literally have a 

`filling up' function. Thus, some of the carbon flux through the TCA cycle is utilized in 

biosynthetic reactions. Forthe oxidative cycle to continue it isessential to replenish these 

intermediates via alternative routes. Commonly, the different isozymes occupy distinct 

intracellular compartments or organelles. These organelles have particular metabolic 

pathways associated with them and these underpin the functions of the organelles. The 

oxidative TCA cycle is located within mitochondria and glyoxylate cycle enzymes are 

found within peroxisomes. Enzymes associated with biosynthesis and anaplerosis are 

usually located in the cytosol. 

Citrate synthase is the rate-limiting step in the TCA cycle. The enzyme, encoded by 

the CIT1 gene, is mitochondrial and part of the oxidative TCA cycle. The CIT2 gene 

product is peroxisomal and part of the glyoxylate cycle. The CIT1 citrate synthase is 

subject to glucose repression. This phenomenon is exhibited by many of the other 

mitochondrial isozymes, for example, aconitase and malate dehydrogenase (MDH1 and 

MDH2). In addition, MDH2 malate dehydrogenase is subject to catabolite inactivation, 

whereas, -ketoglutarate dehydrogenase activity is regulated by catabolite repression. 

Four isozymes of isocitrate dehydrogenase are synthesized by yeast, two 

mitochondrial and two cytosolic. One of each pair is specific for NAD + whereas, the 

other two are specific for NADP + .Only the NAD + -linked enzymes are active in the 

oxidative TCA cycle. The NADP + -linked enzymes probably have anaplerotic roles, as 

does the MDH2 malate dehydrogenase. The peroxisomal MDH3 enzyme appears to be 

involved in the oxidation of NADH, which is produced from -oxidation of fatty acids 

(Section 12.7.1). 

The supply of oxaloacetate for the oxidative TCA cycle (Fig. 12.8) is ensured 

primarily by the action of the anaplerotic enzyme, pyruvate carboxylase. This catalyses 

the condensation of pyruvate and CO2 to form oxaloacetate. The enzyme contains biotin 

as a prosthetic group. Biotin serves as a carrier of carboxyl groups. Energy is provided by 

the breakdown of a molecule of ATP. Pyruvate carboxylase is inhibited by aspartate and 

allosterically activated by acetyl-CoA and ATP. The inhibition is an end-product feedback 

control mechanism since aspartate is derived from the transamination of 

oxaloacetate. Stimulation by acetyl-CoA and ATP ensures that there is always a 

sufficient supply of oxaloacetate to maintain an adequate supply of substrates for citrate 

synthase. Transamination reactions between glutamate and pyruvate to yield alanine and 

oxaloacetate may also be of anaplerotic significance. 


HO CH 

CH2 COO – 

COO – 

H2O 

– 

OOC CH 

HC COO – 

+ 

NADH + H 

Oxaloacetate 

NAD 

MDH3 

Malate 

CoASH 

Fumarate 

2H 

MLS1 

CH 2 

COO – 

COO – 

Acetyl-CoA 

Glyoxylate 

Succinate 

Acetate 

Ethanol 

Fatty acids 

12.5.4 Electron transport and oxidative phosphorylation 

In the terminal stage of the oxidative catabolism of sugars, the reduced redox coenzymes, 

NADH+H + and FADH (reduced FAD; 4.94), arising from the TCA cycle and glycolysis, 

are re-oxidized. The process is mediated by a series of redox carriers and it culminates in 

the reduction of molecular oxygen to water. Together, the redox carriers constitute the 

respiratory or electron transport chain (Fig. 12.9). Consecutive components of the 

electron transport chain have progressively more positive standard redox potentials, 

which facilitates the ordered transfer of electrons. Transfer of electrons down the 

transport chain generates energy, a proportion of which is retained in the form of the 

high-energy bonds of ATP. The process of energy transduction is termed oxidative 

phosphorylation. It can be summarised in the following equation: 

DH2 ‡ 1 

2 O2 ‡ nADP ! D ‡ nATP ‡ H2O 

CO 

CHO 

COO – 

CH 2COO – 

CH 2COO – 

Citrate 

Isocitrate 

DH2 is a hydrogen donor. The value of n is a variable and is dependent on the tightness of 

coupling between respiration and phosphorylation and the nature of the donor. The 

efficiency of phosphorylation is commonly expressed as the P:O ratio, which is the 

number of ATP molecules generated per oxygen atom utilized. 

The redox carriers are a diverse group of compounds that share the common property 

of having a reversibly reducible component. Cytochromes are haemoproteins in which 

the prosthetic group, haem, is a tetracyclic pyrrole, containing an atom of iron, which can 

be reversibly reduced from the ferric to the ferrous form. Ubiquinone (Coenzyme Q) is a 

hydrophobic quinone, which can be reversibly reduced to the quinol form. Iron sulphur 

proteins also undergo transitions between the ferrous and ferric states. Flavoproteins 

ICL1 

COO – 

CH2COO – 

HO C 

CH2COO – 

CH 2COO – 

CH 

HOC H 

Fig. 12.9 The glyoxylate cycle. The genes and enzymes are: ICL1, isocitrate lyase; MLS1, malate 

synthase; MDH3, malate dehydrogenase (cytosolic). Other stages are catalysed by enzymes of the 

tricarboxylic acid cycle. 


COO – 

COO –

contain prosthetic groups flavin mononucleotide (FMN) or flavin adenine dinucleotide 

(FAD). Both contain areversibly reducible isoalloxazine group within the riboflavin 

moiety. 

The electron transport chain consists of five complexes, which are located within the 

inner mitochondrial membrane (Fig. 12.9). The first complex, NADH-CoQ reductase 

accepts electrons from NADH, generated by the mitochondrial TCA cycle. The electrons 

are passed on to apool of ubiquinone causing the latter to be reduced to ubiquinol. The 

ubiquinone pool also accepts electrons from the second complex, succinate dehydrogenase, 

located on the inner surface of the inner mitochondrial membrane. In yeast, this 

dehydrogenase also shows activity towards -glycerophosphate. Ubiquinols are reoxidized 

by transfer of electrons to the third complex, CoQ-cytochrome Creductase. In 

yeast, this complex contains cytochromes band c 1and an iron-sulphur protein. The 

cytochrome cpool mediates transferofelectrons betweenthe thirdand fourth complexes. 

The latter is cytochrome coxidase, which contains cytochromes a, a3 and acopper 

metalloprotein. Cytochrome oxidase completes the process by transferring electrons to 

oxygen, generating water. 

The fifth complex is an ATP synthase, the activity of which is coupled to the energy 

liberated by the controlled flow of electrons. This process is accomplished according to 

the principles of the chemiosmotic theory (Mitchell, 1979). This holds that the electron 

chain complexes are arranged spatially within the mitochondrial inner membrane such 

that as electrons flow down their electrical potential, protons are translocated from the 

inside to the outside of the membrane. Since the membrane is relatively impermeable to 

protons and other charged species, the electrogenic pumping of protons generates a 

transmembraneelectrochemicalpotentialdifference.Thishasbothelectrical(charge)and 

chemical (proton concentration) components. This thermodynamic potential drives the 

synthesis of ATP via areversible proton-translocating ATPase or ATP synthase. 

Complexes 1, 3and 4are associated with proton pumping and hence, indirectly with 

ATP generation, termed sites I, II and III, respectively. Theoretically, each pair of 

electrons traversing the whole of the respiratory chain could generate three molecules of 

ATP. Electrons arising in the mitochondrial matrix from the oxidation of succinate 

bypass the first phosphorylation site. In practice, actual yields of ATP are lower. In some 

yeast genera, for example, Candida utilis, phosphorylation site Iis present during growth 

under certain conditions of nutrient limitation. However, in S. cerevisiae, including 

brewing strains, the existence of site I has been questioned (Guerin, 1991). Unlike 

mammalian cells, mitochondria of S. cerevisiae oxidize exogenous NAD(P)H, directly 

via NAD(P) + dehyrogenases located in the outer surface of the inner membrane. These 

deliver electrons to the common ubiquinone pool and therefore, bypass site I 

phosphorylation. Systems for transporting NADH into mitochondria do exist in S. 

cerevisiae, for example, the malate ± aspartate shuttle (Fig. 12.10). This utilizes a 

combination of malate dehydrogenase and transamination reactions to transfer reducing 

equivalents from the cytosol to the mitochondria. In S. cerevisiae, it serves to control the 

concentration of NADH in the cytosol. 

The individual complexes of the respiratory chain are susceptible to inhibition by a 

variety of compounds. These have been used as tools to identify which components are 

present and their order within the respiratory chain. For example, rotenone inhibits 

complex I. The lack of effect of this compound on oxidative phosphorylation in S. 

cerevisiae is primary evidence for suggesting that complex I is absent in this yeast. 

Cyanide, azide and antimycin A, inhibit complexes III, and IV. Oligomycin inhibits 

complex V, ATP synthase. 


NADH + H + 

NAD + 

CO 

Cytosol Mitochondrial 

membrane 

Inner 

CH 2 

COO – 

COO – 

Oxaloacetate 


CH 2COO – 

CH 2 

C O COO – 

Glutamate 

Aspartate 

1 

Glutamate 

Aspartate 

2 

Malate Malate 

HC CH COO – 

CH 2 

COO – 

H 2N 

CH COO – 

CH 2 

H 2N CO COO – 

CH 2 

COO – 

Oxaloacetate 


1. Glutamate/Aspartate carrier 

2. Dicarboxylic acid carrier 

NADH + H + 

Fig. 12.10 The malate-aspartate shuttle system for transferring reducing equivalents from the cytosol to mitochondria. 


CH 2 

COO – 

NAD +

In many organisms alternative respiratory systems can be detected that are resistant to 

cyanide inhibition. There are two types of cyanide insensitive respiration, which are 

differentiated based on susceptibility to inhibition by salicylylhydroxamic acid (SHAM). 

The SHAM-sensitive pathway has been detected in several yeast strains, including 

Yarrowia lipolytica and stationary phase cultures of many strains, including Candida 

utilis. The SHAM insensitive pathway has been detected in S. cerevisiae, S. pombe, 

Kluyveromyces lactis, Hansenula saturnus and Endomycopsis capsularis. As would be 

predictedfromthelackofsensitivitytocyanidetheserespiratorypathwaysarenotcoupled 

to energy transduction. It is possible that they function as part of cellular redox control. 

12.5.5 Fermentative sugar catabolism 

During growth of S. cerevisiae on glucose and other repressing carbohydrates, the major 

product of sugar catabolism is ethanol. Aproportion of pyruvate derived from glycolysis 

is decarboxylated to acetaldehyde, a reaction catalysed by pyruvate decarboxylase. 

Acetaldehyde is then reduced to ethanol in areaction performed by NAD + -linked alcohol 

dehydrogenase (Fig. 12.11). In this mode of metabolism ATP is derived solely from 

glycolytic substrate level phosphorylation and alcohol dehydrogenase is the major redox 

control route for regeneration of NAD + . 

In S. cerevisiae growing on glucose and to alesser extent other fermentative sugars, 

fermentative metabolism is predominant irrespective of the presence of oxygen. Under 

these conditions, the phenomenon of glucose repression ensures that the genes encoding 

purely oxidative pathways such as respiratory oxidative phosphorylating electron 

transport chain are not expressed (Section 12.5.8). Furthermore, enzymes responsible 

for the utilization of oxidative carbon sources such as ethanol and glycerol are not 

synthesized and mitochondrial development is arrested. If the glucose concentration falls 

to a low level (< 0.2% w/w) the metabolism shifts and becomes `derepressed'. In this 

PDA1 

PDHβ1 

LAT1 

Acetyl-CoA 

CH 3 CO S CoA 

PDX1 

LPD1 

NAD + 

NADH + H + 

CoASH 

CoASH 

ASC2 

ADP ATP 

Acetate 

CH 3 

COO – 

CH 3 

Pyruvate 

CO2 

CO 

COO – 

ALD6 

PDC1 

PDC5 

PDC6 

NAD(P)H + H + NAD(P) + 

Acetaldehyde 

ADH1 

ADH2 

ADH3 

ADH4 

CO2 

Ethanol 

CH 3CHO 

NADH + H + 

NAD + 

Fig. 12.11 Enzymes of pyruvate catabolism. The genes and enzymes for each step are: PDA1, 

PDH 1, LAT1, PDX1, LPD1, pyruvate dehydrogenase complex; PDC1, PDC5, PDC6, pyruvate 

decarboxylase; ADH1, ADH2, ADH3, ADH, 4, alcohol dehydrogenase; ALD6, aldehyde 

dehydrogenase (cytosolic); ACS2, acetyl-CoA synthetase (cytosolic). 


case, the effects of the glucose signal are abolished and induction of the respiratory 

metabolic machinery occurs. In the presence of oxygen, the utilization of oxidative 

substrates, such as ethanol, is coupled to phosphorylation by the electron transport chain. 

The change in metabolism from fermentative and glucose-consuming to oxidative and 

ethanol-consuming is an example of diauxie, or adiauxic shift. In other words, the 

phenomenon of biphasic growth where after the exhaustion of an initial substrate there is 

alag phase during which phenotypic adaptation occurs. This results in asecond period of 

growth during which an alternative substrate is utilized. 

The branch-point between oxidative and fermentative sugar catabolism occurs at the 

level of pyruvate. Pyruvate decarboxylase is encoded by three genes, PDC1, PDC5 and 

PDC6. Of these, PDC1 is the predominant isozyme and is strongly expressed in the 

presence of glucose. PDC5 appears to be fully expressed only if PDC1 is impaired. PDC 

6 expression occurs only during growth on ethanol. Like pyruvate dehydrogenase, 

pyruvate kinase requires the cofactor, thiamine pyrophosphate for activity. 

Yeast alcohol dehydrogenase is encoded by four genes. The ADH1 product is 

associatedwith thereductionofacetaldehydetoethanol sinceitisinducedstronglyinthe 

presence of glucose. ADH2 alcohol dehydrogenase is implicated in ethanol utilization 

since it is repressed by glucose and only active in derepressed cells. The roles of ADH3 

and ADH4 are not clear. The former is amitochondrial enzyme which is induced by 

glucose, although less strongly than ADH1. The ADH4 enzyme requires Zn 2+ for 

activity.Itcanusuallybedetected inbrewingstrainsofS.cerevisiae butnot inlaboratory 

strains. It is possible that in brewery fermentations these enzymes are involved in the 

formation of higher alcohols and/or the reduction of vicinal diketones (Section 12.10). 

Possibly, these enzymes have a role in redox control during anaerobiosis. 

Under fermentative conditions, the cell has a requirement for acetyl-CoA for 

biosynthetic reactions, in particular, syntheses of amino acids and fatty acids. Albeit, the 

total requirement for acetyl-CoA is reduced since the respiratory pathways are not 

operational. Nevertheless, there must be metabolic regulation of carbon flow between 

ethanol formation via pyruvate decarboxylase and pathways leading to the formation of 

acetyl-CoA. In yeast, this may be achieved by regulation of the activities of the pyruvate 

dehydrogenase complex and pyruvate decarboxylase. In addition, a bypass mechanism 

may be implicated, whereby acetyl-CoA is derived from acetaldehyde via the concerted 

action of acetaldehyde dehydrogenase and acetyl-CoA synthetase (Figure 12.11). 

At low glucose concentrations there is a concomitant low concentration of pyruvate. 

Consequently, the pyruvate dehydrogenase route predominates since this enzyme has a 

higher affinity for pyruvate than pyruvate decarboxylase (Postma et al., 1989). At higher 

glucose concentrations, pyruvate concentrations also increase and an increasing 

proportion of the carbon flow is directed towards the formation of acetaldehyde, via 

pyruvate decarboxylase. This is then converted to acetate and acetyl-CoA through the 

bypass route. At still higher glucose concentrations, the acetyl-CoA synthetase reaction 

becomes rate-limiting and ethanol formation occurs, via alcohol dehydrogenase. This 

shift is further favoured by the increase in pyruvate decarboxylase activity due to glucose 

activation. 

Mammalian pyruvate dehydrogenases are subject to regulation via reversible 

phosphorylation. A similar mechanism in yeast has not been demonstrated, conclusively. 

Some control on pyruvate dehydrogenase activity may be exerted at the level of 

transcription. Under fermentative conditions, pyruvate dehydrogenase may be subject to 

limited transcription. The latter is controlled in concert with the enzymes of amino acid 

biosynthesis. 


Pyruvate decarboxylase is acytosolic enzyme, whereas, pyruvate dehydrogenase is 

mitochondrial, as are the enzymes of amino acid biosynthesis which utilize pyruvate and 

acetyl-CoA. For the acetaldehyde bypass to be effective acetyl-CoA must cross the 

mitochondrial membrane. Mitochondrial membranes are accessible to acetate but not 

acetyl-CoA. Acetyl-CoA synthetases and deacylases occur in both the cytosolic and 

mitochondrial compartments. In some yeast strains, acetyl units may be transported into 

mitochondria in the form of acetylcarnitine. 

12.5.6 Gluconeogenesis and the glyoxylate cycle 

Under aerobic conditions, yeast can utilize anumber of non-sugar sources, such as 

ethanol, glycerol and lactate as sole sources of carbon. For growth to proceed under these 

conditions the cell must synthesize glycolytic intermediates, some of which are 

precursors ofessential anabolicmetabolites.This requires glycolysis tooperateinreverse 

and this process is termed gluconeogenesis (Gancedo and Gancedo, 1997). The majority 

of the enzymes of glycolysis are freely reversible. However, phosphofructokinase and 

pyruvate kinase are not. These non-reversible steps are bypassed by two specific 

gluconeogenic enzymes. Fructose 1,6-bisphosphatase catalyses the conversion of 

fructose 1,6-bisphosphate to fructose 6-phosphate and phospho-enol-pyruvate carboxykinase 

catalyses the conversion of oxaloacetate to phospho-enol-pyruvate (Fig. 12.5). 

Glycolysis and gluconeogenesis do not occur simultaneously and therefore the 

direction of carbon flow is regulated. As befits such acrucial crossroad in metabolism, 

control mechanisms are many and stringent. The presence of exogenous glucose at very 

low concentration (0.2mM) totally abolishes gluconeogenesis. Under these conditions, 

the transcription ofFBP1 (fructose 1,6-bisphosphatase) andPCK1 (phosphoenolpyruvate 

carboxykinase) is repressed. The glucose signal does not involve the Ras, cyclic AMP 

system. Phosphorylation of glucose is required since the repression of the gluconeogenic 

enzymes is abolished if hexokinase 1, 2and glucokinase are absent. 

When glucose is added to yeast growing gluconeogenically on an oxidative substrate 

fructose 1,6-bisphosphatase and phosphoenolpyruvate carboxykinase are degraded by 

proteases, aphenomenon termed catabolite inactivation. Both enzymes are reversibly 

inactivated by phosphorylation. Fructose 1,6bisphosphatase is strongly inhibited by 

AMP and fructose 2,6bisphosphate. The latter metabolite is apositive effector of 6phosphofructo-1-kinase. 

The regulation of the enzymes which produce and degrade 

fructose 2,6-bisphosphate are part of a glucose- mediated, cyclic AMP-dependent 

reversible phosphorylation cascade, as described in Section 12.5.8. 

The glyoxylate cycle is apart of the gluconeogenic pathway, which is essential for 

growthontwo-carbonsubstratessuchasethanolandacetate.Essentially,thepathwayisa 

short cut through the TCA cycle that bypasses those irreversible steps that result in aloss 

of carbon as CO2. It involves two specific glyoxylate cycle enzymes, isocitrate lyase and 

malate synthase (Fig. 12.12). Both enzymes are subject to catabolite inactivation when 

glucose is added to yeast growing gluconeogenically. Isocitrate lyase and the MDH3 

malate dehydrogenase are both induced by ethanol and repressed by glucose. The 

glyoxylate cycle is not operative in brewing fermentations. 

12.5.7 Storage carbohydrates 

Under some conditions, the carbohydrates trehalose and glycogen accumulate in yeast and 

under other conditions, they are degraded (Lillie and Pringle, 1980). Since they do not 


ADP + Pi 

H + 

Mitochondrial 

ATPase 

(Complex 5) 

H + 

ATP 

H + 

Site I 

NADH-CoQ 

reductase 

(Complex 1) 

Mitochondrial 

NADH 

NAD(P)H + H + 

NAD(P) + 

NAD(P)H 

dehydrogenases 

Ubiquinone 

pool 

Succinate 

dehydrogenase 

(Complex 2) 

Succinate Malate 

TCA cycle 

H + 

Site II Site III 

H + 

CoQ-cytochrome C 

reductase 

(Complex 3) 

Outer mitochondrial 

membrane 

Cytochrome C 

pool 

Inner mitochondrial 

membrane 

Mitochondrial 

matrix 

O2 

H2O 

Cytochrome C 

oxidase 

(Complex 4) 

Fig. 12.12 Representation of electron transfer and oxidative phosphorylation in mitochondria (adapted from Alexander and Jefferies, 1990). Note the presence of 

the three proton-pumping sites and the reverse flow of protons that permits ATP formation. 


Glucose 

1-phosphate 

UTP 

Glucose 

Glucose 6-phosphate Trehalose 6-phosphate 

PGM1 

PGM2 

UGP1 

UDP-glucose 

GSY1 

PPi GSY2 

Glycogen 

(n–1) 

TPS1 

have structural roles, it has been assumed that both function as reserve materials. This 

appears to be an over-simplistic interpretation. Glycogen is a polymer of -D-glucose with 

a molecular weight of approximately 10 8 . It consists of chains of 10 14 residues of -Dglucose 

joined by 1 ! 4 linkages. These chains are cross-linked by (1 ! 6)- -Dglucosidic 

linkages and therefore there are structural similarities to amylopectin. In 

brewing strains of S. cerevisiae, during fermentation, up to 4% of wort sugars are 

converted to glycogen and it can account for up to 20 30% of the dry weight of the cell 

(Quain and Tubb, 1982). There are two pools of glycogen in S. cerevisiae. The first is 

soluble and its concentration is modulated in response to changes in physiological state. 

The second is only solubilized by treatment with acid. It has a structural role, being 

covalently linked to cell wall -glucans (Arvindekar and Patil, 2002). Trehalose ( -Dglucopyranosyl-1, 

1- -D-gluconopyranoside) is a disaccharide consisting of two molecules 

of D-glucose. In commercial preparations of bakers' yeast, trehalose accounts for 

15 20% of the cell dry weight. In brewing yeast after fermentation, trehalose 

concentrations are usually quite modest, typically 2 3% of the dry weight. In yeast 

recovered from very high-gravity worts (25 ëPlato, SG c. 1106.1), trehalose levels of 

20 25% of the cell dry weight have been reported (Majara et al., 1996a, b). 

Glycogen and trehalose are both synthesized from glucose 6-phosphate (Fig. 12.13). 

Both biosynthetic pathways utilize uridine triphosphate (UTP) to generate the glucosedonor 

molecule, uridine diphosphate-glucose (UDP-glucose). This reaction is catalysed 

by UDP-glucose pyrophosphorylase. Trehalose is synthesized by the action of two 

enzymes, trehalose 6-phosphate synthase (TPS1) and trehalose 6-phosphatase (TPS2). 

Degradation is via the hydrolase, trehalase, two forms of which occur in yeast. The 

neutral form (NTH1) is cytosolic and acidic trehalase (ATH1) is located in vacuoles. A 

second and possibly minor pathway for trehalose has been reported, which utilizes 

H2O 

Pi 

UDP UTP 

GLC3 

ATP ADP 

Pi 

Pi 

TPS2 

NTH1 

ATH1 

Trehalose 

CH2OH HO 

H 

H 

OH 

HO 

O 

H H H 

OH 

O HOH2C O OH 

H OH 

GPH1 

Glycogen Glucose 1-phosphate 

Fig. 12.13 Synthesis and degradation of glycogen and trehalose from and to glucose. The genes 

and enzymes are: PGM1, PGM2, phosphoglucomutase; UGP1, UDP-glucose pyrophosphorylase; 

GSY1, GSY2, glycogen synthase; GLC3, branching enzyme; GPH1, glycogen phosphorylase; 

NTH1, neutral trehalase, ATH1, acidic trehalase (adapted from Francois et al., 1997). 


adenosine diphosphoglucose-dependent trehalose synthase. This is expressed only when 

maltose or galactose is the carbon source. 

Trehalose synthase and phosphatase occur together in a complex. Their gene 

promoters contain sequences that are found in other genes subject to expression by heat 

shock (Section 12.9). The vacuolar trehalase has an acidic pH optimum and is 

constitutive. The cytosolic neutral trehalase is activated by reversible phosphorylation in 

response to cyclic AMP-dependent protein kinase. 

Glycogen biosynthesis begins with areaction catalysed by an initiator protein and 

results in the formation of an -1 !4glycosyl primer molecule from UDPG. The 

initiator protein is produced by two genes, CLG1 and CLG2, which are the equivalent of 

mammalian glycogenin, responsible for de novo glycogen synthesis. The primer is then 

elongated with the formation of -1,4linkages by glycogen synthase. The -1 !6 

cross-linkages are synthesized by glycogen branching enzyme, a transglycoylase. 

Glycogen degradation to glucose 1-phosphate and glucose is accomplished by glycogen 

debranching enzyme and glycogen phosphorylase. 

Glycogen synthase activity is regulated by reversible phosphorylation, possibly under 

thecontrolofcyclicAMP-dependentkinase,althoughotherkinasesmayalsobeinvolved. 

Inthephosphorylatedform,theenzymeislessactive.Dephosphorylationiscatalysedbya 

glycogen synthase phosphatase (GLC7). In the active dephosphorylated form, glycogen 

synthase is subject to allosteric activation by glucose 6-phosphate. Glycogen synthase 

phosphorylase, the glycogen degradative enzyme occurs in aphosphorylated dimeric 

active form and anon-active dephosphorylated tetrameric form. Cyclic AMP-dependent 

kinase may also be involved in the phosphorylation of glycogen synthase phosphorylase, 

however, it appears that this operates via asecond specific glycogen phosphorylasedependent 

kinase. The latter may also be active towards glycogen synthase. 

Accumulationofglycogenandtrehaloseoccurswhengrowthisrestricted.Inaglucoselimitingmedium,thisoccurswhenapproximatelyhalfthe 

glucose hasbeen consumed. In 

glucose-rich media, it occurs in the late exponential phase when another nutrient is 

limitingandduringtheonsetofdiauxie(Croweetal.,1984).Accumulationofglycogenis 

precededbyinductionoftheenzymesresponsibleforitsbiosynthesisandbyactivationof 

glycogen synthase by dephosphorylation. These events are accompanied by other global 

metabolicchanges,whichitissuggestedindicatethattheaccumulationofglycogenispart 

of ageneral nutrient-sensing system (Francois et al., 1997). Thus, the cell responds to 

imminent starvation of an essential nutrient by accumulating glycogen and by reducing 

glycolytic flux, reducing overall protein synthesis and inducing heat shock proteins. 

The Ras-cyclic AMP cascade system is involved in nutrient sensing. Cyclic AMP 

concentration is reduced during growth on glucose. It reaches its lowest level at the onset 

of glycogen accumulation indicating that the cyclic AMP-dependent protein kinase 

(cPKA) could be implicated. However, glycogen accumulation also requires the activity 

of another protein kinase, the Snf1 gene product. This kinase is aglobal regulator whose 

activity is essential for the entire phenomenon of derepression. With regard to glycogen 

accumulation, the activities of cPKA and snf1 are antagonistic. Both of these kinases are 

involved in the regulation of the transcription and post-translational control of genes, 

which include those encoding the enzymes of glycogen synthesis and degradation. 

Inbrewing,glycogenisusedbyyeasttoprovidemaintenanceenergyduringtheperiod 

when yeast is stored in the interval between cropping and re-pitching. Glycogen 

dissimilation accounts for the small decrease in yeast cell dry weight, which can be seen 

inthelatterphaseoffermentationwhengrowthhasceased(Fig.12.1).Inaddition,during 

the aerobic phase of fermentation, glycogen reserves are rapidly mobilized. A linear 


elationship has been demonstrated between the quantities of glycogen utilized and sterol 

synthesized (Quain and Tubb, 1982). 

Trehalose concentrations are low in yeast growing exponentially on glucose. 

Accumulation occurs during the transition between exponential growth and entry into the 

stationaryphase.Thesyntheticanddegradativepathwaysarebothactivesimultaneouslyand 

the actual concentration at any instant represents the balance between the two. The neutral 

trehalaseisactivatedbyphosphorylationbycPKA.Activityofthesyntheticcomplexisalso 

regulated by cPKA. This is not by direct phosphorylation of the enzymes and it seems that 

the cAMP signal pathway regulates transcription of their respective genes. 

Trehalose accumulation in yeast occurs in response to heat shock and some other 

stresses, such as exposure to hydrogen peroxide. It is assumed that this response renders 

the cells more resistant to stress. Trehalose confers resistance to elevated temperature in 

many species (Lillie and Pringle, 1980). This is because trehalose is able to stabilize 

membranes, where it prevents phase transition events in lipid bilayers. For maximum 

effectiveness, it must be present at both the inner and outer surfaces of the membrane 

where it binds to polar heads of phospholipids via the sugar hydroxyl groups. S. cerevisiae 

possessesatrehalosetransporter.Ithasbeensuggestedthataswellasbeingresponsiblefor 

theuptakeofexogenoustrehalose,thecarriermayalsomediateintracellulartransportfrom 

the cytosol to the plasma membrane and periplasm (Eleutherio et al., 1993). 

Following heat shock to S. cerevisiae, trehalose 6-phosphate synthase activity 

increases and neutral trehalase is inactivated. The elevated biosynthetic activity is due to 

increasedexpressionoftheTPS1andTPS2genes.Inaddition,thecatalyticactivityofthe 

biosynthetic and degradative enzymes alters in favour of trehalose accumulation. 

12.5.8 Regulation of sugar metabolism 

Sugar metabolism is highly regulated. The energy yield from individual catabolic 

pathways is very different. During fermentative growth, ATP generation is restricted to 

substrate level phosphorylation. For each glucose molecule oxidized via glycolysis, there 

is a net gain of two molecules of ATP. In the case of yeast utilizing oxidative 

phosphorylation, this value increases to approximately 30 molecules of ATP generated 

per molecule of glucose oxidized. Yeasts may be classified on the basis of their preferred 

mode of sugar catabolism. Obligate aerobes make exclusive use of the respiratory 

pathways and are unable to ferment sugars. They include the genera, Rhodotorula, 

Lipomyces, Cryptococcus, Rhodosporidium and Saccharomycopsis. Facultative anaerobes 

may use both respiratory and fermentative pathways. This group is further 

subdivided based on the proportion of sugars catabolized by each route under aerobic 

conditions. Respiratory types are predominant and dispose of 70% or more of sugars via 

respiration. These include Candida, Hansenula, Kluyveromyces and Pichia. Fermentative 

yeasts are typified by high rates of sugar metabolism of which 10% or less is catabolized 

by respiration. Saccharomyces (including all brewing strains), Brettanomyces and 

Schizosaccharomyces belong to this category. 

Anumber of phenotypic effects are recognized which describe the patterns of sugar 

catabolism in various genera growing under certain defined conditions. These are 

summarized in Table 12.3. The Crabtree effect was originally described in rat ascites 

tumour cells where it was observed that the addition of glucose resulted in a reduction in 

respiration rate (Crabtree, 1929). Subsequently, this was ascribed to competition for ATP 

and inorganic phosphate between glycolysis and respiration. In the presence of glucose, 

cellular requirements for ATP are satisfied by glycolysis. This results in a reduced 


Table12.3 Mechanismsfortheregulationofsugarcatabolisminyeast(BoultonandQuain,2001) 

Mechanism Description 

Short-term Crabtree effect Reduced respiration rate in response to glucose pulse 

Glucose catabolite repression and Suppression of respiration by glucose 

inactivation 

Pasteur effect Reduction in rate of glycolysis under aerobic conditions 

Kluyver effect Obligate aerobic utilization of disaccharides 

Custers effect Aerobic stimulation of rate of glucose fermentation 

requirement for the translocation into the cytosol of ATP produced in mitochondria by 

oxidative phosphorylation. In turn, this restricts the exchange of cytosolic ADP with 

mitochondrial ATP. The resultant reduction in mitochondrial ADP concentration restricts 

oxidative phosphorylation and respiratory rates decline. 

This short-term Crabtree effect occurs within minutes of the addition of glucose. A 

similar phenomenon occurs in yeast and it is accompanied by an immediate increase in 

the rate of ethanol production. The mechanism may be as described in the preceding 

paragraph, however it seems more likely that the effect resides in the relative affinity for 

pyruvate of pyruvate dehydrogenase and pyruvate decarboxylase (Section 12.5.5). 

The so-called long-term Crabtree effect describes the repression and inactivation of 

respiratory enzymes in yeast by the presence of glucose. The underlying biochemistry of 

this effect is totally different from the short-term Crabtree effect. More properly, it is 

described as glucose catabolite repression and inactivation. Obligate aerobic yeasts and 

respiratory facultative anaerobic types exhibit little or no glucose repression. These are 

termed Crabtree-negative or weakly Crabtree-positive, as appropriate. 

S. cerevisiae is aCrabtree-positive yeast. In aerobically growing cultures to which 

glucose (>0.2%w/w) has been added, glycolysis becomes the major energy-yielding 

pathway and ethanol is produced. The expression of several sets of genes are repressed 

(Table 12.4). The respiratory pathways are inoperative even in the presence of oxygen. 

Gluconeogenesis is inhibited, as are pathways associated with the utilization of C 2and C 3 

compounds. Glucose repression is accompanied by changes in cell morphology. Thus, the 

biogenesisofmitochondriaandperoxisomesisinhibited.Inaparallelseriesofevents,some 

ofthepre-formedenzymes,whichareencodedbythegenesofglucose-repressiblepathways, 

are inactivated by proteolysis. This phenomenon is termed glucose catabolite inactivation. 

The molecular basis of glucose catabolite repression is complex and not yet entirely 

elucidated. The evidence suggests that two parallel and antagonistic signalling pathways 

are involved (Stark, 1999). These pathways transmit the signal from the initial glucose 

trigger to the target genes and bring about the phenotypic changes associated with 

repression. In addition, in the absence of aglucose trigger, asignal is transmitted which 

brings about derepression. Thus, derepression is apositive function and is not brought 

about simply by the absence of arepressing signal. 

The initial receptor for the repressing glucose signal is unknown, however, one of the 

genes encoding for ahexokinase (HXK2) is involved. Thus, the repressing signal is not 

transmitted unless glucose is phosphorylated after uptake. No glycolytic genes after 

hexokinase are required for glucose repression and it does not occur in the absence of 

HXK2. There is circumstantial evidence that the Ras cyclic AMP-dependent protein 

kinase(cPKA)signalcascadeisalsoimplicated.Whenglucoseisaddedtoayeastculture 

growing oxidatively under conditions of glucose limitation, there is an immediate 

increaseinintracellularlevelsofcyclicAMP.ThesolefunctionofcyclicAMPinyeastis 

to activate cPKA. The activities of several enzymes, which take part in pathways 


Table 12.4 Genes, electron carriers and enzymes subject to glucose catabolite repression 

Gene Enzyme Metabolic Function 

CYC1 Cytochrome C Respiratory pathway 

COX6 Cytochrome oxidase 

QCR8 Ubiquinol cytochrome C oxidoreductase 

CIT1 Citrate synthase TCA cycle 

ACO1 Aconitase 

KGD1 -Ketoglutarate dehydrogenase 

MDH3 Malate dehydrogenase (cytosolic) 

MAL61 Maltose permease Disaccharide uptake and utilization 

MAL62 Maltase 

SUC2 Invertase 

GAL1 Galactokinase Galactose uptake and utilization 

GAL2 Galactose permease 

GUT1 Glycerol kinase Glycerol utilization 

GUT2 Glycerol 3-phosphate kinase 

FBP1 Fructose 1, 6-bisphosphatase Gluconeogenesis 

PCK1 Phosphoenolpyruvate carboxykinase 

ICL1 Isocitrate lysase Glyoxylate cycle 

MLS1 Malate synthase 

ADH2 Alcohol dehydroegenase II Ethanol utilization 

ACS1 Acetyl-CoA synthetase 

POX1 Acyl-CoA oxidase -oxidation 

POT1 3-Oxoacyl-CoA thiolase 

responsive to glucose repression, are influenced by cPKA. The effect of cPKA may be 

direct or indirect via other intervening kinases. 

The principal components of the signalling pathway mediated by Ras genes are shown 

in Fig. 12.14. The Ras proteins are subject to post-translational modification. This takes 

the form of addition of a GTP residue, a reaction catalysed by guanine-nucleotide 

exchange factor. The latter is the product of another gene (Cdc25p) whose expression is 

controlled by an, as yet, unidentified metabolite. The GTP-Ras protein is as an activator of 

adenylate cyclase and the activity of the latter catalyses the formation of cyclic AMP 

(3 1 ,5 1 -cyclic AMP). 

Cyclic-AMP-dependent protein kinase (PKA) comprises two regulatory sub-units 

(Bcy1P) and two catalytic sub-units (TpK). Cyclic AMP binds to the inactive PKA and 

causes the regulatory sub-units to disassociate. Active PKA is liberated, which is then 

free to phosphorylate target enzymes and so transmit the glucose signal to the various 

target metabolic pathways and produce a response. Cyclic AMP is degraded by 

phosphatases. In addition, cPKA exerts feed back control on cAMP levels by 

phosphorylation of a component of the Ras transduction pathway. 

The antagonistic derepressing signal pathway is mediated by another kinase, Snf1p, 

levels of which increase rapidly in cells starved of glucose. Snf1p occurs as part of a 

multi-protein complex. In the presence of glucose the complex autophosphorylates and 

lacks kinase activity. In the absence of glucose, a protein phosphatase (PP1) is activated 

by an unknown mechanism. This dephosphorylates the Snf1p-containing complex and as 

a result of a conformational shift the auto-inhibition is relieved and kinase activity is 

restored. Snf1p kinase apparently acts on another mediating protein, termed Mig1p. In the 


Glucose 

Cdc25p gene Ras gene Ira1 Ira2 genes 

Guanine-nucleotide 

exchange factor 

GDP-Ras 

protein 

GTP-RAS 

protein 

Cyr1p gene 

Adenylate cyclase 

ATP Cyclic AMP + PPi 

GAP protein 

GTP-ase 

PDE genes 

cAMP 

phosphodiesterase 

Inactive PKA Tpk.Bcy1P Bcy1P.cAMP 

Active PKA 

Tpk 

Phosphorylation of target genes 

Fig. 12.14 The Ras adenylate cyclase signal transduction system. The abbreviations are defined in 

the text. 

unphosphorylated form this protein binds to the promoter region of glucose-repressible 

genes and prevents their expression. When Mig1p is phosphorylated by Snf1p it 

dissociates from the target genes and repression is thereby lifted. 

Glucose catabolite repression is of importance in brewery fermentations. In the initial 

aerobic phase the presence of glucose ensures that metabolism is fermentative. In the 

later stages of fermentation, although the repressing signal may be absent, depending on 

the composition of the wort, anaerobiosis ensures that oxidative respiratory metabolism 

does not develop. The presence of glucose in wort prevents the utilization of the 

predominant sugar, maltose (Section 12.4.1). In all-malt worts this is of small importance 

since glucose concentrations are low, relative to maltose. Thus, during the early aerobic 

phase of fermentation glucose repression prevents the development of respiratory 

capacity and maltose utilization. However, this effect is transitory such that the 

disappearance of glucose and the onset of anaerobiosis are roughly coincident. At this 

stage anaerobiosis prevents the yeast from acquiring respiratory capacity and the 

disappearance of glucose removes the repressing effect such that maltose can be utilized. 

Caution should be exercised with worts containing glucose syrup adjuncts since the 

resultant prolonged repressing signal may prevent maltose utilization throughout much of 

the fermentation. In single-stage immobilized yeast reactors designed for primary 

fermentation glucose repression can have adverse effects. Thus, if throughput rates are 

too high the continuous addition of glucose in fresh wort can repress maltose utilization. 

This problem can be circumvented by the use of continuous systems containing two or 

more discrete stages. In such systems glucose is utilized in the first stage. 


The glucose repression phenomenon is influenced by the availability of other nutrients. 

In the absence of acomplete growth medium, glucose-starved cells exposed to glucose 

exhibitatransientrepressionresponse.Followingadditionofthelimitingnutrient,asource 

of nitrogen for example, the typical glucose repression response is seen. This non-glucose 

response is identical to that mediated by cyclic AMP-dependent protein kinase (cPKA). 

However, the effect still proceeds in mutants lacking the regulatory sub-unit of cPKA. 

Ithas beensuggestedthatyeastsrequire additional controlsthatpreventthefull-blown 

glucose repression response in the absence of acomplete growth medium. When the 

latter condition is satisfied activation of protein kinase A(PKA) occurs and repression 

proceeds. The transmission of the initial trigger from the non-glucose nutrient to protein 

kinaseAoccursviaaregulatorysystemthatdoesnotinvolvecyclicAMP.Thissignalling 

mechanism is described as the fermentable-growth-medium induced pathway (Thevelein 

and Hohmann, 1995). 

ThePasteur effect isthe phenomenon whereby fermentation isinhibited by respiration 

or glycolytic rates decrease under aerobic conditions (Warburg, 1926). The energetic 

yield of respiration is more favourable than fermentation and furthermore, yields of 

cellular biomass per unit of sugar consumed are greater. It would be supposed, therefore, 

that under aerobic conditions yeast would preferentially use oxidative phosphorylation 

for the generation of energy and in consequence reduce rates of glycolysis. The 

magnitude of the Pasteur effect is dependent on the relative respiratory capacities of 

individual yeast strains. Both S. cerevisiae (a fermentative facultative anaerobe) and 

Candida tropicalis (a respiratory facultative anaerobe) growing exponentially on glucose 

under anaerobic conditions exhibit glycolytic rates of approximately 200 M.glucose 

consumed/min./g.dry wt. yeast. Under aerobic conditions, the glycolytic rate in S. 

cerevisiae is virtually unchanged, however, in C. tropicalis it decreases by more than 

90% to approximately 5 M.glucose consumed min 1 g 1 yeast dry weight (Gancedo and 

Serrano,1989). In theformer yeast, acomparatively smallPasteur effect isobservedonly 

in glucose or nitrogen-starved stationary phase cultures. 

The mechanism of the Pasteur effect is obscure and possibly differs depending on 

cultural conditions and the nature of the yeast. In derepressed cells undergoing a 

transition from anaerobiosis to aerobiosis the effect may be due to simple competition for 

pyruvate. Since pyruvate dehydrogenase has ahigher affinity for pyruvate compared to 

pyruvate decarboxylase the presence of oxygen allows carbon to be diverted towards 

respiratorypathwaysandfermentationratesdecline.InyeastsuchasS.cerevisiaewithan 

inherentlylimitedrespiratorycapacity,thiseffectissmall.Decreaseinglycolyticratesby 

aerobiosis appears to involve feed-back control from oxidative phosphorylation. The 

mechanism is unknown although it has been suggested that glycolytic rates might be 

modulated by the effect of phosphate on the activity of phosphofructokinase (Gancedo 

and Serrano, 1989). The proposal is that under anaerobic conditions, flux through 

oxidative phosphorylation is reduced and this results in an increase in phosphate 

concentration. In turn, this activates phosphofructokinase and glycolytic activities are 

stimulated. Other mechanisms regulating sugar metabolism in various genera of nonbrewing 

yeasts have been discovered,for examplethe CustersandKluyver effects.These 

are described elsewhere (Boulton and Quain, 2001). 

12.5.9 Ethanol toxicity and tolerance 

The biochemistry underlying the formation of ethanol from the catabolism of sugars is 

describedinSection12.5.5.Withregardtocommercialfermentations,inwhichethanolis 


a major product of yeast metabolism, the rate of ethanol formation and the maximum 

concentration formed may be important considerations. In the case of the fermentation of 

high-gravity worts the ability of yeast to withstand high concentrations of ethanol is an 

influential factor in strain selection. Brewing yeast strains are exposed to ethanol 

concentrations typically in the range of 3±6% v/v. In high-gravity brewing ethanol 

concentration may be as high as 10% v/v. Wine yeasts are considered more ethanol 

tolerant than their brewing counterparts. In wine fermentations, final ethanol 

concentrations of 10±15% v/v are usual. In extreme cases such as Tokay wines and the 

rice `wine' sakeÂ, the fermentations generate in excess of 20% v/v ethanol. 

The ability to tolerate ethanol is usually considered to have a genetic basis. Exposure 

of yeast to very high ethanol concentrations results in the inhibition of growth and 

ultimately death. There is no precise definition of ethanol tolerance in yeast. Some strains 

are more able than others to withstand the deleterious effects of ethanol. Of course, it is 

possible that intermediates of ethanol formation or other products of fermentation exert 

deleterious effects on yeast. In addition, in order to generate high ethanol concentrations 

during fermentation, it is necessary to provide a high initial concentration of fermentable 

sugar. In this case, the ability to grow under conditions of low water activity may be of 

greater or equal importance to ethanol tolerance, per se. In order to ferment concentrated 

worts it is essential that other nutrients are available in balanced quantities. This is an 

important consideration in high-gravity brewing where the injudicious use of sugar 

adjuncts may result in wort that is deficient in non-sugar nutrients. 

Ethanol inhibits yeast growth in a non-competitive manner. It does not have any 

specific inhibitory effect on glycolytic rate. Many reports describe morphological 

changes in response to ethanol exposure, for example, the development of cell surface 

invaginations and cell shrinkage (Pratt-Marshall et al., 2002). The toxic effects are 

apparently more severe when ethanol is generated endogenously compared to the same 

concentration added to the medium (Nagodawithana and Steinkraus, 1976). This 

observation prompted the suggestion that the effect was a consequence of intracellular 

accumulation of ethanol during fermentation. However, this assertion has been refuted. It 

appears that the plasma membrane is freely permeable to ethanol such that intracellular 

accumulation occurs only during very early fermentation (D'Amore et al., 1988). Other 

aliphatic alcohols also exert inhibitory effects. The severity of the inhibition is 

proportional to the chain-length of the molecule. 

Several mechanisms have been proposed by which the toxic effects of ethanol may be 

exerted. These include non-specific osmotic effects (Jones and Greenfield, 1987) and a 

number of specific target sites (D'Amore et al., 1990; Mishra, 1993). A high concentration of 

intracellular ethanol reportedly denatures some enzymes. Exposure to ethanol has a 

mutagenic effect on mitochondrial DNA as indicated by an increase in the occurrence of 

respiratory petites. The major site for ethanol toxicity appears to be the plasma membrane and 

other intracellular membranes. Several effects have been observed which relate to membrane 

function. These include leakage of cellular components, abolition of membrane proton 

motive potential, inhibition of transport systems and alterations in membrane structure and 

fluidity. The fact that the membrane is the primary target explains the observation that longer 

chain-length alcohols have enhanced toxicity. Thus, there is a concomitant increase in 

hydrophobicity and consequently easier interaction with membrane lipids. 

The observation that endogenously generated ethanol is more toxic than added ethanol 

at similar concentration suggests that other intermediates of ethanol biosynthesis might 

be influential. Potential candidates include short chain fatty acids, higher alcohols, 

acetate and in particular acetaldehyde (Jones, 1987). The argument is perhaps most 


persuasive in the case of acetaldehyde since reportedly, it is an order of magnitude more 

toxic to yeast than ethanol and is the immediate precursor of ethanol (Jones, 1989). 

However, the fact that acetaldehyde can accumulate in yeast cells, under some 

conditions, to agreater concentration than seen in fermentation and apparently with no 

toxic effect provides apowerful counter-argument (Stanley and Pament, 1993). It is 

perhaps most likely that intermediates of ethanol metabolism may contribute to ethanol 

toxicity in asynergistic fashion. 

Yeast cells exhibit various adaptations in response to exposure to ethanol. Ethanol 

elicits astress response such that there is aconcomitant acquisition of thermotolerance 

and barotolerance (Hisada et al., 2002). In addition, there is an increase in levels of 

intracellular trehalose (Mansure et al., 1994). Since trehalose is amembrane-stabilizing 

agent (Section 12.5.7), this response is explained. The lipid composition of membranes is 

altered such that there is an increase in the content of unsaturated fatty acids and 5,7unsaturated 

sterols and adecrease in saturated lipids. Of course, this response is not 

possible in abrewing fermentation where these syntheses are precluded by anaerobiosis. 

The toxic effects of ethanol are reportedly ameliorated by the presence of Mn-superoxide 

dismutase, the mitochondrial enzyme implicated in resistance to oxidative stress Costa et 

al., 1993). This enzyme is induced under oxidative conditions. Therefore, it would not be 

active under the conditions of abrewing fermentation (Section 12.6). 

The tolerance of yeast to ethanol can be influenced by manipulation of the growth 

medium. Predictably, supplementation of the medium with a source of unsaturated fatty 

acids is beneficial in this regard. A considerable body of evidence has now been amassed 

indicating that the addition to growth media of various metal ions provides protection 

against ethanol stress. In particular, calcium and magnesium have been reported to have 

beneficial effects (Dombek and Ingram, 1986; Ciessarova et al., 1996). It has been 

suggested that wort may not contain optimal concentrations of these metal ions (Walker 

et al., 1996; Rees and Stewart, 1997). This can be remedied by supplementation with 

magnesium to ensure that the ratio of this metal ion to calcium is always high. The 

mechanism by which metal ions exert protective effects is not clear. 

12.6 The role of oxygen 

Oxygen presents yeast with both an opportunity and a threat. Facultative anaerobic strains 

such as brewing yeasts have the ability to grow either oxidatively or fermentatively. The 

effects of glucose catabolite repression in Crabtree positive yeast are relieved in the 

absence of a glucose signal. However, development of complete respiratory competence 

requires the presence of oxygen. Thus, aerobiosis provides yeast with the opportunity of 

utilizing the energetically favourable oxidative route for energy production. However, 

oxidative metabolism is accompanied by the generation of potentially harmful reactive 

oxygen radicals. Consequently, yeast must have enzyme systems for removal of oxygen 

radicals and nullifying this potential threat. 

Three classes of genes in S. cerevisiae are recognized based on their response to oxygen 

tension (Zitomer and Lowry, 1992). Some genes are expressed only under anaerobic 

conditions. The role of many of these is unknown, however, some are involved in the 

assimilation of nutrients from the medium which otherwise require oxygen for their synthesis. 

Hypoxic genes are expressed strongly under micro-aerophilic conditions and are 

apparently required for the efficient utilization of low oxygen concentrations. The expression 

of more than 200 genes is required for aerobic respiratory growth (Tzagoloff and Dieckmann, 


1990).Manyoftheseareexpressedonlyunderaerobicconditions.Theseincludecomponents 

of the electron transport chain such as ubiquinone and cytochrome oxidase. 

The signal pathway by which molecular oxygen exerts its effects upon metabolism is 

unknown. However, as described in the previous section in the hierarchy of signalling 

pathways, it plays asubordinate role to glucose repression. Haem is akey intermediate in 

the oxygen-sensing pathway. Formation of the haem precursor, protoporphyrin-IX is via 

the oxidation of coproporphyinogen-III. This is the rate-determining step in haem 

biosynthesis. Consequently, haem content and oxygen tension are directly related (De 

Winde and Grivell, 1993). In addition to its role as aprosthetic group in molecules such 

as cytochromes, haem is an effector metabolite in many pathways that utilise molecular 

oxygen. It is involved in the positive regulation of expression of genes encoding 

respiratory enzymes and those that play apart in protecting the cell against oxygen 

radicals. Conversely, haem represses the expression of several genes that are redundant 

under anaerobic conditions. These include some of those responsible for the synthesis of 

sterols and unsaturated fatty acids. 

Brewing yeasts do not develop respiratory competence under the conditions 

encountered in fermentation. Thus, in the aerobic phase of fermentation, respiratory 

pathways are repressed because of the presence of sugars. In late fermentation when the 

sugars have disappeared and their repressing effects are relieved, anaerobiosis prevents 

the induction of the respiratory enzymes. 

The majority of yeasts require oxygen for growth. In astudy of type species from 75 

genera, it was noted that only 23% could grow under anaerobic conditions on acomplex 

medium supplemented with ergosterol and asource of unsaturated fatty acids (Visser et 

al.,1990).Ofthese,S.cerevisiaewasexceptionalinthatitwascapableofrapidgrowthat 

low oxygen tension. Nevertheless, none of these yeasts, including S. cerevisiae, can grow 

under totally anaerobic conditions unless the medium is supplemented with asource of 

unsaturated fatty acids and sterols (Andreason and Stier, 1953a,b). These essential 

metabolites can be assimilated from the medium or synthesized de novo from 

carbohydrates. Synthesis requires the presence of molecular oxygen. Both of these are 

present in wort at the start of fermentation. 

In brewery fermentations, sterols and unsaturated fatty acids are synthesized during 

the aerobic phase. Cell proliferation during the anaerobic phase of fermentation dilutes 

the pre-formed pools of sterols and unsaturated fatty acids amongst daughter cells. On 

subsequent re-pitching, these lipids must be replenished hence the requirement for 

oxygenation of wort. Failure to provide sufficient oxygen is one of the prime causes of 

slow and sticking fermentations. The quantity of oxygen required for fermentation is 

strain-dependent. In an early study, ale strains were classified as requiring half air 

saturation, air saturation, oxygen saturation or more than oxygen saturation for satisfactoryfermentationperformance 

(Kirsop, 1974). Similar findingshave been reportedfor 

lager yeast strains (Jacobsen and Thorne, 1980). The explanation for these differences is 

related to the spectrum of sterols produced by individual yeast strains (Section 12.7.3). 

The fate of most of the oxygen utilized during the aerobic phase of fermentation is 

unknown. Theoretically 10% is utilized for sterol formation and 15% for the biosynthesis 

of unsaturated fatty acids (Kirsop, 1982). More than 50% is unaccounted for. 

During fermentative growth, yeast forms ATP via cytosolic substrate level 

phosphorylation since the mitochondrial oxidative electron transport is inoperative. 

Many essential energy-requiring enzyme systems are located within promitochondria, the 

undifferentiated organelles characteristic of fermentative yeast. Adenine nucleotides are 

transported between the cytosol and mitochondria via an ADP/ATP translocase. Three 


isozymesofADP/ATP occurinS.cerevisiae,onethatisconstitutive,asecondinducedin 

respiratory cells and athird induced by anaerobiosis (Kolarov et al., 1990). It is proposed 

that the latter enzyme catalyses transfer of ATP from the cytosol into mitochondria 

during fermentative growth. 

Oxygen radicals are highly reactive species, which are implicated in damaging effects 

such as lipid peroxidation, mutagenesis and other degenerative changes associated with 

ageing and senescence. Yeast, in common with other cells, possesses protective 

mechanisms for removing oxygen radicals (Krems et al., 1995). The precursor of sterols, 

squalene, reportedly scavenges free radicals in mammalian cells (Kohno et al., 1995). 

Squalene accumulates in yeast under anaerobic conditions (Table 12.6 on page 447) and 

it could fulfil a similar role. Similarly, reduced glutathione reacts with superoxide, 

hydrogen peroxide and larger hydroperoxides. The two major protective enzymes are 

superoxide dismutase (SOD) and catalase (Fridovich, 1986). These enzymes, acting in 

concert, convert the superoxide radical to oxygen and water. 

2 : O2 ‡ 2H‡ ! O2 ‡ H2O2 

2H2O2 ! 2H2O ‡ O2 

Yeast cells possess two superoxide dismutases, a cytosolic CuZn SOD and a 

mitochondrial Mn SOD. Mutants lacking both isozymes are hypersensitive to oxygen. 

The cytosolic CuZn SOD is constitutive and highly expressed in aerobic cultures and 

those grown on non-fermentable substrates. The mitochondrial Mn SOD is induced by 

oxygen and it is not present in cells growing fermentatively. In brewing yeast, under nongrowth 

conditions during a transition from anaerobiosis to aerobiosis, there was a rapid 

increase in the specific activity of CuZn SOD. The specific activity of the Mn SOD 

increased only after several hours exposure to oxygen (Clarkson et al., 1991). The 

transition was accompanied by a decrease of 5±7% in the viability of the culture. It was 

concluded that the CuZn SOD was protective in anaerobic yeast, whereas the Mn SOD 

was important only in aerobic cultures. 

Two forms of catalase occur in yeast, a cytosolic form (catalase T, CTA1 gene) and a 

peroxisomal form (catalase A, CTT1 gene). Both are haemoproteins. CTA1 is induced by 

oxygen and growth on non-fermentable substrates such as fatty acids. It is repressed by 

glucose. CTT1 is also induced by oxygen and it requires haem for its expression. In 

addition, it is induced in response to stresses such as heat shock, low water activity and 

oxidative stress (Dawes, 1999). These observations have resulted in the suggestion that 

catalase T is involved in hydrogen peroxide removal during the stationary phase, 

whereas, catalase A is protective towards sudden oxidative stress. 

12.7 Lipid metabolism 

Lipids are a diverse group of compounds that are characterized by the common property 

of sparing solubility in water but being soluble in organic solvents. They have important 

structural roles, especially in membranes. Frequently they form part of larger 

macromolecules where the hydrophobic nature of the lipid moiety confers specific 

properties. Many lipids have biological roles in signalling systems, as vitamins and in 

receptor sites on cell surfaces. 

S. cerevisiae, contains relatively modest levels of lipids, typically 5±15% of the cell 

dry weight (Rattray, 1988). The predominant classes of lipid are sterols (both free and 


Table 12.5 Lipid composition (% total lipid) of various strains of S. cerevisiae (adapted from 

Rattray, 1988) 

Strain TAG DAG MAG FFA S SE PL Other 

1 7.4 ± ± 4.1 4.3 21.9 61.7 ± 

2 23.9 12.9 ± 1.9 ± 23.9 26.9 10.4 

3 8.2 1.3 0.9 1.9 6.6 53.3 27.7 ± 

4 7.4 0.3 0.8 0.8 4.7 45.4 38.3 ± 

5 9.6 5.8 ± 2.3 ± 25.3 56.9 ± 

6 40.4 4.3 ± 5.8 ± 19.6 29.9 ± 

7 12.6 4.9 ± ± 3.0 8.4 28.1 ± 

8 8.6 2.3 0.6 1.4 7.2 46.5 33.4 ± 

9 9.5 1.4 1.7 0.6 5.6 49.0 32.2 ± 

10 14.4 ± ± 1.2 1.7 25.4 52.0 5.4 

11 11.8 2.8 3.5 8.8 17.3 26.5 29.2 ± 

12 16.8 2.8 4.4 9.2 11.0 28.3 27.5 ± 

13 15.1 4.9 5.3 15.8 21.1 21.1 16.8 ± 

TAG, triacylglycerol; DAG, diacylglycerol; MAG, monoacylglycerol; FFA, free fatty acids; S, sterol; SE. steryl 

esters; PL, phosholipids. 

esterified), phospholipids and triacylglycerols. Fatty acids, either free or esterified as 

diacylglycerols andmonoacylglycerolsmakeupmostoftheremaininglipid(Table12.5). 

The predominant saturated fatty acids are palmitic (16:0) and stearic (18:0) with smaller 

amounts of myristic (14:0) and lauric (12:0). Unsaturated fatty acids are mainly 

palmitoleic (16:1), oleic (18:1) and linoleic (18:2). 

12.7.1 Fatty acid metabolism 

Fatty acids are synthesized from acetyl-CoA. The latter may arise via several routes. 

Principally, it is derived from glucose catabolism from pyruvate, directly or via 

acetaldehyde, acetate and acetyl-CoA synthetase. It is formed from the catabolism of 

amino acids, leucine, lysine, tryptophan, tyrosine and phenylalanine. In addition, acetyl- 

CoA is the end-product of the -oxidation pathway for the degradation of fatty acids. 

The biosynthetic pathway to fatty acids involves the action of two enzyme systems, 

acetyl-CoA carboxylase and the fatty acid synthase complex. Acetyl-CoA carboxylase 

catalysestheconversionofacetyl-CoAintomalonyl-CoA.Thereactioniscomplexandis 

drivenbythebreakdownofATP.Theadditionalcarbonatomisderivedfrombicarbonate 

ion in areaction involving the coenzyme biotin. 

Enzyme-biotin‡ATP‡HCO3 enzyme-biotin-CO2 ‡ADP‡Pi 

Enzyme-biotin-CO2 ‡acetyl-CoA malonyl-CoA‡enzyme 

Acetyl-CoA carboxylase is encoded by the gene ACC1. Asecond gene ACC2 encodes a 

biotin-apoprotein ligase which catalyses the addition of the biotin moiety to the ACC1 

gene product and converts it from the inactive apo- to the active holo- form. Acetyl-CoA 

carboxylase is considered the rate-determining step in fatty acid biosynthesis. It is 

activated allosterically by citrate and isocitrate. 

Fatty acid synthase (FAS) is amultienzyme complex that catalyses asequence of 

reactions in each cycle of which afatty acid is lengthened by two carbon atoms. (Fig. 

12.15). The two carbon atoms are donated by malonyl-CoA deriving from the activity of 

acetyl-CoA carboxylase. The FAS complex contains acyl carrier protein (ACP) which 

contains 1 0 -phosphopantetheine as a prosthetic group. During the sequence of reactions, 


HOOC CH2 CO S CoA 

HCO3 – 

CH3 CO S CoA 

Acetyl-CoA 

ATP 

NADP + 

NADPH + H + 

CO2 ACP 

Acetoacetyl-S-ACP 

H2O 

NADPH + H + 

Crotonyl-S-ACP 

NADP + 

CoA 

Malonyl-CoA 

Acetyl-S-ACP Malonyl-S-ACP 

D-β-Hydroxybutyryl-S-ACP 

Butyryl-S-ACP 

FAS 7 

cycles 

CoASH 

intermediates and substrates are bound to the SH-group of ACP via thioester linkages. 

Activation of FAS requires activity of phosphopantetheinyl transferase, which attaches 

the 1 0 -phosphopantetheine arm to a serine residue in ACP. 

Each cycle of reactions involves a priming step in which acyl groups are transferred 

from CoA to ACP. This is followed by a condensation reaction where the acyl and 

malonyl groups are combined with the loss of the third carbon atom as CO 2. Finally, there 

are two reductive steps involving the coenzyme, NADPH+H + and an intermediate 

dehydration step. Seven successive cycles of the FAS complex leads to the formation of 

the C16:0 fatty acid ester, palmitoyl-CoA. 

Acetyl-CoA ‡ 7malonyl-CoA ‡ 14 NADPH ‡ 14H ‡ 

CH 3 

CH 3 

CO CH2 CO S ACP 

CH3 CHOH CH2 CO S ACP 

CH CH CO S CoA 

CH3 CH2 CH2 CO S CoA 

Acetyl CoA carboxylase 

! palmitoyl-CoA ‡ 7 CoASH ‡ 7CO2 ‡ 7H2O ‡ 14 NADP ‡ 

Unsaturated fatty acids (UFA) are produced by the action of the enzyme, -9- fatty 

acid desaturase, which inserts a double bond into fatty acid under aerobic conditions. 

Multiple desaturation reactions result in the synthesis of di-and trienoic acids such as 

linoleic (18:2) and linolenic (18:3) acids. The desaturase enzyme is encoded by the gene 

1 

– 

2O2 

Palmitoyl-S-ACP Palmitoyl-CoA Palmitoleic-CoA 

H2O 

Fatty acid desaturase 

Fatty acid 

synthase 

(1 cycle) 

Fig. 12.15 Pathway for the biosynthesis of saturated and unsaturated fatty acids (ACP ˆ acyl 

carrier protein). 


OLE1. The enzyme is located in the endoplasmic reticulum and its activity is linked to an 

NADPH-dependent cytochrome b5 reductase. Activity of the desaturase is regulated by 

the presence of exogenous UFA. Unsaturated fatty acids apparently cause rapid 

degradation of the OLE1 mRNA (Gonzales and Martin, 1996). 

Fatty acids, both saturated and unsaturated, eventually become located in membranes 

where they have structural roles. The relative chain lengths and degrees of unsaturation 

are influenced by environmental conditions. For example, as the growth temperature is 

reduced there is a need to maintain membrane fluidity. This is achieved by increasing the 

degree of unsaturation of the fatty acids in membrane lipid components. In addition to 

desaturases, S cerevisiae also possesses a membrane-bound fatty acid elongation system, 

encoded by the ELO genes. These are independent of the FAS system and are used to 

elongate short chain fatty acids (C10±C12) assimilated from the medium to C16±C18. 

The elongation reactions occur prior to insertion of fatty acids into membranes 

(Schweizer, 1999). 

It would be predicted that traditional lager fermentations performed at relatively low 

temperatures would have an increased requirement for UFA as a consequence of the 

correlation of the latter and membrane fluidity. This possibility appears not to have been 

explored in detail although it has been reported that UFAs are essential for satisfactory 

fermentation performance. This effect was ascribed to the requirement for UFA for 

proper mitochondrial development and function (O'Connor-Cox et al., 1993). UFAs are 

reported to influence the formation of esters (12.10.4). Linoleic acid reduces the 

formation of acetyl esters, reportedly by the repression of ATF1, the gene encoding 

alcohol acetyltransferase (Fuji et al., 1997). 

Fatty acid biosynthesis occurs in the cytosol, although there is evidence that yeast may 

possess a second mitochondrial fatty acid synthase (Schneider et al., 1997). Fatty acid 

degradation in S. cerevisiae occurs in peroxisomes via -oxidation. The process consists 

of a cyclic two-carbon shortening of fatty acids catalysed by a series of enzymes encoded 

by FOX genes. The first and rate limiting step in the pathway is catalysed by an acyl CoA 

oxidase and converts fatty acyl-CoA esters to trans 2, 3-dehydroacyl-CoA esters and 

hydrogen peroxide. The reaction utilizes flavin adenine nucleotide (FAD) as cofactor and 

this transfers electrons directly to molecular oxygen. Hydrogen peroxide is degraded by 

the peroxisomal catalase A. In the next sequence of reactions, catalysed by a 

multifunctional enzyme, trans 2, 3-dehydroacyl-CoA esters are successively modified 

by the action of trans 2-enoyl CoA hydratase and 3-hydroxyacyl CoA dehydrogenase. 

Finally, 3-ketoacyl CoA thiolase releases a molecule of acetyl-CoA leaving a residual 

acyl-CoA chain two carbon atoms shorter than the original. 

The three enzymes of -oxidation are induced during growth on fatty acids under 

aerobic conditions. In addition, the first enzyme, acyl-CoA oxidase is repressed by glucose. 

Fatty acids are transported from the cytosol into peroxisomes via transporters specific for 

medium chain length and long chain length fatty acids. The acetyl-CoA generated by - 

oxidation is transported back into the cytosol via a carnitine/acetylcarnitine shuttle system 

or as citrate. In the latter case, the acetyl-CoA is acted upon by a peroxisomal citrate 

synthase. In view of the susceptibility to glucose repression and requirement for oxygen, - 

oxidation is unlikely to play any part in brewery fermentations. 

12.7.2 Phospholipids 

Fatty acyl-CoA esters are trans-esterified to form acylglycerols via the phosphatidic acid 

pathway. The free 1, 2-hydroxyl groups of glycerol 3-phosphate are acylated by two 


molecules of fatty acyl-CoA to yield a phosphatidic acid. The phosphate is removed from 

the latter by a phosphatase. The resultant diacylglycerol reacts with a third molecule of 

fatty acyl-CoA ester to form a triacylglycerol. Two phosphatidic acid phosphatases occur, 

one microsomal and another that is located in mitochondria. The microsomal enzyme is 

subject to regulation by the Ras cyclic AMP-dependent protein kinase. In non-oleaginous 

yeasts, such as brewing strains, triacylglycerols are synthesized only when the fatty acid 

requirement for phospholipids is satisfied. Triacylglycerols accumulate with steryl esters 

in cytosolic lipid granules. The fatty acids may be mobilized by the action of lipases, 

providing an alternative source of fatty acids for phospholipids synthesis. 

The major phospholipids in yeast are glycerophospholipids. In S. cerevisiae the 

predominant types are phosphatidylcholine, phosphatidylethanolamine, phosphatidylinositol 

and phosphatidylserine (Rattray, 1988). Phospholipids consist of a molecule of 

glycerol in which one of the hydroxyl groups is esterified to phosphoric acid and the other 

two to fatty acids. The phosphate head group form ester linkages with molecules such as 

amino acids and alcohols to give the phospholipids detailed above. The combination of 

polar head group and hydrophobic fatty acyl chains confers amphipathic properties on 

phospholipids, which makes them suitable for insertion into membranes. Inositol 

phospholipids have roles in cellular signalling systems and they are involved in the 

regulation of protein sorting across intracellular membranes. 

Like acylglycerol lipids, phospholipids are synthesized from phosphatidic acids (Fig. 

12.16). Phosphatidylcholine and phosphtidylethanolamine are formed from diacylglycer- 

Acyl CoA 

CoASH 

Choline 

Pi 

CTP 

Pi 

CDP-choline Lysophosphatidic acid 

Phosphatidyl Acyl CoA 

choline 

CoASH 

CMP Pi H2O Phosphatidic 

acid 

CTP Pi 

Diacylglycerol 

Triacylglycerol 

CMP 

CDP-ethanolamine 

Phosphoethanolamine 

ADP 

ATP 

CTP 

Ethanolamine 

Glycerol 3-phosphate 

Phosphatidyl 

ethanolamine 

Acyl CoA 

CoASH 

Pi 

CDP-diacylglycerol 

Phosphatidylinositol 

Phosphatidyl 

glycerol 

L-serine 

CMP 

Phosphatidylserine 

Myo-inositol 

CMP 

Fig. 12.16 Synthesis of phospholipids and acyl-glycerol lipids via the phosphatidic acid pathway. 

CTP ˆ cytidine 5 0 -triphosphate, CDP ˆ cytidine 5 0 -diphosphate, 

CMP ˆ cytidine 5 0 -monophosphate. 


ols. Phosphatidylglycerol (cardiolipin), phosphatidylserine and phosphatidylinositol are 

synthesized from the common precursor, CDP-diacylglycerol. Phospholipids are 

synthesized in reactions that involve the carrier molecule, cytidine 5 0 -diphosphate 

(CDP). Phospholipids are degraded by specific phospholipases that remove the fatty acyl 

groups from the glycerol molecule. 

12.7.3 Sterols 

Sterols, like unsaturated fatty acids, are formed during the initial phase of wort 

fermentationwhenoxygenisavailable.Themostabundantsterolinyeastisergosterol.In 

brewing strains, smaller quantities of zymosterol, episterol, lanosterol, and fecosterol can 

also be detected (Table 12.6). Sterols are essential components of cell membranes where, 

in conjunction with phospholipids, they confer fluidity. Sterols are most abundant in the 

plasma membrane, where they occur in the free form. Some 90% of the sterol in the 

plasma membrane is ergosterol. The smaller quantities of other sterols are probably pools 

of intermediates in the ergosterol biosynthetic pathway. Membranes surrounding 

intracellular organelles contain smaller quantities of sterols. Lipid granules, which 

contain cellular reserves of triacylglycerol, also contain sterols esterified to fatty acids. 

The starting point for sterol synthesis is acetyl-CoA. The latter may arise from the 

catabolism of wort sugars via glycolysis. However, in the early phase of brewery 

fermentations membrane function of pitching yeast may be impaired by sterol depletion. 

Consequently, it has been suggested that the carbon and energy for sterol synthesis could 

be supplied by mobilization of glycogen reserves. Alinear correlation between glycogen 

breakdown and sterol synthesis during the aerobic phase of wort fermentation has been 

demonstrated (Quain and Tubb, 1982). Yeast can utilize exogenous sterols. This occurs 

only under anaerobic conditions when de novo synthesis is precluded by the absence of 

oxygen. This phenomenon has been termed aerobic exclusion (Parks and Casey, 1995). 

By inference, any sterols present in wort would not be utilized during the initial aerobic 

phase of fermentation. 

In the first part of the sterol biosynthetic pathway, three acetyl units are combined to 

form a molecule of mevalonate (Fig. 12.17). Mevalonate is then converted to 3isopentenyl 

pyrophosphate in a sequence of phosphorylation reactions in which three 

molecules of ATP are hydrolysed. Thus, sterol synthesis is expensive in terms of the 

expenditure of metabolic energy. A molecule of 3-isopentenyl pyrophosphate and its 

isomer, 3-3-dimethylallyl pyrophosphate then condense with a loss of pyrophosphate to 

form the monoterpene derivative, geranyl pyrophosphate. This reacts with another 

molecule of 3-isopentenyl pyrophosphate to form the sesquiterpene derivative, farnesyl 

Table 12.6 

data) 

Squalene and sterol composition of lager pitching yeast (S. C. P. Durnin, unpublished 

Component % Dry wt. 

Squalene 1.2 

Ergosterol 0.095 

Lanosterol 0.055 

4, 4-Dimethylzymosterol 0.022 

Zymosterol 0.01 

Ergosta-7, 22-dienol 0.009 

Dihydroergosterol 0.007 


Acetyl-CoA 

PPi 

Geranyl pyrophosphate 

PPi 

Ubiquinone Farnesyl pyrophosphate 

PPi 

Acetyl-CoA 

Acetoacetyl-CoA 

CoA 

Hydroxymethylglutaryl-CoA 

NADPH 

CoA 

Mevalonic acid 

3ATP 

3ADP 

3-Isopentenyl pyrophosphate 

Squalene 

pyrophosphate. This metabolite is also a precursor for haem and ubiquinone biosynthesis. 

Two molecules of farnesyl pyrophosphate condense to give presqualene pyrophosphate, 

which is then reduced by NADPH with a loss of pyrophosphate to form squalene. 

The initial part of the sterol biosynthetic pathway is an anaerobic process. The 

biosynthesis of ergosterol from squalene requires molecular oxygen. The precise steps of 

the biosynthetic pathway differ in individual strains but a generalized scheme is shown in 

Fig. 12.17. Oxygen is utilized in the first step in which squalene is converted to the 

epoxide, 2, 3-oxidosqualene. Some of the subsequent steps utilize cytochrome P450 

oxygenases, which also require molecular oxygen. 

Sterol biosynthesis is highly regulated. The key step in the early part of the pathway is 

the formation of mevalonate from hydroxymethylglutaryl-CoA (HMG-CoA), catalysed 

by HMG-CoA reductase. This step is thought to be the rate-limiting step in the 

biosynthesis of sterols and other isoprenoids (Schweizer, 1999). Two isozymes of HMG- 

CoA reductase occur in yeast, encoded by the genes HMG1 and HMG2. Expression of 

both genes is regulated by oxygen. HMG1 is expressed to the greatest extent under highly 

aerobic conditions, whereas, HMG2 is strongly expressed under conditions of hypoxia. 

The transcription of both enzymes is activated by haem. The modest sterol concentrations 

synthesized by brewing yeast during fermentation probably result from the activity of 

O2 

Lanosterol Squalene epoxide 

Ignosterol 

4,4-Dimethylzymosterol 

CoA 

NADPH 

3,3-Dimethylallyl 

pyrophosphate 

Zymosterol Fecosterol Episterol Ergosterol 

Fig. 12.17 Biosynthetic pathway of ergosterol from acetyl-CoA. 


HO

HMG2 alone. The relatively high sterol concentrations formed in brewing strains of S. 

cerevisiae growing under derepressed conditions probably require the activity of both 

HMG1 and HMG2. 

Regulation of carbon flow up to squalene requires apartitioning of flux between the 

formation of sterols and the biosynthesis of other isoprenoids. This is accomplished by 

controloftheactivitiesoftheenzymesinthe earlypartofthe sterolbiosyntheticpathway 

and the expression of the genes encoding them. Squalene synthase is the first committed 

step in sterol synthesis. It appears that ergosterol and other late intermediates in the sterol 

pathway regulate the biosynthesis of ergosterol by feedback control mechanisms 

controlling the activities of enzymes earlier in the pathway. 

Sterols are reportedly synthesized in the microsomal fraction of the endoplasmic 

reticulum (Schweizer, 1999). Other have asserted that a separate partitioned sterol 

biosynthetic pathway may occur in mitochondria (Casey et al., 1992). To complicate 

matters further, sterol esters are located in cytosolic lipid granules. These membrane 

bound structures also contain triacylglycerols. In addition, sterol C-24 methyltransferase, 

a late enzyme of the sterol biosynthetic pathway is also found here as well as 

intermediates of sterol biosynthesis (Schweizer, 1999). The implication is that there may 

be more than one site of synthesis and in any case, an intracellular sterol transport system 

must exist. It is likely that under conditions of aerobic exponential growth, sterols are 

synthesized in the endoplasmic reticulum, transported to appropriate sites and 

incorporated into proliferating membranes. In stationary phase cells, where membrane 

proliferation has ceased, sterols are esterified by acyl-CoA: sterol acyltransferases and 

stored in lipid granules. If the conditions change such that growth recommences, steryl 

esters are transported to the sites of membrane growth and there they are cleaved by 

steryl ester hydrolases. The resultant free sterol may then be used for membrane 

synthesis. 

Inbrewingfermentationstheextentofsterolsynthesisinyeastismodestcomparedto 

derepressed cells. Pitching yeast typically contains 0.1±0.2% of the cell dry weight as 

sterol. This increases to approximately 1% of the cell dry weight at the end of the 

aerobic phase of fermentation. The same yeast grown aerobically under derepressing 

conditions contains approximately 5% of the cell dry weight as sterol (Quain and Tubb, 

1982). 

12.8 Nitrogen metabolism 

In wort, the major sources of nitrogenous nutrients are amino acids and ammonium ions. 

It would be supposed that amino acids might be incorporated directly into proteins and 

other macromolecules. However, in brewing strains of S. cerevisae, growing 

fermentatively on wort, amino acids are catabolized (Jones and Pierce, 1970). By 

inference, amino acids that are required for the biosynthesis of other macromolecules 

must themselves be synthesized. Amino acids are usually degraded by long and 

convoluted pathways. These pathways provide essential precursors for other nitrogencontaining 

cellular constituents such as purines and pyrimidines. This anabolic 

requirement presumably explains the preference for degradation. The oxidative 

catabolism of amino acids ultimately removes the amino group. Providing the yeast is 

growing under derepressed and aerobic conditions the resultant carbon skeletons can be 

fed into oxidative energy-producing pathways or provide precursors for gluconeogenesis 

(Fig. 12.18). In this way nitrogen and sugar metabolism are coupled and coordinated. 


CH3 CO CH2 CO S CoA 

Acetoacetyl-CoA 

Phenylalanine 

(4.43) 

Tyrosine (4.48) 

Lysine (4.40) 

Alanine (4.24) 

Threonine (4.46) 

Glycine (4.35) 

Serine (4.45) 

Cysteine (4.31) 

Pyruvate 

Acetyl-CoA 

Leucine (4.39) 

Trytophan (4.47) 

Isoleucine (4.38) 

Citrate 

Arginine (4.28) 

Histidine (4.36) 

Glutamine (4.34) 

Proline (4.44) 

Glutamate (4.33) 

2-Oxoglutarate 

Succinyl-CoA 

Fumarate 

Oxaloacetate 

Aspartate (4.29) 

Asparagine (4.30) 

Alternatively, under fermentative conditions some of the products of amino acid 

catabolism may be released into the medium. As aresult of the fermentative growth of 

yeast on wort, these may contribute to beer flavour. The carbon skeletons for amino acid 

synthesis may be derived from glycolysis. They may also be formed from the products of 

the degradation of other amino acids or ammonia may supply nitrogen for the amino 

groups. Thus, S. cerevisiae can utilize ammonia and most individual amino acids as sole 

sources of nitrogen. The biosynthetic pathways are not usually asimple reversal of those 

used for amino acid degradation. Key reactions in amino acid metabolism are 

transaminations where the -amino group of an amino acid is transferred to the - 

carbon atom of an -keto acid. The latter is usually -ketoglutaratic acid. For example, 

the reaction shown below in which the amino group of L-aspartate is transferred to - 

ketoglutarate to form L-glutamate. 

L-aspartate‡ -ketoglutarate oxaloacetate‡L-glutamate 

Isoleucine 

(4.38) 

Methionine 

(4.41) 

Valine 

(4.49) 

Phenylalanine 

(4.43) 

Tyrosine 

(4.48) 

Fig. 12.18 Fate of carbon skeletons of amino acids following oxidative degradation. In some 

cases, as indicated, some amino acids contribute to the pools of more than one intermediate of the 

main sugar oxidative pathways. The numbers in parentheses indicate where the molecular structures 

may be found. 

These reactions are reversible and many amino acids participate inthem. The deaminated 

amino acid is converted to the corresponding -keto acid analogue. These -keto acids 

are precursors of other metabolic by-products, which contribute to beer flavour, for 

example higher alcohols and esters (Section 12.10). In addition, they are themselves used 

to synthesize the corresponding amino acids, as required. The -keto acid analogues of 

some amino acids are shown in Table 12.7. 

Wort amino acids have been classified based on the relative contribution of their 

corresponding -keto acid analogues to the development of abalanced spectrum of 


Table 12.7 Amino acids and their corresponding -keto acid analogues 

Amino acid -keto acid analogue 

Alanine Pyruvic acid 

Aspartate Oxaloacetic acid 

Glutamate -ketoglutaric acid 

Isoleucine -keto- -methylvaleric acid 

Leucine -ketoisocaproic acid 

Phenylalanine Phenylpyruvic acid 

Serine Hydroxypyruvic acid 

Tyrosine Hydroxyphenylpyruvic acid 

Valine -ketoisovaleric acid 

Table 12.8 Classification of amino acids based on the essential nature of the corresponding - 

keto acid analogue (from Jones and Pierce, 1970) 

Class 1 Class 2 Class 3 

Aspartate Isoleucine Lysine 

Asparagine Valine Histidine 

Glutamate Phenylalanine Arginine 

Glutamine Glycine Leucine 

Threonine Alanine 

Serine Tyrosine 

Methionine 

Proline 

flavour compounds (Jones and Pierce, 1970). The initial concentrations in wort of 

members of Class 1 (Table 12.8) were considered relatively unimportant since they 

could be either assimilated from wort or synthesised de novo. Thus, there is no 

shortage of intermediates and the provision of precursors for the synthesis of other 

essential products of metabolism. The initial concentration of Class 2 amino acids was 

considered crucial since in the later stages of fermentation synthesis of these from 

sugars was repressed. Consequently, the -keto acid analogues derived from amino 

acid degradation became the major sources of these amino acid carbon skeletons. A 

shortage of the latter in late fermentation would be predicted to have large effects on 

the metabolism of related by-products and by inference beer quality. Class 3 amino 

acids were considered to be derived exclusively from wort. Deficiencies in these would 

be expected to restrict the synthesis of compounds derived from their -keto acid 

analogues. 

Glutamate and the ammonium ion are central to nitrogen metabolism since they link both 

catabolism and anabolism. The key enzymes are glutamine synthetase, glutamate synthase 

(GOGAT) and glutamate dehydrogenase. These catalyse the reactions shown following. 

NH4 ‡ ‡ glutamate ‡ ATP glutamine ‡ ADP ‡ Pi 

Glutamine synthetase 

glutamine ‡ -ketoglutarate ‡ NADPH ‡ H ‡ ! 2 glutamate ‡ NADP ‡ 

Glutamate synthetase 

glutamate ‡ NAD…P† ‡ H2O -ketoglutarate ‡ NH4 ‡ ‡ NAD…P†H 

Glutamate dehydrogenase 


Glutamate and glutamine are the major sources of cellular nitrogen. Ammonium ions are 

converted to glutamate by glutamate dehydrogenase. Some of the glutamate is then 

converted toglutaminebyglutaminesynthetase. Thelatter enzyme isrequiredfor growth 

on any nitrogen source other than glutamine. Glutamine is an essential precursor for the 

biosynthesis of other amino acids as well as purines and pyrimidines. Yeasts have three 

glutamate dehydrogenases, one which is NAD + -linked (Gdh2). The other two require 

NADP + ascofactor, termed Gdh1 and Gdh3, respectively. The Gdh1 gene product has a 

catabolic function in the formation of glutamate, as described already. The Gdh2 enzyme 

iscatabolicinnatureandservestoproduceammoniaand -ketoglutarateincellsutilising 

glutamate. The role of the Gdh3 enzyme is not clear, although it has been speculated that 

it is involved in anitrogen-sensing pathway (Wilkinson et al., 1996). 

Glutamate synthase performs the same function as the Gdh1, NADP + -dependent 

glutamate dehydrogenase in that it produces glutamate in conjunction with glutamine 

synthetase. Based on differences in kinetic properties it has been claimed that the 

GOGAT system ishigh affinity and low activity, whereasthe Gdh1 system isthe reverse. 

The enzyme systems described are regulated by the availability of nitrogen. Thus, Gdh1 

and Gdh2 activities are inversely related depending on the availability of ammonia. 

Glutamine synthetase activity is induced when glutamate is the sole source of nitrogen 

but low in the presence of glutamine. 

The regulation of amino acid metabolism during growth of S. cerevisiae on an 

undefined medium such as wort is complex. Intermediates and end-products of the 

pathways leading to and from individual amino acids exert control by feed-back 

mechanisms. Inaddition, there are global regulatory mechanisms.Starvation ofanyone of 

several amino acids induces many of the enzymes required for the biosynthesis of several 

amino acids. The phenomenon involves the increased transcription of more than 40 genes 

andit hasbeen termed general amino acid control (Hinnebusch, 1997). With respect to the 

utilization of nitrogen sources, the presence of ammonia or glutamine causes the 

repressionoftheenzymes requiredforthecatabolism ofotheraminoacids.Thisprocessis 

termed nitrogen catabolite repression (Wiame et al., 1985). Other metabolic controls may 

be overlaid on those solely driven by the nitrogen source. For example, in derepressed 

yeast, biosynthesis of serine occurs via glycine, derived from glyoxylate. However, in 

repressed yeast as is the case in fermentation, the glyoxylate cycle is not functional 

(12.5.6) and serine may be formed from glycolysis via 3-phosphoglycerate. 

Glutamine, with other amino acids, provides nitrogen groups for the synthesis of 

purines, pyrimidines and N-acetylglucosamine. The latter is used in the synthesis of 

chitin, astructural component of yeast cell walls. Amino acids are the building blocks of 

proteins. Areview of protein synthesis in yeast may be found in Tuite (1991). Yeast cells 

possess anumber of enzymes responsible for the hydrolytic degradation of proteins, 

termed proteinases. These enzymes fulfil intracellular roles since S. cerevisiae does not 

utilize exogenous proteins. Some proteinases have abroad specificity and are involved in 

long-term protein turnover. Others have very specific substrates and have regulatory 

functions. The latter group catalyse reactions in which target proteins are modified by 

partial proteolysis such that they are reversibly activated or inactivated. These processes 

are distinct from the phenomenon of catabolite inactivation (Section 12.5.6). 

The major site for proteolysis is the cell vacuole. Much of the regulation of both 

specific and non-specific proteolysis involves the sequestration of target proteins into 

vacuoles where they are exposed to proteinases. The majority of proteins are turned over 

very slowly. A small proportion is subject to rapid turnover. The latter group includes 

proteins required only under specific conditions, for example, cyclins. 


The tasks of elimination of damaged proteins and rapid turnover in response to 

physiological changes is performed by the ubiquitin system. Ubiquitin is aprotein that 

becomes attached via lysine residues in target proteins (Finley and Chau, 1991). Target 

proteins are attached to multiple molecules of ubiquitin in reactions catalysed by 

ubiquitin-protein ligase and ubiquitin-conjugating enzymes. The target protein-ubiquitin 

complex is recognized by amulti-enzyme proteolytic complex termed the proteosome. 

The latter catalyses the ATP-dependent degradation of the target protein to amino acids 

and peptides and releases ubiquitin so that it becomes available for further ubquinitation 

reactions. 

12.9 Yeast stress responses 

Yeast acidifies the medium when it is transferred from beer and suspended in water. This 

phenomenon is stimulated by the presence of exogenous glucose. The spontaneous and 

glucose-induced acidification of media is used as a method of assessing yeast 

physiological condition (Sigler and Hofer, 1991). The extrusion of protons is 

accompanied by the excretion of several organic species, including amino acids and 

nucleotides. The process has been termed shock excretion (Lewis and Phaff, 1964). It 

requires the presence of afermentable sugar and presumably represents astress-induced 

transient loss of membrane integrity. 

As yeast progresses through the brewing cycle of storage, pitching, fermentation, 

cropping and storage, it is subject to a number of stresses. These include rapid 

temperature fluctuation, high barometric pressure, high osmotic pressure, low water 

activity, low pH, high ethanol concentration, transient aerobiosis and starvation. Yeasts 

have evolved various metabolic strategies to minimize the deleterious effects of many 

stresses (Sorger, 1991). The ability to withstand an applied stress can be constitutive, for 

example, the differing degrees of ethanol tolerance observed in individual strains of yeast 

(Section 12.5.9). Similarly, different yeast species can grow over different temperature 

ranges, indicating a spectrum of thermotolerance. For example, ale strains of S. cerevisiae 

can usually grow at higher temperatures than lager strains. Resistance to some stresses 

can be induced by non-lethal exposure to the stress in question. The biochemistry of these 

acquired responses is distinct for individual stresses, although often there is overlap. 

Thus, exposure to one stress commonly induces tolerance to a range of other stresses. 

Strategies for overcoming many stresses involve common cellular mechanisms. 

The induced stress response requires a sensing system, a signal transduction pathway 

and expression of target genes. Exposure of yeast to elevated temperatures results in 

cellular dysfunction and ultimately death. These effects are caused via denaturation of 

proteins and membrane damage. When S. cerevisiae undergoes a temperature shift from 

25 ëC to 37 ëC growth is temporarily arrested, trehalose accumulates and the synthesis of 

most proteins ceases. A small group of around 70 so-called heat shock proteins (hsps) are 

induced and produced in high concentrations. After a short period, growth recommences 

and synthesis of hsps continues at a reduced but higher level than non-heat shocked yeast. 

The heat-shocked yeast exhibits elevated thermotolerance. 

The heat shock proteins include those responsible for trehalose biosynthesis. This is 

predictable, bearing in mind the membrane-stabilizing properties of this metabolite. The 

ubiquitin system is activated which facilitates the destruction of damaged proteins (12.8). 

Other hsps function as protein repair systems, ensuring that damaged molecules are 

refolded in their correct conformations. The mechanism by which the heat shock response 


is generated has not been fully elucidated. The hsp genes are regulated by aheat shock 

transcription factor (hsf1P) that binds to the promoter region. The route by which the 

signal from heat shock to activation of the gene encoding hsf1P is not known. However, 

the protein can be phosphorylated and the extent of phosphorylation has been shown to 

correlate with transcriptional activation over arange of temperatures (Sorger, 1991). 

Other stress responses are mediated by aspecific stress response element (STRE) 

whichhasbeenidentifiedinthepromoterregionofseveralstressresponsivegenes.These 

include genesencodingforenzymesinvolved inprotectionagainst amultitude ofstresses 

including,oxidative, lowpH,ethanol,weakorganicacidandosmotic. Activationofthese 

stress genes, via the STRE element, is in response to at least two signal transduction 

pathways. These are the Ras cAMP mediated cascade (Section 12.5.8) and the high 

osmolarity glycerol (HOG) pathway (Section 12.3.1). The common regulatory element 

within anumber of apparently unconnected genes explains how individual stresses can 

induce overlapping responses. 

The relevance of stress responses in yeast to brewing is unclear. The multitude of 

stresses to which yeast is subject during fermentation imply that astress response will be 

exhibited throughout most if not all of the stages of brewing where yeast is present. It is 

reasonable to assume that astress response makes the yeast more resistant to the stresses 

of the brewing process. This may be of particular importance in the case of high-gravity, 

high-volume fermentations. 

12.10 Minor products of metabolism contributing to beer flavour 

Both ethanol and carbon dioxide contribute to beer flavour. The latter has a`mouth 

tingle' character, whereas ethanol imparts a `warming' note to beers. In addition, 

fermentation of wort generates amultitude of other minor products of yeast metabolism, 

many of which contribute to beer flavour. The action of yeast on wort also serves to 

removesomecomponentswhosepersistenceinbeerwouldbeundesirable.Theformation 

of adesirable mixture of flavour-active metabolites in beer is influenced by the choice of 

yeast strain and wort composition. One of the primary aims of fermentation management 

is to control the process to ensure that flavour metabolites are produced in consistent and 

desired quantities. 

The principal flavour metabolites are aliphatic alcohols, aldehydes, organic and fatty 

acids and esters of alcohols and fatty acids. These are formed as by-products of the 

metabolism of sugars and amino acids. The relationships between these classes of 

metabolites are shown in Fig. 12.19. In addition amyriad of other products of yeast 

metabolism contribute to beer flavour. Many of these are excreted by yeast during 

fermentation. However, some are intracellular components that are released in the beer 

either by cell death and autolysis or via shock excretion (Section 12.9). 

12.10.1 Organic and fatty acids 

More than ahundred organic and fatty acids have been identified in yeast (Meilgard, 

1975). Although some of these are derived from wort, many are produced as aresult of 

yeast metabolism. Organic acid formation and excretion contributes to the reduction in 

pH that occurs during fermentation. They confer a`sour' or `salty' taste to beers. The 

most abundant organic acids found in beers and their typical concentrations are shown in 

Table 12.9. Organic acids are largely derived from the incomplete TCA cycle that occurs 


Table 12.9 Beer organic acids (Coote and Kirsop, 1974; Whiting, 1976; Klopper et al., 1986). 

Organic acid Typical concentration in beer (mg/l) 

Acetic 10±50 

Citric 100±150 

Lactic 50±300 

Malic 30±50 

-ketoglutaric 0±60 

Pyruvic 100±200 

Succinic 50±150 

Dicarboxylic acids 

Tricarboxylic acids 

Sugars Amino acids 

Acyl-CoA esters 

Esters 

Oxo-acids 

Aldehydes 

Alcohols 

Vicinal 

diketones 

Fig. 12.19 Relationships between the major classes of yeast-derived beer flavour compounds. 

during anaerobic repressed growth of yeast (Section 12.5.8). In addition, some may 

derive from the catabolism of amino acids (Fig. 12.18). Lactate is derived from the 

reduction of pyruvate. The extracellular concentrations of some organic acids may both 

increase and decrease during fermentation. For example, it has been reported that 

pyruvateisexcretedduringearly fermentation.Atalater stage,thisacidistakenupagain 

and acetate is excreted (Coote and Kirsop, 1974). The maximum exogenous pyruvate 

concentration coincides with the point at which wort free amino nitrogen ceases to be 

assimilated possibly suggesting that the former is derived from the dissimilation of the 

latter (Fig. 12.1). The accumulation of exogenous pyruvate indicates that at least during 

partoffermentation,therateofpyruvatedissimilationtoethanolisslowerthantherateof 

formation of pyruvate. The extracellular formation of two oxo-acids, -acetolactate and 

-acetohydroxybutyrate is of special note since they the precursors of the vicinal 

diketones, diacetyl and 2,3-pentanedione (Section 12.10.2). 

Short and medium chain length fatty acids have unpleasant flavours and they inhibit 

beer foam formation. For these reasons, their presence in beer is undesirable. Generally, 

the medium chain-length fatty acids, principally C16 and C18, of wort are replaced by 

shorter chain-length fatty acids (C6±C10) in beer (Chen, 1980). These short chain-length 

fatty acids are powerful detergents and it seems probable that they are not excreted by 

yeast in acontrolled process. Instead, it is likely that they exit cells as aresult of plasma 

membrane leakage in response to ethanol stress (Section 12.5.9) or in extreme cases 

because of cell death and autolysis. 


12.10.2 Carbonyl compounds 

Carbonyl compounds are abundant in beers, where more than 200 have been detected 

(Berry and Watson, 1987). The concentrations of several aldehydes and the vicinal 

diketones are influenced by yeast metabolism during fermentation and subsequent 

conditioning. As agroup, these generally make anegative contribution to beer flavour 

and aroma. An important requirement of fermentation management is toensure that these 

compounds are reduced to acceptable concentrations. 

Several aldehydes arise during wort production, others are formed as intermediates in 

the biosynthesis of higher alcohols from oxo-acids by yeast (Fig. 12.19). Exogenous 

aldehydes form adducts with sulphur dioxide and in this form they may not be available 

for enzymatic reduction. In this sense, the metabolism by yeast of aldehydes and sulphur 

containing compounds are intimately related. Several yeast reductases each with a 

differing spectrum of activity are involved in the elimination of aldehydes (Debourg et 

al., 1993). These enzymes use either NADH or NADPH as hydrogen donor cofactors. 

The fermentative alcohol dehydrogenase (ADHI) is responsible for the reduction of 

pentanal and pentenal, and an aldose reductase reduces 3-methyl butanal and possibly 

pentanal. In addition, an aldoketoreductase with broad specificity has been detected 

(Laurent et al., 1995). 

Acetaldehyde is of special interest because of its role as the immediate precursor of 

ethanol. It has an unpleasant `grassy' flavour and aroma. Acetaldehyde is formed during 

the early to mid stages of fermentation and thereafter it declines to alow level. In some 

circumstances, it can accumulate during fermentation in concentrations above the flavour 

thresholdof10±20ppm.Theprincipal causes ofhighacetaldehyde concentrations inbeer 

are the use of poor quality pitching yeast, excessive wort oxygenation, unduly high 

fermentation temperature and excessive pitching rates (Geiger and Piendl, 1976). 

S. cerevisiae possesses two acetaldehyde dehydrogenases. One is mitochondrial and 

requires NAD + orNADP + and K + for activity. The second enzyme is NADP + -linked, 

activated by Mg 2+ and is located in the cytosol. It has been proposed that the 

mitochondrial enzyme is functional only during oxidative growth on ethanol (Jacobsen 

and Bernofsky, 1974). However, in a study of the activities of both acetaldehyde 

dehydrogenases during ahigh-gravity lager fermentation this supposition was apparently 

disproved (Fig. 12.20). Here it may be seen that the activity of ADH1 correlated closely 

with the formation of ethanol. Surprisingly, the cytosolic Mg 2+ -dependent acetaldehyde 

dehydrogenase was active only during the early aerobic phase of fermentation. The K + - 

dependent mitochondrial acetaldehyde dehydrogenase was active throughout the whole 

of fermentation. Presumably, the latter enzyme would be instrumental in removing 

acetaldehyde during the later stages of fermentation. The reduction in acetaldehyde 

concentration characteristic of late fermentation would correlate with the concomitant 

increase in the concentration of extracellular acetate described in the previous section 

(Section 12.10.1). 

The concentrations of two vicinal diketones (VDK), diacetyl (2,3-butanedione) and 

2,3-pentanedione are of critical importance in the fermentation of lager beers. Both 

compounds have strong `butterscotch' or `toffee' aromas and tastes. Their presence in 

lagers at concentrations higher than their flavour thresholds of around 0.15ppm and 

0.9ppm respectively, causes an objectionable flavour defect. The now accepted pathway 

is that vicinal diketones arise as by-products of the synthesis of valine and isoleucine 

(Fig. 12.21). Aproportion of the pools of two acetohydroxy acids, -acetolactate and - 

acetohydroxybutyrate is excreted into the fermenting wort. There they undergo 

spontaneous oxidative decarboxylation to form diacetyl and 2, 3-pentanedione. In late 


Fig. 12.20 Specific activities of Mg 2+ -and K + aldehyde dehydrogenases (ALDH), alcohol 

dehydrogenase (ADH1) and changes in the concentrations of acetaldehyde and ethanol during a 

stirred 2.5 litre laboratory anaerobic fermentation using 12 ëP lager wort and a lager yeast strain 

(W. Tessier, unpublished data). 

fermentation, or during the conditioning phase, vicinal diketones are re-assimilated by 

yeast and reduced to form acetoin and 2,3-butanediol from diacetyl and 2,3-pentanediol 

from 2,3-pentanedione. The pattern of VDK formation and dissimilation during 

fermentation is shown in Fig. 12.1. The flavour thresholds of these reduced compounds 

are relatively high and at the concentrations that they are found in beer their presence is 

acceptable. 

The wort FAN concentration and amino acid spectrum influence the formation of 

acetohydroxy acids. Nakatani et al., (1984a,b) derived arelationship between total VDK 

concentration formed (T-VDKmax) and the minimum FAN concentration achieved during 

fermentation: 

T-VDKmax ˆ 

0:161 

FANmin 3:87 ‡0:415 

This relationship was taken to imply that wort composition and fermentation conditions 

should be manipulated to ensure that acontrolled residual FAN concentration was left at 

theendoffermentation. ThisprocedurewouldminimizethemagnitudeoftheVDKpeak. 

Similarly, worts with high concentrations of valine and isoleucine suppressed the 

formation of excessive VDK. Other than using worts with very high FAN concentrations, 

it is difficult to see how the spectrum of individual amino acids could be easily 

manipulated. 

The rate-determining step in the formation of VDK is the spontaneous oxidative 

decarboxylationoftheacetohydroxyacids.Thereactionsproceedrelativelyrapidlyunder 

aerobic conditions. Under anaerobic conditions, metal ions such as Cu 2+ ,Al 3+ and Fe 3+ 

can act as alternative electron acceptors. The process is favoured by acidic conditions 


CH 3 CO CO CH 3 

Diacetyl 

Acetoin 


CH3 CHOH CHOH CH3 CO2 + 2H 

NADH 

NAD + 

CH 3 CO COOH 

Pyruvate 

CH 3 


Thiaminepyrophosphateacetaldehyde 

TPP 

CH3 CO C 

OH 

COOH 

NADPH 

NADP + 

-NH2 

NADH 

NAD + 

2,3-Dihydroxy-isovalerate 2,3-Dihydroxy-3-methylvalerate 

H2O 

α-Keto-isovalerate α-Keto-3-methylvalerate 

-NH2 

CH3 CHOH CHNH2 COOH 

Threonine 

α-Ketobutyrate 

α-Acetohydroxybutyrate 

Valine Isoleucine 

CH3 CH2 CH3 CO C 

OH 

COOH 

CH 3 CHCH 3 CHNH 2 COOH CH 3 CH 2 CHCH 3 CHNH 2 COOH 

CO2 + 2H 

NADH 

NAD + 

CH 3 

CO CO CH2 OH 

2,3-Pentanedione 

2,3-Pentanediol 

CH3 CHOH CHOH CH2 CH3 Fig. 12.21 Pathways for the formation and dissimilation of the vicinal diketones, diacetyl and 2, 3-pentanedione. 


(Inoue et al., 1968). Heating -acetolactate at 70ëC under anaerobic conditions results in 

non-oxidative decarboxylation directly to acetoin. It has been suggested that this can also 

occur at moderate temperatures providing that sufficiently low redox conditions are 

maintained (Inoue et al., 1991). 

Reduction of VDK occurs in late fermentation or during conditioning and it requires 

the presence of viable yeast. Under most circumstances, yeast assimilates and reduces 

free diacetyl very rapidly. Consequently, the VDK which can be detected in wort during 

fermentation is mostly precursor acetohydroxy acid. Before VDK analyses can be 

performed on samples removed during fermentation, they must first be heated to ensure 

that all the acetohydroxy acid is converted to VDK. The majority of brewery VDK 

analyses represent the sum of free diacetyl and -acetolactate. This is understandable 

since -acetolactate is unstable and in the absence of yeast can be considered as 

`potential diacetyl'. An essential aspect of fermentation management is to ensure that 

yeast is not separated from green beer before the pools of free VDK and precursor 

acetohydroxy acids are reduced to an acceptable concentration. 

The physiological role fulfilled by VDK reduction during fermentation is not known, 

other than the possibility that it may represent another route for redox balancing (Section 

12.2). The enzymology of VDK reduction is not well characterized. Several yeast 

reductases, both NAD + and NADP + -requiring show activity in vitro towards diacetyl and 

2,3-pentanedione, however, it is difficult to prove that they fulfil this role in vivo. 

Nevertheless, several reports of the occurrence of specific diacetyl reductases in S. 

cerevisiae have appeared in the literature (Louis-Eugene et al., 1988; Legay et al., 1989; 

Heidlas and Tressl, 1990; Scharwz and Hang, 1994; Murphy et al., 1996). 

In the last of these, it was shown that several brewing strains could be classified into 

two groups based on the differential patterns of thermolability of diacetyl reductases. 

Lager strains all contained a distinct heat stable acetoin reductase and alcohol 

dehydrogenases, which reduced diacetyl but showed no activity towards acetoin. Topfermenting 

ale strains lacked the heat stable acetoin reductase but all contained another 

enzyme, which would reduce both acetoin and diacetyl. 

The ability of yeast to reduce exogenous diacetyl declines during fermentation 

(Boulton and Box, 1999). This was demonstrated by adding exogenous free diacetyl to a 

series of similar stirred laboratory wort fermentations and monitoring the profiles of 

subsequent decline in concentration. The rates of uptake were slower, the later in 

fermentation that the diacetyl was added. At the end of fermentation when VDK 

concentrationswereclosetotheminimumspecificationrequiredbeforetheapplicationof 

cooling, yeast was most deficient in its ability to reduce VDK. This was taken to imply a 

possible decline in rates of diacetyl uptake because of lack of yeast membrane 

competence. 

12.10.3 Higher alcohols 

The formation of glycerol has been described (Section 12.3.1). Several alcohols, other 

than ethanol are formed in beer during fermentation (Engan, 1981). Higher alcohols 

achieve maximum concentrations in fermenting wort at atime roughly coincident with 

the point at which free amino nitrogen falls to aminimum concentration (Fig. 12.1). 

These are termed fusel alcohols because of their occurrence in fusel oil. This is a byproduct 

of the production of ethanol from the fermentation of carbohydrates. Those that 

contribute to beer flavour include n-propanol, iso-butanol, 2-methylbutanol and 3methlybutanol. 

It is considered that they impart a desirable warming character to beers 


such that they intensify the flavour of ethanol. Higher alcohols are the precursors of the 

more flavour active esters. 

Some higher alcohols may derive from the reduction of aldehydes and ketones present 

in wort. Higher alcohols are synthesized from 2-oxo acids. These are decarboxylated to 

form the corresponding aldehyde and then reduced to the alcohol. The alcohol 

dehydrogenases are NAD + -dependent. This has prompted the suggestion that higher 

alcohol biosynthesis represents another mechanism for cellular redox control (Quain and 

Duffield, 1985). The 2-oxo acids arise from carbohydrate metabolism via pyruvate and 

acetyl-CoA, the so-called anabolic route, which forms part of the biosynthetic pathways 

of amino acids. Alternatively, the 2-oxo acid derives from atransamination reaction 

during amino acid utilization. This is termed the catabolic or Ehrlich route to higher 

alcohol formation (Fig. 12.22). 

The relative contribution of each biosynthetic route is determined by wort 

composition, the identity of the alcohol and the stage in fermentation. Thus, n-propanol 

is formed exclusively via the anabolic route since there is no corresponding amino acid. 

Predictably, where the wort has a high content of amino acids the catabolic route is 

favoured. Thus, under these conditions, amino acid synthesis is reduced via feed-back 

inhibition and the pool of 2-oxo acids is generated largely via amino acid catabolism. In 

the reverse situation where the supply of exogenous free amino acid is restricted, 2-oxo 

acids are formed via de novo synthesis from sugars and the anabolic route predominates. 

During fermentation of a typical all-malt wort, it has been reported that the yields of 

higher alcohols from each route are roughly similar (Schulthess and Ettlinger, 1978). 

During the early part of fermentation when free amino acids are relatively plentiful, the 

catabolic biosynthetic route predominates. This is gradually reversed as the concentration 

of assimilable amino nitrogen declines. 

Control of higher alcohol formation is achieved by the choice of an appropriate yeast 

strain and manipulation of fermentation conditions and wort composition. Several authors 

have reported that the choice of yeast strain has the biggest impact and that ale strains 

generally produce more higher alcohols than lager strains (Szlavko, 1974; Engan, 1978; 

Romano et al., 1992). Manipulation of fermentation conditions and wort composition in 

ways that favour increased yeast growth also tend to elevate higher alcohol concentration 

in beer. For example, high wort FAN and high wort dissolved oxygen. Use of high 

fermentation temperatures also favours increased levels of higher alcohols, possibly due 

to alterations in membrane fluidity (Peddie, 1990). Higher alcohol formation can be 

lowered by application of top pressure during fermentation (Rice et al.,1976). 

12.10.4 Esters 

Esters comprise the most important group of flavour-active compounds that are formed 

by yeast during fermentation. More than 100 esters have been detected in beers 

(Meilgard, 1975; Engan, 1981). Esters have fruity/solvent-like aromas and flavours. The 

most abundant is ethyl acetate, which accumulates to concentrations of 10±20 ppm. The 

concentrations of other esters are usually less than 1 ppm. The predominant route for 

formation is via the esterification of ethanol or a higher alcohol and a fatty acyl-CoA 

ester. Two enzymes are involved, an acyl-CoA synthetase and alcohol acyl transferase: 

R1COOH ‡ ATP ‡ CoASH ! R1COSCoA ‡ AMP ‡ PPI 

R1COSCoA ‡ R2OH ! R1COOR2 ‡ CoASH 


-NH2 

TPP 

Acetaldehyde Acetaldehyde-TPP Pyruvate 

Leucine 

α-Ketoiso 

caproic acid 

3-Methyl 

butan-1-al 

3-Methyl 

butan-1-ol 

(CH3) 2 CH CH2 CH2OH NADH + H + 

NAD + 

Acetate 

NADH + H + 

CO2 

β-Isopropyl 

malic acid 

NADH + H + 

NAD + 

Acetyl-CoA 

-NH2 

CoA 

α-Isopropyl 

malic acid 


CoA 

TPP 


Pyruvate 

2,3-Dihydroxyisovalerate 

α-Ketoisovalerate 

Isobutanal 

Valine Isobutanol 

(CH 3) 2 

CH 

CH 2OH 

CO2 

NAD + 

NH2 

CO2 

NADH + H + 

Fig. 12.22 Pathways for the formation of higher alcohols. 

Threonine 

NADP + 

CO2 

NADH + H + 

α-Ketobutyric acid Propanal 

CO2 

NAD + 

-NH2 

NAD + 

NADPH + H + 

α-Aceto-α-hydroxy 

butyric acid 

α,β-Dihydroxyβ-methylvaleric 

acid 

α-Keto-β-methyl 

valeric acid 

2-Methylbutan-1-al 

2-Methylbutan-1-ol 

CH3 CH2 CH(CH3) CH2OH n-Propanol 

Isoleucine 

NADH + H + 

CH 3 CH 2 CH 2OH

Acetyl-CoA and longer chain acyl-CoA esters can arise via the action of pyruvate 

dehydrogenase or acyl-CoA synthetase. During fermentation, the former route is 

unimportant (Section 12.5.5). The importance of the alcohol acyl transferase has been 

confirmed by the observation that mutants lacking the enzyme produce very low 

concentrations of esters (Lyness et al., 1997). 

The synthesis of esters requires the expenditure of metabolic energy suggesting that 

ester formation must fulfil an important metabolic role. It may be amechanism for 

regulating the ratio of acyl-CoA to free CoA (Thurston et al., 1981). Peak ester 

concentrations are reached after the formation of higher alcohols has ceased (Fig. 12.1). 

Rates of ester synthesis are maximal at the mid-point of fermentation coinciding with the 

cessation of lipid synthesis. Thus, when acetyl-CoA cannot be utilized by lipid synthesis, 

theformationofestersprovidesanalternativeuseforthissubstrate.Intermediatesoflipid 

biosynthesis influence ester formation. Supplementation of worts with the unsaturated 

fattyacid,linoleicacid(50mgl 1 )causesadramaticdecreaseinesterformation(Thurston 

et al., 1982). It was suggested that this effect was due to inhibition of alcohol acyltransferasebyunsaturatedfattyacids.Thiseffecthasbeenconfirmedbyothers(Yoshioka 

and Hashimoto, 1982a,b, 1984) and led to the proposal that ester and lipid syntheses are 

inverselycorrelated.Thisissupportedbytheobservationthatincreasingoxygensupplyto 

wort tends to decrease ester synthesis. In this case, oxygen promotes the synthesis of 

unsaturated fatty acids, which in turn reduces the activity of alcohol acyltransferase. 

It now appears that this effect is exerted at amore fundamental level (Malcorps et al., 

1991; Fuji et al., 1997). These reports provided evidence that oxygen and linoleic acid 

causedrepressionofalcoholacyltransferase.Inlaterstudies(DufourandMalcorps,1994) 

the same groups demonstrated the existence of multiple isozymes of alcohol 

acyltransferase. These have different substrate specificity and not all are subject to 

repression by oxygen and unsaturated fatty acids. 

The spectrum of esters produced during fermentation is controlled by the range and 

substrate specificity of the alcohol acyltransferases possessed by individual yeast strains. 

The concentrations of esters produced by given yeast strains can be modulated by factors 

that influence the availability of acyl-CoA esters and lipid biosynthesis. In particular, the 

provision of oxygen is crucial. 


Sulphur-containing compounds in beer produced via yeast metabolic activity arise from 

organic sulphur-containing compounds such as some amino acids and vitamins. 

Alternatively, they may be formed from inorganic wort constituents such as sulphate. 

Sulphate is transported into yeast via aspecific permease. Once in the cellit is reduced to 

sulphite in reactions that require ATP for energy. Thereafter, sulphite is reduced to 

sulphide via an NADP + -dependent reductase. The sulphide is then available for 

incorporation intoavarietyofsulphur-containing organicmetabolites(Fig.12.23).Under 

some circumstances appreciable levels of hydrogen sulphide accumulate in beer. The 

resultant sulphidic taste is an essential part of the flavour of some ales, for example 

Burton pale ales. Over-accumulation of hydrogen sulphide is undesirable. It can be 

controlled by ensuring that beers are exposed to copper, in the form of a piece of 

sacrificial pipe, which allows the formation of an insoluble sulphide. The pool size of Sadenosylmethionine 

has regulatory significance. Thus, its presence in the cell inhibits the 

transcription of all the genes, which encode the enzymes responsible for the uptake of 

sulphate, its reduction to sulphide and the synthesis of S-adenosylmethionine. 


The presence of certain amino acids in wort modulates the metabolism of sulphurcontaining 

compounds (Gyllang et al., 1989). Supplementation of growth media with 

methionine increases the intracellular concentration of S-adenosylmethionine and causes 

the effects described already. Threonine in the medium reduces the activity of 

aspartokinase by feed-back inhibition (Fig. 12.23). This in turn reduces the pool sizes 

of O-acetylhomoserine and methionine. Hence, sulphite levels increase since the 

repressing effects of S-adenosylmethionine are relieved. Isoleucine also causes an 

increase in sulphite since its presence inhibits threonine utilization. 

Sulphite forms adducts with carbonyl compounds. This prompted the suggestion that 

carbohydrate metabolism may influence sulphur metabolism (Korch et al., 1991). These 

authors demonstrated acorrelation between wort glucose concentration and sulphite 

levels in beer. At high glucose levels, there was a concomitant increase in the 

concentrations of pyruvate and acetaldehyde. These carbonyls formed addition 

compounds with sulphite, thereby depriving the methionine synthetic pathway of 

sulphite and resulting in derepression of the same. 

The formation of sulphite during wort fermentation is influenced by the availability of 

amino acids (Dufour, 1991). During early fermentation, aplentiful supply of methionine 

and threonine causes repression of the sulphite synthetic pathway. In the phase of active 

fermentation, depletion of methionine and threonine derepresses the sulphite synthetic 

genes but sulphite does not accumulate because it is fully utilized for the synthesis of 

sulphur-containing amino acids. In mid to late fermentation yeast growth ceases, the 

amino acid pool is fully depleted and sulphite reductase activity declines to alow level. 

Under these circumstances, sulphite accumulation takes place. 

Accumulation of sulphite in beers during fermentation is desirable since it may form 

adducts with potential stalingcarbonylssuch astrans-2-nonenal.Inthisrespect,there isa 

positive correlation between sulphite levels and beer flavour stability. Individual 

carbonyls have varying affinities for sulphite adduct formation. It has been claimed that 

the rate of sulphite formation regulates the proportion of carbonyls bound as adducts and 

those available for reduction by yeast (Dufour, 1991). It is essential for good flavour 

stability that sufficient sulphite is available to prevent the displacement of potential 

staling aldehydes from adducts by irreversible reactions with other beer components such 

as quinones and polyphenols. 

The sulphur-containing compound dimethyl sulphide (DMS) is an important flavour 

compound. At high concentrations, it has arelatively objectionable taste and aroma of 

cooked sweet corn. At moderate concentrations (30±100ppb) it is considered to be an 

essential componentoflagerbeers.TheprecursorofDMSisS-methylmethionine(SMM) 

which is acomponent of malt (Chapter 4). During the conversion of green malt to 

finished malt, SMM is converted to DMS and another related metabolite dimethyl 

sulphoxide (DMSO). The conditions employed during malting influence the proportions 

of DMS and DMSO formed. A kilning temperature greater than 60 ëC is required for 

appreciable DMSO formation (Parsons et al., 1977). SMM is converted to DMS by heat. 

The temperatures required for this conversion occur only during the malt and wort 

production stages of brewing. However, DMS is volatile and much is lost during mashing 

and wort boiling. DMSO is heat stable and persists unchanged through these stages. 

At collection into fermenter, wort contains a mixture of SMM, DMS and DMSO. The 

proportions of each depend upon the raw materials used for wort production and the 

conditions employed in its manufacture. Further conversion of SMM to DMS during 

fermentation is precluded by the low temperature. The residual SMM is assimilated by 

yeast and converted to methionine. Thus, the concentration of DMS in beer is determined 


O-Acetyl serine 

ATP 

PPi 

NADPH 

ADP + NADP + 

NADPH 

NADP + 

Sulphate 

Adenosine 5'phosphosulphate 

(APS) 

ATP 

ADP 

3'-Phosphoadenosine- 

5'-phosphosulphate (PAPS) 

Sulphite 

Sulphide 

ATP sulphurylase 

APS kinase 

PAPS reductase 

Sulphite reductase 

Serine Cysteine Homocysteine 

Methionine 

S-Adenosylmethioine 

O-Acetylhomoserine 

Aspartokinase 

N2H COOH 

CH 

CH2 N 

C 

C 

CH2 HC 

N 

CH2 S CH2 O 

H H 

H H 

OH OH 

N 

CH 

C 

N 

Aspartic acid 

β-Aspartyl phosphate 

Aspartic 

β-semialdehyde 

Homoserine 

Homoserine 

Phosphoric acid 

α-Keto-β-methylvaleric 

acid 

α,β-Dihydroxyβ-methylvaleric 

acid 

α-Aceto-α-hydroxybutyric 

acid 

α-Ketobutyric acid 

Threonine Pyruvate 

Fig. 12.23 Pathways for the assimilation of inorganic sulphur compounds and their incorporation into sulphur-containing amino acids. 


largely during wort manufacture. However, yeast possesses a reductase activity, which 

allows the formation of DMS from DMSO (Gibson et al., 1985). The enzyme is 

apparently subject to nitrogen catabolite repression and is most active in late fermentation 

when the wort FAN concentration is minimal. 


ALEXANDER, M. A. and JEFFRIES, T. W. (1990) Enz. Microb. Technol., 12, 2. 

ANDRE, B. (1995) Yeast, 11, 1575. 

ANDREASON, A. A. and STIER, T. J. B. (1953a) J. Cellul. Comparat. Physiol., 41, 23. 

ANDREASON, A. A. and STIER, T. J. B. (1953b) J. Cellul. Comparat. Physiol., 43, 271. 

ARVINDEKAR, A. U. and PATIL, N. B. (2002) Yeast, 19, 131. 

BARNETT, J. A. (1981) Adv. Carbohydrate Chem. Biochem. 39, 347. 

BARNETT, J. A., PAYNE, R. W. and YARROW, D. (1990) Yeasts, Characteristics and Identification, 2nd edn, 

Cambridge University Press, Cambridge. 

BERRY, D. R. and WATSON, D. C. (1987). `Production of organoleptic compounds'. In Yeast Biotechnology, 

D. R. Berry, I. Russell and G. G. Stewart eds, pp. 345±368. Allen & Unwin, London. 

BLACKWELL, K. J., SINGLETON, I. and TOBIN, J. M. (1995) Appl. Microb. Biotechnol., 43, 579. 

BOULTON, C. A. and QUAIN, D. E. (2001) Brewing Yeast and Fermentation, Blackwell Science, Oxford. 

BOULTON, C. A., BOX, W. G., QUAIN, D. E. and MOLZAHN, S. W. (1999) Proc. 27th Cong. Eur. Brew. Conv., 

Cannes, 687. 

BOURDOT, S. and KARST, F. (1995) Gene, 165, 97. 

BRUINENBERG, P. M., VAN DIJKEN, J. P. and SCHEFFERS, W. A. (1983) J. Gen. Microbiol., 129, 953. 

BUSTERIA, J. R. and LAGUNAS, R. (1986) J. Gen. Microbiol, 132, 379. 

CARTWRIGHT, C. P., ROSE, A. H., CALDERBANK, J. and KEENAN, M. H. (1989) In The Yeasts, Vol. 3 (2nd edn) 

A. H. Rose and J. S. Harrison eds, pp. 5±56, Academic Press, London. 

CASEY, W. M., KEESLER, G. A. and PARKS, L. W. (1992) J. Bacteriol., 174, 7283. 

CHEN, E. C-H. (1980) J. Am. Soc. Brew. Chem., 38, 148. 

CIESSAROVA, V. J., SMOGROVICOVA, D. and DOMENY, Z. (1996) Folia Microbiologica, 41, 485. 

CLARKSON, S. P., LARGE, P. J., BOULTON, C. A. and BAMFORTH, C. W. (1991) Yeast, 7, 91. 

COOTE, N. and KIRSOP, B. H. (1974) J. Inst. Brew., 80, 474. 

COSTA, V., REISS, E., QUINTANILHA, A. and MORADAS-FERREIRA, P. (1993) Archiv. Biochem. Biophys., 300, 

608. 

CRABTREE, H. G. C. (1929) Biochem. J., 23, 536. 

CROWE, J. H., CROWE, L. M. and CHAPMAN, D. (1984) Science, 223, 701. 

CRUMPLEN, R., D'AMORE, T., SLAUGHTER, C. and RUSSELL, I. (1993) Proc. 24th Cong. Eur. Brew. Conv., 

Oslo, 267. 

CRUMPLEN, R. M., SLAUGHTER, J. C. and STEWART, G. G. (1996) Lett. Appl. Microbiol., 23, 448. 

D'AMORE, T., PANCHAL, C. J. and STEWART, G. G. (1988) Appl. Environ. Microbiol., 54, 110. 

D'AMORE, T., PANCHAL, C. J., RUSSELL, I. and STEWART, G. G. (1990) Crit. Rev. Biotechnol., 9, 287. 

DAVELOOSE, M. (1987) MBAA Tech. Quart., 24, 109. 

DAWES, I. W. (1999) In The Metabolism and Molecular Physiology of Saccharomyces cerevisiae J. R. 

Richardson and M. Schweizer eds, pp. 277±326. Taylor & Francis, London. 

DE NOBEL, J. G., KLIS, F. M., RAM, A, VAN UNEN, H., PRIEM, J., MUNNIK, T. and VAN DEN ENDE, H. (1991) 

Yeast, 7, 589. 

DE WINDE, J. H. and GRIVELL, L. A. (1993) Prog. Nucleic Acid Res. and Molec. Biol., 46, 51. 

DEBOURG, A., LAURENT, M., DUPIRE, S. and MASSCHELEIN, C. A. (1993) Proc. 24th Cong. Eur. Brew. 

Conv., Oslo, 437. 

DOES, A. L. and BISSON, L. F. (1989) J. Bacteriol., 171, 1303. 

DOMBEK, K. M. and INGRAM, L. O. (1986) Appl. Environ. Microbiol., 52, 975. 

DUFOUR, J.-P. (1991) Proc. 23rd Conv. Eur. Brew., Conv, Lisbon, 209. 

DUFOUR, J.-P. and MALCORPS, P. (1994) Proc. 4th Aviemore Conf. Malting, Brew. and Dist., I. Campbell 

ed., pp. 137±145, IOB, London. 

ELEUTHERIO, E. C. A., DE ARUJO, P. S. and PANEK, A. D. (1993) Biochim. et Biophys. Acta, 1156, 263. 

ENGAN, S. (1978) Proc. 18th Cong. Eur. Brew. Conv., Zoeterwoede, 28. 

ENGAN, S. (1981) `Beer composition: volatile substances'. In Brewing Science, (J. R. A. Pollock, ed.), vol. 

2, pp. 98±105. Academic Press, London. 

FAERGEMAN, N. J., DIRUSSO, C. C., ELBERGER, A., KNUDSON, J. and BLACK, P. N. (1997) J. Biol. Chem., 272, 

8531. 

FELS, S., RECKELBUS, B. and GOSSELIN, Y. (1999) Proc. 7th IOB Conv., Africa Section, Nairobi, 147. 

FINLEY, D. and CHAU, V. (1991) Ann. Rev. Cell Biol., 7, 25. 


FRAENKEL, D. G. (1982) `Carbohydrate metabolism'. The Molecular Biology of the Yeast, Saccharomyces 

cerevisiae, Vol. 2, Metabolism and Biosynthesis, pp. 1±37. J. N. Strathern, E. W. Jones and J. R. 

Broach eds, Cold Spring Harbor Laboratory. 

FRANCOIS, J., BLAZQUEZ, M. A., ARINO, J. and GANCEDO, C. (1997) Yeast Sugar Metabolism, Biochemistry, 

Genetics, Biotechnology and Applications, F. K. Zimmermann and K.-D. Entian eds, pp. 285±311, 

Technomic Publishing Co., Lancaster, USA. 

FRIDOVICH, I. (1986) Archiv. Biochem. Biophys., 247, 1. 

FUJI, T., YOSHIMOTO, H., NAGASAWA, N. and TAMAI, T. (1996) J. Ferment. Bioeng., 81, 538. 

FUJI, T., KOBAYASHI, O., YOSHIMOTO, H., FURUKAWA, S. and TAMAI, Y. (1997) Appl. Environ. Microbiol., 

63, 910. 

GAMO, F. J., MORENO, E. and LAGUNAS, R. (1995) Yeast, 11, 1393. 

GANCEDO, J. M. (1992) Eur. J. Biochem., 206, 297. 

GANCEDO, J. M. and GANCEDO, C. (1997) `Gluconeogenesis and catabolite inactivation'. Yeast Sugar 

Metabolism, Biochemistry, Genetics, Biotechnology and Applications, F. K. Zimmermann & K.-D. 

Entian eds, pp. 359±378, Technomic Publishing Co., Lancaster, USA. 

GANCEDO, C. and SERRANO, R. (1989) Ènergy-yielding metabolism'. The Yeasts, vol. 3, 2nd edn, A. H. 

Rose and J. S. Harrison eds, pp. 205±259, Academic Press, London. 

GEIGER, E. and PIENDL, A. (1976) MBAA Tech. Quart., 13, 51. 

GIBSON, R. M., LARGE, P. J. and BAMFORTH, C. W. (1985) J. Inst. Brew., 91, 397. 

GONZALES, C. I. and MARTIN, C. E. (1996) J. Biol. Chem., 271, 25801. 

GRENSON, M. (1992) Àmino acid transporters: structure, function and regulation'. Molecular Aspects of 

Transport Proteins. J. De Pont, ed., pp. 219±245, Elsevier Science Publishers, London. 

GUERIN, B. (1991) `Mitochondria'. The Yeasts, Vol. 4, A. H. Rose and J. S. Harrison, eds, pp. 541±600, 

Academic Press, London. 

GUSTIN, M. C., MARTINAC, B., SAIMI, Y., CUTHBERTSON, M. R. and KUNG, C. (1986) Science, 233, 1995. 

GYLLANG, H., WINGE, M. and KORCH, C. (1989) Proc. 23rd Cong. Eur. Brew. Conv, Zurich, 347. 

HEIDLAS, J. and TRESSL, R. (1990) Eur. J. Biochem., 188, 165. 

HINNEBUSCH, A. G. (1997) CRC, Crit. Rev. Biochem., 21, 277. 

HISADA, K., SUZUKI, Y. and TAMURA, K. (2002) Trends in High Pressure Biosci. and Biotechnol., 5, 325. 

HORAK, J. (1986) Biochim. Biophys. Acta, 864, 223. 

HORAK, J. (1997) Biochim. Biophys. Acta, 1331, 41. 

INGLEDEW, W. M. and PATTERSON, C. A. (1999) J. Amer. Soc. Brew. Chem., 57, 9. 

INOUE, T., MASAYAMA, K., YAMAMOTO, Y. and OKADA, K. (1968) Reports of the Res. Lab. Kirin Brew. 

Co., 11, 9. 

INOUE, T., MURAYAMA, H., KAJINO, K., KAMIYA, T., MITSUI, T. and MAWATARI, M. (1991) Proc. 23rd Cong. 

Eur. Brew. Conv., Lisbon, 369. 

JACOBSEN, M. K. and BERNOFSKY, C. (1974) Biochim. Biophys. Acta, 350, 277. 

JACOBSEN, M. and THORNE, R. S. W. (1980) J. Inst. Brew., 86, 284. 

JAZWINSKI, S. M. (1999) Expt. Gerontol., 34, 1. 

JONES, R. P. (1987) Proc. Biochem., 22, 129. 

JONES, R. P. (1989) Enz. Microb. Technol., 9, 334. 

JONES, R. P. and GADD, G. M. (1990) Enz. Microb. Technol., 12, 402. 

JONES, R. P. and GREENFIELD, P. F. (1984) Proc. Biochem., April, 48. 

JONES, R. P. and GREENFIELD, P. F. (1987) Enz. Microb. Technol., 9, 334. 

JONES, M. and PIERCE, J. S. (1970) Proc. 25th Cong. Eur., Brew. Conv., Interlaken, 151. 

KIRSOP, B. H. (1974) J. Inst. Brew., 80, 252. 

KIRSOP, B. H. (1982) Topics in Enzyme and Ferment. Biochem., 6, 79. 

KLOPPER, W. J., ANGELINO, S. A. G. F., TUNING, B. and VERMEIRE, H.-A. (1986) J. Inst Brew., 92, 311. 

KOHNO, Y., EGAWA, Y., ITOH, S., NAGAOKA, S., TAKAHASHI, M. and MUKAI, K. (1995) Biochim. Biophys. 

Acta, 1256, 52. 

KOLAROV, J., KOLAROV, N. and NELSON, N. (1990) J. Biol. Chem., 265, 12711. 

KORCH, C., MOUNTAIN, H. A., GYLLANG, H., WINGE, M. and BREHMER, P. (1991) Proc. 23rd Conv. Eur. 

Brew. Conv., Lisbon, 201±208. 

KREGER-VAN RIJ, N. J. W. (1984) The Yeasts, a Taxonomic Study (3rd edn) Elsevier, Biomedical Press, 

Amsterdam. 

KREMS, B., CHARIZANIS, C. and ENTIAN, K.-D. (1995) Curr. Genetics, 27, 427. 

KRUCKENBERG, A. L. (1996) Arch. Microbiol., 166, 283. 

KURTZMAN, C. P. and FELL, J. W. (1998) The Yeasts, A Taxonomic Study (4th edn) C. P. Kurtzman and J. 

W. Fell, eds, Elsevier, Amsterdam. 

LAGUNAS, R. (1993) FEMS Microbiol. Rev., 104, 229. 

LAURENT, M., GELDORF, B., VAN NEDERVELDE, L., DUPIRE, S. and DEBOURG, A. (1995) Proc. 25th Cong. 

Eur. Brew. Conv., Brussels, 337. 

LEGAY, O., RATOMAHHENINA, R. and GALZY, P. (1989) Agric. Biol. Chem., 53, 531. 

LEWIS, M. J. and PHAFF, H. J. (1964) J. Bacteriol., 87, 1389. 

LILLIE, S. H. and PRINGLE, J. R. (1980) J. Bacteriol., 143, 1384. 


LOUIS-EUGENE, S., RATOMAHHENINA, R. and GALZY, P. (1988) Folia Microbiologica, 33, 38. 

LUYTEN, K., ALBERTYN, J., SKIBBE, W., PRIOR, B. A., RAMOS, J., THEVELEIN, J. M. and HOHMANN, S. (1995) 

EMBO J., 14, 1360. 

LYNESS, C. A., STEELE, G. M. and STEWART, G. G. (1997) J. Amer. Soc. Brew. Chem., 55, 141. 

MAEDA, T., WURGLER-MURPHY, S. M. and SATO, H. (1994) Nature, 369, 242. 

MAGER, W. H. and VARELA, J. C. S. (1993) Molec. Microbiol., 10, 253. 

MAJARA, M., O'CONNOR-COX, E. S. C. and AXCELL, B. C. (1996a) J. Amer. Soc. Brew. Chem., 54, 149. 

MAJARA, M., O'CONNOR-COX, E. S. C. and AXCELL, B. C. (1996b) J. Amer. Soc. Brew. Chem., 54, 221. 

MALCORPS, P., CHEVAL, J. M., JAMIL, S. and DUFOUR, J-.P. (1991) J. Amer. Soc. Brew. Chem., 49, 47. 

MANSURE, J. J. C., PANEK, A. D., CROWE, L. M. and CROWE, J. H. (1994) Biochim. Biophys. Acta, 1191, 309. 

MARDER, R., BECKER, J. M. and NAIDER, F. (1977) J. Bacteriol., 131, 906. 

MARECHAL, P. A., MARTINEZ DE MARANON, I., MOLIN, P. and GERVAIS, P. (1995) Internat. J. Food 

Microbiol., 28, 277. 

MEILGARD, M. (1975) MBAA Tech. Quart., 12, 151. 

MISHRA, P. (1993) `Tolerance of fungi to ethanol'. In Stress Tolerance of Fungi D. H. Jennings, ed, 

pp. 189±208, Marcel Dekker, New York. 

MITCHELL, P. (1979) Science, 206, 1148. 

MURPHY, C., LARGE, P. J., WADFORTH, C., DACK, S. J. AND BOULTON, C. A. (1996) Biotechnol. Appl. 

Biochem., 23, 19. 

NAGODAWITHANA, T. W. and STEINKRAUS, K. H. (1976) Appl. Environ. Microbiol., 31, 158. 

NAKATANI, K., TAKAHASHI, T., NAGAMI, K. and KUMADA, J. (1984a) MBAA Tech. Quart., 21, 73. 

NAKATANI, K., TAKAHASHI, T., NAGAMI, K. and KUMADA, J. (1984b) MBAA Tech. Quart., 21, 175. 

O'CONNOR-COX, E. S. C., LODOLO, E. J. and AXCELL, B. C. (1993) J. Amer. Soc. Brew. Chem., 51, 97. 

OLIVERA, H., GONZALEZ, A. and PENA, A. (1993) Yeast, 9, 1065. 

OURA, E., HAARASILTA, S. and LONDESBOROUGH, J. (1980) J. Gen. Microbiol., 118, 51. 

OZCAN, S. and JOHNSTON, M. (1999). Microbiol. Molec. Biol. Rev., 63, 554. 

OZCAN, S., VALLIER, L. G., FLICK, J. S., CARLSON, M. and JOHNSTON, M. (1997) Yeast, 13, 127. 

PARKS, L. W. (1978) CRC Crit. Rev. Microbiol., 6, 301. 

PARKS, L. W. and CASEY, W. M. (1995) Ann. Rev. Microbiol., 49, 95. 

PARSONS, R. P., WAINWRIGHT, T. and WHITE, F. H. (1977) Proc. 16th Conv. Eur. Brew. Conv., Amsterdam, 

115. 

PEDDIE, H. A. B. (1990) J. Inst. Brew., 96, 327. 

PHAFF, H. J., MILLER, M. W. and MRAK, E. M. (1978) The Life of Yeasts (2nd edn) Harvard University Press, 

USA. 

PIERCE, J. S. (1987) J. Inst. Brew., 93, 378. 

POSTMA, E., VERDUYN, C. SCHEFFERS, W. A. and VAN DIJKEN, P. J. (1989) Appl. Environ. Microbiol., 55, 

468. 

PRATT-MARSHALL, P. L., BREY, S. E., DE COSTA, S. D., BRYCE, J. H. and STEWART, G. G. (2002) Brewers' 

Guardian, 131, 22. 

QUAIN, D. E. and DUFFIELD, M. L. (1985) Proc. 20th Cong. Eur. Brew. Cong., Helsinki, 307. 

QUAIN, D. E. and TUBB, R. S. (1982) MBAA Tech. Quart., 19, 29. 

RAPOPORT, A. I., KHRUSTALEVA, G. M., CHAMMANIS, G. Y. and BEKER, M. E. (1995) Mikrobiologiya, 64, 

229. 

RATTRAY, J. B. M. (1988) `Yeasts' In Microbial Lipids, C. Ratledge and S. G. Wilkinson eds, vol. 1, 

pp. 555±697, Academic Press, London. 

REES, E. M. R. and STEWART, G. G. (1997) J. Inst. Brew., 103, 287. 

RICE, J. F., CHICOYE, E., HELBERT, J. R. and GARVER, J. (1976) J. Amer. Soc. Brew. Chem., 35, 35. 

ROMANO, P., SUZZI, G., COMI, G. and ZIRONI, R. (1992) J. Appl. Bacteriol., 73, 126. 

ROSE, A. H. (1987) `Responses to the chemical environment'. In The Yeasts, 2nd edn, A. H. Rose and J. S. 

Harrison, eds, Vol. 2, pp. 5±40. Academic Press, London. 

ROSE, A. H. and HARRISON, J. S. (1995) `Responses to the chemical environment.' In The Yeasts (2nd edn) 

Vol. 2, Academic Press, London. 

RYDER, D. S., MURRAY, J. P. and STEWART, M. (1978) MBAA Tech. Quart., 15, 9. 

SCHNEIDER, R., BRORS, B., BURGER, F., CAMRATH, S. and WEISS, H. (1997) Curr. Genetics, 32, 384. 

SCHULTHESS, D. and ETTLINGER, L. (1978) J. Inst. Brew., 84, 240. 

SCHWARZ, J. G. and HANG, Y. D. (1994) Lett. Appl. Microbiol., 18, 272. 

SCHWEIZER, M. (1999) `Lipids and membranes'. In The Metabolism and Molecular Physiology of 

Saccharomyces cerevisiae, J. R. Richardson and M. Schweizer eds, pp. 79±155. Taylor & Francis, 

London. 

SERRANO, R. (1989) Ann. Rev. Plant Physiol. and Plant Molec. Biol., 40, 61. 

SIGLER, K. and HOFER, M. (1991) Biochim. Biophys. Acta, 1071, 375. 

SINGH, K. S. and NORTON, R. S. (1991) Arch. Microbiol., 156, 38. 

SORGER, P. K. (1991) Cell, 65, 363. 

SOUMALAINEN, H. and OURA, E. `Yeast nutrition and solute uptake'. (1971) In The Yeasts, A. H. Rose and 

J. S. Harrison, eds, Vol. 2, pp. 3±74. Academic Press, London. 


STANLEY, G. A. and PAMENT, N. B. (1993) Biotechnol. Bioeng., 42, 24. 

STARK, M. J. (1999) `Protein phosphorylation and dephosphorylation' In The Metabolism and Molecular 

Physiology of Saccharomyces cerevisiae J. R. Richardson and M. Schweizer eds, pp. 209±275. 

Taylor & Francis, London. 

STEWART, G. G., ZHENG, X. and RUSSELL, I. (1995) Proc. 25th Cong. of the Eur. Brew. Conv., Brussels, 

403. 

SZLAVKO, C. M. (1974) J. Inst. Brew., 80, 534. 

THEVELEIN, J. M. (1994) Yeast, 10, 1753. 

THEVELEIN, J. M. and HOHMANN, S. (1995) Trends in Biochem. Sci., 20, 3. 

THURSTON, P. A., QUAIN, D. E. and TUBB, R. S. (1981) Proc. 18th Cong. Eur. Brew. Conv., Copenhagen, 

197. 

THURSTON, P. A., QUAIN, D. E. and TUBB, R. S. (1982) J. Inst. Brew., 88, 90. 

TUITE, M. F. (1991) `Protein synthesis'. In The Yeasts, Vol. 3, 2nd edn, A. H. Rose and J. S. Harrison eds, 

pp. 161±204, Academic Press, London. 

TZAGOLOFF, A. and DIECKMANN, C. I. (1990) Microbiol. Rev., 54, 211. 

VAN DER REST, M., KAMMINGA, A. H., NAKANO, A., ANRAKU, Y. and POOLMAN, W. N. (1995) Microbiol. 

Rev., 59, 304. 

VAN ZYL, P. J., KILIAN, S. G. and PRIOR, B. A. (1990) Appl. Microbiol. Biotechnol., 34, 231. 

VISSER, W., SCHEFFERS, W. A., BATENBURG VAN DER VEGTE, W. and VAN DIJKEN, J. P. (1990) Appl. 

Environ. Microbiol., 56, 3785. 

WALKER, G. M. (1998) Yeast Physiology and Biotechnology, John Wiley & Sons, West Sussex, UK. 

WALKER, G. M., BIRCH, R. M., CHANDRASENA, G. and MAYNARD, A. I. (1996) J. Amer., Soc. Brew. Chem., 

54, 13. 

WANG, S. S. and BRANDRISS, M. C. (1987) Molec. Cell. Biochem., 7, 4431. 

WARBURG, O. (1926) Biochemische Zeitschrift, 172, 432. 

WHITING, G. C. (1976) J. Inst. Brew., 82, 84. 

WIAME, J.-M., GRENSON, M. and ARST, H. N. (1985) Adv. Microb. Physiol., 26, 1. 

WILKINSON, B. M., JAMES, C. M. and WALMSLEY, R. M. (1996) Microbiology, 142, 1667. 

YOSHIOKA, K. and HASHIMOTO, N. (1982a) Reports of Res. Lab. Kirin Brewing Co., 25, 1. 

YOSHIOKA, K. and HASHIMOTO, N. (1982b) Reports of Res. Lab. Kirin Brewing Co., 25, 7. 

YOSHIOKA, K. and HASHIMOTO, N. (1984) Reports of Res. Lab. Kirin Brewing Co., 27, 23. 

YOUATT, J. (1993) Crit. Rev. Microbiol., 19, 83. 

ZITOMER, R. S. and LOWRY, C. V. (1992) Microbiol. Rev., 56, 1. 


13 

Yeast growth 


Yeast growthis defined as the coordinated assimilation of nutrients from the mediumand 

subsequent metabolism to yield new biomass. New biomass is generated by increase in 

size of individual cells and by cellular proliferation. The biochemical reactions which 

underpin anabolic metabolism and which result in the synthesis of cellular 

macromolecules are outlined in Chapter 12. The biology of the yeast cell cycle is 

described in Chapter 11. In this chapter the dynamics of yeast populations with respect to 

the influence of cultural conditions are discussed. 

All organisms must proliferate and in so doing promulgate their genotypes via their 

progeny. The formation of beer during the fermentation of wort is aby-product of 

yeast growth. The aim of the brewer is to manipulate conditions to control the growth 

and metabolism of yeast to produce a desired product. In practice, this involves 

exerting appropriate controls to influence the balance between the yields of biomass 

and metabolites. With regard to batch fermentation, maximum fermentation efficiency 

isachievedbyminimizingtheproportionofwortnutrientsusedforbiomassgeneration 

and thereby maximizing the yield of beer. With regard to fermenter cycle times, a 

secondary aim is to ensure minimum residence times. Thus, fermentation management 

requires control of both yeast proliferation and growth rate. However, both of these 

aims have to be tempered by the need to employ conditions that yield beer of the 

desired quality. 

Most brewers ensure the trueness-to-type of production yeast strains by the periodic 

introduction of a new culture derived from a laboratory stock. Sufficient yeast for 

brewing is obtained by performing aseries of fermentations of ever-increasing volume. 

Since the aim is to generate yeast mass and not beer, the conditions are manipulated to 

favour growth. Conversely, in recent years there has been arenaissance of interest in 

continuous fermentation processes, especially using immobilized yeast. These may be for 

primary fermentation or perhaps more commonly for asecondary conditioning process 

such as continuous diacetyl removal. The aim of all of these processes is to minimize cell 

proliferation and use the yeast as a biocatalyst (Boulton and Quain, 2001). 


Most modern brewers use pure cultures of asingle or defined mixture of yeast strains. 

This requires the resources of alaboratory, where stock cultures must be maintained. The 

appropriate apparatus must be available for growing pure cultures to produce sufficient 

yeast to inoculate a brewery propagation vessel. Methods are required to identify 

individual strains to provide assurance that the correct one is being used and that it is not 

contaminated with others. In addition, it is necessary to evaluate stored pitching yeast. At 

its simplest, this may take the form of ameasurement of viability. Increasingly, tests are 

being developed which assess the physiological condition of the viable fraction. 

13.2 Measurement of yeast biomass 

Yeast biomass can be measured in several ways (Table 13.1). Individual methods have 

advantages and disadvantages. No single procedure is absolutely precise and the results 

of individual methods are not always comparable. Care must be taken, therefore with the 

interpretation of results. This confusing situation is explicable in that different methods 

have been developed to fulfil specific needs. Typically, acompromise is made between 

rapidity and precision. Thus, although all methods have some shortcomings, individual 

procedures are useful providing they are used for the application that they were designed 

for. 

Direct methods of biomass determination use two general approaches. The first group 

of methods are those which measure the proportion of weight (or volume) due to cells 

within a suspension. The simplest method is to centrifuge a sample of slurry in a 

graduated tube. The volume fraction of packed yeast cells can be read directly from the 

scale on the tube. More commonly, the cell fraction is recovered by centrifugation or 

filtration from aslurry sample of known weight. The biomass concentration is expressed 

aspercentage wetweight ofyeastsolids. Both proceduresare rapidbutsubjecttoanerror 

if trub is present. This can be minimized by treating with alkali, which dissolves some of 

the non-yeast solid material. 

Assessment of yeast biomass concentration based on wet weight introduces an error 

due to variable amounts of the liquid phase, which are trapped in the interstices of the 

packed cells. This can be overcome by taking asample of slurry of known weight or 

volume, washing the cells in water to remove the suspending medium and drying to 

remove both intra and extracellular liquid. Biomass concentration is expressed as dry 

weight per unit volume of slurry. This method provides ahigh degree of precision, but it 

is time-consuming. The second direct approach is to count the numbers of cells 

suspended in aliquid using amicroscope and ahaemocytometer counting chamber 

(Section 13.10). Competently performed the procedure is precise and repeatable. It is 

rapid and therefore suitable where analyses are required for calculation or checking of 

pitching rates, etc. Since the yeast is examined directly, there is an opportunity to identify 

abnormalities or gross contamination. When used in conjunction with a vital stain the 

proportions of viable and dead cells can be estimated. The result is not affected by trub, 

but errors accrue where the yeast is heavily flocculent or a chain-former. The operator is 

the principal source of error. 

The error associated with visual examination of yeast is eliminated by the use of 

electronic particle counters. These devices rely on suspending yeast in an electrolyte and 

passing the cells through a narrow orifice. Cells are registered in response to a change in 

electrical impedance. Most instruments can discriminate between particles of different 

sizes, reducing the error due to non-yeast solids. Nevertheless, any particle of similar size 


to yeast cells registers in the result. Small yeast flocs or chains of cells introduce a further 

source of error. Problems associated with flocs can be reduced by prior treatment with a 

deflocculating agent such as maltose. This method is rapid and is most suitable for 

checking relatively low yeast counts such as are encountered in newly pitched worts or 

cask beers at rack. 

The classical technique for determining cell concentration is the plate count. A sample 

of the test suspension is serially diluted and aliquots are spread onto plates of selected 

solidified medium. Plates are incubated under suitable conditions for the test organism 

and growth is allowed to proceed until discrete colonies are formed. These are counted 

and it is assumed that each arises from a single cell, therefore the colony count is directly 

related to the cell concentration in the test suspension. This approach has several 

advantages. It is possible to use selective media so that mixed populations of cells can be 

identified. It only detects viable organisms. The precision of the method is low since not 

all viable cells grow to form colonies and flocs or chains do not develop into discrete 

colonies. The method is slow, although it is possible to use a rapid slide culture technique 

where cells are detected as micro-colonies. This reduces incubation times from a few 

days to a few hours. This method provides historical data only and is mainly used for 

checking strain purity and for detecting contaminants. 

Other indirect methods of biomass measurement are calibrated against one of the two 

direct approaches. Thus, an empirical relationship is established between biomass 

concentration and the measured parameter. Most of these approaches have no value in 

brewing. Two methods are used in brewing because they employ sensors which provide 

automatic in-line measurement of biomass concentration. The first detects yeast cells by 

light scattering. The second relies on the dielectric properties of intact cell membranes. 

The optical device uses a sensor that detects yeast in response to near infra-red 

radiation (Reiss, 1986). The control system doses yeast into wort to achieve a predetermined 

set-point of light scattering. This set-point is empirically derived for each 

yeast strain. A dual beam arrangement corrects for light scattering due to non-yeast solids 

in the unpitched wort. The shortcomings of this approach are that it does not correct for 

non-yeast solids present in the yeast slurry, it cannot be used with very flocculent strains 

and it requires a separate viability correction. 

From an electrical standpoint, yeast cells suspended in beer or wort comprise a 

conducting medium and conducting cytoplasm, separated by a non-conducting plasma 

membrane. This juxtaposition allows cells to function as capacitors when subject to 

radiation of a suitable wavelength. In the case of yeast cells this is in the range of radio 

waves (Harris et al., 1987). The measured capacitance is proportional to the total volume 

fraction bounded by membrane within the operating field. Since yeast cells of a given 

strain, grown under defined conditions, are of a relatively constant size, measured 

capacitance is directly proportional to yeast biomass concentration. A biomass probe 

based on this principle is used in production brewing. The measured capacitance can be 

calibrated against cell number or a derivative of cell mass. A separate calibration curve is 

required for each strain. 

It has been successfully applied in systems for the automatic control of both yeast 

pitching and cropping (Boulton et al., 1989; Boulton and Clutterbuck, 1993). This 

method has several advantages. It has a very wide operating range and can be used to 

quantify cell concentrations in pitching yeast slurries without the need for dilution. 

Providing the cell suspension is homogenous, the concentrations of flocculent, nonflocculent 

and chain formers can be determined with equal facility. Non-yeast solid 

materials do not interfere since they do not function as capacitors. For the same reason, 


Table 13.1 Methods for determining biomass concentration 

Method Timing Comments 

Direct methods: 

1. Determination of % wet 

weight of yeast in slurry 

2. Determination of % dry 

weight of yeast in slurry 

3. Cell count with 

haemocytometer and 

microscope 

4. Electronic particle counters 

Rapid 

Slow 

Rapid 

Rapid 


Suitable for direct weight of pressed yeast cake 

Suitableforanalysisofpitchingyeastslurriesbycentrifugationorfiltration(IOB,1997(Section1.15.1,p.9)) 

Errors due to entrained trub 

Not suitable for low yeast counts (cask beer counts, pitched worts) 

Reference method but requires 3 days at 110 ëC for reliable answer 

Errors due to entrained trub 

Suitable for all production analyses but requires skilled operator (IOB, 1997 (Section 1.15.1, p. 9)) 

No error due to trub 

Not suitable for very flocculent strains 

Provides opportunity to view yeast and detect abnormalities or gross contamination 

Can be used to determine viability in conjunction with vital staining 

Suitable for quantifying yeast counts at low concentrations (cask beers, pitched worts etc.) (IOB, 1997 

(Section 1.15.1, p. 9)) 

Not suitable for very flocculent and chain-forming strains 

Possible errors due to entrained trub

Indirect methods: 

1. Plate count Slow Cells inoculated onto solid medium and after incubation colonies are counted. It is assumed that each cell 

gives rise to a separate colony (IOB Methods of Analysis, 1997) 

Not all cells in original suspension may be cultivable 

Errors with flocculent cells 

Can be made selective by choice of suitable medium 

2. Nephelometry Rapid Light scattering at visible wavelengths used for rapid measurement of cell concentrations in laboratory 

research studies. Must be calibrated against other reference method since response is non-linear 

NIR light scattering used in a commercial automatic in-line pitching rate control system (Reiss, 1986) 

3. Analysis of cellular 

constituents 

Slow Biomass concentration expressed relative to the concentration of a macromolecular constituent such as DNA 

or protein 

Only useful as laboratory research tool 

4. Metabolic activity Rapid Biomass concentration inferred from a measure of metabolic activity such as the rate of oxygen uptake, CO2 

generation, exothermy 

Only used as a research tool 

5. Radiofrequency permittivity Rapid Biomass probe which infers cell concentration from capacitance measured in response to radiofrequency field 

(Harris et al., 1987) 

Requires separate calibration for each yeast strain 

Suitable for laboratory and in-line use 

Responsive only to viable fraction of yeast slurries 

Automatic in-line systems for control of cropping and pitching (Boulton et al., 1989; Boulton and 

Clutterbuck, 1993) 


theprobecanbeusedtoquantifyyeastmixed withinertsupportmaterialsinimmobilized 

cell reactors. Yeast cells with disrupted plasma membranes, and which score as being 

non-viable with avital stain such as methylene blue, lose their dielectric properties and 

are not detected. Thus, the probe is responsive only to viable cells. 

For any given yeast strain, grown under defined conditions, cell number and biomass 

weight are positively correlated. However, the relationship is dependent on the 

physiological state of the cell. For example, prolonged storage of pitching yeast results 

in the depletion of intracellular glycogen. In brewing yeast this reserve material can 

account for up to 30% of the cell dry weight (Section 12.5.7). Thus, in such aslurry, a 

plot of cell numbers would show little change with time until apoint is reached when 

cells die and lysis occurs. However, asimilar plot of cell dry weight would show a 

gradual decrease throughout the storage phase as glycogen reserves were dissimilated. 

13.3 Batch culture 

In abatch culture cells are inoculated into amedium with afinite supply of nutrients. 

This is the most common situation in commercial fermentations, including brewing. It is 

the most usual mode of growth in natural habitats. Anumber of distinct phases can be 

recognized (Fig. 13.1). These phases are common to all micro-organisms that reproduce 

by cell division. However, the budding habit of Saccharomyces yeast introduces some 

specific features. The duration of each phase is dependent on cultural conditions, the 

nature of the growth medium and the physiological condition of the inoculum. The lag 

phase represents aperiod of adaptation when cells undergo atransition from one set of 

conditions to another. Thus, cells in the inoculum may have to synthesize the enzymes 

that are needed for the uptake and utilization of substrates present in the medium. In 

addition, abrupt changes in temperature or osmotic potential may induce a stress 

response, which requires aperiod of recovery from before growth commences. 

Thecellspresent inaninoculumarenotidentical.Inthecase ofbuddingyeast,suchasS. 

cerevisiae, there may be differences in physiological condition, stage in the cell cycle and 

cellular age. The duration of the lag phases of individual cells varies, and afinite period is 

required before the metabolic changes associated with adaptation are completed. The result 

is a progressive transition from lag phase to growth phase and hence, the period of 

accelerating growth. The growth rate gradually increases until it reaches amaximum and 

constant value. This is the exponential growth phase where the population increases 

logarithmically. The nature of the organism, the composition of the medium and physical 

conditions such as temperature, pressure and degree of agitation determine the rate. Cells 

removed at this stage and transferred to afresh batch of similar medium will continue to 

grow exponentially without alag phase. Individual cells within the exponentially growing 

population divide at different times, but the graph of biomass against time increases 

smoothly since it represents the averaged concentration for the whole population. Such 

cultures are termed asynchronous. Various experimental procedures can be followed to 

synchronize the cell cycles of the entire population. For example, treatment with amitotic 

inhibitor causes cells to arrest at the same stage in the cell cycle. When the inhibitor is 

removed all the cells progress through the cell cycle in synchrony. Synchrony is usually 

maintained for the first two or three rounds of budding and plots of biomass concentration 

against time are stepped. After this time the culture gradually returns to the usual 

unsynchronized mode. In the case of budding yeast the loss of synchrony is hastened by the 

differingcellcycletimesofmotheranddaughtercells(Section11.7).Theexponentialphase 


Cell concentration 

1 2 3 4 5 6 7 

Time 

Fig. 13.1 Growth of yeast in batch culture on amedium such as wort in which sugar is the 

principal carbon source. For adetailed explanation see the text. The phases indicated are: 1, lag 

phase; 2, period of accelerating growth; 3, exponential growth phase; 4, decelerating growth phase; 

5, stationary phase; 6, diauxic shift phase; 7, second growth phase on ethanol if oxygen present 

(solid line), or death phase if no oxygen present (dotted line). 

of growth may end when a by-product of metabolism increases to an inhibitory 

concentration. More usually it terminates when an essential component of the medium 

becomes depleted. Eventually, the concentration of the essential nutrient falls to alevel at 

which it limits the growth rate and the culture moves into the decelerating phase of growth. 

Whengrowthceasesthecellsenterthestationaryphase.Duringthisphasethebiomass 

remains at aconstant level. This apparent inactivity is illusory since metabolic activities 

of various kinds continue. Aerobic cultures of S. cerevisiae growing on glucose as the 

principalcarbonsourcehaveabi-phasicgrowthcurve,termeddiauxie(Fig.13.1).During 

the first phase, cells grow at the expense of glucose. The effects of catabolite repression 

(Section 12.5.5) ensure that metabolism is fermentative and ethanol is produced. During 

the first stationary phase, the absence of glucose relieves catabolite repression. After a 

period of adaptation and providing oxygen is available the derepressed cells are able to 

utilize the ethanol and asecond phase of growth occurs. 

When no further growth is possible the cells remain in the stationary phase, often 

referred to as the G0 phase. During this time cells undergo changes that maximize their 

chances of survival. Predictably, this involves ashutting down of unnecessary metabolic 

pathways. Rates of protein synthesis are 300-fold lower in stationary phase yeast cells 

compared with those in the exponential phase (Fuge et al., 1994). Other changes occur 

that influence the metabolism and morphology of cells. Glycogen is dissimilated to 

provide carbon and energy for cellular maintenance. Some of the carbon is used to 

synthesize trehalose, which increases tolerance to stress (Section 12.5.7). Cell walls 

become thickened and more resistant to enzymatic degradation. Cells are more resistant 

to heat and desiccation. Cytoplasmic vacuolation becomes more extensive, presumably 

reflecting an increased requirement for turnover of intracellular protein. 

The metabolic changes associated with entry into the stationary phase are mediated at 

the gene level. Probably the RAS cyclicAMP signal transduction pathway controls entry 

into and exit from stationary phase. It is assumed that the cell has amethod for sensing 

the concentrations of nutrients and is able to initiate the necessary changes for entry into 

the stationary phase when it is apparent that the supply of an essential nutrient is about to 

disappearfrom themedium.Certainly,specificgenes appear tobeinducedjustbefore the 

stationaryphasecommences.Othersareexpressedonlyinstationaryphasecells.Therole 

of many of these remains to be elucidated (Werner-Washburne et al., 1996). 


The stationary phase is transitory. The strategy of cells is to adapt to a physiological state 

that offers the longest survival time. During the early stationary phase, readily dissimilated 

sources of carbon such as glycogen are utilized. When these are depleted the cell will, if 

necessary, utilize structural macromolecules. If no nutrients become available the stationary 

phase terminates in death. The gross indication of this is a decrease in biomass 

concentration due to cell lysis; there is a loss of cellular integrity and a concomitant 

release of intracellular contents into the medium. The liberated cellular components can be 

utilized by other, still viable cells and even promote limited, cryptic growth. 

The different growth phases in a batch culture can be described mathematically. The 

rate of increase of biomass (x) with respect to time (t) can be expressed as: 

dx 

ˆ x 13:1 

dt 

Where is a constant termed the specific growth rate and has the unit of reciprocal time 

(h 1 ). During the lag and stationary phases it is equal to zero. It increases throughout the 

phase of accelerating growth and achieves a maximum value ( max) during the 

exponential phase. 

During exponential growth the rate of increase of biomass concentration becomes: 

dx 

dt ˆ maxx 13:2 

After integration and where x0 equals the initial biomass concentration, this yields the 

equation: 

lnx lnx0 ˆ maxt 13:3 

This can be rearranged to give the fundamental exponential growth equation. 

x ˆ x0e … maxt† 

From this equation it follows that a plot of lnx versus t is linear with a slope equal to max. 

The exponential phase of growth is conveniently quantified in terms of the time taken 

for the population to double in size. The doubling time (t D) is equal to: 

tD ˆ log102 

ˆ 0:693 

max 

max 

The number of doublings (n) which occur to reach a final biomass concentration of 

given by: 

is 

x 

n ˆ log2 13:6 

x0 

…x† 

n ˆ 3:32 log10 

x0 

At the end of the exponential phase, the growth rate becomes dependent upon the 

concentration of a growth-limiting substrate [S]. The relationship between growth rate 

and substrate concentration is expressed in the Monod equation: 

ˆ max 

‰SŠ 

Ks ‡ ‰SŠ 

The constant, Ks, is termed the saturation constant and is equal to the substrate 

concentration where equals half max. Plots of against substrate concentration show 


13:4 

13:5 

13:7 

13:8

saturation kinetics. The inflexion point at which growth rate becomes independent of the 

substrate concentration varies with different strains. 

13.3.1 Brewery batch fermentations 

The pattern of growth in abrewery batch fermentation is similar to that shown in Fig. 

13.1. However, some aspects require further discussion. The initial yeast concentration in 

brewery fermentations is relatively high and the subsequent growth extent is modest. 

Thus, inoculation rates in atypical medium gravity (10ëP: SG c.1040) fermentation are 

approximately 3.0g/l wet weight of yeast, equivalent to around 0.6g/l cell dry weight, or 

in terms of cell numbers, approximately 1 10 7 cells/ml. During fermentation there are 

usuallynomorethantwotothreecelldoublings.Theconsequenceofthisisthattheyeast 

in the inoculum plays an important part in the subsequent fermentation. 

The inoculum is usually derived from aprevious fermentation and has been stored for 

aperiod before re-pitching. During the storage period the yeast is starving. It is usual to 

minimize the duration of the storage phase and to reduce metabolic activity by chilling to 

2±3ëC (35.6±37.4ëF). Nevertheless, the physiological state of the inoculum coupled with 

the conditions established at pitch influence subsequent patterns of growth. During the 

lagphasethepitchingyeast(whichisusuallysuspendedinbeer)adaptstoexposuretothe 

nutrients present in wort. In particular, the yeast is exposed to oxygen (see Sections 12.6 

and 12.7). The formation of sterol is accompanied by the concomitant mobilization of 

glycogen reserves (Fig. 13.2). There is alinear relationship between the quantities of 

glycogen dissimilated and sterol synthesized during fermentation (Quain and Tubb, 

1982). The formation of sterol, coupled with glycogen mobilization, is aprerequisite for 

the cells to move from the lag phase to the accelerating growth phase (Section 12.6). 

The events of very early fermentation in relation to the cell cycle are reviewed by 

Boulton and Quain (2001) (see also Section 11.7). Pitching yeast cells are in stationary 

phase (G 0) and unbudded. Transfer to wort induces a relatively synchronized shift to G 1 

phase and progress to START. In the first few hours there is no change in cell number or 

wort gravity. However, cell volume increases by approximately 20% and biomass falls by 

a similar percentage. During this time sterol synthesis takes place at the expense of cellular 

glycogen and molecular oxygen. Within six hours of pitching, almost 90% of the yeast 

population is budding, indicating that nearly all cells have passed into the S phase. The 

increase in budding index (relative number of budded cells) is very rapid, indicating a high 

degree of synchrony during very early fermentation. The achievement of high budding 

index corresponds with the onset of biomass increase and decrease in wort gravity. 

Glycogen 

(% cell dry wt.) 

40 

Glycogen 

0.12 

0.10 

0.08 

0.06 

Sterol 

0.04 

0.02 

0 

0 40 80 120 160 

0 

200 

Time (h) 

30 

20 

10 

Sterol 

(% cell dry wt.) 

Fig. 13.2 Changes in the intracellular concentrations of glycogen and total sterol in yeast 

measured during the course of a laboratory stirred fermentation using a lager yeast strain and 12 ëP 

all-malt lager wort (S. G. P. Durnin, unpublished results). 


The increase in mean cell volume observed immediately after pitching is transitory. As 

the growth rate increases into the exponential phase, the mean cell volume decreases to a 

value similar to or slightly smaller than that seen at pitch. In the exponential phase, the 

population of yeast is made up of typically 25% multi-budded cells, 25% single-budded 

cells and 50% virgin daughters. The budding index also decreases from the initial high 

value and falls close to zero just after the mid-point of fermentation. Cell number 

increases to a constant value at the time that the budding index reaches a basal value. 

Biomass, measured as cell dry weight, declines slightly from the mid-point to the end of 

fermentation, reflecting the dissimilation of glycogen that occurs when growth has 

ceased. 

The decline in budding index to a value close to zero just after the mid-point of 

fermentation indicates that at this time growth is limited by the disappearance of a 

nutrient. In the majority of fermentations, made from all-malt wort, the limiting nutrient 

is probably oxygen and, by inference sterols and/or unsaturated fatty acids. In highly 

oxygenated worts, particularly those containing high levels of sugar adjuncts, nitrogen 

may be the limiting substrate. 

13.3.2 Effects of process variables on fermentation performance 

The batch growth curve is influenced by a number of physical parameters. These are; 

temperature, pitching rate, dissolved oxygen concentration, wort concentration and 

pressure. Adequate control of these parameters is essential to ensure consistent 

fermentation performance and beer quality. Specific effects can be ascribed to 

modulating each parameter in isolation, nevertheless, they are all to some extent 

interdependent. For example, increasing fermentation temperature reduces oxygen 

solubility. Similarly, a decrease in wort concentration increases oxygen solubility. When 

examining the effects of altering a single parameter it is essential to ensure that all others 

remain constant. Some of the effects described will be observed only if the parameters are 

varied over very wide ranges. These extremes will not be seen in real brewery 

fermentations because they are beyond the limits that would be used in practice, but they 

serve as useful illustrations of the underlying principles. 

An increase in the fermentation temperature reduces the time taken to attenuate the 

wort (Fig. 13.3). Thus, temperature exerts its effects primarily on yeast metabolic rate. It 

PG 

50 

40 

30 

20 

10 

25°C 

20°C 

15°C 

10°C 

0 

0 40 80 120 

Time (h) 

160 200 240 

Fig. 13.3 Effect of temperature on fermentation rate. Fermentations were performed at the 

temperatures indicated using pilot scale 8 hl cylindroconical fermenters, a lager yeast strain, pitched 

at a rate of 12 10 6 viable cells/ml into an all-malt lager wort with an OG of 1050 (12.5 ëP) and an 

initial dissolved oxygen concentration of 25 mg/l (Boulton and Box, unpublished data). 


has little effect on the extent of yeast growth during fermentation. Most brewing strains 

have an optimum temperature for growth between 30 and 34ëC (86±93.2ëF) but 

fermentations are maintained at lower temperatures. Superficially, the use of elevated 

temperatures for minimizing vessel turn-round times appears to be an attractive strategy. 

There are several disadvantages. At high temperatures, especially in cylindroconical 

vessels, fermentation is very vigorous. Rates of CO2 evolution are very rapid and losses 

of volatiles, via gas stripping, may become unacceptable. Indeed, beer losses due to 

uncontrollable fobbing may occur. To avoid this it may be necessary to operate 

fermenterswithalargefreeboard.Theconsequentreductioninvesselproductivecapacity 

offsets some of the gains made by shorter vessel residence times. At the end of 

fermentation, the time taken to chill is directly proportional to the fermentation 

temperature. 

Generally, ale fermentations are performed at higher temperatures (18 22ëC; 

64.4 71.6ëF) compared with lager types (8 15ëC; 48.4 59ëF). These temperatures 

are chosen largely because they produce adesired spectrum of yeast-derived beer flavour 

components. Deviating from the normal temperature ranges may produce totally 

unacceptable shifts in beer flavour. Increasing the fermentation temperature leads to 

increases in the concentrations of higher alcohols and esters during fermentation (see 

Sections 12.10.3; 12.10.4). These flavour changes are particularly undesirable in the case 

of pale lager beers. 

The concentration of oxygen supplied in wort at the start of fermentation is one of the 

primary regulators of yeast growth (see Sections 12.6, 12.7). Although wort composition 

is also influential, there is adirect correlation between the initial dissolved oxygen 

concentrationandtheextentofyeastgrowth.Elevation inoxygenconcentration resultsin 

an increased primary fermentation rate. Caution must be exercised in using very high 

oxygen concentrations as a means of reducing fermentation times. High oxygen 

concentrations produce rapid primary fermentations but this may be at the expense of 

ethanol yield. Theincreasedavailabilityofoxygen promoteshigh rates ofyeastgrowth at 

the expense of sugar, which would otherwise be available for ethanol formation (Fig. 

13.4). Since oxygen has adirect effect on the extent of yeast growth, it would be 

predicted that it would also influence the concentrations of flavour metabolites produced 

as aresult of yeast growth. This is indeed the case. With many yeast/wort combinations, 

but not all, there is an inverse correlation between initial oxygen concentration and beer 

ester levels. 

At low values there is adirect correlation between the yeast pitching rate and the rate 

of primary fermentation and the extent of yeast growth. At higher pitching rates an 

inflection point is reached beyond which yeast growth decreases and there is no further 

increase in primary fermentation rate (Fig. 13.5). These observations reflect the 

interrelationships between yeast pitching rate and wort composition, in particular the 

initial dissolved oxygen concentration. Brewery fermentations differ from most other 

commercial fermentations in that the size of the inoculum is high relative to the extent of 

subsequent growth, so the ratio between the initial yeast concentration and available 

nutrients becomes asignificant factor. 

At low pitching rates (below the inflection point), there is adirect relation with 

attenuation rate. Total yeast growth is probably limited by oxygen-dependent lipid 

synthesis. Thus, there is asurplus of oxygen relative to the number of yeast cells present. 

However oxygen is also expended in other non-lipogenic pathways (Section 12.5). The 

small initial yeast population becomes lipid replete but subsequent daughter cells are 

limited by the onset of anaerobiosis. Below the inflection point, yeast growth extent and 


Ethanol. conc. (g/l) 

Fermentation rate 

(time to half-gravity, h) 

80 

Yeast conc. 5 

60 

4 

40 

Ethanol 3 

2 

20 

1 

0 

0 10 20 30 

0 

40 

(Dissolved oxygen) (mg/l) 

100 

80 

Deg. 

fermented 

60 

60 

Fermentation 

40 

40 

20 

rate 

20 

0 

0 10 20 30 

0 

40 

Fig. 13.4 

(Dissolved oxygen) (mg/l) 

Effect of varying initial dissolved oxygen concentration on fermentation rate (expressed 

as time to reach half-gravity) and the formation of ethanol and new yeast growth during the 

fermentation of a 15 ëP lager wort with a lager yeast strain. The temperature was 11 ëC and the 

pitching rate was 15 10 6 viable cells/ml (redrawn from Bamforth et al., 1988). 

Fermentation rate 

(time to half-gravity, h) 

60 

40 

20 

Fermentation 

rate 

Growth 

0 

0 

0 10 20 30 40 

Pitching rate (viable cells/ml × 10 6 ) 

Fig. 13.5 Effect of pitching rate on fermentation rate (time to half-gravity) and new yeast growth 

(weight of crop ± weight of yeast pitched) for 8 hl pilot scale all-malt 15 ëP lager fermentations. The 

initial wort oxygen concentration was 25 mg/l and the pitching rate was 12 10 6 viable cells/ml 

(re-drawn from Boulton and Quain, 2001). 

pitching rate are directly related. With increasing pitching rate a point is reached where 

growth is limited by the nutrient supply available to each yeast cell. The quantity of 

oxygen supplied per yeast cell is probably the limiting nutrient. Attenuation rates remain 

high because of the high yeast population. The patterns of growth and fermentation rates 

shown in Fig. 13.5 will be modified by other parameters such as the initial wort oxygen 

concentration and availability of other nutrients. Thus, at very high initial oxygen 

concentrations the inflection point at which growth begins to decline will be shifted 

towards the right. The inflection will still occur because another nutrient will eventually 

become growth limiting. 


30 

20 

10 

Yeast growth (g/l wet wt.) 

Yeast conc. (g/l dry wt.) 

Degrees fermented

SG 

1050 

1040 

1030 

1020 

1010 

1000 

50% increase 

30% increase 

20% increase 

0 2 4 6 8 

Fermentation time (days) 

Fig. 13.6 The relationship between fermentation time and increasing the original gravity of wort 

from 1033.3 (8.3 ëP) in cylindroconical fermenters (redrawn from Hough et al., 1982). 

Increasing the wort concentration, without changing any other parameters, results in 

an increase in fermentation time. However, providing the pitching rate and wort dissolved 

oxygen concentration are increased pro rata, there is only a small increase in 

fermentation times (Fig. 13.6). This forms the basis of high-gravity brewing in which 

concentrated worts are fermented and subsequently diluted to sales gravity. Fermentation 

profiles are influenced by wort composition, such as variations in the spectrum of 

fermentable sugars. Some of these effects are strain-specific. Lager strains utilize 

maltotriose more rapidly than ale strains (Stewart et al., 1995). It would be predicted, 

therefore that a change in the concentration of this sugar would have different effects 

depending on the nature of the yeast strain. At limiting concentrations, fermentation rates 

are proportional to the concentration of -amino nitrogen (Fig. 13.7). In an all-malt wort 

the supply of -amino nitrogen should be adequate and this dependence should not be 

observed. Deficiencies may arise in high-gravity worts made with a high proportion of 

non-malt adjunct. 

Several distinct effects can be ascribed to pressure. Yeast is subjected to an osmotic 

pressure the magnitude of which is dependent on the concentration of the wort. In highgravity 

worts the osmotic pressure may be as high as 4 10 6 Pa (Owades, 1981). Yeast 

Fermentation rate (h) 

120 

100 

80 

60 

40 

20 

0 

0 20 40 60 80 

α-Amino nitrogen (mg/100 ml) 

Fig. 13.7 Relationship between fermentation time and -amino nitrogen of the wort (redrawn 

from Hough et al., 1982). 


cells appear unaffected by these pressures providing they are not subjected to abrupt 

changes. Osmotic pressures up to 10 8 Paare tolerated by yeast (Gervais et al., 1992). The 

osmotic potential of aqueous media is directly proportional to the concentration of 

dissolved solutes. This parameter is inversely related to the concentration of available 

water and is referred to as the water activity (A w). All organisms are able tolerate agiven 

range of Aw. This parameter appears to be of greater significance to brewing yeast than 

osmotic potential per se (Chapter 12; Section 12.3.1). 

In fermenters yeast is subjected to a hydrostatic pressure that is a function of the height 

of the vessel. In addition, if the exit of CO2 produced during fermentation is restricted, 

the vessel will become pressurized. Very high pressures are deleterious to yeast. Very 

high-pressure treatments (> 10 8 Pa; 100 Bar) have been used to sterilize some foods 

where heat treatments cause deleterious flavour changes. In similar fashion to the dual 

effects of osmotic pressure and water activity, pressurization of fermenters leads to a 

concomitant increase in the concentration of dissolved CO2. The latter may have a greater 

effect on yeast than pressure alone (Thibault et al., 1987). 

Moderate top-pressurization of fermenters (1.2 2.0 10 5 Pa, 1.2 2.0 Bar) has been 

used to reduce yeast growth and control fobbing. This approach has been used where 

changing other parameters, such as the use of very concentrated worts or elevated 

temperature, has resulted in unacceptable perturbations in the formation of flavour 

metabolites due to increased yeast growth (Nielsen et al., 1986, 1987). The effects due 

to pressure are strain-specific. Using a spheroconical fermenter Posada (1978) reduced 

yeast growth by the application of pressure. However, depending on the yeast strain this 

was accompanied by both upward and downward shifts in the concentrations of various 

flavour metabolites. Miedaner (1978) described a protocol for high-temperature 

fermentations in which vessels were allowed to pressurize to approximately 1.8 

10 5 Pa (1.8 Bar) at a point when the wort was approximately 50% attenuated. This 

resulted in a reduction in yeast growth and levels of higher alcohols compared to 

unpressurized lower-temperature fermentations. Comparative analysis of beers suggested 

that this was at the expense of some damage to yeast. Thus, the pH of the trial 

beer was higher and head retention values were reduced, possibly due to the presence of 

greater than normal concentrations of short chain fatty acids. Both of these effects 

suggest yeast autolysis. 

13.4 Yeast ageing 

Most analyses of yeast growth consider the dynamics of whole populations of cells. 

Biomass is estimated by cell number or cell weight. Where population analyses are 

performed they are often rudimentary, for example assessments of the proportions of 

living and dead cells. Where the viable population is analysed, as in the so-called vitality 

tests, most assessments are based on mean values of all the cells within the sample. These 

approaches assume that there is no heterogeneity within yeast populations. Budding 

yeasts differ from organisms that reproduce by binary fission in that individual cells have 

a finite life span determined by the number of times that it buds and its DNA is replicated. 

Replicative age is distinct from chronological age. When cells are unable to bud further 

they become senescent and ultimately they die (Jazwinski, 1999; Powell et al., 1999). 

Replicative age of yeast is measured relatively easily in that each budding event leaves a 

characteristic scar on the cell wall. This can be visualized with fluorophores such as the 

dye, calcofluor. The maximum number of times individual strains are able to bud is 


strain-specific and varies between about 15 and 40, usually +/ about 10. This number is 

termed the Hayflick limit (Hayflick, 1965). 

Ageing is accompanied by morphological changes. The cells gradually take on a 

wrinkled and granular appearance. There is apositive correlation between cell age and 

cell volume (Barker and Smart, 1996), and for alager strain aplot of cellular size versus 

age was linear. There was asixfold difference between the size of young mother and 

senescent cells. Typically, generation times increase sharply just before the onset of 

senescence. Cells entering the senescent phase commonly fail to separate from daughter 

cells. The senescent phase culminates in death. In higher eukaryotes this process is 

termed apoptosis or programmed cell death. It is not arandom process. In multicellular 

eukaryotes cells are continually dying and being replaced. Death appears to be under 

genetic control and occurs when cells have degenerated, possibly as aconsequence of 

damage by reactive oxygen radicals (Madeo et al., 1999). The apoptotic pathway has not 

been positively identified in yeast but it is possible that cells undergo suicide for similar 

reasons. In yeast, death is accompanied by autolysis, literally self-digestion. The process 

involves loss of membrane integrity and the breakdown of cellular macromolecular 

components by avariety of hydrolytic enzymes. 

Yeast cell ageing may serve as amodel for the same process in higher eukaryotes, 

including man. In brewing it may also have relevance to fermentation management. 

Brewing yeastpopulations may beheterogeneouswith respect toage.Given the relation 

between replicative age and cell size it might be supposed that old and young cells 

would have different sedimentation characteristics in the cones of cylindroconical 

fermenters. This was the case in a2000hl fermenter (Deans et al., 1997). Yeast cells at 

the bottom of the cone were, on average, older than those at the top. The fermentation 

performance of the older yeast fraction was significantly poorer compared to younger 

cells. This prompted the suggestion that the first portion of crops from these vessels 

should be discarded. In another investigation (Quain et al., 2001) contrary results were 

obtained. In this case yeast cells of different average replicative age were fractionated 

using sucrose gradient centrifugation. The fermentation performance of each fraction 

was compared. The larger older cells produced fermentations with faster attenuation 

rates compared to the smaller younger cells. It is difficult to reconcile these 

diametrically opposite results. In both investigations there were significant differences 

between fermentation performance and replicative age suggesting that more detailed 

investigations are needed. 

13.5 Yeast propagation 

Theoretically there is no limit to the number of times that yeast may be serially cropped 

and re-pitched. Some traditional breweries, particularly those using top-cropped ale 

strains, have followed this practice for many years without interruption. Most modern 

breweries periodically introduce new cultures of yeast of guaranteed identity and purity 

derived from laboratory stocks. This is done for several reasons. Where several yeast 

strains are used within the same brewery low levels of contamination are inevitable and 

there is aconstant threat of contamination with wild yeasts and spoilage bacteria. Acid 

washing(Section17.6)shouldcontrolspoilagebacteria,butithasnoeffectonwildyeast. 

With prolonged serial fermentation the characteristics of the production yeast may 

change due to genetic instability. Petite mutants, which lack the ability to form functional 

mitochondria, are very common in brewing yeasts. Petites ferment abnormally (Ernandes 


et al., 1993). Other types of genetic instability have been observed in brewing yeast, 

particularly changes from arelatively non-flocculent to aflocculent character (Section 

11.8.2).Bottom-croppingyeastfromcylindroconical fermenterscanselectforthesemore 

flocculent variants. This effect may be exacerbated with repeated serial fermentation. 

Bottom cropping has been associated with other undesirable effects. Prolonged serial 

cropping can result in a progressive enrichment of pitching slurries with trub and other 

non-yeast particulate matter. Not only is this trub added to the next fermentation, it 

results in an underestimate in the calculation of pitching rate where this is determined by 

measurement of spun solids. Bottom cropping may tend to select for larger and therefore 

older cells (Smart and Whisker, 1996; Deans et al., 1997. This effect may contribute to a 

gradual decline in the performance of brewing yeast with generational age. 

Typically, yeast is serially repitched between 5 and 20 times before disposal. This 

wide range reflects the importance that individual brewers place on the need to introduce 

new cultures. Conversely, it reflects the threat that individual brewers consider is posed 

by prolonged serial re-pitching. Propagation attracts both capital and revenue costs. It is 

commonly asserted that newly propagated yeast does not produce standard fermentation 

performance or beer, so there is a natural reluctance to propagate frequently, especially if 

existing yeast lines are performing in a satisfactory manner. The suggestion that newly 

propagated yeast performs poorly is unproven and may simply reflect less than ideal 

propagation plant. The decision to introduce a new culture should be based upon 

microbiological and performance testing of existing yeast. The process should be 

managed so that a new culture is introduced when experience suggests that older cultures 

will be approaching the end of their useful lifecycles. 

Yeast propagation has three elements. Firstly, there is a need to maintain stock 

cultures. Secondly, the stock culture must be used to generate a laboratory culture of a 

scale sufficient to pitch the first brewery culture. Thirdly, the yeast must be propagated 

within the brewery to grow an amount sufficient to pitch the first production scale 

fermentation. 

13.5.1 Maintenance and supply of yeast cultures 

The complexity of the system used for the maintenance of stock cultures depends on the 

size of the operation. It must be a quality assured system in which cultures of guaranteed 

identity and purity are delivered to the brewery. Several levels of complexity are possible. 

Small breweries using a single yeast strain may hold stock cultures at independent thirdparty 

institutions such as the various national collections of yeast cultures. The onus for 

guaranteeing the quality of the supplied yeast is placed, at a cost, on the institution. This 

approach has the advantage of simplicity. Many brewers maintain their own strains. This 

can range from a requirement to look after a single strain at one brewery to multiple 

strains supplied to several breweries. Where a single company has to supply several 

satellite breweries and possibly a number of franchise breweries with a number of yeast 

strains it is convenient to have a dedicated central facility. This facility replaces the thirdparty 

operators and takes on the task of quality assurance of cultures and their supply. 

The satellite breweries have the much reduced burden although still essential task of 

assuring the supply of cultures from their own brewery laboratories into propagator and 

thence production. Alternatively, the central facility may undertake propagation and 

supply bulk yeast to breweries. 

There is a need to store cultures for long periods in such a way that they remain pure, 

at high viability and not subject to genetic change. Several methods are used and they 


fulfil these criteria with varying degrees of success. The simplest method is by periodic 

sub-culture using agar slopes (slants). These consist of small bottles, typically 

containing around 10ml of asuitable nutrient medium, solidified with agar. In order 

to maximize the surface area, the agar is allowed to solidify with the bottle placed at a 

slant, hence the name. The agar is inoculated with apure culture of yeast and incubated 

to provide aprofuse layer of growth on the surface of the agar. Slope cultures are stored 

at 2 4ëC (35.6 39.2ëF) to minimize yeast metabolic activity and prolong the 

maximum storage period. Periodically slopes are sub-cultured by transfer to fresh 

medium. This approach is simple and inexpensive and, providing skilled personnel 

perform it, should not result in loss of purity. It has the major disadvantage that while 

metabolism is slowed by cold storage it is not stopped. Prolonged storage results in loss 

of viability, nevertheless cultures can be held in this way for 4 6months. To allow a 

margin for safety, slopes should be sub-cultured every three months. The most serious 

disadvantage of this method is that over long periods of time and following multiple 

sub-culturing genetic drift and selection of non-standard variants has occurred (Kirsop, 

1991). 

More sophisticated storage methods seek to slow down metabolism further than can 

be achieved by chilling alone and thereby prolong storage times. Apopular method is 

that of lyophilization or freeze-drying. Cultures are rapidly frozen followed by drying 

under vacuum such that water is removed by sublimation. The process is performed in 

glass ampoules, which are sealed when drying is complete. These cultures can be safely 

stored for several months. Reactivation is achieved by breaking the ampoule and 

transferring the dried biomass to fresh liquid medium. This method is widely used but it 

has serious shortcomings. Freeze drying results in alarge overall reduction in viability. 

The fraction that remains viable appears to do so for several months, thereafter but 

usually up to 95% of the original cells die during drying. More worryingly, the viable 

fraction may undergo some degree of genetic disruption during freeze drying (Russell 

and Stewart, 1981). 

The death and deterioration that accompanies freeze-drying is probably caused by the 

formation of intracellular ice crystals (Morris et al., 1988). In other industries, where the 

use of dried yeast is commonplace, yeast is cultivated in amanner that manipulates 

physiology to render the cells less susceptible to the rigours of drying. Thus, cells are 

encouraged to accumulate trehalose, a well-recognized stabilizer of biological 

membranes (Section 12.5.7) and protectants are added during processing. Guldfeldt 

and Piper (1999) reported that ale and lager strains stored in liquid nitrogen retained 

100% viability after seven months storage. The same strains dried under vacuum in the 

presence of glucose suffered losses in viability between 10 and 99.6%. When yeast was 

subjected to an osmotic shock, in the form of exposure to sorbitol (20%w/v) prior to 

drying the viability decrease was reportedly much less. 

The adoption of dried yeast as the method of choice for bakers has also been 

recommended to brewers (Debourg and Van Nedervelde, 1999). Fermentation 

performance and beer quality at pilot and production scale were comparable with those 

produced using conventional pitching yeast. Dried yeast is used in parts of Africa for 

making African-style opaque beer (Chapter 16). It is suggested that the use of dried yeast 

mightreplaceconventionalbrewerypropagation.Itsuseobviatestheneedforpropagation 

andthesubsequentstorageofcroppedpitchingyeast.Theoretically,driedyeastshouldbe 

ofaconsistentphysiologicalconditionandnotsufferthevariabilitythatmightbethecase 

with conventional pitching yeast. On the debit side, bakers' yeast destined for drying is 

grown so that all cells are fully respiratory (Section 13.5). Such cells should be sterol 


eplete and have little or no requirement for oxygenation of worts. Proof is still required 

thatbrewingyeastinthisstatewillgivestandardfermentationperformanceandbeerwith 

anormalvolatilespectrum.Nevertheless,theuseofdriedyeastisanattractiveoptionand 

may well find application where an infrequent supply isneeded such as might be the case 

with franchise brewing, pub brewing and home brewing. 

The most effective (and most expensive) method of preservation and storage of yeast 

cultures is freezing in liquid nitrogen ( 196ëC; 320.8ëF). Cultures must be frozen in a 

controlled manner but once this is achieved the storage potential is measured in years. 

Furthermore,nochangesingenotypehavebeenreported(Kirsop,1991;Quain,1995).As 

with the method for bulk drying bakers' yeast, the cells to be frozen are grown under 

oxidative conditions. Prior to freezing yeast is suspended (c. 30% wet wt/vol) in a 

medium containing glycerol (5% v/v) as acryoprotectant. To maintain high viability the 

rate at which the temperature is lowered must be regulated. The temperature must be 

reduced slowly during the first phase, typically two hours from 20ëC to 30ëC (68 to 

22ëF). After this the yeast is rapidly cooled to 196ëC ( 320.8ëF) by immersion in 

liquid nitrogen. The underlying principle of the process is that during the initial slow 

phase the suspending medium freezes first. Consequently, the osmotic potential of the 

suspending medium increases, a phenomenon assisted by the presence of the 

cryoprotectant. This causes intracellular water to be released into the medium and in 

consequence the cells shrink. The gradual shrinkage and dehydration prevents 

intracellular ice crystal formation. 

Yeast is conveniently frozen in colour-coded sealed straws and held in purpose-built 

liquid nitrogen refrigerators. Yeast is recovered by plunging the straws into warm (37ëC) 

water, broaching using sterile scissors then transferring the thawed slurry to sterile liquid 

medium. This is used to prepare amaster culture from which slope cultures are prepared. 

At the same time checks of strain purity and possibly identity are made. The slopes are 

distributed to breweries for use in propagation. An ISO 9000 quality assured system for 

yeast preservation using liquid nitrogen, recovery of cultures and distribution to satellite 

breweries is described by Quain (1995). 

13.5.2 Laboratory yeast propagation 

The aim of laboratory propagation is to grow apure culture of yeast of sufficient volume 

to pitch the first brewery scale propagation vessel. The process is performed using 

traditional microbiological techniques and mainly glass apparatus, preferably by skilled 

personnel. The importance of this stage is often underestimated. Ensuring that the culture 

ispureisofparamountimportance.Althoughitiscustomarytoconfirm thatthecultureis 

free from contamination the test results may not be obtained until after the first brewery 

stage has been pitched. 

To generate the terminal laboratory propagation cultureit isnecessary togrow aseries 

of intermediate cultures of progressively increasing volume. Atypical protocol is shown 

in Fig. 13.8. Scale up factors are usually around 1:10. The initial stages use ageneralpurpose 

yeast medium such as yeast extract, peptone glucose (YEPG). The terminal 

phase uses sterile brewery wort. The whole process takes around two weeks. Growth of 

yeast is promoted by continuous aeration. Terminal yeast counts should be within the 

range 150±200 10 6 cells per ml at a viablity greater than 98%. The sub-terminal culture 

is carried out using an aspirator flask. The culture is aerated via a glass sinter fitted with a 

sterile gas filter. Good rates of oxygen transfer are encouraged by mechanical agitation 

using a magnetic follower. The outlet on the aspirator is fitted with silicone tubing and a 


Slope culture 

10 ml YEPG 

100 ml YEPG 

2l YEPG 

20l sterile wort 

Brewery seed vessel 

Wash all growth off slope 

with sterile YEPG medium 

Static incubation 

for 24h at 20°C 

Incubate on flask shaker 

for three days at 25°C 

Incubate for three days 

at 25°C with continuous 

agitation and aeration 

Incubate for three days 

at 25°C with continuous 

agitation and aeration 

Fig. 13.8 Protocol for the laboratory phase of yeast propagation. All stages except the terminal 

step use a semi-defined medium (YEPG: yeast extract, 5 g/l; peptone, 10 g/l; glucose, 20 g/l). 

coupling for attachment to the inoculation port on the apparatus used for growing the 

terminal laboratory culture. The latter piece of apparatus must fulfil three functions. It 

must be able to withstand autoclaving during sterilization of the wort. It must be suitable 

for growing pure cultures of yeast to high concentration without risk of contamination. It 

must be suitable for transferring yeast from the laboratory to the brewery, therefore it 

must be of robust construction to withstand the rigours of the production environment. 

Suitable apparatus for carrying out the terminal phase of laboratory propagation is 

shown in Fig. 13.9. The flask has atotal capacity of approximately 25 litres and is 

constructed from stainless steel. Several portstraverseatop plate thatcan beremoved for 

filling with wort and for cleaning. These ports are attached to lines for inoculation, 

sampling, gas inlet and outlet and transfer to the brewery seed vessel. During the growth 

phase the wort is aerated using air delivered via asterile gas filter and astainless steel 

sinter. The wort is agitated using amagnetic follower. Transfer of the culture to the 

brewery seed vessel is achieved using agas cylinder to provide motor gas delivered via 

the exhaust line. All gas lines are protected with sterile filters and all other lines are fitted 

with connectors, which are wrapped during sterilization. These are used to make aseptic 

connections when transfers are made. 

13.5.3 Brewery propagation 

The culture supplied by the laboratory is pitched into asmall seed tank and thence 

through afurther series of vessels of increasing volume until sufficient yeast has been 


Sterile wrapped 

connection to 

brewery seed vessel 

Top plate clamped to 

flask body via sterilizable 

silicone rubber O-ring 

Stainless 

steel flask 

Sterile 

gas filters 

Sterile wrapped 

connection for 

addition of inoculum 

Exhaust gas outlet/ 

transporter gas inlet 

Aeration inlet attached 

to stainless steel candle 

Ports in head plate fitted 

with sterile seals 

Magnetic follower 

Magnetic stirrer 

Fig. 13.9 Components of apparatus suitable for carrying out the terminal growth phase of 

laboratory yeast propagation. The flask has an operating volume of 20 l and a total volume of 25 l. 

generated to pitch the first production fermentation. From amicrobiological standpoint 

propagation is fraught with risk. Any contamination at this stage will have profound 

adverse consequences. The design of the plant, its operation and its internal finish must 

be to the highest hygienic standards. The propagation plant should be located within a 

room designed to minimize the risk of contamination. Access is restricted to essential 

personnel, all internal surfaces are made from hygienic materials, the air in the room is 

filtered and the atmosphere maintained under aslight positive pressure. 

Wort is used as the propagation medium. Preferably, it should be of the same quality 

as that used in fermentation. This is usually not essential from the point of yeast growth 

but as the spent wort will eventually be pitched with the yeast it should match the wort it 

is pitched into. Yeast should not be propagated using high-gravity worts (Cahill and 

Murray, 2000). For ale and lager strains propagated on 7.5, 11.5 and 17.5ëP worts there 

wasaprogressiveincreaseinthemeancellsizeanddecreaseinviability.Bearinginmind 

that ethanol is toxic and that actively growing cells are the most susceptible this is 

unsurprising (Section 12.5.9). Prior to inoculation the wort and vessel must be sterilized. 

The vessel may be sterilized empty and then filled with wort from the hot side of the 

paraflow. Preferably, propagation vessels are provided with a means of boiling wort in 

situ via steam jackets or direct injection of steam. Yeast growth is encouraged by 

provision of air or oxygen. Gas inlet and exhaust lines must be fitted with microbiological 

filters. Sampling and inoculation ports must be capable of aseptic operation, preferably 

sterilized by steaming. 

The performance of the first fermentation should fall within the normal range in terms 

of duration and the extent of yeast growth. The resultant beer should be within 

specification. These goals are not always achieved. Traditional propagators operate on 

the principle that the yeast should, as far as possible, be in the same physiological 

condition as pitching yeast. The assumption is that this will ensure that the newly 

propagated yeast will produce standard fermentation performance when it is transferred 


into the first wort. This is achieved by limiting the quantity of oxygen supplied to yeast 

during the growth phase and controlling the temperature at the same value or slightly 

higher than that used for fermentation. Traditional propagators tend to be of similar 

construction to fermenters. Thus, mechanical agitation is not provided and during the 

growth phase, aeration is not continuous. The downside to this approach is that the yield 

of yeast is low, usually no more than would be obtained from aconventional wort 

fermentation (typically, 50±60 10 6 cells per ml). The low yields associated with 

traditional propagation systems usually require several steps to generate sufficient 

biomass to pitch the first fermentation. Furthermore, scale-up factors between each phase 

are modest, not more than 1:5. Atypical regime would require four stages of 1.5, 6.0, 

30.0 and 150hl, respectively to generate sufficient yeast to pitch up to 800hl (500 UK 

barrels) of wort. The duration of the entire process, including the laboratory phase, takes 

several weeks. Thus, the process is slow, the use of multiple vessels is costly to install 

and operate. The risks of contamination are proportional to the number of vessels used. 

In modern breweries, which may use several yeast strains, traditional propagation 

systemscannoteasilysatisfytherequirementsforyeast.Anadditionalcomplicationisthe 

use of very large fermenters. Commonly, very large fermenters have been installed 

without concomitant upgrading of the propagation plant. To overcome this problem it 

may be necessary to continue using smaller fermenters in parallel with the larger ones 

simply to generate sufficient yeast. Alternatively, large fermenters can be part-filled 

when pitched with newly propagated yeast. 

To satisfy the increased need for propagation in large breweries it has been necessary 

to introduce methods for increasing the yield of yeast and accelerating the process. 

Consider abrewery producing beers with an average fermenter turn-round time of 12 

days, using six yeast strains. Assuming yeast is serially pitched for 15 generations after 

which time anew culture is introduced, the total life-time of each yeast culture would be 

(12 15) ˆ180 days. Assuming apropagation regime which required 35 days to 

complete (14 days laboratory stage +21 days brewery stage) the total time required to 

accommodate all six yeast strains would be 210 days. Clearly, in this case it would be 

necessary to duplicate some of the propagation plant in order to meet the needs of the 

brewery. 

Growing the yeast at a temperature higher than that used during fermentation 

accelerates the propagation process. High yields can be obtained by ensuring that 

conditions are fully aerobic. Aerobic wort propagation performed at 20±25ëC typically 

yields terminal yeast counts of 180±220 10 6 cells/ml. The growth phase requires 24± 

48hto complete. Using atwo-tank system, total turn-round times for the brewery phase, 

including cleaning and filling takes less than seven days. Providing the oxygen supply is 

discontinued when growth ceases, the physiology of the yeast remains catabolite 

repressed and is therefore, similar to that of pitching yeast (Section 12.5.8). 

Several propagation systems have been described (Geiger, 1993; Schmidt, 1995; 

Brandl, 1996; Ashurst, 1990; Boulton and Quain, 1999; Westner, 1999). All are provided 

withameans ofensuring thathigh ratesofoxygen transfercan beachieved.This requires 

two components. Firstly, asystem for delivering sterile oxygen into the growing yeast 

culture, usually a sparge ring, candle or similar device and, secondly, efficient 

mechanical agitation to ensure that oxygen is dispersed throughout the culture. The 

essential features are shown in Fig. 13.10. The aspect ratio of the vessel is relatively high 

to maximize the path-length for oxygen solution. Vessels are jacketed to facilitate 

attemperation by cooling. Either sterile wort is delivered to the wort via an in-line 

pasteurizer/heat exchanger, or wort can be sterilized in situ. Sterile oxygen is delivered at 


the base of the vessel via astainless steel candle. High oxygen transfer rates are ensured 

by provision of alarge mechanical rouser. If desired, the dissolved oxygen concentration 

may be regulated at aset point in afeed-back loop system using output from adissolved 

oxygen probe located just under the surface of the liquid. The process generates 

considerable foam. Vessels with large freeboards contain this. In addition, foam can be 

controlled by application of top pressure and reducing agitation and gassing rates in 

response to atrigger from ahigh-level probe. Propagation in this way is conveniently 

performed within two vessels. The first, seed vessel is asimilar to that shown in Fig. 

13.10. The operating volume should be around 8hl. The second vessel should be sized to 

achieve the desired pitching rate in the first fermentation. For example, apropagation 

vessel with an operating volume of 100hl and aterminal cell count of 200 10 6 /ml 

would generate sufficient yeast to pitch 1300hl of wort at apitching rate of 15 10 6 

cells/ml. In contrast to traditional propagation plant large step-up ratios can be used, 

typically 1:10 to 1:20. 

13.6 Fed-batch cultures 

During fermentation or aerobic propagation using wort, yeast metabolism is catabolite 

repressed by sugars (Section 12.5.8). This limits the yield of biomass to modest values. 

Biomass yields can be dramatically increased by growth under derepressing conditions. 

This can be achieved by using an oxidative carbon source such as ethanol or glycerol. 

Unfortunately,manyyeaststrainsincludingmostbrewingtypesgrowpoorlyonoxidative 

media. Biomass production can be encouraged by fed-batch cultivation, in which the 

nutrient medium is supplied to the growing yeast over aperiod of time. The nutrient feed 

rate is controlled with respect to growth rate so the sugar concentration remains at alow 

derepressing concentration since it is utilized as soon as it comes into contact with the 

cells. In practice, this can be achieved by inoculating amedium containing all necessary 

nutrients but alimited supply of sugar. After inoculation with yeast, growth is allowed to 

proceed until the small concentration of sugar is consumed. At this point additional sugar 

is added to the culture. Sugar addition is exponential at arate that mirrors the batch 

growth curve. Total biomass yields can be increased approximately fivefold using this 

procedure, compared to catabolite repressed cultures. 

Aerobic fed-batch techniques are used for the production of bakers' yeast. In contrast 

to brewing the aim of such processes is to generate biomass. In this respect formation of 

ethanol would be wasteful, hence the reliance on this method. Yeast is usually grown on 

an undefined medium such as molasses (Barford, 1987). The feedstock can be added in 

response to feedback from asensor, which measures the concentration of asugar. In this 

case addition rates are regulated to ensure that the sugar concentration remains at zero or 

avery low value. Most commercial systems add nutrientat afixed exponential rate based 

on empirical observation of the growth of the yeast being cultivated. In order to ensure 

that growth remains oxidative it is essential that conditions are aerobic at all times. Since 

biomass concentrations reach very high values growth vessels must be capable of very 

high rates of oxygen transfer. 

Fed-batchpropagationofyeasthasnotbeenappliedtotheproductionofbrewingyeast 

although it has been proposed (Masschelein et al., 1994; Naudts et al., 1997). 

Theoretically it is an attractive proposition since not only are biomass yields very high 

but derepressed yeast contains high concentrations of the essential membrane lipids, 

sterolsandunsaturatedfattyacids(Section12.7).Thesehighlipidlevelsshouldreduceor 


Cooling 

section 

Wort sterilizer 

Sterile inert gas supply 

and top pressurization 

system 

Heating 

section 

Circulation 

pump 

Cooling 

jacket 

Temp. and 

DO probes 

Wort ring main 

Level probes 

Variable speed 

mechanical 

rouser 

Working 

volume 

Internal 

baffle 

Stainless 

steel sinter 

Sterile oxygen supply 

Pitching 

main 

Fig. 13.10 Two tank aerobic brewery yeast propagation system. The working capacities of each vessel are 5 brl (8hl) and 85 brl (139hl), respectively 

(redrawn from Boulton and Quain, 1999). 


CIP 

Wort in

eliminate the requirement for wort oxygenation. There is areluctance to embrace this 

technology. The use of dried yeast in brewing is becoming more common (Fels et al., 

1999). Such yeast is grown using fed-batch cultivation. It seems likely that as approaches 

like this see more widespread adoption, so might the underlying growth technique. 

13.7 Continuous culture 

Batch cultures are closed systems in the sense that exhaustion of a nutrient or 

accumulation of agrowth-inhibitory metabolite eventually limits growth. If aconstant 

nutrient supply or ameans of removing inhibitory metabolites is provided there is no 

reason why growth should not proceed ad infinitum. This is the underlying principle of 

open or continuous culture systems. Two approaches are possible (Pirt, 1975). Aplug 

flow reactor consists of an extended culture vessel (Fig. 13.11). Acontrolled mixture of 

nutrient medium and inoculum are introduced into the vessel and pumped through it. The 

geometryofthevesselisarrangedsuchthatthereisaminimumofmixingofthecontents. 

As the medium and inoculum pass through the culture vessel the distance travelled is 

proportional to the stage in the growth cycle that would have been reached in a 

conventional batch cultivation. By manipulation of the flow rate, the dimensions of the 

vessel, the composition of the medium, inoculation rate and temperature it is possible to 

control the composition of the medium issuing from the fermenter. The process can be 

made truly continuous by separating the biomass from the process stream exiting the 

culture vessel and re-circulating a proportion to form the inoculum. Control of the 

conditions allow the establishment of asteady state. 

Thesecondandmostcommonapproachtocontinuousfermentationisthechemostat.In 

essence this consists of an attemperated stirred reaction vessel with an inlet medium feed 

and an outlet for removing product (Fig. 13.12). The pipework is arranged such that the 

volume within the reaction vessel remains constant. A variable speed pump controls the 

rate of medium inflow and product outflow. It is assumed that stirring is perfect such that 

incoming medium is instantly and homogeneously distributed throughout the growth 

vessel. Chemostat cultures are initiated by filling the growth vessel with medium. After 

Medium 

Stirrer 

Reactor – tubular helix 

Concentrated biomass return 

Continuous 

centrifuge 

Diverted 

product 

stream 

Product 

stream 

Fig. 13.11 A plug flow continuous reactor. In the scheme shown, a proportion of the biomass is 

recovered by continuous centrifugation and fed back into the flow of in-coming medium. 


inoculation,growthisallowedtoproceedforaperiodtogeneratebiomass.Afterthephase 

of batch growth the continuous phase is initiated by switching on the medium supply. 

The theoretical basis of chemostat operation can be understood starting with a 

consideration of the general growth equation, the derivation of which is described in 

Section 13.3. 

… t† 

xt ˆ xoe 

Medium 

in-flow 

Feed pump 

Stirred 

reactor 

Weir outlet 

Where x t is the cell population at time t; x o is the cell population at zero time and is the 

specific growth rate. The latter function is defined as growth rate expressed as a function 

of the total biomass concentration. Consequently, it has the units of reciprocal time. The 

growth rate is controlled by the same environmental factors that regulate growth in batch 

culture such as temperature and availability of nutrients. Each organism has a maximum 

growth rate ( max) that is determined by the genotype. It is expressed when growth is not 

limited by any external influence. 

The instantaneous growth rate within a chemostat is described by the equation 

13:9 

dx 

ˆ x 13:10 

dt 

The rate of loss of cells from the chemostat is given by 

Product 

out-flow 

Fig. 13.12 Simplified representation of a chemostat. The outlet takes the form of a weir. This 

ensures that the volume in the reaction vessel remains constant. 

dx 

ˆ Dx 13:11 

dt 

D is termed the dilution rate and is equal to the flow rate of incoming fresh medium 

divided by the volume of the growth vessel. It also has the unit of reciprocal time. 

Combining equations 13.10 and 13.11 gives an expression showing the net change in the 

concentration of biomass at any given time. 

dx 

ˆ x Dx ˆ x… D† 13:12 

dt 

From this equation it can be appreciated that for any given dilution rate, the population 

density will increase until a constituent of the growth medium becomes limiting and its 

concentration within the growth vessel tends towards zero. Under these conditions the 

biomass concentration and composition of the spent medium remain constant and are a 

function of the dilution rate. This situation is termed the steady state. As the dilution rate 

is increased the growth rate ( ) increases to accommodate the increased supply of 

nutrients. Eventually, the maximum growth rate ( max) is achieved and any further 

increase in dilution rate results in a progressive loss of biomass, termed washout. 


Both plug flow fermenters and especially chemostats have found widespread 

application as research tools. Chemostats are very powerful since they provide microbial 

biomass with a defined physiological condition. Furthermore, the consequences of 

changing physiological state on the biochemistry of the organism by manipulation of the 

medium composition, or growth conditions, can be readily assessed. Of greater 

significance here, they also form the basis of industrial processes. 

Continuous brewing fermentation is an attractive prospect since it provides amethod 

of harnessing the power of ahighly active yeast population of defined physiological 

condition. Theoretically, it is possible to construct aprocess in which acontinuous infeed 

of wort is rapidly transformed into acontinuous outflow of green beer. The yeast is 

actively growing at all times, thus the lag phase associated with batch fermentations is 

eliminated. The controlled and constant conditions within continuous fermenters should 

be reflected by beer of consistent composition. Since the process is continuous ipso facto 

there is no downtime due to vessel emptying, cleaning and re-filling. Similarly, since 

yeast is not cropped and retained for re-pitching, yeast handling is much simplified. 

In practice, there are also many disadvantages to continuous fermentation. The flow 

ratescanbevariedwithinonlycomparativelynarrowlimitsandonlyasinglebeerquality 

can be produced at any given time. Consequently, continuous systems are inflexible. 

There is aneed for aconstant supply of wort. Most breweries have plant suitable for 

discontinuous wort production (Chapter 6) and therefore a storage facility must be 

installed. The risks of microbial spoilage with bulk wort storage are considerable. The 

risks of contamination extend to the continuous fermenter. The consequences of 

contamination in the reactor are grave since start-up times and establishment of steadystate 

conditions are long. Some brewing yeast strains are genetically unstable (Section 

11.8.2). Selection of mutant strains in continuous fermenters can have disastrous 

consequences.Forexample,someproductionscalecontinuousprocessesrelyontheyeast 

being flocculent for retention within the vessel. Selection of non-flocculent variants 

results in washout. 

The nature of brewery primary fermentation precludes the use of achemostat. The 

assimilation of sugars by yeast growing on wort is an ordered process. The presence of 

glucose inhibits the assimilation of maltose (Section 12.4.1). Furthermore, the 

production of abalanced spectrum of flavour compounds is dependent on controlled 

yeast growth. In achemostat culture, yeast growth extent may be significantly different 

from that seen in batch fermentation. Similarly, the constant addition of fresh wort can 

result in abnormal sugar utilization. For these reasons, continuous primary fermentation 

is best performed using aplug flow type reactor. Unfortunately, it is impossible entirely 

to eliminate back mixing. To avoid these problems and by way of compromise, multistage 

continuous systems may be used. In this case the physical separation of the process 

liquid into discrete reaction vessels allows the essential characteristics of abatch culture 

to be incorporated into asemi-continuous process (Stratton et al., 1994; Williams and 

Ramsden, 1963). Alternatively, the fermenter takes the form of avertical column (tower) 

in which wort is introduced at the base (Seddon, 1975). Use of aflocculent strain not 

only retains yeast within the vessel but also provides a self-generating gradient of 

biomass through which the wort passes. The pinnacle of interest in continuous 

fermentation was reached during the 1960s. During this period many systems were 

developed, some of which were used at production scale. Unfortunately, their use was 

bedevilled by failures. With afew notable exceptions the use of continuous fermentation 

systems was discontinued. Since that time there has been a resurgence of interest, 

especially using immobilized yeast reactors, as described in Section 13.8. Afull review 


of the historical aspects and current status of continuous fermentation can be found in 

Boulton and Quain (2001). 

13.8 Immobilized yeast reactors 

Immobilized yeast reactors are refinements of continuous fermentation systems. In the 

latter, some of the yeast exits from the vessel with the spent medium. This necessitates 

the use of additional plant, usually a continuous centrifuge, for separation of yeast and 

product. In continuous immobilized yeast reactors the yeast is retained within the vessel. 

Fresh medium enters the reactor and passes through the yeast biomass where it is 

transformed into product. The latter exits from the reactor essentially free from yeast. 

Immobilized systems have several advantages compared to conventional continuous 

reactors. Process times are rapid and process efficiencies are high because of the 

combination of elevated biomass concentration and high volume throughput. Thus, 

because yeast loss is restricted, it is possible to use flow rates that would cause washout in 

a conventional chemostat. In an immobilized yeast reactor the relation between dilution 

rate and biomass concentration does not hold since cells are retained. Hence, it is possible 

to have high dilution rates and high biomass concentrations. High biomass concentration 

restricts growth, which also engenders high process efficiencies. High productivity allows 

the use of comparatively small immobilized yeast reactors. Clarification of the product 

stream is simplified because of the retention of biomass. In essence, the yeast in an 

immobilized system functions as a biocatalyst and no actual growth is necessary. The 

process has been defined as cells physically confined or localized within a specific region 

of space with retention of their catalytic activity, if possible or even necessary their 

viability, which can be used repeatedly and continuously (McMurrough, 1995). 

Several systems for immobilization are used (McMurrough, 1995). Retention of 

flocculent yeast within a tower fermenter using upward flow is the simplest method. It is 

of limited use since it can be used only with very flocculent strains and high flow rates 

eventually result in washout of biomass. There are four methods for true immobilization 

of yeast. These are retention by a semi-permeable membrane, attachment to a surface, 

entrapment within a porous polymer and colonization of a porous material. Membrane 

reactors suffer the major disadvantage that the rate of exchange of solutes is slow. In 

consequence, they have not found application in production scale reactors. 

Entrapment of yeast within the matrix of a porous polymer is possibly the most widely 

used method of immobilization. Polymers that have been used include polyacrylamide, 

calcium alginate, -carrageenan, agarose, pectin, chitin and gelatin (Godia et al., 1987). 

Polymeric supports are conveniently formed into beads, typically around 0.3 mm diameter. 

Beads of this dimension offer the best performance with regard to diffusion of nutrients and 

metabolic by products from the surrounding medium to the immobilized cells. The 

mechanical strength of the bead is a function of the degree of cross-linking of the polymer. 

The latter also controls diffusivity of nutrients and metabolites from the medium to the 

yeast cells. The latter parameter is vital to the efficiency of these bioreactors hence it is 

usual to sacrifice some mechanical strength in the interests of productivity. Since the beads 

are also compressible, entrapped yeast bioreactors are usually operated with the process 

flow directed in a vertically upward direction. Beads with very low mechanical strength can 

be disrupted by the evolution of gas bubbles and retention of yeast is comparatively poor. 

Yeast cells can be attached to inert surfaces such as wood chips, ceramics, glass, 

cellulose, stainless steel and various resins (Godia et al., 1987; Ryder et al., 1995). The 


mechanism of attachment is via acombination of electrostatic and hydrophobic binding 

(Mozes et al., 1987). For brewing applications the use of DEAE cellulose has found 

favour. This material is described as granulated derivatized cellulose (GDC) and is sold 

under the trade name Spezyme Õ (Cultor Finland). The beads have adiameter of 0.4 

0.8mm and are capable of bearing ayeast loading of 500 10 6 cells per gwet weight of 

carrier. Following prolonged use the beads can be regenerated by treatment with NaOH 

(2% w/v) at 80ëC (Pajunen, 1995). 

Colonization of porous materials is acombination of entrapment and surface binding. 

Several types of support have been used, the most promising of which are glass beads 

(SIRAN Õ ,Schott Engineering Company, Mainz, Germany) and ceramic rods containing 

amatrix of silicon carbide. The former consists of beads with adiameter of 1 3mm 

prepared from amixture of glass powder and salt. After the beads are manufactured the 

salt is dissolved to leave pores of 60 300 mdiameter. The proportion of each bead 

which is pore accounts for approximately 60% of the total. This gives biomass loadings 

of approximately 15 10 6 cells per gbead (Breitenbucher and Mistler, 1995). Ceramic 

cylinder supports have been described by Krikilion et al., (1995). Each cylinder element 

contains channels through which the process liquid flows. The channels pass through a 

poroussiliconcarbidematrix.Thesurfaceporesofthematrixareapproximately8±30 m 

in diameter and internal pores have adiameter of 100±150 m. The small outer pores 

allow entry of yeast cells, which then grow and colonize the comparatively larger lumen. 

Both glassand ceramicsupports are incompressible and so there islittle restriction on the 

flow rates that can be used. They can be used in reactors with upward or downward flow. 

Similarly, the robustness of the materials protects then from damage due to CO2 

evolution. On the other hand, because cells penetrate into the supports, they are more 

difficult to clean for regeneration purposes. 

All immobilized reactor systems have afinite life, which is limited by the gradual 

build-up of debris and deterioration of the support and the yeast population. The duration 

of the lifespan and the ease and cost of regeneration are important considerations when 

making achoice of support and reactor. Comparatively expensive support materials such 

as glass beads and ceramics require treatment with hot NaOH or an oxidizing agent such 

as hydrogen peroxide, followed by steam sterilization. Relatively inexpensive supports 

such as alginate beads or wood chips can be discarded after asingle use. 

Immobilization influences yeast physiology. From the standpoint of the brewing 

process some of the effects are beneficial others are not. Several reports describe 

enhanced rates of glycolysis and ethanol formation in immobilized cells compared to 

freely suspended cells (Galazzo and Bailey, 1989, 1990; Aires Barros et al., 1987). In 

contrast, rates of growth, as measured by biomass increase are reduced in the case of 

immobilized cells. In one report (Aires Barros et al., 1987) the rate of ethanol formation 

in an immobilized system was 50% greater than that of the same concentration of freely 

suspended cells, but the biomass yield was reduced by 30% and there were increases in 

trehalose andglycogen reservesin immobilized cells. These observations suggest thatthe 

immobilized yeast cells enjoy aprotected environment in agreement with the finding that 

they exhibit an increased tolerance to ethanol (Holcberg and Margalith, 1981). This is in 

accord with the assertion that micro-organisms in nature usually grow in association on 

surfaces, as in biofilms (Section 17.7.1). 

Adverse effects on yeast due to immobilization are probably related to the effects of 

restricted transfer of solutes. Immobilized systems involving colonization of surfaces or 

in polymeric beads provide a continuum of environments from surface to core. Cells that 

are buried deep in the matrix of the bead commonly exhibit morphological abnormalities 


such as pleomorphism, increased ploidy and failure of daughter cells to separate (Aires 

Barros et al., 1987; Koshcheyenko et al., 1983). These cells are likely to be subjected to 

very hostile conditions, such as starvation, anaerobiosis, reduced water activity, high 

ethanol concentration andhigh dissolvedCO 2.Attachmenttosurfacesmay, insomeway, 

disrupts the cell cycle. With carriers that rely on binding of cells to surfaces, 

morphological abnormalities are less common. This suggests that mass transfer has the 

greatest influence. Whatever the precise reason it explains the low growth rate that is 

characteristic of immobilized systems. 

The fact that immobilized systems function as biocatalysts where there is very limited 

growth is both astrength and aweakness. They are most suitable for use in processes 

where limited metabolism is required. Thus, they have been widely used for the 

production of low-alcohol beers by limited fermentation and for continuous diacetyl 

removal. In the first instance, ethanol formation can be discouraged by the use of a 

combination of low temperature and anaerobiosis (Breitenbucher and Mistler, 1995; van 

de Winkel et al., 1991). Debourg et al., (1994) demonstrated that immobilized yeast was 

as effective at removing wort carbonyls as freely suspended cells. Diacetyl removal is 

accomplished by passing green beer, ex-primary fermenter through areactor containing 

immobilized yeast (Pajunen and Gronquist, 1994). The high concentration of yeast 

removes diacetyl very rapidly. It is necessary to heat green beer (c. 10min. at 90ëC; 

194ëF) prior to passage through the bioreactor to ensure that all precursor -acetolactate 

is converted to diacetyl. 

The strictures that apply to the application of continuous processes to primary 

fermentation also hold for immobilized systems. Limited yeast growth and the need for 

ordered assimilation of metabolites to produce balanced quantities of flavour metabolites 

are problematic. Inmostcases the optionof using series ofmultipletanksofimmobilized 

yeast has been adopted in order to physically separate the process stages occurring in 

conventional batch fermentation. A fuller description of prototype and commercial 

immobilized yeast reactors used in brewing may be found in Boulton and Quain (2001). 

13.9 Growth on solid media 

In the majority of immobilized yeast systems the cells grow attached to inert surfaces. In 

this situation the cells obtain nutrients from the surrounding liquid medium and the 

surface merely forms asubstrate for attachment. An alternative growth habit is that of 

yeastgrowingonthesurfaceofnutrientmediasolidifiedwitheitheragarorgelatin.Yeast 

cells multiply and form masses, termed colonies, on the surface of the medium. Colonies 

are roughly circular viewed from above and dome shaped in section. Colony sizes range 

from 1 10 7 cells to more than 10 9 cells depending on the number per plate. The precise 

shapes of the colonies are often characteristic of individual strains and are used as an aid 

to identification. In brewing, the shape and colours of colonies that form on generalpurpose 

media such as Wallerstein Laboratory nutrient agar, can be used to check strain 

purity (Section 17.3.6). As discussed later, the giant colony technique is atraditional 

method for brewing yeast strain differentiation. 

The pattern of growth on solid media by yeast is a function of the genotype of the 

particular strain and how the cells respond to the available nutrients. The kinetics of 

colonial growth has been studied by Kamath and Bungay (1988). The colony increases in 

size at the periphery and in height. The rate of peripheral growth is linear and not 

exponential. It can be expressed by the following equation: 


t ˆKrt ‡r0 

13:13 

Where, rt is the linear radial growth rate, Krt is the radial growth rate constant, r0 is the 

radius at zero time, ris the colony radius at time t. 

In the giant colony technique, yeast is grown on wort medium solidified with gelatin. 

Plates are incubated for 3±6 weeks at temperatures of 15±18ëC; 59±69.4ëF (Hall, 1954). 

Several characteristics of colonies may be of diagnostic significance. Colonies are 

described asbeingmattorshiny.Theprofiles ofcoloniesviewedinsectionandfromabove 

may be simple or very complex. Variations in the peripheral margins of colonies include 

smooth circular, fringed, irregular and lobate. The surfaces of colonies range from smooth 

to concentric or radial striated and deep radial valleyed types. Variations in colonial 

profiles include convex, flat, with acentral dome, wrinkled and crateriform. The variety of 

colonial morphologiesimpliesthatthemassofcellsisheterogeneous. Thisispredictable in 

that cells in the growing colony do not all have equal access to nutrients in the medium and 

to oxygen. The variety of responses to these inequalities in nutrient supply explains the 

characteristics of colonial morphology. For example, the formation of afringe around the 

margin of a colony is a dimorphic response in which cells at the margins adopt a 

pseudomycelial form. This may represent amechanism by which peripheral cells extend 

the area over which they are able to assimilate nutrients. Other shapes represent the 

response ofindividualcellswithinthedevelopingcolony togreaterorlesser concentrations 

of nutrients and metabolic by-products. Colonies are highly organized structures. Agene, 

given the name IRRI, has been isolated that has no known role in cells growing suspended 

in liquid media but is required for colony formation on solid substrates (Kurlanzka et al., 

1999). As with the case of biofilms it is highly likely that colonies represent astrategy in 

which the population ensures survival by co-operative behaviour. 

13.10 Yeast identification 

The taxonomy of yeast, including brewing strains is discussed in Chapter 11 (Section 

11.2). The methods used for the cultivation of yeast for the purposes of identifying 

contaminants are described in Chapter 17 (Sections 17.3.5; 17.3.6). Here the methods 

used for the differentiation of brewing yeasts are described. Methods are roughly 

divisible into two groups, the traditional and the modern. Traditional methods are often 

based on conventional microbiological techniques. They rarely have taxonomic 

significance but have evolved within individual breweries to meet the need for 

identification and differentiation of proprietary brewing yeast strains. Modern techniques 

frequently have taxonomic significance. Usually they have been developed within 

academic research organizations and are not specifically targeted at brewing. Commonly 

they use sophisticated and costly apparatus. Traditional approaches are still used to avoid 

the cost and requirement for skilled operatives needed for many modern techniques. 

Modern methods may be used in confirmatory tests by third parties. However, the use of 

molecular genetic analyses is becoming more commonplace and inexpensive. These may 

soon be seen as the methods of choice for routine brewery QA testing. 

13.10.1 Microbiological tests 

Plating microbiological samples on solid nutrient media is routinely used for enumerating 

microbial populations. By using selective or chromogenic media the method can have 

diagnostic significance. The method most widely used for differentiating brewing from 


wild strains relies on the ability of the latter to grow in the presence of inhibitors such as 

Cu 2+ orthe antibiotic cycloheximide (Section 17.3.6). General-purpose media such as 

WLN help in the differentiation of brewing strains. WLN contains the indicator dye 

bromocresol green and this causes yeast colonies to take on avariety of shades of green. 

The particular shade of colouring is characteristic of some but not all strains. The plates 

mustnotoverloaded.Themethodneedsaskilledpractitionerwithexperienceoftheyeast 

strains being examined. The use ofgiant colony morphology isdescribed in Section 13.8. 

The requirement for several weeks' incubation reduces its usefulness. 

Differentiation of lager and ale colonies can be accomplished by plating out onto a 

medium containing 5-bromo-4-chloro-3-indoyl- -D-galactoside, known as X- -Gal 

(Tubb and LiljestroÈm, 1986). Lager, but not ale, strains contain -galactosidase which 

cleaves the chromogenic substrate to give a product which forms a blue precipitate. 

Colonies of ale yeasts remain colourless whereas lager types assume a blue colour. 

13.10.2 Biochemical tests 

Several tests have been developed, which probe the biochemical properties of individual 

yeast strains. Many of these tests are part of the standard armoury of the microbiologist. 

There is some overlap with the methods discussed in the previous section. Other 

techniques probe aspects of biochemistry that are more directly relevant to brewing 

performance. 

The ability of an organism to utilize selected carbon sources under aerobic conditions 

(assimilation) or anaerobic conditions (fermentation) has been widely used as a method 

of differentiation and identification. Several techniques based on this approach are used. 

They share in common the assessment of growth on a basal nutrient media supplemented 

with a variety of carbon sources. Basal media include peptone water and yeast nitrogen 

base (Section 17.3.6). Liquid media require the use of pure cultures, frequently indicator 

dyes are included and a trap for detecting any CO2 formed. Solid media are also usually 

inoculated with pure cultures. However, with replica plating (Section 17.3.6) it is possible 

to assess mixed populations. Commercial kits are available for identifying particular 

groups of micro-organisms, both bacteria and yeast (for example, API test kits, 

bioMeÂrieux, Marcy-l'Etoile, France). These take the form of strips containing wells each 

of which is filled with a basal medium and a supplement of a different carbon source. 

After inoculation and incubation, growth is indicated by the change in colour of an 

indicator dye. As with many of the general microbiological approaches, these methods 

are best suited to identifying wild yeast strains. One assimilation test does have value. 

Since lager strains possess -galactosidase and ale strains do not, the former are able to 

hydrolyse melibiose and grow on the released sucrose and galactose. 

Strain identification can be based on assessment of performance in laboratory wort 

fermentations. This approach is of obvious value since it assesses the properties that will 

be exhibited during use in the brewery. In this respect, many of the tests are of the 

trueness-to-type variety. Whilst these may not have taxonomic significance they are a 

useful means of checking for strain drift. Several pieces of equipment have been designed 

for carrying out laboratory fermentations. These range from small (100 ml) stirred 

hypovials (Quain et al., 1985) through to larger stirred fermenters with facilities for 

controlling headspace gas. The most commonly used piece of apparatus is the EBC tall 

tube. These are attemperated glass or stainless steel tubes with a capacity of two litres and 

a high aspect ratio (150 cm 5 cm diameter) supposedly reflective of a fermenter. They 

are usually used in banks of several tall tubes to permit simultaneous fermentations. 


Automated versions are available, some of which are fitted with devices for automated 

sampling and data acquisition (Sigsgaard and Rasmussen, 1985; Skands, 1997). 

Fermentation performance in tall tubes is assessed by five criteria; formation of ayeast 

head, formation of a yeast sediment, attenuation rate, degree of attenuation and 

clarification after fining. 

Laboratory fermentations can be used to assess other growth-related properties of 

yeast. For example, the effect of varying temperature. Thus, ale strains on average have 

higher maximum growth temperatures (37 40ëC; 98.6 104ëF) compared with lager 

strains (31.5 34ëC; 88.7 93.2ëF) (Walsh and Martin, 1977). The oxygen requirement 

needed for satisfactory fermentation performance varies between individual yeast strains 

(Section 12.6). Pitching aliquots of wort containing varying dissolved oxygen 

concentration then observing subsequent fermentation performance can assess this 

property. 

13.10.3 Tests based on cell surface properties 

The surface properties of yeast cells underpin two types of differentiating tests. These are 

firstly, those that assess the immunological properties of cells and, secondly, those that 

measure flocculation. 

Immunological analyses rely on interactions between antigens (the yeast cell or 

fraction thereof) and antibodies. The latter are obtained by prior inoculation of the 

antigen fraction into asuitable mammalian host, usually arabbit. Antibodies are raised in 

response to the antigen and these can be isolated and purified. Components of the yeast 

cell wall, principally mannan side chains, have antigenic activity (Section 11.6.1). 

Reactions between antigens and antibodies are specific. The complexes formed can be 

visualized by coupling a fluorescent molecule to the antibody. A much-used method is 

the ELISA technique (enzyme-linked immuno-absorbent assay), in which a membrane is 

coated with a layer of antibody, which is reactive with the antigen of interest. The antigen 

mixture is added to the membrane. The membrane is then washed, which removes all but 

the bound antigen. The complex is then treated with a second batch of antibody, the same 

as that attached to the membrane, although in this case it is conjugated with an enzyme. 

The second antibody fraction binds to the antigen-antibody complex. After a second wash 

to remove unbound antibody, a substrate is added, which is acted upon by the enzyme and 

in doing brings about a colour change. This colour change allows visualization and 

quantification of bound antigen. 

Methods of assessing yeast flocculence have a long history. They cannot be used to 

provide a positive identification but are useful nonetheless for assurance that yeast strains 

are behaving normally. Four methods have gained acceptance within the brewing industry. 

These are the procedures of Burns (Burns, 1941), Gilliland (1951), Hough (1957) and 

Helms (Helms et al., 1953). Of these only the Gilliland and Helms methods remain in 

common use. The Hough method tests for flocculation of aliquots of washed cells 

suspended in a solution of calcium chloride adjusted to pH 3.5 or pH 5.0. Yeast strains are 

classified based on whether flocculation occurs under these conditions. If yes, does 

addition of maltose disperse flocs? If no, does addition of ethanol result in flocculation? If 

no, does the addition of addition of a second yeast strain (NCYC 1108) result in coflocculation. 

In the Helms method a sample of pitching yeast is washed by suspension and 

centrifugation in a solution of calcium sulphate. The washed yeast is re-suspended in 

calcium sulphate adjusted to pH 4.5 in a graduated centrifuge tube. After incubation at 

20 ëC for 20 minutes flocculation is assessed based on the volume of sediment formed. 


Both the Burns and Gilliland procedures assess the occurrence of flocculation after 

yeast is recovered from laboratory wort fermentations. In the Burns procedure, the extent 

of floc formation is recorded when the yeast is suspended in beer, distilled water and 

acetate buffer, pH 4.6. In the Gilliland method yeast is grown on agar medium to obtain 

separate colonies. Fifty colonies are picked off the plate and each is inoculated into 5 ml 

trub-free hopped wort. After three days static incubation, at 27 ëC, each culture is 

examined for the formation of sediment. After pouring off all but 0.5 ml of the liquid the 

yeast is gently agitated and examined. Gilliland classified yeast into four groups on the 

basis of their flocculence characteristics. Class I types were completely dispersed 

throughout. Class II yeast sediment towards the end of the incubation. The sediment is 

granular in appearance when re-suspended. Class III types behave similarly to the Class 

II but the sediment is more difficult to re-suspend and forms large flakes. Class IV yeast 

sediments very early in the growth phase. The sediment re-suspends to form loose 

flakes. 

13.10.4 Non-traditional methods 

Several methods for differentiating yeast strains are relatively new at the time of writing. 

They fall into two broad groups. Firstly, those which are based on an analysis of cell 

composition and, secondly, those which probe the genome of the cell. Methods based on 

cellular composition rely on whole cell analyses, for example, pyrolysis mass 

spectrometry and Fourier transform infra-red spectroscopy. Alternatively, specific subcellular 

fractions can be extracted and analysed, for example proteins and lipids. Methods 

for strain differentiation based on cell composition require that the yeast has been 

cultivated under defined conditions in order to eliminate differences due to variations in 

physiological state. Genetic analyses have the advantage that the composition of the 

genome is relatively constant and independent of physiological state. 

Pyrolysis gas chromatography and pyrolysis mass spectrometry rely on heating a 

sample of biomass in an inert atmosphere to around 550 ëC (1022 ëF). This causes the 

cells to decompose into a mixture of low molecular weight volatile fragments (Goodacre, 

1994). The fragments are separated using gas chromatography or the more powerful 

technique of gas chromatography mass spectrometry (GC-MS). The pattern of fragments 

that are generated are characteristic for individual or closely related groups of strains 

(Timmins et al., 1998). Fourier transform infra-red spectroscopy provides a fingerprint of 

whole cells by measuring the interactions between infra-red radiation and intracellular 

components such as nucleic acids, proteins, membranes and cell wall polysaccharides. 

For microbial cells, the mid-IR range (4000 400 cm 1 ) provides the best resolving 

power. The method is capable of distinguishing bacteria at the strain level (Helm et al., 

1991). The method has been successfully used for differentiating brewing yeast strains 

(Timmins et al., 1998). The proteome of cells can be extracted and separated by 

electrophoresis and visualized using techniques such as Western blotting. Based on 

differences in genotype it would be predicted that proteome analyses would be different 

for individual strains and therefore of taxonomic significance. This has been verified in a 

study of 29 enological strains of S. cerevisiae (van Vuuren and van der Meer, 1987). 

Total fatty acids can be extracted from yeast using a solvent such as a mixture of 

chloroform and methanol. In the form of methyl esters, the mixture of fatty acids can be 

separated using capillary gas liquid chromatography. The spectrum of fatty acid methyl 

esters and their relative abundance has been used to differentiate 13 strains of S. 

cerevisiae (Augustyn and Kock, 1989). 


The most precise methods for strain differentiation are those based on analysis of the 

nucleic acids of the genome. These methods generate so-called genetic fingerprints 

(Chapter 11; Section 11.8.1). Yeast strains can be positively identified and differentiated 

using restriction fragment polymorphism (Schofield et al., 1995), polymerase chain 

reaction (de Barros et al., 1998) and karyotyping (Casey, 1996). 

13.11 Measurement of viability 

Viability of micro-organisms is usually defined as the ability to reproduce. This is a 

useful parameter to determine since it provides apredictive measure of the ability of 

yeast to reproduce when pitched into wort. In traditional microbiology, viability is 

assessed by plating out serial dilutions of suspension of cells onto asuitable medium. 

Each colony that develops after incubation is assumed to have derived from asingle cell, 

therefore the colony count is directly related to the number of viable cells in the original 

suspension. Using this approach the viable cell concentration is usually expressed as the 

number of colony forming units (cfu). The colony counting procedure takes several days 

to obtain aresult. There is arapid version of the test in which micro-colonies are allowed 

to grow on nutrient agar poured into awell in aspecially designed microscope slide. The 

micro-colonies can be counted using amicroscope. This reduces the time taken to obtain 

a result from days to several hours. Even using the rapid method, colony-counting 

techniques aretooslowtobeusedtosupportdecisionsregardingthefitnessofyeasttobe 

pitched. Rapid methods for the determination rely on the use of vital stains. These are 

dyes which are excluded from viable cells but not dead ones. Alternatively, they may be 

taken up by all cells and subject to metabolic modification by living cells. The latter is 

accompanied by acolour change. Several dyes have been used to assess viability. Some 

of these, together with their mode of action are shown in Table 13.2. 

The most used method for the assessment of yeast viability in breweries employs 

methylene blue and ahaemocytometer counting chamber (EBC Analytica Microbiologica, 

1992). The haemocytometer consists of aglass microscope slide, which contains 

two chambers of known volume. The bases of the chambers are divided into anumber of 

small squares to form agrid which facilitates counting. To determine viability asuitable 

dilution of yeast slurry is mixed with an equal volume of a solution of methylene blue and 

placed into the haemocytometer. Viable cells take up methylene blue and reduce the dye 

to the colourless leuco form. Dead cells cannot reduce the dye and stain blue. The relative 

proportions of colourless and blue cells are counted and by calculation the total and 

viable yeast count in the slurry is determined. 

Compared to plate counts, the methylene blue method overestimates viability. While 

at values greater than 90% agreement is reasonable, the disparity becomes increasingly 

marked the lower the true viability (Parkinnen et al, 1976). It has been suggested that 

other dyes are more reliable than methylene blue. In one study, methylene blue (with 

Safranin O counterstain), citrate methylene blue, alkaline methylene blue, citrate 

methylene violet and alkaline methylene violet were compared with plate count 

determinations of viability (Smart et al., 1999). A variety of strains were tested in various 

physiological conditions ranging from exponential phase, through stationary phase, 

starved and heat-killed. It was concluded that citrate methylene violet produced the most 

reliable results. Nevertheless, methylene blue continues to be used widely. Providing the 

analysis is performed by a skilled operator and the viability of the yeast is greater than c. 

80% it provides a reliable result. Counting cells stained with dyes can be automated to 


Table 13.2 Methods for measuring yeast viability 

Method Mechanism Ref. 

Methylene blue staining 

Viable cells reduce dye to colourless form, dead cells stain blue 

[1] 

Methylene blue with Safranin O counterstain 

Fluorescein diacetate 

Bis(1,3-dibutylbarbituric acid trimethine 

oxonol) [DiBAC4] 

Propidium iodide (+ fluorescein diacetate) 

ChemChrome Y 

Mg 1-aniline-8-naphthalene sulphonic acid 

(Mg-ANS) 

2-[N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl) 

amino]-2-deoxyglucose (2NBDG) 

Rhodamine 123 

Counterstain reportedly improves contrast between live and dead cells and possibly differentially stains 

stressed cells 

Esterases in viable cells cleave molecule to release fluorophor, fluorescein 

Anionic dye excluded by viable cells, non-viable cells are fluorescent 

Excluded by live cells, stains dead cells red by binding to DNA. With counterstain, viable cells are 

fluorescent green 

Taken up by viable cells and cleaved enzymically to release fluorophore 

Fluorophore only taken up by viable cells where it binds to proteins 

Fluorescent derivative of glucose only taken up by viable cells 

Cationic fluorophore taken up by viable cells with functional oxidative mitochondria. Not useful for 

viability measurement of anaerobic repressed yeast 

[1] IOB Methods of Analysis, 1997; [2] Jones, 1987; [3] Chilver et al., 1978; [4] Dinsdale et al. 1999; [5] Lloyd and Hayes, 1995; [6] Deere et al., 1998; [7] McCaig, 1990; [8] Oh and 

Matsuoka, 2002; [9] Dinsdale and Lloyd, 1995. 


[2] 

[3] 

[4] 

[5] 

[6] 

[7] 

[8] 

[9]

emove human error, for example, with electronic image analysis (Raynal et al., 1994). 

The radiofrequency permittivity biomass meter described in Section 13.2 is responsive 

only to the viable fraction of yeast slurries. Since it does not respond to the non-viable 

fraction it does not provide a measure of viability. A prototype modified laboratory 

version is described in (Boulton et al., 2001) which uses a combination of radiofrequency 

permittivity and a fluorescent vitality stain to detect both viable and non-viable yeast 

cells. Should this become available commercially it would allow automatic measurement 

of viability. 

Flow cytometry has been used for viability determinations. This apparatus forces a 

suspension of yeast through a small nozzle such that the cells form a single-file stream. 

The cells are counted by passage through a laser beam. In addition, devices of various 

types are provided for the detection of cells stained with specific dyes, most usually 

fluorescent types such as fluorescein diacetate (Lloyd, 1993; Bouix and Leveau, 2001). 

The procedure gave a good correlation between fluorescent staining and a plate counting 

technique. Unfortunately, flow cytometers are costly and not likely to find routine use in 

any but the most lavishly appointed routine quality assurance laboratories. However, the 

instruments are capable of much more than simple measurement of viability as discussed 

in the following section. 

13.12 Assessment of yeast physiological state 

The results of viability tests are used for the calculation of the quantity of yeast to be 

added to wort to achieve a desired pitching rate. They are also used to assess the quality 

of yeast. Typically, an arbitrary value is chosen, usually around 90%, below which the 

yeast is considered unfit for use. The assumption is made that if yeast viability is low then 

the viable fraction is probably stressed. In recent years, a number of additional tests have 

been proposed that seek to probe the physiological state of the viable fraction of yeast 

populations. These are `vitality tests' (Lentini, 1993). The rationale behind the need for 

these tests is that current fermentation practice exposes yeast to a plethora of influences, 

which together have the potential to influence yeast physiology in perhaps unexpected 

ways. These influences have been comprehensively reviewed by Heggart et al. (1999). 

They include environmental effects such as osmotic stress, barometric stress, oxidative 

stress, mechanical stress, pH effects and temperature. These are coupled with genetic 

effects such as mutation and growth-related effects such as yeast cell ageing. In addition, 

nutritional effects including starvation, effect of oxygen, CO 2 and ethanol. These effects 

in combination have the potential to influence yeast physiology in ways that affect 

subsequent fermentation performance. Variable yeast physiology is not detected by 

viability tests. 

It is argued that to assess the cumulative effects of these influences on yeast it is 

necessary to use methods which are more discriminating than simple differentiation between 

viable and non-viable. The results of these tests may be used as the basis of a simple decision 

to use or discard yeast. Preferably they provide a result that is predictive of fermentation 

performance and they should identify appropriate values for parameters such as pitching rate 

and wort dissolved oxygen concentration that will provide optimum fermentation 

performance and consistent beer analysis. Several types of vitality test have been suggested 

which assess different aspects of yeast composition and biochemical function. 

The most relevant measures of yeast condition are those that directly reflect the 

changes that occur during fermentation. It is possible to perform laboratory fermentations 


to predict performance at production scale, however, this is too time consuming to be of 

value. Rapid tests, which have been suggested, include measures of the rates of uptake of 

oxygen, uptake of glucose, evolution of CO2, ethanol formation and exothermy. 

Apparatus has been described that is capable of measuring some of these parameters. For 

example, the Bri vitality apparatus consists of an attemperated incubation chamber fitted 

with an integral dissolved oxygen probe (Kennedy, 1989). Providing a known 

concentration of yeast is placed within the chamber it is possible to measure specific 

rates of oxygen uptake. These are related to yeast sterol content and by implication 

provide a measure of the optimum oxygen requirement for efficient fermentation 

performance(BoultonandQuain,1987).Similarapparatusformeasuringspecificratesof 

CO2 evolution has been described (Muck and Narziss, 1988). 

The acidification power test measures the ability of yeast to acidify the medium both 

spontaneously (AP1) and in response to the addition of glucose (AP2) (Kara et al., 1988). 

Acidification occurs in response to proton extrusion and reflects the activity of plasma 

membrane H + ATPase. Maintenance of transmembrane potential requires metabolic 

activity, therefore the magnitude of AP1 reflects the availability of strorage 

carbohydrates. AP2 is fuelled by exogenous glucose and is indicative of glycolytic flux. 

The results of the acidification test as applied to pitching yeast correlate with subsequent 

fermentation performance (Fernadez et al., 1991; Mathieu et al., 1991; Siddique and 

Smart, 2000). 

The acidification power test has been subject to several modifications. In one (Patino 

et al., 1993) maltose was substituted with glucose based on the fact that this is the major 

sugar component of wort. These authors suggested that the changes should be assessed 

eitherviathemeasureofconductance,orascumulatativeacidifcationpower ±thesum of 

changes in proton concentration over the course of the test. In another approach 

(Isenrentant et al., 1996) the amount of sodiumhydroxide required to maintain aconstant 

pH was determined. This was claimed to avoid the limitations of the standard method 

when using very high-vitality yeast. 

Several cell components, for example glycogen, trehalose and sterol, vary in 

concentration in response to changes in physiological condition. The concentrations of 

many of these cell components are known to be influential on fermentation performance. 

Variation from the norm in the concentrations of some of these cellular components is 

indicative of inappropriate yeast management. 

Metabolism is driven by the energy released by hydrolysis of ATP. The energy status of 

the cell is defined as the adenylate energy charge, which relates the relative concentrations 

of AMP, ADP and ATP (Chapman and Atkinson, 1977). The concentrations of these 

metabolites can be measured using bioluminescence (Section 17.3.1). The approach can be 

expanded to include the concentration of inorganic phosphate by nuclear magnetic 

resonance. Thereby, phosphorylation potential can be determined. Intracellular concentrations 

of glycogen and sterol are related to oxygen requirements in fermentation (Boulton and 

Quain, 1987). Low concentrations of glycogen in pitching yeast are indicative of prolonged 

storage. Coupled with high sterol concentration, low glycogen is indicative of exposure to 

oxygen. Both metabolites can be determined by simple rapid tests. For example, glycogen 

by staining with iodine (Quain, 1981) and sterol by a spectrophotometric procedure using 

the polyene antibiotic, filipin (Rowe et al., 1991). Of these, determination of glycogen 

appears to be the most informative, particularly as an indicator of stressed yeast. The latter 

condition is associated with the presence of elevated concentrations of trehalose (12.5.7). 

Trehalose concentration can be rapidly determined by infra-red reflectance spectroscopy 

(Moonsamy et al., 1996). 


Many aspects of cellular composition and physiological state are amenable to 

investigation using staining techniques. In particular, using biological fluorescent stains 

in conjunction with flow cytometry (Edwards et al., 1996; Lloyd and Dinsdale, 2000). 

Choice of an appropriate stain allows measurement of intracellular pH, glycogen 

concentration, geneological age, ploidy, membrane competence, budding index and 

phase in cell cycle. The cost of flow cytometers continues to ensure that they are largely 

confined to research laboratories. Nevertheless, it is a very powerful technique and no 

doubt widespread adoption in quality assurance laboratories will be accompanied by a 

reduction in price. 


AIRES BARROS, M. R., BARROS, M. R., CABRAL, J. M. S. and NOVAIS, J. M. (1987) Biotechnol. Bioeng., 24, 

1097. 

ASHURST, K. (1990) Brew. Dist. Internat., 21, 28. 

AUGUSTYN, O. P. H. and KOCK, J. F. L. (1989) J. Microbiol. Meth., 10, 9. 

BAMFORTH, C. W., BOULTON, C. A., CLARKSON, S. P. and LARGE, P. J. (1988) Proc. 20th Conv., Institute of 

Brewing (Australia and N. Zealand Section), Brisbane, 209±218. 

BARFORD, J. P. (1987) `The Technology of Aerobic Yeast Growth'. In Yeast Technology, D. R. Berry, I. 

Russell and G. G. Stewart, eds, pp. 200±230, Allen & Unwin, Hemel Hempstead, UK. 

BARKER, M. G. and SMART, K. A. (1996) J. Amer. Soc. Brew. Chem., 54, 121. 

BOUIX, M. and LEVEAU, J.-Y. (2001) J. Inst Brew., 107, 217. 

BOULTON, C. A. and CLUTTERBUCK, V. J. (1993) Proc. 24th Cong. Eur. Brew. Conv, Oslo, 509. 

BOULTON, C. A. and QUAIN, D. E. (1987) Proc. 21st Cong. Eur. Brew. Conv., Madrid, 401. 

BOULTON, C. A. and QUAIN, D. E. (1999) Proc. 27th Cong. Eur. Brew. Conv., Cannes, 647. 

BOULTON, C. A. and QUAIN, D. E. (2001) Brewing Yeast and Fermentation, Blackwell Science, Oxford. 

BOULTON, C. A., MARYAN, P. S., LOVERIDGE, D. and KELL, D. B. (1989). Proc. 22nd Cong. Eur. Brew. 

Conv., Zurich, 653. 

BOULTON, C. A., BOX, W. G., CARVELL, J. and TURNER, K. (2001) Proc. 28th Cong. Eur. Brew. Cong., 

Budapest, 722. 

BRANDL, J. (1996) Brauwelt Internat., 14, 32. 

BREITENBUCHER, K. and MISTLER M. (1995) Proc. EBC Monograph, XXIV, Espoo, 77. 

BURNS, J. A. (1941) J. Inst. Brew., 47, 10. 

CAHILL, G. and MURRAY, D. M. (2000) J. Amer. Soc. Brew., Chem., 58, 14. 

CASEY, G. P. (1996) MBAA Tech. Quart, 33, 1. 

CHAPMAN, A. G. and ATKINSON, D. E. (1977) Adv. Microb. Physiol., 15, 253. 

CHILVER, M. J., HARRISON, J. and WEBB, T. J. B. (1978) J. Amer. Soc. Brew. Chem., 36, 13. 

DEANS, K., PINDER, A, CATLEY, B. J. and HODGSON, J. A. (1997) Proc. 26th Cong. Eur. Brew. Conv., 

Maastricht, 469. 

DE BARROS LOPES M., SODEN, A., MARTENS, A. L., HENSCHKE, P. A. and LANGRIDGE, P. (1998) Int. J. 

Systematic Bacteriol., 48, 279. 

DEBOURG, A. and VAN NEDERVELDE, L. (1999) Proc. 27th Cong. Eur. Brew. Conv., Cannes, 751. 

DEBOURG, A., LAURENT, M., GOOSENS, E., BORREMANS, F., VAN DER WINKEL, L. and MASSCHELEIN, C. A. 

(1994) J. Amer. Soc. Brew. Chem., 52, 100. 

DEERE, D., SHEN, J., VESEY, G., BELL, P., BISSINGER, P. and VEAL, D. (1998) Yeast, 14, 147. 

DINSDALE, M. G. and LLOYD, D. (1995) J. Inst. Brew., 101, 453. 

DINSDALE, M. G., LLOYD, D., MCINTYRE, P. and JARVIS, R. (1999) Yeast, 15, 285. 

EBC ANALYTICA MICROBIOLOGICA (1992) Vol. II, Method 3.1.1.2. 

EDWARDS, C., PORTER, J. and WEST, M. (1996) Ferment, 9, 288. 

ERNANDES, J. R., WILLIAMS, J. W., RUSSELL, I. and STEWART, G. G. (1993) J. Amer. Soc. Brew. Chem., 51, 

16. 

FELS, S., RECKELBUS, B. and GOSSELIN, Y. (1999) Proc. 7th IOB Conv. (Africa Section) Nairobi, 147. 

FERNANDEZ, S. S., GONZALEZ, G. and SIERRA, J. A. (1991) MBAA Tech. Quart., 28, 89. 

FUGE, E. K., BRAUN, E. I. and WERNER-WASHBURNE, M. (1994) J. Bacteriol., 176, 5802. 

GALAZZO, J. L. and BAILEY, J. E. (1989) Biotechnol. Bioeng., 33, 1283. 

GALAZZO, J. L. and BAILEY, J. E. (1990) Enz. Microb. Technol., 12, 162. 

GEIGER, E. (1993) Brauwelt Internat, 11, 430. 

GERVAIS, P., MARCHAL, P. A. and MOLIN, P. (1992). Biochim., Biophys., Acta, 40, 1435. 

GILLILAND, R. B. (1951) Proc. 3rd Cong. Eur. Brew. Conv., Brighton, 35. 

GODIA, F., CASAS, C. and SOLA, C. (1987) Proc. Biochem., April, 43. 


GOODACRE, R. (1994) Microbiol. Europe, 2, 16. 

GULDFELDT, L. U. and PIPER, J. U. (1999). MBAA Tech. Quart., 36, 1. 

HALL, J. F. (1954) J. Inst. Brew., 60, 482. 

HARRIS, C. M., TODD, R. W., BUNGARD, S. J., LOVITT, S. J., MORRIS, J. G. and KELL, D. B. (1987) Enz. 

Microbiol. Technol., 9, 181. 

HAYFLICK, L. (1965) Expt. Cell Res., 37, 614. 

HEGGART, H. M., MARGARITIS, A., PILKINGTON, H., STEWART, R. J., DOWHANICK, T. M. and RUSSELL, I. 

(1999). MBAA Tech. Quart., 36, 383. 

HELM, D., LABISCHINSKI, H., SCHALLEHN, G. and NAUMANN, D. (1991) J. Gen. Microbiol., 137, 69. 

HELMS, E., NOHR, B. and THORNE, R. S. W. (1953) Wallerst. Lab. Commun., 16, 115. 

HOLCBERG, I. B. and MARGALITH, P. (1981) Eur. J. Appl. Microbiol., Biotechnol., 13, 133. 

HOUGH, J. S. (1957) J. Inst. Brew., 63, 483. 

HOUGH, J. S., BRIGGS, D. E., STEVENS, R. and YOUNG, T. W. (1982) Malting and Brewing Science, Vol. 2., 

Hopped Wort and Beer, (2nd edn), Chapman and Hall, London and New York. 

IOB METHODS OF ANALYSIS (1997) Vol. 2, Microbiology, Inst. Brew., Clarges St., London. 

ISENRENTANT, D., GREENANS, W. and VERACHTERT, H. (1996) J. Am. Soc. Brew. Chem., 54, 110. 

JAZWINSKI, S. M. (1999) Expt. Gerontology, 34, 1. 

JONES, R. P. (1987) Proc. Biochem., 22, 130. 

KAMATH, R. S. and BUNGAY, H. R. (1988) J. Gen. Microbiol., 134, 3061. 

KARA, B. V., SIMPSON, W. J. and HAMMOND, J. R. M. (1988) J. Inst. Brew., 94, 153. 

KENNEDY, R. (1989) Brew. Guard., Sept, 57. 

KIRSOP, B. E. (1991) `Maintenance of yeasts'. In Maintenance of Microorganisms and Cultured Cells, A 

Manual of Good Practice, 2nd edn, B. E. Kirsop and A. Doyle, eds, pp. 161±182, Academic Press, 

London. 

KOSHCHEYENKO, K. A., TURKINA, M. V. and SKRYABIN, G. K. (1983) Enz. Microb. Technol., 5, 14. 

KRIKILION, P. H., ANDRIES, M., GOFFIN, O, VAN BEURAN, P. C. and MASSCHELEIN, C. A. (1995). Proc. 25th 

Cong. EBC, Brussels, 419. 

KURLANZKA, A., RYTKA, J., ROZALSKA, B. and WYSOCKA, M. (1999) Yeast, 15, 23. 

LENTINI, A. A. (1993) Ferment, 6, 321. 

LLOYD, D. (1993) Flow Cytometry in Microbiology, D. Lloyd ed., Springer Verlag, Berlin. 

LLOYD, D. and DINSDALE, G. (2000) `From bright field to fluorescent and confocal microscopy'. In 

Brewing Yeast Fermentation Performance, K. A. Smart, ed., pp. 3±9, Blackwell Science, Oxford. 

LLOYD, D. and HAYES, A. J. (1995) FEMS Microbiol. Lett., 133, 1. 

MADEO, F., FROHLICH, E. and LIGER, M. (1999) J. Cell Biol., 145, 757. 

MASSCHELEIN, C. A., BORREMANS, E. and VAN DE WINKEL, L. (1994) Proc. 23rd IOB Conv. (Asia Pacific 

Section), Sydney, 109. 

MATHIEU, C., VAN DEN BERGH, L. and ISERENTANT, D. (1991) Proc. 23rd EBC Conv., Lisbon, 273. 

MCCAIG, R. (1990) J. Amer. Soc. Brew., Chem., 48, 22. 

MCMURROUGH, I. (1995) Proc. EBC Monograph, XXIV, Espoo, 2. 

MIEDENER, H. (1978) Proc. EBC Conv, Symp. Monograph V, Zouterwoude, 110. 

MOONSAMY, N., MOCHOBA, F., O'CONNOR-COX, E. S. C. and AXCELL, B. C. (1996) J. Inst. Brew., 101, 203. 

MORRIS, G. J., COULSON, G. E. and CLARKE, K. J. (1988) Cryobiology, 25, 471. 

MOZES, N., MARCHAL, F. and HERMESSE, M. P. (1987) Biotechnol. Bioeng., 30, 439. 

MUCK, E. and NARZISS, L. (1988) Brauwelt Internat., 1, 61. 

NAUDTS, G., AERTS, G. and ISERENTANT, D. (1997) Proc. 26th Cong. Eur. Brew. Conv, Maastricht, 377. 

NIELSEN, H., HOYBE-HANSEN, I., IBACK, D. and KRISTENSEN, B. J. (1986) Brygmesteren, 2, 7. 

NIELSEN, H., HOYBE-HANSEN, I., IBACK, D., KRISTENSEN, B. J. and SYNNESVEDT, K. (1987) MBAA Tech. 

Quart., 24, 90. 

OH, K.-B. and MATSUOKA, H. (2002) Int. J. Food. Micrbiol., 76, 47. 

OWADES, J. L. (1981) MBAA Tech. Quart., 18, 163. 

PAJUNEN, E. (1995) Proc. EBC Monograph, XXIV, Espoo, 24. 

PAJUNEN, E. and GRONQVIST, A. (1994) Proc. 23rd IOB Conv. (Australia & New Zealand Section), 

Sydney, 101. 

PARKINNEN, E., OURA, E. and SOUMALAINEN, H. (1976). J. Inst. Brew., 82, 283. 

PATINO, H., EDELSEN, C. and MILLER, J. (1993) J. Amer. Soc. Brew. Chem., 51, 128. 

PIRT, S. J. (1975) Principles of Microbe and Cell Cultivation, Blackwell Scientific Publications, Oxford. 

POSADA, J. (1978) Proc. EBC Conv, Symp. Monograph V, Zouterwoude, 207. 

POWELL, C. A., VAN ZANDYCKE, S. M., QUAIN, D. E. and SMART, K. A. (1999) Microbiology, 146, 1023. 

QUAIN, D. E. (1981) J. Inst. Brew., 87, 289. 

QUAIN, D. E. (1995) Proc. 25th Cong. EBC, Brussels, 309. 

QUAIN, D. E. and TUBB, R. S. (1982) MBAA Tech. Quart., 19, 29. 

QUAIN, D. E., BOX, W. G. and WALTON, E. F. 1985) Lab. Practice, 34, 84. 

QUAIN, D. E., POWELL, C., HAMILTON, A., RUDDLESDEN, D. and BOX, W. (2001) Proc. 28th Cong. Eur. 

Brew. Conv., Budapest, 388. 

RAYNAL, L., BARNWELL, P. and GERVAIS, P. (1994) J. Biotechnol., 36, 121. 


REISS, S. (1986) MBAA Tech. Quart., 23, 32. 

ROWE, S .M., SIMPSON, W. J. and HAMMOND, J. R. M. (1991) Lett. Appl. Microbiol., 13, 182. 

RUSSELL, I. and STEWART, G. G. (1981) J. Amer. Soc. Brew. Chem., 39, 19. 

RYDER, D. S., BOWER, P. A. and BROMBERG, S. K. (1995) Proc. EBC Monograph, XXIV, Espoo, 175. 

SCHMIDT, H.-J. (1995) Brauwelt Internat., 13, 130. 

SCHOFIELD, M. A., ROWE, S. M., HAMMOND, J. R. M., MOLZAHN, S. W. and QUAIN, D. E. (1995) J. Inst. Brew., 

101, 75. 

SEDDON, A. W. (1975) MBAA Tech. Quart., 3, 130. 

SIDDIQUE, R. and SMART, K. (2000) `Predicting fermentation performance using proton efflux', In, 

Brewing Yeast Fermentation Performance, K. A. Smart, ed., pp. 46±54, Blackwell Science, Oxford. 

SIGSGAARD, P. and RASMUSSEN, J. N. (1985) J. Amer. Soc. Brew. Chem., 43, 104. 

SKANDS, B. (1997) Proc. 26th Cong. Eur. Brew. Conv., Maastricht, 413. 

SMART, K. A. and WHISKER, S. (1996) J. Am. Soc. Brew. Chem., 54, 41. 

SMART, K. A., CHAMBERS, K. M., LAMBERT, I. and JENKINS, C. (1999) J. Am. Soc. Brew. Chem., 57, 180. 

STEWART, G. G., ZHENG, X. and RUSSELL, I. (1995) Proc. 25th Cong. EBC, Brussels, 403. 

STRATTON, M. K., CAMPBELL, S. J. and BANKS, D. J. (1994) Proc. 23rd IOB Conv. (Asia Pacific Section), 

Sydney, 96. 

THIBAULT, J., LEDAY, D. and COTE, F. (1987) Biotech. Bioeng., 30, 74. 

TIMMINS, E. M., QUAIN, D. E. and GOODACRE, R. (1998) Yeast, 14, 885. 

TUBB, R. S. and LILJESTROÈ M, P. L. (1986) J. Inst. Brew., 92, 588. 

VAN DE WINKEL, L., VAN BEVERAN, P. C. and MASSCHELEIN, C. A. (1991) Proc. 23rd Cong. EBC, Lisbon, 

577. 

VAN VUUREN, H .J. J. and VAN DER MEER, L. (1987) Amer. J. Enol. Viticult., 38, 49. 

WALSH, R. M. and MARTIN, P. A. (1977) J. Inst. Brew., 83, 169. 

WERNER-WASHBURNE, M., BRAUN, E. I., CRAWFORD, M. E. and PECK, V. M. (1996) Molec. Microbiol., 1159. 

WESTNER, H. (1999) Brew. Dist. Internat., Sept. 24. 

WILLIAMS, R. P. and RAMSDEN, R. (1963) Continuous fermentation process and apparatus for beer 

production. British Patent 926847. 


14 

Fermentation technologies 


The object of brewery fermentation is to utilize the ability of yeast cells to convert sugar 

into ethanol and carbon dioxide as the major products of metabolism. The yeast also 

produces aseries of minor metabolites such as esters, higher alcohols and acids that 

contributepositively toflavour.Itisthesemyriadsofminor components thatcharacterize 

abeer brand and make it identifiable to adrinker. The chosen yeast must also control the 

elimination of undesirable flavour components arising from the raw materials or from 

fermentation. Much of this flavour improvement occurs in maturation (Chapter 15). 

Brewery fermentations are discussed in Chapter 12. In this Chapter the technology of 

fermentation is described with the objective of performing the process consistently to 

yield the highest quality beer at lowest cost. Low costs must be demonstrated as low 

capital cost to keep the depreciation element of brewery fixed costs as low as is 

achievable and as low maintenance cost. Minimum beer losses must also be achieved so 

that the maximum amount of beer is derived from the raw materials used. In this way 

brewery variable cost per unit volume of beer produced is lowered. To achieve these 

objectives requires careful selection of the yeast and the appropriate wort composition 

(Chapter 12). It also requires efficient operation of properly designed fermentation 

equipment. In this respect top and bottom fermenting yeasts must be considered. 

The nature of fermentation was not clearly established until the latter part of the 

nineteenth century and so the design of vessels to effectively carry out fermentation has a 

history of about 150 years. This design is fundamentally influenced by the mode of 

separation of the yeast, i.e., top or bottom fermenting yeasts. Cool, bottom fermentation 

was almost entirely restricted to Bavaria until 1840 when the rest of the world was using 

top fermentation (Christian, 1959). Some bottom fermenting yeast was smuggled to 

Czechoslovakia in 1842 and so Pilsen was established as a major brewing centre. About 

three years later a Danish brewer, J. C. Jacobsen, took bottom fermenting yeast from 

Munich to Copenhagen and so the Danish city also became established as important for 

brewing. Of great significance at about the same time was the introduction of a bottom 

fermenting yeast to Pennsylvania in the USA. Its use spread rapidly through the actions of 


immigrant German brewmasters such as Frederick Miller, Bernhard Stroh, Eberhard 

Anheuser and Adolph Coors. The use of bottom fermentation was thus spreading 

throughout Europe and North America at the same time as Pasteur began his 

microbiological research resulting in the development of atheory of fermentation, culture 

techniques for micro-organisms, and the principles of sterilization (Pasteur, 1860, 1876). 

Progress in controlling brewery fermentation could now be made as improvements in the 

microbiologicalaspectsoffermentationcouldfacilitateengineeringdevelopments.Ofcritical 

importanceinbottomfermentationwastemperaturecontrolandinparticularthemaintenance 

oflowtemperatures(0ëC,32ëFandless)forlongperiodsoftime.Thisoriginallyrequiredthe 

use of large quantities of ice until the compressor refrigerator appeared in breweries from 

about1873followingdevelopmentsinAustraliaandGermany.Thenext100yearssawthe 

abandonment of top fermentation throughout the world with the exception of the UK and 

Irelandandrapiddevelopmentsinthelateryearsofthetwentiethcenturytoincreasedbatch 

size fermentations in very large vessels (up to 6,000hl, 3,600imp. brl). 

There has recently been renewed interest in top fermented ale beers produced in small 

batches in `niche' breweries particularly in the USA. There has also been adesire to 

eliminate differences between top and bottom fermentation by seeking semi-continuous 

and continuous fermentation techniques, some of them employing immobilized yeast, 

when the traditional practices of top and bottom separation do not take place. 

Therearemarkeddifferencesinthebatchsizesoffermentationsystemsthroughoutthe 

world. This reflects the market in which the particular brewer operates. International 

brewers producing global brands will have plants with capacities of at least 10m. hl per 

annum(6m.imp.brl)andsometimesconsiderablygreaterthanthisintheUSAandJapan. 

Regional brewers in many countries brew successfully in plants of annual production in 

the range 0.1 to 1.0m. hl (60,000 to 600,000imp. brl) and craft or niche brewers would 

operate at levels of 1,000 to 10,000hl per annum. Fermentation processes are therefore 

successfully carried out in batch sizes ranging from 10 to 6,000hl. It follows that awide 

range of vessel types are used. 

Techniques of fabrication are available to make fermenting vessels of all sizes. 

Practical sizes are limited by economic factors. It is difficult to transport by road, rail or 

barge, vessels much bigger than 2,000hl capacity (1,200imp. brl) and vessels of greater 

volume must be built on site at greater cost. Companies must decide on the optimum size 

for their own operations. Fermentation technology, therefore, embraces astudy of: 

· fermenters for bottom and top fermentations and consequent yeast separation 

· fermentersforcontinuousandsemi-continuousoperationrequiringnoyeastseparation 

· fermentation control systems. 

Some basic principles first need outlining so that vessel design and operation is set in the 

context of the biochemical requirements of successful brewery fermentation (see also 

Chapters 12, 13). 

14.2 Basic principles of fermentation technology 

14.2.1 Fermentability of wort 

The main objective is to ferment wort to the desired gravity; this is often called the 

required degree of attenuation. The proportion of the wort dissolved solids (extract) 

which can be fermented is called the percentage fermentability of the wort: 


Original gravity Final gravity 

Fermentability…%† ˆ 

Original gravity 

The original gravity can be expressed in ëP (Plato), which measures the concentration in 

weight/weight terms as grammes of solids per 100 grammes of wort, or in ëSacch, which 

relates thespecific gravityoftheworttothatofwater takenas1,000.Finalgravitymeans 

thegravityofthewortwhenitisfullyfermentedsuchthataddingmoreyeastorleavingit 

longer will lead to no further fall in gravity. This lowest gravity is often called the 

attenuation limit gravity and when it is reached the beer is said to be fully attenuated. 

Therefore if wort of original gravity 48ëSacch (12ëP) is fermented to a gravity of 

12ëSacch (3ëP) the percentage fermentability is: 

‰48 12=48Š 100 ˆ75% or ‰12 3=12Š 100 ˆ75% 

The alcohol formed in the fermentation has alower density than water and so it 

decreases the final gravity. The final gravity, therefore, does not show the amount of 

extract left in the fermented wort. The attenuation limit gravity referred to above is 

therefore called the apparent attenuation limit and what is calculated by the equation is 

the apparent attenuation of the wort. To measure the real attenuation the alcohol must be 

removed, e.g., by distillation before determining the gravity. The real attenuation is 

approximately 80% of the apparent attenuation. The true factor published by Balling in 

1880 was 0.81, in modern practice the real attenuation can be obtained from the apparent 

attenuation by the use of tables. 

To determine the degree of attenuation achievable from agiven wort the attenuation 

limit is usually measured in the laboratory. Yeast, pre-washed with wort, is mixed with 

filtered wort, which is fermented at 25ëC. Aspecific gravity reading is taken after two 

days and then repeated twice aday until the gravity does not change. This gravity is the 

apparent attenuation limit. It is the highest apparent degree of attenuation that can be 

achieved by fermentation of all fermentable material in the extract of the wort. It is, of 

course, governed by the raw materials chosen to make the wort and the extent of enzyme 

activity in the brewhouse. 

In the brewery fermentation the apparent attenuation limit is not usually reached. 

Some brewers set aspecific attenuation limit for aparticular beer and attempt to reach 

this value in apreset time against apreset fermentation temperature regime. If there is a 

large difference between the final attenuation achieved and the apparent attenuation limit 

as determined in the laboratory then there is fermentable extract present in the beer and 

this represents a risk of supporting infection by yeasts and bacteria. This would 

subsequently cause the beer to go hazy and would create off-flavours. Generally, in 

modern large batch lager fermentations, the objective is to try to ferment the beer to the 

limit. This is not the object in the production of cask ale in the UK when residual 

carbohydrate is required in the beer to allow secondary fermentation and cask 

conditioning to take place (Chapter 21). 

14.2.2 Time course of fermentation 

Bydeterminingthespecificgravityofthewortattimeintervalsonecanfollowthecourse 

of fermentation. Typical fermentation profiles for ale and lager fermentations are shown 

in Fig. 14.1, which also includes typical temperature regimes although these vary 

between breweries. The decline in the specific gravity is matched by the growth of the 

yeast as sugars are metabolized and ethanol produced. The pH value of the wort falls as 


100

Fig. 14.1 Time course of fermentation for ale (a) and lager (b) beers. fa, level of fusel alcohols 

( g/l); e, level of esters ( g/l) t, temperature ëC. (Hough et al., 1982). 

ammonium ions and amino acids are taken from the wort by the yeast and organic acids 

are secreted (Chapter 12). Major flavour compounds, the esters and higher alcohols are 

released into the wort as the yeast grows and so increase in concentration as fermentation 

proceeds. 

The concentration of these flavour compounds is critical for the consistency of the 

beer brand and the flavour profile of the beer (Chapter 20). Brewers need to be aware of 

the decline in concentration of flavour compounds, which can occur towards the end of 

fermentation, and in maturation (Chapter 15). This can occur as volatiles are carried out 

with evolving carbon dioxide (`gas purging') or are re-absorbed by the yeast. Marked 

differences can be seen in the time course profiles of ale and lager fermentations and 

these differences are reflected in the flavour profiles of the resulting beers. It follows that 

the vessels for fermentation must be equipped to regulate these differences consistently 

so that consistent products can be brewed. 

14.2.3 Heat output in fermentation 

Fermentation is an exothermic process. The metabolism of wort sugars in brewery 

fermentation can be represented in general by the summation of the biochemical pathway 

discussed in Chapter 12: 


Glucose‡2Pi ‡2ADP !2Ethanol‡2CO2 ‡2ATP 

In this version of the equation energy is fixed and stored by the yeast cell as ATP. The 

free energy of reaction ( G) can be calculated for each stage of the process (Mahler and 

Cordes, 1969) and anet figure of 157 kJ/mole can be derived. However, the ATP 

generated is used in further reactions in the cell: 

2ATP !2ADP +2Pi Gˆ2 31 ˆ62kJ/mole 

Therefore the overall heat of production can be assessed as 157‡62 ˆ219kJ/mole of 

glucose fermented. If temperature is to be controlled during fermentation to ensure a 

consistent performance then this heat must be removed. The heat produced during a 

typical fermentation can thus be assessed from this figure (Anderson et al., 2000). 

Consider a12ëP wort; this will contain 12.6kg of extract/hl of wort of which, say, 

75% will be fermentable. The molecular weight of glucose is 180 and so 219kJ/mole of 

glucose is equivalent to …219=180† 1000kJ/kg of glucose, i.e., 1217kJ. Therefore 

fermentation of 1000hl of wort would yield in total: 

1000 …12:6 75=100† 1217 ˆ11:5GJ 

The heat is not uniformly given out but will reach apeak at maximum fermentation rate. 

This peak has been variously estimated but avalue with practical worth, is 0.22kg 

extract/hl/h (Fricker, 1978). The maximum cooling load for a1000hl fermenter would 

therefore be: 

1000 0:22 1217 ˆ0:26 GJ/h 

This heat must be removed from the fermenter if the temperature of fermentation is to be 

controlled. The amount of cooling required depends on the temperature range to be 

maintained. Frequently in the production of ales with rapid top fermentation the 

temperature is allowed to rise unrestricted over the first 48 to 60 hours of fermentation. 

Temperatures of over 22ëC (72ëF) can be reached. Cooling is usually then applied to 

lower the temperature to around 15ëC (59ëF) for abrief maturation (Chapter 15) before 

proceeding to conditioning or to cask racking. In a lager fermentation temperature of the 

fermenting beer is usually controlled so that it rises to no more than 13 ëC (55 ëF) after 

pitching at around 8 ëC (46 ëF) and is then lowered by progressive cooling to about 5 ëC 

(41 ëF). The beer will then proceed to the maturation and conditioning stage. An equation 

has been derived to calculate the amount of cooling required which can be applied to any 

fermentation situation (Anderson, et al., 2000): 

Q ˆ M Cp t 

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, 

and t is the required temperature change of the beer. Therefore if a fermenter contains 

1000 hl of beer with a specific heat of 4.05 and it is required to lower temperature from 

13 ëC to 5 ëC, the heat removed will approximately be (assuming gravity to be 1.00): 

1000 100 (13 5) 4.05 = 3.24 GJ 

There is frequently debate between brewers and equipment suppliers about the rate of 

cooling required to remove this quantity of heat. To control temperature during 

fermentation a rate of temperature reduction of 1 ëC/hour is often used. In the above 

example this would equate to: 


1000 100 3.24 =324 MJ/h 

Some brewers require arate greater than this to `crash cool' afermenter to precipitate 

yeast and achieve aconditioning temperature of

Conical nozzle 

with sightglass 

Jacket outlet 

Conical jacket 

outlet 

Thermometer 

CO2 injection 

cock 

Outlet 

cock 

Yeast cock 

with sightglass 

Pipe for CO2 

entry and 

pressure cleaning 

Jacket inlet 

Vessel cleaning 

and pressure 

delivery pipe 

CO2 washing 

lantern 

Conical jacket inlet 

Fig. 14.2 Cylindroconical fermentation vessel, (Hough et al., 1982). 

These vessels are usually 3±4 times taller than their diameter and work at pressures of 

1±1.5 bar above atmospheric. Tank diameters in the UK are usually 3.5 to 4.5m(11.5 

to 15ft.). Heterogeneous fermentations have been observed in tanks much greater than 

20m(66ft.) high and for this reason in recent developments the height of vessels has 

been kept to less than 15m(49ft.). This phenomenon has not been fully explained but 

special circumstances do apply to very large vessels (>2,500hl 1,500imp. brl) in 

relation to temperature gradients and cooling (Section 14.3.3). In classic European 

lager production more squat vessels are used where the ratio of diameter to height is 

3:1, there is atendency for 

increased production of higher alcohols at the expense of esters. This may be caused 

by increased amino acid utilization caused by the increased beer circulation generated 

byrapidcarbondioxideproduction.Increasingthesizeofthevessel canlowercostper 

unit volume. Doubling the size of the vessel leads to acost increase of about 35% 

(Maule,1977).Ingeneralthegreaterthevolumetosurfacearearatiothelowertheunit 


volume cost will be. Large vessels are sited in the open. These are subject to planning 

regulations in some countries. 

Cylindroconical vessels cannot be completely filled for primary fermentation. Alarge 

volume of foam is formed by the evolution of carbon dioxide and this could cause 

pressure release valves to block. The headspace volume of the tank should therefore be at 

least 25% of the pitched wort volume. From this discussion it is difficult to generalize 

about the optimum size and aspect ratio of fermenters. As ageneral rule filling the vessel 

shouldnot take longer than12 hours, irrespective ofthenumberofbrews beingruntothe 

vessel for fermentation. Continuous filling with fresh wort will lead to increased 

production of -acetolactate and result in enhanced maturation time for diacetyl removal 

(Chapter 15). It is useful in abrewery to have arange of sizes of vessels. If the brew 

length is 1000 hl (600 imp. brl) then vessels of 4,000±5,000 hl (2,400±3,000 imp. brl) 

capacity will be suitable but it will be advantageous to have vessels of 250 hl (150 imp. 

brl) size for propagation and some 500±1,000 hl (300±600 imp. brl) vessels for new 

product development and trial work. 

14.3.2 Construction of cylindroconical vessels 

Originally fermenting vessels for bottom fermentation were constructed of mild steel 

with a glass or epoxy resin lining. This lining had to be frequently inspected to ensure its 

integrity. Mild steel was also prone to rusting and modern vessels are almost always 

constructed of chrome-nickel stainless steels. 

Metals and design 

Generally the steels used are stainless and austenitic, i.e., containing carbon, which forms 

a carbide with the gamma form of iron, which is normally only stable at high 

temperatures but can be stabilized at normal temperatures by the inclusion of elements 

such as chromium, nickel, and molybdenum. These steels are often referred to in brewing 

by the general classification V2A and V4A, but these categories cover a series of 

different alloys. The class can be subdivided into AISI 304 (V2A group) and AISI 316 

(V4A group), which are the steels in most common use. Resistant properties of 316 steels 

are enhanced by the inclusion of molybdenum (Table 14.1). A further category of steels, 

designated 321 contain titanium. Normally 304 stainless is used. However V2A steels are 

not fully resistant to chloride ions or to pH values < 4.5. This is not usually a problem for 

fermenters but with liquors having high chloride contents 316 can be specified, but it is 

much more expensive than 304 steel. 

A very important factor is the surface smoothness of the steel that can be achieved in 

manufacture. It should be as smooth as is possible so that indentations cannot provide areas 

for potential microbial contamination. Cold rolling of stainless plate will yield a surface 

with a `roughness' (Ra) of 0.6 to 0.8 m. In some systems of work this type of finish is 

designated `2b'. Electro-polishing can lower Ra to 0.3 to 0.4 m but this will be at greater 

cost and is not always specified by brewers. Vessels are now almost always produced to 

Table 14.1 Composition of austenitic stainless steels (Barnes, 2001) 

Type Carbon (%) Chromium (%) Nickel (%) Molybdenum (%) 

304 0.03±0.06 17.5±19.0 8.0±12.0 ± 

316 0.03±0.07 16.5±18.5 10.0±14.0 2.25±3.0 


standards of design that can be specified by the brewer at the time of making an enquiry to 

purchase. Choice is thus facilitated which can concentrate on price aspects. Generally three 

design standards are recognized in Europe and North America: AD MerkblaÈtter (Germany), 

ASME (American Society of Mechanical Engineers) VII div 1(USA), and BS 5500 (UK). 

Vessels operating at pressures >0.5 bar will normally be subject to national regulations 

relating to pressure vessels and will require inspections for insurance purposes. 

Cooling jackets 

The fermenter must be equipped with acooling system to remove heat generated during 

fermentation and to allow the control of temperature to the required profile. If the vessel 

is to be used for maturation as well (Section 15.2.5) then there is the requirement to cool 

to less than 0ëC (32ëF) and to hold this temperature for several days. Systems of direct 

and indirect cooling can be used. In direct cooling ammonia gas is used as the refrigerant 

and cooling is achieved by expansion and evaporation of the liquefied gas. In indirect 

cooling asecondary coolant (such as IMS, industrial methylated spirit), generated from a 

refrigeration plant is used. This coolant is circulated through the fermenting vessels and 

thencereturnedtotheplant.Coolingjacketsareconstructedsothatgoodheatexchangeis 

possible and several designs are available (Fig. 15.2). The choice often depends on the 

expertise of the proposed manufacturer. Limpet coils are wound onto the vessel surface 

and seam welded, they have a relatively heavy construction. An alternative is the pressed 

corrugated `profile' plate (Barnes, 2001), which can be pre-formed and spot welded to the 

vessel shell. In some versions this is called a `dimple' jacket. There is the further 

alternative of a `quilted' panel, which is hydraulically inflated after spot welding to the 

vessel surface. This system has a low volume and is favoured with direct expansion 

cooling using ammonia. If properly specified and manufactured the design of the cooling 

jacket has little impact on fermentation performance. A fully equipped fermenter will 

probably have two cooling sections on the cylinder of the tank and a further section on 

the cone. 

Vessel fittings 

Vessels are filled and emptied from below, reducing oxygen ingress. Vessels are fitted 

with pipes for the addition of wort, the removal of yeast, and the removal of beer. There 

are also pressure relief and vacuum relief systems, which are fitted into a top-plate 

assembly. Cleaning-in-place (CIP) fluids must also be introduced and this is usually 

through the top plate although, for ease of access, the valve may be situated at the bottom 

of the tank. 

One of the main factors in the operation of these vessels is ensuring the integrity of the 

pipe-work systems so that yeast, wort, beer and cleaning fluids are, when required, 

handled separately and not allowed to mix. This is frequently achieved by using a `dial-apipe' 

system where connections are made manually. The different lengths of the Ushaped 

connection pipes make wrong connections impossible. The pathways so made are 

opened and closed by manual or remotely operated butterfly valves. 

In large modern fermenting `tank farms' the vessels are often permanently connected 

by the pipe-work system. Valves are then either connected to every tank or collected into 

a `valve routeing block' to which every tank is connected and which is automatically and 

remotely controlled. The problem in this system is valve leakage which can cause severe 

damage if, say, cleaning fluid leaks into the beer. The risk is reduced by the use of 

`double seat' valves that provide leakage indication. This assembly contains an upper and 

lower valve controlled by springs and separated by a small space. Opening the valve is 


achieved by compressed air, which overcomes the pressure of the springs. On release the 

compressed air escapes and the pressure of the springs takes over, this forces first the 

upper and then the lower valve to close. The small space between the valves still exists 

and if avalve is not properly sealed liquid will flow through the space and out of the 

assembly through apipe. Leaking is thus detected. These systems are now favoured with 

the desire to lower manpower costs in breweries, but are only of use if closely examined 

frequently! 

Carbon dioxide produced during fermentation must be removed from the vessel 

irrespective of whether this gas is collected for further use (Section 15.4.3). This gas 

pressure must, therefore, be released and the vessel must also be protected against the 

development of avacuum. CIP fluids must be safely and effectively introduced into the 

tank. For ease of manufacture and ease of operation it is now common practice to 

incorporate these fittings into atop plate of about 1m(3ft.) diameter. If the vessels are 

standing in the open air this top plate must be protected against the weather. 

Pressure relief is usually by asimple weight-operated valve (Fig. 14.3), although 

spring controlled systems are sometimes used. Excess pressure can develop from vessel 

filling, and expansion of fermentation gases when hot CIP is used. A reduction in 

pressure can occur on emptying the vessel, from the reaction of carbon dioxide with 

causticdetergents andwhencold liquid entersthetank after hot cleaning.Hot cleaningof 

fermenting vessels is always dangerous to the integrity of the tank and, if possible, it is 

recommended that fermenters are cleaned at temperatures of

adual planar system where the nozzles rotate on the head whilst the whole head rotates 

on the supply pipe (see Fig. 14.6 on page 525). 

Insulation 

Fermenters require insulation. Outside tanks are insulated against ambient conditions, 

which obviously vary considerably throughout the world. Indoor tanks will also require 

some insulation to lessen the demands on the temperature control system. Insulation is 

best added to the tank during manufacture for vessels up to asize of around 2000hl 

(1200imp. brl). 

Insulation materials usually contain chloride ions and so achloride inhibiting layer 

must be applied to the tank to protect the stainless steel prior to the application of the 

insulation. There is a wide range of insulation materials available and individual 

companiesoftenhaveverydistinctpreferences.Thebasicchoiceisbetweenpolyurethane 

foam and phenolic foam. Phenolic foam is less flammable although polyurethane can 

incorporate fire retardants, but these contain chloride. Inorganic materials such as glass 

fibre or mineral wool are completely non-flammable but are less good insulators. The 

thickness of the insulation used also varies widely. The capital cost of additional 

insulation is small in relation to the total cost of the vessel and so it is foolish to 

compromise here. Polyurethane foams are often applied in thicknesses of 100 to 150mm 

(4 to 6in.), whereas phenolic foams have been claimed to be effective at 75mm (3 in.) 

(Anderson, et al., 2000). Some brewers will specify amaximum permitted temperature 

rise in, say, summer conditions of

fermentations. For higher-gravity fermentations (say >12ëP, 1048ëSacch) it is now 

common to use oxygen instead of air. Concentrations of up to 30mg/l are possible but 12 

to 20mg/l are normal. Over oxygenation is sometimes thought to be impossible because 

oxygen is so rapidly utilized by the yeast. However, with some strains of yeast excess 

oxygen in the wort can lead to overgrowth and lower ethanol yields. It is therefore good 

practice to match the oxygen requirement of each particular yeast strain (Chapter 13). 

Oxygen is normally added after wort cooling but before the yeast. Anumber of types 

of injection system are in use. Sintered ceramic candles can cause the injection of avery 

fine stream of bubbles to achieve intimate mixing but these systems are difficult to clean. 

Aventuripipeisveryeffective.Airoroxygenisintroducedjustbeforetheconstrictionin 

the pipe and then mixes thoroughly with the wort in the turbulent flow resulting from the 

subsequent increase in diameter of the pipe (see also Chapter 10). 

Yeast for pitching has normally been derived from previous fermentations and 

maintained in yeast storage vessels as a slurry at < 5 ëC (41 ëF). The essence of successful 

pitching is the measurement of the quantity of the yeast to be pitched into the wort 

(Chapter 13). This can be achieved by volume, mass, or weight but these methods rely on 

the slurry being of constant composition and do not compensate for yeast viability. 

Methods are now available for making these compensations. 

· Volumetric methods. These methods are simple, cheap and can be successfully 

carried out by relatively unskilled personnel. A calibrated vessel can be used to 

measure the volume or more effectively a volumetric flow meter such as a magnetic 

flow meter. Peristaltic pumps have also been used. Account must be taken of the 

viability and concentration of the yeast to achieve good results. The method is affected 

by carbon dioxide gas trapped in the yeast slurry. This cannot be assessed and hence 

markedly different carbon dioxide concentrations in different slurries will affect the 

amount of yeast pitched into the wort. 

· Mass methods. The problem of trapped carbon dioxide can be avoided by the use of a 

mass meter, e.g., of the Coriolis type. Again account must be taken of the viability and 

concentration of yeast in the slurry. Methods of this type are gaining use in breweries 

often in conjunction with instrumental systems to measure yeast concentration. 

· Weight methods. This is a common and effective system. If used with yeast slurries 

the weight method requires the yeast storage vessels to be installed on load cells so the 

vessel contents can be weighed. The method is also commonly used with pressed yeast 

cake, which is subsequently slurried in chilled water prior to addition. The main 

problem with these methods is the reliability of the load cells that do not work well in 

the conditions of the yeast storage area (wet and cold!). As with other methods account 

must be taken of viability and concentration of the yeast. 

As simply applied these methods suffer from the inherent variability of yeast slurries 

in terms of viability and concentration. The number of cells in a yeast slurry can be 

estimated in a laboratory using a Coulter counter. This figure can then be used to estimate 

the volume of slurry needed to give a particular yeast count in the wort. The Coulter 

counter will not distinguish between live and dead cells. To do this many brewers will 

apply a compensation factor to the amount of slurry to be pitched. Viability will be 

assessed by the methylene blue staining method. In this method dead cells stain blue and 

viable cells remain colourless (Chapter 13). The method has, however, been considered 

unreliable for viabilities of < 85% (Institute of Brewing, Analysis Committee, 1971). 

There have been recent attempts to improve methods of determining concentration and 

viability of slurries by using new technologies. There has also been the attempt to link 


these methods into automatic systems of yeast pitching in breweries to avoid laboratory 

involvement and devolve the control of the process to brewery personnel. The objective 

here has been to improve the consistency of brewery fermentations and achieve more 

predictable attenuation and flavour volatile production. Askid-mounted instrument has 

been described (Teass, 2000) that will provide instantaneous cell counts in the 0to 2 

billion cells/ml range. Results obtained with the instrument correlated well with 

laboratory measurements. The volume to be pitched is controlled by aflow meter but 

corrections for viability still have to be made. Other instruments purport to measure 

viablecellconcentration.Abiomasssensormeasures thedielectricalpermittivityofyeast 

cell suspensions. This system utilizes aradio-frequency signal to create an electrical field 

through which the yeast cell suspension can flow. The yeast slurry acts as adielectric in 

that the electric field gives rise to no net flow of electric charge but only to a 

displacement of that charge. Areading of the displacement angle can be correlated with 

the concentration of intact cells. The assumption is made that intact cells are viable cells. 

On-line and off-line instruments have been developed using this principle (Carvell et al., 

1998). The sensor can be calibrated to different yeast strains used for different brewery 

fermentations. 

The actual pitching rate used varies considerably between breweries and rates of 5to 

20millioncells/ml ofwort are common depending onthe specificgravity ofthe wort. An 

optimum level is considered to be 10 to 12 million cells/ml and this should result in a 

reproduction rate for lager yeast of 3to 5times (Stewart and Russell, 1998). 

Temperature control 

The heat output during fermentation has been discussed (Section 14.2.3). It follows that 

for reproducible lager fermentation in abrewery this excess heat must be removed by the 

cooling system so that the temperature of the fermenting wort can be controlled to a 

chosen profile. The beer must also be cooled at the end of fermentation to achieve the 

maturation temperature (Chapter 15) and to help the sedimentation of the yeast. 

All three methods of known heat transfer can affect brewery fermentations. 

Conduction occurs when heat is transferred through the vessel wall, e.g., from the 

insulation. Radiation can occur from, say, direct sunlight on the vessel surface. But the 

most important factor to deal with is convection arising from movements within the 

fermenting liquid itself. Convection can be caused by density gradients arising from 

unequal temperature distribution in the vessel. These convection currents are enhanced by 

the movement of the liquid caused by the natural evolution of carbon dioxide bubbles 

during the fermentation of the wort sugars. A complex system thus operates in the 

fermenter which can, however, be subjected to basic physical analysis. The rate of heat 

transfer can be expressed in an equation derived from Fourier's Law: 

Q ˆ U.A. t 

Where Q is the rate of heat flow or conductivity, W, measured in calories/second, A is the 

area of heat transfer in m 2 , U is the heat transfer coefficient in W/m 2 /ëC and t is the 

temperature difference in ëC. 

For a fermenting vessel the term A is the area of the wall of the vessel subject to the 

temperature difference (Andrews, 1997). This could be the whole of the vessel area if the 

effects of insulation are being considered or the area exposed to the cooling jackets if 

considering the effect of coolant. U is difficult to calculate and is dependent on a 

combination of factors relating to fluid viscosity and density and conductivity and fluid 


Density (g/cm 3 ) 

1.000 

0.9999 

0.9998 

–2 0 2 4 6 8 

Ice starts 

forming 

Point of maximum 

density of beer 

Direction of cooling 

convection currents 


Point of maximum density 

of water, 3.98°C 

Fig. 14.4 Change in density of fermenting beer with change in temperature (Barnes, 2001). 

velocity at the vessel wall. Extraneous materials on the vessel wall, e.g., fouling by yeast 

deposits will also affect this fluid velocity and conductivity. 

If a temperature difference is now established between the fermenting beer and the 

vessel wall by means of a coolant, the beer adjacent to the vessel wall (sometimes called 

a beer film) will move downwards as the wall temperature will be lower than the beer 

temperature. A temperature profile will thus be set up between layers in the beer and the 

coolant. This will result in the establishment of density gradients further causing 

convection currents. 

An added factor to be considered in the temperature distribution in fermenters is the 

inversion temperature (Fig. 14.4). Water is most dense at 4 ëC (39.2 ëF). Water warmer 

than 4 ëC will rise and so on cooling a large tank, cold water (or fermenting beer) will 

initially flow down the tank wall until it reaches 4 ëC (the inversion temperature) when 

the flow pattern will reverse and the fluid will rise. The temperature of maximum density 

of beer is affected by alcohol content and extract. The temperature of maximum density 

of a 12 ëP (1048) beer is about 2.5 ëC (36.5 ëF) whereas a 16 ëP (1064) beer would have a 

maximum density at 1 ëC (33.8 ëF); lower gravity beers will obviously be closer to 4 ëC at 

maximum density. 

It is a real practical challenge to achieve a uniform temperature distribution in a 

cylindroconical fermenting vessel by using cooling jackets on the vessel and cone walls. 

We are dealing with the efficiencies of heat transfer from the beer film adjacent to the 

vessel wall into the bulk of the beer and heat transfer to the coolant from the beer film. 

There will also be the effect of the thickness of the vessel wall itself although as this is 

usually about 5 mm (approx 0.2 in.) the thermal conductivity is very high (4,000 W/m 2 / 

ëC). The driving force in temperature distribution is the convection currents. Having a 

large temperature difference between the coolant and the vessel wall will increase the rate 

of heat transfer and potentially improve temperature uniformity in the vessel. However, 

this is limited as freezing of the beer on the vessel wall must be avoided. Sophisticated 

systems have employed variable temperature of coolant using a lower temperature during 

fermentation than at the end and so avoiding freezing, however, these systems are too 


complex to operate in most breweries and the temperature of coolant chosen is usually a 

compromise choice at about 5 ëC (23 ëF) to provide optimum cooling without freezing. 

Good thermal conductivity in the beer film adjacent to the vessel wall is to be 

encouraged. The strength of convection currents will theoretically be increased by a large 

temperature difference between the bulk of the beer and the film and the release of carbon 

dioxide creating turbulence at the peak of fermentation. But at the end of fermentation, 

movement in the vessel is again solely as a result of density differences which, at the 

temperature of maximum density will be at a minimum. A typical U value during 

fermentation for the beer side film transfer to the beer bulk would be 400 W/m 2 / ëC but 

this will be less when carbon dioxide production declines. 

Heat transfer to the coolant is easier to assess than heat transfer from the beer film to 

the bulk of the beer. A high velocity of coolant (1 m/s) to create turbulence is the most 

important factor in efficient heat transfer. A low viscosity is also favourable which is 

often the reason to prefer ethanol in the form of industrial methylated spirit (IMS) to 

propylene glycol: 

Viscosity of IMS, 7.5 cP at 3 to 5 C (27 to 23 F† 

Viscosity of propylene glycol, 18 cP at 3 to 5 C (27 to 23 F) 

A heat transfer coefficient of 2340 W/m 2 /ëC is achievable with IMS and about 1900 with 

propylene glycol. Ammonia is very effective as a primary refrigerant used as a 

compressed gas and of course has zero viscosity as far as these comparisons are 

concerned and can create high turbulence during evaporation resulting in a heat transfer 

coefficient of 4,000. The overall heat transfer coefficient therefore must be calculated 

from the sum of the individual coefficients: 

1=U ˆ 1=Ubeer side ‡ 1=Ucoolant side ‡ 1=Uwall resistance ‡ 1=Ufouling 

This equation includes a contribution of the effect of fouling on heat transfer on both the 

beer side and coolant side of the system. In practice, with secondary refrigerants such as 

IMS there is very little fouling on the coolant side. This is more of a factor in direct 

expansion systems with a gas like ammonia, which needs a lubricating oil on internal 

surfaces of the cooling jackets. 

There have been few reports on investigations of temperature distribution in large 

tanks in the last 25 years (Andrews, 1997) and many of the design proposals of brewery 

engineers relate to the theory discussed above and experimental investigations of the 

1970s (Maule, 1976, 1986). From these investigations and unpublished results (Barnes, 

2001) it is clear that the natural convection currents are inadequate to provide totally 

uniform cooling irrespective of the position of the cooling jackets on the vessel or the 

nature of the coolant used. Beer in the upper volume of a tank may hardly change its 

temperature throughout a cooling regime. Temperature probes in the lower zones of the 

tank may indicate that temperature control is being achieved but this is often not the case 

for all the beer in the tank. This is clearly unsatisfactory and can lead to inconsistencies in 

flavour volatile production and attenuation. 

This has prompted experiments in agitation of the contents of the vessel. Small amounts 

of a gas such as air or carbon dioxide can be introduced to the base of the vessel for this 

purpose. When carbon dioxide was injected through a sintered stainless steel candle at 100g/ 

min. for four minutes (Lemer et al., 1991) improvements in the uniformity of temperature 

and subsequent cooling were achieved. Mechanical rousers are in use in several breweries 

but they are difficult to maintain and present a potential source of microbial infection. A 


novel proposal (Andrews, 1997) is to construct split cooling jackets so that only half of the 

circumference is providing cooling as the temperature is approaching that of maximum 

density, i.e., that at which flow inversion occurs. This will cause an increase in heat transfer 

coefficient, as fluid movement would be maintained. Temperature gradients would then be 

generatedhorizontallyandnotverticallythusavoidinglayeringinthetank.Agitationduring 

primary fermentation is disliked by some brewers because of the consequent increase in 

fermentation rate and increased organic acid production and lower beer pH value. It is, 

however, important to pay attention to the uniformity of temperature distribution in large 

cylindroconical vessels for regularity of beer flavour development. 

In asecondary coolant system the coolant is circulated by pumps through the cooling 

jacketsonthe vessels.The coolant itself iswarmedandisreturnedtoarefrigerationplant 

for further cooling. This system is thus ìndirect'. In the refrigeration plant the secondary 

coolant is cooled by aprimary coolant, which evaporates and so takes heat from its 

environment. Chlorofluorocarbons (CFCs) have frequently been used for this purpose in 

breweries. The compounds are given an `R' number, e.g., R22, where the number relates 

tothenumberofcarbon,hydrogenandchlorineatoms,butthesesystemsofnomenclature 

are sometimes confused. These compounds have ozone-depleting effects and are 

gradually being replaced by ammonia according to internationally agreed protocols. 

However, ammonia is avery corrosive and toxic gas causing acute irritation of the 

respiratory pathway. It is also explosive when mixed with air at high temperatures. It 

follows that rigorous safety procedures must be in place in breweries where this gas is 

used in the refrigeration system. 

Ammonia gas can be used directly as the primary refrigerant in adirect expansion 

cooling system (Fig. 14.5). Liquid ammonia is pumped through the cooling jackets at a 

defined pressure and about 10% of the mass evaporates causing cooling. The mixture of 

gas and liquid is returned to a receiver. The gas is then re-compressed and condensed to a 

liquid and can be returned through the cooling jackets. Direct expansion systems 

consume between 35 and 45% less energy than indirect systems but the dangers of using 

ammonia remain and in many breweries indirect systems are in use with the trend to 

replace the primary refrigerant CFCs in these systems with ammonia. 

It is evident that control of temperature in the whole of a cylindroconical vessel is 

difficult. The position of the temperature probes is critical so that correct control of 

Ammonia vapour 

to compressors 

Pressure control signal 

to compressors 

Liquid ammonia 

from condenser 

Oil return to 

compressors 

LC 

PC 

Ammonia 

receiver 

Liquid 

ammonia 

For clarity, vessel is shown with a single jacket 

TC 

Ammonia liquid and vapour 

Fig. 14.5 Direct expansion cooling of a fermenting vessel; PC, pressure control; LC, level control; 

TC, temperature control (Anderson et al., 2000). 


coolant flow is achieved. The probe should be installed at aposition between 30 and 50% 

of the height from the base of the vessel and about 500mm (20in.) into the vessel from 

the wall to avoid influence from the temperature of the wall itself. In some systems a 

series of probes are installed and the overall temperature computed from these inputs. 

However, attention to ensuring the uniformity of temperature by understanding the 

gradients occurring in the vessel as aresult of convection cannot be too strongly stressed. 

Cleaning of vessels 

Cylindroconical vessels are now almost always cleaned by in-place cleaning systems 

(CIP).TheprinciplesofCIParediscussedinChapter17.CIPisexpensiveandpotentially 

a major contributor to brewery variable cost. The vessels have a high organic soil level 

(yeast deposits) and therefore respond to cleaning with hot caustic soda (1±2%). 

However, to save costs the bulk of the soil should be removed with jets of water in an 

impact system. Various designs of impact jet are available but the most effective for 

fermenters are the so-called multi-planar heads in which the nozzles on the head rotate in 

one plane whilst the head itself rotates at right-angles on the support pipe (Fig. 14.6). 

Liquid should not collect in the base of the vessel or effective cleaning of this area is lost. 

This liquid is usually removed by scavenge pumps which must be correctly sized and 

have a capacity greater than the supply system and be capable of pumping with an empty 

supply system (self-priming). CIP cycles are now very carefully controlled by computer 

to optimize the use of materials and many variations can be set. A basic sequence would 

be based on: 

· Pre-rinse with cold recovered water (say 15 ëC, 59 ëF) with the objective of removing 

as much of the soil as possible. The soiled water would pass to the effluent system. 

Time 10±12 minutes. 

· Rinsing with hot (say 60 ëC, 110 ëF) caustic detergent (1±2% NaOH). Detergent will be 

re-circulated to a tank and will be replaced only when depleted (carbonate formed). 

Time 20±25 minutes. 

Sprayhead rotates 

in horizontal plane 

Sprayhead suspended 

by CIP supply pipe 

Nozzles rotate in 

vertical plane 

Fig. 14.6 A multi-planar CIP cleaning head (Anderson et al., 2000). 


· Post-rinsewithcleanwatertoremovedetergentresidues.Asterilantcouldbeaddedat 

this stage (Chapter 17). Water from this stage is recovered as pre-rinse. Time 10±12 

minutes. 

The importance of the pre-rinse stage in minimizing detergent use cannot be over 

emphasized. Hot caustic soda solutions must not be applied to tanks containing carbon 

dioxide or rapid dissolution of the gas and implosion of the tank will follow. Vacuum 

relief valves (section 14.3.2) will provide some protection against this but must not be 

relied on as acontrol measure. 

Acids can be used to clean fermenters. Over time beer stone (calcium carbonate/ 

oxalate) will collect on vessel surfaces and on bends in pipe-work and sometimes in 

pumps. In this situation cleaning with 0.5±1.0% nitric acid or sometimes phosphoric acid 

is effective. Caustic detergent with asequesterant such as EDTA (ethylene diamine tetra 

acetate), which strips the calcium from the stainless steel and keeps it in solution is 

effective but is more expensive than acid. Acid cleaning systems possibly lower the 

overall time of cleaning compared to systems using caustic soda (Barnes, 2001). 

In fermentation vessel control systems CIP is often interlocked with aliquid level 

detector in the base of the tank. It is ruinous if aCIP sequence is started when beer is still 

in the tank. Double seat valves are therefore increasingly used in automated systems to 

provide security of operation and protection against leakage. 

14.4 Top fermentation systems 

Lager beers produced by bottom fermenting yeasts are by far the most widespread beer 

types throughout the world; consequently the bulk of development work on fermentation 

technologyhasbeenoncylindroconicalvesselsforbottomfermentingyeasts.Howeverin 

the UK and Ireland, ale and stout are the traditional beers and these are normally 

produced with strains of the top fermenting yeast Saccharomyces cerevisiae as are some 

Belgian and German beers. Fermentation systems developed to allow for the property of 

separating the yeast from the top of the fermenter at the end of fermentation. Traditional 

ale is cask conditioned (Chapters 21 and 23). Asecondary fermentation takes place inthe 

cask to provide condition to the beer. This ale is almost entirely produced by top 

fermentation. Ale can also be sold in kegs as a chilled and filtered product (Chapter 21). 

Some of this ale is produced by bottom fermentation in processes that are now difficult to 

distinguish from those of lager fermentation. Traditional ale brewers would regard 

producing ale by bottom fermentation with suspicion, but its proponents would cite the 

ease of separation of the yeast as the overriding issue. 

Removal of the yeast at the end of fermentation is traditionally by manual skimming 

or suction of the froth or `head' directly from the bulk of the fermented wort in the vessel. 

Developments of traditional systems have resulted in more ingenious methods for yeast 

separation. Examples of these methods are the Yorkshire Square system and the Burton 

Union system. 

14.4.1 Traditional top fermentation 

Vessels and rooms 

Traditional top fermentation utilizes a single vessel that is open to the atmosphere to 

facilitate yeast removal. The vessels were traditionally shallow (2 to 4 m, approx 6.5 to 


Fermenting 

beer 

Walkway 

Re-circulated air 

Louvres 

Air-cooling 

heat exchanger 

13ft. deep) and could be round, square or rectangular in shape (Fig. 14.7). Vessels have 

been made of many materials including wood, stone, slate, aluminium, cast iron, mild 

steel, copper, reinforced concrete and stainless steel. Vessels made of wood, cast iron, 

mild steel or concrete were usually lined with afurther material to assist cleaning. 

Linings were made of vitrified enamel, pitch, and various plastics (with or without the 

incorporation of fibreglass) and epoxy resins. Nearly all these linings had adverse 

features, mostly the possibility of tainting the beer. Some were fragile and needed 

frequent repair. Almost all top fermenting vessels built since the 1960s have been made 

of stainless steel, usually of type 304 (Section 14.3.2). 

Top fermenting vessels have traditionally been small (80 to 1000 hl; 50 to 550 imp. 

brl). Vessels have normally been grouped together in fermenting rooms. To lower the risk 

of microbial infection the surfaces in the fermenting rooms must be smooth and also, 

most importantly, accessible to easy cleaning. Walls are normally tiled or finished with 

polypropylene sheeting and floors are tiled or covered with asphalt or terrazzo. There 

must be a sufficient fall on the floor to allow for drainage and the drains must be 

constructed with traps to avoid odours. Condensation on ceilings is often a problem as 

condensate can fall into the fermenting beer. To avoid this the whole fermenting room is 

often air-conditioned. 

Normally a fermenting room has a false floor between the vessels usually about 600 to 

900 mm (23 to 36 in.) below the tops of the vessels and about 2.5 m (8 ft.) above the true 

floor. The space between the false floor and the true floor is often called a `shell' room. 

This space is utilized for circulating attemperated air and for mains and pipes. Air from 

above, heated by fermentation and containing carbon dioxide, can be aspirated into the 

shell room space, mixed with fresh air and cooled through a heat exchanger and then reintroduced 

above the false floor. Ideally the room should be kept at a temperature 

between 15 and 18 ëC (59 to 64 ëF). 

Fan 

Arrows – forced 

air circulation 

Fresh air 

Fig. 14.7 An open square fermenting vessel (after Hough et al., 1982). 


Carbondioxideevolvedfrom open fermentersisamajorhealthhazard.Themaximum 

concentration for exposure during an eight-hour shift is set at 0.5% (sometimes called the 

threshold limit value, TLV) in many countries including the UK and Germany. At 1to 

2% carbon dioxide, blood composition changes and oxygen access to the brain becomes 

restricted. At concentrations above 2% respiration rate increases in an attempt to 

compensate for the shortage of oxygen and dizziness is likely to be felt. Unconsciousness 

and death will follow at concentrations above 8%. Many accidents arising from exposure 

to carbon dioxide have occurred in brewing and accidents continue to occur. It is, 

therefore, extremely important to take great care when working with open fermenters. 

Carbon dioxide, being heavier than air, will collect in the walkways between vessels as it 

spills over from the fermenter. Asystem of positive air displacement must be used to 

ensure that this air is removed and if re-circulated (see above) must be enriched with 

fresh air to lower the carbon dioxide concentration to

topfermentedalesthanforlagers.Moremodern vessels,facilitatingeasiercleaning,have 

side wall cooling. 

The yeast is removed from the vessel by skimming; this is usually by suction but was 

once done manually with paddles. This removal of yeast may take place once, twice or 

three times depending on the conditions and the type of yeast in use. The first yeast crop 

is frequently dirty and will contain trub, particularly if adropping system is not in use. 

This crop will also be the least attenuative of that collected and it is often discarded. The 

following crops will be more attenuative and cleaner and these crops are usually retained 

for re-pitching although this relies on the experience and skill of the brewer. Yeast 

remaining on the base of the vessel (grounds) is normally discarded as this does not 

display the characteristics required for successful top fermentation, i.e., ayeast that 

attenuates rapidly and rises from the beer easily into the head at the completion of 

primary fermentation. 

Yeast removed from the fermenterby suction is held in avesselcalled ayeast back. In 

traditional systems, the yeast slurry would be processed through afilter press (Chapter 

15), to separate yeast from the entrained beer, `barm' ale, which would then be added 

back to the beer at racking usually after pasteurization. The yeast would be discharged 

from the press and held in trays or trucks in arefrigerator before re-use. The yeast can 

also be held without pressing, as aslurry at

Fish 

tail 

Pump for 

rousing 

For yeast 

removal 

Manhole 

Organ pipe 

Fig. 14.8 Yorkshire square fermenting vessel (Hough et al., 1982). 

flocculent. Beers of characteristic flavour are produced, which often at low gravities 

(

Enclosed vessels have been built (Griffin, 1996; Ogie, 1997) to allow full CIP, more 

effective microbiological control and the potential to collect carbon dioxide gas. Of major 

importance has been the novel method of yeast collection reducing the manual operations 

of the vessels. Vacuum yeast collection tanks are connected to a pair of yeast skimming 

points set in the top deck of the vessel. Suction operates through these points for a set 

period until most of the yeast has been removed. Spray lances are then used to deliver 

bursts of high pressure water across the deck to direct the remaining yeast to the 

skimming outlets. Each burst lasts about 10 seconds and is set to ensure the yeast 

concentration is sufficient for re-pitching and that water consumption is kept as low as 

possible. The base of the fermenter is fitted with pop-up spray jets (pintle valves), which 

operate to remove the grounds once the beer has been run to maturation or racking. All 

operations are remotely controlled by computer and the vessels are essentially automatic. 

Thus the traditional features of Yorkshire square fermentation are maintained with the 

minimum of manual involvement. 

14.4.3 Burton Union fermentation 

This system of fermentation is associated with the Burton-on-Trent area in England and 

was devised for the production of pale ales with powdery, i.e., non-flocculent yeasts. 

Wort is collected and pitched in a collecting vessel and then transferred at the peak of 

fermentation (36 to 40 hours) to the set of Union vessels. These vessels are oak casks of 

capacity 153 imp. gallons (7 hl) arranged in two adjacent rows of 12 vessels beneath an 

inclined, cooled `top' trough. The individual casks contain cooling coils. At the top of 

each cask is a swan-neck pipe, which can discharge into the trough (Fig. 14.9). Carbon 

dioxide gas carries yeast up through the swan-neck and it falls into the top trough. Yeast 

tends to sediment in the trough and beer collects at the lower end and is returned by side 

tubes to the casks. As fermentation completes virtually all the yeast has passed to the top 

trough. The beer in the casks is discharged into a second trough running below the row of 

casks and thence to racking backs. Beer transferred to racking is very clear and the yeast 

count is adjusted by the position of the outlet tap on each cask. 

a1 a2 s 

b 

c 

f 

d 

e 

s 

h h 

Fig. 14.9 Burton Union fermentation system (a 1 ) attemperator water (beer); (a 2 ) attemperator 

water (yeast); (b) side rod; (c) waste water; (d) top trough; (e) bottom trough; (f) feeder; (s) swannecks; 

(h) bottom tap; (m) side tap; (n) sample tap (Hough et al., 1982). 


s 

m 

n

As with Yorkshire squares, beer of characteristic flavour is produced and some brewers 

maintain that true Burton pale ale can only be made in this way. In 1990 one company 

invested in a new version of the Burton Union system constructed in stainless steel 

comprising four sets of 30 150 gallon Unions. Burton Union systems require alot of 

space for relatively low throughput and have ahigh capital cost. Labour involvement is 

high because, essentially, all the cleaning is manual. However, good quality yeast for repitching 

is invariably produced along with beer that is easy to fine (Chapter 15) and the 

opportunity is presented to exploit the method of production for enhanced marketing of the 

brand. 

14.5 Continuous fermentation 

Once the biochemistry of fermentation began to be understood towards the end of the 

nineteenth century (Pasteur, 1860; Chapter 13), some brewing scientists began to think 

they had limited control over the fermentation process. Surprisingly this stimulated a line 

of investigation on the continuous culture of yeast at high concentrations in which an 

active metabolic state was maintained and the problems of batch separation of yeast from 

beer avoided (DelbruÈck, 1892). Continuous fermentation of beer was thus attempted 

before 1900 and by 1906 at least five separate systems had been described, some of 

which were patented (Van Rijn, 1906). This patent (Van Rijn, 1906) is of some interest 

because in it are described processes involving the cascading of partly fermenting beer 

from one open tank to another, an idea that was picked up much later on in the 1950s 

(Wellhoener, 1954). These early systems of continuous fermentation did not achieve 

commercial success for reasons that are not entirely clear. It seems probable that the 

inability to avoid infection by micro-organisms was one reason and perhaps another was 

the resistance to change by the established brewers of the day. 

As sales of beer increased in the 1950s and 1960s there was renewed interest in the 

potential of continuous fermentation particularly in New Zealand, Canada and the UK. At 

this time the brewing scientist was often represented at main board level in brewing 

companies and as such had considerable influence on company strategy. There was thus 

enhanced knowledge in companies of the brewing process, which could be coupled with 

developments in electronic process control equipment. As a result there was real hope for 

the commercial success of continuous systems with the advantages comprising: 

· lower capital cost 

· lower working capital because of less beer in process, as a result of faster throughput 

· lower product cost as a result of lower beer losses, more ethanol and less yeast 

· lower fixed costs because of less manpower as a result of less cleaning and automatic 

fermenter control. 

Companies attempted to exploit the technology and gave considerable support to their 

scientific proponents. By the early 1970s about 4% of UK beer production was by some 

form of continuous fermentation and the outstanding technology of Morton Coutts was 

established in New Zealand (Coutts, 1957). But many of the systems fell quickly out of 

use and by the late 1970s and early 1980s the only production scale continuous 

fermenters in use were and are in New Zealand. 

The decline of continuous fermentation reflected the rise in the use of the large batch 

cylindroconical fermenting vessel as brewers responded to the `threat' of continuous 

fermentation by finding improved batch methods. Developments in the design and 


operation of these vessels in the 1970s were rapid. At the same time the influence of 

marketing was becoming manifest in brewing. After apeak in brewing sales in Europe in 

1979, economic recession in the early 1980s led to asearch for more innovative ways of 

expanding beer sales. This resulted in new product development often using different yeast 

strains, and the desire to present the public with agreater choice of drinks. Inevitably this 

meant that some of these drinks were produced in low volumes and some failed to achieve 

market success. Many continuous fermentation systems did not have the flexibility to 

handle this type of production, which was much easier in batch processes. Acontinuous 

system was geared to produce ahigh volume of asingle type of product. It was also 

realized that to operate continuous systems successfully required highly trained personnel 

often working shifts and this led, paradoxically, to increased costs. The situation was 

perhaps different in the less developed market conditions of New Zealand where Morton 

Coutts was able to foster and develop his system to commercial success over many years. 

Into the 21st century the use of immobilized yeast systems has demanded a 

reassessment of the value of continuous fermentation and this has been allied with huge 

developments in process control using computers, which were not available in the 1970s 

and 1980s. The physiological and biochemical behaviour of yeast in continuous 

fermentation is discussed in Chapter 13. In this Chapter practical aspects are considered. 

14.5.1 Early systems of continuous fermentation 

There are anumber of different ways of classifying continuous yeast culture and hence 

fermentation systems (Herbert, 1961; Chapter 13). Yeast can emerge in an òpen' system 

with the beer or in a`closed' system when it is retained with the beer. In ahomogeneous 

system the yeast and fermenting beer is intimately mixed in astirred fermenter, whereas 

in heterogeneous systems there will be separation and concentration of the yeast away 

from the beer in avessel. The cascade open system of the series of inter-connecting 

vessels was described in 1954 (Wellhoener, 1954). Substrate (wort) entered the first 

vessel and product (beer) emerged from the last. Residence time was about 25 days and 

the system was never fully adopted on the commercial scale. But some open systems did 

achieve limited production success around 1970. 

Stirred tank fermenters 

Asystem (Bishop, 1970) operated in four UK breweries achieving an output of 32,000hl 

per week ( 20,000imp. brl). There were two stirred fermenters in series and a 

sedimentation vessel for collecting the yeast (Fig. 14.10). Yeast was not recycled and the 

residence time was around 15 hours. Other UK breweries operated similar systems. The 

problems for further enhancement were as discussed above and to achieve outputs 

comparable to those of the cylindroconical vessels of the day (with afive day turn round 

time); the stirred fermenters would have to be increased sixfold in size. This was not 

commercially feasible. 

Tower fermenters 

Tower fermenters were developed by the APV Company in the 1960s and were used in 

breweriesinthe1970s(Portno,1973).Thisheterogeneoussinglevesselsystem(Fig.14.11) 

utilized the ability of flocculent yeast strains to sediment and so maintain a high 

concentration of cells in the system. Wort was pumped into the base of the vertical tube. The 

sedimentary yeast formed a plug in the base of the vessel and the wort permeated through the 

plug. Some of the yeast was carried up the tower by the flow of wort and the fermentation 


Wort in 

Pump 

Oxygen 

column 

Sterilizer 

Stirrer 

drive 

Stirrer 

drive 

Fob 

breaker 

Cooling 

coil 

Beer 

outlet 

Yeast out 

CO 2 outlet 

Sedimentation 

vessel 

Fig. 14.10 Stirred tank continuous fermentation, two stirred fermenters in series and a yeast 

sedimentation vessel (Bishop, 1970). 

b 

e 

a 

h 

c 

d 

b 

Head 

8.2 m 

Top 

7.0 m 

26.3 brl 

Middle 

3.9 m 

14 brl 

Bottom 1.2 m 3.25 brl 

Fig. 14.11 Continuous tower fermenter; (a) wort collection; (b) impeller pump; (c) flow meter; (d) 

control valve; (e) flash pasteurizer; (f) tower; (g) yeast separator; (h) beer receiver; (j) CO2 

collecting vessel (after Hough et al., 1982). 


f 

g 

j

proceeded as the wort rose. The rate of wort injection was adjusted so that at the top of the 

tower the wort was fully fermented. The tower contained baffles to control the upward 

movement of the yeast, which can be too rapid as aresult of carbon dioxide evolution in 

large bubbles. Yeast fell from suspension at the top of the tower and an inclined chute near 

thebeeroutflowassistedthis.Thesystemwasnotadaptablefortheuseofpowderyyeastand 

therefore could not be used for producing beers associated with these yeasts. 

It was desired to produce the same crop of yeast as with conventional fermentation so 

that these could be mixed if required. This required adjustments in wort composition as 

yeast in high concentration grows slowly and does not assimilate the normal amount of 

nitrogenous material from the wort. Worts with high levels of fermentable carbohydrate 

andlowlevelsofassimilablenitrogenwere usedtoyieldbeersofnormalcomposition.This 

in turn leads to complications and increased costs in the management of the system. There 

were further difficulties with abnormal production of acids, esters, higher alcohols and 

diacetyl. Tower fermenters were also prone to infection with lactic acid bacteria requiring 

wort to be pasteurized before fermentation and so adding to cost. Residence times in the 

fermenter were as low as four hours for ale fermentations but like the stirred tank 

fermenters and for similar reasons, prolonged production performance was not achieved. 

14.5.2 The New Zealand system 

In view of the difficulties described above it is aconsiderable achievement that the 

system developed by Morton Coutts (Coutts, 1957, 1958) remains acommercial success. 

This is attributable to sound basic design, and unfailing commitment and skill. 

Significantly,theNewZealandmethod isanintegralpartofacontinuousbrewingsystem 

(Fig. 14.12) and it comprises: 

· ahold-up vessel 

· two stirred fermenters 

· ayeast separator. 

In the hold-up vessel the wort is aerated and mixed with yeast and beer recycled from the 

first fermenter. The residence time is about four hours and considerable yeast growth 

occurs. The volume of the hold-up tank is 230hl (140imp. brl) and the following 

fermenters are 1,637hl (1,000imp. brl) and 409hl (250imp. brl) respectively. 

Temperature is maintained at 15ëC (59ëF) and the throughput is 70hl per hour. Wort 

of 18ëP (72 ëSacch) is fed to the hold-up vessel and is diluted in the first fermenter to 

13ëP (52ëSacch). The specific gravity of the beer emerging is about 3ëP (12ëSacch). 

Total residence time is about 30 hours and continuous runs of up to one year have been 

achieved. The process hasundergone anumberof changes since its introduction (Stratton 

et al., 1994). These have focused on improved microbiological control and process 

automation, which have resulted in major efficiency gains. Several plants are now 

operating(in2003). This systemhas been asuccessbecausesoundbeerstruetothe brand 

type have been consistently brewed and the system has not been required to produce 

different beers on an intermittent basis. 

14.5.3 Continuous primary fermentation with immobilized yeast 

Technology 

Afundamental principle of continuous culture of yeast is the maintenance of the yeast in 

high concentration in its substrate (Chapter 13). All continuous brewing fermentation 


Air 

From 

lauter 

tun 

Filtered wort storage 

1°C (34°F) 

Hold-up 

vessel 

To finished 

beer 

collection 

Intermediate 

boiling 

Hops 

vessel Hop 

strainer 

Wort 

boiler 

Hot trub 

Yeast for propagation 

Stabilizing vessel 

0°C (32°F) 

CF 1 CF 2 

Fermented beer return for pH reduction Yeast in suspension 

control 

Chilling 

CO2 

Sucrose 

priming 

Diatomaceous 

earth filter 

Hop Caramelized 

extract sugar Finings 

Fermented wort 

O.G. dilution 

Surplus yeast and 

precipitated material 

Yeast 

separator 

Cooling 

and 

chilling 

Fig. 14.12 New Zealand system of continuous fermentation, (CF1) continuous fermenter 1, (CF2) 

continuous fermenter 2 (Coutts, 1957, 1958). 

systems must have ways of achieving this objective. However, yeast in high 

concentration grows slowly and the metabolic effects on the wort are different from 

those achieved by batch fermentation at normal yeast concentrations. This factor, 

together with the proneness of continuous systems to become infected with beer spoilage 

bacteria limited the commercial development of continuous fermentation. This situation 

has changed with the availability of immobilized cell technology that offers the 

possibility of real heterogeneous fermentation. 

Immobilizationinvolvesthephysicalconfinementorlocalizationofwhole intactyeast 

cells in afixed position whilst retaining normal viability and metabolic activity (Chapter 

13). The yeast cells are no longer free and therefore, in a fixed position, can be 

maintained in high concentration. The most common immobilization technique (Stewart 


and Russell, 1998) is to trap the yeast in amatrix. This is usually apolymer that forms a 

gel around the cells. Many polymers have been used: alginate, carrageenan, agar, pectin, 

polyacrylamide and silica gel, etc. After trapping the cells on the polymer, growth in a 

nutrient medium takes place to ensure that the matrix is fully saturated with yeast cells. 

Themixtureisthen`gelled'intosheetsandsubsequentlycutintoparticlesofdesiredsize. 

Spherical beads of 0.3±3.0 mm in diameter are normally used. The matrix is porous 

enough to allow the free diffusion of substrates and fermentation products. The gel must 

have mechanical strength so that it does not split in operation and allow yeast cells to 

leave the matrix. This is often as aresult of the evolution of carbon dioxide gas bubbles 

from the entrapped cells. This was amajor problem in early immobilized yeast systems. 

Somesystemshaveusedimmobilizationonpre-formedporousornon-poroussupports 

such as kieselguhr, perlite, wood chips, PVC chips, glass fibres, stainless steel and silica. 

These systems reduce mass transfer problems in that the cells are in closer contact with 

the substrates, but they are more prone to physical disruption particularly by carbon 

dioxide gas bubbles. Like any other continuous system, to be commercially viable an 

immobilized yeast system must offer repeatable advantages over batch systems. This 

means considerable work to optimize immobilization procedures, and mass transfer to 

ensure consistentlyhigh rates of fermentation. Recent reportssuggest these aims can now 

be achieved and that immobilized yeast bioreactors are capable of achieving rapid 

reproducible primary fermentation of wort over continuous periods of many months. 

Initially immobilized systems were primarily used for accelerated maturation of beer 

(Chapter 15) with rapid removal of diacetyl but now primary fermentation is achievable. 

Operation 

Early attempts at primary fermentation were not successful. This was partly as a result of 

prejudice arising from the perceived failure of free cell systems. But enthusiasts 

continued to experiment and received some encouragement from reports from the fuel 

alcohol industry (Margaritas and Merchant, 1984) which suggested reduced costs and 

operating consistency were achievable with immobilized yeast bioreactors. Mass transfer 

limits were the problem in early immobilized technology (Masschelein et al., 1985) and it 

proved difficult to produce beers directly flavour matched to beers from batch 

fermentations. This was partly owing to the use of worts with too low levels of free 

amino nitrogen (FAN) resulting in unbalanced concentrations of higher alcohols and 

esters. There was also low oxygen tension in early immobilized systems which was 

advantageous for ethanol production but not for the balance of beer flavour volatiles. 

In a Japanese two-stage process a free cell stirred fermenter is followed by an 

immobilized yeast bioreactor (Inoue, 1995). Yeast growth occurs in the stirred vessel, 

which allows appropriate utilization of FAN similar to that of conventional batch 

fermentation. Full attenuation was then achieved in the anaerobic bioreactor where the 

yeast cells were trapped on porous ceramic beads. Centrifugation of the beer between the 

two stages was often carried out. The system was combined with a maturation column 

(Chapter 15) to complete beer production in three to five days. From a 20-litre pilot plant, 

scale up to 100 hl has been achieved which has operated for two years. Whilst the system 

was resistant to bacterial infection it did not achieve the efficiencies hoped for and it did 

not offer significant improvement over batch fermentation. Capital costs were higher than 

anticipated and so were revenue costs as a result of higher beer losses in centrifugation 

and increased energy usage. 

A novel approach to solving the problem of efficient mass transfer has been developed 

in Canada (Mensour et al., 1997). This system utilizes a gas lift draft tube bioreactor (Fig. 


Stainless steel 

head plate 

Inoculated beads 

Sparging gas 

Plant wort 

‘Green’ beer 

Thermal jacket 

Bioreactor 

Draft tube 

Fig. 14.13 Gas lift draft tube bioreactor (Mensour et al., 1997). 

14.13) with carrageenan gel beads. Good mixing is achieved with minimum shear on the 

matrix; hence mass transfer is considerably improved. Amixture of air and carbon 

dioxide was used as the sparging gas. The proportion of air affected the flavour profile of 

the resulting beer and by taste panel evaluation air concentrations of between 2±5% 

yielded acceptable products though not precisely matched to conventionally fermented 

beer. Again the system can be combined with accelerated maturation to produce finished 

beer in two days. 

An alternative two-stage process has been developed in Belgium (Masschelein and 

Andries,1996).Theimmobilizedyeastbioreactorprecedesacylindroconicalstirredtank. 

The yeast is immobilized in rods of porous sintered silicon carbide of length 900mm and 

diameter 26mm. Fermenting beer is circulated through the internal channels of the 

carbiderods.Greenbeerisdrawnfromthetopofthereactortothestirredtankinwhichit 

can also be circulated from the tip of the cone to apoint on the cylinder just above the 

start of the cone. The fermentation in the stirred tank is completed by yeast cells, which 

have escaped from the immobilized bioreactor. Beers of acceptable flavour have been 

produced at rates of 135hl/hl of fermenter volume over acontinuous period of six 

months. 

Further developments have been reported from Finland (Andersen et al., 1999; 

Pajunen et al., 2000) where there has been considerable progress on achieving 

satisfactory maturation with immobilized yeast (Chapter 15). This work has concentrated 

on solving the problems associated with temperature control and the disruptive effect of 

carbon dioxide gas bubbles on the stability of the matrix. A single reactor is used 

operating at a high enough pressure to keep carbon dioxide gas in solution. The gas is 

removed in a separate de-gassing unit and the beer is circulated around the fermentation/ 

de-gassing loop to complete attenuation. The heat generated in fermentation is removed 

with a heat exchanger. The system has operated in pilot scale at 50 litres per hour to yield 

beer in 20 hours. 

Future 

Work on continuous fermentation with immobilized yeast bioreactors is characterized by 

the great innovation and enthusiasm of its devotees. Nevertheless, problems remain over 

the widespread adoption of the technique. Rapid fermentation tends to result in the 


production of higher levels of diacetyl than are normal in brewery fermentations. This 

diacetylcanberemovedusingimmobilizedyeasttechnology(Chapter15).Somebrewers 

view this as an extra, avoidable process and are deterred from using continuous rapid 

primary fermentation. This does not detract from using immobilized yeast for rapid 

diacetyl removal in maturation. The use of agenetically modified yeast could overcome 

this problem of increased diacetyl production but this is unlikely to achieve public 

acceptance for some time, if ever. The external addition of -acetolactate decarboxylase 

would also help but some brewers wish to avoid the use of added enzymes. 

There is also the problem of the diversity of brands produced by some brewers. If the 

production of different beers requires the use of different yeasts then the inflexibility of 

continuous fermentation is manifest. Successful continuous fermentation requires the 

steady production of asingle brand with great consistency (Section 14.5.2). To achieve the 

undoubted potential benefits of continuous fermentation would require different brands to 

be produced from varying wort composition or from treatments post-fermentation. 

Optimistic conclusions for the future were drawn in an EBC symposium on 

immobilized yeast reported in 1997 (Linko et al., 1997), nevertheless in the same year 

99.9% of the World's beer was produced in batch fermentation systems (Masschelein, 

1997). 

14.6 Fermentation control systems 

The control of temperature in fermentation has received much attention (Section 14.3.3). 

These control systems are reactive to the exothermic nature of fermentation and are now 

highly developed. However, until recently little attention has been paid to the control of 

brewery fermentation in relation to the chemical changes which are taking place. These 

changesinclude thefallinspecificgravityandinpHvalueofthewortandtheproduction 

of ethanol and carbon dioxide. 

In the general literature of biotechnology the control of industrial fermentations to 

yield many products is considered in depth. Brewers have not always made best use of 

this information, but any control system, no matter how innovative, must justify itself in 

terms of its cost effectiveness for the production of potable beer (Moll, 1983; Dauod, 

1987). 

14.6.1 Specific gravity changes 

The most widely investigated methods of fermentation control have been based on the 

fall in specific gravity as wort is fermented to yield beer. In its most basic form this 

method involves taking arepresentative sample of wort and measuring its attenuation 

limit using, e.g., EBC Analytica IV method. This can be done throughout fermentation to 

construct atime-course picture of attenuation during the fermentation process (Section 

14.2.2). The system can be automated such that the density of the wort is determined by 

direct methods (Fig. 14.14); this is, of course, asystem of automated measurement rather 

than control. Wort density could also be measured by refractometry or by the use of the 

oscillating U-tube. These methods have been developed into control systems when the 

fall in gravity has been used to cause an external change such as the onset of fermenter 

cooling. This has the potential of earlier use of cooling and hence savings in process time. 

In large cylindroconical fermenters pressure difference measurements have been used 

to monitor the density change. As sugars are fermented to ethanol and to carbon dioxide 


CIP 

supply 

FV 

CIP 

supply 

Sample 

pump 

Specific gravity 

signal to control 

system 

Mass 

flowmeter 

so the density drops and the pressure in the fermenter will change in proportion. Pressure 

transducers are used to measure this change and by linking to acomputer the extent of 

fermentation can be followed. This technique has also been applied in new Yorkshire 

square fermenters (Griffin, 1996). 

14.6.2 Other methods 

It was demonstrated in the 1970s (Alford, 1976) that the partial pressure of carbon 

dioxide in exhaust fermenter gas was an excellent approximation to the partial pressure 

exertedbydissolvedcarbondioxideintheliquidinthefermenter.Thisprovidedthebasis 

for apotential control system to assess the extent of fermentation by measurement of 

carbon dioxide partial pressure in the exhaust gas. Acontrol method relating to the 

production of carbon dioxide has been described (Dauod et al., 1989), but has not 

achieved widespread adoption. 

The use of dielectrical permittivity for the control of yeast pitching rate has already 

been discussed (Section 14.3.3 and Chapter 13). The commercially available probe can 

also be used to detect viable yeast cells in streams of yeast emerging from the cone of 

MF 

ø25mm line size 

Fig. 14.14 Specific gravity monitoring system using a sample loop; the product sampled is 

fermenting wort in the specific gravity range 1000±1100 ëSacch, a signal is sent from the mass 

flowmeter to the control system, sample pump must operate in both directions to allow for CIP; 

(FV), fermenting vessel (Barnes, 2001). 


cylindroconical vessels during cropping (Boulton and Clutterbuck, 1993). This biomass 

probe can be used in-line to automatically control yeast removal from the fermenter at the 

end of fermentation. The proportion of the yeast crop most suitable for re-pitching can 

therefore be retained. This technique could also be of direct financial benefit in allowing 

the reduction of yeast storage capacity. 

Methods of fermentation control involving the measurement of pH value are also 

available (Barnes, 2001). A sample is withdrawn from the fermenter the pH is measured 

and, if it is too high, oxygen can be introduced to the fermenter automatically. This can 

be further linked into modulation of the coolant control valves to direct the temperature in 

the fermenting vessel. Possibly, future developments in fermentation control will focus 

on assimilating a number of measurements: density, carbon dioxide production, ethanol 

production, pH value into a computer and causing outputs to achieve automatically the 

desired time course of the fermentation. This will mainly result in improvements in the 

control of cooling by better modulation of coolant valves, with consequent savings in 

energy. The capital cost of such systems will, however, be high and anticipated paybacks 

must be carefully assessed before proceeding with installation. 

14.7 Summary 

The major developments in fermentation technology in the latter part of the 20th century 

and into the present day have focused on the cylindroconical fermenting vessel, which 

has been operated in batch sizes of up to 6,000 hl (3,600 imp. brl). The development of 

the vessel has reflected the increased dominance in the brewing world of beers produced 

by bottom fermenting yeasts. Recent developments of the cylindroconical vessel have 

included improvements in yeast pitching control to maximize the pitching of viable yeast 

and improved temperature control and cleaning. Overall fermentation control systems, 

where changes in the state of the fermenter can be effected automatically arising from the 

measurement of primary fermentation parameters such as density, are not in wide use. 

The difficulty of achieving cost effectiveness in this area provides encouragement for the 

proponents of continuous fermentation technologies. 

Continuous primary fermentation with free cell systems is difficult to exploit 

consistently and commercially. Large volumes of an established brand in a consistent 

market, and skill and enthusiasm of operators are needed for success. This has been 

achieved only in New Zealand. Immobilized yeast technology would seem to be the most 

likely continuous system for widespread applicability. Extremely rapid fermentation is 

achievable (20 hours) and this can be allied to rapid maturation also with immobilized 

yeast. However, a problem remains, the perceived inflexibility of continuous systems and 

inability to deal with a range of beer brands requiring different yeasts. Nevertheless 

developments in the use of immobilized yeast for continuous primary fermentation will 

occur, as there is real potential for major reduction in brewery variable cost when this 

technique succeeds. 


ALFORD, J. S. (1976) Can. J. Microbiol. 22 (1), 52. 

ANDERSEN, K., BERGIN, J., RANTA, B. and VILJAVA, T. (1999) Proc. 27th Congr. Eur. Brew. Conv., Cannes, 

771. 


ANDERSON, R. G., BRITES SANCHES, A., DEVREUX, A., DUE, J., HAMMOND, J., MARTIN, P. A., OLIVER-DAMEN, 

B. and SMITH, I. B. (2000) Fermentation and Maturation Manual of Good Practice, European 

Brewery Convention, Zoeterwoude, The Netherlands. 

ANDREWS, J. M. H. (1997) Ferment, 10, 309. 

BARNES, Z. C. (2001) Personal Communication. 

BISHOP, L. R. (1970) J. Inst. Brewing, 76, 173. 

BOULTON, C. A. and CLUTTERBUCK, V. J. (1993) Proc. 24th Congr. Eur. Brew. Conv, Oslo, 509. 

CARVELL, J., ODDI, L. and HARDING, C. (1998) Proc. 25th Conv. Inst. Brew. (Asia Pacific Section), Perth, 

201. 

CHRISTIAN, A. H. R. (1959) Brewers' Guard. 88, 43. 

COUTTS, M. W. (1957) British Patents, 872391±872400. 

COUTTS, M. W. (1958) Australian Patent, AU 216618. 

DAUOD, I. (1987) Brewers' Guard. 116 (6), 14. 

DAUOD, I., DYSON, R., IRVINE, J. and CUTHBERTSON, R. C. (1989) Proc. 22nd Congr. Eur. Brew. Conv., 

Zurich, 323. 

DELBRUÈ CK, M. (1892) Wochensch. Brau., 9, 695. 

FRICKER, R. (1978) Brewers' Guard. 107, 28. 

GRIFFIN, S. R. (1996) Brewers' Guard. 125 (8), 12. 

HERBERT, D. (1961) Continuous Culture of Microorganisms, SCI Monograph, SCI London, 12, 21. 

HOGGAN, J. (1977) J. Inst. Brewing, 83, 133. 

HOUGH, J. S., BRIGGS, D. E., STEVENS, R. and YOUNG, T. W. (1982) Malting and Brewing Science 2nd 

Edition Volume 2, Aspen, Gaithersburg, Maryland. 

INOUE, T. (1995) Proc. 25th Congr. Eur. Brew. Conv., Brussels, 25. 

INSTITUTE OF BREWING ANALYSIS COMMITTEE (1971) J. Inst. Brewing. 77, 181. 

KNUDSEN, F. B. and VACANO, N. L. (1972) Brewers' Digest, 47 (7), 68. 

LEMER, C., TAEYMANS, D. and MASSCHELEIN, C. A. (1991) Proc. 23rd Congr. Eur. Brew. Conv., Lisbon, 

329. 

LINKO, M., VIRKAJAVI, I., POHJALA, N., LINBORG, K., KRONLOF, J. and PAJUNEN, E. (1997) Proc. 26th Congr. 

Eur. Brew. Conv., Maastricht, 385. 

MAHLER, H. R. and CORDES, E. H. (1969) Biological Chemistry, New York, Harper-Row. 

MARGARITAS, A. and MERCHANT, F. (1984) Crit. Reviews in Biotehnology, 1, 339. 

MASSCHELEIN, C. A. (1997) J. Inst. Brewing, 103, 103. 

MASSCHELEIN, C. A. and ANDRIES, M. (1996) Brew. Distill. Internat. 27 (7), 16. 

MASSCHELEIN, C. A., CARTIER, A., RAMOS-JEUNNEHOMME, C. and ABE, L. (1985) Proc. 20th Congr. Eur. 

Brew. Conv., Helsinki, 339. 

MAULE, D. R. (1976) The Brewer, 62 (5). 

MAULE, D. R. (1977) The Brewer, 63 (6), 204. 

MAULE, D. R. (1986) J. Inst. Brewing, 92, 137. 

MENSOUR, N. A., MARGARITAS, A., BRIENS, C. L., PILKINGTON, H. and RUSSELL, I. (1997) J. Inst. Brewing, 

103, 363. 

MOLL, M. (1983) Proc. 19th Congr. Eur. Brew. Conv., London, 272. 

NATHAN, L. (1930) J. Inst. Brewing, 36, 538. 

OGIE, P. J. (1997) Brew. Distill. Internat, 28 (6), 32. 

PAJUNEN, E., RANTA, B., ANDERSEN, K., LOMMI, H., VILJAVA, T., BERGIN, J. and GUERCIA, H. (2000) Proc. 

27th Conv. Inst. Brew. (Asia Pacific Section), Singapore, 91. 

PASTEUR, L. (1860) Annales de Chimie et de Physique, 3 e Serie. 58, 323. 

PASTEUR, L. (1876) Etudes sur la bieÁre, Paris, Gauthier-Villers. 

PORTNO, A. D. (1973) Brewers' Guard. 102 (7), 33. 

STEWART, G. G. and RUSSELL, I. (1998) Brewer's Yeast, Institute of Brewing, London. 

STRATTON, M. K., CAMPBELL, S. J. and BANKS, D. J. (1994) Proc. 23rd Conv. Inst. Brew. (Asia Pacific 

Section), Sydney, 196. 

TEASS, H. A. (2000) Tech. Quart. MBAA, 37 (1), 37. 

VAN RIJN, L. A. (1906) British Patent, 18045. 

WELLHOENER, H. J. (1954) Brauwelt, 94, 44, 624. 


15 

Beer maturation and treatments 


Beer, at the completion of primary fermentation is said to be `green'. It contains little 

entrained carbon dioxide, it is hazy and its taste and aroma are inferior to beer that is 

ready for sale. In order to refine green beer it must be matured or conditioned. This 

maturation processtakesplacein closed containers inthe brewery and beer treated inthis 

way is called brewery conditioned beer. This is most of the beer produced and sold in the 

world, the exception being cask conditioned beer produced primarily in the UK (Chapter 

21).The process, also called `lagering' whenreferring tobottom fermented beers, used to 

occupy several weeks or even months, but now is often completed in one to two weeks 

and sometimes in considerably less. It follows that accelerated processes of maturation 

have been developed. Traditionally, maturation involves asecondary fermentation and is 

effected by the small amount of yeast remaining in the beer when it is transferred from 

the fermenting vessel. This yeast can utilize fermentable carbohydrates remaining in the 

beer at the end of primary fermentation or small quantities of fermentable carbohydrate 

added in the form of `priming sugar'. In some systems wort is added to provide the 

fermentable material or actively fermenting wort when the process is called `krausening'. 

The carbon dioxide that is produced dissolves in the beer because the vessel is closed and 

the beer becomes `conditioned'. Other gases and volatile substances are produced during 

maturation which are deleterious to beer flavour, e.g., hydrogen sulphide and some 

diketones. Occasional, but systematic, release of pressure on the maturation vessel will 

allow the venting of these compounds to atmosphere. 

Green beer is hazy as well as having an unacceptable flavour. During maturation, 

clarification of the beer takes place. This is by natural sedimentation in the cold 

( 1ëC, 30ëF) of protein and polyphenol complexes, but this process can be enhanced 

and considerably hastened by physical and chemical means and this is now common 

brewery practice. Stabilization of the beer is also an important aspect of maturation. The 

objective is to ensure that turbidity owing to chemical precipitation or growth of microorganisms 

does not occur or, in the case of chemical precipitation, does not recur when 

the beer is clear and stable. During maturation, treatments can be made to the beer to 


adjust its flavour and colour by the use of caramel orother colouring materials and by the 

use of various post-fermentation hop treatments for both bitterness and aroma. 

The final treatment for beer before packaging is filtration and beers which have been 

effectively matured and stabilized by whatever means are easier to filter. This is of 

importance in large breweries where interruptions in beer filtration are extremely 

damaging to overall efficiency. The separation of sediment or `tank bottoms' from the 

maturing beer either by, or before filtration is, therefore, crucial to success. The disposal 

of this sedimented material is becoming aproblem in some areas, e.g., the UK. 

Beers brewed at higher than sales gravity are diluted before packaging and this often 

occurs after maturation but under the same departmental control. The department in the 

brewery in which maturation and clarification is carried out is often referred to as the 

process department. Beer changes or treatments after primary fermentation but before 

packaging therefore comprise: 

· maturation; flavour and aroma changes 

· stabilization against non-biological haze 

· carbonation 

· biological stabilization (pasteurization or sterile filtration, Chapter 21) 

· clarification and filtration. 

Certain special beer types can also derive from further treatments post fermentation. 

These include low-alcohol and non-alcoholic beers and so-called ice beers. 

15.2 Maturation: flavour and aroma changes 

The flavour changes that take place as abeer matures are profoundly important in 

developing the character and hence the brand identity of the beer. Successful brands 

generally have stable flavours and so can be recognized by consumers. This applies to 

national and international brands but even local brands will be expected to display 

consistent taste. Flavour improves during the maturation process but this flavour 

improvement is difficult to characterize and optimize. There is the added factor of the 

effect of oxygen, which will generally cause adverse flavour changes, and so any 

discussion of flavour maturation must include astudy of ways of preventing oxidation. 

Maturation is carried out in many different ways and it is difficult, therefore, to establish 

the underlying principles. What is clear is that processes of secondary fermentation and 

subsequent cold storage are involved that frequently took several months but now are 

usually completed in around two weeks or less, so investment in the fixed assets of 

storage tanks has been reduced. 

15.2.1 Principles of secondary fermentation 

Secondary fermentation permits continued activity by the yeast at areduced rate limited 

by the low temperature and the lower yeast count in the beer. Traditionally after primary 

fermentation (Chapter 12) the beer would pass to the conditioning or maturation vessel 

and would contain 1±4 million yeast cells/ml of beer and about 4ë of gravity (1.1% 

fermentable extract). There are many temperature regimes which are subsequently 

applied, and they represent compromises between promoting production of carbon 

dioxide and hence providing condition to the beer and allowing the removal of 

undesirable flavour compounds. The beer was cooled, traditionally to 8 ëC (46 ëF) at the 


end of primary fermentation to remove most of the surplus yeast before transfer to the 

warm maturation vessel. In this process the remaining yeast becomes re-suspended and 

there is asmall uptake of oxygen, which activates the yeast to start the slow secondary 

fermentation. This results in the conversion of many unwanted flavour compounds into 

flavourless products (O'Rourke, 2000). 

Flocculentyeastsseparateeasilyattheendofprimaryfermentationandconditionscan 

be adjusted such that sufficient yeast can be retained in the beer to effect the flavour 

changes required in maturation (Chapter 12). Powdery yeasts, not separating effectively, 

may ferment too fully in secondary fermentation and remove all residual extract and may 

remain in suspension making clarification difficult. These different situations provide 

constant challenges to fermentation and maturation management. In any event the yeast 

must have access to fermentable carbohydrate for the process to succeed. This 

carbohydrate is provided, as above, by residual gravity in the beer or by the addition of 

sugar by priming or by krausening. Krausening is the addition of wort from the active 

`krausen' stage of the primary fermentation usually at 5±10% by volume of the green 

beer. In shorter secondary fermentation regimes yeast activity must be intense to achieve 

carbonation, purging of the undesirable volatiles, removal of all residual oxygen and 

chemical reduction of many compounds. This leads to immediate improvement of flavour 

and aroma and flavour stability. 

15.2.2 Important flavour changes 

Several important groups of compounds have been identified as changing during the 

maturation of beer with consequent positive effect on beer flavour. The most important 

are: diketones (especially diacetyl), sulphur compounds, aldehydes, and volatile fatty 

acids. 

Diketones 

Diacetyl and 2,3-pentanedione are produced in all brewery fermentations (Chapter 12). 

Diacetyl in particular has an intense sweet, butterscotch flavour. This cannot be tolerated 

in lager beers and the concentration in finished beer should be < 0.1 mg/l. A period of 

warm conditioning (2±3 days at 14±16 ëC, 59 ëF) is very effective in reducing the diacetyl 

content of beer. The precursors of diacetyl and 2, 3-pentanedione, -acetolactate and - 

acetohydroxybutyrate, are excreted by the yeast and are non-enzymically converted in the 

green beer to the vicinal diketones by oxidative decarboxylation. The level of the 

acetohydroxy acids in beer is a function of the yeast strain and is enhanced by conditions 

of rapid yeast growth. Yeast cells will not assimilate exogenous acetohydroxy acids but 

will readily take up and reduce diacetyl and 2, 3-pentanedione to acetoin and 2, 3-pentane 

diol, which have no adverse flavours. The rate of this reaction is dependent on the yeast 

strain, how it has been stored, and its age. This forms the basis of effective diacetyl 

removal from green beer; the yeast must be in a healthy metabolic state to carry out the 

reduction efficiently. Frequent causes of inability to control the concentration of vicinal 

diketones in beer are yeast of the wrong strain or yeast in poor health perhaps accelerated 

by the too rapid onset of fermenter cooling, causing the yeast to separate. If active yeast is 

not present diacetyl will not be reduced. 

The effective and reproducible removal of vicinal diketones from beer is important in 

overall brewery efficiency. Delays in diacetyl removal result in delays in filtration and 

hence delays in the supply of beer to packaging. This has stimulated work to examine 

ways of accelerating the removal of diketones. The most effective way remains the 


choice of an appropriate strain of yeast. However, in some cases strains are used which 

have other desirable properties but are inherently poor at diketone removal. The use of a 

commercially available -acetolactate decarboxylase enzyme has been proposed 

(Hanneman, 1999). It was shown, in tube fermentations and full-scale brewery trials, 

that maturation was accelerated and the need for cold beer lagering eliminated orreduced 

to an absolute minimum. The optimal dose of enzyme depends on the yeast strain and 

wort composition. The enzyme is derived from Bacillus subtilis, which contains agene 

from Bacillus brevis. The enzyme is now used in many parts of the world (Jepsen, 1991). 

It is not approved for use in the UK. 

Anovel method of maturation, allowing very rapid reduction of the diacetyl content, has 

been described using immobilized yeast in abioreactor (Pajunen et al., 1991; Pajunen and 

JaÈaÈskelaÈinen, 1993, see Chapters 13 and 14). Yeast cells are immobilized on beads of 

DEAE-cellulose (diethylaminoethylcellulose). The brewing yeast is first separated from the 

green beer by centrifugation to give avery low yeast count. It is then heated to 90ëC 

(195ëF) for 8±10 minutes to convert all the -acetolactate to diacetyl. The beer is then 

passed through the immobilized yeast reactors and the diacetyl is converted to acetoin. The 

beer is then ready for filtration and packaging. The technique has been used commercially 

(Pajunen and JaÈaÈskelaÈinen, 1993) to operate at an annual volume of one million hl. In this 

system there are four bioreactors with amaximum total flow of 14m 3 /h, which corresponds 

to the centrifuge capacity and the time needed for rapid emptying of primary fermenters. In 

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 

dioxide produced during beer maturation. There is some evidence that hydrogen sulphide 

is more effectively purged from beers where carbon dioxide production is vigorous 

(Zangrando and Girini, 1969). 

Aldehydes 

Acetaldehyde in particular can affect beer flavour. This arises from the oxidation of 

ethanol and can occur if transfer of the beer from primary fermentation to maturation is 

carelessly carried out giving the opportunity for oxygen uptake by the beer. During 

normal maturation the acetaldehyde concentration will decrease to 2±7 mg/l. Acetaldehyde 

can be detected at about 10 mg/l in a lightly flavoured pilsen-type beer, when it 

gives a flavour of green apples. It is less easily detected in ales and is a characteristic of 

some ale flavours. Concentrations of > 35 mg/l should be avoided. 

Volatile fatty acids 

Beer conditioning temperature is very important in determining the excretion of C4 to C10 

fatty acids. Synthesis of short chain fatty acids by yeast stops at the onset of maturation. 

C8 fatty acid increases in concentration during fermentation and this is replaced in 

maturation by C10 acid. Glycerides and phospholipids are synthesized during maturation 

and so there is a general trend for a reduction in volatile acids as ageing proceeds. This 

trend can be reversed if maturation is extended too far. There can then be a rise in the 

concentration of free fatty acids owing to the hydrolysis of reserve glycerides, with 

consequent adverse effects on flavour. If a high maturation temperature is maintained for 

too long then there can be a slow excretion of C10 acid (capric acid), which has a flavour 

threshold of 10 mg/l and this is undesirable. Maturation is seldom controlled specifically 

from the viewpoint of controlling volatile fatty acids. Provided that the warm 

conditioning is not extended beyond the time needed to reduce the concentration of 

vicinal diketones, and cold storage is controlled to the time required for stabilization of 

haze precursors, there should not be a problem. 

15.2.3 Techniques of maturation 

Technique varies widely from brewery to brewery. In general, for bottom fermented beers 

the technique must ensure the production of a balanced beer flavour with the minimum 

concentration of diacetyl. For brewery conditioned top fermented beers, the technique 

centres around creating the required haze stability as the more robust flavour of the top 

fermented ale derives from primary fermentation and diacetyl reduction is not a problem. 

Lager methods 

Beer is transferred to a separate tank for maturation in traditional lagering methods. 

Before the end of primary fermentation cooling is applied to the cone of the fermenter to 

achieve a temperature of 5 ëC (41 ëF). The remainder of the beer above the cone (at least 

95% of the volume) is at a higher temperature to ensure effective diacetyl removal. 

Cooling below 5 ëC (41 ëF) is not necessary and runs the risk of cooling the beer beyond 

its point of maximum density when inversion of flow around the fermenter may occur 

(Andrews, 1997). About 24 hours after applying cooling an initial removal of yeast is 

usually carried out. This yeast is usually discarded. When diacetyl reduction is complete 

the remainder of the beer is slowly cooled to 5 ëC (41 ëF) to complete the maturation by 

adjusting the flavour volatiles. A sudden fall in temperature must be avoided or the shock 


may induce the yeast to excrete protease enzymes that could be detrimental to foam 

stability. This cooling may take from two to nine days. At the end of this period asecond 

removal of yeast is usually made. The beer is now completely cooled to at least 1ëC 

(30ëF). The period of colloidal stabilization now takes place at 1to 2ëC (30 to 28ëF) 

for two to three days. Afinal yeast removal is made before filtration. The colloidal 

stabilization temperature must be maintained throughout the whole vessel. 

There are many variations on this technique which would still be regarded as 

traditional. Indeed the times can be much longer (Miedaner, 1978) with storage times at 

1ëC (30ëF) being up to six weeks. In this situation up to 40% of the capital cost of the 

brewery could be in the tanks required to condition beer (Coors, 1977). Fermentation and 

maturation can alsobecarriedout inonevessel soavoiding transfer betweenprimaryand 

secondary fermentation. 

Because of the high capital costs associated with building tanks and the working 

capital of stored beer, brewers have sought to reduce maturation times. Yeasts with the 

ability to reduce diacetyl rapidly have been used and warm storage times have been 

almost eliminated. There are almost as many variations on maturation technique as there 

are breweries. At least one successful system comprises: fermenter filling 20 hours, 

primary fermentation 72 hours, warm storage 48 hours, cooling 48 hours, cold storage at 

1ëC (30ëF) 36 to 48 hours. Total time is less than ten days. These systems can produce 

sound beer. However things can go wrong and they are susceptible to variations in raw 

material quality. To escape from periods of uncertainty over maturation and hence 

flavour stability, brewers need to keep in mind the traditional principles of the process. 

One of the most successful systems for reducing maturation time seems to be the system 

of using immobilized yeast in abioreactor (Section 15.2.2). 

Ale methods 

The traditional method for maturation and conditioning of ale is cask conditioning 

(Chapter 23). Now many types of ale are brewery conditioned and filtered and sold in 

kegs (Chapter 23). Ale fermentations are rapid and vigorous and usually completed in 48 

to 60 hours (Chapter 12) at temperatures of up to 24ëC (75ëF) before being cooled to 

Turbulence causes the loss of carbon dioxide and the uptake of oxygen. Beer must 

therefore be moved gently through correctly sized pipes. 

Movement of beer from one tank to another, e.g., from fermenter to maturation vessel 

is often encouraged by applying top pressure of carbon dioxide to the donor tank after 

flushing the mains and the receiving tank with the gas. But carbon dioxide is expensive 

and the cheaper nitrogen gas is now often used. The high density of carbon dioxide, 

relative to air, can be exploited. Athin blanket of carbon dioxide can be produced above 

the beer in the donor tank, and atop pressure of air can be used to move the beer out of 

the tank. Asmall amount of carbon dioxide can then be injected into the beer en route to 

the receiving tank where the gas escapes to provide the blanket of carbon dioxide above 

the beer now in the receiving tank. As the tank fills the air within it is displaced upwards 

and does not come into contact with the beer. Alternatively the receiving tank can be 

filled with deaerated liquor (containing

Caramels used in brewing are electropositive and can have acolour of 50,000ëEBC 

and contain 65% solids. In Europe, caramels are classified and those used in brewing are 

designated E150c caramels. These caramels can contain 4-methyl imidazole which is 

toxic to rabbits, mice and chickens (Hasimoto, 1973) and more recent research has 

revealed the presence of 2-acetyl-4-tetrahydroxybutyl imidazole (THAI), which has been 

found to react as a anti-pyridoxine factor in rats leading to lympholeucocytopaenia 

(Leach, 1989). For these reasons caramels used in brewing must have THAI levels 

fermentation and so change the aroma of beer. Some of these preparations contain resins 

as well as oils and are known as oil-rich extracts. They must be added to beer before 

filtration to avoid the possibility of haze developing in the beer during storage. 

Recently (Marriott, 1999) there have been further developments to produce aroma 

products that are completely soluble in bright beer (see also Chapters 7and 8). These can 

give potentially 100% utilization and so acompletely reproducible effect on aroma. 

Essences have been produced that can reproduce the effect of late copper hopping and 

even the dry hopping of cask ale (Chapter 7). Essentially the insoluble hydrocarbon 

fraction of the oil must be removed to produce asesquiterpeneless oil. This oil can then 

be further fractionated to produce individual products with `spicy' or `floral' characters. 

These essences can be added to bright beer at rates of 50 to 100 g/l and have profound 

effects on aroma and flavour. They are normally supplied as 1% solutions in ethanol. 

Blending 

Beers can be blended post fermentation and this yields more uniform products and 

provides further opportunities to deliver avolume of beer to its final specification. 

Recovered beer can also be added post fermentation but this is atechnique requiring 

considerable care, often demanding excessive pasteurization of the beer with a 

consequent adverse effect on flavour. For this reason recovered beer is best added 

before fermentation, e.g., to the whirlpool (Section 15.5). 

Sulphur dioxide 

Sulphur dioxide is both anatural product of fermentation (Chapter 12) and can be added 

to beer after fermentation. It provides a measure of protection against flavour 

deterioration by oxidation and has bacteriostatic properties. Maximum levels of sulphur 

dioxide in beer are usually governed by statute and vary in different countries. In the 

European Union the maximum permitted level is 20mg/l (as SO2), except for cask 

conditionedbeerwhenthelevelis50mg/l.Sulphurdioxide isusuallyaddedassodiumor 

potassium metabisulphite and can so be added along with finings or priming sugar. 

Water 

It is an accepted strategy in many breweries to brew at ahigher gravity than that at which 

the beer is subsequently sold (Chapter 6). Many studies have been made on the effect of 

gravity (wort strength) on the properties of the resultant beer and the physiological health 

of the yeast after subsequent generations at high gravity (>1060 or 15ëP) (Pfisterer and 

Stewart, 1975; Stewart et al., 1999). Dilution of the fermented high gravity beer is best 

effected after beer filtration and the quality of the diluting water used is of the utmost 

importance (Chapter 3). This needs to be of brewing quality because this water is to be 

drunk by the consumer of the product. It must be free from taint, sterile and deaerated. 

Thisisnormallyachievedbytreatingthewaterbypassingitthroughapurpose-builtplant 

involving atrap filter to remove solids, acarbon filter to remove chlorine residues and 

inert gas stripping or vacuum stripping to remove oxygen. Sterilization is often effected 

by the use of uv light or by filtration through asterilizing sheet filter (0.45 m). The 

oxygen level in the water should be

the beer stream following accurate measurement of density often by the oscillating Utube 

method. 

15.2.5 Maturation vessels 

Vessels for the maturation of ales or lagers are normally cylindrical in shape. They are 

similar to fermentation vessels (Chapter 14). They can be horizontal or vertical in aspect. 

Tanks can be made of stainless steel, mild steel with aglass, enamel or plastic epoxy 

lining, or more rarely of aluminium. Horizontal tanks are usually of 100 to 500hl (60± 

300 imp. brl) capacity, but the vertical cylindroconical tanks can be up to 6,500hl (4,000 

imp. brl). Horizontal tanks are frequently built inside the brewery in warm or cold 

conditioning rooms. Vertical tanks are externally clad and usually stand in the open. 

Tanks are normally fitted with impellers for mixing and with atemperature control 

system.Temperatureisusuallycontrolledbyacoolingjacketsuppliedwithbrine,ethanol 

as industrial methylated spirit (IMS), or propylene glycol. Ethylene glycol has been used 

but it is highly toxic. Older designs incorporated internal cooling coils supplied with 

chilled water or IMS. Frequent inspection of the integrity of the coils is essential if a 

coolant other than water is used, as small leaks would be disastrous for beer quality. 

The escape of carbon dioxide gas is limited by closing the vessel and hence pressure 

regulation is required. Devices are used to control gas pressure up to 1.4 bar (20lb./in. 2 ). 

Water column or mercury column manometers were once used but now weight loaded 

valves are employed, which can be set to open and release pressure at adefined level 

(Fig. 14.3). 

Materials of construction and vessel size 

Construction is similar to that of fermenting vessels (Section 14.3.2). The preferred metal 

of construction would now be type AISI 304 stainless steel, it is rare that conditions are 

such that the much more expensive AISI 316 is needed. The inner surface of the tank 

shouldbeassmoothaspossiblesoastoprovidenosurfaceindentationsforthelodgingof 

soil (Section 14.3.2). 

Size is important when the vessels are to be used for fermentation and maturation, i.e., 

uni-tanks (Fig. 15.1). This relates to the hydrostatic pressure on the yeast during 

fermentation and agenerally acceptable height for afermenter is now not greater than 

15m(about 50ft.). In apurpose-built maturation tank there is no such restriction and 

tanks up to 30m(almost 100ft.) and even 40m(130ft.) in height have been built. Size 

usually relates to brewery throughput and brew lengths. Arule of thumb is that the size 

should be equivalent to ahalf-day production, larger tanks take too long to fill and will 

contain beers of variable age and hence potentially variable final flavour. Tank diameters 

are usually 3.50±4.75m(11.5 to 16ft.) and the cone angle is 60ë to 75ë. The ratio of 

diameter to beer height in the cylinder can vary from 1:1 to 1:5. The volume of the 

headspace in the tank relates to the work to be done. Avessel used solely for cold 

maturation will require aheadspace volume of 5% of the total, whereas if diacetyl 

removal in the warm is included the volume excess should be 10%. Afully equipped 

fermenting vessel will need 25% headspace. 

Cooling 

The details of temperature control in fermentation were discussed in Chapter 14 and 

many of the principles apply to the maturation vessel. Chilled water is unsuitable for use 

as a coolant below 2 ëC. Propylene glycol has the disadvantage of being extremely 


TPA 

SB 

PT 

Cooling 

panel 

Cooling 

panel 

Cooling 

panel 

TP 

Gross volume 

Working 

volume 

viscous (13.4mPa at 0ëC). Ethanol (IMS) is often the preferred coolant, but it should be 

noted that both IMS and propylene glycol will oxidize slowly to acid products which will 

attack mild steel pipework and cause sludge deposits. Exposure to air should therefore be 

minimized and hence solution strengths should be maintained (30% for ethanol and 40% 

for propylene glycol). Sometimes acorrosion inhibitor is included with the secondary 

coolant. There has been renewed interest in direct expansion systems using aprimary 

refrigerant such as ammonia. Ammonia has no ozone depleting effect and therefore has 

this major advantage over the fluorinated hydrocarbons, which are now being withdrawn 

as refrigerants. However ammonia is avery corrosive, dangerous and pungent gas and 

brewers still tend to favour indirect systems (Section 14.3.3). 

Several designs of cooling jacket are available (Chapter 14). The requirements are 

more simple than in the fermenter or combined vessel and the choice often depends on 

the manufacturer's capability. Coils can be wound onto the vessel section and seam 

welded (limpet coils). An alternative uses aprofile plate, which is preformed and spotwelded 

on tothe vessel,avariation ofthe preformed plate is the `dimple jacket', which is 

also preformed, and spot welded (Fig. 15.2). An inflated `quilted jacket' is afurther 

alternative most frequently used in direct expansion systems with ammonia. Most vessels 

will have two cooling sections on the cylinder of the vessel and adimple jacket on the 

cone. The cone section cooler will usually be used at the end of primary fermentation to 

aid yeast flocculation. 

TP 

Bottom manway 

Fig. 15.1 Schematic representation of dual-purpose fermenting vessel (1600hl, 1000 imp. brl); TP, 

temperature probes; PT, pressure transducer; TPA, top plate assembly containing a pressure relief 

valve (0.8 bar, 12 lb./in. 2 ), anti-vacuum valve (55 mm WG), CO 2/CIP combination valve, top pressure 

transmitter, access hatch, inspection light, trace heating and cabin for weather protection; SB, spray 

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 

(66 lb./in. 2 ); Height 16.0 m; Diameter 4.25 m; Working temperature, 1 to 90 ëC (30±190 ëF); Working 

pressure, 0.7 bar (10 lb./in. 2 ); Wort depth 13.5 m; Weight empty 14 t; Weight full 212 t; Gross volume 

1922 hl; Working volume, 1640 hl; Freeboard volume 282 hl (17% of working volume); Cone volume 

147 hl (9% of working volume; included cone angle 70ë) (Barnes, 2001). 


TP 

10.5 m 

3.0 m 12.0 m 

1.0 m

Plain dimple Limpet 

Pressed 

corrugated 

Inflated 

panel 

Fig. 15.2 Designs ofcoolingjackets fordual-purpose fermenting vessels: plainjacket, notusually 

used because it requires thick vessel construction to resist external pressures; limpet, provides a 

flow-path for coolant, but again relatively heavy; pressed corrugated, lighter construction than 

limpet and widely used for secondary refrigerant systems; inflated panel, low volume and suited to 

primary refrigerant systems acting as an evaporator (Barnes, 2001). 

Refrigerationloadinabreweryisaconsiderablefixedcostandthereforethedesignofthe 

controlsystemtoensurethemostefficientuseofenergyisimportant.Thenumberofpumps 

for secondary coolant should be minimized in the design. Avariable speed drive in the 

secondary coolant system will maintain aconstant pressure differential in the supply and 

will help to minimize losses. There should also be the facilityto shut downthe system if no 

vesselsrequirecooling.Thepositionofthetemperaturesensoriscriticaltoefficientcooling. 

Thesensorshouldbepositionedatbetweenone-thirdandone-halfofthevesselheightfrom 

thebottomandshouldprotrudeabout500mm(20in.)intothetanktoavoidanytemperature 

effectfromtheinternalsurfaceofthetank.Informationfromthesensorinamodernbrewery 

will be fed back to acomputer at which the desired temperature regimes may be set. 

Cleaning-in-place (CIP) 

Cylindroconical maturation vessels are cleaned by CIP systems. The principle of this 

cleaning is very similar to that described for fermenting vessels (Section 14.3.3) and 

discussed in overall detail in Chapter 17. Maturation vessels do not generally have such a 

high organic soil level (yeast deposits) as fermenting vessels and therefore respond to 

cleaning with acids, which is usually cheaper than caustic alkali based systems using a 

sequesterant. However, as with fermenters to save costs of detergent, the bulk of the soil 

should be removed with jets of water in an impact system. The jets used vary enormously 

in design but are often similar to those used for cleaning fermenting vessels, where the 

nozzles on the head rotate in one plane whilst the head itself rotates at right-angles on the 

support pipe (Fig. 14.6). The basic sequence of CIP is similar to that used for fermenters 

(Section 14.3.3) with slightly shorter time sequences. 

Insulation 

Similar principles apply to those discussed in Chapter 14 in relation to fermenting 

vessels. Horizontal tanks are usually contained within aroom in the brewery which is 


temperature controlled. The tanks themselves are not insulated. Vertical tanks require 

insulation. Insulation materials often contain chloride ions and a chloride barrier is 

needed against the stainless steel of the vessel to prevent attack of the surface by chloride 

(Section 14.3.2). 

15.3 Stabilization against non-biological haze 

Competition between brewers is intense and the quality and consistency of their beers is 

paramount. This demands that the beers following maturation should not only have 

desirable, stable flavours but must also display stability with respect to haze, i.e., the 

beers must be bright and remain so during the period from dispatch from the brewery to 

drinking. Therefore, in addition to removing yeast, beers must have the precursor 

constituents of haze removed to ensure long-term stability. Beer haze and its chemistry 

are considered in Chapter 19. In this Chapter we focus on the removal of haze-forming 

materials to ensure the production of astable beer. 

Arange of substances can cause non-biological haze in beer: 

· -glucans, which can often lead to hazes not easily seen by eye but which cause high 

levels of light scattering in 90ë haze meters 

· -glucans (starch), which can behave similarly to -glucans 

· pentosans, which may be derived from wheat based adjuncts 

· dead bacteria from malt 

· oxalate from calcium deficient worts. 

However, the most common, important and troublesome type of non-biological haze is 

that deriving from the cross-linking of proteins and polyphenols and it is the elimination 

of the precursors of these polymers to which beer stabilization treatments are directed. 

The most effective beer treatment with respect to haze stability is the cold storage of 

the beer for about seven days at 1 to 2ëC (30±28ëF). This technique allows a 

reduction in the cost of other beer treatments designed to remove potential haze-forming 

proteins and polyphenols. However, brewers frequently wish to accelerate the process of 

haze stabilization and achieve greater stability than is possible with cold storage alone. 

15.3.1 Mechanisms for haze formation 

Colloidal haze in beer arises from the formation of protein-polyphenol complexes during 

beer storage (Gopal and Rehmanji, 2000, and Chapter 19). Fresh beer contains acidic 

proteinsand numerous polyphenols. These can come together byloosehydrogen bonding 

but theassociations formed aretoo small tobeseen bythe nakedeye. Thesepolyphenols, 

called flavanoids, can further polymerize and oxidize to produce condensed polyphenols, 

which have been called tannoids (Chapon, 1994). These tannoids can `bridge' by 

hydrogen bonding across anumber of proteins to form areversible chill haze (Fig. 15.3). 

This haze forms at around 0 ëC (32 ëF) but redissolves when the beer is warmed to 15 ëC 

(59 ëF). After further storage of the beer strong bonds can form between the tannoids and 

proteins and irreversible, permanent haze is formed. The rate at which this haze is formed 

and its extent of formation depends on the raw materials used in wort preparation and the 

process conditions. This `model' suggests that effective stabilization should be achieved 

by removing from the beer the constituents of the haze, i.e., the `tannin sensitive' proteins 

and/or the polyphenols. 


Simple 

flavanoids 

Protein 

Oxidized flavanoids 

= tannoids 

Proteins 

Tannoids 

Proteins 

Chill haze 

Permanent haze 

Fig. 15.3 Models of chill and permanent hazes development in beer (Gopal and Rehmanji, 2000). 

An alternative model of haze-formation has been proposed (Siebert et al., 1996 and 

Chapter 19). This suggests there are afixed number of binding sites on haze-forming 

proteins (proline residues) and that haze-forming polyphenols have two binding sites, 

through which they can jointo two adjacent protein molecules. If there is an excess of 

protein over polyphenol then the polyphenol is involved in binding just two protein 

molecules together and these dimers do not constitute insoluble complexes. If the 

amount of polyphenol greatly exceeds that of protein then there is a shortage of protein 

binding sites and again haze complexes will not be formed. Hazes are therefore formed 

when there are equivalent amounts of protein and polyphenol in the beer. This model 

suggests an alternative strategy for the prevention of haze, i.e., substantially increase the 

amount of either protein or polyphenol. This is not a favoured approach and most 

brewers will seek to reduce levels of either the proteins or polyphenols or most likely 

both. 

15.3.2 Removal of protein 

All of the haze-forming protein in beer comes from malt. Proteins which are particularly 

liable to cause haze are rich in proline and have molecular weights > 10,000. A 

concentration of 2 mg/l will give a haze value in beer of > 1.0 EBC formazin units 

(Chapter 19) which will give a perceptible turbidity to the beer. However, other proteins 

can also form haze and some of these have good foam potential. It is not, therefore, a 

simple matter to categorize the proteins responsible for haze-formation in beer. It can 


simply be concluded that the presence of hydrophobic groups on the surface of protein 

tertiary structure increases the capacity to form haze and the capacity to improve foam. 

Lowering the overall protein concentration will increase haze shelf-life. The use of low 

nitrogen malts (TN 40%) have inherently poor 

foam potential. It is thus theremoval of protein thatisimportant. Proteins can essentially be 

removed by hydrolysis; usually by enzymes, by precipitation or by adsorption. 

Hydrolysis 

The usual enzyme to use is papain derived from papaya fruit although bromelin from 

pineapple and an acidic protease from Bacillus subtilis have been used (Chapter 2). The 

optimum dose is 5±15 10 6 units/hl (1 unit is the amount of enzyme generating 1 g 

soluble tyrosine/hour under rigidly standardized test conditions). This usually translates to 

0.5±4.0g/hl depending on the formulation. The enzyme needs acontact time and is usually 

added to the beer during maturation. Some preparations can be added to final beer but they 

must be sterile and totally soluble. This treatment is cheap but is risky. Proteolytic enzymes 

will damage beer foam proteins and some can survive pasteurization so will continue to 

have an effect on foam during beer storage. There can also be an adverse effect on flavour 

stability by the liberation of free thiol groups from peptides in the beer. 

Precipitation 

Since proteins form haze complexes with polyphenols it is not surprising that aclass of 

polyphenol can be used for the removal of haze sensitive protein. Thus anionic tannic acid 

reacts with cationic proteins to form an insoluble complex. This is as aresult of interaction 

between ketone groups on the tannin and nucleophilic groups (SH, NH 2)on the protein. The 

most effective tannins are known as gallotannins extracted from gall nuts or from the sumach 

tree.Thetanninpreparationwouldnormallybeaddedenroutetomaturationat0ëC(32ëF)at 

5±9g/hlandwouldrequireatleast24hourscontacttimeforfulleffectiveness.Essentiallythe 

tannins act as aprecipitant of the tannin sensitive proteins (Anderson et al., 2000) and when 

used in this way in the cold tank alarge volume of `tank bottoms' is formed. This requires 

careful movement of the beer from above the sediment or the use of acentrifuge or filter to 

separate beer from the tank bottoms (Section 15.5). If this is not done losses of beer will be 

high. For this reason the use of tannic acid for protein removal has generally gone out of favour 

in recent years. However, improved preparations of gallotannins have become available 

(Musche and de Pauwe, 1999). These high molecular weight gallotannins can be dosed into 

the beer before filtration into the inlet buffer tank to the filter at 0 ëC (32 ëF) at rates of 2±4 g/hl. 

A contact time of 5±25 minutes is claimed to be sufficient. This method of treatment avoids 

the production of tank bottoms but is limited by the relatively short filtration run achieved and 

for this reason centrifugation is still often used before the filter. Gallotannins have the added 

property of removing metals from beer (Musche and de Pauwe, 1999). Iron, aluminium, lead 

and copper can all be removed by gallotannin treatment and filtration. The metals can be 

lowered in concentration to below the levels at which the properties of beer are harmed. 

Adsorption 

Silica gels are used as protein adsorbents. They are produced for brewers in two main 

commercial forms: hydrogel (> 30% water) and xerogel (< 10% water). To produce a 


silica gel, sulphuric acid and sodium silicate solution are mixed under controlled 

conditions to produce a `hydrosol' of colloidal polysilicic acid. These colloidal species 

can be condensed to produce a hydrogel in which the water phase is immobilized in the 

silica matrix (McKeown and Earl, 2000). The hydrogel can then be further processed and 

then milled to produce a material of defined particle and pore size and surface area. 

Optimum pore size is in the range 30±120 m diameter with a surface area of around 

800 m 2 /g. 

Alternatively, the hydrogel can be dried before milling to produce a xerogel where the 

surface area can range from 300±800 m 2 /g with again pore sizes in the range 30±120 m. 

Protein adsorption takes place on so-called silonol (SiOH) sites on the silica gel and 

adsorption capacity is a function of the number of these sites. Generally, though, the 

number of sites exceeds the quantity of available protein. 

Silica gels can be added to beer in maturation but their most attractive characteristic is 

their usefulness when dosed into the beer stream prior to filtration. Adsorption is 

therefore achieved with a very short contact time (80% of the adsorption is achieved in 

< 3 min.) between the silica and the beer. This also provides the opportunity to treat only 

a proportion of the beer in a maturation vessel, i.e., that proportion destined for a long 

shelf-life (> 40 weeks). In this situation the filtration quality of the gel is as important as 

its adsorbent property. The best filterability is obtained with large size gel particles but 

these are not so good for protein adsorption. Generally, hydrogels are good adsorbents but 

because of small particle diameter (< 10 m) are relatively poor filter aids. Xerogels have 

larger diameter gel particles (up to 100 m) and so are less good at adsorption than 

hydrogels but have superior filtration characteristics. Rates of addition are normally 

between 50 and 80 g/hl. Brewers have to make a choice of which material to use 

depending on their circumstances. It is claimed (McKeown and Earl, 2000) that it is now 

possible to manufacture small particle sized xerogels that will deliver high filtration rates 

as well as displaying the optimum adsorption characteristics of a small sized hydrogel 

(< 10 m). 

Beer proteins are associated with the good foaming potential of beer. There has 

therefore been concern over the effects of silica gels on foam and head retention. The 

amino acid sequence of the protein determines its hydrophilic/hydrophobic balance. It is 

generally accepted that hydrophilic proteins are haze-formers and hydrophobic proteins 

promote foam. 

The selectivity of different silica gel preparations for removing haze-forming proteins 

has been studied in both all-malt and high adjunct (maize) beers (Guzman et al., 1999). 

Total protein adsorption was influenced by beer type but adsorption of haze-forming 

protein was not. Further, it was possible to manufacture a xerogel which showed 

selectivity for haze-forming, hydrophilic proteins. Maximum uptake appeared to occur 

around 100 m pore diameter. 

An essentially different method of using the adsorptive properties of silica derivatives 

to remove protein is to use the compound in the form of the liquid silica hydrosol, the 

intermediate in the manufacture of the hydrogel or the xerogel (Green et al., 2000). The 

hydrosol is a bluish opalescent liquid with a specific area of 300 m 2 /g in relation to its 

SiO 2 content. Essentially, if added to beer the hydrosol forms a gel by cross-linking the 

SiO 2 particles. The gel flocculates and sediments and so adsorbs protein, which settles 

out as tank bottoms. The silica sol can be added to the beer in the maturation tank. 

Careful mixing is required and considerable tank bottoms are produced which must be 

drawn off before filtration or removal by centrifugation. The sol can also be added with 

the body feed to the filter when it is effective in lowering haze but will not contribute to 


filtration performance like the gel. Further development work on silica sol and gel 

structure in relation to the removal of hydrophilic haze-forming protein from beer seems 

likely to be worthwhile. This technique remains very useful and has wide acceptance by 

brewers. 

15.3.3 Removal of polyphenols 

There is still some concern from brewers that techniques relying solely on the removal of 

protein for haze stabilization will affect beer foam. Techniques to remove polyphenols 

from beer have also been developed. The polyphenols responsible for haze-formation by 

interacting with proteins derive from malt with asmall contribution from hops. Thus the 

possibility exists of eliminating polyphenols from beer by eliminating them from malt. 

Two basic methods of polyphenol removal have therefore been practised; use of 

polyphenol-free malt and the much more widespread adsorption technique applied to 

wort or beer. 

Adsorption 

Early work involved the use of synthetic materials such as `Nylon 66', which were 

effective polyphenoladsorbents.FurtherworksawthedevelopmentofthepolymerPVPP 

(polyvinylpolypyrrolidone). This is across-linked polymer, which is insoluble in water, 

alcohol and acid (Gopal and Rehmanji, 2000) and hence has ahigh surface area for 

adsorption of haze-forming polyphenols. This occurs at the surface of the material by 

strong hydrogen bonding. It is proposed that the structure of the synthetic polymer limits 

internal bonding thus maximizing the number of external sites and reducing the 

concentration of PVPP needed for effective stabilization. 

PVPP can be employed as a`single use agent' and in this situation the insoluble 

PVPP-polyphenol complex is removed on the filter with kieselguhr depth filtration. 

PVPP for single use is ground to afine powder with alarge surface area to weight ratio 

and is used at rates of 15±25g/hl. The PVPP can be added to the beer stream along with 

kieselguhr at 0ëC (32ëF) and, like silica gel, ashort contact time is effective, although in 

this case about ten minutes is needed from the point of contact with the beer to removal 

on the filter. Some brewers prefer to add single use PVPP to the maturation tank. This is 

usually at rates of 10±15g/hl and alarge dense grade of powder is used that promotes 

sedimentation.Athoroughmixingofthetankcontentsbypumpedrecirculationisneeded 

for optimum effectiveness. The bulk of the PVPP-polyphenol complex is removed from 

the base of the tank after sedimentation, with final removal of suspended matter on the 

filter. 

PVPP can also be used as aregenerable product. Washing with hot caustic soda 

solution breaks the PVPP-polyphenol bonds. Addition rate is normally 30±50g/hl. The 

powder is of larger particle size and has greater mechanical strength than that for single 

use. The usual technique is to use the regenerable PVPP in ahorizontal leaf filter or a 

candle filter (Section 15.5). Apre-coat of PVPP of 1±2mm (0.04±0.08in.) depth is 

layered onto the screens in the filter and PVPP is dosed into the beer stream by a 

proportioning pump. A rate of 10 hl/m 2 /h can be achieved and the precise rate is matched 

to that of the kieselguhr filter which follows. The treatment run will end when PVPP has 

filled the space between the screens of the filter. Spent PVPP is regenerated by 

circulating a solution of 1±2% sodium hydroxide at 60±80 ëC (140±175 ëF) through the 

PVPP filter bed for 15±30 minutes. The filter cake is then flushed with water at 80 ëC 

(175 ëF) to lower the pH value. 


This is normally followed by a rinse with dilute acid, usually phosphoric (1% w/w) or 

nitric (0.3% w/w), until the pH value of the solution leaving the filter is about 4.0. This 

treatment removes carbohydrates and calcium salts trapped in the debris of the filter aid. 

The filter is then washed with cold water to achieve a neutral effluent. Finally water and 

carbon dioxide gas are used to displace the regenerated PVPP to the dosing tank. This can 

be assisted by spinning the screens in the filter to provide a centripetal force. Process 

losses are about 1% and these are made up by adding more PVPP to the dosing tank 

before repeating the cycle of treatment. There is greater capital cost in the use of 

redeemable PVPP but, for breweries with large outputs of beer requiring stabilization for 

long shelf-life (> 40 weeks), this is a cheaper option than single-use PVPP. 

Generally, it has been accepted that the most effective stabilization occurs, whether 

with PVPP or silica gel when combined with the use of temperatures of 0 to 2 ëC (32± 

28 ëF) and at times in excess of 48 hours. However, recent work suggests that PVPP can 

deliver stabilization at 4 ëC (39 ëF) which is equivalent to that obtained at 0 ëC when 

assessed by shelf-life measurement (Byrne et al., 1999). This could be of considerable 

significance in energy saving. 

Proanthocyanidin free malt 

Polyphenols in beer in the main derive from malt. The most reactive of the polyphenols 

are proanthocyanidins or anthocyanogens and are found in the barley testa. It would 

therefore seem feasible that if malt could be produced from a barley free of 

anthocyanogens then haze stability might be achieved without the intensive postfermentation 

treatments described above. 

Barley breeding work was carried out in the Carlsberg laboratories in the early 1970s 

with the objective of inducing mutations in barley so that anthocyanogen producing genes 

were ineffective (Jende-Strid, 1997). It should be noted that the resulting barleys were not 

produced by genetic transformation by the transplanting of genes from one species to 

another. The induced mutations were transmitted to commercial cultivars by classical 

techniques of crossing and re-selection. In the last 30 years over 700 mutants of winter 

and spring barleys have been produced (Sole, 2000). The problem has been the 

commercial acceptability of the varieties from both an agronomic and malting and 

brewing standpoint. Some varieties showed no dormancy and so lodged and effectively 

malted in the field and so were useless for commercial malting. There was no doubt 

however that beers could be brewed from anthocyanogen-free malts, which had haze 

shelf-lives equivalent to those beers stabilized with PVPP and/or silica gel. 

The latest varieties to be trialled are Caminant, Chamant and Prominant in mainland 

Europe and Clarity in the UK. There has been relatively little interest in other parts of the 

world. Clarity appears to have sound agronomic properties and has been grown 

successfully on the light lands of eastern England. Clarity malt has been trialled in pilot 

breweries at Brewing Research International in the UK and at commercial breweries. 

Advantages have been demonstrated over conventional malts in lower polyphenol levels 

in wort and in haze shelf-life tests (Sole, 2000) and in the opportunity to use higher 

temperatures for stabilization (+4 ëC compared to 1 ëC). The problem has been in lower 

extract yield in the brewhouse, which has often been at least 1%, and this has limited 

acceptability of Clarity malt for some brewers. This has been counteracted in some 

situations by using Clarity malt as a proportion of the grist (say 50%). 

Full acceptance of the use of anthocyanogen-free malts as an effective means of 

stabilizing beer awaits the demonstration of the economics of the whole system in 

relation to the costs of PVPP treatment. 


15.3.4 Combined treatments to remove protein and polyphenols 

Refined silica gel treatments are very effective at selectively removing haze-forming 

protein and PVPP will effectively remove the haze-forming polyphenols from beer. 

However,there is stillsome riskthat foam potential can be reduced byexcessive removal 

of protein and there is abelief that polyphenols contribute anti-oxidant properties to beer 

and give improved texture to beer in the mouth. These considerations have been used as 

reasons to favour astabilization approach dependent on the removal of defined amounts 

of both protein and polyphenol. PVPP can be used with papain (Esnault, 1995) and is 

added with the enzyme. The dangers of excessive proteolysis with papain remain. PVPP 

cannot be added with gallotannins as it complexes immediately and will not adsorb the 

polyphenols in the beer. It can be added after gallotannin treatment (Musche and de 

Pauwe, 1999) when it will adsorb polyphenols and will complex residual tannic acid that 

can contribute aharsh astringency to the beer. 

The most effective combined treatments have seen PVPP used along with silica gel 

added from separate dosing vessels (McMurrough et al., 1999). Combined stabilization 

was shown to be more cost effective than the intensive treatments needed when PVPP or 

silica gel are used alone. Recently acombined PVPP/silica xerogel (3:7) preparation has 

been described (Gopal et al., 1999) that works to reduce proteins and polyphenols in a 

single addition at 80g/hl. This treatment is used without PVPP regeneration and achieves 

stabilization at lower dose rates than the treatments used alone which required silica gel 

at 95g/hl and PVPP at 20g/hl. Filtration characteristics on the kieselguhr filter were also 

excellent and better than hydrogels used alone. 

The relative costs of treatments to remove protein and polyphenols from beer vary 

over time. Systems dependent on single use PVPP have high revenue costs. But if small 

volumes are being treated then the capital cost of recovery systems is avoided. Silica 

xerogels were expensive but recent developments have seen prices falling. The most 

revenue cost effective treatment is probably gallotannin treatment combined with 

recoverable PVPP, but this will almost certainly require the use of acentrifuge to process 

tank bottoms and there is considerable capital cost in the recovery system. In the final 

assessment of which treatment to use, cost will be only one factor. Brewers will rely on 

their experience of the ways their beers behave to achieve the maximum stability 

required. 

15.3.5 Hazes from other than protein or polyphenols 

Haze in beer can sometimes be caused by carbohydrate or bacterial cell wall material or 

calcium oxalate. Attempts to remove these hazes at the maturation stage are usually 

failures. These hazes do not respond to stabilization treatments appropriate for proteins/ 

polyphenols. Beer hazes derived from carbohydrates or oxalate certainly can be 

controlled or eliminated by attention to raw materials or the conditions of mashing. 

Sufficient calcium must be present in the wort to precipitate oxalate from the malt or 

other grist materials. 

Theachievementofoptimumstarchconversiondependsonadequateamylaselevelsin 

the maltbut moreespecially onthe attainment ofgelatinization temperatures inthe mash. 

This should result in no problems with -glucan hazes in beer. Alow-temperature stand 

in the mash (

contamination during germination. This should be avoided with proper specification and 

dealing with reputable maltsters. All these aspects are discussed in other parts of this 

book. 

15.4 Carbonation 

Carbon dioxide is avery important constituent of beer. It imparts sparkle and `mouth 

feel' and sharpness associated with its properties as an acid gas. The concentration of 

carbondioxide inbeerforsaleiscarefully controlledtoensurethatconsumers ofthebeer 

can drink aconsistent product. Beers that lack carbon dioxide, particularly lager beers, 

are dull and lifeless and are said to lack condition and be flat. The carbon dioxide is the 

gas produced naturally in primary and secondary fermentation and that added to the beer 

by `carbonation'. 

At the end of primary fermentation the concentration of carbon dioxide in beer can 

vary from about 2g/l (1 vol/vol) in ashallow ale fermenting vessel up to 5g/l (2.5 vol/ 

vol) in adeep cylindroconical vessel. During secondary fermentation the level of carbon 

dioxide will increase. If acontrolled secondary fermentation is not carried out then 

carbon dioxide will normally have to be added to the beer. Such is the control now 

demanded for the precise level of carbon dioxide in finished beer that the gas levels are 

frequently adjusted after primary and/or secondary fermentation. The amount of carbon 

dioxide that will dissolve in beer depends on temperature and pressure. 

15.4.1 Carbon dioxide saturation 

Enclosed fermenters and maturation vessels are fitted with automatic gas pressure 

regulators that can be set at apredetermined pressure. Assuming that the gas above the 

beer is pure carbon dioxide then the gas that will dissolve in the beer at equilibrium is 

shown in Fig. 15.4. It can be seen that at agiven temperature and pressure aparticular 

equilibrium condition willbe reached. Increasing pressure willlead toalinear increasein 

the weight of carbon dioxide dissolving in the beer. An increase in temperature will give 

anon-linear decrease in the amount of gas dissolved (Fig. 15.5). These factors derive 

from Henry's Law, which states that the concentration of gas in the liquid phase is equal 

to the imposed pressure on the gas divided by Henry's constant that is temperature 

dependent. This is of fundamental importance in practical brewing and it must be 

remembered that as temperature rises substantially less carbon dioxide will dissolve in 

beer. It is therefore extremely important to keep beer as cold as is practicable after 

fermentation to keep carbon dioxide in solution and avoid the necessity for excessive 

artificial carbonation. It is normally the convention in brewing to state the concentration 

of carbon dioxide as volumes of gas at standard temperature and pressure per volume of 

beer. A gram molecule of a perfect gas occupies 22.4 litres at STP and so one volume of 

carbon dioxide is equivalent to 0.196% carbon dioxide by weight. 

At constant temperature and pressure the amount of carbon dioxide dissolving in beer 

will be a function of the time of contact between beer and gas and will decrease 

exponentially as equilibrium is approached. Increasing the surface of the beer exposed to 

gas can increase the rate of dissolution and a very shallow layer of beer will approach 

saturation more quickly. Further, if bubbles are created then a very large surface area is 

presented for gas transfer into the surrounding beer. In deep tanks, hydrostatic pressure 

also affects the concentration of carbon dioxide in beer (about 0.5 vol. increase/metre 


Pressure (psig) 

60 

50 

40 

30 

20 

10 

0 

Pressure (bar) 

4 

3 

2 

1 

0 

0 5 10 15 20 25 30 


35 40 45 50 55 60 65 70 75 80 85 


Fig. 15.4 Relationship between equilibrium values for dissolved carbon dioxide, temperature and 

pressure (Hough et al., 1982). 

Pressure (psig) 

80 

60 

40 

20 

10 

Pressure (bar) 

5 

4 

3 

2 

1 

depth) and the concentration at the base of the tank is higher than at the surface. 

Convection currents in the beer affect this gradation. 

Beer is capable of holding carbon dioxide in a supersaturated state, so rapid release of 

pressure or increase in temperature does not immediately lead to attainment of the 

appropriate equilibrium. This is advantageous when serving highly carbonated beers (say 

3 vol. CO 2) from bottles. When this beer is poured into a glass, the gas does not normally 

effervesce uncontrollably (`gushes') but releases carbon dioxide slowly. Knowledge of 

3.0 

2.5 

2.0 

1.8 

1.4 

1.0 

4 vol. 3 vol. 

Carbon dioxide content (volumes) 

10 20 30 40 50 


40 50 60 70 80 90 100 110 120 


2 vol. 

1 vol. 

Fig. 15.5 Effect of temperature on pressure for fixed contents of carbon dioxide. Note that an 

increase in temperature gives a non-linear decrease in the amount of gas that dissolves, viz., say at 3 

bar the equilibrium occurs for 4 vol. at 15 ëC, 3 vol. at 26 ëC and for 2 vol. at 41 ëC (Hough et al., 

1982). 


0.6 

0.4 

0.2 

Carbon dioxide content (% w/v)

Henry's Law thus provides the understanding for controlling the carbon dioxide content 

of the beer prior to filtration and subsequent packaging. For standard lager beers it is 

normal to aim for around 5g/l (2.5 vol. or 0.5%) of carbon dioxide in the beer. The 

pressure release valve will normally be set to give aslightly higher carbon dioxide 

content to compensate for the subsequent losses in filtration and packaging. 

Inevitably, some carbon dioxide produced during secondary fermentation will escape. 

This escaping gas is an important factor in beer maturation because it exerts acleansing 

effect by purging volatile fermentation products and so accelerating the flavour 

improvement process. 

15.4.2 Carbon dioxide addition 

All processes after secondary fermentation should be designed to keep carbon dioxide in 

solution in the beer. Thus beer should be kept cold and under the appropriate pressure of 

carbon dioxide to prevent gas release. However, situations will arise when temperature 

will rise or pressure will fall and gas will escape. The carbon dioxide must then be 

replaced in the beer before packaging. This is frequently achieved after filtration during 

transit to the bright beer tanks. 

This addition of carbon dioxide can occur whilst chilling beer through aplate heat 

exchanger (Fig. 15.6) and so can take advantage of the turbulence of the beer in creating 

good conditions for gaseous exchange. Apurpose-designed carbonation unit (Fig. 15.7) 

can also be used. This consists of a long pipe usually in the form of U-tube bends through 

which the beer flows. Carbon dioxide is injected as fine bubbles and the uptake, even in 

this form, can take a considerable time. The carbon dioxide must be the purest form 

available and no oxygen must be introduced. The injection unit must be easy to clean and 

must be cleaned regularly. Carbon dioxide can also be added ìn-vessel' but this is 

frequently less efficient and more difficult to control. A `carbonation stone' is sometimes 

used to ensure production of fine bubbles of carbon dioxide to aid dissolution in the beer. 

This technique is sometimes described as `gas washing' and provides an opportunity for 

the removal of oxygen and unwanted flavour volatiles as well as carbonation. After 

`washing' the vessel must be sealed to allow pressure build-up and the dissolution of 

carbon dioxide. There are a number of problems associated with externally added carbon 

dioxide and it is good practice to avoid this technique as far as possible. 

Care must also be taken not to dissolve more carbon dioxide in the beer than the 

specification allows. Reducing carbon dioxide levels is difficult without creating froth or 

fob. This reduction can be achieved by gas washing with careful bubbling of oxygen-free 

Conditioning 

tank 

Coolant 

CO 2 

Plate heat 

exchanger 

chiller 

Cold tank 

Fig. 15.6 Carbonation and chilling of beer òn the run', from conditioning tank to cold tank 



a 

Beer 

b 

e 

c 

nitrogen gas through the beer. Nevertheless, this technique is often a source of 

reintroduction of oxygen to beer with adverse effects on haze and flavour stability and 

loss of foaming potential. 

It mustnever be assumed that over carbonation of abeer reduces the uptake of oxygen 

if the beer is exposed to air. This uptake will be afunction of the difference in partial 

pressures between the beer and the air in contact with it and is not influenced by the 

carbon dioxide content of the beer. Turbulence in movement should be avoided for beer 

which has been carbonated, either naturally or by external addition. Turbulence will lead 

to the loss or `break-out' of carbon dioxide, the potential pick-up of oxygen and loss of 

foam potential because of protein denaturation at the gas/liquid interfaces. 

15.4.3 Carbon dioxide recovery 

The cost of carbon dioxide gas varies but it is generally expensive and certainly costs 

more than nitrogen. Before contemplating the economics of carbon dioxide recovery in a 

brewery the use of nitrogen gas as an alternative tocarbon dioxide for ensuring anaerobic 

handling of beer should be maximized. On this basis the amount of carbon dioxide 

required in breweries varies considerably from about 1.3 to 2.0kg of carbon dioxide/hl of 

beer produced. Of course carbon dioxide is essential for many purposes and there is the 

opportunity to collect the gas from enclosed vessels during fermentation. An essential 

aspect of recovery systems is that the gas must be purified if it is to be added to beer. 

Carbon dioxide is normally collected through afob tank into an inflatable balloon 

(Fig. 15.8), and from there it passes through water scrubbers and carbon purifiers before 

being compressed at the liquefying pressure of 18±22 bar (265 to 320lb./in. 2 ).This 

creates considerable heat and the compressed gas must then be cooled and so liquefied. 

The gas is dried through alumina driers and then stored as liquid until required for use 

when it can be restored to the gaseous state through an evaporator. Up to 2.0kg of carbon 

dioxide per hl of beer can be recovered by this technique and this could represent the 

whole of abrewery requirement. However, recovery is often much less than this as a 

resultoflossesandcleaningofthegas.Manybrewersexpectrecoverysystemstoprovide 

around 60% of requirement. 

d 

Beer 

f CO 2 

Fig. 15.7 Carbonation unit; (a) venturi nozzle, (b)carbon dioxide control valve,(c) CIP valve, (d) 

carbon dioxide sensor, (e) carbon dioxide dissolving area, (f) carbon dioxide measuring device 

(Kunze, 1999). 


FV 

Fob tank 

CO 2 compressor 

control switch 

Steam 

CO 2 

evaporator 

Condensate 

CO 2 gas 

Balloon 

Liquid CO 2 

tank 

Excess 

pressure 

device 0–3 bar 

Ammonia 

compressor 

and cooler 

Booster 

compressor 

15°C 

Pump 

Water cooler 

Twin activated 

alumina driers 

Water 

scrubber 

17 bar 

Water 

Compressor 


intercooler 

Fig. 15.8 Equipment for collecting, purifying, storing, and releasing carbon dioxide (Hough et al., 1982). 


Twin activated 

carbon purifiers

15.5 Clarification and filtration 

Clarification of beer involves the removal of yeast and the sedimented protein and 

polyphenol haze material derived from the beer stabilization techniques and cold break 

(Section15.3). Beerflavour is considerably morestablewhenit containssuspended yeast 

as the yeast promotes strongly reducing conditions. Total removal of yeast should 

therefore be delayed to the last possible moment before packaging. In cask beer yeast is 

never totally removed and so this beer is particularly robust with respect to flavour 

stability. There is, however, something of adilemma here because to achieve satisfactory 

filtration, beer going to the filter must contain

days from the end of primary fermentation. As noted, considerable effort has been 

directed to accelerating the flavour maturation of beer and the development of haze 

stability. The separation of yeast cannot be allowed to become rate limiting. 

Considering Stokes' Law from the viewpoint of sedimentation of yeast in beer, it 

would be difficult to lower the viscosity of the solution without a detrimental effect on 

beer quality. Likewise to create a marked difference between the density of the particles 

and the surrounding beer could alter the path of flavour maturation. An obvious method 

to improve the rate of sedimentation would be to increase the value of the factor 2r 2 , i.e., 

to increase the size of the yeast cells by agglomeration, i.e., causing them to clump 

together. This can be achieved by the use of isinglass finings. Isinglass finings are 

prepared from the dried swim bladders of certain fish (often sturgeon), which live in the 

estuaries of tropical rivers such as the Mekong and the Amazon (Leach and Barrett, 

1967). To cope with large changes in water density as a result of experiencing fresh and 

salt water conditions alternately these fish develop very large bladders to adjust their 

buoyancy. These bladders are composed of almost pure collagen and they are one of the 

purest forms of protein found in nature. An isinglass solution is obtained by soaking the 

dried bladders in dilute solutions of cold tartaric, sulphurous and sometimes malic acid 

for periods up to six weeks (a process known as cutting). A turbid, colourless, viscous 

solution results, which contains soluble collagen, gelatine, (the denaturation product of 

collagen) and some insoluble material. Only the soluble collagen contributes to fining 

action and methods are available for determining the collagen content of fining solutions 

(Leach and Barrett, 1967). 

Effective fining is achieved as a result of the positive charge on the collagen molecule 

that exists as a triple helix of complex stereochemistry rich in basic amino acids. Yeast 

cells are thought to flocculate as a result of surface proteins (lectins) bridging with 

carbohydrate side chains (mannan) on neighbouring cell walls (Shiel, 1999). It has been 

suggested that isinglass functions as an extracellular lectin with the carbohydrate moieties 

on yeast cell walls bearing a negative charge bridging to -NH3 + sites on the collagen 

molecule. In this way yeast cells will clump together and so the factor 2r 2 in Stokes' Law 

will be enhanced with a corresponding increase in sedimentation rate. Breweries can use 

isinglass either as a liquid purchased from the manufacturer or as shredded swim bladders 

ready for cutting, with acid, in the brewery. 

Larger molecules of soluble collagen are likely to be more effective at fining than 

smaller molecules. Large molecules will have more charged sites per molecule and 

therefore will react more readily with and cross-link the negatively charged yeast cells. 

Finings from different sources do not differ much with respect to total charge or charge 

distribution but the constituent collagen molecules do differ in size and size distribution. 

The size distribution is a function of the ease with which collagen is broken down by acid 

and the extent to which aggregates of molecules are held together by weak physical and 

chemical forces. 

The collagen triple helix forms a long rigid rod of 1.5 nm diameter and the individual 

chains exhibit frequent sharp twists owing to the presence of adjacent imino acids, 

proline and hydroxyproline. As a result of the high hydroxyl group content there is strong 

hydrogen bonding between the three chains. Young tissues contain monomers of 300 nm 

length and 300,000 molecular weight. 

The molecular size of the soluble collagen can be estimated by measuring the intrinsic 

viscosity (Leach and Barrett, 1967): 

‰ Š ˆ KM 


Table 15.1 Comparison of fining properties of isinglass from different sources (proportion of 

effective finings derived from isinglass ˆ A B=100%) 

(Leach and Barrett, 1967) 

Type of leaf Form Soluble Soluble Intrinsic 

nitrogen (%) collagen (%) viscosity [ ] 

A B (dl/g) 

Karachi Flock 80 86 15.4 

Karachi Shredded 78 96 21.9 

Brazil Lump Flock 81 97 18.2 

Brazil Lump Shredded 84 93 18.7 

Long Saigon Flock 62 87 16.5 

Long Saigon Shredded 57 92 2.6 

Round Saigon Shredded 97 94 26.5 

Penang Shredded 20 98 16.8 

where Kis aconstant depending on the particular solute/solvent system, Mis the average 

molecular weight and is aconstant dependent on molecular shape and rigidity, for 

isinglass is approximately 2 

\ p ˆKM 

Intrinsic viscosities are independent of concentration and are high for finings 

compared to proteins in general. If beer is easy to fine then in citrate solution [ ]should 

be >16 and it is found that different types of isinglass show little difference in fining 

action. Beer that is more difficult to fine needs finings where [ ]is >20 and then 

differences are manifest between different forms of isinglass (Table 15.1). Collagen 

rapidly denatures to inactive gelatine when the temperature rises and fining solutions 

must be stored at temperatures of 4±10ëC (39±50ëF). Finings can be used effectively at 

temperatures up to 14ëC (57ëF) and work best when the temperature of the beer is rising 

slightly. This particularly applies to the use in cask beer when the finings are normally 

added at racking prior to despatch from the brewery (Chapter 21). 

The use of isinglass finings in the preparation of brewery conditioned beer is not 

universal. It is largely a UK practice deriving from its effectiveness for cask beer. There 

is, however, renewed interest in the USA and the use of finings is increasing in Australia 

and Africa. Isinglass finings improve foam stability of beer by removing lipid material 

such as fatty acids and phospholipids, which are foam negative. This is an important 

secondary characteristic. The use of isinglass along with silica hydrogel has also recently 

been described (Shiel, 1999) and improvements in filter runs (by as much as threefold) 

and actual beer shelf-life were demonstrated. This seems likely to be as a result of the 

huge removal by isinglass of particles < 4.5 m in diameter that will cause blockage of the 

filter and have the potential to cause haze in the beer. 

Sedimentation of yeast in the presence of isinglass is rapid and compact easily 

removable tank bottoms are formed. This may require further processing by 

centrifugation or filtration to recover entrapped beer and finally separate the yeast if 

this is deemed economically sensible. It is clearly established (Shiel, 1999) that isinglass 

is completely removed during the brewing process and is not detectable in finished beer. 

Centrifugation 

The other powerful component of Stokes' Law is the effect of the acceleration owing to 

gravity. Centrifuges increase the gravitational force and so will allow more rapid 


separation and removal of the yeast and other particles. Centrifuges are often used where 

isinglass finings are not used or they are used in conjunction with isinglass to separate 

yeast and beer in tank bottoms. A centrifuge operating at 5,000 rpm can effect an 8,000 

increase in the gravitational force. The general relationship being: 

g ˆ ! 2 r 

where g is the gravitational force, ! is the rotational speed in radians/second and r is the 

radius in metres. 

Sedimentation of the particles is also enhanced if the settling distance is reduced and 

this is an important feature of centrifuges. Centrifuges were developed from 

considerations of the behaviour of solids and liquids, and liquids of different densities 

in balanced tanks. If such a balanced tank is rotated about its vertical axis centrifugal 

force supplements the acceleration owing to gravity. Discs can be inserted in the 

centrifuge to reduce the path distance of sedimentation. Two main types of centrifuge are 

now used in breweries for the separation of yeast from beer prior to filtration, selfcleaning 

clarifiers and decanter clarifiers. Self-cleaning clarifiers can operate at relatively 

low solids contents to process complete tanks of beer, say between fermentation and 

maturation, when other means of yeast separation such as isinglass finings and 

sedimentation are not available or are not fast enough. They do not operate well on tank 

bottoms where the solids content can be very high. In this situation the decanter machine 

would be used. 

· Self-cleaning clarifiers. In this type of centrifuge the solids are discharged at intervals 

whilst the centrifuge is operating at full speed (Fig. 15.9). The bowl contains discs 

separated by spaces of 0.5±2.0 mm (0.02 to 0.08 in.) by distance pieces known as 

`chaulks'. The yeast and other solids slide on the disc surface to the periphery and collect 

on the surface of the bowl. Clarified beer moves towards the centre of the machine and is 

pumped to the next processing stage, which is normally the maturation vessel when the 

centrifuge is inserted in the line between fermentation and maturation vessels. 

Depending on the solids load the rate of operation can vary from 40 to 600 hl/h. Yeast 

is removed by a mechanism whereby the spinning bowl separates momentarily into two 

parts at the rim and the solids are ejected. This ejection can be on a simple predetermined 

time basis, irrespective of the presence of sludge, in which case beer losses 

can be high. The self-sensing type of machine uses a hydraulic differential pressure 

Solids 

Feed 

Clarified liquid 

Solids 

Fig. 15.9 Self-cleaning clarifier centrifuge (Hough, et al., 1982). 


Rotor 

drive 

Outlet for 

liquid 

Scroll 

Outlet for 

solids 

mechanism for sensing the accumulation of solids in the bowl. These machines offer 

better control of beer losses but must be constantly maintained. A third system works by 

optically monitoring the clarity of the beer leaving the machine. 

· Decanter clarifiers. If the solid content of the beer to be processed is very high, as 

can be the case in tank bottoms (up to 60% by volume), a decanter centrifuge can be 

used (Fig. 15.10). These machines employ a rotating screw in a casing with discharge 

of solids at one end and processed beer at the other. They normally operate at speeds 

of 40 hl/h. 

Centrifuges have a number of advantages for yeast removal and clarification of beer. 

They have a small space requirement and can be sterilized and maintained sterile. They 

can be hermetically sealed and this is essential to exclude oxygen. Fitting seals achieves 

this where the rotating parts of the machine adjoin the stationary parts. Centrifuges have 

no requirement for filter aids and no active adsorption is involved. If a constant solids 

load is presented it is possible to operate a centrifuge continuously for an indefinite 

period. Yeast separated by centrifugation varies between 13 and 25% dry matter. 

Centrifuges have a very high requirement for electrical energy. There is a high inertia 

of the rotating parts and when slurry is discharged an equivalent volume of beer entering 

the machine is brought to rotational speed in a very short time. Motors have to be sized to 

meet these maximum loads, which on a 200 hl/h centrifuge would be 22.5 kW. The 

normal running rate of a centrifuge will consume about 0.35 MJ of energy per hl of beer 

processed. Centrifuges are noisy, frequently exceeding 85 dB(A) and so ear protection 

must be worn when inspecting or working on them. This sometimes deters maintenance 

and inspection. The machines are complex and difficult to maintain and spares are costly. 

These are real disadvantages in a modern brewery operating at low fixed cost. A further 

problem is the rise in temperature of the beer and yeast, which occurs during 

centrifugation (up to 3 ëC, 5 ëF). This can lead to a loss in carbon dioxide and a need to rechill 

the beer. The physiological condition of the yeast is adversely affected by shear 

forces and the temperature rise and yeast collected by centrifugation is far less suitable 

for re-pitching than that collected by natural sedimentation. 

Filtration 

Yeast can be separated from beer and beer recovered by various types of filtration: the 

yeast press, rotary vacuum filtration and cross-flow filtration. Filtration techniques are 

normally used for the processing of tank bottoms. Tank bottom beer (sometimes called 

`barm ale') may represent 1±3% of the total volume of the beer in the tank. It is thus a 

high process loss to discard the whole of this beer. This is often not of such good quality 

Feed 

Fig 15.10 Decanter centrifuge (based on drawing received from Alfa-Laval Ltd.). 


as the bulk of the beer in the tank. It usually has ahigher pH value and contains more 

amino nitrogen and higher alcohols. Recovered beer must be blended at some stage with 

primary tank beer. This should be carried out as early as is possible in the process as the 

recovered beer will contain yeast autolysate. 

The point of addition will vary from brewery to brewery depending on the available 

equipment and the quantity of beer. A good point of addition is the beginning of 

fermentation when the actively fermenting pitching yeast will quickly absorb the 

metabolic products of the autolysed yeast in the recovered beer with no deleterious effect 

onoverall beer quality. Earlier additions intothe brewhousearealsopossible.The capital 

cost of making these additions and indeed recovering the yeast may be high and the 

overall economics of the process must be assessed when deciding on the optimum 

method inaparticularbrewery.Inmanypartsofthedeveloped worldthecostsoftreating 

effluent containing yeast and beer are so high that it will be essential from afinancial 

viewpoint to separate yeast and recover beer from tank bottoms before discharging 

effluent to drain (Chapter 3). 

· Yeast press. In traditional ale brewing in the UK it was common practice to press 

yeast after skimming from the fermenter to separate it from barm beer. The pressed 

yeast was stored as adry cake in arefrigerator or was slurried in chilled water until 

reused. This was agood way to recover barm beer for addition back to the process and 

the pressed waste yeast in aconvenient form to sell for the manufacture of yeast 

extracts or animal feed. Yeast was also sold to the distilling industry in this form. 

Yeast presses are made of gun metal and contain fabric cloths. Two cloths form a 

chamber on either side of a backing plate. The yeast slurry is pumped into the 

chambers and the beer recovered from channels in the backing plate. Presses are 

labour intensive and slow, requiring manual cleaning of cloths and two-man operation 

for yeast removal. In many breweries the traditional press has been discarded and the 

yeast is used and stored as barm (i.e. as aslurry in beer). 

There have been recent developments in yeast press design demanding a 

reconsideration of this technique for tank bottom processing. Inflatable diaphragm 

plates made of polypropylene have been introduced which can squeeze the cake and 

considerably shorten the time of pressing. New presses can be in-place cleaned at 

80ëC (175ëF) and can be operated by one man. Afully automated press has also been 

described (Anderson et al., 2000) in which astack of vertical plates is used with a 

continuous polypropylene cloth driven by power rollers fed around each of the plates. 

In the modern press the pressure applied rises from 4to 19 bar (60±280lb./in. 2 )as 

the chamber is filled and kieselguhr or perlite is sometimes added to improve 

filterability. Therecoveredbeer cancontain0.1to0.5 10 6 yeast cells/ml ofbeer and 

the concentration of the pressed yeast is between 25 and 40% dry weight. The market 

value of pressed yeast varies but it can normally be sold at aprofit. The beer is bitter 

and yeasty andcontains high levelsofamino acids andpolyphenols. It can be added at 

the start of fermentation but in some breweries it is pasteurized before addition. If 

possible the beer should be added after wort boiling but before wort cooling and 

pitching. Experience shows that beer from ayeast press should not be added at a 

greater rate than 5%. 

· Rotary vacuum filter (Fig. 15.11). This machine can operate continuously and 

consists of a rotating drum covered by a filter sheet on which is deposited a pre-coat of 

perlite or kieselguhr. Beer is sucked into the drum and yeast collects as a layer on the 

surface, which is then removed by knives to a tank underneath the filter. The 


Outlet 

Part of tank 

receiving slurry 

Pipes collecting filtrate 

Precoat 

Septum 

Septum support 

concentration of the recovered yeast is usually 10 to 25% dry matter. The beer has 

similar properties, although it contains more oxygen, than that from the yeast press 

and can be returned to the process in asimilar way. 

· Cross-flow filter. In this system the yeast/beer mixture is circulated under pressure 

over amembrane. For the system to be successful shear at the membrane surface is 

necessary to prevent fouling and amechanism for the removal of solids is essential. 

Shear can be created by high volume flow at the membrane surface, but these systems 

have not always been commercially successful because of the high energy input 

needed and the high pumping rate leading to yeast cell damage and consequent 

adverse flavour effects in the beer. A device has been described (Chalmers and 

Haughney, 1998) in which the shear at the membrane surface is created mechanically 

by vibrational energy. The basis of the design is the use of atorsional spring mass 

system for the creation of vibrational shear at the membrane surface (Fig. 15.12). A 

motor rotatesashaft with an eccentric weight. Themotion ofthiscausesthe motionof 

asecond weight, which has arelatively high mass moment of inertia, this is called the 

seismic weight. The motion ofthe seismic weight is translated through atorsion bar or 

spring to a membrane element assembly. The element assembly vibrates at a 

frequency of 50±60Hz and the energy requirement is 0.03 to 0.15kW/m 2 . 

The membrane assembly can contain up to 40m 2 ofmembrane area and it is made 

to operate three streams: feed, permeate (or filtrate), and retentate. The element 

assembly has atop and bottom end plate and membrane elements with spacers on the 

inner and outer diameter. The spacers provide an open channel for the distribution of 

thefeedfluid(beercontainingyeast, e.g.,tankbottoms)acrossthemembranefrom the 

outer diameter to the inner diameter. As this happens the filtrate (beer) passes through 

the membrane and is directed to acentre channel. Fluid that does not pass through the 

membrane (yeast) is concentrated and leaves the membrane element assembly as the 

retentate. 

Membrane quality is very important for the success of the system. Membranes have 

been made from ceramic tubes and hollow fibres or flat sheet polymers. Initial trial 

work with the vibrating membrane filter used aPTFE membrane, each element of 

which was 0.4 m 2 ;aten element system therefore presented 4m 2 ofseparating 

membrane. Beer/yeast slurries of up to 80% solids (by volume) were processed. For 

tank bottoms where the solids content was >60% by volume, dilution with water 

improved performance. Loss of carbon dioxide can be prevented by pressurizing the 

systemandoperating withabackpressureonthepermeateof1.0to1.4bar(15±20lb./ 

in. 2 ).Yeast in lager and ale tank bottoms was concentrated to adry solids content of 

20% (80% by volume). Pressure drop across the membrane was not significant and 

Knife 

Fig. 15.11 Rotary vacuum filter (Hough et al., 1982). 


Divided tank 

(portion under 

knife receives 

solids)

Permeate port 

Motor mounting 

plate 

Drive shaft 

enclosure 

Motor 

Retenate ports 

Filter pack 

Tubing support 

Torsion bar 

Node clamp 

Shipping clamp 

Fig. 15.12 Vibrating membrane filter (Chalmers, and Haughney, 1998). 

successful permeate flows were obtained at pressures of 0.8 bar (12 lb./in. 2 ). 

Commercial systems have now been developed that can operate at average flow 

rates of up to 2,500 l/h which could serve a brewery producing 4,000,000 hl of beer a 

year of which 6% was barm (tank bottoms). Cleaning the membrane assemblies can be 

effected with caustic soda solutions (0.2 M at 60 ëC; 140 ëF). The process has an 

extremely low energy requirement of 0.004 kWh/l of recovered beer. This factor, 

coupled with the compactness of the machine, the ease of automation and low 

manpower input makes the use of a vibrating membrane filter an attractive option for 

separating yeast and beer in barm and tank bottoms. 

15.5.2 Beer filtration 

The final process to consider in beer treatment, prior to packaging, is filtration. This is the 

clarification of the beer to a standard that is acceptable for sale. The process involves the 

removal of any remaining yeast cells and the removal of precipitated protein and 

polyphenol haze material. The beer must be rendered stable so that visible changes do not 

occur during its commercial (shelf) life, which could be up to 52 weeks from the date of 

packaging. To be successful, beer coming onto the filter must contain < 0.2 million yeast 

cells/ml of beer and so the processes discussed above are crucial for the success of the 

final beer filtration. 

The driving force for filtration is the pressure difference between the filter inlet and 

outlet. Pressure is always greater at the inlet and the pressure difference is an indicator of 

how much the filter is resisting filtration. An increase in this pressure difference indicates 

the approach of the end of a filter run. In commercial brewery practice this is an 

important factor and long filter runs are essential to overall brewery efficiency. 


Avery important factor in successful filtration is the chilling of the beer. The lower 

the temperature the more cold trub and chill haze will form. Filtration will remove this 

material, provided that the beer temperature does not rise in the filter itself. Beers 

emerging from maturation must therefore be maintained at 2 to 1ëC (28±30ëF) 

through filtration. The turbidity (see Chapter 19) of the beer leaving the filter must be 

< 0.5 ëEBC. In good filter practice losses of colour, extract, bitterness and foam potential 

should be minimal. 

Care must also be taken to avoid oxygen pick-up after maturation and during the 

filtration process. At the end of maturation the oxygen concentration should be 

< 0.01 mg/l. This concentration can be maintained with sound practice. Deaerated/ 

deoxygenated water must be used for introducing powder pre-coats on powder filters and 

carbon dioxide is used as the counter pressure gas to move beer from maturation through 

to the filtration process. Air must be completely removed from all pipework before the 

passage of beer. 

As with beer sedimentation there have been attempts to define filtration in 

mathematical terms following empirical studies. It was found in 1856 that: 

Q ˆ ' PA 

LM 

where Q is the flow rate in ml/second, ' is the permeability factor, P is the pressure 

differential in dynes/cm 2 , A is the area of the filter medium in cm 2 , M is the viscosity of 

the liquid in poises, and L is the thickness of the filter medium in cm. Filter throughput 

may therefore be increased by increasing ' which relates to the composition of the filter 

material. Alternatively, the applied pressure and filter area may be increased and the filter 

thickness and liquid viscosity decreased, this latter factor cannot be achieved in practice 

without increasing temperature, which could result in the re-solution of haze polymers. A 

high viscosity is sometimes indicative of the presence in solution of high molecular 

weight -glucan, -glucan or yeast polysaccahride which will adversely affect filtration 

and if this is the case the cause should be sought in examining malt quality and 

brewhouse performance. In considering any type of brewery filtration system the above 

equation should be studied to optimize performance. 

Different mechanisms of filtration can be used: 

· Sieving or surface filtration in which the particles are trapped in pores in the filter 

medium and retained in a layer. Filtration quality improves with time but the volume 

flow decreases continuously. 

· Depth filtration in which a separation medium, e.g., kieselguhr is used on a support 

and which causes the particles in the beer to take a very elongated route through a 

large surface area. The particles are retained by mechanical sieving because of size 

and will gradually block the pores in the medium and so reduce flow rate and the 

particles can also be retained by adsorption as a result of electrical charge effects. 

Traditionally, surface filtration in breweries was associated with sheet filters. 

However, adsorption can occur on some sheets as demonstrated when some yeasts and 

bacteria fail to penetrate sheets when the pore size should permit it, as the fibres hold the 

negatively charged micro-organisms electrostatically. Breweries frequently employed 

double pass filtration where the beer was filtered twice through discrete systems. This 

could be two sets of sheet filters of decreasing pore size or a depth filter followed by a 

sheet filter (called a polishing filter). These systems were sometimes associated with high 

beer losses and the addition of oxygen to the beer and there is increasing use of single 


pass filtration to prepare beer for packaging. Of course if sterile filtration is to be used 

prior to packaging then there will be aseparate filter for this purpose (Chapter 21). The 

two different mechanisms of filtration can be effected using sheets or by using cloths on 

which amedium (powder) is deposited to create the depth. Genuine surface filtration is 

also achieved with membranes (Chapter 21), which can be made of many polymers, e.g., 

polyamide, polyethylene, and polycarbonate. 

Sheet filtration 

The standard filter sheet in breweries is 60 62cm and the largest normally 100 

100cm. Alarge single-ended filter press would have 240 filter plates allowing afiltration 

rate of 120hl/h. This low throughput rate has limited the use of sheet filtration for 

primary filtration in large breweries. Sheet filters are now usually used only as second 

polishing filters following depth filtration with powders or are dispensed with 

completely. There have been developments in the design of filter plates to ensure the 

maximum surface area of filter sheet is presented to the beer (Fig. 15.13). Stainless steel 

plates are now usually used having corrugated inserts as supports. Flow rates of > 2.0 hl/ 

m 2 /h have been achieved. 

Filter sheets were originally made from a mixture of cellulose and asbestos fibres. 

Sheets have also incorporated kieselguhr for over 70 years. Recently perlite, glass fibres 

and cotton fibres have been used. Asbestos is not now used because of the carcinogenic 

properties of some types of asbestos. It is fair to comment that the elimination of the use 

of asbestos took a long time in many breweries. Asbestos was useful because it offered 

adsorption through electrical charge as well as surface filtration. This has been replaced 

in some applications by the incorporation into the sheet of aluminium oxide or zirconium 

oxide fibres. PVPP can also be incorporated into filter sheets to impart additional 

stabilization to the beer by removal of polyphenols (15.3). In large throughput systems 

washing the sheet with a solution of 0.5% sodium hydroxide regenerates the PVPP. 

Sheet filters generally operate at low flow rates. If attempts are made to increase the 

flow rate, pressure can force yeast and haze particles off the fibres and through the filter 

into the beer. For this reason pressure differentials not greater than 0.7 bar (10 lb./in. 2 ) 

are often maintained. When this pressure differential is exceeded then sheets are cleaned 

by back-washing or are replaced. Sheet filters are often sterilized with steam (0.6 bar; 

9 lb./in. 2 ). 

Sheet filtration does, therefore, have a number of drawbacks. The sheets cannot be 

regenerated indefinitely and so operating costs are high. The filter occupies a lot of space 

if substantial throughput is required and it is not easy to automate. It is also very sensitive 

to variable solids levels in the beer and will quickly block. As a primary beer filter the 

sheet filter has usually been replaced by the powder filter. 

Powder filtration 

The most successful and cost effective beer clarification is achieved by powder filtration. 

This was developed from the earlier pulp or mass filtration in which the mechanism is 

primarily depth filtration. Many breweries employ this system for single pass filtration 

and are able to deliver beers to packaging consistently at < 0.5 ëEBC haze units. In a 

powder filter the powder (filter aid) is coated onto a support and provides a tortuous path 

through which the beer passes giving many opportunities for the trapping and adsorption 

of particles. 

Two types of powder are commonly used: kieselguhr or perlite. 


(a) 

Beer out 

Beer in 

(b) 

Outlet pressure 

gauge 

Outlet sight 

glass and 

sampling valve 

Outlet and valve 

Valve for draining 

In-plates Out-plates 

Filter sheet 

Plate Sheet Plate 

· Kieselguhr is a diatomaceous earth, which is mined from Miocene period deposits in 

Europe and North and South America. It consists of skeletons of marine algñ 

containing silicon dioxide. Kieselguhr powders for use in brewing are prepared by 

drying and milling the mined raw material. Most effective filtration was achieved with 

the use of calcined kieselguhr prepared by heating the raw material in rotating drums 

at 600 to 800 ëC (1,100±1,450 ëF). However this substance is classified as highly 

dangerous when inhaled and can give rise to the disease of silicosis. Equipment is 

needed for automatic slitting of bags and transfer to slurry tanks to avoid manual 

handling. Uncalcined kieselguhr, prepared by drying at < 400 ëC (750 ëF) represents 

only a moderate risk and is now usually preferred. However, some uncalcined 

kieselguhr can contain traces of iron and other metals. Numerous grades of kieselguhr 

are produced from fine through to medium and to coarse. The finer the kieselguhr the 

better is the clarification but the speed of filtration is less. Coarser grades give a rapid 

flow rate but poorer clarification. Kieselguhr usage varies from 70 to 220 g/hl. It is an 

(c) 

Inlet pressure gauge Inlet sight glass 

and vent valve 

Venting 

valve 

Inlet and 

valve 

Tray Tray 

Drain 

(d) (e) 

Compression 

screw 

Valve for 

venting and 

entry of steam 

Recess for 

compression 

screw 

Drain valves 

Fig. 15.13 Details of a sheet filter; (a) vertical section, (b) single plate front view, (c) alignment of 

plates and sheets, (d) control end of machine, (e) compression end of machine (Hough et al., 1982). 


expensive product and it is also expensive to dispose of as slurry, usually to landfill 

sites where it is an unsatisfactory uncompacted infill. 

· Perlite is avolcanic material, mostly composed of aluminium silicate, obtained from 

Greek islands. Raw perlite is heated to about 750ëC (1,400ëF), which causes bursting 

of the particles yielding glassy structures. These are milled to afree flowing powder, 

which is about 30% lighter per unit volume than kieselguhr. Perlite represents alow 

risk to health but, because of its low density, it disperses easily in the air and creates 

nuisance dust. At low pH values (

(c) 

Direction 

of flow of 

liquor to 

be filtered 

3 Filter cake of 

removed impurity 

and filter aid 

particles 

Direction 

of flow of 

filtered 

liquor 

1 Filter cloth 

2 Precoat of filter 

aid particles 

Fig. 15.14 Principles of kieselguhr filtration (Hough et al., 1982). 

(a) 

(b) 

Filter sheet 

Plate Frame Plate Frame 

Pre-coat 

Bodyfeed 

Fig. 15.15 Plate and frame filter; (a) side view, (b) automatic mechanism for moving individual 

plates for opening and closing, (c) arrangement of plates and frames and the flow of beer. (Courtesy 

of Alfa-Laval Ltd.). 

into the void in the plate from which it discharges (Fig. 15.15). The sheets are 

washable and have a long life. After filtration the plates are separated and the 

kieselguhrisdislodgedfrom thesheets byspraying.Insomeplateandframefiltersthe 

sheet is protected by adisposable cellulosic `nappy liner', which does not add to 

pressure differentialbut doesprolong the life ofthe sheetand aids kieselguhrdisposal. 

· Leaf filters. Leaf filters (or screen filters) have aseries of stainless steel leaves fitted 

either vertically or horizontally inside afilter body (Fig. 15.16). In avertical leaf filter 

both sides of the support are coated with filter aid whereas in the horizontal leaf filter 


Beer in 

Beer out 

Vertical filter 

only the upper surface is used. The kieselguhr adheres to the stainless steel septa 

because of the pressure at which the beer is forced into the filter. Even coating of the 

supports is not easily achieved particularly with the vertical leaf filter in which the 

kieselguhr tends to slip downwards. Anew type of support for horizontal leaf filters 

hasbeendescribed(Oechsleetal.,2000).Thisisastainlesssteelmembraneof0.4mm 

gauge and an optimized slot width of 35 mand alength of 2mm. The slots widen 

towards the filtrate side, which reduces blocking. In trial work this support was 

effective in beer filtration without the use of an initial coarse kieselguhr pre-coat. This 

resulted in major savings in kieselguhr use as the first pre-coat can be up to 800gof 

kieselguhr/m 2 of filter. Time was also saved and longer filtration cycles were 

demonstrated with beer clarities better than the controls delivered from conventional 

pre-coat filtration. 

There is also the potential for health improvements with the elimination of the 

coarse kieselguhr pre-coat, which in some breweries is composed of calcined 

kieselguhr, which is the most risky for health. Increasingly breweries have to cope 

with the filtration of more different beer brands, some in very low volumes. This is a 

result of greater consumer pressure and the proliferation of licence brewing where 

some companies will brew another company's brands. Losses must be minimized and 

the flexibility and automation of the filter is very important. Single pass filtration in a 

horizontal leaf filter is effective in achieving these aims (Thilert, 1999). Cleaning leaf 

filters is effected by spraying the leaves and filter body and in the case of the 

horizontal filters spinning the leaves to deposit the spent kieselguhr into aholding 

tank. Sometimes the last remains of the kieselguhr must be removed by pressure 

spraying from abuilt-in spray bar as aresult of which this small amount of material 

goes down the drain. This volume must be minimized because entrapped beer will 

contribute to COD of the effluent and the powder will raise suspended solids, both 

important factors in the charging formulñ for effluent in some countries (Chapter 3). 

a 

e 

Horizontal filter 

Fig. 15.16 Vertical and horizontal leaf filters; horizontal filter: (a) beer inlet, (b) filter support, (c) 

perforated shaft, (d) filtered beer outlet, (e) sediment outlet (Courtesy of Alfa-Laval Ltd.). 


b 

c 

d

(b) 

(a) 

Division plate 

Inlet 

Sludge outlet 

Filter head 

Outlet 

Filter body 

Fig. 15.17 Candle filter or Metafilter; (a) candle filter, (b) detail of single candle (Hough et al., 

1982). 

· Candle filters. A candle filter is a cylindrical, vertical pressure vessel containing 

many filter elements (Fig. 15.17). Each element consists of a rod of Y cross-section 

around which annular discs are stacked. The discs are made so that liquid can 

penetrate between them and then flow along the channels between the holes in the 

discs and the recesses of the Y-section rod. Kieselguhr powder builds up between 

adjacent filter discs to provide the surface area for depth filtration. Between 500 and 

700 candles can be arranged in a cylindrical housing to create a very large filter. 

Views vary considerably on the relative advantages of different types of filter. It is 

difficult to draw firm conclusions but some principles emerge. To avoid losses and to 

avoid oxygen pick-up, single pass filtration should be used wherever possible. On this 

basis powder filtration is usually regarded as being superior to sheet filtration. Powder 

filters have been compared for a number of parameters (Harding, 1977). Flow rates 

through filters are usually about 5 hl/m 2 but slower rates will ensure more effective 

particle removal. Plate and frame filters are more easily pre-coated and less susceptible to 

pressure surges than leaf or candle filters but the greater volume of the filter means higher 

beer losses, an important factor. 

Sterilization of the filter, usually achieved with steam, is easier in candle filters and 

leaf filters than in plate and frame filters which have a large mass of metal. Kieselguhr 

removal is easiest with the new generation of horizontal leaf filters. (Thilert, 1999). These 

filters are well suited to low volume filtration for a series of different brands with 

minimal losses and oxygen pick-up. 

It is difficult to generalize on capital costs but candle filters are usually the cheapest 

and horizontal leaf filters the most expensive. The revenue cost varies considerably 


depending on the conditions of use. Clearly any system avoiding the use of afirst precoat 

(Oechsle et al., 2000) may offer alower cost advantage. 

15.6 Special beer treatments 

Many beers are brewed at ahigher gravity (or alcoholic strength) than that at which they 

are subsequently sold. These beers are diluted after fermentation and usually after 

filtration just prior to packaging (Section 15.2.4). Fermentation at high gravity was 

considered in Chapter 12. Some beers are sold as low-alcohol or no-alcohol products and 

the alcohol is removed after fermentation. Low-alcohol beers can also be produced by 

changes to the mashing process or to fermentation and all types will be discussed in this 

section. There has also been interest in the sale of beers produced by freezing to remove 

water and yield a beer with a smooth flavour and elevated alcohol content. These are 

known as ìce beers'. A further `special' type of beer is the so-called diet beer, which has 

a very low dextrin level (but often high alcohol), and although this is not strictly a postfermentation 

`treatment beer' it will be reviewed in this section. 

15.6.1 Low-alcohol and alcohol-free beers 

The production of alcohol-free and low-alcohol beers has a long history and patents on 

the processes involved go back over 100 years. Marketing and sale of these beers has 

varied in intensity throughout this period. Low-alcohol beers were produced in 

considerable volume at the time of the First and Second World Wars as a result of the 

shortage of raw materials and prohibition in the USA from 1919 to 1933 stimulated 

production. There has been renewed interest, since about 1978, because of legislation 

relating to the driving of motor vehicles and health considerations leading to some beliefs 

in the advantages of drinking less alcohol. There is also a trade in the export of nonalcoholic 

beer to Islamic countries where the sale of alcohol is banned. These situations 

lead to the development of a healthy market and most major brewers included lowalcohol 

and alcohol-free beers in their product portfolios. The market for these beers has 

recently come under pressure both from aggressive marketing from soft drink companies 

and from so-called àlcopops' in which alcohol is mixed with some type of fruit extract. 

This has led to brand losses and there are now fewer types of beer available. However the 

competition has resulted in marked flavour improvements in those brands which have 

survived. In 1992, in Europe, the market for low-alcohol beer was 4% of the total 

alcoholic drinks market but it has shown no growth since this time. 

Legal definitions of what constitute low-alcohol and alcohol-free beer varies from 

country to country. A low-alcohol definition will allow an alcohol content of 0.5 to 1.2% 

v/v whereas an alcohol-free beer should contain < 0.5% v/v alcohol. Production methods 

for these beers involve either the removal of alcohol in a post-fermentation treatment or 

the restriction of alcohol production during the brewing process. Removal of alcohol can 

be effected by vacuum distillation, vacuum evaporation, dialysis and reverse osmosis. 

Restriction of alcohol production can involve choice of grist materials, control of 

mashing schedules, stopped fermentation, and use of special yeasts. 

Vacuum distillation 

In this process (Regan, 1990), the beer to be de-alcoholized is heated to 50 ëC (122 ëF) in a 

plate heat exchanger and is then de-esterified under high vacuum. Volatile components 


evaporate from the beer and are collected in a mixing tank. The de-esterified beer is then 

separated from alcohol in a vacuum column at about 40 ëC (104 ëF) and passed to the 

mixing tank where it is recombined with the volatile components. This method has 

produced beers with sound flavours but is now seldom used because the high temperatures 

involved do tend to make the consistent production of high-quality beer difficult. 

Vacuum evaporation 

This process (Attenborough, 1988; Narziss et al., 1992; Regan, 1990) has developed from 

considering the difficulties of vacuum distillation. The temperatures used are lower than 

with vacuum distillation and the residence time under evaporation is less. It is easier to 

produce de-alcoholized beers of consistent quality with this method. A process described 

by Alfa-Laval involves centrifugal evaporation. Beer is pumped over the internal surface 

of a rotating conical heat exchanger and forms a film (0.1 mm) over this surface. The beer 

remains in this position for 0.5 to 1.0 seconds and reaches a temperature of 30±40 ëC (85± 

104 ëF). The concentrated, de-alcoholized beer is collected at the periphery of the cone 

where it is drawn by suction to a cooler. The aroma compounds are retained in the beer by 

this technique and this is the real advantage compared to the vacuum distillation method. 

The alcohol evaporates and the vapour passes through the centre of the cone to a 

condenser. The process, taking only about ten seconds, is repeated several times if it is 

required to lower the alcohol content to < 0.5%. 

An APV system utilizes a triple effect falling film evaporator. The maximum temperature 

of the beer is about 40 ëC (104 ëF) and the residence time in the system is three to five 

minutes. These plants are in wide use and can give throughputs of 200 hl/hour. Warmed beer 

is carefully heated in three evaporators in series and the alcohol vapour in each case is 

collected in a pressure reduction vessel and then condensed in a spray condenser. After this 

triple process the beer, at 0.3% v/v alcohol, is cooled in a heat exchanger counter-current to 

incoming beer, which is so warmed. The beer leaves the plant at about 1 ëC (34 ëF). 

Dialysis 

In this method (Attenborough 1988; Niefind, 1982; Regan, 1990) the alcohol is removed 

by pumping beer through a membrane at a pressure of just over 2 bar (30 lb./in. 2 ). The 

membrane is normally a hollow fibre with a very thin wall. The membranes are collected 

into modules of several thousand and sealed at both ends. The beer is pumped uniformly 

through the membranes and the dialysate (water containing alcohol) passes through the 

hollow fibres in the opposite direction. The rate of dialysis is directly proportional to the 

concentration gradients formed and inversely proportional to the size of the molecules. 

Equilibrium is reached when the alcohol concentration is the same on both sides of the 

membrane. Alcohol is therefore removed from the dialysate by continuous distillation at 

reduced pressure and so the process of alcohol removal can continue. Separation of beer 

and alcohol occurs between 1 and 6 ëC (33 and 43 ëF) and so quality of the resultant dealcoholized 

beer is good. However, important flavour esters pass out of the beer with the 

alcohol. The dialysate, therefore, passes to a rectification column for removal of alcohol. 

The alcohol-free fraction containing esters is reincorporated into the low-alcohol beer. 

This is a complex process and considerable skill is required to produce beers of consistent 

flavour. Nevertheless the process is in use throughout the world. 

Reverse osmosis 

This process uses filtration at high pressure (30 to 60 bar) through a semipermeable 

membrane. The membranes are made of cellulose acetate, nylon or other polymers and 


allow the passage of small molecules such as water and ethanol and hold back the larger 

molecules. The high pressure used forces the water and alcohol against the natural 

osmotic pressure of the beer through the membrane. The flavour and aroma compounds 

mostly remain in the beer although in some systems the water/alcohol mixture (permeate) 

is rectified and the alcohol-free fraction, which does contain some volatiles, is added back 

to the beer. In other systems the permeate is used without rectification for sparging. 

The high pressure used causes heating and the equipment must be cooled so that the 

beer temperature does not rise above 15 ëC (60 ëF). The membrane modules must possess 

a large surface area to achieve commercial flow rates of at least 25 hl of low-alcohol beer 

per hour. About 60 l/m 2 /h of flow is possible so to achieve the above capacity around 20 

modules would be needed and the plant would be expensive. Cleaning the membranes is 

essential for optimum performance. 

Control of mashing 

Clearly, if mashing can be performed to produce a wort of low fermentability then there 

is the possibility of fermenting this wort to yield a beer of low alcohol content (Muller, 

1990). These methods were formerly associated with intensely `worty' flavours in the 

resultant beers which were cloying and not `moreish'. To reduce worty flavour the 

amount of malt in the mash must be reduced and replaced with starchy adjunct, which can 

provide 40 to 70% of the extract. Mashing is carried out to restrict amylolysis, by using 

temperatures of 70±80 ëC (160±175 ëF), which produces high levels of non-fermentable 

dextrins. The pH value of the wort is also usually artificially lowered with the use of 

phosphoric or sulphuric acids. Used on their own, methods relying solely on the 

restriction of amylolysis are seldom successful for the production of non-alcohol beers 

and they have to be combined with further measures to control the fermentation. It is also 

essential to ensure vigorous wort boiling to lower the levels of aldehydes, which in the 

absence of normal levels of ethanol, will spoil flavour. 

Control of fermentation 

The essence of these methods is to lower ethanol production by restricting fermentation. 

This can be done by stopping yeast activity before fermentation is complete, by using a 

special strain of yeast, by temperature control, or by controlling contact time of the yeast 

with wort. Fermentation can be stopped early by removal of yeast by filtration or 

centrifugation or by using a plate heat exchanger to provide a thermal shock and so kill 

the yeast. These methods are difficult to control (Brenner, 1980). A flocculent yeast at 

low pitching rate must be used and these products require maturation for at least ten days 

at 1 ëC (30 ëF) to ensure an acceptable flavour. 

In a patented so-called `cold contact' process, yeast is mixed with wort at 0.5 ëC 

(31 ëF) for 48 hours and circulated to ensure good mixing (Schur, 1983). There is virtually 

no ethanol production but carbonyl compounds are reduced in concentration and so worty 

flavour is reduced. After yeast removal the product can be matured for packaging. 

Fermentation can also be controlled by special yeasts, e.g., Saccharomycodes ludwigii, 

that ferment only glucose, fructose and sucrose, which comprise about 15% of the 

carbohydrate in an all malt wort. The resultant beer, therefore, contains < 0.5% ethanol 

but tastes sweet because of the high residual maltose and maltotriose content. 

The problem remains with any of these restricted fermentation methods that control is 

difficult and the beers often have a worty taste that limits their appeal. These problems 

can be reduced in continuous fermentation processes with immobilized yeast cells. A 

successful system is in use in The Netherlands (Mensour et al., 1997) in which the yeast 


is immobilized on DEAE cellulose in packed beds as a result of ionic binding between 

the negatively charged yeast cells and the positively charged carrier. Lactic acid is added 

to the wort before fermentation to lower the pH value to about 4.0 and so restrict the 

growth of bacteria. The wort thus treated is allowed to percolate through the reactor at 

1 ëC (30 ëF). Under these conditions the yeast preferentially metabolizes glucose. 

Maltose and maltotriose are not easily consumed as a result of the repression by glucose 

of their transport systems. A product of < 0.1% alcohol is produced. It is low in carbonyl 

compounds but contains some esters associated with normal beers and has good flavour 

stability. The system requires cleaning and re-sterilizing twice a year. A similar system 

has been described (Aivasidis et al., 1991) using sintered glass beads instead of DEAE 

cellulose as the support medium. 

Use of spent grains 

These methods (Attenborough, 1988) utilize spent grains to produce worts of low 

fermentability. The grains can simply be extracted with water or by acid hydrolysis or can 

be extrusion cooked. Fermentation is normally at a gravity of 8 ëP (32 ëSacch). Again, 

these beers require long maturation times (at least 14 days) to yield acceptable flavours. 

It is difficult to generalize on the best method to produce low-alcohol beer of 

acceptable flavour that will persuade the drinker to drink more of the product. Dealcoholized 

beers often contain lower concentrations of potential flavour-spoiling 

aldehydes than those produced by restricted fermentation but they lack the higher 

alcohols that can contribute positively to flavour. Further development of these drinks 

will take place only if the market grows. 

15.6.2 Ice beers 

Brewers have long experimented with ice beer. In Germany kegs of beer were 

deliberately `frozen' in the winter when an ice wall would form on the inside of the 

container and an increasingly strong beer would result. These were dangerous beers to 

drink particularly as the starting point was often a bock beer of high gravity (16 ëP; 

64 ëSacch)! 

In the 1980s many brewers were interested in lowering costs as sales began to fall after 

the peaks of the late 1970s. One idea was to concentrate beer at the brewery by removing 

water and to reconstitute it at the point of sale, thus lowering distribution cost. An 

obvious method of removing water was to freeze the beer, thus removing pure water as 

ice. The beer flavour components remained in the beer in a more concentrated form. 

Much work was carried out but, for a number of reasons associated with capital and 

revenue costs and the lack of a market, as a production process the method looked 

doomed to fail. Then, in Canada, the Labatt Company provided a whole new angle to the 

process. Labatt realized there was a powerful market association with the concept of ìce' 

beer and a new beer type was born. There was now a real market to drive product 

development. By the late 1990s most major brewers had produced their own ice brands 

and sales increased, backed by huge advertising spends. In 1997 in the USA alone over 

32 million hl of ice beer was produced and volumes were increasing by 4% year on year. 

The rate of growth of production of these beers has slowed but nevertheless ice beer is 

now an important segment of the alcoholic drinks market. 

The higher the original gravity of the beer the lower the temperature at which the beer 

freezes. The important point for ice beer is that the beer does not freeze homogeneously 

but as the temperature falls below 0 ëC, water separates as pure ice. Some compounds, 


a 

d 

b 

c 

CO 2 

e 

NH 3 

NH 3 

NH 3 

NH 3 

insoluble at low temperatures, such as some proteins and polyphenols also separate whilst 

alcohol and flavour volatiles concentrate. The key to a successful commercial process is 

to remove the ice as it is formed so it does not remain in one place and so restrict the 

process (as with casual ice bock production!). This is achieved by moving the beer during 

the cooling stage. 

In the Labatt process (Fig. 15.18) fully fermented beer is cooled and then centrifuged 

to remove yeast. The beer is further cooled and then pumped through three heat 

exchangers to lower the temperature to 4 ëC (25 ëF). Small ice crystals form and the 

beer is then moved to the recrystallizer where the small crystals deposit on larger ones 

already present. The ice crystals can be filtered from the beer or removed in a 

hydrocyclone. A finite amount of ice is always retained in the recrystallizer. The beer of 

high alcohol content is held in a storage tank and adjusted to the required alcohol content 

with sterile, deaerated, carbonated water. The alcohol content of the processed beer is 

normally higher than the starting beer and the removal of polyphenols gives the beer a 

characteristically smooth full taste. It is this property that is appreciated by drinkers. The 

process is expensive and the success for the brewery depends on the market allowing the 

charging of a higher price for the product. Further developments in ice beers will depend 

on how the market develops and what prices can be sustained for the beers. 

15.6.3 Diet beers 

These are not strictly post-fermentation treatment beers but are discussed here to complete 

this section on special beer types. The basis of this type is a beer low in carbohydrate. A 

higher proportion of fermentable carbohydrate is therefore made available to the yeast than 

is the case in standard brewery fermentations. The resulting beers have a higher ethanol 

content but lower dextrin levels from a given original extract compared to normal beers. It 

should be noted that these beers seldom have lower calorific values than normal beers, 

merely lower carbohydrate contents. It follows that these beers are usually derived from 

fermentations of 100% apparent attenuation. Mashing can be adjusted by extending the 

NH 3 

f f f 

NH 3 NH 3 NH 3 

Fig. 15.18 Ice beer plant, Labatt/Niro; (a) route from fermentation, (b) dropping cooler, (c) 

centrifuge, (d) yeast, (e) beer cooler, (f) heat exchanger, (g) recrystallizer, (h) route to conditioning. 

(Kunze, 1999). 


h 

CO 2 

g

time and by a low temperature stand at 50 ëC (122 ëF) for at least 30 minutes but attenuation 

limits of more than 90% are seldom produced. Enzymes must therefore be added to the 

fermenter to degrade residual dextrins during fermentation and this can be in the form of 

malt flour or diastatic malt extract. The -amylase, -amylase, and limit dextrinase so 

added results in the degradation of dextrin in the fermenting wort. 

In some countries the addition of enzymes of fungal origin is permitted (glucoamylase 

and pullulanase). Beers produced using fungal enzymes tend to be more biologically and 

non-biologically stable. Fermentation of these highly fermentable worts yields beers of 

very high alcohol contents and virtually no residual carbohydrate. The beers can be 

diluted to the appropriate alcohol content for sale. Pasteurization must be effectively 

carried out if using malt enzymes, as almost certainly bacteria will be introduced into the 

wort from the malt flour. 

Work at Brewing Research International in the UK in the mid 1990s (Baxter, 1995) 

resulted in the genetic modification of a brewing yeast strain to include a glucoamylase 

gene from a non-brewing yeast. This strain was approved for use and is used in-house for 

the production of low carbohydrate beers. But, as the result of general public disquiet 

about the use of genetic manipulation, there have been no commercial developments 

using this strain. There seems to be a continuing, if static, market for this type of beer 

particularly as it is often assumed (wrongly!) to be a less fattening beer. 

Finally, the term `light' beer is sometimes confused with diet beer. Light beers have no 

real generic definition but are merely beers of low-alcohol or low-sugar content. As a 

result what constitutes a light beer can vary enormously. Beers described as light beers 

can have alcoholic strengths of from 2.5 to 4.0% abv but usually have a dextrin content of 

about 1% and a calorific value of 25 to 30 kcal/100 ml. As such, if drunk as an alternative 

to normal beer, they may be less fattening. 

15.7 Summary 

The post-fermentation treatment of beer is critical in yielding a product that is fit for sale 

and meets the consumer's expectation of a quality drink. The aim has to be to delight the 

consumer so that he will return to the product and drink it again. Flavour, clarity and 

stability of beer are improved by the post-fermentation treatments that have been 

discussed. This is also the area of the whole brewing process where emphasis can be 

placed on the development of special beer types. 

There is increasing interest in the manipulation of beer properties post-fermentation to 

produce different beer brands from essentially the same brewhouse and fermentation 

techniques. This can result in considerable saving in capital and revenue cost with fewer 

requirements to invest in expensive maturation storage. In this respect further 

development in immobilized yeast technology with its potential for accelerated flavour 

improvement and production of novel beers would seem to be worthwhile. 


AIVASIDIS, A., MANDREY, C., ELLIS, H. G. and KATZKE, M. (1991) Proc. 23rd Congr Eur. Brew. Conv, 

Lisbon, 569. 

ANDERSON, R. G., BRITES SANCHES, A., DEVREUX, A., DUE, J., HAMMOND, J., MARTIN, P. A., OLIVER-DAMEN, 

B. and SMITH, I. B. (2000) Fermentation and Maturation Manual of Good Practice, European 

Brewery Convention, Zoeterwoude, The Netherlands. 


ANDREWS, J. M. H. (1997) Ferment, 10, 309. 

ANNESS, B. J. (1980) J. Inst. Brewing, 86, 134. 

ATTENBOROUGH, M. W. (1988) Ferment, 1 (2), 40. 

BARNES, Z. C. (2001) Personal communication. 

BAXTER, E. D. (1995) Ferment, 8, 307. 

BRENNER, M. W. (1980) Tech. Quart. MBAA. 17, 185. 

BROWN, D. G. W., CLAPPERTON, J. . F. and MEILGARD, M. C. (1978) J. Amer. Soc. Brewing Chemists, 36, 73. 

BYRNE, H., MATHEWS, S., MADIGAN, D., KELLY, R. J., MCENROE, C. and HARMEY, D. (1999) Proc. 7th Conv. 

Inst. Brew (Africa Section), Nairobi, 55. 

CHALMERS, S. and HAUGHNEY, H. ( 1998) Proc. 25th Conv. Inst. Brew (Asia Pacific Section), Perth, 165. 

CHAPON, L. (1994) Brewers' Guard. 123 (12), 21. 

COORS, J. H. (1977) The Practical Brewer, Master Brewers' Association of the Americas. Madison, 

Wisconsin. 

DICKENSON, C. J. and ANDERSON, R. G. (1981) Proc. 18th Congr. Eur. Brew. Conv., Copenhagen, 413. 

ESNAULT, E. (1995) Brewers' Guard., 124 (1), 25. 

GARDNER, D. J. S. (1993) Ferment, 6, 279. 

GOPAL, C., REHMANJI, M., MOLA, A., NARAYANAN, K., TRINH, T. and WHITTINGHAM, J. (1999) Proc. 7th 

Conv. Inst. Brew. (Africa Section), Nairobi, 62. 

GOPAL, C. and REHMANJI, M. (2000) Brewers' Guard., 129 (5). 

GREEN, H., SHAW, R. and CANDY, E. (2000) The Brewer, 86, 201. 

GUZMAN, J. E., MCKEOWN, I. . P., GLEAVES, M., STEWART, G. G. and DOYLE, A. (1999) Tech. Quart. MBAA. 

36, 227. 

HANNEMAN, W. (1999) Tech. Quart. MBAA. 36, 167. 

HARDING, J. A. A. (1977) Brewers' Guard., 106 (9), 90. 

HASIMOTO, N. (1973) Ann. Rept. Res. Labs. Kirin Brew. Co., 16, 1. 

HOUGH, J. S., BRIGGS, D. E., STEVENS, R. and YOUNG, T. W. (1982) Malting and Brewing Science Volume 2, 

2nd Edition, Aspen, Gaithersburg, Maryland. 

JENDE-STRID, B. (1997) Proc. 25th Congr. Eur. Brew. Conv, Maastricht, 101. 

JEPSEN, S. (1991) 4th Meeting on the Industrial Applications of Enzymes, Barcelona. 

KUNZE, W. (1999) Technology Brewing and Malting, Int. Edition, (Translated by Wainwright, T.) VLB, 

Berlin. 

LEACH, A. A. (1989) Ferment, 2, 33. 

LEACH, A. A. and BARRETT, J. E. (1967) J. Inst. Brewing, 73, 246. 

LONG, D. E. (1995) Ferment, 8, 239. 

MARRIOTT, R. (1999) Proc. 7th Conv. Inst. Brew. (Africa Section), Nairobi, 140. 

MCKEOWN, I. P. and EARL, G. J. (2000) Brewers' Guard., 129 (6). 

MCMURROUGH, I., MADIGAN, D., KELLY, R. and O'ROURKE, T. (1999) Food Technology, 53, 58. 

MENSOUR, M. A., MARGARITIS, A., BRIENS, C. L., PILKINGTON, H. and RUSSELL, I. (1997) J. Inst. Brewing, 

103, 363. 

MIEDANER, H. (1978) The Brewer, 64 (2), 33. 

MULLER, R. (1990) Ferment, 3 (4), 224. 

MUSCHE, R. A. and DE PAUWE, C. (1999) J. Inst. Brewing, 105, 386. 

NARZISS, L., WULFINGER, H., STICH, S. and LAIBLE, R. (1992) Brauwelt, 51/52, 2650. 

NIEFIND, H. (1982) Monatsshrift fuÈr Brauerei, 35, 90. 

OECHLSE, D., ASCHER, R. and FEIFEL, K. (2000) Tech. Quart. MBAA., 37, 377. 

O'ROURKE, T. (2000) Brewers' Guard., 129 (2), 29. 

PAJUNEN, E. and JAÈ AÈ SKELAÈ INEN, K. (1993) Proc. 24th Congr. Eur. Brew. Conv. Oslo, 559. 

PAJUNEN, E., GROÈ NQUIST, A. and RANTA, B. (1991) Proc. 23rd Congr. Eur. Brew. Conv., Lisbon, 361. 

PFISTERER, E. and STEWART, G. G. (1975) Proc. 15th Congr. Eur. Brew. Conv., Nice, 255. 

REGAN, J. (1990) Ferment, 3 (4), 235. 

SHIEL,P. (1999) Proc. 7th Conv. Inst. Brew. Conv (Africa Section), Nairobi, 76. 

SCHUR, F. (1983) Proc. 19th Congr. Eur. Brew. Conv., London, 353. 

SIEBERT, K. J., TROUKHANOVA, N. V. and LYNN, P. Y. (1996) J. Agric. Food Chem., 44, 80. 

SOLE, S. M. (2000) Ferment, 13 (4), 25. 

STEWART, G. G., BRYCE, J. H., COOPER, D., MONAGAS, M. and YOUNIS, O. (1999) Proc. 7th Conv. Inst. Brew 

(Africa Section), Nairobi, 100. 

THILERT, T. (1999), Tech. Quart. MBAA. 36, 427. 

ZANGRANDO, T. and GIRINI, G. (1969) Proc. 12th Congr. Eur. Brew. Conv., Interlaken, 445. 


16 

Native African beers 


African beers almost certainly have ancient origins, and may have originated in Egypt or 

Mesopotamia, where beers were being made by at least 3500 BC, and probably much 

earlier (Briggs, 1998). The names given to African beers are often unacceptable or 

inexact. Thus beers brewed in southern Africa have been called Kaffir or Bantu beers (but 

elsewhere in Africa beers are made by peoples who are not Bantu). Another term is 

opaque beers but not all African beers are truly opaque (e.g. Nigerian òtika') and, on the 

other hand, some European-style beers are at least turbid (wheat beers; some ales 

consumed with conditioning yeast in suspension) and others, such as stouts, are not 

transparent. The term `sorghum beers' is inexact as sorghum (raw grain or malt) is 

sometimes wholly or largely replaced with maize or millets or wheat or barley, and 

indeed in some cases bananas or manioc (cassava) serve as starchy adjuncts. Bouza may 

be made with wheat, barley or millets. Native names are also confusing in that different 

names are used for similar products by different tribes and within one language group 

different names are used for different types or qualities of beers (Daiber and Taylor, 

1995; Dendy, 1995; Haggblade and Holzapfel, 1989; Harris, 1997; Miracle, 1965; 

Novellie, 1966, 1968, 1977; Novellie and De Schaepdrijver, 1986; Peterson and Tressler, 

1965; Schwarz, 1956). Among the best known of these names are utshwala (Zulu) and 

joala (Basuto). 

Traditionally, beers are made by women brewsters, as was the case in mediaeval 

Europe, and they may be consumed with some ceremony. However, in southern Africa, 

as urbanization occurred, men moved into towns as casual labour, leaving their 

womenfolk behind. To meet the demand for opaque beers commercial brewing began 

(around 1908±1910) in Bulawayo and Durban. Since then, in some periods, the rate of 

increase in production has risen at an astonishing rate. For example, in South Africa, at 

one stage production increased by 26% in one year and while in 1953/4 production was 

20 million imp. gallons (0.909 million hl) in 1965/6 production was 120 million imp. 

gallons (5.46 million hl; Novellie, 1968). At the same time much larger volumes of beer 

were produced in homes and small-scale `village' breweries. Estimates of more recent 


opaque beer production in different countries are in millions of hectolitres/year (Harris, 

1997). At first the industrialized production of these beers was not straightforward, and 

many difficulties were encountered. To overcome these problems the CSIR (the Council 

for Scientific and Industrial Research) in South Africa established, around 1953, a 

research organization to investigate the bases of malting sorghum and beer production. 

The publications from this group, and its successors, provide nearly all the available 

information on the science of opaque beer production. Industrial production is spreading 

into other countries (Harris, 1997). In many areas the traditional brewing methods differ 

significantly from those used in southern Africa, and no doubt many brewing methods 

have not been described. Brewing may occur in the home, for home consumption or for 

sale, or it may be produced in a factory. Factory brewing seems to be carried out by men, 

a change that parallels the historical move from home-based brewsters (women) to 

industrial brewers (men) in Europe. 

16.1.1 An outline of the stages of production 

Stages usually distinguished in the production of African beers are the selection of the 

raw materials, malting (usually sorghum or millets or, less usually, maize), grinding, 

souring, cooking (with adjuncts), mashing (or conversion), straining, fermentation, and 

packing, distribution and consumption. The souring fermentation stage produces a 

desired level of acidity, caused by lactic acid. The second fermentation produces (mainly) 

alcohol and carbon dioxide. The beers are consumed while they are warm and are still 

fermenting and effervescent and so their composition is continually altering. Because of 

the continuing production of carbon dioxide the beers are held in vented containers. If not 

consumed in a day or two they become flat, too acidic and poorly flavoured, at least 

partly because acetic acid accumulates, and they are rejected. 

The short shelf-lives of African beers create many commercial production and 

distribution problems. There are major differences between African- and European-style 

beers. African beers are rarely or never flavoured by herbs (in contrast to hopped 

European-style beers), complete starch conversion is avoided, and brewing does not 

produce an excess of yeast. Indeed factory brewing is a net consumer of yeast. Then the 

beers are always consumed, with the yeast they contain, while still fermenting and they 

are mostly opaque because of suspended yeast, starch granules and small particles of 

cereal grains, which are maintained in suspension by the rising bubbles of carbon dioxide 

and the high viscosity of the beer. The high viscosity is caused by gelatinized but 

incompletely degraded starch. It seems that most drinkers are more concerned with the 

flavour and body of a beer than with the (changing) alcohol content. South African beers 

are described as being as refreshingly sour as yoghurt, with a characteristic fruity odour. 

Alcohol contents of 1±8% have been noted, but values of 2.5±4.5% seem to be usual. 

Colours vary from a pale buff to a pinkish-brown, or elsewhere may even have a reddish 

tinge. pH values may be 3.3±3.6, lactic acid contents about 0.26% and total solids are 

around 6%. However, there are wide variations in beer composition, particularly in homebrewed 

beers (Haggblade and Holzapfel, 1989; Harris, 1997; Novellie, 1966; Novellie 

and De Schaepdrijver, 1986). 

16.1.2 Bouza 

Bouza (bouzah, bowza, etc.) is a bread-beer, made in Egypt and the Sudan from wheat, 

barley or millets, using methods supposedly resembling those used by the ancient 


Mesopotamians and Egyptians (Briggs, 1998; Morcos et al., 1973). Coarsely ground 

grain, sometimes mixed with a little malt, is mixed with water and some leaven or some 

yeast in sourdough. After standing the dough is moulded into loaves, which are lightly 

cooked. About one-third more grain is malted and then, often after drying in the sun, the 

malt (green or dry) is ground up with water and lumps of the broken up loaves. The 

mixture begins to ferment, either spontaneously or after the addition of some older bouza. 

After a period of active fermentation the mixture is filtered, for example through a 

horsehair sieve. The introduction of air at this stage checks the fermentation, but this is 

soon resumed. The drink is thick and yeasty, pale yellow, acidic and with a characteristic 

odour. The pH may be 3.5±4 and the alcohol content 4±5.5 g/100g. It must be consumed 

quickly, before deterioration begins (Briggs, 1998). 

16.1.3 Merissa 

In the Sudan a beer called merissa is produced. It has been said that sometimes the women 

chew some of the grain and spit the mix into the mixture, so adding salivary -amylase, 

which may accelerate starch degradation; others do not mention this practice. Dirar (1978) 

describes a complex scheme for making merissa. Some sorghum grain is malted, dried and 

ground to a flour. Raw sorghum grain is ground to a fine flour, which is divided into three 

equal lots, each of which is processed differently. The first lot is lightly cooked, `halfcooked', 

to a grey powder. The second lot is well cooked to give a brown paste. These two 

solid materials are mixed on leaves and allowed to cool. The third portion is wetted with just 

enough water to moisten it, and is set aside for about 36 hours, when a spontaneous, mainly 

lactic fermentation occurs. The acidified dough is strongly cooked in a steel container with 

repeated mixing until it is dark brown, is extremely sour and has a pleasant caramel flavour. 

It is cooled to room temperature and mixed with about 5% malt flour, water and some good 

merissa. Fermentation is well established after 4±5 hours. This material is too acid to drink. 

Portions of the combined two-thirds cooked flour, mixed with malt flour, are added in 

increments to the, strongly fermenting, acid fraction, without stirring them together. After 

8±10 hours. fermentation the mixture is filtered through cloth. The liquid is consumed while 

it is still fermenting. The solids removed by straining are fed to cattle. The product has a pH 

of about 4, and an alcohol content of about 5%. 

16.1.4 Busaa and some other beers 

Busaa, and similar drinks, are made in Kenya, Uganda and Tanzania ((Nout, 1980; 

O'Rourke, 2001). In making busaa maize grits are mixed with water and allowed to stand 

for 2±3 days, at about 25 ëC (77 ëF), for souring. Malted finger millet (Eleusine coracana) 

is prepared by steeping for 8±24 h, then germinating for 2±3 days, also at about 25 ëC 

(77 ëF), followed by drying in the sun for 1±2 days and coarsely grinding. The soured 

maize dough is cooked on steel sheets over a charcoal fire at 65±75 ëC (149±167 ëF) for 

three hours. The cooled maize soured material is broken into lumps and one part is mixed 

with 1.5 parts of water and 0.1 part of malt flour. The main fermentation proceeds for 2±4 

days. The mixture is then strained and consumed within one day, as it deteriorates rapidly 

and the increasing acidity (sometimes approaching 2% lactic acid) is unpleasant and 

causes the suspended solids to precipitate. When consumed beers may contain 0.5±1% 

lactic acid and 2±4% (v/v) alcohol. 

Investigations into this process indicate that temperature control and the use of pure 

cultures of Lactobacilli and yeast could be used to advantage in the souring and 


fermentation stages, giving more stable products that could be kept longer if preserved by 

end-fermentation or pasteurization in bottle. The latter was preferred. Beers made by 

related processes include ajou (Uganda) and mbweje (Tanzania) (O'Rourke, 2001). In the 

former a paste of ground millet and water is soured by burying bags of it in the ground for 

5±7 days. In the latter bananas are used as starchy adjuncts. In the foregoing examples 

some materials are well-cooked, giving the possibility of adjusting the flavours of the 

beers by varying the intensity of the cooking. This is not so in subsequent examples, 

where water-grain mixtures are only boiled. 

The preparation methods of other African beers have been described, otika from 

sorghum (Ogundiwin, 1977), pito from maize or sorghum (Ekundayo, 1969), oyokpo 

from millet (Pennisetum typhoideum; Iwuagwu and Izuagbe, 1985) and burukutu, also 

from sorghum (Faparusi, 1970; Faparusi et al., 1973) all in Nigeria. The preparation 

methods for sorghum beers in the Cameroons and Togo have also been described 

(Chevassus-Agnes et al., 1976; Perisse et al., 1959). However, most information is 

available for the brewing methods used in southern Africa. 

16.1.5 Southern African beers 

A reliable method of brewing, in use by the Zulus in 1907 (quoted by Fox, 1938) used 

sorghum for preference, but millets and/or maize might also be used. The process required 

considerable skill. Grain sewn into sacks was steeped in running water for 1±2 days 

(sorghum) or up to 4 days (maize), longer periods being used when the weather was cool. 

Grain was sprouted, still in the sack or in a pot, for 2±5 days, maize taking longer, until 

shoot growth was judged adequate (1.9 cm, (0.75 in.), for sorghum, 1.3 cm, (0.5 in.), for 

maize). Usually the malt was dried in the sun or inside a hut, but sometimes it was used 

undried. Initially unground grain or (better) a 50:50 mixture of grain and malt was soaked 

in water for a day. After draining the wet grain was finely ground to a paste, between 

stones, in the morning and the dough was moulded into lumps. In the afternoon the paste 

was just covered with boiling water in a pot and cold water was added to adjust the 

temperature according to the brewster's judgement. As the mixture slowly cooled so 

spontaneous acidification occurred. Next day the water was collected from above the 

dough and was boiled with more water while the dough itself was mixed with fresh boiling 

water and was mixed to a thin porridge which was added to the boiling water. After the 

boil, of 20±40 minutes, (longer periods being needed for maize), the mixture had 

thickened because the starch present had gelatinized. The mixture was too thick to pour 

from a spoon. Most of the mixture was allowed to cool quite slowly, but a small amount 

was cooled quickly and was mixed with ground malt, when starch conversion and a 

spontaneous fermentation began, creating a `starter culture'. When the main mash was 

cool enough more ground malt, in an amount exceeding the initial amount of grain by 

about 25%, was mixed in, together with the fermenting `starter culture', initiating a rapid 

onset of fermentation. When fermentation was vigorous the mixture was strained through 

a woven grass strainer. The filtered liquid continued to ferment while the strainings were 

reserved for making a `small beer'. Fermentation went on for 1±2 more days before the 

beer was consumed. Beyond this period the product spoilt. Analyses of these beers gave 

estimates of solids contents of 5±13%, alcohol contents of 0.5±8.0% (v/v; usually 4% 

when fresh), crude protein contents of 0.7±1% and mineral salts contents of 0.18±0.36%. 

Others reported solids contents of up to 20%. These beers were regarded as foods, which 

constituted most or all of men's diets in some seasons, and were an alternative to porridges 

or acidified porridges which were made without the second, alcoholic fermentation. 


Other, less complicated methods have been quoted (Haggeblade and Holzapfel, 1989; 

Novellie and De Schaepdrijver, 1986). For example, maize broken up by pounding is 

mixed with boiling water and is left for a day, when a spontaneous lactic fermentation 

occurs. More water is added, the mixture is boiled for 2±3 h, and is left to cool. When 

cool a roughly equal amount of pounded sorghum malt is mixed into the soured maize 

adjunct mixture. After a fermentation period of 24 h the mixture is strained and is ready 

for consumption. In a `generalized' scheme for home or small-scale brewing for sale 

malted sorghum, or less usually malted pearl millet (Pennisetum typhoides) or finger 

millet (Eleusine coracana) is broken up using a mortar and pestle, or by grinding between 

stones or is pulverized in a hammer mill. In southern Africa commercially prepared and 

ground sorghum malt may be purchased. For souring some of the malt is mixed with 

water and is heated for 30±90 minutes. Heating may be in traditional clay pots, or in iron 

or steel containers. The mixture is allowed to cool overnight, when a spontaneous lactic 

acid fermentation begins, and the mixture is soured. Next day the sour is mixed with more 

water and ground sorghum, sorghum malt, maize or maize grits. The mixture is cooked 

by bringing it to the boil and boiling for 2±7 h. The mixture is cooled overnight, 

sometimes by dividing it between several shallow dishes. During this time the mixture 

thickens as the gelatinized starch tends to set. In the morning mashing and fermentation 

begin when more malt flour is mixed in, sometimes together with some good beer, which 

provides an inoculum of yeasts and other microbes. This stage may be carried out in 

wooden, metal, clay or plastic containers. After about two days fermentation the beer is 

strained to remove coarse particles by passing it through a woven-grass, bag-like 

container (which may be squeezed to recover more liquid) or a metal screen. When 

fermentation has resumed the beer is ready for consumption. 

16.2 Malting sorghum and millets 

In southern Africa sorghum malts are preferred but elsewhere in Africa malts may also be 

made from various millets (Briggs, 1998; Haggblade and Holzapfel, 1989; Miracle, 1965; 

Nout and Davies, 1982; Novellie, 1968; Novellie and De Schaepdrijver, 1986; Nzalibe 

and Nwasike, 1995). All these plants belong to the grass family. Technically the grains 

are caryopses, fruits in which the ovary wall remains investing the seed as the pericarp. It 

seems that maize is malted only as a last resort. In `home' brewing the chosen grain is 

steeped sewn into rush bags, sacks or held in baskets, in running water, or in pots, jars, 

tubs, gourds, calabashes or other vessels, in still water. Steeping times vary from 1±3 

days. The grain is drained and placed in jars or baskets lined with leaves or on mats, and 

is covered with leaves and left to germinate. From time to time it is watered. When 

growth is far enough advanced, in 2±6 days, the grain is usually dried in the sun before 

use, although sometimes the malt is used undried. It is used with the roots and shoots still 

attached. All the grain species occur in many varieties of widely differing malting 

qualities and characters. Sorghums with grains having thousand corn dry weights (TCW) 

of 7±61 g are known, but values of 10±38 g are more usual, these grains having 

dimensions of 3±5 mm (0.118±0.197 in.) by 2±5 mm (0.079±0.197 in.). Sorghum varieties 

vary greatly in their malting qualities. 

Maize grains are larger, for example dent corn may have TCWs of 150±300 g and be 

12 mm by 8 mm by 4 mm (0.472 by 0.315 by 0.157 in.). This grain was introduced into 

Africa, perhaps in the 16th century, and grains, or materials made from them, are 

common brewing adjuncts. The various millets have much smaller grains. This creates 


problems for ìndustrial' maltsters. Like sorghum, millets have many different names 

(Briggs, 1998). Pearl millet (Pennisetum typhoides) has the largest grains (TCW 5±10 g; 

3±4 mm, 0.118±0.157 in., length). Finger millet (Eleusine coracana) is frequently malted. 

Common millet (Panicum miliaceum) grains are about 3mm (0.118 in.) long and have a 

TCW of about 6 g. Acha, or fonio (Digitaria exilis) grains have a TCW of only 0.65 g. 

In southern Africa, the first industrial malting of sorghum followed village practices 

and to some extent this is still the case (Daiber and Taylor, 1995; Haggblade and 

Holzapfel, 1989). Grain is steeped for about 16±18 h in concrete tanks, metal drums or 

barrels, and then, after draining, is spread out, in the open, on slightly sloping concrete 

floors in beds 13±90 cm (5.12±35.4 in.) thick. Here it is covered with wet sacks. At 

intervals the grain is unevenly wetted by hosing, and it may be turned by hand. In warm 

weather the grain is spread more thinly and in cold weather the bed is thickened to favour 

heat loss and heat retention respectively. Evidently temperature control is inadequate, as 

is regulation of the wetting, `sprinkling'. Growth is very irregular, the grain at the top of 

the bed being poorly grown, while that at the base, on the floor, is overgrown. When 

growth is judged to be sufficient the sacks are removed and the grain is spread more 

thinly to dry in the sun. This process is used to make malt that is sold after grinding and 

makes a very irregular product. It often carries a high load of microbes. Apart from 

sometimes covering the floors with roofs, but with no side walls, and sometimes using 

steeps containing formaldehyde (see below) this form of malting seems to have advanced 

very little. 

Malting for the larger breweries is carried out indoors, under more controlled and 

hygienic conditions. The grain is thoroughly cleaned before use and may be washed. 

Malting plant is regularly cleaned. Steeping, for 16±24 h (or even as little as 6 h), is 

usually carried out in tanks that may be aerated. The moisture content finally achieved 

is about 35%. To control microbes the grain may initially be steeped in a solution of 

formaldehyde or sodium hypochlorite (an agent that taints barley malts, giving them an 

àntiseptic' flavour). The formaldehyde treatment was originally adopted to deal with 

high-tannin, birdproof sorghums that are so rich in tannins they inactivate and 

insolubilize the malt enzymes during mashing and inhibit the souring process, so 

blocking brewing. For the first four hours of steeping, the grain is immersed in a 

solution of formaldehyde (0.02±0.08%, depending on the tannin content of the 

particular grain), then it is thoroughly rinsed and steeping is completed in fresh water 

(Daiber, 1975; 1978). Recent studies indicate that steeping for longer periods of up to 

40 h at 25±30 ëC (77±86 ëF), with air-rests, gives superior malts (Dewar et al., 

1997a,b). Evidently there would be advantages to using temperature-controlled steeps, 

equipped for carbon dioxide extraction, during air-rest periods. 

Germination is carried out in modified Saladin boxes. The grain bed, which may be 

1.5 m (4.92 ft.) deep, rests on a perforated base through which temperature-controlled and 

humidified air can be blown to cool the grain. Because the grain grows so vigorously the 

airflow must be larger than that used for barley, around 1,000±1,200 m 3 /h/tonne grain. 

Because of the extensive embryo growth the volume of the grain bed increases greatly 

during germination. The grain should be at 24±30 ëC (75.2±86 ëF). Turning may be 

mechanical but, because the seedlings are so easily damaged, it may be carried out better 

by hand, the grain being shovelled from one compartment to the next. During turning the 

grain may be also sprinkled with water as required. After 5±7 days, or 4±6 days under 

ideal conditions, the malt is dried by blowing warm air, at 50±60 ëC (122±140 ëF), up 

through the grain. The temperature is kept low to minimize the destruction of enzymes, 

and avoiding transfer to a kiln avoids damage to the green malt. Experience with barley 


suggests that a lower initial drying-air temperature (40 ëC; 104 ëF) would be beneficial for 

enzyme survival. 

The main objectives of malting sorghum are to generate the enzymes needed for 

mashing and to supply sufficient soluble nitrogen to support the Lactobacilli during 

souring and the yeast during the fermentations. Sometimes sugar levels are also 

considered. Although it is sometimes measured, the yield of extract is largely ignored. 

This is because most of the extracted materials in sorghum beers are derived from the 

adjunct(s) used. Malting losses, with seedling roots and shoots retained with the malt, are 

10±20% dry basis (Daiber and Taylor, 1995), but using some conditions they are very 

much higher (particularly if the seedling tissues are discarded, as is the case for malts 

intended for making lager beers), and so the maltster must strike a balance between malt 

quality and losses. 

The criteria that must be met are the level of diastatic power (DP), the free amino 

nitrogen (FAN), and a low tannin level (Daiber and Taylor, 1995; Daiber et al., 1973). In 

addition the beer should not have an appreciable level of mycotoxins or cyanide from the 

malt. Sorghum seedlings generate the cyanogenic glycoside dhurrin, which is 

enzymically degraded to glucose, p-hydroxybenzaldehyde and hydrogen cyanide (prussic 

acid). Toxic levels of prussic acid precursor occur in sorghum seedlings and some baking 

and steaming processes reduce this to low levels (Dada and Dendy, 1987). When 

sorghum was germinated for six days at three different temperatures the cyanide contents 

of the shoots increased continually to six days, to 614 ppm, at 25 ëC (77 ëF), peaked after 

two days germination at 666 ppm, at 30 ëC (86 ëF), and fell after two days at 35 ëC (95 ëF), 

when it was 385 ppm (Panasiuk and Bills, 1984). However, for reasons that are not clear, 

toxic levels of prussic acid do not appear in sorghum beers (Glennie, 1983). 

The grain chosen for malting must be of an acceptable variety, be clean and 

undamaged and germinate well, usually at least 92±95%. As sorghum germinates the DP 

rises from a negligible value as both - and -amylases are synthesized in the embryo. 

There are variations, but generally 60±70% of the DP is due to the former enzyme. This 

contrasts with the situation found in barley malts, in which the DP is higher and the - 

amylase activity is relatively much higher (50±80% of the DP), and it exists preformed in 

the starchy endosperm (Briggs, 1998). There are problems in extracting the enzymes 

from some sorghum malts, due to the presence of tannins, and often 2% peptone is 

included in the extraction medium when DP is to be determined. -Glucosidase (maltase) 

is insoluble but active. Its importance in brewing is unclear and its activity is not 

determined on a routine basis. Its pH optimum is unusually acidic, pH 3.8 (Taylor and 

Dewar, 1994) and so it is likely to act in mashing and produce the relatively high levels of 

glucose found in the worts. 

The DP levels of sorghum malts (20±60 SDU/g malt) are low compared to those of 

barley malts (150±200 SDU/g malt). The units of activity used, sorghum diastatic units 

(previously KDUs), are small and 1 SDU approximately equals 0.5 ëL. Many malts were 

and are produced with too low DP activities, often caused by growing the grain at too low 

temperatures. Commercial brewers now specify minimum DP values in malts, (often 28± 

33 SDU/g). Higher levels of enzymes and FAN in malts are obtained by malting highnitrogen 

grain and choosing smaller grains, which contain larger proportions of embryo 

tissues. In sorghum all the diastase is synthesized in the embryo, and apparently there is 

no contribution from the aleurone layer. Sorghum grains, unlike barley, wheat and some 

millets, do not respond to added gibberellic acid to any appreciable extent. 

The level of FAN in sorghum malt is important because this represents the nitrogencontaining 

nutrients needed by the microbes (Pickerell, 1986). The largest part of the 


FAN is from the shoots and roots, which are ground up together with the rest of the grain, 

before mashing. FAN levels, like DP, increase with malting time. During malting 

proteolytic activity increases relatively little and this enzymic activity is poorly 

extractable but carboxypeptidase activity, which is concentrated in the embryo, increases 

substantially (Dewar et al., 1997c, d; Dewar and Taylor, 1995; Taylor, 1991). Both 

enzymic activities have acidic pH optima. 

The structure of the sorghum grain alters, i.e., undergoes modification, during malting 

but the cell walls of the endosperm appear to remain largely intact, while losing their 

physical strength. Modification can be assessed by measuring the porosity or 

compressibility of the grain (which both decline as modification proceeds) or, most 

simply, by measuring its specific gravity, which also declines (Daiber et al., 1973). These 

characteristics are not determined routinely. 

Relatively few studies have been reported aimed at optimizing sorghum malting for 

the production of African opaque beers. Other studies have been directed towards making 

sorghum malts for brewing clear, lager-type beers (Briggs, 1998). The variables that have 

been studied are steeping duration and temperature, the degree of sprinkling during 

germination, the temperature and duration of germination. The malt characteristics 

investigated have been the DP, FAN, hot water extract (total soluble solids) and malting 

loss. Extracts have been determined in various ways, e.g., after mashing for two hours at 

60 ëC (140 ëF). Values obtained after mashing at 45 ëC and then 70 ëC (113 and 158 ëF) are 

higher (Daiber and Taylor, 1995). In connection with brewing opaque beers malting 

losses are determined with the seedling tissues retained with the malt, so the losses 

encountered are mainly respiratory losses with a little due to leaching losses incurred 

during steeping. 

Early experiments indicated that malting temperatures of 20 ëC (68 ëF), or less, were 

much too low for maximum development of DP (Novellie, 1966). Systematic studies 

confirmed various other results, e.g. that for grain steeped and grown at about 30 ëC 

(86 ëF), 18 hours was the optimum steeping time and 4±5 days the best germination 

period for malting (Pathirana et al., 1983). In a study with germination times of up to six 

days, at temperatures of 24±36 ëC (75.2±96.8 ëF), with fixed initial steeping conditions 

but three levels of watering during germination, it was found that at steep out the 

moisture content was 33.7%, and after six days germination the moisture contents of the 

low, medium and highly wetted samples were 42.8%, 60.8% and 77% respectively 

(Morrall et al., 1986). Entire malt was used in this study and total soluble solids 

(èxtract') were determined with two-hour mashes at 60 ëC (140 ëF), conditions which 

approximate to those used in mashing when making opaque beer. DP increased rapidly 

during the first four days of germination, the highest yield of 46.6 SDU/g being obtained 

after five days germination, at 24 ëC (75.2 ëF) in the medium-wetted grain, but 28 ëC 

(82.4 ëF) gave nearly as good results. At this stage the malting loss was 9.9%, the FAN 

was 129 mg/100 g malt and the total soluble solids value was 73.5%. 

At higher temperatures DP values were lower and under a number of conditions DP 

values peaked and then declined, as also happens in malting millets and wet-grown barley 

(Briggs, 1998). All the variables influenced FAN, which increased with increasing 

moisture content and germination time. The maximum level reached was 180 mg FAN/ 

100 g malt, after six days germination, at the highest moisture content, in grain grown at 

32 ëC (89.6 ëF), when the malting loss was about 20%, the DP was 38 and the total soluble 

solids value was 66.1%. The maximum level of total soluble solids, 75.5%, was obtained 

from grain grown for six days at 24 ëC (75.2 ëF), with a medium moisture content, when 

the malting loss was 13.6%, the DP was 46.3 and the FAN was 148 mg/100 g malt. 


Malting losses (seedling tissues retained with the malt) increased with germination time 

and increasing moisture content but were relatively little influenced by germination 

temperature. 

Clearly, strict control of malting conditions is necessary to produce malt in the best 

possible yield, with the desired specifications. Usually it will be necessary to compromise 

since, for example, conditions chosen to give maximum levels of FAN are associated 

with high malting losses and low DP values. Air-rests are beneficial when warm water 

steeping is tested (Ezeogu and Okolo, 1994, 1995). Further studies indicate that with 

steep aeration or air-rests longer steeping times, up to around 40 hours, are better than the 

usual shorter times found by earlier investigators (Dewar et al., 1997 a, b, c). Malt DP 

increased with steeping temperature up to 30 ëC (86 ëF), and of the temperatures tested, 

FAN and extract peaked at a steeping temperature of 25 ëC (77 ëF). Aeration increased the 

yields of extract and FAN. Using optimized steeping conditions, with air-rests, and two 

different wetting schedules, it was found that generally the optimum germination 

temperature was between 25 and 30 ëC (77 and 86 ëF). The roots and shoots contributed 

up to 61% of the whole malt FAN. In the winter, in South Africa, temperatures can fall 

below 18 ëC (64.4 ëF). It was concluded that much sorghum malt is made under seriously 

sub-optimal conditions. 

Ways have been sought for reducing the losses of dry matter that occur when sorghum 

is malted. Applications of potassium bromate and ammonia have not been beneficial, and 

warm-water steeping, at 40 ëC (104 ëF), have given equivocal results, which varied with 

different grain samples (Ezeogu and Okolo, 1994, 1995). When dilute alkaline steeps 

(0.1% sodium hydroxide) were tested, probably with the initial object of extracting 

unwanted tannins, the moisture content and quality of the malts obtained were increased, 

even when low-tannin grains were treated (Dewar et al., 1997a, 1999; Ezeogo and Okolo, 

1999; Okolo and Ezeogu, 1996). However, there were varietal differences in responses to 

alkaline steeping. 

While various millets are malted in villages it seems unlikely that they are now malted 

commercially. The small sizes of the grains make them inconvenient to handle and, in 

pneumatic malting plants, they tend to block the slots of the false floor and form dense 

layers that obstruct the passage of conditioning air. However, in the past millets were 

malted mixed with sorghum, which gave a more open bed of grain. It was believed that a 

more favourable mixture of enzymes was obtained from the mixture. Unlike sorghum, 

some millets respond to external doses of gibberellic acid. The studies carried out on the 

malting of millets have been in connection with the preparation of foodstuffs (Briggs, 

1998). 

16.3 Brewing African beers on an industrial scale 

The first attempts at brewing African beers on an industrial scale were made in southern 

Africa, using primitive equipment and without fully understanding the principles 

involved (e.g. Novellie, 1968; Oxford, 1926; Schwartz, 1956; Young, 1949). The failure 

rate was high. There have been great improvements in both understanding and brewery 

performance since the early years, and industrial scale brewing is spreading to other 

countries (Harris, 1997). Milled malts and adjuncts (maize grits, whole maize meal, degermed 

sorghum grits, or whole sorghum), may be delivered directly to the brewery. Of 

the several brewing systems in use the most common in South Africa is the Reef Process 

(Daiber and Taylor, 1995; Haggblade and Holzapfel, 1989; Harris, 1997; Novellie and De 


Schaepdrijver, 1986). In one version of this process milled sorghum malt (0.28 t; 

617.3 lb.) is slurried with water (25 hl; 550 imp. gal.) and is held at 48±50 ëC (118.4± 

122 ëF) for 8±18 hours. Older `sour', kept under conditions that maintain the lactic acid 

bacteria in a rapid, logarithmic state of growth, is mixed in to seed the mixture with 

thermophilic Lactobacilli. The process is stopped when the pH falls to 3.3 or less and the 

lactic acid concentration is about 0.8%. 

During this process some amylolysis and proteolysis occur, with increases in sugars 

and FAN. Alternatively, outside South Africa, commercially prepared lactic acid may be 

used to provide acidity. In the next stage, cooking, the sour is mixed with maize grits or a 

sorghum adjunct (2.3 t; 5,071 lb.) and water (168 hl; 3,696 imp. gal.) and the mixture, at 

pH 3.6±4.0, (depending on the product), is boiled for two hours at atmospheric pressure 

or for shorter times under pressure at temperatures up to 110 ëC (230 ëF). During the 

subsequent cooling a little malt may be added, at a temperature of about 80 ëC (176 ëF), to 

thin the material, by beginning to liquefy the starch so that transfer is easier. Water (13 hl; 

286 imp. gal.) and ground malt (0.62 t; 1,367 lb.) is mixed in and conversion takes place 

in this cooled, acidic mash, which is continued at 60 ëC (140 ëF) for two hours, at a pH of 

3.6±3.8. Under these conditions the granular maize and sorghum starch does not 

gelatinize and the activities of the amylases are limited by the low pH so sugar production 

is limited. Proteolysis occurs and the level of FAN increases to a useful extent. The 

thinned mash is then `strained', at 60 ëC (140 ëF), either through a screen, or by 

centrifugation followed by passage through a vibrating screen. The collected strainings 

weigh about 3 t, 6,614 lb., and contain about 1.008 t, 2,222 lb., of dry material, which 

represents about 30% of the grist solids. The screenings are sold, wet or after drying, for 

cattle food. This wasteful process has proved difficult to improve. Although it is possible 

to re-mash the screenings with added enzymes to obtain a secondary wort, this finding 

seems not to have been exploited. The objective of straining is to remove coarse particles 

of more than about 0.25 mm (about 0.01 in.) width. 

After straining the `wort', about 200 hl, 4,399 imp. gal., cooled to 28±30 ëC (82.4± 

86 ëF), is transferred to a stainless steel fermenter of 150 or 270 hl (3,300 or 5,939 imp. 

gal.) capacity, and is pitched with a selected, pure, dried culture yeast (5.5 kg; 12.1 lb.). 

At this stage the wort contains 6±7% fermentable sugars (glucose, maltose and 

maltotriose in the approximate ratio 1:3:1), 3% soluble dextrins, more than 1% of 

gelatinized starch and about 2% ungelatinized starch. The optimal conditions for 

proteolysis are 51 ëC (123.8 ëF) and pH 4.6, and so it is not surprising that FAN increases 

during mashing. At the end of the mash about 30% of the FAN was generated during 

mashing, the other 70% being from the malt and having been generated during souring. 

As malts are prepared with higher DP values the ratio of adjunct to malt tends to increase, 

carrying with it the risk that the mash will contain too little FAN to adequately support 

the yeast growth during fermentation. Perhaps 100 mg FAN/litre is the minimum safe 

concentration, when the FAN is derived from sorghum malt. Fermentation proceeds for a 

selected time (usually 8±24 h), then the actively fermenting beer is sold either on draught 

or in small waxed cardboard or larger polyethylene containers. All containers are vented 

to allow the escape of the continuously generated carbon dioxide. Unfortunately this is 

apt to allow beer to escape as well. 

Compared to `clear beers' opaque beers are viscous (e.g. 15 mPa.s) and are rich in 

fusel oils. Other analyses of commercial South African beers (ranges, with the mean 

values in brackets) are: alcohol, 2.4±4.0% w/w (3%), total solids 2.6±7.2%,w/w (4.9); 

insoluble solids, 1.6±4.3% (2.3); lactic acid, 164±250 mg% (213); pH3.2±3.9 (3.5); 

volatile acids as acetic acid g/100ml, 0.012±0.029 (0.026), total nitrogen 0.065±0.115% 


(0.084) (Novellie, 1968; Novellie and De Schaepdrijver, 1986). Other reported values 

differ significantly. Part of the art in this brewing system is to produce the desired levels 

of nutrients for the lactic acid bacteria and the yeast, to gelatinize some of the starch, but 

not to degrade the starch so much that the beer is `thin'. The residual starch is relatively 

poor in amylose and the side chains of the amylopectin are roughly halved in length 

(Glennie, 1988). 

The iJuba process gives a less viscous beer, which is made to meet the preferences 

of the Zulu people. Starch conversion is carried further in this process. Water (165 hl; 

3,630 imp. gal.), maize grits (1.275 t; 2,811 lb.), ground sorghum grain (0.795 t; 

1,753 lb.) and sorghum malt (`pre-malt'; 0.105 t; 232 lb.), or a microbial amylase, are 

combined and the mixture is heated and finally boiled, for two hours, at its natural pH, 

when enzyme activity ceases, starch is gelatinized and microbes are almost all 

destroyed. The pre-malt or microbial amylase is added to liquefy some of the starch 

and reduce subsequent handling problems. The mix is cooled, to 60 ëC (140 ëF), and 

sorghum malt (1.1 t; 2,425 lb.) is added and, after two hours conversion, which occurs 

more rapidly than in the Reef process, at this `natural' pH, water is added and the mash 

is allowed to sour for around four hours at 50 ëC (122 ëF) to a pH of 3.8±4. The souring 

may be initiated by the addition of a pure culture of Lactobacillus delbruÈckii 

(leichmannii) or part of a previous sour. A small amount of ground malt (0.105 t; 

232 lb) is then added and the mixture is heated to boiling, when the lactic acid bacteria 

are killed and all enzyme activity is terminated. Following this heat treatment the 

mixture is cooled to 40 ëC (104 ëF), more sorghum malt (0.105 t; 232 lb.) is added, and 

the mixture is strained and cooled to 28 ëC (82.4 ëF). Strainings amount to 2.5 t, 5,512 

lb., less than in the Reef process because starch liquefaction is more advanced and so 

the viscosity of the mixture is less, allowing a `cleaner' separation of the coarser 

materials. Active dried yeast (3.5 kg; 7.72 lb.) is added to the wort (about 200 hl; 4,400 

imp. gal.) and fermentation proceeds for 8±48 h, at 28 ëC (82.4 ëF), before the 

fermenting beer is packaged and distributed. In contrast to the Reef process the 

amylases convert the mash at the natural pH, when they are much more active than in 

the acid conditions used in the Reef process mash. In addition the heat treatment 

effectively sterilizes the mash before the last malt addition, straining and pitching. 

The Kimberley style of brewing is intermediate between the Reef and the iJuba styles, 

and was developed to allow brewing with poor-quality sorghum malts with low DP 

values. The first stages involve two process streams. In the `sour stream' sorghum malt 

(0.9 t; 1,984 lb.) is mixed with water (25 hl; 550 imp. gal.) and the mixture is held at about 

49 ëC (120.2 ëF) for 18 h, when the pH falls to 3.2. The microbes are then killed, and 

enzymes are inactivated, by heating the sour to 85 ëC (185 ëF). In the `main' process 

stream maize grits (2.3 t; 5,071 lb.) and water (168 hl; 3,696 imp. gal.) are mixed and 

sometimes some microbial -amylase is added to `thin' the starch as it gelatinizes. The 

mixture is then boiled for two hours, at its natural pH. The cooked material is mixed with 

water (13 hl; 286 imp. gal.) and ground sorghum malt (0.62 t; 1,367 lb.), and conversion 

proceeds for 2 h at 60 ëC (140 ëF) at the `natural' pH, 5.8±6.0. After mixing in the heated 

sour from the other process stream, when the fall in pH checks the activities of the 

amylases, the mixture is strained, yielding 3 t, 6,614 lb., of strainings, and the wort is 

pitched with 5.5 kg, 12.1 lb., of dried yeast. Fermentation and distribution are carried out 

in the usual ways. 

While the three processes described, including minor variations of them, are the most 

usual others are also used (Diefenbach, 1996; Harris, 1997). In some of these processes 

`sours' are replaced with lactic acid. In the split sour Chibuku process maize and sorghum 


grits are cooked together with some of a sorghum malt sour (about 40%), and more sour 

(60%) is added after a conversion stage, which is carried out on the cooked material after 

the addition of sorghum malt and more water. After heating to 80 ëC (176 ëF), at pH 3.7, 

the material is subjected to a second conversion stage, with added fungal amyloglucosidase, 

for 30 minutes at 60 ëC (140 ëF). The amyloglucosidase degrades starch and 

dextrins to glucose. The next stages are straining, cooling and fermentation. 

A predictable problem is continuing activity of the amyloglucosidase in the beer. In 

the Chibuku Zimbabwe/Botswana process -amylase and lactic acid are added to the 

maize and water cook and amyloglucosidase and sorghum malt are used in a second 

conversion stage, the enzymes needed in the first stage being provided by sorghum malt. 

So in this process milled maize (1.8 t; 3,968 lb.) is mixed with water (95 hl; 2,090 imp. 

gal.), lactic acid (8 litres; 1.76 imp. gal., 80%), and amylase. After a boil for 1.5 h at 

pH 5.0, water (60 hl; 1,320 imp. gal.) is added and sorghum malt (0.46 t; 1,014 lb.) is 

added to the cooled mixture and conversion takes place for two hours, at 60 ëC (140 ëF) 

and pH 5.5. After centrifugal straining the wort is pasteurized by heating to 80 ëC 

(176 ëF). The mixture is cooled to 60 ëC (140 ëF) and is incubated with amyloglucosidase 

and more sorghum malt (0.02 t; 44.1 lb.) for 0.5 h. Following this second conversion the 

wort is cooled and pitching and fermentation are carried out, at 26 ëC (78.8 ëF). In the 

Chibuku/Zimbabwe/Malawi process malt is not used at all. Extract is derived exclusively 

from milled maize. Acidity is provided by lactic acid and the enzymes employed are 

microbial -amylase and amyloglucosidase. The product, by usual criteria, is not a beer 

but a beer substitute (Harris, 1997). 

Comparisons between `home' and industrial opaque beer brewing practices and 

between these and the methods used in making `clear beers' are interesting, but opaque 

beer home brewing does not seem to have been studied scientifically. Obvious 

differences are the scales of operation and the near total lack of control, other than the 

brewster's judgement, in home brewing. Industrial brewers have better raw materials of 

more uniform quality so, for example, they specify that malts should have DPs of 28±35 

SDUs and water contents of less that 10%. The more exact temperature control available 

in industrial breweries has advantages at every stage of operation. The souring process is 

under better control, both because by operating at around 50 ëC (122 ëF) only 

thermophilic bacteria are encouraged to grow, and inoculation may be with a pure 

culture of a Lactobacillus, and because pH and acidity can be measured. Not all 

thermophilic bacteria are desirable in the sour. In home brewing the souring occurs in a 

cooling environment, so many different bacteria may grow, thermophiles being 

succeeded by mesophiles, with the generation of less lactic acid and the formation of 

unwanted flavours and aromas. 

In home brewing the souring and alcoholic fermentations are not clearly separated, 

rather they overlap. The lactic acid is said to help the softening of the endosperm during 

cooking, so facilitating the release and gelatinization of the starch granules. Final beers 

should contain both gelatinized starch and ungelatinized granules. Starch gelatinization 

temperature ranges are maize, 62±74 ëC (143.5±165.2 ëF), sorghum, 69±75 ëC (156.2± 

167 ëF), millets, 54±85 ëC (129.2±176 ëF) and barley, 60±62 ëC (140±143.6 ëF) (Briggs, 

1998). Consequently in mashes carried out at 60 ëC (140 ëF), the sorghum amylases will 

not attack granular sorghum, millet or maize starches to any appreciable extent. In several 

of the mashing processes the acidic pH values (pH 3.6±4.0) limit the activities of the 

sorghum amylases ( -amylase optimum pH 4.5±5; -amylase optimum 5.2±5.5; Daiber 

and Taylor, 1995), acting on the gelatinized starch. As with all -amylases, calcium ions 

form an integral part of the sorghum enzyme and inclusion of calcium salts in the 


mashing liquor stabilizes the enzyme and enhances its activity, as is the case with the 

barley enzyme during mashes for making clear beers (Taylor, 1989, 1992). 

Commercially, the alcoholic fermentation is initiated by pitching with pure, top 

fermenting yeast, rather than the mixture of microbes present on the sorghum malt and 

brewing vessels used in home brewing, which give unpredictable fermentations. Even so, 

industrial fermentations contain microbes from the raw materials and these are major 

contributors to spoilage. During the alcoholic fermentation acidification continues and so 

the pH continues to fall. Hetero-fermentative, mesophilic bacteria produce some lactic 

acid and also a range of other products, with adverse effects on flavour. As fermentation 

slackens, and the generation of carbon dioxide declines, air gains more ready access to 

the liquid, oxidative changes can occur and some alcohol is oxidized to acetic acid, which 

confers an unwanted, vinegary flavour and aroma. Yeast may die and autolyse, reducing 

the competition with, and supplying nutrients for, contaminating microbes. Some 

microbes may ultimately form a pellicle on the surface of the beer. Thus spoilage is 

inevitable in several days and this creates major problems for supplying beers over long 

distances and in meeting the variable demands for beer, which peak at weekends and at 

the ends of each calendar month. 

16.4 Attempts to obtain stable African beers 

The irregular and short shelf-lives of African beers, say three days in summer and five 

days in winter in South Africa, create problems of supply and distribution and ensure that 

breweries have excess brewing capacity for much of the time. This they must have to 

allow them to meet the high demands at weekends. In an older type of brewery the shelflife 

of the product may be three days, while in a modern, scrupulously cleaned, dust 

controlled plant, using pure cultures of yeast and perhaps of Lactobacilli, the shelf-life 

may still be only five days. Various attempts have been made to produce stable beers but 

with only limited success. Experimentally, a stable form of Kenyan beer was prepared by 

pasteurization and a clear Nigerian beer, oyokpo, was stabilized with benzoic acid 

(Daiber and Taylor, 1995; Iwuagwu and Izuagbe, 1985; Nout, 1980). Most studies 

concern southern Africa (Harris, 1997; Haggblade and Holzapfel, 1989; Novellie and De 

Schaepdrijver, 1986). Chilling the wort to 14±16 ëC (57.2±60.8 ëF) before pitching with 

yeast slows the fermentation, and lengthens the shelf-life of the beer. However, this 

approach has problems as the product is consumed warm by choice and refrigerated 

transport and dispensing equipment is not available. Storing under a top pressure of 

carbon dioxide has been proposed, but apparently this is not used. Heavily -irradiating 

the malt greatly reduces the population of microbes and increases the shelf-life of beer 

from about four to six days. Again, this process seems not to be used. 

Successful ways around the instability problem involve the use of beer powders or 

concentrated worts. Techniques using pasteurization of worts or beers have not worked so 

well. Beer powders consist of dry, finely milled sorghum malt, dry yeast and pre-cooked 

maize meal. The cooking is by steam injection, which gelatinizes at least some of the 

starch. Some formulations may also contain some acidified material. `Brewing' consists 

of dispersing the powder in warm water. Conversion, acidification (by bacteria from the 

malt) and alcoholic fermentation begin rapidly and simultaneously. The product is ready 

for consumption in 4±8 h and has a shelf-life of about one day. This process resembles the 

production of the millet beer, busaa (Nout, 1980). The beer is of poor quality. In part this 

is because initially the acidity is low, giving spoilage organisms a chance to multiply 


efore the pH falls. The powder is stable and so can easily be transported and stored, 

allowing beer to be produced in remote locations, at short notice and in small or large 

amounts. Another approach is to make wort, with a low solids content, in an approved 

fashion and then, after straining, concentrating it in a film evaporator to around 50% 

solids (Harris et al., 1999). It is possible to spray-dry this material. The concentrated wort 

is syrupy and contains solids. It is relatively stable, even to spoilage by osmophilic yeasts. 

Because it is concentrated and stable it can be widely distributed relatively easily. To 

brew syrup, 5 kg (11 lb.), is diluted to 25 litres (5.5 imp. gal.) and is pitched with brewing 

yeast. The product can be a high-quality beer. 

Both batch- and tunnel-pasteurization treatments failed, partly because the heating 

gelatinized the granular starch in the beer, making it excessively viscous. Partial 

microbiological control was obtained by flash-pasteurizing wort at 72±76 ëC (161.6± 

168.8 ëF) for 12±15 s using plate heat exchangers. Viscosity and flavour were scarcely 

altered, but because of the solids present complete sterilization was not achieved. Beer 

could be flash-pasteurized more successfully, at 75±80 ëC (167±176 ëF) for 20±25 s. 

However, all the carbon dioxide was removed and so the beer was `flat', that is, it lacked 

effervescence and `tingle'. Carbonating this product failed, apparently because of 

difficulties caused by the high solids content. This flat beer has limited acceptability but 

some is sterile packaged and sold. Maintaining sterility has been difficult but where it has 

held the product has a long shelf-life. Another approach has been to bottle pasteurized 

beer with added yeast and a calculated amount of sugar, (a process closely similar to the 

traditional British process of conditioning ìn bottle' with added priming sugars and a 

secondary yeast). In the South African process ìn bottle conditioning' occurs as the yeast 

ferments for 5±12 days and the pressure of carbon dioxide rises to 20±30 psi, being 

limited by the amount of sugar added. The beer has an unusual flavour but it is acceptable 

and relatively stable (Harris, 1997). Bottles are said to have burst, possibly because of 

over-dosing with sugar, permitting an excessive accumulation of carbon dioxide, with a 

consequent excessive rise in pressure. 

16.5 Beer composition and its nutritional value 

Opaque beers are valuable foodstuffs, and are the best of all the alcoholic beverages in 

this respect. However, there is no reliable way to decide exactly how important they are 

in practice. There are several reasons for this. Home-brewed beers almost certainly vary 

very greatly in their compositions (but these are unknown) and the natures and 

proportions of the raw materials used, the quantities of beer drunk, when (consumption 

varies with the season) and by whom, and the nutritional status of the drinkers all vary 

(Daiber and Taylor, 1995; Haggblade and Holzapfel, 1989; Heerden, van, 1989a, b; 

Novellie and De Schaepdrijver, 1986). Many years ago it was noted that in some seasons 

the men of particular tribes lived on beer alone for extended periods. As there is a 

gradation between soured, acidified, essentially alcohol-free porridges and acidified 

beers, and as the alcohol contents of the beers are variable and are continually increasing 

as fermentation continues, no conclusions regarding alcohol consumption can be drawn. 

With commercial beers the alcohol contents, at the time of sale, are relatively low 

(generally 3% or less) compared to clear beers (3±9%) and very much less than in wines 

(11±13%) or spirits (often 40%). 

The nutrients present in beers, particularly the vitamins, are derived from the raw 

materials, Lactobacilli and the yeast. Some of the materials initially present in the wort 


are taken up by the Lactobacilli and the yeast and may be converted into different 

substances. Some nutrients are destroyed by the brewing process, while those that are 

present in the beer may not be `bio-available' and cannot be used by the consumer. The 

compositions of several commercial beers have been reported. Ranges (and mean values) 

for eight beers were: energy content, in kJ/litre, 1,530±1,840 (1,651); alcohol content, in 

g/litre, 16±32 (25.4); crude protein, g/litre, 4±9 (5.4); fat, g/litre, trace 1 (trace); crude 

fibre, g/litre, trace 1 (trace); ash, g/litre, 1±2 (1.13); carbohydrate, g/litre, 32±59 (47.6). 

Starch contents of five beers ranged between 27.8 to 32.7 g/litre (mean 29.7) and the 

values for three vitamins in 15 beers, all in mg/litre, were thiamine 0.13±0.36 (0.24); 

riboflavin, 0.30±0.47 (0.39) and for nicotinic acid, 2.32±3.74 (2.93) (Heerden, van, 

1989b). Some contents of mineral elements, ranges and mean values all in mg/litre were 

potassium, 145±438 (280); sodium, 7±46 (21); calcium, 22±57 (38); magnesium, 69±174 

(111); copper, 0.11±0.26 (0.18); iron, 0.9±2.1 (1.4); manganese, 0.8±2.2 (1.4) and zinc, 

1.0±1.9 (1.4) (Novellie and De Schaepdrijver, 1986). Phosphorus contents, in other 

estimates on five beers, in mg/litre, were 96±565 (218). 

Although particular substances are present in beers it does not follow that they are 

available to the imbiber. The living yeast in opaque beers takes up many vitamins quickly 

and these, together with those already present in the micro-organisms, are then largely 

unavailable. Heat treating or pasteurizing the beer damages or kills the yeast and the 

vitamins are released into solution and are then available. Analysis of beer crude protein 

shows that a significant amount is present, with a `plant-like' distribution of amino acids, 

which do not contain an ideal mixture of nutritionally essential amino acids but which, 

nevertheless, has a good proportion of the essential amino acid lysine, which is probably 

derived from the yeast. As the lysine may be largely contained within the yeast cells, its 

availability is in doubt. 

Some phosphorus could be present as phytic acid (meso-inositol hexaphosphate, 

(4.156)), a substance with strong chelating properties that can limit the nutritional 

availability of metal ions. However, and in contrast to many other cereal-based products, 

phytate was not detected in significant amounts in these beers. The replacement of 

sorghum grain adjunct with refined maize grits (from which the nutrient-rich embryos 

and aleurone layers have been removed) in both commercial and home brewing, and the 

progressive reduction in the proportion of sorghum malt used in the grist in commercial 

brewing (facilitated by the increasing DP values of sorghum malts and the 

supplementation of these malts with microbial amylases) are changes which have 

reduced the nutritional values of beers, in particular the B vitamin contents, and have led 

to the proposal that maize grits should be replaced with the traditional whole sorghum 

grain and the beers be supplemented with vitamins, thiamine and vitamin C (ascorbic 

acid; Heerden, van, 1989 a, b). Home-brewed beers, made with higher proportions of 

sorghum malt than are used in commercial brewing, are probably nutritionally superior. 

In addition to the B vitamin content these beers almost certainly contain higher levels of 

desirable crude fibre since the `home' straining procedures are less stringent than those 

used in commercial brewing. Probably each litre of commercial beer consumed can 

provide 10% of the daily recommended protein requirement and 14% of the energy 

requirement of a moderately active man and is a useful source of the B vitamins thiamine, 

riboflavin and nicotinic acid as well as the minerals iron, zinc, manganese, magnesium 

and phosphorus. The starch is a good energy source but the alcohol is less good. Drinkers 

usually consume about two litres (3.52 imp. pints) of opaque beer/day. 

Specifications for commercial beers are likely to include values for total solids, crude 

protein and lactic acid contents, pH, alcohol content when sold, (3% or less in South 


Africa), and an upper limit on volatile acids (as acetic acid). In addition the products must 

be free of pathogenic organisms, must not have a tendency for the solids to separate and 

precipitate, have an appropriate viscosity, appearance, (colour, opacity and foam), and be 

acceptable as judged by a taste panel. 


BRIGGS, D. E. (1998) Malts and Malting. Aspen Publishing, Inc., Gaithersburg, MD, 796 pp. 

CHEVASSUS-AGNES, S., FAVIER, J. C. and ORSTOM, A. J. (1976) Cahiers Nutr. Diet., 11 (2), 89. 

DADA, L. O. and DENDY, D. A. V. (1987) Tropical Sci., 27, 101. 

DAIBER, K. H. (1975) J. Sci. Fd. Agric., 26, 1399. 

DAIBER, K. H. (1978) Special Report BB114. Manual on the treatment and malting of birdproof grain 

sorghum. The Sorghum Beer Unit, CSIR, South Africa. 12 pp. 

DAIBER, K. H. and TAYLOR, J. R. N. (1995) in Sorghum and Millets: Chemistry and Technology. (Dendy, 

D. A. V. ed.), A. A. C. C. St Paul, Minn. p. 299. 

DAIBER, K. H., MALHERBE, L. and NOVELLIE, L. (1973) Brauwissenschaft, 26 (7), 220, 248. 

DENDY, D. A. V. (ed.) (1995) Sorghum and Millets; chemistry and Technology. Amer. Assoc. Cereal 

Chemists. St. Paul, Minn. 

DEWAR, J. and TAYLOR, J. R. N. (1995) Proc. 5th Sci. Tech. Conv. Inst. of Brewing, (Central and Southern 

African Sect.), Victoria Falls, p. 93. 

DEWAR, J., OROVAN, E AND TAYLOR, J. R. N. (1997a) J. Inst. Brewing, 103, 283. 

DEWAR, J., TAYLOR, J. R. N. and BERJAK, P. (1997b) Proc. 6th Conv. Inst. of Brewing (Central and 

Southern African Sect.), Durban, p. 29. 

DEWAR, J., TAYLOR, J. R. N. and BERJAK, P. (1997c) J. Cereal Sci., 26, 129. 

DEWAR, J., TAYLOR, J. R. N. and BERJAK, P. (1997d) J. Inst. Brewing, 103, 171. 

DEWAR, J., DONALDSON, S. and TAYLOR, J. R. N. (1999) Proc. 7th Conv. Inst of Brewing (African Sect.), 

Nairobi, p. 217. 

DIEFENBACH, M. (1996) Brauwelt Internat., 14 (5), 431. 

DIRAR, H. A. (1978) J. Food Sci., 43, 1983. 

EKUNDAYO, J. A. (1969) J. Food Sci., 4, 217. 

EZEOGU, L. I. and OKOLO, B. N. (1994) J. Inst. Brewing, 100, 335. 

EZEOGU, L. I. and OKOLO, B.N. (1995) J. Inst. Brewing, 101, 39. 

EZEOGU, L. I. and OKOLO, B. N. (1999) J. Inst. Brewing, 105, 49. 

FAPARUSI, S. I. (1970) J. Sci. Food Agric., 21, 79. 

FAPARUSI, S. I., OLOFINBOBA, M. O. and EKUNDAYO, J. A. (1973) Zeits. f. Allg. Mikrobiologie, 13 (7), 563. 

FOX, F. W. (1938) J. S. African Chem. Inst., 21, 39. 

GLENNIE, C. W. (1983) J. Agric. Food Chem., 31, 1295. 

GLENNIE, C. W. (1988) Starch/StaÈrke, 40 (7), 259. 

HAGGBLADE, S. and HOLZAPFEL, H. (1989) in Industrialization of Indigenous Fermented Foods 

(Steinkraus, K. H. ed.). Marcel Dekker Inc. New York, p. 191. 

HARRIS, R. N. (1997) Proc. 6th Conv. Inst. of Brewing (Central and Southern African Sect.), Durban, 

p. 89. 

HARRIS, R. N., MATHIBA, K., JOUSTRA, S. M., VILJOEN, C. R. and YENKETSWAMY, C. (1999) Proc. 7th Conv. 

Inst. of Brewing (African Sect.), Nairobi, p. 218. 

HEERDEN, I. V. VAN (1989a) J. Inst. Brewing, 95, 17. 

HEERDEN, I. V. VAN (1989b) Proc. 2nd Sci. Tech. Conv. Inst. of Brewing (Central and Southern African 

Sect.), Johannesburg, p. 293. 

IWUAGWU, Y. O. U. and IZUAGBE, Y. S. (1985) J. Appl. Bacteriol., 59, 487. 

MIRACLE, M. P. (1965) in Food Technology the World Over (Peterson, M. S. ed.). Avi Publishing, 

Westport, Conn. p. 107. 

MORCOS, S. R., HEGAZI, S. M. and EL-DAMHOUGY, S. T. (1973) J. Sci. Food Agric., 24, 1157. 

MORRALL, P., BOYD, H. K., TAYLOR, J. R. N. and VAN DER WALT, W. H. (1986) J. Inst. Brewing, 92, 439. 

NOUT, M. J. R. (1980) Chem. Mikrobiol. Technol. Lebensm., 6, 137, 174. 

NOUT, M. J. R. and DAVIES, B. J. (1982) J. Inst. Brewing, 88, 157. 

NOVELLIE, L. (1966) Internat. Brew. Distill., 1 (1), 27. 

NOVELLIE, L. (1968) Wallerstein Labs. Communs., 31, 17. 

NOVELLIE, L. (1977) in Proc. of a Symp.on Sorghum and Millets for Human Foods. (Dendy, D. A. V. ed.), 

The Tropical Products Institute, London. p. 73. 

NOVELLIE, L. and DE SCHAEPDRIJVER, P. (1986) in Progress in Industrial Microbiology. 23. Microorganisms 

in the production of food. (Adams, M. R. ed.), Elsevier, Amsterdam, p. 73. 

NZALIBE, H. C. and NWASIKE, C. C. (1995) J. Inst. Brewing, 101, 345. 


OGUNDIWIN, J. O. (1977) Brew. Distill. Internat., 7 (6). 40. 

OKOLO, B. N. and EZEOGU, L. I. (1996) J. Inst. Brewing, 102, 79, 277. 

O'ROURKE, T. (2001) Brewer Internat., 1 (10), 46. 

OXFORD, T. (1926) J. Inst. Brewing, 32, 314. 

PANASIUK, O. and BILLS, D. D. (1984) J. Food Sci., 49, 791. 

PATHIRANA, R. A., SIVAYOGASUNDARAM, K. and JAYATISSA, P. M. (1983) J. Food Sci. Technol., 20 (3), 

108. 

PERISSE, J., ADRIAN, J., REVET, A. and LE BARREÂ , S. (1959) Ann. de la Nutrit. et de l'Aliment., 13 (1), 1. 

PETERSON, M. S. and TRESSLER, D. H. (1965) Food Technology the World Over. 2. South America, Africa, 

Middle East and Asia, Avi Publishing, Westport, Conn. p. 130. 

PICKERELL, A. T. W. (1986) J. Inst. Brewing, 92, 568. 

SCHWARTZ, H. N. (1956) J. Sci. Food Agric., 7, 101. 

TAYLOR, J. R. N. (1989) Proc. 2nd Sci. Tech. Conv. Inst. of Brewing (Central and Southern African Sect.), 


TAYLOR, J. R. N. (1991) Proc. 3rd Sci. Tech. Conv. Inst of Brewing (Central and Southern African Sect.), 


TAYLOR, J. R. N. (1992) J. Amer. Soc. Brew. Chem., 50, 13. 

TAYLOR, J. R.N. and DEWAR, J. (1994) J. Inst. Brewing, 100, 417. 

YOUNG, R. S. (1949) J. Inst. Brewing, 55, 371. 


17 

Microbiology 


Thedesignofthebreweryplantandoperationoftheprocessmustbesuchthatproduction 

yeast strains remain segregated with no possibility of inter-mixing. In addition, 

contamination with foreign micro-organisms must be prevented. These goals can be 

achieved by ensuring high standards of hygiene within the brewery. Brewery plant is 

constructedfromstainlesssteelforeaseofcleaning.Themodernprocesstendstobefully 

enclosed to ensure that amicrobiological barrier is maintained between process liquids 

and the external environment. All critical parts of the plant are fitted with automatic and 

efficient cleaning in place (CIP) systems. The microbiological integrity of the brewing 

process must be confirmed with appropriate testing. For acomplex process such as 

brewing this necessitates the adoption of asampling plan to ensure that all stages are 

checked where there is arisk of introduction of contaminants. The samples must be 

representative of the process stream they are taken from. It follows that the sampling 

devices must be fit for this purpose. Analysis of the samples may use classical 

microbiological techniques, in other words, inoculation into a suitable medium, 

incubation and scoring for growth. This `classical' approach is a valuable aid to 

validating the microbiological integrity of the process. Several selective media are in 

common usage, which have been developed specifically for the isolation and 

identification of brewery contaminants. Several days are usually required before aresult 

is obtained, therefore the data is `historical' and cannot be used for immediate hygiene 

control. Nevertheless, routine microbiological testing is valuable when used for trend 

analysis. 

Microbiological testing of some samples must cope with the presence of production 

yeast strains. Methods for the identification and differentiation of brewing strains are 

described in Chapter 13 (Section 13.10). Several selective media have been devised 

which allow the identification of bacteria and non-brewing, so-called wild yeast 

contaminants in the presence of high concentrations of brewing yeast. Validation of 

cleaning cannot be usefully checked with conventional microbiological methods since 

these are too slow for production requirements. Rapid procedures have been developed to 


meet this need. These have been designed to detect instantly the presence of microorganisms 

and soiling. 

17.2 The microbiological threat to the brewing process 

Micro-organisms may exert adverse effects on the brewing process both directly and 

indirectly. The direct effects are the obvious ones of contamination of wort or beer with 

foreign organisms. Oxygenated wort represents acomparatively rich source of nutrients 

capable of supporting the growth of awide range of micro-organisms. The presence of 

hops is advantageous since trans-humulone, ( )-humulone and colupulone are inhibitory 

to many bacteria by virtue of their ability to act as ionophores (Verzele, 1986). 

Nevertheless, many micro-organisms, including yeast, are capable of growth in their 

presence. The effects of contamination range from comparatively minor changes in beer 

flavour and fermentation performance through to gross flavour defects and superattenuation 

of worts. Once pitched the wort is, to some extent, protected by the yeast 

since many contaminants, if present at low levels, are not able to compete. Beer is a 

comparatively poor growth medium. The nutrients are limited to the small residue that 

remains after fermentation is completed. Beer is arelatively hostile environment to many 

micro-organisms. The antiseptic properties of hop compounds are augmented by ethanol. 

Low redox and acid pH provide additional protection against many potential spoilage 

organisms. Ethanol is apowerful inhibitor of microbial growth. Low and zero alcohol 

beers have a much increased susceptibility to spoilage compared to their alcoholic 

counterparts. 

Several bacterial and some yeast species are capable of growth in beer. This can cause 

the formation of hazes, surface pellicles and many undesirable changes in beer flavour 

and aroma. The outward symptoms of these infections have been long recognized and 

many are characterized as `diseases' of beers. These are usually descriptive of the 

changes in flavour and appearance. 

Micro-organisms exert indirect undesirable effects on brewing in three ways. Firstly, 

growth on raw materials can produce undesirable changes such that the materials do not 

behave normally. Secondly, the growth of contaminants on raw materials can generate 

microbial metabolites, which can persist into the brewing process and exert deleterious 

effects. Thirdly, very heavily contaminated raw materials can introduce microbial 

biomass that persists into green beer. Although dead, the cells can cause beer filtration 

problems and even beer hazes if filtration is deficient. 

From a microbiological standpoint, the brewing process is divisible into the steps 

leading up to wort production followed by those that include fermentation and subsequent 

beer processing. The copper boil separates these two parts. This process step serves many 

functions, one of which is to sterilize wort (Chapters 9, 10). It follows that some 

microbiological contamination can be tolerated in the process steps preceding the copper 

boil,although withthe caveat thatthe raw materials must bewithinspecification. The steps 

after the copper boil are those in which the risk of contamination is highest and where the 

greatest caution must be exercised. Any brewing raw material that is capable of supporting 

microbial growth has the potential to produce unwanted metabolites that can persist 

through the brewing process and produce adverse effects. Water is aspecial case in that 

microbial metabolites can be introduced even though the organism does not have direct 

contact with any other brewing materials. To counteract this threat all water, especially 

from wells, should be carbon filtered to remove organic contaminants (Chapter 3). 


Sugar syrups may become tainted as aresult of the growth of osmophilic moulds and 

yeasts (Chapter 2). Growth on the surface of the syrup can occur where inappropriate 

storage conditions allow the formation of condensation resulting in alocalized sugar 

dilution and lower water activity. Growth of contaminants may proceed since the 

inhibitory effects of high osmotic potential are reduced. This should be guarded against 

by storage of syrups at an appropriate temperature and preferably under an inert gas such 

as carbon dioxide or nitrogen. 

Malt and adjuncts derived from cereals present the greatest threat. Poor control of the 

steps involved in the manufacture of these ingredients can result in mould growth. 

Species ofmoulds from generasuch as,Alternaria,Aspergillus, Cladosporium, Fusarium 

and Rhizopus have all been reported to produce adverse effects (Flannigan, 1999). 

Excessive mould growth produces many metabolites that can produce off-flavours and 

aromas in beers. Terms such as molasses, stale, burned and winey have all been used to 

describe the effects. In addition, changes in colour may also occur. The metabolites 

producing these effects also result in increased nitrogen levels in worts and beers. In 

extreme cases, beer hazes may be generated. 

The most widely recognized defect ascribed to the growth of mould on malts is that of 

gushing. This phenomenon occurs in bottled beers where on broaching there is asudden 

loss of carbon dioxide with concomitant uncontrolled foaming. Studies have demonstrated 

that culture filtrates of several moulds, especially Fusarium spp, were capable of 

inducing gushing when added to beers (Amaha et al., 1974; Kitabatake and Amaha, 

1974). Small polypeptides have been isolated that are apparently responsible for the 

phenomenon. In one case aconcentration as low as 0.05ppm was sufficient to produce 

the effect (Kitabatake and Amaha, 1974). Moulds capable of producing gushing-inducing 

metabolites are commonly those that also produce mycotoxins. Indeed, some, but not all, 

mycotoxins have been shown to be capable of inducing gushing. More than 200 distinct 

mycotoxins have been isolated from various fungi. They appear to function as facilitators 

of fungal pathogenesis. The most common are the trichothecenes of which around 150 

have been recognized. Chemically, they are tetracyclic sesquiterpenes, the most common 

being nivalenol, deoxynivalenol and T-2 toxin. All are potent inhibitors of protein 

synthesis and possibly they disrupt membrane function. Trichothecenes are heat stable 

and therefore capable of surviving through the wort boil. They are toxic to humans and 

animals. Purely from abrewing standpoint, if present at high concentration they inhibit 

yeast growth. It has been suggested that they could in some circumstances be acause of 

sticking fermentation (Boeira et al., 1999a,b). 

After wort boiling the microbiological integrity of the process is dependent upon good 

hygienic practice. The efficiency of CIP systems is of paramount importance to ensure 

that contamination is not introduced by unclean plant. Uninoculated wort, either in 

fermenter or propagater is at the greatest risk. This must be pitched as soon as possible 

afteritsproductioninordertominimizetheriskofinfection.Althoughbeerisarelatively 

poor substrate, asubstantial range of micro-organisms are capable of growth in it. It 

follows, therefore, that after wort, bright beer represents the second most vulnerable 

material. Entry of microbial contaminants can occur at any stage where liquid or gaseous 

additions are made to primary process streams. Some of these are illustrated in Fig. 17.1. 

At the end of the process, the beer must be packaged in a way that renders it 

microbiologically stable throughout its expected shelf-life. In the case of small-pack 

products, in bottle or can, the most common option is to use a tunnel pasteurization 

process. The temperature and contact time must be controlled to ensure that each 

individual package receives the desired heat treatment. This requires careful design and 


Laboratory 

propagation 

Brewery 

propagation 

Break 

down 

liquor 

O2/air 

Conditioning tank 

Filtration 

Bright beer tank 

Clean fill Sterile fill 

Tunnel 

pasteurizer 

Small pack 

beers 

Malt Liquor 

Hopped wort 

Wort collection Additives 

Fermentation 

Green beer 

Bulk beer 

tanking 

Flash 

pasteurizer 

Acid 

washing 

Yeast cropping 

Racking 

tank 

Cask 

Keg Dispense 

Pitching 

yeast 

Primings 

Finings 

Other additives 

Fig. 17.1 Outline of the complete brewing process indicating steps in which there is a potential for 

microbiological contamination. 

operation in the event of a packaging line stoppage. It is common practice in this situation 

to cool the pasteurizer to safeguard product already inside against heat degradation due to 

over-pasteurization. When product flow recommences, it is essential to delay forward 

movement of product within the pasteurizer until correct operating temperatures are 

attained. 

For certain small-pack beers where the container cannot withstand heat and for keg 

beers the product is pasteurized in-line. This process is as efficacious as tunnel 

pasteurization, however it introduces extra risks. Thus, contamination is possible from the 

container or from the plant situated between the pasteurizer and package. Special 

precautions must be taken to ensure that this does not occur. Some flavour degradation is 

inevitable whenever beer is heated. To eliminate this it is becoming more common to 

package aseptically. In this case the in-line pasteurization step is replaced by sterile 

filtration. The equipment must be scrupulously clean and operated to the highest 

standards of hygiene. 

In the case of draught beers, the possibility of microbiological spoilage extends 

beyond the brewery and into retail establishments. Beers in cask are to some extent 

protected by the endogenous flora of brewing yeast. Nevertheless, casks are vented to the 


atmosphere for dispense and therefore open to the entry of contaminants. In awellmanaged 

cellar this should not occur. Dispense systems for cask and keg beers are 

possible routes for contamination. In particular, the possibility of biofilm development 

must be guarded against by the use of appropriate cleaning regimes. It is essential to 

consider the microbiological implications of introducing new products, plant and 

processes into the brewery. Introduction of anew raw material has the potential toimport 

awhole new spectrum of microbial contaminants, which have not been encountered 

hitherto. This is of special note where the production of beverages other than beers is 

introduced into abrewery. For example, the use of non-sterile fruit concentrates, which 

are ingredients in flavoured alcoholic beverages. If these are used in the same plant as 

beer, great caution should be exercised to ensure that the coexistence of these distinct 

productstreamsis microbiologicallyrobust.Inparticular,it should be realized that media 

designed for the detection of typical brewery contaminants may not be suitable for nonbrewing 

micro-organisms. 

Itisimportanttoconsideranymicrobiologicalimplicationswheremodificationstothe 

brewingprocessaremade.Amove from pasteurization tosterile filling heightensthe risk 

of microbiological failure and the precautions to prevent this need to be made 

correspondingly more stringent. However, in some instances apparently unrelated 

changes can have unexpected consequences. For example, on quality grounds there has 

been a gradual tightening in specification regarding maximum dissolved oxygen 

concentrations in product both in process and package. This decreases the overall risk of 

spoilage by preventing the growth of obligate aerobes. On the other hand, it provides a 

better selective medium for obligate anaerobes. In fact, bacterial infections of beer by 

anaerobes such as Pectinatus have been recorded only in relatively recent times and are 

taken to reflect the gradual reduction of in-process oxygen exposure (Section 17.3.3). 

17.3 Beer spoilage micro-organisms 

Beer is liable to spoilage by arange of micro-organisms, both bacteria and yeasts. 

Spoilage results in the formation of hazes, undesirable flavours and aromas. Although 

beerisrenderedunpalatable,thegrowthofcontaminantsdoesnotgenerallyleadtohealth 

risks. There are some exceptions to this general statement, as discussed subsequently, 

however pathogenic micro-organisms do not survive in beer. 

17.3.1 Detection of brewery microbial contaminants 

Routine microbiological testing in the brewery is usually restricted to enumerating 

populations. In many situations, no contamination whatsoever should be detected. In 

other cases, some contamination is inevitable. Maintaining arecord of the numbers of 

micro-organisms detected provides auseful method of assessing the general cleanliness 

of the brewery environment, the robustness of cleaning regimes and the microbiological 

integrity of the process. Methods must be capable of detecting low concentrations of 

contaminants in isolation (as in bright beer) or in the presence of high concentrations of 

othermicro-organisms(asindetectionoflowlevelsofbacteriainpitchingyeastslurries). 

In traditional practice, microbial contamination isdetected by taking asuitable sample 

from the brewery and inoculating it into an appropriate solid or liquid microbiological 

medium. Appropriate sampling devices and procedures have been developed for the 

routinetestingofprocess liquids,gases andsurfaces.These are describedinSection 17.5. 


After incubation, to allow any micro-organisms togrow to adetectable concentration, the 

cultures are examined for the presence or absence of growth. Growth can be detected via 

the visible formation of hazes in liquids or as colonies on solid media. Many selective 

media are used that contain components that allow the growth of specific strains and not 

others. Descriptions of common microbiological media used in brewing are provided in 

Section 12.3.5. 

Growth can be quantified by performing cell counts. This may be via direct 

microscopic enumeration using a counting chamber. Microscopic examination of 

contaminants provides a useful method for preliminary identification. More commonly, 

cell concentrations are determined by making colony counts where serial dilutions of the 

test sample are streaked out into solid media or collected on a membrane which is then 

placed on solid medium. After incubation under appropriate conditions, any colonies that 

arise are assumed to have formed from single cells. Therefore, the colony count is 

directly proportional to the cell concentration in the original sample. 

Conventional microbiological techniques will continue to have a place in brewery 

laboratories because they are relatively inexpensive and do not require sophisticated 

apparatus. They suffer the drawback of slowness and produce data that is of historical 

interest only. More rapid techniques are needed for validating the microbiological 

integrity of the brewing process in real time. Rapid methods are of two general types 

(reviewed by Russell and Dowhanick, 1999). Firstly, those that require a growth stage 

before the organisms can be detected. Secondly, those able to detect very low levels of 

contamination in samples without the need for pre-treatment. All rapid detection methods 

rely on three general principles, used alone or in combination, for their operation. These 

are pre-concentration, low threshold of detection and specificity. Some rapid methods, 

particularly those with a high degree of specificity, combine elements of detection and 

identification. These are described below. 

Low levels of contaminants in samples containing little extraneous solid material can 

be concentrated by filtration through a sterile membrane filter. The approach can be 

applied to both liquids and process gases. The membrane must be sufficiently porous to 

allow throughput of a large sample volume but have a pore size small enough to retain 

bacteria. Typically, 0.22 or 0.45 m membrane filters are used for this purpose. The 

membrane is transferred to a Petri dish containing a suitable solid nutrient medium, 

incubated and examined for growth. The procedure can be made rapid by examining the 

membrane under a microscope and looking for micro-colonies. Commonly, membranes 

are stained to aid visualization of the colonies. Preferably fluorescent stains are used that 

are incorporated into the medium. This approach has the advantage that viable and nonviable 

cells can be differentiated and cells can be recovered for further analysis. The 

micro-colony method can produce a result within approximately 24 hours, thus saving 

several days compared to traditional techniques. 

The direct epifluorescent filter technique (DEFT) uses a membrane filtration step to 

concentrate low concentrations of contaminants. No pre-growth stage is required because 

any cells trapped on the membrane are visualized by microscopic examination after 

staining with a fluorescent dye, usually acridine orange. This dye binds to single-stranded 

RNA molecules, which are plentiful in viable cells, and produces orange fluorescent 

cells. Dead cells are deficient in single stranded RNA and these stain green due to the 

natural fluorescence of double-stranded RNA. Other dyes have also been used such as 

berberine sulphate (in conjunction with acridine orange), aniline blue, Viablue and some 

tetrazolium salts. Reportedly, these produce improved staining reactions, which make 

easier differentiation between microbial cells, inanimate debris and the membrane. The 


procedure can be automated using computer-assisted image analysis. Results can be 

obtained within 30 minutes. The sensitivity of the approach is such that a single microbial 

cell can be detected on a membrane. 

Procedures that rely on pre-incubation and a low threshold of detection use indirect 

methods for detecting microbial growth. In clear liquid media the presence of microbial 

cells can be detected using turbidometry or spectrophotometry. Using appropriate 

apparatus changes can be detected before visual turbidity becomes apparent. The growth 

of micro-organisms is an exothermic process. Apparatus has been developed that detects 

growth-related exothermy as an electrical current via the intermediary of a thermocouple. 

Impedometric devices detect changes in electrical properties of the medium brought 

about by microbial growth. As growth proceeds, the uptake of nutrients and the formation 

of extracellular ionic metabolites result in a change in the conductance and capacitance of 

the medium. Commercial apparatus has been developed capable of measuring these 

changes and thereby rapidly detecting microbial growth (for a review, see Fleet, 1992). 

The time required to reach the threshold of detection of microbial growth using these 

methods is dependent upon the size of the inoculum and the growth rate of the organism 

therefore the time required to obtain a result is dependent on the nature of the sample and 

the concentration and identity of the contaminants. The procedures can be used in forcing 

tests where the undiluted sample is assessed, for example, analysis of bright beer. 

Alternatively, samples may be inoculated into a suitable growth medium. The presence of 

low levels of bacterial contamination in the presence of high concentrations of yeast, for 

example in pitching slurries, can be accommodated by the incorporation of 

cycloheximide to inhibit yeast growth. Compared to traditional microbiological 

techniques the rapid procedures use expensive and relatively sophisticated apparatus. 

A decision must be made as to whether or not the size of the brewing operation merits the 

required capital investment. 

All living organisms contain adenosine triphosphate (ATP). Therefore, the presence of 

ATP is indicative of biological activity. ATP is easily detected using the phenomenon of 

bioluminescence (Simpson, 1999). The firefly, Photinus pyralis contains an enzyme, 

luciferase. In the presence of oxygen and ATP, luciferase catalyses a reaction in which a 

substrate termed luciferin (6-hydroxybenzothiazole) is oxidized to oxyluciferin. During 

the reaction ATP is hydrolysed to AMP. For each mole of ATP hydrolysed, a photon of 

light with an emission maximum of 532 nm is released. 

ATP ‡ luciferin ‡ O2 ! AMP ‡ oxyluciferin ‡ CO2 ‡ PPi ‡ light 

The bioluminescence reaction has been utilized for many years as the basis of an ATP 

determination. It is now routinely applied to the detection of ATP in industrial 

environments as a means of validating cleaning regimes (Ogden, 1993). Early devices 

relied on sampling using a sterile swab followed by rinsing into a cuvette, addition of 

appropriate reagents and insertion into a bioluminometer to obtain a reading. These 

approaches needed skilled laboratory personnel. More modern versions utilize all-in-one 

arrangements, where sampling device and reagents are incorporated into a single unit. 

Samples may be taken in the form of swabs to check the cleanliness of surfaces or as 

liquid taken from terminal rinse water. The measuring unit usually includes a data capture 

system to facilitate the maintenance of records. Operation does not require any skill other 

than the ability to follow simple instructions. 

Several commercial ATP bioluminometers are now available. Unfortunately, there has 

been no agreement to standardize the results of the measurement. Arbitrary values of 


ioluminescence, termed relative light units (RLU), are used. The output from individual 

instruments differs for a constant ATP concentration. It is necessary, therefore, to 

calibrate and set standards for each instrument. Typically, a range of target values in 

RLUs is established for each location under test. The lower limit is classed as a pass, an 

intermediate value indicates caution and should prompt a check of the CIP system. A 

value above an upper limit defines a fail and indicates the need for a check of the CIP 

system and repeat clean. 

Correlation between luminescence and microbiological counts is usually poor because 

ATP concentration varies widely between different species and the same species in 

different physiological states. In general, yeast cells contain around 100 times more ATP 

than bacterial cells (Hysert et al., 1976). In brewing the situation is more confusing since 

beer also contains significant and very variable levels of ATP. Thus, Simpson et al. 

(1989) reported mean levels of ATP in beer of 5 nM but with a range of 0.01 to 100 nM. 

For this reason, bioluminometry cannot differentiate between soil and microbial 

contamination. It also follows that it is not as sensitive as some direct microbiological 

analytical techniques. The DEFT technique is capable of detecting a single cell on a 

membrane. ATP bioluminescence apparatus designed for hygiene testing can detect 

between 5 and 250 cells (Boulton and Quain, 2001). Despite the issues regarding 

sensitivity it is the method of choice for validating cleaning regimes both in the brewery 

and for dispense lines in licensed premises. 

Refinements of the ATP bioluminesence method have allowed the detection of 

individual microbial cells in beer. For example, the Sapporo Breweries in Japan have 

developed apparatus termed the MicroStar RMDS-SPS (Rapid Microbe Detection 

System ± Sapporo Special) in which beer samples are passed through a membrane filter. 

Reagents are sprayed directly onto the membrane and the bioluminescence due to any 

trapped micro-organisms detected by a photomultiplier linked to a computer for data 

collection. The apparatus can be used to detect very low concentrations of beer spoilage 

bacteria by pre-incubation of the membrane on a suitable nutrient medium for two days, 

prior to performing the ATP analysis (Takahashi et al., 1999a). 

17.3.2 Identification of brewery bacteria 

Even in the best-managed brewery occasional microbiological failures do occur. In these 

instances, identifying the organism can be a valuable aid in tracing the nature of the 

process failure. Identification may be made according to taxonomic principles. In 

addition, it may be of equal value to make identification based on more pragmatic 

grounds. For example, differentiation of production and wild yeast strains. Bacteria 

encountered during brewing are classified and identified using classical microbiological 

techniques such as cellular morphology, possession or lack of motility, colonial 

morphology when cultivated on solid media and biochemical properties. Bacterial 

morphologies are classified as being rod-shaped (bacilli) or spherical (cocci). The size of 

bacilli and cocci and whether or not the cells are borne singly, in chains or clusters are all 

diagnostic of individual species. The examination of cellular morphology is aided by the 

use of biological stains. These assist with visualizing cells for microscopic examination. 

The response of individual species to some stains is of taxonomic significance. The 

most widely used of these is the Gram stain. Gram positive bacteria are stained purple by 

treatment of a heat fixed smear with the dye, crystal violet. The procedure involves 

subsequent steps in which the smear is treated with a solution of iodine and potassium 

iodide (Gram's iodine) followed by washing with ethanol and counter-staining with a 


pink dye,safranin.Gram negative cells are stained pink becausethe ethanol step removes 

the complexofcrystalvioletandGram'siodine from fixedcells. Thedifferentialstaining 

response of the two bacterial groups is due to differences in the structures of the cell 

walls. All bacterial cell walls contain peptidoglycan as astructural component. In Gram 

positive bacteria this forms amuch thicker layer compared to Gram negative types. In 

consequence, in the former the complex of crystal violet and Gram's iodine remains 

trapped following treatment with ethanol. 

The biochemical properties of bacteria can be assessed using anumber of tests. Some 

of these are summarized in Table 17.1. These tests determine fundamental aspects of the 

physiology of individual strains such as the ability to utilize various sources of carbon 

and nitrogen and to grow aerobically or anaerobically. Other procedures look for the 

possession of specific enzymes. For example, bacteria that possess catalase are able to 

decompose exogenous hydrogen peroxide. The ability to produce specific metabolic end 

products can be probed by anumber of tests. These usually take the form of preparing a 

culture of the bacteria under investigation. The presence of aparticular metabolite is 

confirmed by the addition to the culture of reagents that bring about acolour change. 

Commonly, these tests detect the formation of metabolites that would cause recognizable 

defects in infected beers. 

Several techniques have been developed that allow the rapid identification of bacteria. 

These use the same principles as those methods used to differentiate yeast strains 

(described in Chapter 13). For completeness, abrief overview of the most promising 

methods is given in Table 17.2. Acomprehensive review may be found in Gutteridge and 

Priest (1999). Undoubtedly, conventional microbiological techniques in conjunction with 

hygiene testing via bioluminescence will continue to be the system of choice of quality 

control for many brewers. However, rapid methods of detection and identification are 

likely to become increasingly important as brewers move from quality control to quality 

assurance.Thisstrategyrequiresthatspecificspoilageorganismsareidentifiedassoonas 

possible since they pose the greatest potential threat. This is all the more important since 

the trend towards packaging of beers aseptically without pasteurization seems likely to 

grow. 

TheapplicationoftherapidtechniquesoutlinedinTable17.2tobeerspoilagebacteria 

has been reported. For example, use of DNA polymerase technology (Dimichele and 

Llewis, 1993; Tsuchiya et al., 1992a,b), use of membrane filtration and automatic 

detection of bacteria with specific fluorochromes and image analysis (Yasui and Yoda, 

1997), electrophoretic characterization of lactate dehydrogenases of Lactobacillus brevis 

(Takahashi et al., 1999b), and identification of lactic acid bacteria using monoclonal 

antibodies (Yasui and Yoda, 1997). 

17.3.3 Gram negative beer spoiling bacteria 

Abrief description of the Gram negative bacteria associated with beer spoilage and the 

defects produced by their growth is given in Table 17.3 (review, Fleet, 1992). The stages 

in the brewing process at which these bacteria exert their effects and the defects that are 

produced are dependent upon the physiological capabilities of the organisms. The 

descriptions of the bacterial species in Table 17.3 provide an immediate indication of the 

stages in brewing where the results of their spoilage may become evident. Thus, acetic 

acid bacteria are obligate aerobes and produce acetic acid from ethanol. The 

concentration of ethanol that may be tolerated varies between strains. In early work, 

Shimwell (1936) reported that none could grow at ethanol concentrations greater than 6% 


Table 17.1 Traditional microbiological techniques used in the identification of brewery bacteria 

Test Basis 

1. Carbon source utilization Score for growth on minimal medium supplemented with various sole carbon sources. 

2. Nitrogen source utilization Score for growth on minimal medium supplemented with various sole nitrogen sources. 

3. Production of acid and/ 

or gas 

As for (1), but medium supplemented with methyl red, formation of a red colour indicates the formation of acid. Incorporation of 

a small inverted (Durham) tube in the medium allows the detection of the formation of gas (bubble formation). 

4. Catalase test Pour solution of hydrogen peroxide onto surface of slope culture. Catalase positive bacteria produce copious frothing. 

5. Indole test Ethanolic solution of dimethylamidobenzaldehyde in presence of HCl and potassium persulphate added to peptone water culture. 

The presence of indole is indicated by the formation of a red coloration. 

6. Voges-Proskauer (VP) test The presence of acetoin and diacetyl in peptone water culture indicated by the formation of a red coloration following the 

addition of -naphthol and creatine. 

7. Formation of hydrogen 

sulphide 

Bacterial (and yeast) colonies develop black coloration when grown on nutrient medium and then overlaid with paper soaked in 

lead acetate solution. 

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 

colour following the addition of a solution of -naphthylamine containing acetic and sulphanilic acids. 


Table 17.2 Rapid methods for the identification of brewery bacteria (Boulton and Quain, 2001) 

Test Principle 

1. Genomic analysis Hybridization of unknown genome with DNA or RNA probe(s) from known organism. 

2. Proteomic analysis Extraction of proteome and analysis via electrophoresis. Identification via comparison with electrophoretograms made from the 

proteomes of known organisms. 

Analysis of extracted proteome and identification of selected proteins via binding to specific labelled antibodies. 

Whole cell pyrolysis under inert atmosphere and analysis of sub-cellular fragments via gas chromatography (Py-GC) or mass 

spectroscopy (Py-MS). 

3. Analysis of cellular 

components 


Whole cell extraction of fatty acids and identification from GC profiles of fatty acid methyl esters (FAME). 

Analysis of whole cells via Fourier transform infra-red spectroscopy (FT IR). 

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 

concentrations up to 10% v/v (De Ley et al., 1984) and Gluconobacter oxydans can 

survive 13% v/v (Magnus et al., 1986). Some strains can grow under micro-aerophilic 

conditions.Theyaretolerantofethanol,hopresinsandlowpH.Typically,theyspoilbeer 

where some oxygen is present, as might be the case in licensed premises where air is 

allowedtoentercasks.Beerlinesanddispenseequipmentarefrequentlycontaminatedby 

acetic acid bacteria. Spoilage becomes evident in the form of surface pellicles, turbidity 

and ropiness. The latter refers to the formation of extracellular polysaccharide material, 

which can be seen suspended as slime in the infected beer. Infected beer becomes acid 

and off-flavours develop. Acetic acid bacteria are ubiquitous in brewery and licensed 

premises. However, they should be easily controlled by the use of appropriate hygiene 

regimes. In particular, dispense systems must be kept scrupulously clean. The best 

safeguard against acetic acid bacterial infection of beers is to eliminate oxygen. 

Members of the Enterobacteriaceae associated with spoilage (Obesumbacterium, 

Rahnella, Citrobacter and Klebsiella) are related to those such as Escherichia coli that 

are commonly found in the gut of mammals. For this reason they are referred to as 

coliforms. It should be stressed that none of those that are found as contaminants in the 

brewing process are pathogens. They are facultative anaerobes. In other words, they are 

capable of growth under both aerobic and anaerobic conditions. Inthe context of brewing 

this increases the number of locations where spoilage can occur. However, other 

metabolic constraints limit the niches that these bacteria are able to occupy. All of the 

species of beer spoilage Enterobacteriaceae are tolerant of hop resins and they can 

ferment arange of sugars but cannot utilize ethanol. In consequence they are wort 

contaminants and capable of exerting deleterious effects during fermentation. They do 

not spoil beers. 

Members of the Citrobacter and Klebsiella genera are sensitive to ethanol and do not 

survive beyond the end of fermentation. Rahnella and Obesumbacterium are more 

tolerant of ethanol but nevertheless do not survive very high-gravity fermentations. In 

lower-gravity fermentations they can persist and indeed grow with the yeast. Rahnella 

and especially Obesumbacterium can be cropped with the yeast, survive during storage 

and, if steps are not taken to remove them, they can then infect future fermentations. 

Growth of coliforms on wort during fermentation produces avariety of tastes and aromas 

ranging from sweet/honey/fruity through to vegetable/faecal. Amultitude of bacterial 

metabolites is responsible for these flavour changes. These include various esters, higher 

alcohols, organic acids, acetaldehyde, diacetyl, acetoin, dimethyl sulphide and dimethyl 

disulphide. This range of end products is areflection of the metabolic versatility of these 

bacteria. In addition to the common pathways for the metabolism of glucose via 

glycolysis and the hexose monophosphate shunt (Chapter 12, Figs 12.5 and 12.7) some 

Enterobacteria possess an alternative route, the Entner-Duodoroff pathway. In the latter, 

glucose is degraded to give 2-oxo-3-deoxy-6-phosphogluconate, which is then cleaved to 

form pyruvate and glyceraldehyde 3-phosphate (Fig. 17.2). 

Two major fermentative routes occur in different genera of the Enterobacteriaceae. 

These differ based on the relative formation of organic acids and acetoin plus 2,3butanediol 

(Fig. 17.3). In mixed acid types the major end products are organic acids and 

only small amounts of acetoin and 2, 3-butanediol are formed. These types, which include 

Citrobacter and Obesumbacter, test positive with the methyl red test but negative with 

the Voges-Proskauer test. The non-mixed acid group, which includes Klebsiella and 

Rahnella produce high concentrations of acetoin and 2, 3-butanediol and in consequence 

they test positive with the Voges-Proskauer procedure. 


Table 17.3 Gram negative beer spoilage bacteria (Van Vuuren, 1999) 

Bacterial type Description Effects of growth in brewing process 

Acetic acid bacteria ± Acetobacter 

A. aceti 

A. liquefaciens 

A. pastorianus 

A. hansenii 

Acetic acid bacteria ± Gluconobacter 

G. oxydans 

Zymomonas 

Z. mobilis 

Obesumbacterium (Hafnia) 

O. proteus 

Citrobacter 

C. freundii 


Slightly curved or straight rods up to 4 m in length. Cells are 

pleomorphic and occur in pairs or chains. Some species are 

motile obligate aerobes and catalase positive. Capable of 

oxidizing ethanol. 

Similar morphology to Acetobacter. Obligate aerobes, catalase 

positive, ethanol is oxidized to acetic acid. Ethanol is not 

oxidized. 

Short fat rods, which occur singly, in pairs, chains or rosettes. 

No endospores are formed. Some species are motile others are 

not. They grow anaerobically but are catalase positive and 

tolerate aerobiosis. Glucose and fructose (but not maltose) are 

fermented to form ethanol. The optimum growth temperature is 

25±30 ëC. 

Short, fat, pleomorphic rods. They are catalase positive and 

ethanol tolerant. Growth in wort produces dimethyl sulphide, 

higher alcohols and diacetyl. Nitrate or nitrite are reduced to 

form carcinogenic nitrosamines. 

Slender straight rods occurring singly or in pairs and usually 

motile. Cells are catalase positive and are facultative anaerobes. 

Citrate is used by most but not all species. Glucose is fermented 

to form mixtures of organic acids (lactate, pyruvate, isocitrate 

and succinate). Relatively ethanol intolerant. 

Form hazes or pellicles in beers containing oxygen. 

Products of metabolism include acetic acid and 

acetate. 

As for Acetobacter. 

Exclusive to ale breweries where spoilage causes 

`rotten apple' flavour due to the formation of 

hydrogen sulphide and acetaldehyde. 

Contaminant of pitching yeast which, if present, 

grows with yeast during fermentation and results in 

slow attenuation rates and high pH beer. Gives rise 

to fruity/parsnip off-flavours. 

Rare contaminant in fermentations where it causes 

accelerated attenuation rate and produces increased 

organic acids and DMS. They are killed in late 

fermentation by the presence of ethanol.

Enterobacter (Rahnella) 

R. aquatilis 

E. agglomerans 

Klebsiella 

K. terrigena 

K. oxytoca 

Pectinatus 

P. cerevisiiphilus 

Megasphaera 

M. cerevisiae 


Short squat rods, which may be motile. Glucose is fermented to 

produce acid and gas. Strains are positive in the VP test (Table 

17.2). 

Slender straight capsulated, non-motile rods occurring singly or 

in short chains. They are facultative anaerobes and ferment 

glucose to produce acid and gas. 

Very slender curved rods occurring singly or in pairs. Older 

cells are elongated. They are motile and obligately anaerobic. 

Obligately anaerobic slightly elongated 

non-motile and non-spore forming cocci occurring singly or in 

short chains. They are relatively ethanol intolerant. 

In the brewing process it behaves in a similar manner 

to Obesumbacterium and is a contaminant of 

pitching yeast. It is relatively intolerant to ethanol 

and survives more readily in top cropping ale 

fermentations. Abnormally high diacetyl levels are 

produced in contaminated worts. 

Ferulic acid in wort is decarboxylated to produce 4vinylguaiacol. 

This imparts a phenolic off-flavour to 

beer. The reaction is also catalysed by some wild 

yeasts. 

Contaminants of small-pack beers where oxygen 

levels are low. Produces hydrogen sulphide and other 

sulphur compounds. 

Spoilage is restricted to low oxygen environments 

where the ethanol concentration does not exceed c. 

4% v/v. Putrid aromas and tastes occur due to the 

formation of hydrogen sulphide and other 

sulphur-containing metabolites.

Malate 

2H 

Glucose 


2-oxo-3-deoxy-6-phosphogluconate 


3-phosphate 

Pyruvate 

Indole-negative Klebsiella strains produce phenolic off-flavours in beers as a 

consequence of the formation of 4-vinlyguaiacol via the decarboxylation of ferulic acid. 

The latter is a phenolic acid present in wort and the reaction is similar to that which is 

catalysed by some wild yeast infections. The increased concentration of dimethyl 

sulphide (DMS) associated with many Enterobacteriaceae infections does not derive 

from the degradation of sulphur-containing wort amino acids (Wainwright, 1972). In the 

case of R. aquatalis, DMS was formed via the reduction of dimethyl sulphoxide (McCaig 

and Morrison, 1984). 

ATP 

ADP 

H2O 

Fig. 17.2 The Entner-Duodoroff Pathway. 

Glucose 

Oxaloacetate Phosphoenolpyruvate 

CoA-SH 

Acetate 

CO2 

Lactate Pyruvate 

2H 

Acetyl-CoA 

4H 

CoA-SH 

Ethanol 

Formate 

CO2 + H2 

Pyruvate 

CO2 

CO2 


Acetoin 

2H 


Fig. 17.3 Glucose metabolism by bacteria belonging to the genus Enterobacteriaceae (redrawn 

from Van Vuuren, 1999). 


CO2

Of all the Gram negative beer spoilage bacteria, Obesumbacterium proteus poses the 

greatest risk to the brewing process as aresult of its role in the formation of Nnitrosamines 

(Smith, 1994). These compounds are powerful animal carcinogens. Nonvolatile 

N-nitrosamines (apparent total N-nitroso compounds, ANTC) are formed by 

reactions between wort amines and nitrite. Nitrate is present in all worts and O. proteus 

is capable of reducing it to nitrite thereby providing the precursor for ANTC formation. 

TheabilityofO.proteustogrowinwortandsurviveintotheyeastcrophasnecessitated 

the introduction of procedures for ensuring that it is eliminated before re-pitching. The 

process of acid washing, in which pitching yeast is subjected to controlled acidification, 

accomplishes this. Disinfection of pitching yeast by acid washing relies upon the 

relative tolerance and sensitivity of yeast and bacteria, respectively to low pH (see 

Section 17.6). 

Zymomonas mobilis tolerates oxygen but grows under anaerobic conditions. It 

ferments glucose and fructose but not maltose. Unlike most of the Enterobacteriaceae it 

toleratesethanolandreportedlysurviveshigh-gravityfermentationsinwhich12±13%v/v 

ethanol are formed (Magnus et al., 1986). It has arelatively high optimum growth 

temperature of 25±30ëC. For this reason, it tends to be a more common spoilage 

bacterium in ale breweries as opposed to those fermenting lager worts at lower 

temperatures. Infected worts develop a characteristic rotten apple odour due to the 

formation of acetaldehyde. In addition, ethanol, acetic acid, lactic acid, acetoin and 

glycerol are formed (Van Vuuren, 1999). The formation of and high tolerance to ethanol 

has made Zymomonas the organism of choice for many industrial alcohol production 

processes. This fact emphasizes the threat that the organism poses to brewing. 

The Gram negative cocci, Megasphaera spp. and the Gram negative rods, Pectinatus 

spp. are contaminants of beer. They are tolerant of hop resins but their potential for 

spoilage is limited by virtue of their absolute requirement for anaerobiosis. For this 

reason they tend to be found in finished beers. Megasphaera strains produce several 

organic and fatty acids, notably butyric acid and some acetic, isovaleric and valeric. In 

addition, hydrogen sulphide is generated (Engelmann and Weiss, 1985). Their potential 

for beer spoilage is restricted by their sensitivity to ethanol (>2.8% v/v) and acid pH 

(Haikara and Lounatmaa, 1987). It is considered that unpasteurized, low-alcohol beers 

aremostpronetospoilagebyMegasphaera.Nevertheless,severalweeksmayberequired 

before turbidity becomes evident. 

Pectinatus strains are also obligate anaerobes but are more tolerant of ethanol. Their 

presence in pitching yeast has been reported (Haikara, 1989) butinfection of beer via this 

routeisofvery rare occurrence.Spoilage ofbeerbyPectinatusresultsintheformationof 

high concentrations of hydrogen sulphide with its putrid odour and development of 

turbidity. Various fatty acids, especially propionic and acetic, together with some acetoin 

are also produced. 

17.3.4 Gram positive beer spoiling bacteria 

Gram positive bacteria associated with beer spoilage are either rods or cocci, which 

together are termed the lactic acid bacteria. In addition, some members of the genera 

Bacillus and Micrococcus have been isolated from beers, although their status as true 

spoilage bacteria is questionable. Nevertheless they are included for completeness (Table 

17.4). Apart from their reaction to the Gram stain this group of spoilage bacteria is 

distinguished from Gram negative types in that they are on average less resistant to the 

antiseptic effects of hop resins. However, this distinction is not absolute and there is 


Table 17.4 Gram positive beer spoilage bacteria (Priest, 1999) 

Bacterial type Description Effect on brewing process 

Lactobacillus 

L. brevis 

L. casei 

L. plantarum 

L. fermentum 

L. buchneri 

L. delbruÈckii 

Pediococcus 

P. damnosus 

(syn. P. cerevisiae) 

P. inopinatus 

Bacillus 

B. coagulans 

Micrococcus 

M. kristinae 

Slender non-motile anaerobic rods that do not form 

endospores. They lack catalase but can tolerate oxygen and 

low pH. Some strains are resistant to hop resins. They usually 

have fastidious nutritional requirements. Fermentative growth 

produces mainly lactic acid (homofermentative types) or 

mixtures of lactic acid, acetic acid, ethanol and carbon dioxide 

(heterofermentative types). 

Gram positive non-motile cocci occurring singly, in pairs or as 

tetrads/short chains. Originally they were known as sarcinae, 

although aggregates of eight cells are rare. They are catalase 

negative but can tolerate some oxygen and grow under 

microaerophilic conditions. Most strains are homofermentative 

and many are resistant to hop resins. They are ethanol tolerant. 

Large motile rods which form endospores. They are catalase 

positive and aerobic/facultatively anaerobic. They are 

thermoduric and thermophilic but sensitive to hop resins and 

cannot grow in media with a pH lower than c. 5.0. 

They are catalase positive and usually obligate aerobes (M. 

kristinae is a facultative anaerobe). They are sensitive to acid 

pH and hop resins. 


Produce turbidity in infected beers. Some strains produce 

extracellular polysaccharides, which appear as visible `ropes' in 

infected beer. Sour/acid off-flavours are generated. 

Spoilers of fermenting worts and beers where they produce 

hazes, acidity and high concentrations of diacetyl. Historically, 

the latter was referred to `sarcina sickness'. 

The endospores allow them to survive wort boiling. They are 

able to grow in hot (55±70 ëC) sweet wort where they produce 

lactic acid. They are inhibited by hop acids and low pH and do 

not cause beer spoilage. 

Common contaminants in breweries but their sensitivity to hop 

resins and intolerance of acid pH prevent beer spoilage.

significant variability. Thus, some members of the lactic acid bacteria are resistant to hop 

resins, whereas Micrococcus and Bacillus spp. are sensitive. 

The lactic acid bacteria are an important group. Apart from having the potential to 

spoil foods they are used industrially to make fermented dairy products such as yoghurt. 

Others are of clinical significance. A review of these bacteria may be found in Priest 

(1999). The genus Lactobacillus contains members that are genetically diverse and 

further revision and sub-division is likely. Original classifications were based upon the 

mode of fermentative growth and temperature relations. Heterofermentative types were 

placed within the Betabacterium genus. Homofermentative types were subdivided into 

thermophilic strains (Thermobacterium) and mesophilic strains (Streptobacterium). 

These groups are still mentioned but do not have taxonomic significance. 

Lactococcal bacteria are classified into several genera. Historical classifications 

placed many of them into the Streptococci on the basis of facultative anaerobiosis and 

homofermentative physiology. These have now been sub-divided into the Streptococcus 

sensu stricto, Enterococcus (indicators of faecal contamination, includes S. faecalis), 

Lactococcus (includes the dairy types, S. lactis), Vagococcus (motile types resembling 

Lactococcus). Heterofermentative cocci occurring in pairs or short chains are now 

classified as Leuconostoc. Homofermentative cocci that divide in two planes to produce 

pairs or tetrads are classified as Pediococcus. 

Beer spoilage by lactococcal bacteria is restricted to Pediococcus, the most common 

species being P. damnosus. This bacterium is found commonly as a contaminant of wort 

and beer. It is not found in brewing raw materials (Priest, 1999), suggesting that it is 

particularly well adapted to the environment of the brewery. A second species, P. 

inopinatus, has also been isolated from breweries but is also found in non-brewing 

habitats. P. inopinatus has been isolated from pitching yeast but rarely from beer. Of the 

two species, P. damnosus is the more resistant to hop resins and it persists through 

fermentation into finished beer. 

The potential for Lactobacillus to spoil beer is dependent upon the relative sensitivity 

of individual strains to hop resins. Simpson and Fernandez (1992) determined the 

minimum concentration of trans-isohumulone required to inhibit the growth of 42 strains 

of Lactobacillus. The bacteria could be classified into three groups; sensitive types 

inhibited by 20 M trans-isohumulone, an intermediate group in which the minimum 

inhibitory concentration was 20±40 M and a third group capable of growth in the 

presence of up to 180 M trans-isohumulone. Only the third group could be isolated from 

beer. In a later paper (Simpson and Fernandez, 1994), the same authors concluded that 

resistance to at least 90 M trans-isohumulone was necessary for Lactobacillus strains to 

qualify as beer spoilers. Resistance to hop resins is pH dependent. Simpson (1993) 

reported that increase in pH decreases the toxic effect of hop iso- -acids. In the author's 

view, an increase in pH of as little as 0.2 units could reduce the protective effect of hop 

resins by as much as a half. 

Confirmation that Lactobacillus strains are beer spoilers is notoriously difficult to 

demonstrate since usually they will not grow on beer. Cultivation may be facilitated by 

successive pre-incubation of strains in media containing increasing proportions of beer. 

Similar results are obtained if the beer is substituted with 45 M trans-isohumulone 

(Simpson and Fernandez, 1992). This training procedure is time consuming. A more 

rapid approach has been developed in which MRS medium is supplemented with 20 M 

trans-isohumulone (Simpson amd Hammond, 1991). This medium suppresses the growth 

of non-beer spoiling Lactobacillus strains but permits the growth of those capable of 

spoilage. The ability of some strains to tolerate hop resins is probably a consequence of 


the possession of a plasmid-borne gene, termed hor A. Thus, Sami et al., (1997a) 

demonstrated that of 61 strains containing hor A, only two could not grow in hopped 

beer. Similarly, only one out of 34 hor Anegative strains grew in beer. The same group 

hasproposedthat rapididentificationofbeer spoiling Lactobacillusstrains ispossiblevia 

detection of hor Ausing PCR DNA technology. From ataxonomic standpoint, at least 

nine species of Lactobacillus have been isolated from beer (Sami et al., 1997b). Those 

most commonly encountered are L. brevis, L. casei, L. curvatus, L. plantarum and L. 

delbruÈckii. 

Infections of beer by Pediococci are characterized by the formation of high 

concentrations of diacetyl, accompanied by a reduction in yeast growth and low 

fermentation rates. Historically, this was called sarcina sickness, a reference to the 

similarity between Pediococci and true octuplets of Sarcina spp. Extracellular 

thixotrophic polysaccharide slimes may also be formed resulting in the formation of 

visible `rope'. Infections by Lactobacillus produce similar symptoms to Pediococcus. 

Growth in beer produces `silky turbidity'. Rope may be formed by some strains. As with 

Pediococcus acid is produced although the most noticeable flavour defect is the 

formation ofdiacetyl.Alllacticacid bacteria rely onthe small concentrations ofnutrients 

available in beer for their growth. The presence of priming sugars, such as sucrose or 

fructose, provides areadily assimilable source of sugar. In the absence of simple sugars, 

maltotriose and maltotetrose may be utilized by some strains. L. diastaticus, now not 

considered to be aseparate species and placed within L. brevis, is capable of utilizing 

dextrins. As its name suggests it has the potential to produce superattenuation of worts 

during fermentation. 

Lactic acid bacteria are catalase negative anaerobes. They do not possess superoxide 

dismutase (Archibald and Fridovich, 1981). Since they lack these standard mechanisms 

for nullifying the toxic effects of reactive oxygen radicals, it is surprising that they can 

tolerate exposure to oxygen. In fact, they use acombination of NADH oxidases and a 

pseudocatalase for removing peroxide ions (Johnston and Delwiche, 1965). Superoxide 

radicals are apparently scavenged by high intracellular concentrations of Mn 2+ . 

In homofermentative strains (Pediococcus spp., L. casei, L. plantarum and L. 

delbruÈckii) the major product of sugar metabolism is lactic acid. In this group sugars are 

metabolized via glycolysis. Pyruvate, derived from glycolysis, is reduced to lactate via 

the action of NADH-linked lactate dehydrogenase. Heterofermentative strains (L. brevis) 

produce amixture of end products, including lactate, glycerol, ethanol and acetate. These 

strainsutilize the phosphoketolasepathway, inwhich, glucosecatabolismproceedsvia 6phosphogluconate 

which, following adecarboxylation reaction, forms the pentose sugar, 

xylulose 5-phosphate. These reactions are those that form the initial part of the hexose 

monophosphate shunt (Section 12.5.2). Apart from this being an alternative to glycolysis, 

it provides aroute by which pentoses can be catabolized. Phosphoketolase catalyses the 

cleavage of xylulose 5-phosphate to yield, after the addition of another phosphate group, 

amolecule of triose phosphate and one of acetyl phosphate. From these intermediates 

glycerol, acetate, ethanol and pyruvate are formed (Fig. 17.4). 

The formation of diacetyl by lactic acid bacteria contamination during fermentation 

does not exclusively use the same pathway as that employed by brewing yeast. In yeast, 

diacetyl derives from the spontaneous oxidative decarboxylation of -acetolactate. 

Diacetyl is then reduced to acetoin and 2,3-butanediol (Section 12.10.2). In lactic acid 

bacteria, the same sequence of reactions can occur but some strains also possess - 

acetolactate decarboxylase, which produces acetoin directly, without the intermediary of 

diacetyl. A second route also operates in which diacetyl is synthesized directly from the 


Maltose 

Glucose 

Glusose 

1-phosphate 

Glucose 

6-phosphate 

Fructose 

1,6-bisphosphate 


3-phosphate 

Pyruvate 

Lactate 

6-Phosphogluconate 


3-phosphate 

Pentoses 

Xylulose 

5-phosphate 

Acetyl 

phosphate 

Glycerol Pyruvate Acetaldehyde 

Lactate Ethanol 

Acetate 

Fig. 17.4 Homofermentative and heterofermentative sugar metabolism in lactic acid bacteria (for 

details see text). 

TPP 

Pyruvate α-Acetolactate 

Acetaldehyde-TPP 

activated form of acetaldehyde (acetaldehyde thiamine pyrophosphate) and acetyl-CoA 

(Speckman and Collins, 1973; Fig. 17.5). 

17.3.5 Beer spoilage yeasts 

Spoilage by yeasts is potentially a serious problem since many such contaminants are 

capable of occupying the same ecological niche as production strains. However, since 

they have different genotypes from the production strain their activities can produce a 

variety of defects in process and product. Traditionally, contaminants are referred to as 

wild yeasts. The concept of `wildness' is imprecise. Thus, contaminants may range from 

non-Saccharomyces yeast strains through to accidental mixing of production brewing 

strains. In the latter case, the mixing of ale and lager strains can be especially 

problematic. A good working definition of wild yeast is any yeast not deliberately used 

and under full control (Gilliland, 1971). This all-embracing definition allows for all 

possibilities, from the use of pure monocultures through to those rare fermentations that 

rely on spontaneous contamination. The similarity of wild yeasts to production strains can 

make them difficult to detect. Although their presence may be signalled by major changes 

CO2 

Acetyl-CoA CoA-SH 

Diacetyl 

Fig. 17.5 Pathways for the formation of diacetyl by lactic acid bacteria. 


to product and process, it is equally possible that much less apparent and subtle defects 

can be caused. Acid washing of pitching yeast, which reduces bacterial contamination is 

not effective with wild yeast. Avoidance of contamination by yeast is entirely dependent 

upon the maintenance of high standards of hygiene. 

Beer spoilage yeasts may be considered as Saccharomyces and non-Saccharomyces 

types. The Saccharomyces wild yeasts pose the greatest threat since they are most similar 

to production strains. Ipso facto, they have the ability to colonize the same range of 

habitats as production yeast strains. Their similarity to production strains makes 

differentiation difficult. The taxonomic classification of Saccharomyces beer spoilers is 

now of little practical significance in that many strains originally given the status of 

species, based on characteristics of relevance to spoilage, have now been assigned to S. 

cerevisiae. The discussion in this section will be confined to a description of the effects of 

contamination. Species names will be used only to provide a historical context. The 

effects of inadvertent mixing of production strains are difficult to predict. Some likely 

outcomes are changes in flavour, cropping behaviour, fining behaviour and attenuation 

rate and extent. These are immediately obvious changes. More subtle changes from the 

norm, particularly if the level of contamination is low, may be much more difficult to 

detect, especially if there is a gradual change in the level of contamination with 

successive fermentations. 

More dramatic defects are caused by contamination with specific Saccharomyces 

yeasts. Certain Saccharomyces strains and some members of the genera Kluyveromyces, 

Pichia and Williopsis synthesize so-called killer factors. These are toxins, also known as 

zymocins that have no effect on the producing strain but are rapidly lethal to other 

susceptible strains of the same species (Young, 1987; Magliani et al., 1997). Some 

zymocins are ionophores. They exert their lethal effects by disrupting the plasma 

membrane of target cells such that their ability to retain ions is destroyed. Others inhibit 

DNA synthesis. 

Contamination of fermentations with killer yeast has the potential for catastrophic 

disruption of the brewing process. They appear to be rare in breweries. In a survey of 964 

species, representing 28 genera, 59 were found to produce killer factors. Of these, more 

than half were Saccharomyces strains (Sami et al., 1997a). Most of these were laboratory 

haploids and only four were brewing types (Philliskirk and Young, 1975). Occasional 

infections with killer yeast of an early continuous primary fermentation system were 

reported by Maule and Thomas (1973). The problem was severe, since infection levels of 

less than 3% of the total yeast population were sufficient to virtually eliminate the 

production strain. 

Killer factors might be harnessed for a useful purpose. An early proposal was to 

genetically modify brewing strains by the introduction of a killer factor (Hammond and 

Eckersley, 1984). The use of a `killer' brewing yeast strain would be an aid in the 

prevention of contamination. This option has not been pursued since brewers will not use 

genetic engineering. No doubt that it is an interesting approach but it would never be a 

substitute for good hygiene. In addition, it might be a risky strategy where several yeast 

strains are used within the same brewery. 

Most brewing strains are unable to utilize dextrins and these persist in beer where they 

contribute to fullness and mouth-feel. Some strains, originally classified as S. diastaticus 

but now placed with S. cerevisiae, possess glucoamylase and in consequence can utilize 

dextrins. Contamination of fermentations with diastatic yeasts leads to superattenuation 

of the wort and beers with abnormally low present gravity. Occasionally, diastatic yeasts 

have been used to produce so-called `light' beers. However, such strains commonly 


OH 

OCH3 

CO2 

possess other, undesirable characteristics. The removal of dextrins is more usually 

accomplished by the direct addition to wort of preparations of glucoamylase 

(amyloglucosidase). Contamination of unpasteurized bottled beer with diastatic yeast is 

potentially hazardous, since abnormally high concentrations of carbon dioxide can 

develop with the consequent risk of bottle explosions. 

Many diastatic yeast strains possess a gene termed POF, an acronym for phenolic offflavour 

(Ryder et al., 1978). This gene encodes for the enzyme phenolic acid 

decarboxylase. This enzyme decarboxylates wort phenolic acids such as ferulic and 

cinnamic acids to produce 4-vinyl guaiacol and styrene (Fig. 17.6). These compounds 

impart a medicinal or clove-like taste and aroma. They are an essential part of the flavour 

of some beers, for example, many wheat beers. In most cases the presence of these 

compounds is a serious defect. 

Many non-Saccharomyces yeasts are routinely found in breweries. Most cannot 

compete with brewing yeasts and hence do not usually gain a foothold within the process. 

Many potential contaminants are not particularly tolerant of ethanol, few are able to 

ferment sugars and many cannot grow under anaerobic conditions. Their relatively poor 

adaptation to brewing conditions means that the threat they pose is small. Spoilage by 

non-Saccharomyces yeast is generally restricted to some raw materials and to the aerobic 

phase of fermentation. More opportunities for spoilage occur in licensed premises. Ales 

in cask and dispense systems used for all draught beers can provide a semi-aerobic 

environment in which many non-Saccharomyces yeasts can grow. In this sense these 

yeasts are opportunistic contaminants which flourish where poor hygiene and bad 

practice combine to provide the conditions for growth. Scrupulous cleaning of dispense 

equipment and prevention of air ingress into casks by dispensing under blankets of inert 

gas minimize the risks. 

Non-Saccharomyces yeasts encountered in both the brewery and licensed premises 

include representatives of the following genera. Cryptococcus and Rhodotorula are 

commonly detected but unless conditions are grossly atypical are not able to spoil wort or 

beer. Candida, Kluyveromyces, Pichia and Torulaspora are opportunistic spoilers during 

the aerobic phase of fermentation and in unpasteurized cask beers. Most are obligate 

aerobes although Candida and Torulaspora are capable of poor growth under anaerobic 

OH 

CH CH COOH 

CH CH2 

Ferulic acid 4-Vinylguaiacol 

CH CHCOOH CH CH2 

CO2 

Cinnamic acid Styrene 

OCH3 

Fig. 17.6 Formation of 4-vinylguaiacol from ferulic acid and styrene from cinnamic acid by wild 

yeast possessing the phenolic off-flavour (POF) gene. 


conditions. Strains of Pichia which colonize cask ales maximize their opportunity for 

utilizing any oxygen in the gas space by forming surface films. Zygosaccharomyces, 

especially Z. bailii and Z. rouxii are osmotolerant strains and can cause spoilage of bulk 

sugar syrups. Of all the non-Saccharomyces yeasts, Brettnanomyces and Dekkera 

probably pose the greatest threat to unpasteurized beers. Both are able to ferment sugars 

to form ethanol, however oxygen stimulates fermentation, (Custers effect, Section 

12.5.8). Beer spoilage by these strains is characterized by the formation of high 

concentrations of acetic acid. 

17.3.6 Microbiological media and the cultivation of micro-organisms 

In order to cultivate micro-organisms in the laboratory it is necessary to provide 

favourable growth conditions and it may also be important to control conditions such that 

they are inimical for organisms whose growth is not desired. This requires the provision 

of asuitable source of nutrients, possibly the addition of growth inhibitors and control of 

the physical environment. 

The key environmental parameters are temperature and gas supply. The former is 

regulated by the use of thermostatically controlled microbiological incubators. 

Maintenance of temperature to +/ 1ëC within arange of ambient to 55ëC is usually 

adequate. In many instances cultivation at room temperature (18±25ëC) is sufficient. The 

gas supply is controlled to differentiate between aerobes and anaerobes. For aerobic 

cultures incubation is in the presence of air. For liquid cultures it may be necessary to 

improve oxygen transfer from the air to the growing micro-organisms by incubating on 

devices which shake the culture flask. Commonly, flask shakers and thermostatically 

controlled incubators are combined into asingle piece of apparatus. For larger volume 

liquid cultures, where shaking is not practicable or efficient, air or oxygen can be 

introducedbybubblinggasintothemediumthroughasinterorcandlefittedwithasterile 

filter. The provision of mechanical stirring further improves gas transfer rates. 

Maintenance ofanaerobic conditions requires specializedequipment.Typically,thisis 

ajar in which solid or liquid cultures are placed. Air can be removed by evacuation or by 

replacement with an inert gas such as nitrogen. Visual confirmation of anaerobiosis is 

provided by incorporation of asolution of methylene blue, which is colourless in its 

reduced form. Early devices relied on flushing out air with hydrogen and removing the 

final vestiges of oxygen via combustion using aplatinum catalyst. Modern approaches 

use commercially available kits that remove oxygen chemically. Typically, these are 

added to an anaerobic jar and activated by the addition of water. For example, one such 

kit contains sodium borohydride and sodium bicarbonate. Addition of water generates 

hydrogen and carbon dioxide, which flushes out air. For incubation of one or asmall 

number of plates, these kits are available as self-contained re-sealable foil pouches 

containing deoxygenating chemicals. Enrichment of the atmosphere within an anaerobic 

culture with carbon dioxide promotes the growth of many organisms, for example, lactic 

acid bacteria. The choice of solid or liquid medium depends upon the application. Liquid 

media are often used for the production of pure cultures, as with yeast propagation 

(Section 13.5.2). They are useful where the yield and growth rate must be maximized. 

Solid media are useful for characterizing microbial populations. They consist of a 

nutrient medium solidified with agar or gelatin. Normally the medium is held in a Petri 

dish (Prescott et al., 1996). 

Solid media serve several useful purposes. Microbiological samples can be plated out 

such that individual cells are separated from their neighbours. During incubation colonies 


develop and it may be assumed that each derives from asingle cell. Counting colonies 

allows enumeration of the microbial population within the sample. The appearance of the 

coloniesoftenhasdiagnosticsignificance.Incorporatingcomponentsintomediathatalter 

the appearance of some colonies and not others, depending on the identity of the microorganism, 

can enhance this. In this way plate cultures can be used to check that cultures 

are pure. Conversely, they can provide visual indication that the microbial population in 

the original sample was mixed. Separation of mixed populations into distinct colonies 

facilitates subsequent purification by allowing removal of chosen colonies and transfer to 

fresh medium. 

General-purposemediaaredesignedtosupportthegrowthofmanydifferentmicrobial 

species. They are useful for detecting heterogeneous populations and may be used to 

assess total microbial loadings in particular locations. Commonly these are complex 

media that contain nutrients that many micro-organisms are able to utilize. Complex 

media are not chemically defined. Usually they contain one or more components, which 

are general sources of amino acids, proteins, vitamins and metal ions. These are 

supplemented with specific sources of carbon such as sugars. Examples of general base 

ingredients are yeast extract, peptone and clarified meat extracts. Some complex media 

are tailored to a particular application. For example, in brewing several media 

formulations include wort or beer. 

Selective media are formulated to promote the growth of certain groups of microorganisms 

but not others. They are used in conjunction with selection via control of the 

environment, such as manipulation of temperature and provision or exclusion of oxygen. 

Selective media are used in many ways. Defined media contain only identified and often 

chemically simple components in combinations that support the growth of some microorganisms 

but not others. Media can be made more selective by the incorporation of 

inhibitors. These can be very general purpose, for example, the addition of selected 

antibioticsthatinhibityeastbut allowthegrowthofbacteria, such asmight beusedinthe 

assessment of contamination. Other inhibition protocols might prevent the growth of 

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) 

Medium Application Comments 

MYPG General-purpose medium for yeast and bacteria. Malt extract, yeast extract, peptone, glucose. 

Nutrient agar/broth General-purpose medium for bacteria although many yeasts 

will also grow. 

WLN (Wallerstein 

Laboratory nutrient 

medium) 

General-purpose medium for yeast but many bacteria will also 

grow. 

Contains yeast extract, peptone, NaCl and Lab-Lemco (clarified 

meat extract). Acidification with HCl to pH 4.0 suppresses 

growth of many bacteria. 

Incorporates pH indicator, bromocresol green which allows 

differentiation of some yeast strains from the colour of 

colonies. 

MYPG ± copper Wild yeast, both Saccharomyces and non-Saccharomyces. Addition of hydrated copper sulphate (200 mg.l 1 ) prevents 

growth of brewing yeast. 

WLD (Wallerstein 

Laboratory Differential) 

General-purpose medium for bacteria. Addition of cycloheximide (15 mg.l 1 ) suppresses growth of 

brewing and some wild yeast. Addition of isomerized hop 

extract (400 mg.l 1 ) suppresses growth of spore-forming bacilli. 

Lysine medium Selective medium for non-Saccharomyces wild yeasts. Synthetic medium in which lysine is the principal source of 

nitrogen. 

Crystal violet medium Selective medium for some wild yeasts. Crystal violet (20 g/ml 1 ) suppresses the growth of brewing 

yeast, variable with others. 

Yeast nitrogen base Carbon assimilation medium for yeast. Minimal medium to which various carbon sources are added. 

Potato dextrose agar General-purpose medium for yeasts and moulds. Contains glucose and potato extract. 

Hopped wort agar General-purpose medium for yeasts, suppresses growth of 

some Gram positive bacteria. 

Hopped (1040) wort solidified with agar. 

Yeast morphology agar Assessment of yeast colonial morphology. Complex defined medium. 


Melibiose medium Differential medium for ale and lager strains. Only lager strains grow. 

Frateur's medium Differential medium for Acetobacter and Gluconobacter spp. Contains yeast extract, ethanol and calcium carbonate. 

Acetobacter colonies produce clearing due to acid formation 

from ethanol. Gluconobacter deposit chalk round clearing due to 

continued growth on ethanol with formation of CO2. 

Carr's medium Differential medium for Acetobacter and Gluconobacter spp. Contains yeast extract, ethanol and bromocresol green. 

Acetobacter strains produce acid, Gluconobacter strains degrade 

acid following prolonged incubation. 

Dadd's and Martin's 

Medium 

MRS (Man, de Rogosa, 

Sharpe medium) 

Medium for isolation of Zymomonas spp. Glucose, yeast extract, peptone supplemented with ethanol (3% 

v/v) and cycloheximide (50 mgl 1 ). Adjust to pH 4.0 with HCl. 

Selective medium for lactic acid bacteria. Buffered peptone, Lab-Lemco. glucose, plus Mg 2+ , Mn 2+ . Made 

selective by the addition of 2-phenylethanol and cycloheximide. 

Modified MRS contains maltose, MRS + beer makes medium 

more suitable for application in brewing. 

Raka-Ray Recommended selective medium for lactic acid bacteria. Yeast extract, tryptone, liver concentrate, maltose, fructose, 

glucose, metal salts. Supplemented with sorbitan mono-oleate, 

cycloheximide and 2-phenylethanol. 

NBB 

(Nachweismedium fuÈr 

BierschaÈdliche Bacterien) 

Selective medium for lactic acid bacteria. Buffered glucose, maltose, peptone, yeast extract, meat extract. 

Supplemented with beer and an indicator, chlorophenol red. 

Pre-reduced PYG agar Detection of Megasphaera. Peptone, yeast extract, glucose medium which is autoclaved and 

placed in anaerobic jar whilst still molten. Allowing agar to set 

in anaerobic conditions maintains reducing conditions. 


comparative assessment of results is possible. Successful analysis of the test samples 

confirms that brewery laboratories are able to recover, cultivate and identify typical 

brewery contaminants. 

17.4 Microbiological quality assurance 

Traditional microbiological analyses do not easily fit into a conventional quality control 

system. Analyses based on sampling, plating and incubation require three to seven days 

to yield results. This is too long. Packaged product must be dispatched to trade as quickly 

as possible and with a minimum of stock holding. On the other hand, the brewery must 

ensure that product is wholesome and will not deteriorate during its intended shelf-life. 

There is an obvious correlation between poor quality and lost sales. From a broader 

perspective, brewers have an obligation to exercise due diligence and guarantee the 

harmlessness of their products. 

The time lag between sampling and result means that product may have reached the 

consumer before problems become evident. Fortunately, beers do not support the growth 

of pathogens. However, the economic implications of a major product recall could be 

catastrophic. To counter this threat there has been a move towards systems quality 

assurance. Quality control (QC) is based upon sampling the final product and ensuring 

that specifications are achieved before dispatch. Quality assurance (QA) ensures that 

production systems are sufficiently robust that product quality and integrity are 

guaranteed. In other words, quality systems are designed to be preventative rather than 

based on inspection. Most brewers use a hybrid of QC and QA. This is partly a 

consequence of the conservatism of the brewing industry. From a microbiological 

standpoint, QA guarantees the integrity of product and therefore eliminates the 

requirement for final product testing. Most brewers accept this argument but prefer to 

follow a more visceral approach and demand an apparent clean bill of microbiological 

health for finished product. 

Microbiological quality assurance systems consist of a number of essential elements, 

The process is divided into separate defined sub-processes. A sample plan is constructed 

for each sub-process. This considers where samples should be taken, what their size and 

frequency should be and what analyses should be performed on them. Results of analyses 

must be judged against appropriate specifications. A permanent record of the results must 

be maintained. Results are linked with individual batches of product, since there must be 

full traceability throughout the entire process from raw materials to finished product. 

Complex processes such as brewing require the use of formal quality systems, which 

are subject to external and independent accreditation. Commonly, the ability to operate to 

the ISO 9000 quality standard is used within the UK brewing industry. Such systems 

ensure product wholesomeness and also satisfy food safety legislation, for example the 

European Food Hygiene Directive and United Kingdom Food Safety Act (White, 1994). 

Within the quality systems, the formal assessment and management of risks is achieved 

using HACCP analysis (hazard analysis and critical control points). HACCP is integrated 

into the broader ISO 9000 system (Kennedy and Hargreave, 1997). It is a management 

system introduced to ensure food safety (Mundy, 1997). Usually it has been applied 

where processes present hazards to health from either physical or chemical agents. 

Microbiology has not usually been included except where there might be a health risk 

from pathogenic organisms. It is beginning to be used as a method of microbiological 

quality assurance where the risk is limited to product wholesomeness. 


Hazard analysis 

Critical control point (CCP) 

analysis 

Establish critical limits 

Establish monitoring 

procedures 

Establish corrective 

actions 

Establish verification 

procedures 

Establish record keeping and 

documentation procedures 

Fig. 17.7 Elements of HACCP analysis. 

There are several elements to HACCP analysis. The starting point is to draw up a 

detailed flow diagram of the process under consideration. Once constructed, the process 

flow diagram must be verified to ensure that all relevant steps have been included. All 

HACCP analyses contain seven parts (Fig. 17.7). In the first part each step in the process 

is assessed and its inherent risks are identified. In the second step the identified risks are 

graded to identify those that are critical control points (CCPs). A CCP is a process step 

which, if not under proper control, has the potential to cause injury or illness to 

consumers. For microbiological control of the brewing process the concept of the CCP is 

widened to include a potential to result in the sale of product that is not wholesome. The 

third part of the analysis is to set critical limits for each CCP. For example, in the case of 

a pasteurized beer, the critical limits would be the time and temperatures needed to ensure 

that the product is rendered microbiologically stable. 

The fourth step is to establish monitoring procedures to ensure that the critical limits 

are adhered to and the CCP is in control. In the case of the pasteurized product this would 

be a permanent record of the times and temperatures to which all batches of product had 

been exposed. The analysis of this type of step would include procedures to ensure the 

veracity of measurements, such as proof of calibration of thermometers. Monitoring 

should allow for the identification of trends where the CCP might be moving towards 

being out of control. Early identification of such trends allows corrective actions to be 

taken. 

The fifth part of the analysis establishes corrective actions should a CCP be found to 

be out of control. To continue the example of the pasteurized product, this would include 

procedures to segregate product, which it was suspected might not have received the 

specified heat treatment. The procedures to be followed in these circumstances must be 


detailed in the HACCP plan. They must happen and not be subject to discussion. In the 

example cited the first priority would be ensure that suspect product could not be sent to 

trade. The absolute requirement for reliable systems of traceability and labelling can be 

readily appreciated. Once suspect product has been isolated a more leisurely examination 

of the problem and consideration of its fate can be undertaken. 

The penultimate step in the HACCP plan is to establish verification procedures. These 

are of several types and their precise nature depends upon the detail of the process. They 

must include two elements. Firstly, a day to day examination of the product to ensure that 

it meets pre-established specifications. Secondly, regular and preferably independent 

audits must be performed to guarantee the integrity of the process. Finally, the entire 

HACCP plan must be documented and a system of record keeping set up. 

17.5 Sampling 

An appropriate method must be used for obtaining a sample that is representative of the 

part of the process under investigation. The sampling device must operate in a manner 

that ensures that there is no contamination from the operator or from any other part of the 

process. The sample must be of an appropriate size to ensure that subsequent 

microbiological analyses yield significant results. Similarly, the frequency of sampling 

must be such that the microbiological quality of the process is assured. The results of 

analyses must be compared against predetermined specifications. Any results that fall 

outside a specified range must prompt investigations and where necessary cause 

corrective actions to be made. A record of results must be maintained for the purposes of 

traceability and for trend analysis. 

17.5.1 Sampling devices 

Sampling devices must be designed and used to ensure that each sample is taken in an 

aseptic manner. For permanent installations such as sample cocks that are fitted to vessels or 

process pipework, the design and location must be such that they are properly cleaned during 

CIP. Sample points should be in accessible locations. Prior to withdrawing the sample it 

must be possible to sterilize the internal and external surfaces of the device. This is best 

achieved by steam, but this is rarely available. Portable gas burners can be used to flame 

sample cocks, although care must be taken to avoid damage to rubber seals, etc. Where heat 

labile components are present external surfaces and internal components can be sterilized by 

flooding with a 70% v/v solution of methanol or industrial methylated spirits (IMS). IMS 

may be ignited (with care!) to facilitate sterilization. All apparatus used to collect samples, 

such as tubing and containers must be sterile and wrapped to prevent contamination. 

Sample cocks are of three types. The safest is the diaphragm type. These consist of a 

port mounted flush into the surface of the tank or pipe. The port is sealed with a rubber 

membrane held in place by a stainless steel nut. The sample is withdrawn by piercing the 

membrane with a sterile hypodermic needle attached to a piece of sterile tubing. The 

sample is run into a suitable sterile container for transport to the laboratory. These sample 

ports must not be heated by flaming and the membrane must be sterilized by flooding 

with IMS. Membranes must be replaced regularly. 

Plug-type sample cocks achieve a seal via a stopper containing a hole that can be 

rotated to align with the aperture in the sample port. A combination of good engineering 

and a thin layer of silicone-type grease achieves a watertight seal between the body of the 


device and the stopper. Sterilization is achieved by acombination of heat and flooding 

withIMS.Theoutletfromthesamplingdevicehasaconnectionforahosethroughwhich 

the sample is collected. The hose is the most microbiologically suspect part of the 

sampling device. Prior to use it must be sterilized by immersion in IMS. 

Valve-type sample cocks have ascrew thread mounted spindle that in the closed 

position seals the port. Unscrewing the spindle moves it outwards, unseals the port and 

allows liquid to flow. Typically, these sample cocks have two outlets mounted vertically 

to each other. Prior to sampling, the whole assembly is sterilized by passing steam 

through the top outlet and allowing the condensate todrain through the lowerone. Again, 

care must be taken to ensure that contamination is not introduced from careless handling 

of hoses. Both plug- and valve-type sample cocks are suitable for withdrawing aseptic 

liquid samples of any volume. However, in the hands of the unskilled, they are more 

prone to result in contamination during sampling compared to diaphragm samplers. 

Where comparatively large volumes must be sampled, because the expected microbial 

loading is low, it is convenient to use sterile membrane filters. The process liquid must 

not contain appreciable amounts of particulate material. Sample volumes are typically 1± 

5litres. Usually membranes with apore size of 0.45 Mare used. They are placed within 

afilter holder which has appropriate fittings for attachment to the sample point. The 

membrane filter is attached, with the usual aseptic precautions, to asample cock (either 

diaphragm- or valve-type). Typically ameasured volume is allowed to pass through the 

membrane, but slowly, so as to avoid excessive gas breakout. Membranes are collected 

still in the holder, returned to the laboratory, removed and overlaid onto a plate 

containing asuitable solidified medium. Membranes are usually printed with agrid 

which, after incubation, facilitates counting of any colonies. By calculation the microbial 

count in the original sample is determined. This method can also be used for line drip 

tests where asmall proportion of the process flow through apipe is diverted so that it 

passes through amembrane, as described. Where very low, or zero counts are expected 

membranes may be left in place for several hours. 

Themembraneapproachisusedwhereverthereisaneedtoconcentratemicro-organisms 

fromalargevolumeofliquidorgas.Suitableapparatusisavailableforasepticprocessingof 

liquid samples in the laboratory. This procedure is used for processing pasteurized and 

packaged beers. For example, keg/cask samples and whole bottles and cans. It can also be 

usedforhygienetesting,forexample,confirmationofCIPbytestingoffinalrinseliquors,or 

saline rinse samples of steamed kegs. Rinse samples can be analysed using conventional 

microbiological techniques. More commonly, the ATP bioluminescence procedure is used 

(Section 17.3.1). Alternatively, the cleanliness of surfaces can beassessed by wiping with a 

sterile swab. Swabs are rinsed in sterile saline and any microbial contamination in the liquid 

assessed by conventional techniques or via bioluminesence. 

Process gases such as CO2, N2, air or O2 are sampled using bespoke apparatus such as 

the Hollandaer and Dalla Vale device. This consists of a sterilizable housing, which 

accommodates a 45 mm plate of nutrient medium. The device is attached aseptically to a 

suitable sampling point such that the gas is allowed to flow over the surface of the plate 

for a predetermined period of time. Plates are removed and incubated to detect growth. 

Results are expressed as colony forming units (cfu) per unit time. A criticism of the 

method is that prolonged gassing dries out plates and leads to underestimates of counts. 

This can be avoided by using gassing times of no more than 60 seconds. Alternatively, 

the sampling device can be replaced with an Ehrlenmeyer flask sealed with a bung 

through which pass inlet and outlet tubes. The latter is fitted with a sterile gas filter. The 

flask contains sterile saline solution and the gas flow to be sampled is allowed to bubble 


through the liquid for apredetermined time. Microbial contamination is determined from 

samples of the saline plated out onto appropriate media. 

Environmental microbial surveys should be performed to check loadings in sensitive 

areas such as packaging halls or cask racking plants. Atmospheric contamination can be 

assessed simply by exposing agar plates of appropriate media in the area of interest. Any 

organisms falling onto the plates that grow provide a measure of contamination. 

Apparatusdesignedforairsamplingtakestheformofasterilehousinginwhichaplateof 

nutrient agar medium is placed. The device, which is portable and hand held, draws a 

known volume of air across the surface of the plate. The plate is removed and incubated 

and resultant microbial colonies counted. 

17.6 Disinfection of pitching yeast 

Pitching yeast acts as a reservoir for low levels of bacterial contamination. Serial 

cropping and re-pitching provides a route for infection of successive fermentations. 

Providing the level of infection is low and the pitching yeast has high viability and 

vitality the effects on fermentation performance and beer quality will be small. However, 

as described in Section 17.3.3, the common contaminant of pitching yeast, O. proteus, 

has the ability to produce potentially carcinogenic nitrosamines from nitrite. To avoid this 

hazard it is common practice to disinfect pitching yeast prior to pitching. 

Several approaches have been used to disinfect yeast each of which rely on agents that 

have the ability to selectively kill bacteria but not yeast. Obvious agents for this are 

antibiotics, which are used in an analogous manner in selective media. In two early 

studies (Gray and Kazin, 1946; Case and Lyon, 1956) it was demonstrated that tyrothricin 

and polymyxin B could be used to selectively kill bacteria in the presence of yeast. 

Superficially, the use of antibiotics is attractive. No effect on yeast would be expected, 

antibiotics are effective at low concentrations and their effect would persist into 

fermentation. However, this application of antibiotics was suggested at a time when the 

risks of selecting for resistant strains of bacteria were not appreciated. It is now 

recognized that introduction of antibiotics into foodstuffs is irresponsible and this 

practice has never been implemented. During the 1980s a brief resurgence of this 

approach took place. The use of a bacteriocin, nisin, was promoted as a bacterial 

disinfection agent for use in brewing (Ogden, 1987). Nisin is a small polypeptide 

synthesized by strains of Lactobacillus lactis. It exerts its toxic effects by disrupting the 

membranes of susceptible cells. It is used in many food industries, particularly those 

producing dairy goods where it is a legally acceptable preservative (E234 in Europe). 

Nisin has the same advantages as antibiotics with the additional properties of being 

relatively heat stable and retaining activity at acid pH. It has the serious disadvantage, 

from a brewing standpoint, in that it is lethal for many Gram positive bacteria but shows 

no activity against most Gram negative types. This implies that it would not be suitable 

for controlling O. proteus. For this reason, as well as reluctance by brewers to add 

`foreign' substances likely to persist into product, the use of nisin has not found favour. 

Disinfection of pitching yeast is commonly achieved by acid washing. This process 

kills susceptible bacteria, but has no effect on wild yeasts. Yeasts are relatively more 

tolerant to acid conditions than most common bacterial contaminants. The process has a 

long history, Pasteur suggested that treating with tartaric acid could reduce the levels of 

potentially harmful bacteria in pitching yeast. In a typical modern process, yeast slurries 

are treated with a food grade acid such that the pH is reduced to approximately pH 2.2 


(+/ 0.1) and held for 2±4 h at 3 ëC (+/ 1 ëC). Several acidulants are used, the most 

common being phosphoric acid. Mineral acids including sulphuric, nitric and 

hydrochloric are also used. The efficacy of the treatment is improved by incorporating 

oxidizing agents such as ammonium persulphate. Thus, ammonium persulphate (0.75% 

w/v) plus phosphoric acid, pH 2.8, was more effective than treatment with acid alone at a 

pH of 2.2 (Simpson, 1987). 

Whether acid washing has deleterious effects on brewing yeast is controversial. 

Certainly, the process must be properly controlled. Individual brewers decide on 

optimum conditions for their particular yeast strains. Experience suggests that the 

conditions described above are close to best practice. Addition of the acidulant should be 

done in a manner that avoids localized high acid concentration therefore the yeast slurry 

should be roused mechanically throughout the whole process and the acidulant dosed in 

gradually to ensure proper dispersion. Commonly, specific acid washing and pitching 

tanks are used. These have the advantages that yeast sufficient for pitching is treated with 

acid and there is no residue to dispose of. Acid washing tanks can be fitted with pH 

probes to monitor the addition of acidulant. The siting of the probe in relation to the acid 

dosing point should be such that there is no possibility of over-dosing. Where in-tank pH 

probes are not used it is preferable to remove a sample of slurry and perform a titration to 

determine the quantity of acidulant required to achieve the target pH. Pitching into wort 

terminates the acid washing via dilution with wort. If there is a process delay such that 

pitching is delayed it is best to avoid prolonged exposure to acid conditions by the 

addition of food grade sodium hydroxide solution to a pH of 4.0±4.5. 

Acid washing has a dramatic effect on the appearance of many pitching yeast slurries. 

Commonly there is a marked reduction in viscosity. This has been attributed to 

deflocculation. It has been suggested that this might be implicated in a improvement in 

fermentation performance by acid washed yeast (Jackson, 1988). Compared with nonacid 

washed yeast this author reported that high-gravity worts were fermented more 

rapidly, the duration of the diacetyl rest was reduced and beers contained lower 

concentrations of acetaldehyde. More usually, deterioration of brewing yeast subject to 

acid washing has been observed. The effects of washing with acidified ammonium 

persulphate on 16 yeast strains were studied (Simpson and Hammond, 1989). Although 

no gross effects were observed, changes to the cell surface occurred and there was 

leakage of cellular contents. Using scanning electron microscopy it was demonstrated 

that the surface of acid washed yeast acquired characteristic `blebs'. Prolonged treatment 

produced slurries that were sticky, possibly indicating some cell lysis. 

It was suggested that where yeast was obviously deteriorated it should not be acid 

washed. This meant yeast that was heavily contaminated with bacteria, had been 

recovered from a previous slow fermentation or from a high-gravity fermentation where 

the alcohol concentration exceeded 8% v/v. The reasoning is that stressed yeast is less 

well able to withstand the additional stress of acid washing. Generally, yeast with a 

viability judged by methylene blue staining of less than 80% should not be acid washed. 

Where a choice exists, low viability yeast that is obviously stressed will not be chosen for 

repitching. 

17.7 Cleaning in the brewery 

The ability to keep process plant scrupulously clean is an essential prerequisite of good 

hygiene. This is a complex and frequently neglected part of the brewing process. Too 


often when new plant and processes are installed cleaning is considered as an 

afterthought. Tanks and complex individual pieces of plant are usually well equipped 

with CIP systems. However, cleaning the pipework connecting these vessels is often not 

subjected tosufficient scrutiny. There is little benefit in having clean tanks if the attached 

pipework is areservoir for taints or infection. Similarly, removable items such as flexible 

hoses and fittings of various sorts are frequently poorly cleaned and act as foci for 

infecting otherwise clean plant. For trouble-free operation cleaning must be treated as an 

integral part of the brewing process. 

Cleaningisrequiredforthreereasons.Firstly,theremovalofanysoilingwhichhasthe 

potential to introduce taints into products. Secondly, the removal of soiling which has the 

potential to adversely affect the operation of parts of the plant, for example, building up 

of scale on heat exchangers. Thirdly, disinfection to eliminate any risk of microbial 

spoilage. This section concentrates on the latter aspect of cleaning although inevitably 

there is much overlap with the first two. Acomprehensive review of brewery cleaning 

regimes and some useful definitions can be found in Singh and Fisher (1999) (Table 

17.6). 

The cleaning challenge changes throughout the brewing process (Fig. 17.8). At the 

malting stage the spoil is largely restricted to particulate materials and plant needs to be 

physically clean (although precautions must be taken to control mould growth). During 

the wort production stages the soil consists of protein, dextrins, sugars, minerals, tannins 

and hop materials. Heating during wort boiling can generate scale (beerstone, calcium 

oxalate) which becomes baked onto surfaces. There is no microbiological threat and so it 

is sufficient to render the plant chemically clean. From the wort cooling stage onwards 

the soil consists primarily of beer and yeast. Throughout all of these stages the process 

plant must be both chemically and microbiologically clean. This requirement extends into 

licensed premises where draught beers are dispensed. 

Chemical cleaning relies on treatments capable of removing soils associated with wort 

and beer. These soils contaminate the surfaces of tanks and pipework as films that remain 

after product streams have drained away, so with the exception of baked-on scale, most 

chemical soils are loosely attached to the surfaces of brewing process plant. 

Microbiological soils are more complex and potentially much more tenacious. Natural 

mixed populations of micro-organisms are commonly found attached to surfaces as 

biofilms (Quain, 1999; Stickler, 1999). Indeed, this may be the preferred mode of growth. 

Micro-organisms in biofilms, attached by extracellular polysaccharides, are organized so 

as to make best use of available nutrients whilst occupying a protected location. Many 

surfaces, including stainless steel, are readily colonized. Many common beer spoilage 

bacteria are adept at forming biofilms (Czechowski and Banner (1992). In addition to the 

threat posed in the brewery, biofilms are also important causes of spoilage in licensed 

premises. In particular, cleaning of beer dispense lines must be rigorous to ensure that 

biofilms are not allowed to form. 

Biofilms are of significance to brewing for two reasons. Firstly, once formed they are 

difficult to remove, and they can act as permanent sources of contamination capable of 

seeding successive batches of product. Secondly, and perhaps more importantly, analysis 

of suspended microbial counts may dramatically underestimate the true magnitude of 

contamination. Usually assessment of contamination is based on measurement of the 

free-living population. Hygiene tests should be based on the swabbing of surfaces rather 

than analysis of rinse water. Cleaning regimes must be sufficiently vigorous to prevent 

biofilms forming. 


Table 17.6 Cleaning regimes ± definitions (adapted from Singh and Fisher, 1999) 


Physical cleanliness Visually clean. 

Chemically clean Cleaned surface imparts no contamination to product. Cleaned surface wets completely with clean water and forms a continuous 

film. 

Microbiologically clean Absence of microbiological contamination. 

Cleaning-in-place (CIP) Automatic cleaning of plant without the need to dismantle. 

Soil Any substance in the wrong place but typically residues from process liquids and solids. 

Detergent A chemical cleaning agent, often with surfactant properties for removing soil from surfaces. 

Disinfection Treatments that kill micro-oganisms and reduce loadings to a desired concentration. It does not imply total killing (sterilization). 

Production sterility Term used in industrial microbiology to describe a state in which plant or processes are disinfected to produce conditions which 

do not result in spoilage. 

Chemical sterilant A chemical disinfectant. 

Sanitization Process which is a combination of cleaning and disinfection. 


Process step Soil Standard of 

cleanliness 

Malting 

Brewhouse 

Wort cooling 

Fermentation 

Conditioning 

Bright beer tank 

Packaging 

Draught dispense 

Particulate Physical 

Wort 

Beer + yeast 

Chemical 

Chemical + 

microbiological 

Fig. 17.8 Nature of the soil and the standard of cleanliness required at various stages in the 

brewing process. 

17.7.1 Range of cleaning operations 

Many traditional breweries use open vessels and these may still be cleaned manually using 

buckets and brushes. However, attempts have been made to improve the efficiency of 

cleaning, using less manpower. Undoubtedly, this has encouraged the adoption of 

automatic cleaning regimes. Thus, traditional open vessels, such as fermenters, are 

commonly fitted with detachable hoods that are fitted to enclose the vessel during 

automatic cleaning. Detachable fittings such as hose adapters, valves, swing bends, 

components of fillers, flexible hoses, etc., may be cleaned manually followed by immersion 

in soak baths filled with a suitable disinfectant. Surfaces of complex equipment such as 

fillers and crowners are commonly cleaned with foams or gels which have excellent wetting 

properties and an ability to penetrate into hard-to-reach areas. With bottle and can filling 

machines foaming sprays are often part of an automatic cleaning system of external 

surfaces. Environmental cleaning can be a neglected area, particularly with the increased 

enclosure of the brewing process. Nevertheless in sensitive areas, such as filling halls and 

yeast rooms, walls, floors and ceilings should be cleaned with high-pressure sprayers. 

Individual pieces of plant tend to be fitted with dedicated automatic cleaning-in-place 

(CIP) systems such as various types of vessel and associated equipment such as heat 

exchangers, centrifuges and filters. Intervening pipework may be cleaned as part of CIP 

circuits with vessels or other pieces of plant. Some pipe runs may need to be cleaned 

independently. The proximity of cleaning fluids and beer streams requires a high standard 

of engineering to ensure that no cross-contamination can occur. In sophisticated and 

complex brewery processes, cleaning and other operations are usually fully automatic and 

computer controlled. These systems include fail-safe arrangements to ensure that possible 

conflicts are eliminated. On the other hand fully automatic systems are often inflexible. 


The use of enclosed equipment introduces abarrier between the process and the 

external environment. Together with improved cleaning this has improved microbiological 

standards. However, these improvements have been costly. CIP systems are 

expensive in both capital and revenue, therefore it is essential that each system is 

appropriatefor itsjob.Onceasystemhas been chosen,itsoperationmust beoptimizedto 

maintain abalance between efficiency of cleaning and cost of the operation. Several 

factors impact on the efficiency of cleaning including the vigour of the delivery of the 

cleansing agents to the plant being cleaned. The duration of the treatment, the 

temperature and nature of the chemical agents used are all influential. 

Proper design of brewery plant is an essential prerequisite of good cleaning. Items 

such as tanks are fitted with spray-balls, which are designed to distribute jets of cleaning 

fluids such that all internal surfaces are cleaned. The jets emanating from spray-balls 

must be able to reach all parts of the vessel. Two types of spray-ball are used. Lowpressure 

types rely on being placed in an appropriate location to ensure adequate 

coverage ofsurfaces.High-pressurespray-ballsare causedtorotatebythe incomingfluid 

and thereby ensure that all surfaces are covered. Cleaning fluid is collected via the 

pipework of the piece of equipment being cleaned and re-circulated back through the 

spray-balls. Internal metal surfaces must be polished to asuitable standard to ensure that 

therearenocrevicestoharbourmicrobialcells.Similarly,internalweldsandfittingssuch 

as sample cocks must not provide shadow areas where soil cannot be removed. The rate 

of delivery of cleaning fluids must be balanced with rates of draining so that flooding 

does not occur and thereby negate the scouring action of the spray-balls. Often this is 

accomplished by delivering the cleaning agent in aseries of bursts. The duration of 

successive cleaning treatments must allow for proper draining to avoid intermixing. 

Design and the velocity and turbulence of cleaning fluids influence cleaning of 

pipework. Proper cleaning of pipes requires complete filling with fluid to ensure wetting 

of all surfaces. The diameter of the pipe, the velocity of the fluid flow and the aspect of 

the pipe relative to the fluid flow influence filling. The larger the diameter of the pipe the 

higher the flow rate must be to ensure complete filling. As arule of thumb, apipe with a 

diameter of 3in. (7.62cm) requires aflow rate of at least 2.2m.sec 1 toensure complete 

filling (Fig. 17.9). Vertical runs of pipe work are adequately filled providing the liquid 

flow is upwards. Where the fluid flow is vertically downwards even greater velocities are 

needed to ensure complete flooding. In the case of athree-inch pipe the vertical downflow 

rate must be increased to at least 8.5m.sec 1 toensure complete wetting. This 

problem can be ameliorated by slightly reducing the diameter of vertical pipe runs. 

Alternatively, non-return valves may be used. Some common problems associated with 

pipe runs and remedies are shown in Fig. 17.10. 

The duration of cleans and the temperature at which cleaning agents are delivered is a 

balance between efficiency and cost. An increase in both parameters is associated with an 

increase in cost. The efficacy of alkaline detergents both for removing soil and for killing 

micro-organisms is increased by elevated temperatures. In order to achieve a given level 

of cleanliness and microbial kill, there is a rough negative correlation between detergent 

strength and temperature. Similar relationships exist between both of these parameters 

and the duration of the clean. Several authors maintain that soil that is formed hot should 

be removed hot (Platt, 1986, Singh and Fisher, 1999;). On this basis, most of the 

brewhouse, wort paraflows and pasteurizers should be cleaned hot. However, most areas 

which receive heavy yeast soils, such as yeast storage tanks and propagation plants, tend 

to be cleaned hot. Preliminary rinses and acid detergents are usually applied cold, 

whereas alkaline detergents are used hot. Irrespective of the operation, the guiding 


Velocity (m/2) 

3.5 

3.0 

2.5 

2.0 

1.5 

1.0 

Horizontal pipes full 

Horizontal pipes part-full 

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Inches 

2.54 5.1 7.5 10.2 Centimetres 

Fig. 17.9 

Pipe diameter 

Relationship between pipe diameter and velocity of fluid flow required to ensure that 

horizontal pipe runs remain full of liquid. 

(a) 

Inverted pocket 

traps air 

Joints and flanges can 

allow air into the system 

(b) 

High point 

traps air 

Vertical drop 

never fills 

Unchoked drain 

allows air into system 

High-point vent 

to bleed air Restricting 

vertical drop 

helps main to fill 

Non-return on drain 

keeps air out 

Fig. 17.10 Poorly designed pipework (a) and some possible improvements (b) (Diagram courtesy 

of A. Mielenewski). 

principle must always be achievement of a desired clean within an acceptable time and at 

an acceptable cost. Singh and Fisher (1999) conveniently quantified the total cleaning 

energy required as the sum of chemical energy (strength of cleaning agent), mechanical 

energy and thermal energy. 


17.7.2 CIP systems 

Several variations are used, all sharing some common features. Cleaning agents are 

stored in tanks, which are attached to the pieces of equipment and pipework that need to 

be cleaned. Cleaning agents are delivered from storage tanks to the plant by dedicated 

pumps. Usually asmaller scavenging pump returns fluids to the cleaning tanks to form a 

CIP circuit. Additional pipework allows the liquid flow to be directed to drain, as 

appropriate. The whole arrangement is termed a CIP set. Very sensitive pieces of 

equipment may have their own dedicated CIP sets. More usually, one or asmall number 

of CIP sets serve particular parts of the brewery process. The total number of CIP sets 

available influences the capacity and flexibility of the cleaning operation. Large numbers 

of CIP sets increases the cost of the operation. CIP may be manual or automatic. Manual 

systems are flexible and allow one or asmall number of CIP sets to service many 

different parts of the brewery. The nature of the cleaning regime can be varied to suit the 

type of soil that needs to be removed. Automatic regimes are inflexible and more 

expensivethanmanualtypes,howevertheytendtogiveamoreuniformcleaningprocess. 

Total dump (total loss or single use) systems do not recover any of the cleaning fluids or 

rinses. They are profligate in terms of chemical and water usage and produce the most 

effluent, so they are used only in areas of heavy soil where cleanliness and microbiological 

hygiene are of paramount importance, e.g., in yeast propagation plant. They may also be 

used in conjunction with caustic-based detergents where there is ahigh concentration of 

CO 2.Typically, single use CIPsets are dedicated to single pieces of plant.They are located 

in close proximity to the piece of equipment they are required to clean. 

Partial and total recovery systems save some costs by recovering aproportion of the 

cleaning fluids. In these systems some or all of the dilute detergent is retained, after 

topping up with fresh detergent to ensure that it is of the correct strength it is re-used. 

Similarly, aproportion of rinse waters may be returned to the CIP set and re-used. These 

systems are used where the cleaning task is less onerous, for example, bright beer tanks. 

CIP sets have amanual or automatic means of setting the cycle times for individual steps 

in the cleaning process. Athermometer and thermostat is needed to set the temperature 

for hot treatments. Where caustic-based detergents are used it is usual to monitor the 

strength by means of aconductivity probe. This may be used as part of an automatic loop 

system for dosing concentrated detergent into feed tanks. Care must be taken when 

assessing caustic concentration based on conductivity measurements (Section 17.7.5). 

CIP treatments consist of a number of individual steps. At the start the plant being 

cleaned must be empty so that there is no dilution of cleaning agents. Similarly, the plant 

must be fully drained between individual treatments to ensure that there is no intermixing 

and consequent reduction in the effectiveness of liquids. The process starts with a prerinse 

to remove heavy soil. Often pulsed rinsing is use to improve the efficiency of soil 

removal. The pre-rinse is sent to drain and in recovery systems, the liquid used is often 

derived from a recovered final rinse from a previous CIP. The main detergent clean 

follows the pre-rinse. This is the longest part of the process in which the liquid is 

circulated through the CIP circuit, typically 30±60 minutes. In total dump systems the 

detergent is sent to drain. In recovery systems it is retained and used in subsequent cleans. 

Where alkaline detergents are used an acid rinse may follow to neutralize residual alkali. 

The process is completed by a terminal rinse, which typically includes a sterilant. 

Terminal sterilants include peracetic acid, chlorine dioxide and, now more rarely, sodium 

hypochlorite. In some cases it is not appropriate to use a sterilant since it may affect the 

process or the beer, for example, propagation plant and packaging lines. In these cases 

sterility can be assured by a final treatment with wet anaerobic steam. 


17.7.3 Cleaning agents 

Soil removal is much improved by the use of detergents. These act synergistically with 

physical cleaning processes by loosening soils from surfaces. Several roles performed by 

detergents are recognized (Singh and Fisher, 1999). Physical effects include wetting of 

surfaces,dispersionoflarge agglomeratesofsoiltoformfinelydivided particulatematter 

and suspension of soil in solution. Chemical effects include dissolution of inorganic 

mineral scale by acid detergents, saponification of lipophilic components by alkaline 

detergents and hydrolysis of proteins. 

For the removal of heavy soils the most commonly used detergents are those based on 

sodium hydroxide. Typically caustic soda is used at aconcentration of 2±5% w/v at a 

temperature of 70±90ëC. It is avery good cleaning agent and, when hot, is an effective 

biocide and efficient at removing biofilms (Czechowski and Banner, 1992). It has two 

major disadvantages. Caustic soda reacts with CO 2to produce sodium bicarbonate. In 

parts of the brewery, such as fermenters where there are high concentrations of CO2, this 

can dramatically reduce the effectiveness of caustic detergents. Thus, 1m 3 ofCO2 at 1 

bar and 20ëC will neutralize 2kg of NaOH (Gingell and Bruce, 1998). To avoid this 

problem it is necessary either to ensure that vessels and pipework are vented to remove 

CO2 or to appreciate that losses will occur and increase the dosage of caustic detergent 

accordingly. Alternatively, acid-based detergents can be used in areas where CO2 will be 

aproblem. 

Caustic soda reacts with the salts that cause hardness in water to form insoluble 

precipitates in the form of calcium carbonate and magnesium hydroxide. The precipitates 

can accumulate on the surfaces in the form of scale. The problem is exacerbated in areas 

subjected to hot cleaning such as parts of the brewhouse. To counter this detergents are 

routinely supplemented with chemical additives termed sequestrants. These are of two 

types (Table 17.7). Stoichiometric sequestrants form soluble chelates with metal ions. 

They break down mineral precipitates in soils or prevent their formation such that scale 

formation is minimized. Threshold sequestrants do not prevent the formation of water 

hardness precipitates, rather they modify their crystal structure such that they do not 

adhere to surfaces and form scale. 

Acidic detergents have no reaction with CO2 and, therefore, are suitable for use with 

fermenters and associated plant. Commonly they are used cold. Their use for cleaning of 

fermenters is not popular since, compared to hot caustic detergents, they are less efficient 

at removing heavy soils. However, they do remove scale and, where this becomes a 

problem, they may be chosen for occasional use. Acid and caustic detergents can be used 

sequentiallyinheavilysoiledlocationssuchasfermenters.Inthiscase,acausticpre-rinse 

isused toremove the worst ofthe soilfollowed by anacid-basedclean. Acidic detergents 

are often used for keg washing. The most commonly used acid is phosphoric acid. This is 

sometimes supplemented with nitric acid. When used caution must be exercised since 

acidic detergents are corrosive. 

Theefficacyofbothacidandalkalinedetergentscanbeenhancedbytheincorporation 

of surfactants. These are molecules that have both hydrophobic (non-polar) and 

hydrophilic (polar) groups. They improve wetting by reducing the surface tension of 

liquids. This is achieved by their tendency to adopt aconfiguration in which the nonpolar 

groups become aligned towards the surface of liquids and the polar groups lie in an 

inward-facing orientation. Above certain concentrations (the critical micellar concentration) 

the non-polar groups form aggregates termed micelles. These trap lipophilic soil 

particles and prevent them from re-adhering to surfaces. Water-insoluble surfactants are 

defoaming and are used where the formation of foam is undesirable, such as bottle 


Table 17.7 Sequestrants used in CIP cleaning agents (after Singh and Fisher, 1999) 

Generic sequestrant Examples Action 

1. Stoichiometric sequestrants 

Amino carboxylic acids Ethylenediaminetetraacetic acid (EDTA) 

Nitrilotriacetic acid (NTA) 

Chelator of metal ions forming stable complexes and 

helping to remove scale or prevent its formation. 

Hydroxycarboxylic acids Derivatives of gluconic acid Chelator of Ca 2+ , Fe 3+ , Al 3+ . Effectiveness is 

increased in the presence of free NaOH and activity 

is maintained at high temperature. They are used in 

the brewhouse and for bottle washing. 

Polyphosphates Sodium tripolyphosphate 

Sodium hexametaphosphate 

2. Threshold sequestrants 

Phosphonic acid derivatives Amino,tris-(methylenephosphonic acid) (ATMP) 

1-Hydroxyethane diphosphonic acid (HEDP) 


Relatively insoluble chelators often used as 

constituents of powdered detergents. 

Modifiers of the crystal structure of precipitates so 

that they do not adhere to surfaces and form scale. 

Used in conjunction with caustic soda.

washers. Conversely, water-soluble surfactants produce stable foams and are used in 

foam cleaners. Surfactants are able to disrupt biological membranes and consequently 

many are powerful biocides. 

Anionic surfactants have negatively charged dissociated groups. They have the 

properties of soaps and are used as conveyor belt lubricants. Conversely, cationic 

surfactants have positively charged dissociated groups. Some are used as corrosion 

inhibitors in water treatment systems, for example in pasteurizer sprays. The most common 

cationic surfactants are quaternary ammonium compounds. They have disinfectant activity. 

Amphoteric surfactants have acharge that is dependent on the pH of the medium they are 

added to. They have good biocide activity and are used in soak baths or as foam sprays. 

Non-ionic surfactants are electrically neutral and are used as CIP additives. 

Disinfectants (often called sanitizers) have several applications in ensuring microbiological 

cleanliness in brewing. Where process sterility is important, roughly from wort 

cooling onwards, disinfectants are incorporated into terminal rinses at the end of CIP to 

ensure that microbiological loadings remain at low levels. They are used in soak baths for 

the same reason. Disinfectants are used in foam spray cleaners and are used as additives to 

cleaning agents for some manual operations. They are incorporated into non-process water 

to prevent microbial growth. Several types are available depending on the application. 

Disinfectants should have anumber of attributes (Singh and Fisher, 1999). They should 

exhibit biocide activity towards abroad spectrum of micro-organisms at low concentration, 

be inexpensive and have no effect on plant, product or other chemical agents that they may 

comeintocontactwith.Nonemeetalloftheseneedsandsomecompromisesmustbemade. 

Several halogen-containing compounds exhibit disinfectant properties due to their 

oxidizing properties (see Chapter 3). Those containing chlorine are most commonly used in 

brewing. Two forms are used, sodium hypochlorite and chlorine dioxide. In alkaline 

solutions, sodium hypochlorite is relatively stable and it is supplied in this form. At acid pH 

(< 5.0) it forms hypochlorous ions, the active form of chlorine which is a powerful biocide. It 

is used at a concentration of 50±300 mg.l 1 as a terminal sterilant in rinse liquors. Hypochlorite 

has several disadvantages the most serious of which is its ability to cause corrosion. It 

is quite unstable especially at low pH (< 5.0) where free chlorine is generated. Sodium 

hypochlorite reacts with some organic compounds to form chlorophenols and chloramines. 

These can produce taints. In addition, trihalomethanes are potential carcinogens. 

Some of the disadvantages of hypochlorite are avoided by chlorine dioxide 

(Cadwallader, 1992). This is becoming increasingly popular and is tending to replace 

hypochlorite as the disinfectant of choice for use in terminal rinses. In fact, it is often 

used to treat all water used throughout the brewing process (Chapter 3). Usually it is 

synthesized in special generating equipment in which HCl is mixed with a solution of 

sodium chlorite. The reaction between these two chemicals forms chlorine dioxide, which 

is dosed into water at a concentration of 0.1± 0.5 mg.l 1 . At this concentration it is nearly 

three times more effective as a biocide than hypochlorite. It is non-toxic, does not 

produce taints and does not cause corrosion. 

Iodine is complexed with surfactants to form preparations termed iodophores. These 

are powerful disinfectants and are used at concentrations of 10 mg.l 1 available iodine in 

soak baths and in spray cleaners. At high concentrations they can be corrosive and impart 

taints and so care must be taken to ensure that they are diluted according to the supplied 

instructions. Preparations containing bromine (0.5 mg.l 1 available bromine) are used in 

the treatment of re-circulating water. 

Peracetic acid (CH3 CO OOH) is often incorporated into final rinses as a terminal 

sterilant. It is a powerful disinfectant owing to the fact that it decomposes to form acetic 


acid and hydrogen peroxide. It is used cold at a concentration of 75±300 mg.l 1 . It is 

difficult to handle in concentrated form, it has an unpleasant odour and it is corrosive. For 

these reasons the bulk supply must be stored in portable tanks in a bunded area. The 

storage area should be vented since over time peracetic acid decomposes and generates 

gaseous oxygen. The difficulties with handling mitigate against its use and increasingly it 

is being replaced with chlorine dioxide. 

Non-oxidizing disinfectants used in brewing include quaternary ammonium cationic 

surfactants, amphoteric surfactants and biguanides. All are used as components of soak 

baths. Each has strengths and weaknesses. They are too foam active to be used in CIP. 

Quaternary ammonium compounds, used at a concentration of around 200 mg.l 1 , show 

little activity against Gram negative bacteria and they can impart unpleasant fishy taints. 

Amphoteric types are equally effective against Gram negative and Gram positive bacteria 

at concentrations around 1,000 mg.l 1 . They are ineffective against moulds and yeast. 

Biguanides are derived from guanidine. They are active disinfectants towards a similar 

spectrum of micro-organisms as amphoteric surfactants but only within the range of 

pH 3.0±9.0. Biguanides are also used to disinfect recirculating water supplies. 

17.7.4 Cleaning beer dispense lines 

The microbiological integrity of small pack beers must be guaranteed throughout their 

shelf-lives. For draught products microbiological quality assurance must extend beyond 

the brewery to the premises where they are dispensed. There is no benefit to be gained in 

ensuring the highest standards of quality up to the point at which beer leaves the brewery 

without extending this to the management of hygiene within licensed premises. Central to 

this is the proper cleaning of dispense lines and associated equipment. 

Adequate cleaning of beer lines prevents the formation of biofilms. Materials best 

suited for use in beer lines have been investigated (Thomas and Whitham, 1996). A 

variety of materials were investigated for their ability to support biofilms. Nylon and 

medium-density polythene were the least favourable supports and these are recommended 

for use in beer lines. PVC was the least suitable material. Best practice for line 

cleaning has been reviewed (Treacher, 1995). Several proprietary line cleaning agents are 

available and all share common features. They are supplied as concentrates and require 

dilution with water to a concentration of 1% v/v. They contain sodium hydroxide as the 

primary detergent together with sodium hypochlorite (250 mg.l 1 chlorine) for 

disinfection purposes and a sequestrant for water softening. Several contain indicator 

dyes to provide a visual indication that cleaner is present in beer lines. 

Beer lines should be cleaned at least every seven days. First the line is flushed with 

clean water to eliminate any beer. It is then filled with detergent and left to soak for at 

least 20 minutes and no more than 30 minutes. During this time beer coolers should be 

switched off to prevent freezing. During the soak period the detergent action can be 

improved by agitation and half way through the soak the detergent should be replaced 

with a fresh supply. Finally, the lines should be flushed with clean water until no more 

cleaner can be detected, e.g., with indicator papers. The water should then be chased 

through with beer at which point the line is ready for re-use. During the soaking period it 

is essential that the lines and dispense equipment are completely filled with cleaning 

fluid. Any venting apparatus, etc., should be fully open during the initial filling 

procedure. Coupling devices, etc., must be removed and cleaned separately before the 

lines are soaked. Extending the soak time beyond the recommended period can damage 

dispense equipment and lead to taints being imparted to beers. 


17.7.5 Validation of CIP 

It is essential that procedures are followed to ensure that CIP processes are carried out 

correctly. Two types of validation procedure are used. Firstly, checks must be made to 

confirm that the conditions employed during the cleaning process were within the 

predetermined specification. Secondly, the cleanliness of the plant must be assessed 

against predetermined standards. 

CIP checks include cycle times, temperatures and the strengths of cleaning chemicals. 

Assuring that the correct concentration of caustic soda-based detergent is used is worthy 

ofspecialcomment. Commonly,the strengths ofsolutions ofcausticsoda areassessed by 

measurement of conductivity. Such readings can be misleading since conversion of 

sodium hydroxide to sodium bicarbonate following exposure to carbon dioxide does not 

produce a change in this parameter. Preferably, the concentration of caustic soda 

solutions should be checked by off-line titration. 

The cleanliness of plant after CIP can be checked using ATP bioluminescence 

(Section 17.3.1). Where possible this should be supplemented with visual checks to 

ensure that spray-balls are functioning correctly and no shadow areas exist. Recently the 

use of a video camera, termed the `Topscan', mounted in the top of cyclindroconical 

fermenters has been recommended for examining vessel cleanliness (Wasmuht and 

Weinzart, 1999). Validation of CIP is essential when new plant is commissioned. Since 

CIP is a costly and time-consuming process it is necessary to employ conditions that 

provide the desired level of cleaning at the lowest cost. 

17. 8 References 

AMAHA, M., KITABATAKE, K., NAKAGAWA, A., YOSHIDO, J. and HARADA, T. (1974) Bull., Brew. Sci., 20, 35. 

AMERICAN SOCIETY OF BREWING CHEMISTS (1992) Methods of Analysis, 8th edn, American Soc. Brew. 

Chem., St. Paul, Minnesota. 

ARCHIBALD, F. S. and FRIDOVICH, L. (1981) J. Bacteriol., 146, 928. 

BOEIRA, L. S., BRYCE, J. H., STEWART, G. G. and FLANNIGAN, B. (1999a) J. Inst. Brew., 105, 366. 

BOEIRA, L. S., BRYCE, J. H., STEWART, G. G. and FLANNIGAN, B. (1999b) J. Inst. Brew., 105, 376. 

BOULTON, C. A. and QUAIN, D. E. (2001) Brewing Yeast and Fermentation, Blackwell Science Ltd., 

Oxford. 

BRIDSON, E. Y. (1998) Oxoid Manual, 8th edn, Oxoid Ltd., Hampshire, UK. 

CADWALLADER, S. D. (1992) Ferment, 4, 380. 

CASE, A. A. and LYON, A. I. L. (1956) J. Inst. Brew., 62, 477. 

CZECHOWSKI, M. H. and BANNER, M. J. (1992) MBAA Tech. Quart., 29, 86. 

DE LEY, J., SWINGS, J. and GOSSELE, F. (1984) `Key to the genera of the family Acetobacteraceae'. In 

Bergey's Manual of Systematic Bacteriology, 9th edn, Vol. 1., N. R. Krieg and J. G. Holt, eds, 

Williams and Wilkins, Baltimore, p. 268. 

DIMICHELE, L. J. and LLEWIS, M. J. (1993). J. Amer. Soc. Brew. Soc., 51, 63. 

ENGELMANN, U. and WEISS, N. (1985) Systematics and Appl., Microbiol., 6, 287. 

EUROPEAN BREWERY CONVENTION (1998), Analytica Microbiologica Vol. II 5th edn, H. C. Verlag, 

Germany. 

FLANNIGAN, B. (1999). `The microflora of barley and malt'. In Brewing Microbiology, F. G. Priest and I. 

Campbell, eds, Aspen Publishers, Inc., Gaithersburg, Maryland, pp. 83±126. 

FLEET, G. (1992) Crit. Rev. Biotechnol., 12, 1. 

GILLILAND, R. B. (1971) J. Inst. Brew., 77, 276. 

GINGELL, K. and BRUCE, P. (1998) Proc. 23rd IOB Conv., Asia Pacific Sect., Perth, 134. 

GRAY, P. P. and KAZIN, A. D. (1946) Wallerstein Lab. Commun., 9, 115. 

GUTTERIDGE, C. S. and PRIEST, F. G. (1999) `Methods for the rapid identification of microorganisms'. In 

Brewing Microbiology, F. G. Priest and I. Campbell, eds., Aspen Publishers, Inc., Gaithersburg, 

Maryland, pp. 239±270. 

HAIKARA, A. (1989) Proc. 22nd Cong. Eur. Brew. Conv., Zurich, 537. 

HAIKARA, A. and LOUNATMAA, K. (1987). Proc. 21st EBC Cong., Madrid, 473. 

HAMMOND, J. R. M. and ECKERSLEY, B. W (1984) J. Inst. Brew., 90, 167. 

HYSERT, D. W., KOVECSES, F. and MORRISON, N. M. (1976) J. Am. Soc. Brew. Chem., 34, 145. 


INSTITUTE OF BREWING (1997) Methods of Analysis, Vol. 2, Microbiological, I. O. B., Clarges St., 

London. 

JACKSON, A. P. (1988) Tech. Quart. Master Brewers Assoc. Americas, 25, 104. 

JOHNSTON, M. A. and DELWICHE, E. A. (1965) J. Bacteriol., 90, 347. 

KENNEDY, A. L. and HARGREAVE, L. (1997) Proc. EBC Symp. Monograph XXVI, Stockholm, 58. 

KITABATAKE, K. and AMAHA, M. (1974) Bull. Brew. Sci., 20, 1. 

MAGLIANI, W., CONTI, S., GERLONI, M., BERTOLOTTI, D. and POLONELLI, L. (1997) Clin. Microbiol. Rev., 

10, 369. 

MAGNUS, C. A., INGLEDEW, W. M. and CASEY, G. P. (1986) J. Am. Soc. Brew. Chem., 44, 158. 

MAULE, A. P. and THOMAS, P. D. (1973) J. Inst. Brew., 79, 137. 

MCCAIG, R. and MORRISON, M. (1984) J. Am. Soc. Brew. Soc., 42, 23. 

MUNDY, A. P. (1997) Proc. EBC Symp. Monograph XXVI, Stockholm, 141. 

OGDEN, K (1987) J. Inst. Brew., 93, 302. 

OGDEN, K. (1993). J. Inst. Brew., 99, 389. 

PHILLISKIRK, G. and YOUNG, T. W. (1975) Antonie Van Leeuwenhoek, 41, 147. 

PLATT, D. (1986) Brew. Dist. Internat., 16, 20. 

PRESCOTT, L. M., HARLEY, J. P. and KLEIN, D. A. (1996) Microbiology, McGraw-Hill, USA. 

PRIEST, F. G. (1999) `Gram positive brewery bacteria'. In Brewing Microbiology, F. G. Priest and I. 

Campbell, eds., Aspen Publishers, Inc., Gaithersburg, Maryland, pp. 127±161. 

QUAIN, D. E. (1999) Proc. 27th EBC Cong., Cannes, 239. 

RUSSELL, I. and DOWHANICK, T. M. (1999). `Rapid detection of microbial spoilage'. In Brewing 

Microbiology, F. G. Priest and I. Campbell, eds., Aspen Publishers, Inc., Gaithersburg, Maryland, 

pp. 209±235. 

RYDER, D. S., MURRAY, J. P. and STEWART, M. (1978) MBAA Tech. Quart., 15, 79. 

SAMI, M., YAMASHITA, H. and HORONO T. (1997a) J. Ferment. Bioengin., 84, 1. 

SAMI, M. YAMASHITA, H., KADOKURA, H., KITAMOTO, K., YODA, K. and YAMASAKI, M. (1997b) J. Am. Soc. 

Brew. Chem., 55, 137. 

SHIMWELL, J. L. (1936) J. Inst. Brew. 42, 585. 

SIMPSON, W. J. (1987) J. Inst. Brew., 93, 313. 

SIMPSON, W. J. (1993) J. Inst. Brew., 99, 405. 

SIMPSON, W. J. (1999) Brewers' Guardian, May, 24. 

SIMPSON, W. J. and FERNANDEZ, J. L. (1992) Lett. Appl. Microbiol., 14, 13. 

SIMPSON W. J. and FERNANDEZ, J. L. (1994) J. Am. Soc. Brew. Chem., 52, 9. 

SIMPSON, W. J. and HAMMOND, J. R. M. (1989) J. Inst. Brew., 95, 347. 

SIMSON, W. J. and HAMMOND, J. R. M. (1991) Proc. 23rd EBC Cong., Lisbon, 185. 

SIMPSON, W. J., HAMMOND, J. R. M., THURSTON, P. A. and KYRIAKIDES, A. L. (1989) Proc. 23rd EBC Cong., 

Lisbon, 185. 

SINGH, M. and FISHER, J. (1999). `Cleaning and disinfection in the brewing industry'. In Brewing 

Microbiology, F. G. Priest and I. Campbell, ed., Aspen Publishers, Inc., Gaithersburg, Maryland, 

pp. 271±300. 

SMITH, N. A. (1994) J. Inst. Brew., 100, 347. 

SPECKMAN, R. A. and COLLINS, E. B. (1973) Appl. Microbiol., 26, 744. 

STICKLER, D. (1999) Curr. Opinion in Microbiol., 2, 270. 

TAKAHASHI, T., NAKAKITA, Y., MONJI, Y., WATARI, J. and SHINOSTUKA, K. (1999a) Proc. 27th Cong. EBC, 

Cannes, 259. 

TAKAHASHI, T., NAKAKITA, Y., SUGIYAMA, H., SHIGYO, T. and SHINOTSUKA, K. (1999b) J. Biosci., Bioeng., 

88, 500. 

THOMAS, K. and WHITHAM, H. (1996) Proc. EBC Symposium, Edinburgh, Monograph XXV, 124. 

TREACHER, K. (1995) Brew. Guard., Aug., 19. 

TSUCHIYA, Y., KANO, Y. and KOSHINO, S. (1992a) J. Am. Soc. Brew. Chem., 51, 40. 

TSUCHIYA, Y., KANEDA, H., KANO, Y. and KOSHINO, S. (1992b) J. Am. Soc. Brew. Chem., 51, 64. 

VAN VUUREN, H. J. J. (1999) `Gram negative spoilage bacteria'. In Brewing Microbiology, F. G. Priest and 

I. Campbell, eds., Aspen Publishers, Inc., Gaithersburg, Maryland, pp. 163±191. 

VERZELE, M. (1986) J. Inst. Brew., 92, 32. 

WAINWRIGHT, T. (1972) Brewers' Digest, 47, 78. 

WASMUHT, K. and WEINZART, M. (1999) Brauwelt Internat., 17, 512. 

WHITE, F. H. (1994) Proc. EBC Symp., Monograph XXI, Nutfield, 2. 

YASUI, T. and YODA, K. (1997) Appl. Environ. Microbiol., 63, 4528. 

YOUNG, T. W. (1987) `Killer yeasts'. In The Yeasts, 2nd edn, Vol. 2. A. H. Rose and J. S. Harrison, eds, 

Academic Press, London, pp. 131±164. 


18 

Brewhouses: types, control and economy 


The objective of this chapter is to consider the evolution of brewhouses and to consider 

briefly the diversity of design of breweries that are operated today. Control systems and 

economic aspects of operation will be considered. 

Brewhouse equipment must operate in a reproducible way to yield wort with a 

composition that is expected from an understanding of the biochemistry of the process 

(Chapter 4). In large breweries (say, greater than 500,000hl, or 300,000 imp. brl annual 

volume) of great importance is the ease with which brewhouse equipment can be 

automated, however, there is a warning here. Biochemists study enzymes in model 

systems at controlled pH value and temperature often approaching ideal conditions. The 

environment in abrewery mash is not ideal and sometimes unpredictable results will 

occur. The level of automation should, therefore, be such that manual intervention can 

easily take place when things go wrong. This will avoid the production of a series of 

worts of poor quality with consequent adverse quality and economic effects. 

A good brewhouse, then, is a set of vessels which will allow the biochemical reactions 

of brewing to take place in a controlled way as close as possible to the ideal environment 

in which the enzymes have been mostly studied. Engineers, therefore, should utilize the 

experience of brewers and scientists so that the equipment they design and sell will 

effectively support the critical enzyme activities needed for wort production. 

18.2 History of brewhouse development 

Visiting different breweries reveals considerable variation in the methods of construction 

and the layout of the plant. Pioneering work in the late 19th century by Horace Brown 

and others (Anderson, 1993) stimulated brewers to think even more carefully about 

brewhouse design. Economic pressures also became more acute at this time and 

competition in the brewing industry became a reality. Directors of Companies demanded 

adequate financial returns on the capital employed in the brewery and looked to make 


profit from the brewing of beer. In the late 19th century most brewers were brewers for 

sale and did not own the premises in which beer was sold. The cost of beer production, 

therefore, now came to be understood and its control was vital to Company success. This 

happened in Europe, particularly in the UK and in North America (Anderson, 1993). 

Brewers were thus forced to look critically at the design of the plant they were using to 

produce their beer. 

To lower costs it was appreciated that economies of scale were important and larger 

breweries (>300,000hl pa) were built, mainly in the UK. In the middle of the 19th 

century in Germany, however, there were around 15,000 breweries and many of these 

were extremely small, concentrating on brewing for one retail outlet. Some of these 

breweries, only partially modernized, exist today. However, it has been much easier to 

introduce new technology and lower costs in larger breweries and this has been the 

fashion in the UK, North America, Scandinavia and Japan throughout the 20th century. 

Developments in the USA were, of course, severely restricted by the introduction of the 

prohibition of alcohol that lasted from 1919 to 1933. 

The choice of raw materials for brewing has also influenced developments. Brewers 

established themselves commercially in the German States in the 15th century and 

municipal laws closely controlled their activities. To avoid the use of substitute 

materials,someofwhichhadbeeninjurioustohealth(someflavourings replacinghops), 

the purity law, the Reinheitsgebot, was introduced in 1516 and has applied throughout 

Germany since 1906. The law declares that beer can be made only from barley, water 

and hops. This has strongly influenced the development of brewing technology in 

Germanyinthelast100years.TheReinheitsgebotappliedinmanycountriesintheearly 

stages of brewing development, particularly as so many German brewers emigrated. But 

by 1870 American brewers had experimented with maize and rice mash tun adjuncts to 

economic advantage and did not go back to barley malt alone. In the UK there was a 

plentiful supply of sugar from the West Indies and again this was used to advantage to 

lower costs. Against this background the different trends in brewhouse improvement 

must be set. 

18.2.1 The tower brewery lay-out 

The development of the steam engine by James Watt in 1765 provided the basis for the 

industrialrevolutioninVictorianBritain.Brewerswerequicktorelatethescientificwork 

ofHoraceBrowntoWatt'sgreatinventionandalsotothepioneering workofJouleinthe 

early 1840s (working in the laboratory of the family brewery in Salford) leading to the 

First Law of Thermodynamics, and hence the concept of the mechanical equivalence of 

heat. 

Brewers in the 19th century quickly realized that making more use of gravity could 

lower costs in the brewhouse. The concept of the `Tower' brewery was born with 

consequent savings in energy and manual labour. The need for excessive pumping of 

worts was eliminated, resulting in quality improvements as potential sources of infection 

and oxidation were lessened. 

Tower breweries had aflat roof supporting water cisterns and tanks (Fig. 18.1). A 

considerable weight had to be supported on a small area. The buildings, therefore, had to 

be very substantially constructed with solid foundations and thick walls. These methods 

of construction were immediately available in the mid to late 19th century through to the 

1950s (Jeffery, 1956), when cheaper building materials were developed. This meant that 

the excessive demands of the civil construction of the Tower brewery became very 


Mill 

Grist 

case 

A 

B 

C 

Dust 

quencher 

Hot liquor 

Malt screen 

Mash tun 

Slotted 

false bottom 

Malt store 

Cold liquor tank 

tanks 

A – Mashing liquor 

B – Sparge liquor 

C – Underlet liquor 

Steel’s masher 

Sparge arms 

Rakes 

Spend 

safe 

Underback 

Hop store 

Cellar 

Malt 

hopper 

Condenser 

Steam 

inlet 

Wort 

pump 

Primings 

dissolving 

vessel 

Copper 

Wort 

pump 

Steam 

jacket 

Sugar dissolving 

vessel 

Heater 

Steam outlet 

Hop sparge 

Hop back 

Slotted 

false bottom 

Wort 

receiver 

Primings cooler 

Primings storage 

vessel 

Enclosed 

wort 

refrigerator 

Spreader 

W.W. attemperator 

liquor: inlets and outlets 

Suction yeast Parachute 

skimmer yeast skimmer 

W W 

Attemperator 

Fermenting 

Racking 

back 

vessels 

Yeast 

collecting 

vessels 

Yeast 

press 

Fig. 18.1 Sectional representation of a traditional small ale `Tower' brewery, operating with one 

elevation of the wort to the Wort Boiler (Jeffrey, 1956). 

expensive and few were built after the 1960s. In most Tower breweries pumping only 

once elevated the wort. This was usually after separation from the extracted malt when 

the wort would be pumped to the wort boiler. The wort boiler would be situated at the top 

of the brewery and then the wort would fall by gravity to the hop separator (hop back) and 

then to a wort receiver prior to cooling. 

These Tower brewhouses were built to be as fire resistant as possible and the use of 

wood was avoided. The buildings were essentially brick and concrete with steel being 

used in later developments. One of the essentials of a Tower brewery was a hoist or lift 

from the bottom to the top of the building. This in itself created a problem for fire 

prevention and the lift had to be installed in a brick shaft with self-closing doors of sheet 

iron on each floor. 

Tower breweries also had very small vessels by the standards of the late 20th and early 

21st centuries. A large wort boiler would be no more than 325 hl (200 imp. brl) in 

capacity whereas today a wort boiler could have a volume of 1,000 hl (600 imp. brl) or 

more. In big brewing companies economies of scale tend to demand large vessels and 

these breweries are often producing huge volumes of one or two brands. A Tower 

brewery would simply be too expensive to construct to satisfy these requirements. For a 

modern brewery a horizontal lay-out using cheaper building materials offers a lower-cost 

solution. 


18.2.2 The horizontal brewery lay-out 

As often is the case when examining developments in an historical context no system is 

found to be entirely designed to one set of rules. We therefore have breweries that have 

partial characteristics of vertical or horizontal design. Small breweries, particularly on 

mainland Europe, were often of two-vessel design (Fig. 18.2) having one vessel as a mash 

tun and a separator and a second vessel as a mash cooker and wort kettle. The lay-out of 

these breweries was usually horizontal. It was natural, therefore, that when these plants 

were extended to four or more vessels that these were laid out on the horizontal. The 

construction of the building to house the vessels is made much simpler in this situation. 

This usually means a steel-framed building with brickwork to first-floor level and then 

some form of mild steel profile cladding and glass to the upper levels. Fireproofing is 

simpler and there is no need for a complex design of lift shaft. Costs of construction are 

much reduced and the money can be spent on ensuring the best quality of vessels, 

pipework and control systems. The Tower brewery allowed savings in manpower and 

energy use but this was overtaken by high building costs and the subsequent ease of 

automation of very large horizontally laid out plant allowed further savings, which more 

than compensated for any energy increases. 

Mash 

mixer 

Grist 

Mash or 

wort boiler 

Lauter tun 

Section 

Fig. 18.2 Two-vessel brewhouse with mash mixing vessel on the left, mash and wort boiler in the 

centre and lauter tun on the right (Briggs et al., 1981). 


Plan

18.3 Types of modern brewhouses 

Amodern horizontal brewery will have acombination of vessels: 

· amash conversion vessel or tun 

· amash cooker 

· awort separation device such as alauter tun or mash filter 

· awort boiler 

To these vessels could be added: 

· aheated wort buffer tank 

· ahop separator such as awhirlpool. 

ThedetailedoperationofallthesevesselsisconsideredinChapter6andtherelatedscience 

is discussedinChapter 4.Fromthese Chapters theessential technology ofthelargemodern 

brewhouse emerges and this is obviously the most significant for brewing: science and 

practice. Different types of brewhouse are, however, also successful in the 21st century. 

Brewhouses have to produce a volume of wort of the right quality to support 

fermentation as frequently as is required by the market in which the brewing company is 

operating. This, of course, derives from the brands of beer the brewery is expected to 

produce. The requirements from amicro- or pub brewery are very different from a 

brewery producing very large volumes of anational or international brand. Abrewery 

producing high volume brands might be of asize to produce between 3.5 and 10 million 

hl (2 to 6million imp. brl) of beer per year. There then would be arange down in size 

throughmedium-sizedplantsof0.5to0.8millionhl(0.3to0.5millionimp.brl)tomicrobreweries 

as small as 1,000hl (600 imp. brl) per year or even less. 

All these types of plant are important in the brewing industry of today; this is from 

where the great variety of beer that we enjoy derives and this has been an important change 

in the last 20 years. We have seen the emergence of the global brand produced in many 

countries to very strictly defined principles and the rise of the so-called boutique or niche 

brands produced to satisfy avery local demand. There have been casualties on the way and 

many breweries in Europe and North America and other parts of the developed world have 

closed. There is almost apolarization from the global to the local which large companies 

try tobridgewiththe adage`thinklocal actglobal' butitis oftendifficult toreconcile these 

conflicting factors. This has implications for astudy of the diversity of brewhouses in the 

21st century. The driving force for any business, however, is that it must be financially 

successful and the yield and quality of wort from abrewhouse is important to both the 

brewer of the global brand and the micro-brewer serving avery local community. 

In large brewhouses (Chapter 6), brewers may be trying to cope with the demands of up 

to 14 brews in one day. There are very different demands facing brewers operating other 

brewhouses. These other brewhouses can be broadly categorized as experimental 

brewhouses and micro- or pub breweries. Beer produced in the former category does not 

alwayshavetobecommerciallyacceptablewhereasbeerfromapubbrewerymusthavethe 

highest acceptability. The requirements of the two brewhouses are therefore quite different. 

18.3.1 Experimental brewhouses 

These brewhouses can be used solely for training purposes or can be used for a 

combination of training and raw material evaluation and new product development. It is 

often difficult to flavour match beers produced on the small scale to those produced on 


the production plant. New product development, therefore, is frequently carried out on 

the large scale and is correspondingly very expensive! However, raw material evaluations 

such as those of malt prepared from a new barley variety or new hop varieties are usually 

successful. Indeed in the UK, pilot scale trials coordinated by The Institute and Guild of 

Brewing are crucial in gaining approval of new varieties of both barley and hops. These 

trials have operated for the last 30 years and have prevented barleys of poor brewing 

quality becoming widely grown. Beers produced in experimental trials can sometimes be 

blended back into mainstream brews at say 10% to minimize costs. 

Experimental breweries can vary considerably in size. Breweries from 5 to 200 litres 

have been described (De Clerk and De Clerk, 1965; Baetsle, 1983). And there is also the 

genuine pilot brewery that can be between 10 and 100 hl in brewlength (Moll and 

Midoux, 1985). The greater the size means the greater the likelihood results of trial brews 

will be closer to commercial results. Most international brewers will have pilot plant 

facilities and all training institutes and research organizations will have this type of 

equipment. Results must be interpreted with care and the skill and experience of the 

brewer in charge of the equipment is of major importance. The brewing equipment must 

be capable of reproducible operation and it is highly desirable to carry out a statistical 

evaluation so that the value of the least significant difference between the various 

parameters is known. Only in this way will truly meaningful results be obtained. 

18.3.2 Micro- and pub breweries 

The international brands of beer are highly specified and extremely consistent. To ensure 

this consistency and to provide the reproducibility necessary to brew the beers all over the 

world these beers tend not to have strong flavours. They tend to be bland and in particular 

have low bitterness and lack hop character. The international brands are also tending to 

taste more and more similar. Beer drinkers now demand more choice in their beers and 

have looked for more original flavours, usually with higher levels of bitterness. This has 

been particularly driven by changes in the USA (Lewis and Lewis, 1996) since 1985 when 

the micro-breweries started to emerge. This was partly as a result of legislation in 1982 that 

legalized `Brewpubs'. Micro-brewing in the USA still occupies a small volume but it has 

grown at 50% per year since 1985 to take over 2% of the whole market. The structure of the 

industry is complex with a large variation in the size of individual companies but genuine 

micro-brewers are driven by real passion for what they do. Consequently their enthusiasm 

has had an impact on the major brewers who have sought to cash in on market opportunities 

by introducing their own boutique brands often under disguised labelling. This has further 

fuelled competition and development. The strongest markets for micro-brewers in the USA 

are in California, Oregon, New England and Florida. 

Most micro-breweries in the USA use infusion mash tuns for extraction of the malt 

and wort separation. This equipment is relatively low cost and allows a link to the British 

tradition of ale brewing. Indeed this is the normal choice throughout the world. 

Frequently the brewing equipment is on display to the public consuming the beer. The 

breweries vary considerably in size often from about 10 to 50 hl (6 to 30 imp. brl) in 

brewlength with three to five brews per week. Consumers will tolerate some variation in 

the beer in the interests of presumed authenticity but will not tolerate any hint of 

infection. Sanitation and microbiological control in these breweries is therefore of 

paramount importance and if this fails then this is the most frequent cause of 

unacceptable beer. Modern micro-brewing apparatus usually has associated CIP 

equipment and this must be used rigorously. Yeast must also be checked before pitching 


Table 18.1 Comparison of analyses of generic national lager and generic micro-brewed lager in 

the USA (Lewis and Lewis, 1996) 

Parameter National lager Micro-brewed lager 

Original gravity (ëP) 11.5 15.5 

Alcohol by weight (%) 3.8 6.0 

Bitterness (IBU) 12 35 

Colour (ëEBC) 4 25 

Table 18.2 Japanese consumption and market share of beer and happoshu (Malone, 2001) 

Volume (000 hl) 1996 1997 1998 1999 2000 

Beer & happoshu 72,485 72,095 72,351 72,264 71,764 

Beer 69,770 67,929 62,561 58,327 55,510 

Happoshu 2,715 4,167 9,790 13,937 16,254 

Market share (%) 1996 1997 1998 1999 2000 

Beer 96 94 87 81 77 

Happoshu 4 6 14 19 23 

for purity. Bottling equipment is often of poor quality. In addition to ensuring that it is 

effectively cleaned, every effort must be made to limit the occurrence of headspace air 

that has exceeded 4 ml in some cases with disastrous consequences causing oxidation and 

rapid flavour deterioration. Micro-brewed beers are often prepared with 100% malt and 

as such may be inherently less stable than national beers. This is sometimes offset by 

their normally much higher hop content (Table 18.1). 

There have been some interesting developments in Japan (Malone, 2001), where a 

mature beer market was in need of some innovation. The licensing laws in Japan were 

changed in 1994 allowing breweries to make a minimum of 600 hl (360 imp. brl) of beer 

per year. There are now around 300 small breweries producing local beer (ji-biru). To 

some extent growth of these breweries has been stimulated by the growth of local sake 

breweries (ji-zake). There has also been very significant growth in `happoshu' (sparkling 

drinks). These products have less than 25% malt in the grist and are subjected to much 

lower rates of duty. The drinks thus retail at lower prices and have grown in popularity 

(Table 18.2). This has created a market opportunity for more flavoursome micro-brewed 

beers that will grow, particularly if the Government moves to reduce the differences in 

duty payments between happoshu and beer. 

The majority of the small breweries in Japan produce top-fermented beers but some 

have tried to produce German-style lagers in this way and have found this difficult, with 

dire commercial consequences. A shakeout of breweries is now under way but it seems 

likely that growth above the current 0.5% market share of ji-biru will continue. 

Throughout the world the most successful micro-breweries tend to be those producing 

draught products for consumption with a very short shelf-life of no more than four weeks. 

Problems arise when producing small pack products and trying to extend the shelf-life to 

six months and beyond. This usually results in oxidized beers that fail dismally when 

compared to beers brewed in large breweries. This comparison is best avoided. Microand 

pub-brewed beers taste best when drunk fresh and this is the unique selling point to 

exploit. The most successful of these operations keep things simple and use a two-vessel 

brewery with a mash tun and a kettle that can double as a mash kettle and wort kettle as 

necessary. 


As the 21st century progresses it will be interesting to see how the micro-brewing 

industry develops. International brands of beer will continue to be heavily advertised 

throughout the world. To compete, local beers will need to be brewed in scrupulously 

clean plants by dedicated, well-trained professionals, they will have strong and 

characteristic flavours and be drunk as fresh as is possible. 

18.4 Control of brewhouse operations 

A key feature of any good brewhouse is reproducibility. An environment must be 

provided in which the biochemical reactions can take place in a controlled and optimized 

way. This was originally achieved by close operator involvement in constant checking of 

times and temperatures and taking remedial action where appropriate. Things are now 

different. There are very few operators, maybe only two on a 12-hour shift scheduled for 

12 brews a day, and they are likely to be led by a team leader controlling the whole of the 

process from raw material intake to the end of fermentation. The departmental manager 

may be a remote figure frequently most concerned with achieving the financial plan 

targets and writing the monthly report. 

18.4.1 Automation in the brewhouse 

In this increasingly common situation automation of the sequence of brewhouse 

operations is indispensable. This automation is no substitute for the brewhouse operator 

knowing what is happening in the vessels and his training in this respect is crucial. Very 

expensive mistakes can ensue if automation is slavishly followed when things are going 

wrong. The big change in automation in the last 20 years has been the gradual replacement 

of hard-wired relay logic systems with the silicon chip based programmable logic 

controllers. This has been coupled with the very rapid development of personal computers 

(PCs), which has provided user-friendly operator interfaces and the ability to store huge 

amounts of data. Concurrent with all this have been big reductions in the capital cost of 

equipment, as the `Microsoft' operating systems have come to dominate the world. The 

developments in PCs have facilitated and made cheaper industrial systems' development 

in what has truly been a buyer's market. A major problem is that systems have advanced so 

fast that obsolescence has become a factor that the brewer cannot ignore. The language of 

the control systems expert has also become difficult and obscure for the brewing 

technologist and this has sometimes led to increased expense when supplier and customer 

have not understood one another. The old adage, `the customer is always right' still holds 

but it is imperative in this area that the customer sets out very precisely what is required of 

the system he is buying. There is no substitute for the `Design Brief' and the Ùser 

Requirement Specification', which must form the basis of the Contract. The brewer (the 

customer) should remain the most important person in the deal. 

Systems are usually designed bespoke and are not universal. However it is possible to 

make some comment on the most successful systems. The way the system is put together 

is called architecture, which describes the links through from plant level to a management 

information system (MIS). Several important steps can be recognized. 

Sensors 

The success of the system, of course, starts with measurements. No amount of clever 

software will make up for an inaccurate or imprecise measurement. The quality of the 


initial measuring is paramount. Fortunately, in the brewhouse we usually want to measure 

temperature, pH value and pressure and equipment for these measurements is cheap and 

reliable. 

Programmable logic controllers (PLCs) 

The control at plant level is effected through PLCs. The PLC is a computer that is 

programmed to handle simple input and output instructions, e.g., òpen', `close'. Inputs 

are the signals from sensors and outputs are the instructions to the machines; pumps, 

motors, valves, etc. Normally a brewery company will specify one type of PLC from one 

manufacturer and this is highly desirable for consistency over the whole brewery site and 

for future changes to the system. 

Supervisory control and data acquisition systems (SCADA) 

PLCs link machines together and send out simple instructions to make changes in the 

process according to how they have been programmed. The operator requires information 

on what is happening and needs to store information for future use. This is provided 

through a SCADA system that can communicate with PLCs by an industrial Ethernet. 

The SCADA system can now have a very user-friendly interface through a PC (which can 

use familiar Microsoft Windows NT software) in the brewhouse. The SCADA system can 

also be linked to high-level MIS that can further be linked nationally or even 

internationally providing Head Office with direct access to an individual brewery. 

It is now possible to use this system architecture to hold the recipes of the brewing 

process and to schedule the brewhouse operations (Cooper et al., 2002). A series of 

reports can be obtained that could detail actual achieved mash temperatures, transfer 

times and run-off profiles, etc. Studying these reports allows predictions to be made and 

correlations to be sought which can lead to future brewhouse improvements. 

All these developments make the operation of the brewhouse easier to control in these 

days of very low manpower levels. It must again be stressed how important the training 

of the operator is so that timely intervention may be made when indicated by the control 

system. The development of the qualification the `General Certificate in Brewing and 

Packaging' (formerly the Foundation Certificate) by the Institute and Guild of Brewing is 

a particularly important step in this respect (Brookes, 1999). 

18.4.2 Scheduling of brewhouse operations 

The brewhouse is the most important part of the brewery in relation to its overall 

capacity. The number of brews that can be carried out in one day and the length of the 

brewing week are the relevant factors. Different businesses will have very different 

requirements. The driving force for the micro- or pub brewery will be to have the beer 

presented to the customer in the pub in the freshest possible state. There will be 

seasonality here and the micro-brewer will need to be able to increase production in the 

summer to meet demand. However it is unlikely that there will be more than two or three 

brews performed in one day and the brewing week will probably be five days but those 

days may include the weekend so as to ensure maximum exposure to the public if 

brewing vessels are on view. This will give two days for thorough cleaning which is 

essential to maintaining product quality. 

The situation is very different in the large brewery producing national or international 

brands. It is now frequently the case that beer will not be delivered direct to the customer 

from the brewery but will go to a regional distribution centre where stock will be held 


(Chapter 22). The brewery's staff often feels remote from the customer in this situation. 

The brewery's success will be measured against its ability to produce beer to aplan and 

this plan will be derived from historical sales data and marketing forecasts. The plan will 

often be given to the brewery from acentral planning department who will monitor 

progress. The role of the brewery is clear. It must deliver the plan. 

The core of the plan then will be the number of brews that can be consistently 

delivered from the brewhouse. Working arrangements in the brewhouse will probably be 

based on some form of 168-hour cover where teams will cover the whole week and take 

rest days on a rotating basis. The weekends are not significant since any overtime 

payments are likely to be built in to an overall salary for the job. This makes scheduling 

easier. However, it is still very important to build in to the programme sufficient time for 

cleaning,maintenance andcalibrationofplantanditisunlikelythattheplan willdemand 

more than six-day brewing. There will also be time needed for maintenance and six-day 

brewing cannot be sustained indefinitely. Trade is usually such that there is considerable 

seasonality in demand with peak brewing being required in the summer and at national 

holiday periods or periods of Company sponsorship of major events such as football 

championships. These external factors become the driving force to the plan and influence 

the calculation of brewhouse capacity. The maximum capacity of the brewhouse will be 

the number of brews in aday length of brewing week number of weeks of brewing 

in ayear. To be prudent in calculating seasonality we might express the foregoing as 10 

6 46 ˆ2,760 brews per year. The planning department would build this information 

into its model of operations and hence the production and distribution plan would be 

derived. 

The actual volume of beer produced depends, of course, on the strength of wort 

delivered from the brewhouse and the amount of post-fermentation dilution carried out. 

This relates to the types of beer being brewed and the nature of the grists. Scheduling of 

brewhouse operations will usually include scheduling the intake of raw materials and this 

will also frequently be arranged centrally in large brewing groups. To service alarge 

brewery as illustrated above, considerable quantities of raw materials will require to be 

delivered each week: 

Let us assume that each brew is 1,000 hl ( 600 imp. brl) at agravity of 1048ë 

(48ëSacch, 12ëP). Let us also assume that the grist is 90% white malt and 10% sugar 

(Section 18.5). Each brew will require approximately 15 tonnes of malt and 1.5 tonnes of 

sugar solution. For asix-day week at ten brews per day this amounts to 900 tonnes of 

malt and 90 tonnes of sugar per week therefore we require around 36 loads of malt and 4 

loads of sugar solution per week. 

This represents aconsiderable schedulingoperationand requiresclose liaisonbetween 

the maltster, the sugar supplier and the brewer. To be prudent the brewer will probably 

want to hold at least one week's stock of raw materials on site and so will require at least 

ten 100-tonne malt bins and three to four 20-tonne sugar tanks. Of course there may also 

be arequirement for delivery and storage of smaller quantities of special malts and mash 

tun adjuncts (see Chapter 5). 

In recent years this whole subject has received much attention from planners and 

software experts and computer systems with the general title of materials requirements 

planning (MRP) and the further development of manufacturing reserve planning (MRPII) 

are now commonplace. The objective is to pick up the principles of `just in time' (JIT) 

production to minimize the stock holding of raw materials, those in-process and finished 

beer in the brewery. This in turn reduces working capital. Software systems in use include 

PRMS, PRISM and SAP. These systems can be extremely useful particularly in providing 


information to suppliers of raw materials. But they are no substitute for physical stock 

checks in the brewery to ensure that the materials required and apparently in stock are 

actually present! 

18.5 Economic aspects of brewhouses 

Fixed costs in breweries relate to manpower, utilities, repairs and depreciation and other 

costs such as rates and other local taxes. Variable costs relate to the efficiency in the use 

ofmaterials and losses of product incurred during processing. Of prime importance tothe 

efficiency of the brewhouse then, and hence to brewery variable cost, is the quantity of 

soluble material extracted from the grist. The more efficient this process the greater will 

be the contribution to the fixed costs of running the brewery, where contribution is 

defined as the income less the variable costs (sometimes called gross profit). 

The strength of wort is, of course, judged by the quantity of material in solution, the 

extract. This is usually measured as the specific gravity relative to water at aspecific 

temperature (see also Chapter 4), or by relating to the concentration (%w/w) of asucrose 

solution having the same specific gravity using the scales of Balling or Plato. A pale malt 

ground and extracted in cold water will yield 16±22% of its dry matter as soluble extract 

(the cold water extract, Chapter 4). The same malt ground and extracted at 65 ëC (149 ëF) 

will yield 75±83% of its dry matter as extract (the hot water extract). The extra substances 

yielded by hot water mashing derive from enzymic attack on initially insoluble materials, 

mainly starch. This forms the nub of potential economic improvement in the brewhouse. 

If increases in soluble extract can be consistently achieved then there will be reductions 

in the raw material bill, lower variable cost and brewhouse operations will be judged a 

financial success. Consider the preparation of 1,000 hl of wort at a gravity of 1,048 ë 

(48 ëSacch, 12 ëP), from a grist of 90% white malt and 10% sugar: 

Let the extract of the white malt be 300 l ë/kg (as is). Let the extract of the sugar 

solution be 320 l ë/kg (as is). Let the extract efficiency of the brewhouse be 97% (i.e. in 

normal circumstances, 97% of the extract as indicated by laboratory analysis would be 

expected to be extracted into brewery wort). The amount and strength of wort to be 

collected is: 

1000 48 ˆ 48; 000 hl …4;800;000 l † 

the quantity of materials required will be: 

white malt …4,800,000 0:9† 0:97 300 ˆ 14,845 kg 

sugar …4,800,000 0:1† 320 ˆ 1,500 kg 

The sugar is not affected by the 97% extract recovery because it is added directly to 

the wort boiler and is assumed to be utilized at 100%. Clearly, if the extract efficiency 

can be consistently raised to 98% then 1% less malt will be required, i.e., 148 kg of malt 

less. Over the course of 2,000 brews in a year this would amount to around 300 tonnes of 

malt, which could equate to variable cost savings of around US $96,000 per year. This is 

a considerable sum and illustrates the great importance of striving to achieve consistent 

improvement in brewhouse performance by achieving operating conditions to optimize 

enzyme activity. 


18.6 Summary 

Brewhouses are the heart of every brewery whether large multi-national or niche brand 

pub brewery. It is the extraction of the grist that is crucial to economic success and to 

yeast performance and beer quality. Brewhouses can be of different types and large 

brewhouses are capable of being operated with high efficiency and consistency whilst 

small brewhouses are capable of being controlled to yield worts for beers with unique and 

stronger flavours. Further improvements are likely to be restricted by the amount of 

money now being invested in basic research. In the long run this will be a short-sighted 

policy. The long-term winners will be those organizations that invest in studying how 

enzymes cope best with the heterogeneous conditions of the brewery mash. 


ANDERSON, R. G. (1993) Ferment, 6, 191. 

BAETSLE, G. (1983) Cerevisia, 7, 11. 

BRIGGS, D. E., HOUGH, J. S., STEVENS, R. and YOUNG, T. W. (1981) Malting and Brewing Science Volume 1, 

2nd edn, Chapman and Hall, London and New York. 

BROOKES, P. A. (1999) Proc. 7th. Conv. Inst. Brew. (Africa Section), Nairobi, xiii. 

COOPER, T. J., BARNES, Z. C. and MCFARLANE, I. K. (2002) Proc. 27th Conv. Inst. Brew. (Asia Pacific 

Section), Adelaide, 1. 

DE CLERK, J. and DE CLERK, E. (1965) Tech. Quart. MBAA., 2, 183. 

JEFFERY, E. J. (1956) Brewing, Theory and Practice, Nicholas Kaye, London, 18. 

LEWIS, M. J. and LEWIS, D. J. (1996) Proc. 6th Int. Brew. Tech. Conf., Harrogate, 227. 

MALONE, R. (2001) Brewers' Guard., 130 (9), 24. 

MOLL, M. and MIDOUX, N. (1985) Tech. Quart. MBAA., 22, 67. 


19 

Chemical and physical properties of beer 

19.1 Chemical composition of beer 

Beer, the final product of the brewing process, is designed to be drunk. It is acomplex 

mixture; well over 450 constituents have been characterized, and, in addition, it contains 

macromolecules such as proteins, nucleic acids, polysaccharides and lipids. Together all 

theseconstituentsproducethecharacterofbeer.Inthischapterwereviewthecompounds 

that have been found in beer but delay discussion of the influence they have on beer 

flavour until the next chapter. However, to prevent duplication, data on the taste 

thresholds of the compounds found are included in the Tables given in this Chapter. A 

useful introduction is given in the monograph Beer: Quality, Safety and Nutritional 

Aspects (Baxter and Hughes, 2001). Jurado has reviewed beer styles and provided 

analyses for Pilsener (2002a), Munich Helles (2002c), brown ales (2001a), strong ales 

(2002e), pale ales (incl. IPA) (2002e), red beers (2002b), wheat beers (2001b) and 

seasonal and special beers (2002d). The ranges of analytical values found are collected in 

Table 19.1. A. Piendl (Weihenstephan) has published analyses of beers from all over the 

world. In most cases the units used by the authors quoted have been retained. It is often 

not known whether parts per million (ppm) is mg/l or mg/kg. Parts per billion (ppb) is 

parts per 10 9 ( g/l or g/kg). 

Beer constituents can be divided into volatile and non-volatile components. The 

volatile components have greater vapour pressure and are responsible for the bouquet or 

aroma of beer. They are concentrated in the headspace above the liquid in a closed 

container and will pass into the distillate if the beverage is distilled. The complex mixture 

of volatile components either in the headspace or in a solvent extract of the beer can be 

resolved by gas-liquid chromatography, using either packed or capillary columns, and the 

components identified by mass spectrometry (GC-MS). The non-volatile constituents 

include inorganic salts, sugars, amino acids, nucleotides, polyphenols and hop resins 

together with macromolecules such as polysaccharides, proteins and nucleic acids. Such 

compounds are usually resolved by high precision liquid chromatography (HPLC). 

Before the advances in chromatography available today, brewers developed numerous 

methods of analysis to control their process and the quality of their products and many of 


Table 19.1 Beer analyses 

100% Malt Munich-style Brown Strong IPA Pale ales Red beers Wheat beers Seasonals 

Pilsener Helles ales ales and specials 

Ref. Jurado Jurado Jurado Jurado Jurado Jurado Jurado Jurado Jurado 

(2002a) (2002c) (2001a) (2002e) (2002e) (2002e) (2002b) (2001b) (2002d) 

No. of samples 18 33 22 8 17 58 23 27 34 

Specific 1.00682± 1.00258± 1.00599± 1.00676± 1.00739± 1.00471± 1.00627± 1.00449± 1.00342± 

gravity 1.01411 1.10166 1.01991 1.02715 1.01646 1.01680 1.01986 1.02385 1.03471 

(1.009740) (1.01485) (1.0132) (1.01236) (1.01326) 

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 

extract (2.48) (2.04) (3.36) (3.15) (2.67) 

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 

(% w/w) (4.30) (4.11) (3.97) (3.88) (4.04) 

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 

(% w/w) (4.30) (3.92) (5.20) (4.93) (4.53) 

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 

gravity ( o P) (11.97) (11.89) (12.855) (12.42) (12.34) 

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 

of fermentation (65.4) (68.4) (61.194) (62.0) (65.0) 

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 

(158.0) (156.2) (171.9) (165.7) 163.7) 

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 

(4.33) (4.37) (4.33) (4.21) (4.46) 

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 

(Lovibond) (6.0) (4.2) (32.8) (23.9) (11.2) 

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 

(IBU) (31.0) (18.4) (26.6) (21) 17.2) 

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 

(0.0) (0.04) (0.04) (0.04) (0.05) 

Sodium (Na, ppm) 3±45 11±50 21±106 24±322 14±113 13±123 17±100 14±86 13±150 

(34) (26) (39) (43) (43) 

Dimethyl sulphide (ppb) 3±73 ± ± ± ± ± ± ± ± 

(50±55) 

SO2 (ppm) ± 0±3.2 ± ± ± ± ± ± ± 


Table 19.2 Analyses of British beers (Anon., 1960, 1967) 

Quality OG Alcohol Unfermented Isohumulones 

(% v/v) matter (%) (mg/l) 

Draught bitter 1030.9±1045.3 3.0±4.6 27±45 20±40 

Draught mild 1030.7±1036.5 2.5±3.6 29±48 14±37 

Light ale (bottle or can) 1030.6±1038.9 2.9±4.0 30±40 16±38 

Best pale ale 1040.3±1050.3 4.3±6.6 21±43 19±55 

Brown ale 1030.2±1040.6 2.5±3.6 43±55 16±28 

Stout ± Guinness 1040.0±1046.1 4.4±5.1 30 55±62 

± Mackeson 1044.3±1047.6 3.7±3.8 49 27±31 

Strong ales 1065.9±1077.7 6.1±8.4 32±44 25±43 

Lagers 1029.7±1036.3 3.3±3.6 35±39 20±32 

these are retained in Analytica-EBC and by the Institute of Brewing and the American 

Society of Brewing Chemists (Section 1.15.1, p. 9). Some representative analyses are 

given in Tables 19.1 and 19.2. 

Another way to classify the organic constituents of beer is with reference to the 

heteroatoms present. Beers contain only trace amounts of hydrocarbons, the majority of 

the constituents contain carbon, hydrogen and oxygen. Small amounts of nitrogencontaining 

constituents are present of which the proteins are important for the physical 

properties of beer. Only low levels of sulphur-containing compounds are found but 

volatile sulphur compounds have low thresholds and so small amounts can influence the 

flavour of beer. Indeed, abnormally high levels of sulphur compounds can be responsible 

for off-flavours. Volatile sulphur compounds can be examined by gas chromatography 

usingeitheraflame photometricdetectororaSievers'chemiluminescentdetector bothof 

which are specific for sulphur-containing compounds. In general different beers contain 

different proportions of the same constituents. Only when novel raw materials are used 

will novel constituents be found in the beer as, for example, Belgian fruit beers such as 

Framboise. However, accidental or deliberate contamination of wort or beer with foreign 

micro-organisms may well produce new metabolites and flavours. 

19.1.1 Inorganic constituents 

The most abundant constituent of beer is water, the medium in which, in bright beer, all 

theotherconstituentsaredissolved.Brewingliquornormallycontainsonlytraceamounts 

of organic matter and the desirable cations and anions required in the liquor are reviewed 

inChapter3.Duringthebrewingprocessothersaltswillbeextractedfrommaltandhops, 

some ions may be precipitated on the break and others may be absorbed by the yeast so 

that the inorganic salts present in beer are very different from those in the liquor used. 

Data for the inorganic constituents of beer are collected in Table 19.3. 

The major ions are the cations potassium, sodium, calcium and magnesium and the 

anions chloride, sulphate, nitrate and phosphate. There are agreed international methods 

for measuring sodium, potassium, calcium and magnesium in beer by atomic absorption 

spectroscopy. In addition the ASBC gives methods for determining calcium and 

magnesiumbytitrationwithEDTA.Internationalmethodsarealsoavailabletodetermine 

the anions in beer, chloride, sulphate, nitrate and phosphate by ion chromatography. 

There is also an international method for chloride by conductometry, and Analytica-EBC 

gives agravimetric method for sulphate and an enzymatic method for nitrate. Excess 

nitrate in beer is undesirable as potentially nitrates can be reduced to nitrites, which with 


Table 19.3 Inorganic constituents of beer 

Constituent Source Concentration Reference 

range mg/l (ppm)* 

Aluminium German 0.1±1.24 Postel et al. (1983) 

Arsenic Lagers 0.02 Binns et al. (1978) 

Spanish 3.1±8.2 Cervera et al. (1989) 

Others 1.8±11.2 Cervera et al. (1989) 

Cadmium German 0.0002±0.020 Postel et al. (1983) 

Spanish 0.031±0.397 YbaÂnÄez et al. (1989) 

Others 0.095±0.677 YbaÂnÄez et al. (1989) 

Calcium British 40±140 Paul and Southgate (1978) 

German 3.8±102 (32.7) Postel et al. (1974) 

Lagers 10±135 (36) Binns et al. (1978) 

Chromium European (0.0072) Robberecht et al. (1984) 

Spanish 0.004±0.022 Farre et al. (1987) 

Cobalt Spanish 0.00005±0.00079 YbaÂnÄez et al. (1989) 

Others 0.00005±0.00038 YbaÂnÄez et al. (1989) 

Copper British 0.3±0.8 Paul and Southgate (1978) 

German 0.04±0.80 (0.19) Postel et al. (1972b) 

0.02±1.55 Postel et al. (1983) 

Spanish 0.0064±0.0603 (0.029) YbaÂnÄez et al. (1989) 

Others 0.010±0.040 (0.027) YbaÂnÄez et al. (1989) 

Lagers 0.01±0.41 (0.11) Binns et al. (1978) 

Iron British 0.1±0.5 Paul and Southgate (1978) 

German 0.02±0.84 (0.02) Postel et al. (1972a) 

0.04±1.55 Postel et al. (1983) 

Lagers 0.04±0.44 (0.12) Binns et al. (1978) 

Wheat beer (0.63) Postel et al. (1972a) 

Lead German 0.003±0.024 Postel et al. (1983) 

Spanish 0.0014±0.0056 (0.0028) YbaÂnÄez et al. (1989) 

Others 0.0008±0.0025 (0.0017) YbaÂnÄez et al. (1989) 

Lagers 0.06 Binns et al. (1978) 

Magnesium British 60±200 Paul and Southgate (1978) 

German 75±250 (114) Postel et al. (1974) 


Manganese German 0.04±0.51 (0.20) Postel et al. (1973) 

Mercury 0±0.0008 Donhauser et al. (1987) 

Nickel 0±0.26 Brenner et al. (1965) 

Phosphorous British 90±400 Paul and Southgate (1978) 

Australian 96±304 (196) Bottomley and Lincoln (1958) 

Potassium British 330±1100 Paul and Southgate (1978) 

German 396±562 (476) Kieninger (1978) 

Mexican 220±358 Canales et al. (1970) 


Selenium 0±0.0072 (0.0012) Donhauser et al. (1987) 

Silicon 10.2±22.4 Anderson et al. (1995) 

Sodium British 40±230 Paul and Southgate (1978) 



Tin German 0.010±0.020 Postel et al. (1983) 

Zinc German 0.01±1.48 (0.10) Postel et al. (1975) 

Lagers 0.01±0.46 Binns et al. (1978) 

German 0.01±0.26 Postel et al. (1983) 

ANIONS 

Chloride British 150±984 Paul and Southgate (1978) 




Constituent Source Concentration Reference 

range mg/l (ppm)* 

Fluoride British 0.08±0.71 Warnakulasuriya et al. (2002) 

German 0.08±0.64 (0.15) Postel et al. (1976) 

Nitrate German 1.4±101.3 (34.0) Postel (1976) 

German 13±43 Gmelch et al. (1989) 

Phosphate British 260±400 Paul and Southgate (1978) 


Sulphate British 150±400 Paul and Southgate (1978) 


African lager 125±260 Shah (1975) 

* Average values in parentheses. 

amines can form carcinogenic N-nitrosamines (see later). The EEC limit for nitrates in 

drinking water is 25 mg/litre. 

Trace amounts of many metals are essential for yeast growth whereas larger amounts 

may be toxic and may be limited by law. When specific limits for beer are not specified 

the limits for potable water are usually applied. In Britain, the levels of arsenic and lead 

are limited to 0.2 mg/kg (ppm) and the Institute of Brewing describes methods for 

ensuring these limits are met. International methods for determining iron, copper and zinc 

by atomic absorption spectroscopy are published as well as spectrophotometric methods 

for iron and copper. The UK Food Standards Committee recommends a limit for copper 

in beer of 7.0 mg/kg. Nickel can be estimated in beer spectrophotometrically by 

measuring the colour formed with dimethylglyoxime (Analytica-EBC). In the 1960s 

cobalt was added to beer to improve foam stability and prevent gushing. However, this 

was found to cause acute heart disease in heavy drinkers (in excess of 20 pints beer/day) 

so the practice was discontinued (Long, 1999). 

Carbon dioxide is a natural product of fermentation and beers contain 3.5±4.5 g/l. 

Supersaturated beers and naturally conditioned beers may contain as much as 6 g/l but gas 

will be evolved as soon as the pressure is released. The sensory threshold of carbon 

dioxide is about 1 g/l so the amount present will influence the flavour of beer. Analytica- 

EBC describes a titrimetric and an instrumental method for measuring the amount of CO2 

present. The Institute of Brewing describes a manometric method and the ASBC gives 

two methods, one for beer in tanks and the other for beer in bottles or cans. 

19.1.2 Alcohol and original extract 

In the European Economic Community (EEC) the strength of beer and other alcoholic 

drinks, is expressed as alcohol by volume (ABV), that is the ratio of the volume of the 

ethyl alcohol contained in the liquor to the volume of the liquor including the ethyl 

alcohol (expressed as a percentage to one decimal place). Further the EEC directed in 

1987 that the ABV àlc x.y% vol' shall appear on the label. In many countries excise duty 

is levied on the basis of the ABV. In Britain duty is payable on beers with more than 

1.2% ABV. Before 1993 duty in Britain was calculated from the original gravity of the 

wort fermented. Beers can contain between 0.05% ABV, in an alcohol-free beer, up to 

about 12.5%. In Germany bock beers must contain more than 6.0% ABV and doublebock 

beers more than 7.5% ABV. According to Glaser (2002), the strongest beer is 

Utopias MM II with 24% ABV. Analytica-EBC, the Institute of Brewing and the ASBC 


provide several methods for measuring alcohol in beers but national Excise authorities 

may specify their own modifications. 

Thedistillationmethodisusuallyregardedasthereferencemethodtobeusedinanycases 

ofdispute.Hereanaccuratelymeasuredquantityoffilteredbeer(say100.0mlinavolumetric 

flask at 20ëC or 100.00g) is washed into asuitable distillation flask (400±500ml capacity) 

and distilled, taking care not to char the residue in the distillation flask, until c. 85ml is 

collected (inthesame volumetricflask). Thisisthenmadeupto100.00ml(or100.00g)and 

thespecificgravityinairat20ëC/20ëCismeasuredinasuitablepycnometer(Reischauer)or 

specificgravitybottletofiveplacesofdecimals.APaardensitometerorasimilarinstrument, 

whichmakethismeasurementelectronicallyisnowcommonlyused(seeIoB,ASBC,Mundy, 

1996). Several tables exist to convert the specific gravity of the distillate to give the alcohol 

content.TheEECrecommendthatthetablesoftheOrganisationInternationaledeMetrologie 

Legale (OIML) should be used but they refer to specific gravity in vacuo. The EBC has 

therefore provided an alcohol table based on the specific gravity measured in air and give 

polynomials for alcohol and extract (Rosendahl and Schmidt, 1987): 

A% m=m ˆ517:4…1 SGA†‡5084…1 SGA† 2 ‡33503…1 SGA† 3 

where Ais %alcohol by weight and SGA is the specific gravity of the distillate in air at 

20ëC/20ëC. 

To convert A% m/m to A% v/v (ABV): 

Alcohol %v=v ˆ 

A%m=m SGA 

0:791 

where 0.791 is the specific gravity of ethanol at 20ëC/20ëC. 

The American Society of Brewing Chemists and the British Excise Authorities 

provide their own tables (for part of the latter see Table 19.4). 

Alcohol may also be determined by catalytic combustion using a Servochem 

Automatic Beer Analyser (SCABA). The injected beer is divided into two streams, one 

enters a Paar U-tube densitometer, the other passes down a column as a falling film where 

the alcohol is removed as a vapour with a countercurrent of air and passed over an 

alcohol sensor. After calibration with known standards, the onboard computer will 

display the % alcohol either m/m or v/v when the results are found to agree with the 

distillation method (Freeston and Baker, 1993). Schropp et al. (2002) evaluated an NIR 

procedure for measuring the alcohol content as in the Alcolyzer instrument. 

The amount of alcohol present can also be determined from the refractive index of the 

media. This is much quicker than the distillation method but requires standardizing to 

determine the constants A, B, and C in the regression equation: 

Alcohol % …v=v† ˆ A…SR WR† B…PG† C 

where SR is the refractive index of the beer, WR is the refractive index of water, and PG 

is the present gravity of the beer {1000 (SG 1.00000)}. The instrument is thermostatted 

and measurements are usually made at 20 ëC. Ethanol can also be determined by gas 

chromatography using a flame ionization detector and direct injection on to a suitable 

column (Poropak Q, 15% Carbowax 20 M, SGE BP20 or CP Wax 52 CB), after a known 

addition of n-butanol as an internal standard. The amount of alcohol is calculated by 

comparison of the peak areas (see also Clarkson et al., 1995). 

For alcohol-free or low-alcohol beers (< 0.008%) an enzymatic method is given based 

on the Boehringer test kit. The alcohol is oxidized first to ethanal and then to ethanoic 


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) 

0 1 2 3 4 5 6 7 8 9 

788.16 100.0 ± ± ± ± ± ± ± ± ± 

780 ± ± ± ± ± ± ± ± ± 99.84 

790 99.64 99.44 99.24 99.04 98.83 98.63 98.40 98.18 97.96 97.74 

800 97.51 97.28 97.05 96.81 96.57 96.33 96.09 95.84 95.59 95.33 

810 95.08 94.82 94.56 94.30 94.03 93.76 93.49 93.22 92.94 92.63 

820 92.38 92.09 91.80 91.52 91.22 90.93 90.63 90.33 90.03 89.73 

830 89.42 89.12 88.81 88.49 88.18 87.86 87.55 87.23 86.90 86.58 

840 86.25 85.53 85.60 85.26 84.93 84.59 84.24 83.92 83.58 83.23 

850 82.89 82.54 82.20 81.85 81.50 81.14 80.79 80.43 80.08 79.72 

860 79.35 78.99 78.63 78.26 77.89 77.53 77.16 76.78 76.40 76.03 

870 75.66 75.28 74.90 74.51 74.13 73.75 73.36 72.97 72.58 72.19 

880 71.79 71.40 71.00 70.60 70.20 69.80 69.39 68.99 68.58 68.17 

890 67.76 67.34 66.93 66.51 66.09 65.67 65.25 64.83 64.40 63.97 

900 63.54 63.11 62.68 62.24 61.80 61.36 60.92 60.47 60.03 59.58 

910 59.13 58.67 58.22 57.76 57.30 56.84 56.37 55.90 55.43 54.96 

920 54.48 54.00 53.52 53.03 52.54 52.05 51.55 51.05 50.55 50.04 

930 49.53 49.01 48.49 47.97 47.44 46.91 46.36 45.82 45.26 44.71 

940 44.15 43.58 43.00 42.41 41.82 41.23 40.61 39.99 39.37 38.73 

950 38.08 37.42 36.75 36.07 35.37 34.67 33.95 33.21 32.47 31.69 

960 30.92 30.12 29.32 28.48 27.55 26.80 25.92 25.04 24.14 23.24 

970 22.32 21.40 20.47 19.54 18.60 17.68 16.74 15.84 14.93 14.03 

980 13.14 12.27 11.40 10.56 9.72 8.90 8.09 7.30 6.52 5.76 

990 5.01 4.27 3.54 2.83 2.13 1.44 0.77 0.10 ± ± 

997.15 0.00 


acid with nicotinamide adenine dinucleotide (NAD ‡ ) and the reduction of the cofactor is 

measured spectrophotometrically at 340 nm: 

C2H5OH ‡ NAD ‡ ˆ CH3CHO ‡ NADH ‡ H ‡ 

CH3CHO ‡ NAD ‡ ‡ H2O ˆ CH3CO2H ‡ NADH ‡ H ‡ 

As mentioned above, duty in the United Kingdom used to be levied on the gravity of 

the wort fermented. Accordingly the Institute of Brewing provide a method to determine 

the original gravity and Analytica-EBC and the ASBC give a method to find the original 

extract. Both are based on the distillation method for alcohol when, as well as the 

distillate, the residue in the distillation flask is diluted to the original volume and the 

specific gravity is measured. In the British method, the number of degrees of gravity by 

which the distillate is less than the specific gravity of distilled water is called the spirit 

indication of the distillate. From the `Mean Brewery Table' (Thorpe and Brown, 1914; 

HM Customs and Excise, 1997; Statutory Instrument No. 1146, 1979) the degrees of 

extract that must have been fermented to produce the spirit indication is read off and 

added to the gravity of the residue to give the original gravity. Such a Table is necessary 

because in any fermentation carbohydrate is used for yeast growth and the production of 

metabolites other than ethanol, so the yield of ethanol is always less than that predicted 

by Gay-Lussac's equation: 

C6H12O6 ! 2C2H5OH ‡ 2 CO2 

Any radical change in the ratio of yeast growth to alcohol production from that used in 

compiling the `Mean Brewery Table' could result in beers in which the original gravity, 

determined by the above method, is higher than that actually employed. Thus, the use of 

larger fermentation vessels and methods of continuous fermentation results in beers in 

which the measured OG is higher than the gravity of the wort employed. Similarly, the 

production of large amounts of secondary metabolites can alter the results. Belgian 

lambic beer contains a high level of volatile acids and this is taken into account in the 

Belgian method for determining the original extract. 

In the EBC and ASBC methods for determining the original extracts of beers, the 

apparent extract (E % w/w) is determined from the specific gravity of the filtered beer, 

the alcohol content (A % w/w) from the specific gravity of the distillate, and the real 

extract (ER% w/w) is determined from the specific gravity of the distillation residue 

made up to the original volume. From Balling's equation when the extract in the original 

wort (% w/w) (ˆ ëPlato) ˆ p 

or 

p ˆ 100 

2:0665 A ‡ ER 

100 ‡ 1:0665 A 

p ˆ ER ‡ E 

‡ ER 

q 

where q is the correcting factor. The ASBC also define: 

Real degree of fermentation ˆ 

Apparent degree of fermentation ˆ 

100… p ER† 

p 

100… p E† 

p 


For small breweries without laboratory facilities HM Customs and Excise (1997) 

provide an approximate formula to determine alcohol by volume: 

%ABV ˆ…OG PG† f 

whereOGistheoriginalgravityofthewort,PGisthepresentgravityofthebeer,andfis 

afactor connecting change in gravity with alcoholic strength. Unfortunately fis not 

constant and varies from 0.125 for weak beers to 0.135 for very strong beers. For the 

majority of popular UK beers fis 0.128±0.129. Brewers using this method should obtain 

confirmatory testing by an independent analyst at least annually. 

19.1.3 Carbohydrates 

The total carbohydrates remaining in beer can be estimated spectrophotometrically with 

anthrone in 85% sulphuric acid (Analytica-EBC, ASBC, IoB) for arange of beers values 

between 0.89±5.98% as glucose were found. Fully attenuated low carbohydrate `lite' 

beers, originally brewed for diabetic patients, are now generally available with 

carbohydrate contents of 0.4±0.9% w/v as glucose. Of the carbohydrates present in 

wort, glucose (4.1), fructose (4.2), sucrose (4.3), maltose (4.4) and maltotriose (4.5) will 

usually be fermented. Under-attenuating yeast strains will not ferment maltotriose while 

super-attenuating strains will partly ferment maltotetraose (4.6) but, in general, beers will 

contain only low levels of fermentable sugars other than those added as primings (Table 

19.5). Nevertheless trace amounts of many other sugars have been detected in beer 

including the monosaccharides; D-ribose (4.12), L-arabinose (4.11), D-xylose (4.10), Dmannose 

(4.8) and D-galactose (4.9), the disaccharides isomaltose (4.13), cellobiose 

(4.18) and kojibiose, and the trisaccharides panose (4.14) and isopanose (4.15). Data for 

these and other oligosaccharides are given in Table 19.6. 

Analytica-EBC and ASBC give methods for the fermentable carbohydrates in beer by 

HPLC. A detailed study of the dextrins in Tuborg lager beer was made using gel 

chromatography (Enevoldsen and Schmidt, 1974). The carbohydrates present were 

resolved into the following fractions (degree of polymerization-glucose units and 

percentage of the total carbohydrates): DP 1±3, 7.4%; DP 4, 13.1%; Group I(DP 5±10), 

22.7%; GroupII(DP11±16),16.7%;GroupIII(DP17±21),9.7%;GroupIV(DP22±27), 

6.2%; Group V (DP 28±34), 4.0%; and higher dextrins (DP>35), 15.2%. Debranching 

studies with pullulanase indicated that the dextrins in Group Iwere either linear or single 

branched, while those in Groups II, III and IV contain two, three and four -(1±6)linkages 

respectively. The majority of the -(1±6)-linkages in amylopectin appear to 

survive the brewing process. 

If -(1±3)(1±4)-D-glucan, which makes up 70% of the barley endosperm cell wall 

(Section 4.4.3), is not completely broken down during malting, it may survive into beer 

where it can precipitate and lead to filtration problems. Only -glucans with molecular 

weights in excess of 200,000 daltons are said to precipitate. Two methods of analysis for 

-glucans are given in Analytica-EBC, one enzymatic using lichenase and the other using 

the fluorochrome Calcofluor. A range of beers contained 0±650 mg/l of -glucan; 10/36 

had no -glucan but a beer which precipitated a -glucan gel contained 1,900 mg/l. By 

colorimetric methods beers were found to contain 0.13±0.21 % of fructose and fructosans 

and 0.25±0.39 % of pentosans. Schwarz and Han (1995) found the arabinoxylan content 

of a number of beers was in the range of 514±4,211 mg/l. 


Table 19.5 Sugar content of commercial beers shown as percentage (w/v) in sample (Otter and Taylor, 1967) 

Type of beer OG SG Fructose Glucose Sucrose Maltose Maltotriose Maltotetraose Total 

(hydrate) 

1 Pale ale 1050 1011 Nil 0.06 Nil 0.54 0.28 0.04 0.92 

2 Brown ale (primed) 1032 1012 1.0 1.0 Trace Trace 0.2 0.4 2.6 

3 Stout (primed) 1033 1013 0.53 0.61 Trace Trace 0.08 0.06 1.28 

4 Sweet stout (primed) 1045 1022 0.6 1.2 Trace Trace 0.6 0.3 3.6* 

5 Pale ale 1068 1019 0.01 0.01 Nil 0.7 1.7 0.4 2.9 

6 Strong ale 1085 1026 Trace Trace Nil 0.16 0.21 0.12 0.49 

7 Lager 1032 1007 Nil Nil Trace Trace Trace Trace Trace 

8 Lager (export) 1045 1008 Nil Trace Nil Trace 0.28 0.18 0.46 

9 Stout (conditioned) 1044 1008 Nil Nil Nil Nil Trace Nil Trace 

10 Ale 1038 1002 Trace 0.8 Nil Nil Trace Trace 0.8 

11 Lager 1040 1003 0.18 0.49 Nil Nil Nil Nil 0.67 

12 Lager 1046 1003 Nil Trace Nil Trace Trace Trace Trace 

13 Lager 1045 1004 Nil 0.27 Nil 0.17 0.24 0.09 0.77 

14 Lager 1052 1011 Trace 0.15 Nil 0.13 0.16 0.14 0.58 

15 Lager 1045 1008 Nil Nil Nil 0.25 0.33 0.20 0.78 

* Contained also lactose (hydrate) 0.9% 


Table 19.6 Oligosaccharides in beer (g/100ml as glucose) 

Lager Lager Diabetic 

(Danish)* (German)y lagery 

Pentose 

Fructose 

Glucose 

± 

} 0.02 

0.019 

0.015 

± 

0.052 

± 

0.008 

Isomaltose 0.08 0.102 0.098 

Maltose 0.07 0.188 0.143 

Panose ± 0.036 0.066 

Maltotriose 0.17 0.315 0.193 

4- -Isomaltosyl-D-maltose ± 0.049 0.100 

Maltotetraose 0.30 0.187 0.043 

Maltopentaose 0.08 0.144 0.100 

Maltohexaose 0.15 0.130 0.039 

Maltoheptaose 0.15 0.063 0.035 

Malto-octaose 0.17 

1.560 0.065 

Maltonoaose 0.15 

Higher dextrins 1.06 ± ± 

Total 2.40 2.830 0.092 

* Gjertsen (1955). 

y Silbereisen and Bielig (1961). 

} } 

19.1.4 Other constituents containing carbon, hydrogen and oxygen 

Non-volatile 

Manyofthesecomponentsareproductsofyeastmetabolism.Mostoftheintermediatesin 

the metabolic pathways discussed in Chapter 12 have been detected in beer. 

Quantitatively glycerol is important and arange of 436±3,971mg/l has been found; the 

highest level in aspecial beer of OE 27.3 g/100ml. In general, top fermented beers had 

higher glycerol levels than Pilsen-type beers (Klopper et al., 1986). Significant amounts 

of higher polyols have not been found but beer contains butane-2,3-diol (up to 280 mg/l) 

and smaller amounts of pentane-2,3-diol together with 3-hydroxybutan-2-one (9.16, 

acetoin, 3±26mg/l) and 3-hydroxypentan-2-one. These are reduction products of the 

volatile vicinal diketones (see later). Peppard and Halsey (1982) found 2,4,5-trimethyl- 

1,3-dioxolane (19.1) in beer (c. 0.1 mg/l); this is the cyclic acetal formed between 

butane-2,3-diol and ethanal (acetaldehyde). Similar 1,3-dioxolanes formed between 

butane-2,3-diol and isobutanal and isopentanal were also detected. Another non-volatile 

alcohol found in beer is tyrosol (19.2), the Ehrlich pathway degradation product of 

tyrosine. Canadian lager had 22.1±29.4mg/l while ales had 3.0±13.6mg/l, well below the 

taste threshold of 200ppm (McFarlane and Thompson, 1964). 

Non-volatile acids found in beer are listed in Table 19.7. In addition, Klopper et al. 

(1986) found pyruvic acid (1±127mg/l), malic acid (6±136mg/l), lactic acid (10± 

1362mg/l) and citric acid (6±211mg/l) in arange of beers; the highest levels of lactic 

acid(andaceticacid)werefoundinBelgianàcid'beers.Thelevelofoxalicacidinbeers 


isimportant becausetheinsolublecalciumsaltmaycauseahazeand/orpromotegushing. 

Beers contain trace amount of lipids. ASwedish beer (12ëPlato) was found to contain 

(mg/l): triglycerides, 0.1±0.2; diglycerides, 0.1; monoglycerides, 0.1±0.3; sterol esters, 

0.01; free sterols, 0.01±0.02 and free fatty acids C 4±C 10, 10±15; and C 12±C 18, 0±0.5 

(AÈyraÈpaÈaÈet al., 1961). Similar data were found for other beers. The free fatty acids are 

volatile but the addition of another substituent usually results in loss of volatility. 

Autoxidation of linoleic acid gives rise to isomers of dihydroxy- and trihydroxyoctadecenoic 

acids. The concentration of these acids (Table 19.7) is greater than that of 

linoleic acid itself. These hydroxy acids are potential precursors of 2-trans-nonenal, 

which contributes acardboard flavour to stale beer. The influence of lipids on foam and 

head retention is discussed later. 

Also included in Table 19.7 are the phenolic acids present in beer which, with other 

polyphenols, are extracted from malt and hops (Sections 4.8 and 8.4). Polyphenols give 

colours with ferric salts and Analytica-EBC and the ASBC give a non-specific 

specrophotometric method for polyphenols based on the colour formed at 600nm with 

ferric ammonium citrate. (‡)-Catechin (4.138) can be used as standard. In addition 

Analytica-EBC give aspectrophotometric method for flavanoids based on the colour 

formed with p-dimethylaminocinnamaldehyde: again (‡)-catechin may be used as 

standard. By HPLC the polyphenols quercetin (8.58, RˆH)(36±148ppm), rutin (8.58, 

R ˆ -L-rhamnosyl-6- -D-glucosyl) (1.5±7.7ppm), catechin (26Ð141ppm) and 

epicatechin (4.139) (8.5±127ppm) have been quantified in beer (Qureshi et al., 1979). 

Beers also contain proanthocyanidins (anthocyanogens) of which the dimeric 

procyanidin B-3 (4.143) predominates (0.5±4.0ppm). Whittle et al. (1999) examined 

the polyphenols in sixty beers by HPLC with electrochemical detection. They found over 

twenty procyanidin dimers and trimers in beer but no tetramers or pentamers although 

thesewerepresentinbarleyextracts.Theyalsoobservedwhichpeakswereremovedwhen 

the beer was treated with excess polyvinylpyrrolidone (20g/l). As the name implies, 

proanthocyanidinsontreatmentwithacidformthecolouredanthocyanidinpigments( max 

c. 545 nm) but the reaction is not straightforward and the yield of pigment from different 

products varies. Polyphenols, particularly proanthocyanidins, react with proteins during 

the storage of beer to produce non-biological haze (see later) but the yields of 

anthocyanidins, liberated with acid, do not correlate with the shelf-life of the beer. The 

haze potential of beers has been estimated nephelometrically by the haze formed after 

treatment with either cinchonidine sulphate, polyvinylpyrrolidone 700 or tannic acid. 

CloselyrelatedtothepolyphenolsarethehopresinsdiscussedinChapter8.Themajor 

bittering principles in beer are the cis- and trans-isomers of isocohumulone, isohumulone 

and isoadhumulone (8.40). The individual isomers may be resolved by HPLC but for 

routine analysis they are usually estimated together from the light absorption of an isooctane 

extract of acidified beer at 275nm (Analytica-EBC, IoB and ASBC). To avoid 

making assumptions about the nature of the bittering principles, the absorbance is 

multiplied by 50 and the result given as (International) Bitterness Units (IBU or BU). 

Nevertheless, in beers brewed with fresh hops IBU approximate to mg iso- -acids/l. 

Beers maycontain 10±60IBU(Tables19.1and19.2) andexceptionally 100IBU(Glaser, 

2002). The sensory detection threshold of the iso- -acids is c. 5±6 mg/l and the 

tetrahydro- and hexahydroiso- -acids are even more bitter (Weiss et al., 2002). 

Volatile 

Although trace amounts of the volatile constituents of malt and hops may survive wort 

boiling, the majority of the volatile constituents of beer are fermentation products. After 


Table 19.7 Non-volatile acids in beer 

Acid (ppm) Berlin Pilsner (a) German (b) British (c) American (d) Flavour threshold (e) 

C2 Glycolic ± ± ± ± 

Oxalic (4.151) ± 9.9±22.8 ± 0.5±3.0 

C3 D-Lactic ± 20±200 ± ± 

L-Lactic ± 40±152 ± ± 

Lactic (4.150) 188 ± 44±292 ± (400) 

Pyruvic (4.146) ± 42±75 10±104 ± (300) 

Malonic 0.02 ± ± ± 

C4 Succinic (4.149) 48 ± 36-166 ± 

Fumaric (4.148) ± ± ± ± 

Malic (4.152) ± 55±105 14±97 ± 

Oxaloacetic ± ± ± ± (500) 

Tartaric ± ± ± ± (600) 

C 5 2-Hydroxy-3-methylbutyric 0.26 ± ± ± 

Levulinic ± ± ± ± 

2-Methylfumaric (Mesaconic) ± ± ± ± 

Glutaric 0.01 ± ± ± 

2-Hydroxyglutaric ± ± 0±17 ± 

Citramalic ± ± ± (5.9±15.2) 

2-Oxoglutaric (4.147) ± ± 0±20 

C6 2-Hydroxy-3-methylpentanoic 0.29 ± ± ± 

2-Hydroxy-4-methylpentanoic 0.33 ± ± ± 

Adipic ± ± ± ± 

Kojic ± ± ± 5±78.2 

Citric (4.153) ± 130±230 56±158 ± 

Isocitric ± ± ± ± 

Oxalosuccinic ± ± ± 

C7 Benzoic 0.45 ± ± ± 

2-Hydroxybenzoic 0.02 ± ± 1.00±9.0 

4-Hydroxybenzoic 0.13 ± ± ± 

2-Hydroxyheptanoic 0.06 ± ± ± 

3,4-Dihydroxybenzoic 2.4 ± ± 6.3±29.0 

2,5-Dihydroxybenzoic (Gentisic) ± ± ± 2.8±12.7 


Pimelic 0.01 ± ± ± 

Gallic (4.125) ± ± ± 12.0±30 360 

C 8 Phenylacetic 0.93 ± ± ± 2.5 

2-Hydroxyoctanoic 0.04 ± ± ± 

3-Hydroxyoctanoic 0.07 ± ± ± 

2-Ethylhexanoic ± ± ± 10.2 11±32 

4-Hydroxyphenylacetic 0.04 ± ± ± 

Suberic 0.15 ± ± ± 

Vanillic 2.4 ± ± 0.3±1.5 80 

Phthalic 0.02 ± ± ± 

C9 Phenylpropionic 0.01 ± ± ± 

trans-Cinnamic 0.5 ± ± 1.0±8.3 

cis-Cinnamic < 0.01 ± ± (trans + cis) 

Phenyl-lactic 1.2 ± ± 

4-Hydroxyphenylpropionic 0.02 ± ± ± 

trans-p-Coumaric 1.9 ± ± 8.21 520 

cis-p-Coumaric 0.02 ± ± (trans + cis) 

Caffeic ± ± ± 1.4±8.0 690 

Azeleic 1.5 ± ± ± 

C10 3-Hydroxydecanoic 0.16 ± ± ± 

trans-Ferulic 4.6 ± ± 1.7±20.8 660 

cis-Ferulic 1.1 ± ± (trans + cis) 

C11 Undecanedioic 0.13 ± ± ± 

Sinapic ± ± ± 0.7±3.6 

C12 Dodecanedioic 0.18 ± ± 

C16 Chlorogenic ± ± ± 1±11.2 

C18 Trihydroxyoctadecanoic 

9,12,13-10-trans ± 4.9±9.0 ± 

9,12,13-11-trans ± 1.0±2.4 ± 

9,10,11-12-trans ± 0.4±0.7 ± 

[a] Tressl et al. (1975) 

[b] Mandl and Piendl (1971) 

[c] Coote and Kirsop (1974) 

[d] Qureshi et al. (1979) 

[e] Meilgaard (1975) 


Table 19.8 Volatile constituents of beer 

(a) Alcohols [a] 

Concentration Flavour threshold 

(ppm) [b] (ppm) [c] 

C 1 Methanol ± 10,000 

C 2 Ethanol ± 14,000 

C 3 Propanol ± 800 

Propan-2-ol ± 1,500 

C4 Butanol ± 450 

Butan-2-ol ± 16 

2-Methylpropanol 4.8 200 

C5 Pentanol 0.15 (80) 

Pentan-2-ol ± 45 

2-Methylbutanol 84.0 70 

3-Methylbutanol (2 Me +3Me) 65 

Furufuryl alcohol 1.20 3,000 

C6 Hexanol 0.33 4.0 

Hexan-2-ol ± 4.0 

Hex-2-enol 0.025 13 

Hex-3-enol 0.020 15 

C7 Heptanol ± 1.0 

Heptan-2-ol 0.015 0.25 

Benzyl alcohol ± 900 

C 8 Octanol ± 0.9 

Octan-2-ol 0.005 0.04 

Oct-l-en-3-ol 0.030 0.2 

2-Phenylethanol 1.8 125 

Tyrosol see text 200 

4-Vinylphenol 0.025 ± 

C9 Nonanol ± 0.08 

Nonan-2-ol 0.01 0.075 

4-Vinylguaiacol 0.10 0.3 

C 10 Decanol ± 0.18 

Decan-2-ol 0.005 0.015 

Linalol ± 0.08 

-Terpinol ± 2.0 

Nerol ± 0.50 

C 12 Dodecanol ± 0.40 

Dodecan-2-ol ± ± 

(b) Aldehydes [a] 


(ppm) [d] (ppm) [c] 

C2 Acetaldehyde See Table 19.10 10 

Glyoxal ± ± 

C3 Propanal ± 30 

Prop-2-enal (acrolein) 1.6 15 

Pyruvaldehyde ± - 

C 4 Butanal 10.9 (1.0) 

2-Methylpropanal ± (1.0) 

But-2-enal (crotonal) 1.33 8 




(ppm) [d] (ppm) [c] 

C5 Pentanal 2.7 0.5 

2-Methylbutanal ± 1.25 

3-Methylbutanal 7.0 0.6 

Pent-2-enal 0.59 ± 

Furfural (25.0) 150 

C6 Hexanal 1.6 0.3 

Hex-2-enal 0.36 0.5 

5-Hydroxymethylfurfural see text 1,000 

C7 Heptanal 1.2 0.05 

Hept-2-enal 0.08 0.0005 

Benzaldehyde 10 2 

C8 Octanal 1.4 0.04 

Oct-2-enal 0.03 0.0003 

2-Phenylacetaldehyde (5) 1.6 

C9 Nonanal 3.7 0.015 

Non-2-enal 0.07 0.0003 

C 10 Decanal 0.9 0.005 

Dec-2-enal trace 0.001 

C11 Undecanal 0.4 0.002 

C12 Dodecanal 0.2 0.002 

(c) Acids [a] 


(ppm) [e] (ppm) [c] 

C1 Formic ± ± 

C2 Acetic 175 

C3 Propionic 150 

C4 Butyric 0.62 2.2 

2-Methylpropionic 1.1 30 

Crotonic ± ± 

C5 Pentanoic (Valeric) 0.03 8 

2-Methylbutyric ± ± 

3-Methylbutyric 1.3 1.5 

Pentenoic ± ± 

C6 Hexanoic (Caproic) 

Hex-2-enoic 

2.5 

} 

8 

0.01 

Hex-3-enoic 

± 

1.3 

4-Methylpent-3-enoic 0.32 ± 

3-Carbethoxypropionic 0.22 ± 

C7 Heptanoic (Oenanthic) 0.03 ± 

Hept-2-enoic < 0.01 ± 

4-Methylhex-2-enoic ± ± 

Benzoic 0.45 ± 

C8 Octanoic (Caprylic) 6.1 13/15 

6-Methylheptanoic ± ± 

Oct-2-enoic < 0.01 ± 

Phenylacetic 0.93 2.5 




(ppm) [e] (ppm) [c] 

C9 Nonanoic (Pelargonic) 0.02 

Non-2-enoc




C8 Methyl heptenoate ± ± 

Ethyl hexanoate 0.95 0.23 

3-Methylbutyl propionate 0.15 0.7 

Hexyl acetate 0.025 1.4 

Methyl 4-methylhex-2-enoate 0.06 ± 

Ethyl hexenoate ± ± 

Ethyl nicotinate 1.4 ± 

C9 Methyl octanoate ± ± 

Ethyl heptanoate ± ± 

Amyl butyrate ± 0.6 

3-Methylbutyl butyrate ± ± 

3-Methylbutyl 2-methylpropionate 0.140 ± 

Heptyl acetate 0.025 1.4 

EthyI heptenoate ± ± 

Ethyl benzoate 0.01 ± 

C10 Ethyl octanoate 1.50 0.9 

Butyl hexanoate ± ± 

2-Methylpropyl hexanoate 0.025 ± 

Amyl 3-methylbutyrate ± ± 

3-Methylbutyl 3-methlbutyrate 0.040 ± 

Hexyl butyrate 

Octyl acetate 0.030 0.5 

2-Phenylethyl acetate 1.625 3.8 

C11 Ethyl nonanoate ± 1.2 

Amyl hexanoate ± ± 

Pent-2-yl hexanoate ± ± 

3-Methylbutyl hexanoate 0.420 0.9 

Heptan-2-yl butyrate ± ± 

Nonyl acetate 0.005 ± 

Ethyl nonenoate 0.045 ± 

2-Phenylethyl propionate 0.010 ± 

Ethyl cinnamate 0.005 ± 

C12 Ethyl decanoate 0.190 1.5 

Butyl octanoate ± ± 

3-Methylbutyl heptanoate 0.020 ± 

Hexyl hexanoate ± ± 

Octyl butyrate ± ± 

2-Phenylethyl butyrate ± ± 

3-Phenylethyl 2-methypropionate 0.015 ± 

3-Methylbutyl benzoate ± ± 

Ethyl dec-4-enoate 0.035 ± 

Ethyl deca-4, 8-dienoate 0.015 ± 

C13 2-Methylbutyl octanoate 0.015 ± 

3-Methylbutyl octanoate 0.830 2.0 

2-Phenylethyl 3-methylbutyrate 0.015 ± 

C 14 Ethyl dodecanoate 0.030 3.5 

3-Methylbutyl nonanoate ± 2.0 

Hexyl octanoate ± ± 

Octyl hexanoate ± ± 

2-Phenylethyl hexanoate 0.075 ± 

2-Phenylethyl hexenoate ± ± 





C15 2-Methylbutyl decanoate 0.005 ± 

3-Methylbutyl decanoate 0.095 3.0 

3-Methylbutyl dec-4-enoate 0.020 ± 

C 16 Ethyl tetradecanoate ± 2.5 

Hexyl decanoate ± ± 

2-Phenylethyl octanoate 0.015 ± 

C 17 3-Methylbutyl dodecanoate 0.010 ± 

(e) Lactones [a] 



C 4 4-Butanolide 1.300 ± 

C6 4-Hexanolide 0.0200 

4, 4-Dimethylbutan-4-olide 0.160 ± 

4, 4-Dimethylbut-2-en-4-olide 1.750 ± 

C7 4-Heptanolide 0.015 ± 

C8 4-Octanolide 0.020 ± 

C 9 4-Nonanolide 0.320 ± 

C10 4-Decanolide 0.020 0.4 

C 11 4-Dihydroactindiolide 0.030 ± 

(f) Ketones [a] 



C 4 Butan-2-one 0.018 (80) 

Butane-2, 3-dione 0.058 0.15 

3-Hydroxybutan-2-one 0.420 (50) 

C5 Pentan-2-one 0.020 (30) 

Pentan-3-one ± (30) 

Pentane-2, 3-dione 0.012 0.9 

3-Hydroxypentan-2-one 0.050 ± 

2-Methyltetrahydrofuran-3-one 0.025 ± 

2-Methyltetrahydro-thiophen-3-one 0.005 ± 

C6 Hexan-2-one ± 4.0 

3-Methylpentan-2-one 0.060 0.4 

4-Methylpentan-2-one 0.12 ± 

2-Acetylfuran 0.040 ± 

C 7 Heptan-2-one 0.110 2.0 

C 8 Octan-2-one 0.010 (0.25) 

6-Methylhept-5-en-2-one 0.050 

C 9 Nonan-2-one 0.030 (0.2) 

C10 Decan-2-one ± 0.25 

C11 Undecan-2-one 0.001 0.4 



(g) Hydrocarbons [a] 



C6 Cyclohexane ± ± 

C7 3, 4-Dimethylpent-2-ene ± ± 

Toluene ± ± 

C8 2, 2-Dimethylhexane ± ± 

o-Xylene 0.020 ± 

Styrene 0.070 ± 

C 9 Methylethylbenzene 0.020 ± 

C 10 i-Butylbenzene ± ± 

-Pinene ± ± 

Limonene ± ± 

Myrcene ± 0.013 

Naphthalene 0.015 ± 

C15 Carophyllene ± ± 

[a] Drawert and Tressl (1972) 

[b] Tressl et al. (1978) 

[c] Meilgaard (1975) 

[d] Greenhoff and Wheeler (1981) 

[e] Tressl et al. (1975) 

ethanol, discussed above, the largest group of volatile constituents are the higher 

alcohols; those identified in beer are listed in Table 19.8(a) By distillation the higher 

alcohol fraction may be separated when it is known as fusel oil. Thus, the distiller has a 

closercontroloverthehigheralcoholcontentofhisbeveragethanthebrewer.Gin,vodka 

and grain whisky have low levels of higher alcohols while malt whisky and brandy, 

produced in pot stills, usually have higher levels of these congeners. 

The principal higher alcohols found in beer are 3-methylbutanol (isoamyl alcohol), 2methylbutanol 

(active-amyl alcohol), 2-methylpropanol (isobutyl alcohol), propanol, 

(propyl alcohol) and -phenylethanol (phenethyl alcohol) (Table 19.9). Greenshields 

(1974) found that the level of higher alcohols in home-brewed beers and wines was ten 

times higher than the level in commercial products. The major volatile constituents of 

beer are most conveniently examined by gas chromatography. The results given in Table 

19.9 (Morgan, 1965) were obtained by GC after distillation and extraction of the volatile 

productsintoether.DirectinjectionofbeerontoasuitableGCcolumnminimizessample 

preparation but requires frequent replacement of the top layer of apacked column or of 

the glass wool in the injection port due to the deposition of beer solids. Although the 

precision is not high (Baker, 1989) aheadspace method of analysis of the major volatiles 

in beer has been approved giving values for acetaldehyde, propanol, isobutanol, 

methylbutanols, ethyl acetate and ethyl hexanoate (Fig. 19.1). 

In order to identify the minor volatile constituents of beer it is usually necessary to 

examine adistillate or solvent extract which can be fractionated further by adsorption 

chromatography (Tressl et al., 1975, 1978, Lermusieau et al., 2001). 

Also included in Table 19.8(a) are 4-vinylphenol and 4-vinylguaicol (4.134, 4hydroxy-3-methoxystyrene), 

which are regarded as off-flavours in most beers. However, 


Table 19.9 Principal volatile constituents of beer (Morgan, 1965) 

Concentration Range 

B.p. ( o C) Stout Pale ale Brown ale Lager 

Ethanol (% v/v) 78 2.0±8.9 3.4±4.0 2.1±4.5 2.8±3.2 

(mg/l) 

n-Propanol 97 13±60 31±48 17±29 5-10 

2-Methylpropanol 108 11±98 18±33 11±33 6±11 

2-Methylbutanol 128 9±41 14±19 8±22 8±16 

3-Methylbutanol 131 33±169 47±61 28±77 32±57 

-Phenylethanol 220 20±55 36±53 19±44 25±32 

Ethyl acetate 77 11±69 14±23 9±18 8±14 

Isoamyl acetate 139 1.0±4.9 1.4±3.3 0.4±2.6 1.5±2.0 

4-vinylguaicol, which has aclove-like, spice-like flavour, provides part of the essential 

character of Weizenbier and Rauchbier. Arange of ale and lagers contained 0±0.09mg/l 

4-vinylguaicol, below the threshold value of 0.3mg/l, but Weissbier contains 0.8± 

1.5mg/l (McMurrough et al., 1996). These phenols are formed by decarboxylation of 

trans-p-coumaric acid (4.129) and trans-ferulic acid (4.131) respectively. This may 

occur thermally during kilning or wort boiling or enzymatically during fermentation. 

The capacity of yeasts to decarboxylate cinnamic acids (Pof‡ phenotype) is strong in 

wild strains of Saccharomyces but absent from lager-brewing yeasts and most alebrewing 

yeasts. However, it is present in the yeasts used in the production of wheat 

beers.Thelevelof4-vinylguaicolinbeerdeclinesduringstorage(withahalflifeofc.60 

days at 18ëC) presumably due to the formation of 4-(1-ethoxyethyl)guaiacol. 

Decarboxylation of cinnamic acid will give styrene (Table 19.8(g)), which is reported 

tobecarcinogenic.AccordinglyPof‡strainsofyeastareavoidedinbrewingmostbeers. 

Onlylowlevelsofaldehydesarefoundinbeer(Table19.8(b)).AsdiscussedinChapter 

12, ethanol and the higher alcohols are formed by reduction of the corresponding 

aldehydes by the enzyme alcohol dehydrogenase. Acetaldehyde (ethanal) is the major 

aldehyde in beer and some values are given in Table 19.10. Acetal (1,1-diethoxyethane), 

40 

1 

2 3 

4 

5 

7 

6 

8 

0 0 16 

10 

9 

11 

12 

13 

14 

GRAPH KEY 

40 mV % 

16.00 min. 

1. Acetaldehyde 

2. Dimethyl sulphide 

3. Acetone 

4. Ethyl acetate 

5. Methanol 

6. Isobutyl acetate 

7. Ethyl butyrate 

8. n-Propanol 

9. Isobutanol 

10. Isoamyl acetate 

11. Internal standard 

12. Isopentanols 

13. Ethyl hexanoate 

14. Ethyl octanoate 

Fig. 19.1 Typical chromatogram of volatile compounds in beer (Institute of Brewing, Methods 

of Analysis, 1997) 


Table 19.10 Acetaldehyde content of beers (Otter and Taylor, 1971) 

Beer Acetaldehyde content (ppm) 

(mean value in parentheses) 

British lager 2.3±28.2 (8.7) 

Foreign lager 0±13 (5.4) 

Light and pale ales 3.8±33.8 (8.2) 

Primed beers 3.0±37.2 (15.2) 

Irish stout 0.5±10.0 (4.0) 

American beers 2±18 (9.1) 

formed by condensation of ethanal with two molecules of ethanol, has been detected in 

beer. During the storage of bottled beer higher alcohols are oxidized to aldehydes by 

melanoidins; these aldehydes have much lower threshold values than the parent alcohols 

and can produce off-flavours. As mentioned above, the cardboard flavour of stale beer is 

though to be due to 2-trans-nonenal and 5-methylfurfural (9.22). Among the heterocyclic 

compounds formed during wort boiling (Chapter 9), 5-hydroxymethylfurfural (9.8, R= 

CH2OH), 5-methylfurfural and furfural (9.17) itself are found in beer. One survey of over 

300 German beers reported 0.5±4 ppm of 5-hydroxymethylfurfural (maximum value 7.8 

ppm) (Thalacker and Kaltwasser, 1978) but other workers found higher levels increasing 

with beer colour; one dark German beer contained 71.5 ppm (Kieninger and Birkova, 

1975). Lower levels of furfural are found (

ethyl acetate and 107±483ppm of ethyl lactate (Van Oevelen et al., 1976). The majority 

of the esters in beer are fermentation products but traces of esters from hop oil may be 

present. Methyl 4-decenoate and methyl 4,8-decadienoate, present in hop oil, are 

transesterified during fermentation to give the corresponding ethyl esters in beer. No 

doubt other esters in wort are transesterified during fermentation to give ethyl esters in 

beer.Lactones(Table19.8(e))arecyclicestersofhydroxyacids.Dihydro-2(3H)-furanone 

(19.3) and dihydro-5-methyl-2(3H)-furanone may be regarded as lactones. 

Ketones (Table 19.8(f)) are, like aldehydes, carbonyl compounds but they are not 

major fermentation products. Many of those present in beer may be derived from hop oil 

or hop resin degradation products. An exception may be 2-acetylfuran which occurs at 

higher levels in ales (90±97ppm) than in lagers (4±12ppm). 4-Hydroxy-3(2H)-furanone 

(19.4) derivatives also contribute to beer flavour. 5-Methyl-4-hydroxy-3(2H)-furanone 

has ameaty and brothy flavour note but the level in beer (0.12±0.77ppm) is less than the 

flavour threshold (8.3ppm). 2,5-Dimethyl-4-hydroxy-3(2H)-furanone has a sweet 

caramel flavour (threshold 0.16ppm) so the concentration in beer (0.19±2.73ppm) will 

cause it to influence flavour. 2-(or 5)-Ethyl-5-(or 2)-methyl-4-hydroxy-3(2H)-furanone 

also has asweet caramel flavour with alower threshold (0.02 ppm) than the dimethyl 

compound. It has been detected in some beers but not in others; in all cases the 

concentration was less than the threshold level (Mackie and Slaughter, 2000). 

Vicinal or -diketones have two carbonyl groups on adjacent carbon atoms. The most 

important vicinal diketone (VDK) in beer is diacetyl (9.14, butane-2,3-dione) which is 

accompanied by smaller amounts of pentane-2,3-dione. These diketones produce 

butterscotch flavours with thresholds of 0.07±0.15mg/l and 0.9mg/l respectively. 

Quantities in excess of 0.15mg/l of diacetyl are said to produce an off-flavour in lager 

beer but higher levels may be acceptable in ales and stouts. Many methods have been 

proposed for the analysis of diacetyl and vicinal diketones. The most specific involves 

gas liquid chromatography with an electron capture detector which is sensitive to vicinal 

diketones but not tothe majority ofotherbeer constituents (Harrison et al., 1965a,b). The 

Institute of Brewing has described methods using both packed and capillary columns 

(Buckee and Mundy, 1994) and the latter, which uses hexane-2,3-dione as internal 

standard, has been accepted by the EBC. The analyses given in Table 19.11 were 

obtained using such methods. In addition there are several colorimetric methods for 

determining vicinal diketones. The vicinal diketones are by-products of the biosynthesis 

of the amino acids valine and leucine (Fig. 12.21). One of the intermediates, - 

acetolactate, decomposes to diacetyl on heating so analytical methods which involve a 

preliminary distillation, especially those of fermenting wort, give high results. Headspace 

analysis avoids distillation and using this technique, with electron capture detection, it 

was found that during fermentation of an Irish stout the maximum level of vicinal 

diketones (0.6 ppm) was found 44 h after pitching but thereafter the level fell to about 

0.1 ppm. Yeast strains differ in the amount of diacetyl they produce and the choice of 

yeast strain is probably the most important factor in controlling the level of VDK in beer 


Table 19.11 Diacetyl and Pentane-2,3-dione content of beers (mg/l) (Harrison et al., 1965a,b) 

Quality* Diacetyl Pentane-2,3-dione 

Barley wine (4) 0.11±0.40 0.04±0.08 

Lager (9) 0.02±0.08 0.01±0.05 

Ale (9) 0.06±0.30 0.01±0.20 

Stout (5) 0.02±0.07 0.01±0.02 

Stout (1) 0.58 0.26 

* Number of samples in parentheses. 

(Portno, 1966). In particular respiratory deficient `petite mutants' of yeast produce large 

quantities of diacetyl and the off-flavour produced in beers infected with Pediococci (the 

so-called beer sarcina) is due to diacetyl. During fermentation ketones may be reduced to 

secondaryalcohols.Thusreductionofdiacetylgivesfirst3-hydroxybutan-2-one(acetoin) 

and then butane-2,3-diol which are much less volatile. They are also much less potent 

flavouring agents with threshold values of 17 and 4,500mg/l respectively. The addition 

of actively fermenting wort to beer may reduce the diacetyl content. The hop oil 

constituents which have been detected in beer are discussed in Chapter 8. 

19.1.5 Nitrogenous constituents 

Non-volatile 

During the 20th century the total nitrogen content of barley, malt and beer was usually 

measured by the Kjeldahl method where the sample is digested with concentrated 

sulphuricacid and asuitablecatalyst andthe nitrogenous constituentsare brokendown to 

ammonium sulphate. After dilution of the digest, the ammonia is liberated and distilled 

intostandardacid.AlthoughtheKjeldahlmethodhasbeenautomateditstillemploystoxic 

and hazardous reagents and today is being replaced by the older, but now automated, 

Dumas combustion method (Buckee, 1994, 1995, 1997; Johnson and Johansson, 1999). 

Herethesampleiscombustedinthepresenceofoxygenatabout1,000ëCtogiveoxidesof 

nitrogenwhicharecatalyticallyreducedtonitrogen.Otherproductsofcombustionsuchas 

carbondioxide andwater areremovedby selective absorptionandtheremaining nitrogen 

measured in athermal conductivity cell. The Dumas method consistently gives slightly 

higher total nitrogen values than those given by the Kjeldahl method but it long been 

known that the Kjeldahl method does not deal efficiently with certain types of compound 

(Buckee,1995).Free -aminonitrogenisdeterminedbyaninternationallyagreedmethod 

measuring the purple colour formed with ninhydrin at 570nm. 

The total nitrogen content multiplied by the factor 6.25 is often expressed as `protein'. 

Mostbeerscontain 300±1,000mg/ltotal-Nequivalent to0.11±0.63% protein.Anall-malt 

Burton strong ale contained 1,840mg/l total-N equivalent to 1.15% protein. There 

appears to be no universally accepted definition of protein. Some authorities classify 

proteinsbyfunction,othersbytheirmolecularsize.Itisagreedthathydrolysisofproteins 

gives polypeptides, peptides and eventually amino acids (terms such as `proteoses and 

`proteids' are even less well defined). One text suggests that peptides may contain up to 

10 amino acid residues, polypeptides 11±100 residues and proteins more than 100 

residues but Bamforth (1985) suggests that the term protein should be restricted to an 

undegraded molecule with an independent and unique identity such as, for example, a 

moleculeofanenzyme.Onthisbasishesuggestedthatfewproteinssurviveintobeerand 

most of the nitrogen is present as polypeptides. Williams et al. (1995) compared seven 


methods for determining the protein concentration of beer including both classical and 

automated Kjeldahl analyses. They concluded that the protein dye-binding assays 

(Coomassie Brilliant Blue and Pyrogallol Red-Molybdate methods) gave reproducible 

values for the protein concentration consistent with the results indicated by electrophoresis. 

Many new techniques have been used to study the proteins/polypeptides present in 

beerprincipallywiththeaimofcharacterizingthefractionsassociatedwithfoamstability 

and haze formation. Dialysis with Visking tubing retains compounds with molecular 

weight greater than 5,000 which accounts for approximately half of the total, N. Size 

exclusion chromatography (gel filtration) of beer shows the presence of polypeptides 

across the range of molecular weights from 2,000 to >100,000 but the majority have Mr 

Table 19.12 Amino acid analyses of beers (amino acids as g -amino N/ml) 

Indian pale ale Brown ale Draught bitter Mild ale Stout Stout 

OG 1042.8 1031.5 1040.8 1036 1045 1045 

Alanine (4.24) ± 0.19 0.43 ± 0.2 ± 

Ammonia 1.07 0.92 1.88 1.36 1.7 1.8 

Arginine (4.28) ± ± 0.18 ± ± ± 

Aspartic acid (4.29) ± 0.04 0.16 0.4 0.2 0.2 

Glutamic acid (4.33) ± 0.04 0.14 0.2 ± ± 

Glycine (4.35) 0.21 0.01 0.08 0.19 ± ± 

Histidine (4.36) 0.14 ± ± ± ± ± 

Isoleucine (4.38) ± ± 0.01 ± 0.4 ± 

Leucine (4.39) ± ± 0.03 ± 0.4 ± 

Lysine (4.40) 0.42 0.31 0.50 0.32 0.2 0.6 

Methionine (4.41) ± ± 0.24 ± ± ± 

Phenylalanine (4.42) ± 0.04 0.49 0.07 ± ± 

Serine (4.45) ± ± 0.15 0.05 0.2 0.3 

Threonine (4.46) ± ± 0.66 ± 0.2 0.2 

Tryptophan (4.47) 1.77 1.62 2.06 1.62 0.5 1.2 

Tyrosine (4.48) ± 0.50 0.35 0.07 ± ± 

Valine (4.49) ± ± 0.03 ± 0.4 0.7 

Total -amino N 3.61 3.74 7.39 3.88 4.4 5.0 

Proline 28.25 13.32 28.53 22.05 38.0 40.1 

-Amino N and proline 31.86 17.06 35.92 25.93 42.4 45.1 

Total N 276 336 459 ± ± ± 


Table 19.13 Nucleotides in beer (Qureshi et al., 1979) 

Concentration ( g/ml) Taste threshold (ppm) 

5 0 -Cytidine monophosphate 1.8±7.8 ± 



5 0 -Adenosine monophosphate 12.4±67.6 ± 

5 0 -Guanosine monophosphate 1.1±4.9 35 


5 0 -Uridine monophosphate 2.8±9.1 1.7±4.5 


3 0 -Uridine monophosphate 1.5±10.3 ± 

2 0 -Uridine monophosphate 9.3±26.2 ± 

3 0 -Guanosine monophosphate 1.6±5.1 ± 

5 0 -Thymidine monophosphate 2.0±11.9 ± 

5 0 -Inosine monophosphate 1.5±4.7 120 

3 0 -Inosine monophosphate 1.3±8.1 ± 

3 0 -Thymidine monophosphate 1.4±6.6 ± 

2 0 -Guanosine monophosphate 2.4±6.7 

also examined the small peptides (Mr

Table 9.14 Purines, pyrimidines and nucleosides in beer 

(Qureshi et al., 1979) 

Cytosine (4.58) 

Concentration ( g/ml) 

11±24.6 

Cytidine 18±41 

Guanine(4.57) 0.2±3.2 

Adenine (4.56) 0.8±5 

Uracil (4.59) 1.0±4.6 

Uridine 21±70.3 

Adenosine 12.5±24.3 

Xanthine (4.63) 2.8±9.7 

Inosine 1.0±2.4 

Guanosine 45±139 

Thymidine 7±19.8 

Table 19.15 Amides in dark German beer (Tressl et al., 1977) 

N,N-Dimethylformamide 

N,N-Dimethylacetamide 

N-Methylacetamide 

N-Ethylacetamide 

N-(2-Methylbutyl)acetamide 

N-(3-Methylbutyl)acetamide 

N-Furfurylacetamide 

H.CONMe2 

CH3.CONMe2 

CH3.CONHMe 

CH3.CONH.CH2Me CH3.CONH.CHMe.CH2 CH2Me 

CH3.CONH.CH2 CH2. CHMe2 

Concentration (ppb) 

15 

10 

+ 

20 

10 

25 

120 

N-(2-Phenylethyl)acetamide CH3.CONH.CH2CH2C6H5 15 

(10ppb), have been associated with stale grainy flavours in beer at 2,000 and 5ppb 

respectively (Palamand and Grigsby, 1974). Some amides present in beer are given in 

Table 19.15 but both they and the heterocyclic compounds formed during wort boiling 

(Chapter 9) are slightly volatile. Nevertheless, many survive into beer (Table 19.16) 

although the concentrations found are usually below the threshold values. Low levels of 

pyrroles have also been found in beer (Table 19.17). 

Volatile 

The biogenic amines and polyamines in beer have been reviewed by KalacÏand KrõÂzÏek 

(2003). At the pH of beer ammonia and the volatile amines found in beer (Table 19.18) 

are present astheir non-volatile salts. Ammonia isthe most abundantvolatile nitrogenous 

constituent. The mean value in arange of US beers was 14.6ppm (3±33ppm) compared 

with 21.3ppm (0±33ppm) in imported beers (Owades and Jacevac, 1959). At the low 

concentrations found the volatile amines will have little influence on the flavour of beer. 

Afterammonia,dimethylamineisthemajorcomponentofthisclass.Tyramine(4.68),the 

decarboxylation product of tyrosine, has been detected in beer. Patients being treated for 

depression with monoamine oxidase inhibitors must avoid alcoholic drinks and yeast 

products due to the build up of toxic levels of tyramine. Traces of ethyl carbamate 

(urethane, H2N.CO2C2H5), which is reported to be carcinogenic, have been found in 

many fermented and distilled beverages. In asurvey of 933 samples of Scotch whisky 

15±115ppb (mean 43ppb) of ethyl carbamate were found but in asimilar survey of 69 

beers, Canas et al. (1989) found that 30 beers had between one and four ppb, two had 

between 9and 13ppb but in the majority ethyl carbamate could not be detected. 


Table 19.16 Some heterocyclic bases present in beer (Harding et al., 1977) 

As mentioned above, secondary amines, such as dimethylamine, can react with oxides 

of nitrogen to form carcinogenic N-nitrosamines: 

H3C H3C 

\ \ 

2 NH + N2O3 ! 2 N.NO + H2O 

/ / 

H3C H3C 

(4.78) (4.85) 

Concentration (ppm)* Threshold in light ale (ppb) 

Pyridine (9.33) 

2-Acetylpyridine (9.42) 100 

3-Acetylpyridine 

Pyrazine (9.34) 19±48.5 

Methylpyrazine 201±279 (410±419) 1000 

2,3-Dimethylpyrazine 0.9±6.5 20 



Ethylpyrazine 5.9±11.7 

2-Ethyl-5-methylpyrazine 5.4±35 

2-Ethyl-6-methylpyrazine 0.5±3.4 

Trimethylpyrazine 0.7±7.7 100 

2-Ethyl-3,5-dimethylpyrazine 6.5±24.6 

2-Ethyl-3,6-dimethylpyrazine 1.9±13.8 50 

2-Ethyl-5,6-dimethylpyrazine 

Tetramethylpyrazine 9.2±62.0 200 

Acetylpyrazine 8.1±19.2 100 

6,7-Dihydro-5H-cyclopentapyrazine 

2-Methyl-6,7-dihydro-5H-cyclopentapyrazine 

Thiazole (9.32) 

2-Furanmethanol (9.20) 

Furfural (9.17) 

2-Acetylfuran (9.25) 4±97 80 000 

Dihydro-2(3H)furanone 

Dihydro-5-methyl-2(3H)furanone 

5-Methyl-2-furfural (9.22) 

2-Thiophenecarboxaldehyde (9.19) 

2-Acetylthiophene (9.27) 

5-Methyl-2-thiophenecarboxaldehyde (9.24) 

* Qureshi et al. (1979) 

N-Nitrosodimethylamine has been found in beer and many other foodstuffs. German 

beers were found to contain 0±68 ppb (mean 5.9 ppb) N-nitrosodimethylamine while 

American beers had 0±14 ppb (mean 5.9 ppb). The highest levels were found in a dark 

strong German lager (maximum 47 ppb) and in Rauchbier (maximum 68 ppb). N- 

Nitrosodiethylamine and N-nitrosoproline have also been found in beer. After much 

research it was found that the nitrosamines were formed during the kilning of malt 

especially in direct fired kilns. Here the mixture of nitrogen oxides, NO x, react with the 

amines present, probably hordenine, to form the nitrosamine. With this knowledge the 

level of nitrosamines in beer has been greatly reduced. Nevertheless in 1980 the US Food 

and Drug Administration set a limit of not more than 5 ppb of N-nitrosodimethylamine in 

beer and the ASBC describe a method to measure nitrosamines in beer. 


Table 19.17 Pyrrole derivatives in beer (Tressl et al., 1977) 

Concentration (ppb) 

Pyrrole + 

2-Methylpyrrole 1800 

2-Formylpyrrole (9.18) 30 

2-Acetylpyrrole (9.26) 1400 

2-Acetyl-5-methylpyrrole 10 

2-Formyl-5-methylpyrrole (9.23) 110 

2-Pyrrolidone 10 

1-Methyl-2-pyrrolidone + 

1-Acetylpyrrole + 

1-Furfurylpyrrole 10 

Indole + 

Table 19.18 Volatile amines present in beer (Slaughter and Uvgard, 1971) 

Concentration (ppm) 

Methylamine (4.64) 

Ethylamine(4.65) a 

0.03±2.12 

n-Propylamine 

n-Butylamine 

Isobutylamine (4.66) b 

0.05±0.10 

sec-Butylamine 

n-Amylamine 

Isoamylamine 

Hexylamine 

1,3-Diaminopropane 

1,4-Diaminobutane (4.74, putrescine) 

1,5-Diaminopentane(4.75, cadaverine) 

N,N-Dimethyl-1,4-diaminobutane 

Dimethylamine (4.78) b 

0.07±0.78 

Diethylamine 

Di-isobutylamine 

Pyrrolidine (4.71) 

Trimethylamine (4.79) b 

0.02±0.06 

Tripropylamine 

N,N-Dimethylbutylamine 

Ethyl carbamate (4.87, urethane) 0±0.013 

Ethanolamine 

p-Hydroxybenzylamine c (4.80) 0.16±0.72 

Tyramine (4.68) 0.15 

Histamine (4.69) d 

0.08±0.55 

a Palamand et al. (1969) 

b Koike et al. (1972) 

c Slaughter and Uvgard (1972) 

d Chen and Van Gheluwe (1979) 

19.1.6 Sulphur-containing constituents 

Beers contain 100±400ppm of sulphate (Table 19.3). The major non-volatile organic 

sulphur compounds in beer are the amino acids cyst(e)ine and methionine and the 

peptides and proteins which contain them. Some of these compounds will survive into 

beer. In addition malt contains S-methylmethionine (4.157) and dimethyl sulphoxide 

(4.160) which are precursors of dimethyl sulphide (4.158). Hops may be asource of 

sulphur; they may be dusted with elemental sulphur before burr but the burning of 


sulphur on the oast is less common today. The sulphur compounds in hop oil are 

discussed in Chapter 8. 

The analysis of volatile sulphur compounds is quite difficult as additional compounds 

may be formed if the sample is heated or exposed to light and/or oxygen. Headspace 

analysis using aflame photometric detector is probably the most satisfactory technique. 

Peppard (1985) describes purging beer (sulphur) volatiles, their absorption in aPoropak 

Qtrap, and their subsequent desorption onto afused silica WCOT capillary column. The 

volatile sulphur compounds which have been identified in beer are listed in Table 19.19. 

Brewery fermentations can produce up to 10mg/l of sulphur dioxide and sodium (or 

Table 19.19 Volatile sulphur compounds in beer 

Compound Typical levels Flavour Flavour 

( g/l) threshold description 

( g/l) 

Hydrogen sulphide 1±20 5 Sulphidic, rotten eggs 

Sulphur dioxide 200±20,000 >25,000 Sulphitic, burnt match 

Carbon oxysulphide ± ± ± 

Carbon disulphide 0.01±0.3 >50 ± 

Thioformaldehyde ± ± ± 

Methanedithiol (Dithioformaldehyde) ± ± ± 

Thioacetone ± ± ± 

Methanethiol 0.2±15 2.0 Putrefaction, drains 

Ethylene sulphide 0.3±2.0 20 ± 

Ethanethiol 0±20 1.7 Putrefaction 

Propanethiol 0.1±0.2 0.15 Putrefaction 

1,1-Dimethylethanethiol ± ± ± 

2-Furfurylmercaptan ± ± Rubbery 

Dimethyl sulphide 10±100 30 Sweetcorn, tin 

Diethyl sulphide 0.1±1.0 1.2 Cooked vegetables 

Dimethyl disulphide 0.1±3 7.5 Rotten vegetables 

Diethyl disulphide 0±0.01 0.4 Garlic, burnt rubber 

Dimethyl trisulphide 0.01±0.8 0.1 Rotten vegetables, 

onion 

n-Butyl methyl sulphide 1 ± ± 

Methionol 50±1,300 2,000 Raw potatoes 

Methional 20±50 250 Mash potatoes, souplike 

Methyl thioacetate 5-20 50 Cabbage 

Ethyl thioacetate 0-2 10 Cabbage 

3-Methylthiopropionic acid ± ± ± 

Ethyl 

3-Methylthiopropionate 5±180 ± ± 

3-Methylthiopropyl acetate ± ± ± 

2-Methyltetrahydro-thiophen-3-one ± ± ± 

S-Methyl 

2-Methylbutanethiolate ± 1 ± 

S-Methyl 

4-Methylpentanethiolate ± 15 ± 

S-Methyl hexanethiolate ± 1 ± 

4-(4-Methylpent-3-enyl)-3,6-dihydro- 

1,2-dithiine ± ± ± 

3-Methyl-2-butene-1-thiol 0.001±0.1 0.01 Skunk, leek-like, light 

struck 

After Baxter and Hughes (2001) with additions 


potassium) metabisulphate may be added as apreservative either during conditioning or 

at racking. However, EEC regulations (1987) have limited the level of SO2 in packaged 

beer to 20mg/kg and to 50mg/kg in cask conditioned beers. In contrast, 160±400mg/kg 

SO 2areallowedinwineand200mg/kgincider.Muchofthesulphurdioxideinbeerisin 

abound form such as the acetaldehyde bisulphite compound. There are several approved 

methods for the analysis of sulphur dioxide in beer. Analytica-EBC and the Institute of 

Brewing use the classicMonier-Williams distillation method inwhich the SO2isdistilled 

in acurrent of nitrogen or carbon dioxide into hydrogen peroxide and the sulphuric acid 

formed titrated with standard sodium hydroxide. The ASBC describe amethod whereby 

the colour which is restored to acid-decolourised p-rosaniline hydrochloride is measured. 

In a collaborative trial (Baker and Upperton, 1992) these two methods gave better 

precision than the IoB rapid method or the EBC dithiobisnitrobenzoic acid method. 

However, concern has been expressed about the possible carcinogenic nature of prosaniline. 

Analytica-EBC also describes an enzymatic method for sulphur dioxide. Sulphite is 

oxidized to sulphate by the enzyme sulphite oxidase (SO2.OD): 

SO2.OD 

SO3 2 ‡O2 ‡H2O !SO4 2 ‡H2O2 

The hydrogen peroxide produced is reduced by the enzyme NADH-peroxidase (NADH- 

POD): 

NADH-POD 

H2O2 ‡NADH‡H ‡ 

!NAD ‡ ‡2H2O 

and the NADH is measured by its absorption at 340nm. The level of SO2 in beer seldom 

exceeds the flavour threshold and declines during storage with ahalf life between 37 and 

221 days; in one beer ahalf life of 1,038 days was found (Ilett et al., 1996). 

The hydrogen sulphide present inbeer is also partlyin abound form but the total level 

may exceed the threshold level. At this low level the flavour impact is not unpleasant and 

is characteristic of the flavour of ales especially cask conditioned ales which may have 

had potassium metabisulphite added as apreservative. Lager yeasts are less efficient in 

reducing SO2 to H2S than ale yeasts. In avigorous fermentation much of the H2S will be 

removed with the carbon dioxide. Yeast autolysis or microbial infection, e.g., by 

Zymomonas spp., can also produce hydrogen sulphide. The amounts of methanethiol, 

ethanethiol and propanethiol in beer (Table 19.19) probably exceed their flavour 

thresholds. 

Dimethyl sulphide (4.158, DMS) is an important component of the flavour of lager 

beers. The concentration in arange of beers, given in Table 19.20, shows the higher 

levels in lager beers. As mentioned above it is mainly formed by the breakdown of Smethylmethionine 

present in malt. Lightly kilned lager malts will retain more Smethylmethionine 

than ale malts. Some brewers specify the level of S-methylmethionine 

in the malt they buy. DMS formed during kilning and wort boiling will be lost to the 

atmosphere but that formed in the whirlpool and during later processing is likely to 

remain in the wort unless vapour stripping is employed. Some DMS will be purged with 

the CO2 during fermentation but some will survive into beer. During kilning some Smethylmethionine 

is converted into dimethyl sulphoxide (DMSO) which is not as volatile 

as DMS but soluble in water so it will survive into wort. Here some yeast strains are 

capable of reducing it back to dimethyl sulphide. Late or dry hopping with whole hops, 


Table 19.20 Dimethlyl sulphide content of beer (ppb) 

British ales 14 Sinclair et al. (1970) 

British lagers 16±27 Sinclair et al. (1970) 

Continental lagers 44±114 Sinclair et al. (1970) 

Beer from green malt 80 Anderson et al. (1975) 

German lagers 32±205 (av. 94) Narziss et al. (1979) 

Diet beer 46±98 (av. 63.5) Narziss et al. (1979) 

Canadian ale 92 Hysert et al.(1979) 

Canadian lager 114 Hysert et al. (1979) 

Canadian low-alcohol beer 82 Hysert et al. (1979) 

British 41±75 ± 

German 141±153 ± 

United States 59±106 ± 

will add as much as 15 ppb to the beer dimethyl sulphide level (Moir, 1994). Hegarty et 

al. (1995) found that 2-phenylethanol can suppress the flavour intensity of dimethyl 

sulphide. Dimethyl sulphide in beer is usually estimated by headspace gas chromatography 

with acapillary column. Aflame photometric detector is specific for sulphur 

compounds but aflame ionization detector gives similar results for DMS and also allows 

measurement of acetaldehyde, ethyl acetate, n-propanol, isobutanol and isoamyl alcohol 

(Analytica-EBC, IoB, Dupire, 1999) (see also Mundy, 1991). 3-Methylthiopropionaldehyde 

was found to contribute the worty flavour to alcohol-free beers (Perpete and Collin, 

1999). The same compound is the precursor of dimethyl trisulphide. 

As discussed in Chapter 8, photolysis of the iso- -acids in the presence of asensitizer 

(riboflavin) and asuitable sulphur donor can produce alight- or sun-struck flavour. This 

is due to 3-methyl-2-butene-1-thiol (8.48, prenyl mercaptan) which produces askunklike, 

leek-like aroma with an extremely low threshold of 10ng/l (0.10ppb). To avoid this 

off-flavour few beers are now bottled in clear glass bottles. Alternatively the beer may be 

bitteredwitheithertetrahydroiso- -acidsor -iso- -acidswhicharenotaffectedbylight. 

19.2 Nutritive value of beer 

The nutritional aspects of beer are discussed by Baxter and Hughes (2001). The calorific 

(caloric, energy) value of beer is calculated from the alcohol, carbohydrate and protein 

contents. The (UK) Food Labelling Regulations (1980) give the following figures for 

calculating calorific values: 1gcarbohydrate (as monosaccharide) ˆ3.75 kilocalories, 

1gprotein ˆ4kcal, 1gfat ˆ9kcal, and 1galcohol ˆ7kcal. Accordingly the IoB give 

the formula: 

Calorific value …kcal=100ml† ˆ7…A†‡3:75…C†‡4…P† 

where Aâlcohol content/100 ml, Cˆtotal carbohydrate (as glucose)/100ml, and Pˆ 

protein content/100ml (Martin, 1982). Alternatively, the result may be expressed in 

kilojoules (1kcal ˆ4.184kJ): 

Energy value …kJ=100ml† ˆ29…A†‡17…C†‡17…P† 

Some typical energy values are given in Table 19.21. It can be seen that alcohol 

contributes more energy than the carbohydrates so low carbohydrate `lite' beers can have 


Table 19.21 Calorific (energy) values of beer 

Beer type Energy Value 

kJ/100 ml kcals/100 ml 

`Lite' beers 75±110 20±26 

Lagers 85±125 20±30 

Ales 114±160 25±38 

Stouts/strong beers 150±300 35±70 

From Baxter and Hughes (2001) 

Table 19.22 Vitamin content of beers 

Vitamin 

(mg/l) 

Range in beers (mg/l) Typical level in UK beer 

Niacin 3±20 7.7 

Riboflavin 0.07±1.3 0.3 

Pyridoxine (B6) 0.13±1.7 0.5 

Foliates 0.03±0.10 0.05 

Biotin 0.007±0.018 0.01 

B12 0.09±0.14 0.1 

Pantothenic acid 0.5±2.7 0.9 

Thiamine 

After Baxter and Hughes (2001) 

0.002±0.14 0.03 

high energy values due to the alcohol present. However, in some countries the term `lite' 

refers to low-alcohol beers. 

The ASBC use slightly different figures to calculate the caloric content of beers (i.e. 1 

gcarbohydrate ˆ4kcal and 1g alcohol ˆ6.9kcal) and take into account the ash content 

of the beer so that: 

Caloric content …kcal=100ml beer† ˆ6:9…A†‡4…B C† 

where AÂlcohol (% by weight), BˆReal extract (% by weight), and Câsh content 

(%byweight).Beerisarichsourceofcertain vitamins(Table19.22).BaxterandHughes 

(2001) discuss the contribution that beer can make to the required daily intake of the 

vitamins. 

19.3 Colour of beer 

The physical properties assessed by abeer drinker include colour, clarity, viscosity and 

foam. Obviously if the beer is drunk directly from acan or bottle these properties will 

have less impact on the consumer. The colour of beer is largely due to the melanoidins 

and caramel present in the malt and adjuncts used but further caramelization can take 

place during wort boiling (see Chapters 9and 10). Minor adjustments of the colour of 

beer can be made by the addition of caramel either to the copper or with the primings. 

Other contributors to the colour of beer are oxidized polyphenols especially in the 

presence of trace metals such as iron or copper. In pale beers the yellow vitamin 

riboflavin (

Extinction 

Wave numbers (×10 

25 24 23 22 21 20 19 18 17 16 15 

–3 ) cm –1 

Wave numbers (×10 

25 24 23 22 21 20 19 18 17 16 15 

–3 ) cm –1 

(a) (b) 

0.9 

0.8 

0.7 

0.6 

0.5 

0.4 

0.3 

0.2 

0.1 

0 

f 

b 

c 

d 

a 

A new series of glass discs were released in 1950 and adopted by the EBC. The discs (2± 

27 EBC colour units) were matched against the beer in either 5, 10, 25 or 40 mm cells. 

However, it is necessary that every person entrusted to the measurement of the colour of 

worts and beers in this way is known to be free from colour blindness: 10% of the male 

and 1% of the female population do not have perfect colour vision. Ishihara (1964) has 

provided tests for colour blindness. In view of these difficulties spectrophotometric 

measurements at one wavelength were adopted. Analytica-EBC chose measurements at 

430 nm in a 1 cm cell when: 

EBC Colour ˆ A430 25 

Originally the Institute of Brewing chose measurements at 530 nm as being more 

suited to ales but later adopted 430 nm. The ASBC also take measurements at 430 nm but 

in a half-inch cell when: 

Color …ASBC† ˆ A430 10 

e 

Key 

a Pale ale 1 

b Pale ale 2 

c Pale ale 3 

d Light ale 1 

e Light ale 2 

f Pale ale 4 

417 435 455 477 500 526 556 589 625 667 


Nevertheless, single wavelength measurements can provide only limited information. In 

the human eye cone cells in the centre of the retina respond to either red (c. 600 nm), 

green (c. 550 nm) or blue (c. 450 nm) light and send three signals to the brain which are 

there integrated and interpreted as colour. Hence tristimulus values are being increasingly 

used to measure the colour of beers (Sharpe et al. 1992, Smedley, 1992, 1995). 

When evaluating colour the energy of the illuminant light source across the visible 

spectrum (`colour temperature') must be taken into consideration. The Commission 

Internationale de l'Eclairage (CIE) have defined a number of standard illuminants: 

Illuminant B (colour temperature, 4,900 K) represents sunlight, Illuminant C (6,800 K) 

represents average daylight, and Illuminant D65, which is most commonly used, 

represents daylight with some UV correction. The CIE defines colour in terms of three 

parameters: hue (hë), value (L*), and chroma (C*) as illustrated in the CIELAB colour 

Extinction 

0.8 

0.7 

0.6 

0.5 

0.4 

0.3 

0.2 

0.1 

0 

e 

d 

c 

b 

a 

Key 

a-d Largers 

e Stout 

417 435 455 477 500 526 556 589 625 667 


Fig. 19.2 (a) Spectra of commercial ales; (b) spectra of lager and stout (Hudson, 1969). 


space (Fig. 19.3). `Hue' is the term ascribed to what we generally consider as the 

prominent `colour' of an object. The hues, red, yellow, green, blue and purple form a 

continum in the horizontal plane of the colourspace which is sometimes called the colour 

wheel (Sharpe et al., 1992). It should be noted that the hue (h) is a measure of angular 

rotation. The value (L*) is a measurement of `light' and `dark' and is a vertical 

displacement from the north pole (white) to the south pole (black). The chroma (C*) 

measures what observers call `dull' or `vivid' and is a horizontal displacement of the 

vertical axis of the colour space towards the circumference. For dealing with differences 

in colour between two samples it is generally more convenient to express locations within 

the colour space using Cartesian co-ordinates. Here, hue and chroma are combined into 

two parameters a* and b* so: 


Chroma …C † ˆ a 2 

‡ b 2 

Hue …h † ˆ tan 1 …b =a † 

Commercial tristimulus transmission instruments are available which record colour in 

terms of L*, a* and b*. There is a linear relationship between L* and EBC colour units 

but tristimulus measurements can distinguish between beers showing the same EBC 

colour values (Smedley, 1995). The ASBC have recommended the use of tristimulus 

analysis for measurement of beer colour (Cornell, 2002). 

19.4 Haze 

Blue 

Green 

Chroma 

White 

Black 

Value 

Yellow 

Fig. 19.3 A representation of the CIELAB colour space showing the relationship between 

value, hue and chroma (Smedley, 1995) 

Clarity is the absence of haze. Most drinkers expect their beer to be bright and clear and 

may reject cloudy beer untasted. Beer hazes are of two types: biological and nonbiological. 

Infection of bright beer with either bacteria or wild yeasts will produce a 

biological haze due to the growth of the invading organism when the beer will usually 

become sour and unacceptable. With the use of pasteurization and sterile filtration, 

Hue 


Red

iological infection and haze formation is fairly rare. However, sterile beers kept for a 

length of time will develop anon-biological haze. The rate of development of such hazes 

determinestheshelf-lifeofbottledandcannedbeers.Beforebeershowsanyhazeatroom 

temperature (20ëC) it may form achill haze if cooled to 0ëC, Such chill hazes will 

redissolve at 20ëC when permanent hazes will remain. Chill hazes are obviously more of 

aproblem with lager beers, which may be served as cold as 4ëC, than with ales. Beer 

haze has been reviewed by Bamforth (1999) and by Siebert (1999). 

19.4.1 Measurement of haze 

The amount of chill haze isolated for an EBC collaborative analysis from three different 

breweries was between 1.4 and 8.1mg/l while the permanent haze accounted for 6.6± 

14.6mg/l (Carrington et al., 1972). The amount of permanent haze increases with time of 

storage and yields of up to 44mg/l have been reported but beers are commercially 

unacceptable long before the haze can be measured gravimetrically. When light is passed 

through asuspension of acoloured precipitate and the amount of reflection is negligible, 

the light absorption gives ameasure of the turbidity according to Lambert's law. With 

white precipitates, such as beer haze, much of the light is reflected so measurements are 

made of the light reflected at agiven angle to the incident light (nephelometry). The 

angle chosen and the size of the haze particles are the two most important factors which 

determine the amount of haze perceived (Thorne and Nannested, 1960). Most of the 

commercial instruments in use today take measurements at 90ëto the incident light: the 

exception is the Monitek 251 which takes readings at aforward angle of 13ë. The 

wavelength of the incident light also varies between instruments (350±860nm) and some 

instruments are more sensitive to colour than others (Buckee et al., 1986). Mundy and 

Boley (1999) concluded that there are significant differences between the haze values 

obtained on the same sample with different instruments, confirming that the results 

obtained on different instruments cannot be directly compared. The EBC, IoB and ASBC 

all use formazin, prepared by the reaction between hydrazine sulphate and 

hexamethylenetetramine, as the primary haze standard. Unfortunately the EBC and the 

ASBC have adopted different scales so: 

10;000 ASBC Formazin Turbidity units ˆ145 EBC Formazin haze units 

1EBC haze unit ˆ69 ASBC haze units 

Earlier, barium sulphate (Helm) and fullers' earth were used as standards. A 

comparison of the different scales with visual assessment is given in Fig. 19.4. Styrenedivinylbenzene 

copolymer suspensions (AEPA) have also been used as turbidity 

standards but samples from different suppliers had different mean particle sizes. Coulter 

counter analyses showed that the majority of the particles in the Formazin haze standard 

were between 1.5 m and 2.5 m diameter (mean 2.1 m) (Morris, 1987). In most cases 

visual assessment of beer haze correlates well with instrument readings for light scattered 

at 90ë but some beers which appear bright to the eye give substantial meter readings. 

Such beers were said to contain ìnvisible' or `pseudo' haze. These `pseudo' hazes were 

not observed with instruments using 13ë forward light scattering. Morris (1987) 

investigated the relationship between haze and particle size. He found that the 90ë 

(Radiometer) haze meter gave results which can be closely related to turbidity for 

particles > 0.5 m in diameter but for particles < 0.5 m it was somewhat oversensitive. 

A high reading will therefore be obtained for a suspension containing particles < 0.5 m 


Comparison of haze scales 

It must be stressed again that no exact equivalence must be expected between different 

haze scales or between visual observations made in different laboratories. However, a rough 

comparison between haze scales can be made with the help of this nomogram and with the 

relationship: 

1 EBC Formazin Haze Unit = 40 Helm units = 69 ASBC Formazin units 

10 

EBC 

units 

9 

8 

7 

6 

5 

4 

3 

2 

1 

Helm 

units 

400 

350 

300 

250 

200 

150 

100 

50 

ASBC 

units 

700 

Verbal 

assessment 

diameter if the calibration was originally carried out with particles > 0.5 m, which 

explains the observation of ìnvisible' or `pseudo' haze. 

19.4.2 Composition and formation of haze 

The most common non-biological hazes found in beer are formed by interactions between 

proteins and polyphenols. Chill hazes contain between 45.5 and 66.8% of proteins, the 

hydrolysates of which are rich in glutamic acid, proline, arginine and aspartic acid 

residues. Alkaline hydrolysis of beer hazes produces a range of phenolic acids including: 

ferulic, sinapic, vanillic, syringic, gallic, protocatechuic and caffeic acids (Harris, 1965). 

In contrast, acid treatment of beer hazes liberates the pigments cyanidin and delphinidin 

showing the presence of proanthocyanidins in beer haze. However, none of the 

proanthocyanidins found in malt, hops and beer contain a methoxy group so the ferulic, 

vanillic, sinapic, and syringic acids produced on alkaline hydrolysis must come from 

another source probably lignin. Barley straw lignin contains 16.4% methoxyl so, on the 

assumption that all the methoxyl in haze is derived from lignin and that this lignin has the 

same composition as barley straw lignin, hazes contain 5.7±7.9% lignin (Harris, 1965). 

Hazes contain 0.7±3.3% of ash rich in many metals (Hudson, 1955). Some of the metals, 

e.g., copper, iron and aluminium, are concentrated in haze up to 80,000 times the level in 

the residual beer. Hazes also contain 2±4% of glucose and traces of the pentoses 

arabinose and xylose. 

600 

500 

400 

300 

200 

100 

Very hazy 

Hazy 

Slightly hazy 

Very slighly hazy 

A1 brilliant 

Brilliant 

Fig. 19.4 Comparison of haze scales (Analytica-EBC, 1998). 


During wort boiling (Chapter 9) some protein is removed in the hot break or trub but 

protein-polyphenol complexes, which dissociate at 80ëC, are not found in the hot break but 

do occur in the cold break. Proteins not removed during wort boiling may survive 

fermentation and persist into beer where they may cause haze. However, not all proteins 

nor all polyphenols are haze-active. The suggestion that hydrophilic proteins are 

responsible for haze and hydrophobic proteins are necessary for head retention is probably 

too facile. Haze-active proteins or polypeptides in beer are mainly derived from the barley 

prolamins or hordeins, which are alcohol soluble and rich in proline. Indeed proline 

residues appear to be necessary for haze formation. In amodel system with catechin and 

different polypeptides, haze increased linearly with the mole %proline in the polypeptide, 

the most haze being formed with polyproline. Polypeptides which contain little or no 

proline produce little or no haze. However, Ishibashi et al. (1996) found that antibodies 

raised against foaming polypeptides were specific for those compounds but antibodies 

raised against haze reacted with both haze and foaming polypeptides. It is suggested that 

albumin- and globulin-derived material may adhere to particles formed originally from 

hordein-derived polypeptides. Yang and Siebert (2001) investigated dyes that bind to 

proteins andfoundthatbromopyrogallolredgavethebestindicationofhaze-activeprotein. 

Similarly, not all the polyphenols present in beer are haze-active; proanthocyanidins 

and flavanols are the most important. Catechin, epicatechin and gallocatechin give small 

amounts of haze with abeer haze-active protein, but the dimers, e.g., procyanidin B-3, 

produce much more haze and the trimers more again. As mentioned above, tetramers and 

pentamers do not survive into beer but they may reform by oxidation as beers age. Of the 

dimers, prodelphinidin B-3 produces more haze than procyanidin B-3. The amount of 

haze formed depends on the concentrations of both the protein and the polyphenol and 

their ratio. Siebert et al. (1996) produced amodel for protein-polyphenol interactions 

(Fig. 19.5) in which both the polyphenols and the haze-active proteins have afixed 

number of binding sites (shown as two and three respectively in Fig. 19.5). When the 

number of polyphenol ènds' equals the number of protein binding sites, the largest 

network with the largest particles will be formed which will show the greatest amount of 

lightscattering.Withanexcessofprotein,thepolyphenolwillformabridgebetweentwo 

peptide chains but there will be insufficient for further bridges. With an excess of 

polyphenol relative to protein, all the protein binding sites will be occupied and it is 

unlikely that the free end of the polyphenol will find avacant protein site for further 

cross-linking. 

Occasionally hazes which differ from the normal protein-polphenol haze are found in 

beer. These include calcium oxalate hazes (the solubility of calcium oxalate is only 

6.07mg/l at 13ëC) and carbohydrate hazes such as retrograded starch ( -glucan), - 

glucans and pentosans (Chapter 15). 

19.4.3 Prediction of haze and beer stability 

The conditions to which any particular beer package will be subjected cannot be known 

beforehand; therefore two types of test are adopted by most breweries: 

1. Long-term storage at a temperature related to that likely in trade with an examination 

for biological and non-biological haze at the end of a defined period (e.g. three or six 

months). 

2. Accelerated haze production under defined high temperature conditions, to give an 

early indication of the liability to non-biological haze. 


[Polyphenol] = [Protein] 

[Polyphenol] > [Protein] 

Polyphenol molecule 

Protein molecule with fixed 

number of polyphenol binding 

sites (i.e. haze-active) 

Fig. 19.5 Conceptual mechanism of protein-polyphenol interaction (Siebert et al., 1996). 

Copyright (1996) American Chemical Society 

The EBC recommendation is to measure the increase in non-biological haze at the end of 

a defined period at a defined temperature. Brewers will adopt different conditions to suit 

their beers but for comparison the EBC recommends that six bottles should be cooled to 

0 ëC overnight and the initial haze read in the morning at 0 ëC. The bottles are then placed 

in an upright position and without agitation, in a bath at 60 ëC for 48 hours. At the end of 

this time they are again cooled at 0 ëC overnight and the haze measured the following 

morning at 0 ëC The initial and final haze values are reported in EBC Formazin units. 

The ASBC measure the total haze after chilling in a similar manner and describe 

accelerated chill haze tests. They suggest that 12 weeks or three months is a reasonable 

time in which beer can be expected to be sold. Accordingly, they recommend storing 24 

bottles or cans at 22 ëCÔ2 ëC in a vertical position without agitation for eight weeks. 

Three containers are then cooled overnight at 0 ëC and the haze measured at 0 ëC. Each 

week, up to 12 weeks, three more containers are cooled and the haze measured. For the 

accelerated chill haze test three containers are held at the selected temperature (40, 50, or 

60 ëC) for one week. The samples are then cooled at 0 ëC for 24 hours and the haze 

measured. (Note that caution should be exercised when the test is run at 60 ëC because 

there is some danger that bottles may burst and cans may rupture.) The results of the 

accelerated chill haze tests, in Formazin Turbidity Units, are compared with the haze 

formed after 12 weeks at 22 ëC to determine which forcing temperature gives results that 

correlate best with room temperature storage. Thereafter only that forcing temperature 

need be used routinely. 

These accelerated shelf-life test take at least 48 hours or one week to give results. 

More rapid predictions of beer stability are sometimes obtained using chemical 


precipitants. The IoB describe methods for sensitive protein in beer, alcohol chill haze in 

beer, and for the saturated ammonium sulphate precipitation limit of beer. 

19.4.4 Practical methods for improving beer stability 

Brewers wish to increase the shelf-life of their beers and adopt various strategies to 

achieve this, i.e., removal of protein, removal of polyphenols or removal of aportion of 

each (Chapter 15). This last method is usually employed since some proteins and some 

polyphenols are necessary for the character of the beer. To reduce the level of proteins 

some (lager) brewers used to treat their beers with proteolytic enzymes (papain, 

bromelain,ficin,orenzymesfromBacillussubtilis)butsuchtreatmentislikelytoremove 

both haze-active and foam-active proteins. Similarly, treatment with bentonite, removes 

both types of protein but silica hydrogels are more specific for haze-active proteins 

(Siebert and Lynn, 1997). Apperson et al. (2002) studied the silica absorption of hazeactive 

beer proteins and found that non-activated small pore volume silica (silica B) was 

one of the most efficient absorbers of isolated haze protein. Alternatively, tannic acid is a 

relatively efficient precipitant of haze-active proteins in beer. 

Proanthocyanidin-free barley varieties have been bred and malted which, when used 

with apolyphenol-free hop extract, give colloidally stable beers. However, most brewers 

still rely on absorbants to remove excess polyphenols. After nylon and polyvinylpyrrolidone 

(PVP), cross-linked polyvinylpyrrolidone, polyvinylpolypyrrolidone (PVPP) is 

now the agent of choice (McMurrough, 1995) often in conjunction with silica hydrogels. 

Two types of PVPP are available. The first, single-use PVPP which is amicroized white 

powder with ahigh surface/weight ratio, is used in afilter aid bodyfeed dosing regimen. 

The second type is regenerable PVPP used either in impregnated sheets or within a 

horizontal leaf pressure vessel. These materials are regenerated by treatment with 1±2% 

sodium hydroxide, followed by washing and neutralization. According to Bamforth 

(1999), this last treatment is the cheapest. Siebert and Lynn (1998) compared polyphenol 

interactions with PVPP and haze-active protein. They concluded that the mechanisms by 

which haze-active polyphenols attach to PVPP and to haze-active proteins are similar but 

not identical. 

19.5 Viscosity 

Dynamic viscosity is defined as the resistance to shear flow within aliquid. Kinematic 

viscosity is ameasure of the time taken by aliquid to flow through an orifice under 

gravity. Both are measured by an international method using the time of flow in an 

Ostwald viscometer at 20.0ëC. At this temperature the dynamic viscosity of water is 

1.002cP (centipoises) and the kinematic viscosity is 1.00cS (centistokes). The SI unit for 

dynamic viscosity is the Pascal-second when 1cP ˆ0.001Pa.s. The EBC use a20% w/v 

sucrose solution as an additional standard (1.945 cP) but the Institute of Brewing, 

following BS188: 1957, use a 3.0% w/v solution of glycerol. Some values are: worts (SG 

1,030±1,100) 1.59±5.16, lager (SG 1,007) 1.45; and stout (SG 1,009) 1.96 cP. The method 

can be extended to measure the viscosity of liquid sugars and brewing syrups using a 

viscometer of the appropriate range. 

The viscosity of wort and beer is influenced by the macromolecules present, Sadosky 

et al. (2002) found that arabinoxylan, -glucans and dextrins all increased the viscosity of 

model solutions with the dextrins having the largest effect. 


19.6 Foam characteristics and head retention 

The vast literature on this subject has been reviewed by Bamforth (1985), Baxter and 

Hughes (2001) and Evans and Sheehan (2002). The European Brewery Convention 

(1999) have published the proceedings of a symposium on Beer Foam Quality, held in 

Amsterdam in 1998. Bamforth (1985) defined beer foam quality as a combination of its 

stability, quantity, lacing (adhesion or cling), whiteness, `creaminess', and strength. 

Further, he provided evidence (2000) that consumers differ, along regional and gender 

lines, in their requirements for the amount of beer foam, its stability and whether it laces 

the glass. In England, draught beers from Burton-on-Trent are served with little or no 

foam but in the north-east a rich creamy head that overflows the glass is expected. This 

can cause problems when the glass is a legal measure. However, the judgment of UK law 

is that the customers' preference should be catered for. Thus the head can be considered 

an integral part of the pint or the drinker can demand a full pint of liquid. 

Beer foams are colloidal systems comprising a continuous liquid phase and a 

discontinuous gas phase; the bulk density of the system approaches that of a gas rather 

than that of a liquid. Beer foam physics have been discussed by Ronteltap et al. (1991) 

and by Prins and van Marle (1999). Ronteltap et al. (1991) concluded that foam 

characteristics are determined by four key processes: bubble formation, drainage, 

coalescence, and disproportionation. Bubble formation requires a suitable site for 

nucleation. Bubbles may be generated in beer either by dispersal or condensation. The 

simplest dispersal system involves injection of a gas from a capillary. A spherical bubble 

forms at the tip of the capillary and will be released when its buoyancy is greater than the 

surface tension effects that hold the bubble to the tip of the capillary. Bubbles formed in 

this way are of similar size whereas those generated from a sinter will be more 

heterogeneous and more likely to undergo diproportionation. Homogeneous condensation 

occurs, for example, when a crown cork is removed from a bottle of beer at, say, 5 ëC, 

when rapid gas expansion may cause the temperature to drop locally as low as 36 ëC. 

Much more likely is heterogeneous condensation when the gas already present (usually 

air in the case of beer dispense) is expanded by diffusion of gas (either carbon dioxide or 

nitrogen) from the solution into the gas phase. As the bubble grows it experiences greater 

buoyancy until it breaks away from the nucleating surface leaving a small pocket of 

residual gas to begin the process again. So the nature and amount of gas in solution 

influence the beer foam (Fisher, et al., 1999) as does the angle of dispense. 

Etched glassware and small particles may provide nucleation sites but lose their 

effectiveness when totally wetted or not scrupulously clean. Agitation of beer can 

produce nucleation sites by cavitation, the instantaneous formation of a vacuum when the 

liquid separates from the vessel wall. Ultrasound can promote cavitation but often with 

uncontrollable gushing. Bubble formation involves an increase in the surface area of the 

liquid which is opposed by the surface tension of the liquid but bubbles remain stable 

because of materials of low surface tension within the bubble wall which have both 

hydrophobic and hydrophilic areas (e.g. proteins). Once a bubble leaves its nucleation site 

it will rise through the beer until it reaches the beer-air/foam interface. Surface active 

materials in the bubble wall may facilitate this movement. The nature of the gas within 

the bubble is important and the pressure within the bubble is inversely proportional to its 

diameter so gas will pass from smaller bubbles to larger ones (disproportionation). If the 

gas is soluble in the liquid film the passage of gas between the bubbles is faster. Thus 

foams containing oxygen and nitrogen produce smaller bubbles and a more stable foam 

than those containing carbon dioxide, which is more soluble in the liquid film. As the 


spherical bubble leaves the beer-foam interface it will be pushed upwards by younger 

bubbles arriving from below. As it moves to the top of the head the liquid between the 

individual bubbles starts to drain away, the fresh wet foam is converted to dry foam and 

finally only asolid network of bubble walls remains which may be deposited from the 

liquid phase on to the glass showing cling, lacing, or foam adhesion. To the consumer 

foam drying or drainage results only in amodest change in foam volume but bubble 

coalescence or disproportionation, with the formation of larger coarser bubbles, is amore 

obvious sign of foam destabilization. 

19.6.1 Methods of assessing foam characteristics 

Methods have been devised for measuring many foam parameters such as foamability, 

foam stability, foam drainage, cling, viscoelasticity, lateral diffusion, film thickness and 

bubblesizebutmostmeasurementsconcentrateontherateoffoamcollapseor,inversely, 

the duration of head retention or foam stability. It is obvious that when dispensing beer 

either draught or from abottle or can, one can make arough assessment of foam stability 

and it has been suggested that the views of apanel on this are closer to the consumers' 

view of foam behaviour than the tests discussed below. All measurements should be 

made in atemperature-controlled room as the collapse of beer foam is very sensitive to 

temperature. Scrupulously clean glassware should be used in all the tests, for example, 

that washed with 2% trisodium phosphate solution. 

Blom (1937) produced foam in atared separating funnel by passing carbon dioxide 

through aChamberland candle. The method is therefore applicable to worts and aqueous 

solutionscontainingfoam-activesubstances.Whenthefunnelisfulloffoam,thebeerisrun 

off and the funnel weighed. The beer is again run off and the residual foam weighed after 

intervals of one, two, three, and four minutes. Blom observed that the collapse of the foam 

was analogous to afirst-order chemical reaction and the foam stability (K) was given by: 

Kˆ 1 

t log 

x 

…a x† 

where tˆtime (min.), aînitial weight of the foam, and xˆfinal weight of the foam 

after time t. By substituting a/2 for xin this equation ,Kcan be calculated from the halflife 

of the foam tÝ. Blom found that beers with ahalf-life of 90shave excellent head 

retention but values less than 80sindicate poor head retention. Head retention values for 

arange of British beers, found using Blom's method, is given in Table 19.23. In Table 

19.24 changes in head retention throughout the brewing process are recorded. 

If air is passed continuously through a porous membrane into a liquid such as wort or 

beer there is a correlation between the volume of air (V) passing in time (t) and the 

average amount of foam produced ( ) so that: 

X ˆ t 

P is a measure of the life of a bubble in the foam. Ross and Clarke (1939) calculated that 

X t 

ˆ 

2:303 log 

b ‡ c 

2 

where t is the time in seconds of the stationary phase of the head, b is the volume (ml) of 

the beer formed in that period, and c is the volume of beer given by the residual head and: 


Table 19.23 Head retention (Blom) for production beers (Curtis et al, 1963) 

X ˆ1:44t1=2 

OG t1/2 (s) 

Pale ale 1,035 75 

Pale ale 1,055 77 

Brown ale 1,035 77 

Strong ale 1,080 73 

Sweet stout 1,046 89 

Table 19.24 Changes in head retention (Blom) through the brewing process (Curtis et al.,1963) 

OG 1,035 1,055 

Wort 115 ± 

Beer at rack 86 94 

Fined beer 82 88 

One day in bottle 78 ± 

One week in bottle ± 79 

One month in bottle 80 ± 

They concluded that was aproperty of the beer foam independent of the method used 

to produce it, the dimensions of the container and the temperature within alimit of 2ëC. 

However, this linear logarithmic decay breaks down when 95% of the foam has returned 

to the liquid phase. 

The American Society of Brewing Chemists have adopted amethod of Helm (1933), 

who worked at Carlsberg, in which beer is poured into acylindrical separating funnel 

(75mm o.d.) until the foam reaches the 800ml mark. After 30 seconds, all the beer that 

has separated is drawn off, the stopwatch started and after 200 seconds the beer that has 

separated in that period is drawn off (`b' ml) and the time `t' noted (225±230s). The 

remaining foam is then collapsed, with either isopropyl or butyl alcohol (2ml) delivered 

with afine pipette, and collected (`c' ml after deducting the volume of the defoaming 

agent). The sigma ( )value is then calculated according to Ross and Clarke's formula 

given above. 

The Institute of Brewing have adopted amodification of Rudin's (1957) method. The 

apparatus (Fig 19.6) consists of ajacketed foam tube (26±28mm dia.) andat least 350mm 

tallmountedoveraporosity3glasssintereddisc.Degassedbeerisaddedtothe10cmmark 

andfoamedwithcarbondioxidetothe325mmmark.Asthefoamcollapsesthetimetaken 

for the foam/beer boundary to traverse the distance between the 50mm and the 75mm 

marks gives ameasure of the half life of the foam. The logarithmic rectilinear collapse of 

foamsformedwithcarbondioxideisfourtimesfasterthanthatforbeersfoamedwithairor 

nitrogen so traces of air either in the CO2 used for foaming or introduced into beer by 

pouring can cause departures from aregular logarithmic decay. The Institute of Brewing 

alsoapprovethemeasurementoffoamstabilityusingaNIBEMmeter(Klopper,1973)(Fig 

19.7).Thefoamisdispensedintoastandardglassoverwhichthemeterheadwithacentral 

electrode and four shorter needle electrodes is mounted. When the needle electrodes are in 

the foam the conductivity between the longer and shorter electrodes switches off the 

servomotor. As the foam collapses the conductivity is broken the servomotor engages and 

lowers the electrodes until they touch the surface of the foam again After a `wait' period, 

the time for the foam to collapse over 10, 20, and 30 mm is measured for a beer containing 


Water 

outlet 

20°C 

Side 

arm (F) 

Trap (D) 

Jacketed 

foam tube 

Water 

inlet 

Sinter 

(b) 

Edwards 

needle 

valve (C) 

Flow 

indicator 

Needle valve (A) 

CO2 from cylinder 

fitted with reducing 

valve 

Fig. 19.6 Rudin head retention apparatus (Institute of Brewing, Methods of Analysis, 1997). 

> 3.4 g/l CO2. For a beer containing < 3.4 g/l CO2 the collapse is measured over 5, 10, and 

15 mm. The IoB Analysis Committee (Sharpe, 1997) found the precision of both the Rudin 

and the NIBEM methods was independent of the of the foam stability of the sample. 

However, the two methods ranked three beers differently, no doubt due to the different 

principles involved. A later model, the NIBEM-T meter, has protection against air 

movement and automatic temperature compensation. It gave better repeatability and 

reproducibility for the determination of foam stability in beer and has been accepted by the 

EBC Analysis Committee (Ferreira, 2003). 

Servomotor to 

raise and lower 

electrodes 

120 mm 

60 mm 

Electrodes 

Foam 

Beer 

Fig. 19.7 NIBEM foam stability apparatus (after Klopper, 1973). 


The American Society of Brewing Chemists also describe a foam flashing method for 

bottled beers. Under a positive pressure of 29 lb CO2 the beer is foamed through a 

0.79 mm (0.031 in.) orifice and 200 ml of foam are collected. The volume of the beer 

collapsed from the foam in 90 seconds is measured (B1) and the remaining foam is 

collapsed with isopropyl alcohol (2 ml). If the total volume of beer produced from the 

collapse of 200 ml of foam, after subtraction of 2 ml of isopropyl alcohol, B2 is then: 

Foam Value Units…FVU† ˆ 200…B2 B1† 

B2 

Several workers prefer to use a standard orifice since it has been observed that variations 

in the porosity of Chamberland candles and sintered glass discs can influence the 

observed head retention. Ault et al. (1967), Klopper (1973), and Jackson and Bamforth 

(1982) have described methods for measuring cling, lacing or foam adhesion. 

19.6.2 Beer components influencing head retention 

Pure liquids do not give stable foams and many investigations have been carried out to 

characterize the surface-active substances in beer responsible for good head retention but 

it is obvious, for example, from the use of alcohols to collapse beer foams, that beer 

contains both positive and negative foam factors. Undoubtedly the most important 

positive foam factors in beer are polypeptides. Evans and Sheehan (2002) showed that 

measurement of beer proteins by the Bradford Coomassie Blue dye binding assay, which 

only measures proteins with MW > 5,000, correlated well with Rudin Head Retention 

Values. Earlier, Narziss and RoÈttger (1973) had found a good correlation between foam 

stability (Ross and Clarke, 1939) and the concentration of nitrogenous material with MW 

> 12,000 found by gel filtration. Slack and Bamforth (1983) fractionated beer proteins by 

hydrophobic interaction chromatography on Octyl-Sepharose CL-4B and found that the 

most hydrophobic polypeptides were the most foam active. 

At least two barley proteins, protein Z and lipid transfer protein 1, survive the brewing 

process more or less intact. Protein Z (Mr c. 40 kDa), which accounts for 10±25% of the 

non-dialyzable protein in beer, can be resolved into two isoforms, Z4 (80%) and Z7 

(20%). Evans et al. (1999) used ELISA to measure the levels of protein Z4, protein Z7, 

BSZ7b, and lipid transfer protein 1 (LPT 1) in 25 different malts which were subjected to 

pilot or small-scale brewing trials. Regression analyses correlated the foam stability 

(Rudin) of the beer with protein (Coomassie blue), protein Z4 (ELISA), free amino 

nitrogen (FAN), -glucan, arabinoxylan, and viscosity. The levels of protein Z7 and LPT 

1 were not correlated with the HRV: LTP 1 influences the quantity of foam generated 

which is not measured by Rudin's method. By multiple linear regression analysis Evans 

et al. (1999) found that the level of malt protein Z4 and the wort -glucan level predicts 

72% of the variation in foam stability. 

Jegou et al. (2000) also found that LTP 1 and LTPb in barley grain survived in a 

modified form into beer. The modifications, which may influence foam promotion 

properties, include glycation with a number of hexose units and reduction of disulphide 

bonds. Some hordein-derived fragments are also enriched in beer foam (Evans and 

Sheehan, 2002). Varg et al. (1999) identified a 17 kDa protein in barley which is 

important for beer foam stability; this protein is partly modified into a more active form 

during mashing and/or wort boiling. 

High molecular weight non-starch polysaccharides such as -glucans and arabinoxylans 

increase beer viscosity thereby slowing the drainage of liquid from the foam and 


improving its stability. In the trial brews mentioned above (Evans and Sheehan, 2002) 

viscosity was significantly correlated with HRV and with the non-starchy polysaccharides 

inbeer. However,high levels of -glucan and arabinoxylanare negativelycorrelated 

with beer filtration rate so the improvement of foam quality by increasing the level of 

non-starchy polysaccharides is not advised. Evans and Sheehan (2002) found in atrial 

brewthatadditionofBioglucanase andBio-cellulaseduringmashingdegraded morethan 

75% of the -glucan and reduced the viscosity and HRV of the beer but not the lacing 

index. In contrast, Lusk et al. (2001) found that no foam enrichment or foam stabilization 

occurred with -glucan; it did not collect in the foam and treatment with -glucanase did 

not alter beer properties. 

The hop iso- -acids are concentrated into beer foam. For example, when lager 

containing 25±26ppm iso- -acids (50l) was foamed, the collapsed foam (2 l) contained 

93±120 ppm iso- -acids (Bishop et al., 1974). Of the individual iso- -acids, 

isohumulone and isoadhumulone are concentrated into the foam more than isocohumulone. 

Similarly, unhopped beer bittered (to 21.0 BU) with isohumulone had abetter head 

retention( ˆ132)thanthatbitteredwithisocohumulone( ˆ115)(Difforetal.,1978). 

The reduced iso- -acids stabilize foam more than their unsaturated parents. Baker (1990) 

added reduced iso- -acid preparations to unhopped beer and found that, in particular, 

foam collapse times of beers treated with tetrahydroiso- -acids and hexahydroiso- - 

acids increased significantly with each incremental addition. The foam produced with 

hexahydroiso- -acids was unnaturally dense but tetrahydroiso- -acids can be added to 

beer to improve the head retention. The amounts that can be added are limited by their 

more potent bitterness (Table 8.3) but the addition of 3±5mg/l of tetrahydroiso- -acids 

can effectively stabilize beer foam. 

Similarly, Weiss et al. (2002), using aNIBEM foam stability meter, found that -iso- 

-acids slightly improved the foam stability but the tetrahydroiso- -acids showed amore 

dramatic effect. In the presence of iso- -acids metal ions improve beer foaming. Hughes 

and Simpson (1995) proposed that metals such as Mn 2+ ,Al 3+ ,and Ni 2+ cross link the iso- 

-acids (Fig 19.8) to strengthen the bubble film. It is well known that polyphenols react 

with proteins but despite this they do not appear to have a large effect on foam stability. 

In contrast, melanoidins can form stable foams in the absence of protein as they slow the 

drainage of liquids from the foam. Finally, the pH of the beer influences foam stability; 

the lower the pH the more stable the foam. This may be due to the fact that at lower pH 

values the iso- -acids are less ionized and more hydrophobic. 

Traditionally, the only gas in beer was carbon dioxide but in 1964 Guinness 

introduced a draught stout with a mixture of carbon dioxide and nitrogen for dispense. 

The nitrogen gave the beer its characteristic creamy head since it produces smaller 

bubbles than CO2. Afterwards some ales were dispensed in a similar manner. Later, in the 

UK, small plastic devices called widgets, were put into beer cans along with liquid 

nitrogen. Nitrogen is not very soluble in beer and diffuses into the cavity in the widget. 

When the can is opened and the top pressure released, the nitrogen gas within the widget 

at the bottom of the can forces its way through the beer producing considerable amounts 

of creamy foam. A warm can may show drastic overfoaming! 

Alcohol-free beers usually have unstable foams and additions of small amounts of 

ethanol (c. 1%) enhances foamability and foam stability but this declines at higher 

concentrations. Ethanol is weakly surface active but at the concentration in beer (2.3±5.5%, 

0.5±1.2 M) it is likely to be challenging foam stability. Lienert (1955) showed that the 

straight chain higher alcohols, and their acetate esters, all destroy foam, the effect 

increasing with the length of the carbon chain. But the major potential foam inhibitors in 


O 

O 

HO 

R 

H 

O 

O 

M 2+ 

O 

O 

H 

(n – 2) + 

Fig. 19.8 Possible structure of iso- -acids chelated to ametal cation (Baxter and Hughes, 

2001, Fig. 2.2). Reproduced by permission of The Royal Society of Chemistry 

beerarelipids (Dickie etal.,2001).Lipidsinbeercanbeeitherintrinsic,derivedfrommalt 

andyeast lipids,or extrinsic, derived fromdirty glassware, fattyfoodsor cosmetics. Few of 

thelipidspresentinthegristarethoughttosurviveintobeersotheintrinsiclipidsarelikely 

to be derived from the yeast. For example, asignificant negative correlation was found 

between the HRV and the free fatty acid content of beer. Bamforth and Jackson (1983) 

foundthat,inparticular,decanoicanddodecanoicacidsandunsaturatedC 18acidsdisrupted 

beerlacing. Monopalmitin was moredisruptivethandipalmitin andtripalmitin (3ppm)had 

apositive effect on lacing. Roberts et al. (1978) observed that after the addition of lipid to 

beer there was a large drop in the HRV but on stirring the beer the HRV recovers 

approaching the original value. This is thought to be due to the presence of lipid binding 

proteins in the beer, which have yet to be characterized. Lipid binding proteins have been 

characterized in wheat (puroindolines) and barley (hordoindolines )(Mr c. 13kDa) but it is 

not known whether they survive the brewing process. 

19.6.3 Head retention and the brewing process 

Dilution experiments show that all-malt worts provide an excess of foam-positive 

substances so that dilution of the grist with nitrogen-free adjuncts or sugars should not 

affect the head retention significantly. The use of unmalted cereals such as wheat flour or 

flaked barley in the grist improves head retention in the beer but maize and rice have to 

be processed before use to reduce the level of lipid. Brewhouse procedures can also 

influence the level of lipid in the wort. The last runnings from the mash tun may be rich 

inmaltfat.Strainmasters(Nootertuns)givewortswithhighlipidcontentsbutifthemash 

worts are recycled the level of lipids is reduced (Chapter 6). Hop-boiling improves head 

retention unless so prolonged that all the foam stabilizing proteins are coagulated. The 

hop back is more efficient than the hop separator for eliminating lipids. 

During the brewing process the largest loss of head retention occurs during 

fermentation (Table 19.24). This is due to the loss of foam stabilizing material into the 

foam and yeast crop and to the formation of ethanol, higher alcohols, C6 C12 fatty acids 

and other negative foam factors. More foam stabilizing factors are lost on top-fermenting 

yeasts than bottom-fermenting strainssothe choice ofyeaststrainisimportant. Foam can 


R 

OH 

O 

O

O 

HO 

be controlled by the use of enclosed fermentation vessels and by recycling part of the 

fermenting wort and using it to sparge the yeast head and foam (Thompson et al., 1965). 

Yeast should not be allowed to autolyse in contact with the beer as the fatty acids 

liberated will destroy the foam stability. Fining, by means of isinglass, always improves 

head retention and the finished beer should always be handled gently. Finally, pipes, taps 

and glassware used in dispense must be kept scrupulously clean and well rinsed as traces 

of detergent can reduce head retention. 

As mentioned above the brewer can improve head retention by choice of malt rich in 

protein Z4 and by the addition of wheat flour to the grist. Foam quality can also be 

improved by the post-fermentation addition of tetrahydroiso- -acids. Propylene glycol 

alginate (PGA, Fig. 19.9) is apermitted additive to improve foam stability. It is formed 

by partial esterification (80±90%) of alginic acid with propylene oxide. Typical levels of 

application are c. 50mg/l. 

19.7 Gushing 

O 

O 

O 

HO 

OH 

HO 

O 

OH O 

CO2Na 

Gushing, wild, overfoaming or jumping beer, as it is variously called, is an undesirable 

quality in packaged beer (Gardner, 1973). Abeer is said to gush when, on releasing the 

overpressure, innumerable minute bubbles appear throughout the volume of the beer 

which rapidly expand and displace the contents of the bottle. In severe cases as much as 

three-quarters of the contents may be lost. Outbreaks of gushing may be of two types: (i) 

Sporadic or transitory or (ii) Epidemic or serious. Transitory gushing may be related to 

minor changes in the production process and is usually susceptible to specific cures. 

Epidemic outbreaks of gushing may affect several breweries in the same area or those 

using acommon source of raw material. Many factors have been implicated in gushing: 

since it is overfoaming an excess of the materials involved in head formation may be 

responsible.Simon(1998) thinksthat overcarbonation may be responsible for somecases 

of gushing but other workers say that, although overcarbonation and rough handling may 

stimulate gushing in sensitive beers, it will not induce it in normal beers. As noted above, 

beers carbonated with amixture of carbon dioxide and nitrogen (and awidget) are more 

likely to gush than those carbonated with CO2 alone. 

Krause (1936) suggested that prolonged shaking beats many microbubbles from the 

headspace into the beer and these microbubbles attract surface active constituents into 

their interfaces. When the gas dissolves the hydrophobic surface active shells may not 

dissolve and remain to act as nuclei for CO2 evolution. Gardner (1973) has reviewed the 

evidence in support of Krause's idea but the nature of the nuclei is still not understood. 

The nuclei have not been observed by electron microscopy and, by this technique, no 

difference could be seen between precipitates collected from gushing and non-gushing 

O 

O 

O 

O 

HO 

Fig. 19.9 Idealized structure of propylene glycol alginate (Baxter and Hughes, 2001, Fig. 2.5). 

Reproduced by permission of the Royal Society of Chemistry. 


HO 

OH 

O

eers. Turbid beers do not have an increased tendency to gush. A correlation has been 

found between surface viscosity and gushing. Substances known to promote gushing 

increase the viscosity which is reduced when antigushing agents are added. A high 

surface viscosity does not cause gushing but appears to correlate with the existence of 

stable nuclei. 

Transitory outbreaks of gushing have been observed due to the precipitation of 

microcrystals of calcium oxalate while other outbreaks have been associated with the 

presence of heavy metals, in particular, iron, nickel, tin and molybdenum. Such outbreaks 

can often be cured by the addition, if allowed, of EDTA to the beer; EDTA forms 

complexes with calcium and the heavy metals. Iron and nickel cause gushing only in the 

presence of iso- -acids but cobalt shows little tendency to cause gushing. Indeed, in one 

outbreak of gushing the addition of 0.2±1.0 ppm of cobalt dramatically reduced the 

incidence of gushing but in another it was without effect. In any case the addition of 

cobalt causes toxicity problems with heavy drinkers (Long, 1999). 

Severe outbreaks of gushing have been associated with the introduction of new 

season's malt especially when this has been made from barley harvested under wet 

conditions. Such barleys/malts are often contaminated with moulds especially Fusarium 

spp. Both Narziss et al. (1990) and Niessen et al. (1992) found that the most common 

fungal contaminant of barley, associated with gushing, was F. graminearum. Other 

species associated with gushing include F. avenaceum, F. culmorum, F. oxysporum, 

Alternaria alternata, and species of Penicillium, Mucor and Rhizopus Wheat infected 

with Microdochium nivale var. major also produces beverages with gushing properties. 

Scanning electron spectroscopy showed that with Fusarium species the fungal hyphae 

and mycelia tended to concentrate in the furrow and tip of the grain where they were not 

readily apparent to the naked eye. 

Similarly, in the USA Schwarz et al. (1996) found that malt samples infected with 

Fusarium species produced beers with a propensity to gush and that the levels of 

deoxynivalenol and ergosterol (Fusarium metabolites) strongly correlated with the amount 

of gushing. However, it is thought unlikely that deoxynivalenol and ergosterol promote 

gushing themselves but, rather, a polypeptide produced from the barley or malt by the 

micro-organism is the active agent. Obviously it is undesirable to use weathered barley for 

malting but, if necessary, the addition of formaldehyde to the steep liquor, if allowed, gives 

malt that does not produce a gushing beer. Other successful treatment of wild beer include 

the use of absorbents such as kaolin (1,000 g/l), bleaching earth (200 g/l), fullers' earth 

(200 g/l), and nylon (140 g/l) or to increase the hop rate. 

The observation that certain isomerized hop extracts provoked gushing led to a 

detailed examination of the gushing potential of individual components. The -acids and 

hulupones suppressed gushing while the iso- -acids and the humulinic acids had no 

influence on gushing behaviour. The most potent gushing agent found among the hop 

compounds was dehydrated humulinic acid [2-isovaleryl-4-(3-methyl-2-butenylidene) 

cyclopentane-1, 3-dione] (Laws and McGuinness, 1972). This occurred only rarely in 

isomerized extracts but 25 ppm provoked gushing in most commercial beers. Moir et al. 

(1991) prepared an isomerized extract which produced 38% gushing in beer. The unusual 

constituents responsible for the gushing activity were dimers of the iso- -acids formed by 

the action of oxygen on the iso- -acids in the presence of iron. Other hop compounds 

reported to provoke gushing include the abeo-iso- -acids and the hexahydroiso- -acids. 

Hop extracts contain variable amounts of fatty acids of which long chain saturated fatty 

acids promote gushing while unsaturated acids suppress gushing. 



ANDERSON, I. W., MOLZAHN, S. W., ROBERTS, N. B., BELLIA, J. and BIRCHALL, D. (1995) Proc. 25th Congr. 

Eur. Brew. Conv. Brussels, 1995, 543. 

ANDERSON, R. J., CLAPPERTON, J. F., CRABB, D. and HUDSON, J. R. (1975) J. Inst. Brewing, 81, 208. 

ANON (1960) Which, p. 167. 

ANON (1967) Which, p. 366. 

APPERSON, K., LEIPER, K. A., MCKEOWN, I. P. and BIRCH, D. J. S. (2002) J. Inst. Brewing, 108, 193. 

AULT, R. G., HUDSON, E. J., LINEHAN, A. J. and WOODWARD, J. D. (1967) J. Inst. Brewing, 73, 558. 

AÈ YRAÈ PAÈ AÈ , T., HOLMBERG, J. and SELLMANN-PERSSON, G (1961) Proc. 8th Congr. Eur. Brew. Conv. 

Vienna, 1961, 286. 

BAKER, C. D. (1989) J. Inst. Brewing, 95, 267. 

BAKER, C. D. and UPPERTON, A. M. (1992) J. Inst. Brewing, 98, 461. 

BAKER, C. W. (1990) Tech. Quart. MBAA, 27, 33. 

BAMFORTH, C. W. (1985) J. Inst. Brewing, 91, 370. 

BAMFORTH, C. W. (1999) J. Amer. Soc. Brew. Chem., 57, 81. 

BAMFORTH, C. W. (2000) J. Inst. Brewing, 106, 229. 

BAMFORTH, C. W. and JACKSON, G. (1983) Proc. 18th Congr. Eur. Brew. Conv. London, 1983, 331. 

BAXTER, E. D. and HUGHES, P. S. (2001) Beer: Quality, Safety and Nutritional Aspects. Cambridge, The 

Royal Society of Chemistry pp. xiv + 138. 

BECH, L. M., VAAG, P., HEINEMANN, B. and BREDDAM, K. (1995) Proc. 25th Congr Eur. Brew. Convn. 

Brussels, 1995, 561. 

BERNSTEIN, L. and LAUFER, L. (1977) J. Amer. Soc. Brew. Chem., 35, 21. 

BINNS, F., ENSOR, R. J. and MACPHERSON, A. L. (1978) J. Sci. Food Agric., 29, 71. 

BISHOP, L. R., WHITEAR, A. L. and INMAN, W. R. (1974) J. Inst. Brewing, 80, 68. 

BLOM, J. (1937) J. Inst. Brewing, 43, 251. 

BOTTOMLEY, R. A. and LINCOLN, G. J. (1958) J. Inst. Brewing, 64, 50, 53. 

BRENNER, M. W., BERNSTEIN, L., BLICK, S. R. and BLENKINSHIP, B. K.. (1965) Proc. A. M. Amer. Soc. Brew. 

Chem., 187. 


BUCKEE, G. K. (1995) Ferment, 8, 357. 


BUCKEE, G. K. and MUNDY, A. P. (1994) J. Inst. Brewing, 100, 247. 

BUCKEE, G. K., MORRIS, T. M. and BAILEY, T. P. (1986) J. Inst. Brewing, 92, 475. 

CANALES, A. M., DE BANCKS, N. M. and GARZA, T. L. I. (1970) Proc. A. M. Amer. Soc. Brew. Chem., 75. 

CANAS, B. J., HAVERY, D. C., ROBINSON, L. R., SULLIVAN, M. P., JOE, F. R. and DIACHENKO, G. W. (1989) J. 

Assoc. Off. Analyt. Chem., 72, 873. 

CARRINGTON, R., COLLETT, R. C., DUNKIN, I. R. and HALEK, G. (1972) J. Inst. Brewing, 78, 243. 

CERVERA, M. A., NAVARRO, A., MONTORO, R., CATALA, R. and YBAÂ NÄ EZ, N. (1989) J. Assoc. Off. Analyt. 

Chem., 72, 282. 

CHEN, E. C.-H. and VAN GHELUWE, G (1979) J. Amer. Soc. Brew. Chem., 37, 91. 

CLARKSON, S. P., ORMROD, I. H. L. and SHARPE, F. R. (1995) J. Inst. Brewing, 101, 191. 

COOTE, N. and KIRSOP, B. H. (1974) J. Inst. Brewing, 80, 474. 

CORNELL, J. (2002) J. Amer. Soc. Brew. Chem., 60, 215. 

CURTIS, N. S., CLARK, A. G. and OGIE, P. J. (1963) J. Inst. Brewing, 69, 30. 

DALE, C. J. (1990) Ferment, 3, 217. 

DALE, C. J. and LYDDIATT, A. (1994) J. Inst. Brewing, 100, 173. 

DALE, C. J. and YOUNG, T. W. (1989a) J. Inst. Brewing, 95, 35. 

DALE, C. J. and YOUNG, T. W. (1989b) J. Inst. Brewing, 95, 89. 

DICKIE, K. H., CANN, C., NORMAN, E. C., BAMFORTH, C. W. and MULLER, R. E. (2001) J. Amer. Soc. Brew. 

Chem., 59, 17. 

DIFFOR, D., LIKENS, S. T., REHBERGER, A. J. and BURKHARDT, R. J. (1978) J. Amer. Soc. Brew. Chem., 36, 63. 

DONHAUSER, S., WAGNER, D. and JACOB, F. (1987) Monats. fuÈr Brauwissenschaft, 40, 328. 

DRAWERT, F. and TRESSL, R. (1972) Tech. Quart. MBAA, 9, 72. 

DUPIRE, S. (1999) J. Inst. Brewing, 105, 265. 

ENEVOLDSEN, B. S. and SCHMIDT, F. (1974) J. Inst. Brewing, 80, 520. 

EUROPEAN BREWERY CONVENTION (1999) Monograph 27. Symposium on Beer Foam Quality, 

Amsterdam, October 1998. NuÈrnberg: Hans Carl, p. 220. 

EVANS, D. E. and HEJGAARD, J. (1999) J. Inst. Brewing, 105, 159. 

EVANS, D. E. and SHEEHAN, M. C. (2002) J. Amer. Soc. Brew. Chem., 60, 47. 

EVANS, D. E., SHEEHAN, M. C. and STEWART, D. C. (1999) J. Inst. Brewing, 105, 171. 

FARRE, R., GIMENO, M. J. and LAGARDA, M. J. (1987) J. Inst. Brewing, 93, 394. 

FERREIRA, A. A. (2003) J. Inst. Brewing, 109, 400. 

FISHER, S., HAUSER, G. and SOMMER, K (1999) Eur. Brew. Convn. Monograph No. 27 Beer Foam Quality, 



FOOD LABELLING REGULATIONS, THE (1980) SI 1849: 1980. Schedule 5. 

FREESTON, M. J. and BAKER, C. D. (1993) J. Inst. Brewing, 99, 377. 

GARDNER, R. J. (1973) J. Inst. Brewing, 79, 275. 

GJERTSEN, P. (1955) Proc. 5th Cong, Eur. Brew. Convn., Baden-Baden, p. 37. 

GLASER, G. (2002) Modern Brewery Age, 53 (12), 6, 32. 

GMELCH, F., MAIER, J., ROSSBAUER, G. ZWACK, F. and NAST, D. (1989) Brauwelt, 129 (45), 2156, 2158± 

2162. 

GREENHOFF, K. and WHEELER, R. E. (1981) J. Inst. Brewing, 87, 35. 

GREENSHIELDS, R. N. (1974) J. Sci. Food Agric., 25, 1307. 

HARDING, R. J., NURSTEN, H. E. and WREN, J. J. (1977) J. Sci. Food Agric., 28, 225. 

HARRIS, G. (1965) J. Inst. Brewing, 71, 292. 

HARRISON, G. A. F., BYRNE, W. J. and COLLINS, E. (1965a) Proc. 10th Congr. Eur. Brew. Convn. Stockholm, 

1965, p. 352. 

HARRISON, G. A. F., BYRNE, W. J. and COLLINS, E. (1965b) J. Inst. Brewing, 71, 336. 

HEGARTY, P. K., PARSONS, R., BAMFORTH, C. W. and MOLZAHN, S. W. (1995) Proc. 25th Congr. Eur. Brew. 

Convn. Brussels, 1995, 515. 

HEJGAARD, J. and KAERSGAARD, P. (1983) J. Inst. Brewing, 89, 402. 

HELM, E. (1933) Wochensch. Brau., 50, 241. 

HM CUSTOMS AND EXCISE (1997) Notice 226. Beer Duty (including Update 1, December 2000) 

HUDSON, J. R. (1955) J. Inst. Brewing, 61, 127. 

HUDSON, J. R. (1969) J. Inst Brewing, 75, 164. 

HUGHES, P. S. and SIMPSON, W. J. (1995) Cerevisia Biotechnol., 20, 35. 

HYSERT, D. W., MORRISON, N. M. and JAMIESON, A. M. (1979) J. Amer. Soc, Brew. Chem. 37, 30. 

ILETT, D. S., BURKE, S. and SIMPSON, W. J. (1996) J. Sci. Food Agric., 70, 337. 

ISHIBASHI, Y., TERANO, Y., FUKUI, N., HONBOU, N., KAKUI, T., KAWASAKI, S. and NAKATANI, K. (1996) J. 

Amer. Soc. Brew. Chem., 54, 177. 

ISHIHARA, S. (1964) Tests for Colour Blindness, Kanehara, Shuppan, Tokyo, Japan. 

JACKSON, G. and BAMFORTH, C. W. (1982) J. Inst. Brewing, 88, 378. 

JEGOU, S., DOULIEZ, J. P., MOLLE, D., BOIVIN, P. and MARION, D (2000) J. Agric. Food Chem., 48, 5023. 

JOHNSON, B. A. and JOHANSSON, C.-G. (1999) J. Inst. Brewing, 105, 360. 

JURADO, J. (2001a) The Brewer International, 1 (10), 41. 

JURADO, J. (2001b) The Brewer International, 1(12), 39. 

JURADO, J. (2002a) The Brewer International, 2 (2), 20. 

JURADO, J. (2002b) The Brewer International, 2 (4), 30. 

JURADO, J. (2002c) The Brewer International, 2 (6), 29. 

JURADO, J. (2002d) The Brewer International, 2 (9), 33. 

JURADO, J. (2002e) The Brewer International, 2 (12), 21. 

KAERSGAARD, P. and HEJGAARD, J. (1979) J. Inst. Brewing, 85, 103. 

KALACÏ , P. and KRIÂZÏ EK, M. (2003) J. Inst. Brewing, 109, 123. 

KIENINGER, H. (1978) Brauwelt, 18, 616. 

KIENINGER, H. and BIRKOVA, V. (1975) Brauwelt, 115, 1250. 

KLOPPER, W. J. (1973) Proc. 14th Congr. Eur. Brew. Convn., Salzburg, p. 363. 

KLOPPER, W. J., ANGELINO, S. A. G. F., TUNING, B. and VERMEIRE, H. A. (1986) J. Inst. Brewing, 92, 225. 

KOIKE, K., HASHIMOTO, N., KITAMI, H. and OKEDA, K. (1972) Reports of the Research Laboratories of the 

Kirin Brewery Co. Ltd., 15, 25. 

KONDO, H., YOMO, H., FURUKUBO, S., FUKUI, N., NAKATANI, K. and KAWASAKI, Y. (1999) J. Inst. Brewing, 

105, 293. 

KRAUSE, B. (1936) Svenska BryggarefoÈrenings MaÊnadsblad, 51, 221. 

LAWS, D. R. J. and MCGUINNESS, J. D. (1972) J. Inst. Brewing, 78, 302. 

LERMUSIEAU, G., BULENS, M. and COLLIN, S. (2001) J. Agric. Food Chem., 49, 3867. 

LIENERT, H. (1955) Proc. 5th Congr. Eur. Brew. Convn., Baden-Baden, p. 282. 

LONG, D. G. (1999) J. Inst. Brewing, 105, 79. 

LUSK, L. T., GOLSTEIN, H. and RYDER, D. (1995) J. Amer. Soc. Brew. Chem., 53, 93. 

LUSK, L. T., DUNCOMBE, G. R., KAYE, S. B., NAVARRO, A. and RYDER, D. (2001) J. Amer. Soc. Brew. Chem., 

59, 183. 

MACKIE, A. E. and SLAUGHTER, J. C. (2000) J. Inst. Brewing, 106, 209. 

MANDL, B. and PIENDL, A (1971) Proc. 13th Congr. Eur. Brew. Convn., Estoril, p. 343. 

MARTIN, P. A. (1982) J. Inst. Brewing, 88, 320. 

MCFARLANE, W. D. and THOMPSON, K. D. (1964) J. Inst. Brewing, 70, 467. 

MCMURROUGH, I. (1995) Ferment, 8, 39. 

MCMURROUGH, I., MADIGAN, D., DONNELLY, D., HURLEY, J., DOYLE, A. M., HENNIGAN, G., MCNULTY, N. and 

SMYTHE, M. R. (1996) J. Inst. Brewing, 102, 327. 

MEILGAARD, M. C. (1975) Tech. Quart. MBAA, 12, 151. 

MOIR, M. (1994) EBC Monograph XXII. Symposium on Hops, Zoeterwoude, p. 165. 

MOIR, M., DOBSON, G. and SEATON, J. C. (1991) Proc. 23rd Congr. Eur. Brew. Convn., Lisbon, p. 225. 


MORGAN, K. (1965) J. Inst. Brewing, 71, 166. 

MORRIS, T. M. (1987) J. Inst. Brewing, 93, 13. 

MUNDY, A. P. (1991) J. Inst. Brewing, 97, 45. 

MUNDY, A. P. (1996) J. Inst. Brewing, 102, 69. 

MUNDY, A. P. and BOLEY, N. (1999) J. Inst. Brewing, 105, 75. 

NARZISS, L. and ROÈ TTGER, W. (1973) Brauwissenschaft, 26, 261. 

NARZISS, L., MIEDANER, H. and BOURJA, T. (1979) Brauwissenschaft, 32, 62. 

NARZISS, L., BACK, W., REICHENEDER, E., SIMON, A. and GRANDL, R. (1990) Monatsschrift fuÈr 

Brauwissenschaft, 43, 296. 

NIESSEN, L., DONHAUSER, S., WIEDENEDER, A., GEIGER, E. and VOGEL, H. (1992) Brauwelt, 132 (16/17), 

702. 

OTTER, G. E. and TAYLOR, L. (1967) J. Inst. Brewing, 73, 570. 

OTTER, G. E. and TAYLOR, L. (1971) J. Inst. Brewing, 77, 467. 

OWADES, J. L. and JACEVAC, J. (1959) Proc. Ann. Meet. Amer. Soc. Brew. Chem., p. 18. 

PALAMAND, S. R. and GRIGSBY, J. H. (1974) Brewers Digest, Sept. pp. 58, 90. 

PALAMAND, S. R., HARDWICK, W. A. and MARKL, K. S. (1969) Proc. Ann. Meet. Amer. Soc. Brew. Chem., 

p. 54. 

PAUL, A. A. and SOUTHGATE, D. A. T. (eds) (1978) McCance and Widowsons' The Composition of Foods. 

4th edn HMSO, London. 

PEPPARD, T. L. (1985) J. Inst. Brewing, 91, 364. 

PEPPARD, T. L. and HALSEY, S. A. (1981) J. Inst. Brewing, 87, 85. 

PEPPARD, T. L. and HALSEY, S. A. (1982) J. Inst. Brewing, 88, 309. 

PERPEÁ TE, P. and COLLIN, S. (1999) J. Agric. Food Chem., 47, 2374. 

PORTNO, A. D. (1966) J. Inst. Brewing, 72, 193, 458. 

POSTEL, W. (1976) Brauwissenschaft, 29, 39. 

POSTEL, W., DRAWERT, F. and GUÈ VENC, U. (1972a) Brauwissenschaft, 25, 341. 

POSTEL, W., DRAWERT, F. and GUÈ VENC, U. (1972b) Brauwissenschaft, 25, 391. 

POSTEL, W., DRAWERT, F. and GUÈ VENC, U. (1973) Brauwissenschaft, 26, 46. 

POSTEL, W., DRAWERT, F. and GUÈ VENC, U. (1974) Brauwissenschaft, 27, 11. 

POSTEL, W., GORG, A. and GUÈ VENC, U. (1976) Brauwissenschaft, 29, 132. 

POSTEL, W., GORG, A., DRAWERT, F. and GUÈ VENC, U. (1975) Brauwissenschaft, 28, 301. 

POSTEL, W., MEIER, B. and MARKERT, R. (1983) Monatschrift fuÈr Brauwissenschaft, 36, 360. 

PRINS, A. and VAN MARLE, J. T. (1999) EBC Monograph 27. Symposium on Beer Foam Quality, 


QURESHI, A. A., BURGER, W. C. and PRENTICE, N. (1979) J. Amer. Soc. Brew. Chem. 37, 153. 

ROBBERECHT, H., VAN SCHOOR, O. and DEELSTRA, H. (1984) J. Food Sci., 49, 300. 

ROBERTS, R. T., KEENEY, P. J. and WAINWRIGHT, T. (1978) J. Inst. Brewing, 84, 9. 

RONTELTAP, A., HOLLEMANS, M., BISPERINK, C. G. J. and PRINS, A. R. (1991) Tech. Quart. MBAA, 28, 25. 

ROSENDAL, I. AND SCHMIDT, F. (1987) J. Inst. Brewing, 93, 373. 

ROSS, S. and CLARKE, C. L. (1939) Wallerstein Labs. Commun., 6, 46. 

RUDIN, A. D. (1957) J. Inst. Brewing, 62, 506. 

SADOSKY, P., SCHWATZ, P. B. and HORSLEY, R. D. (2002) J. Amer. Soc. Brew. Chem., 60, 153. 

SCHROPP, P., BRUDER, T. and FORSTNER, A. (2002) Monatsschrift fuÈr Brauwissenschaft, 55, 212. 

SCHWARZ, P. B. and HAN, J.-Y. (1995) J. Amer. Soc. Brew. Chem., 53, 157. 

SCHWARZ, P. B., BEATTIE, S. and CASPER, H. H. (1996) J. Inst. Brewing, 102, 93. 

SHAH, S. K. (1975) J. Inst. Brewing, 81, 293. 

SHARPE, F. R. (1997) J. Inst. Brewing, 103, 277. 

SHARPE, F. R., GARVEY, T. B. and PYNE, N. S. (1992) J. Inst. Brewing, 98, 321. 

SHEEHAN, M. C. and SKERRITT, J. H. (1997) J. Inst. Brewing, 103, 297. 

SHIMIZU, C., NAKAMURA, Y., MIYAI, K., ARAKI, S., TAKASHIO, M. and SHINOTSUKA, K. (2001) J. Amer. Soc. 

Brew. Chem., 59, 51. 

SIEBERT, K. J. (1999) J. Agric. Food Chem., 47, 353. 

SIEBERT, K. J. and LYNN, P. Y. (1997) J. Amer. Soc. Brew. Chem., 55, 73. 

SIEBERT, K. J. and LYNN, P. Y. (1998) J. Amer. Soc. Brew. Chem., 56, 24. 

SIEBERT, K. J., TROUKHANOVA, N. V. and LYNN, P. Y. (1996) J. Agric. Food Chem., 44, 80. 

SILBEREISEN, K. and BIELIG, K. (1961) Proc. 8th Congr. Eur. Brew. Convn., Vienna, p. 421. 

SIMON, R. (1998) Brauwelt, 138, 1244. 

SIMPSON, W. J. and HUGHES, P. S. (1994) Cerevisia Biotechnol., 19 (3), 39. 

SINCLAIR, A., HALL, R. D., THORBURN-BURNS, D. and HAYES, W. P. (1970) J. Sci. Food Agric., 21, 468. 

SLACK, P. T. and BAMFORTH, C. W. (1983) J. Inst. Brewing, 89, 397. 

SLAUGHTER, J .C. and UVGARD, A. R. A. (1971) J. Inst. Brewing, 77, 446. 

SLAUGHTER, J. C. and UVGARD, A. R. A. (1972) J. Inst. Brewing, 78, 322. 

SMEDLEY, S. M. (1992) J. Inst. Brewing, 98, 497. 

SMEDLEY, S. M. (1995) J. Inst. Brewing, 101, 195. 

SéRENSEN, S. B. and OTTESEN, M. (1978) Carlsberg Res. Commun., 43, 133. 


THALAKER, R. and KALTWASSER, I. (1978) Monatsschrift fuÈr Brauerei, 31, 20. 

THOMPSON, C. C., CURTIS, N. S., GOUGH, P. E. and RALPHM D. J. (1965) Proc. 10th Congr. Eur. Brew, Conv., 

Stockholm, p. 305. 

THORNE, R. S. W. and NANNESTED, I. (1960) J. Inst. Brewing, 66, 388. 

THORPE, E. and BROWN, H. T. (1914) J. Inst. Brewing, 20, 569. 

TRESSL, R., KOSSA, T., RENNER, R. and KOÈ PPLER, H. (1975) Monatsschrift fuÈr Brauerei, 28, 109. 

TRESSL, R., RENNER, R., KOSSA, T. and KOÈ PPLER, H. (1977) Proc. 16th Congr. Eur. Brew. Convn., 


TRESSL, R., FRIESE, L., FENDSACK, F. and KOÈ PPLER, H. (1978) J. Agric. Food Chem., 26, 1422. 

VAAG, P., BECH, L. M., CAMERON-MILLS, V. and SVENDSEN, I. (1999) Proc. 27th Congr. Eur. Brew. Convn., 

Cannes, p. 157. 

VAN OEVELEN, D., DE L'ESCAILLE, F. and VERACHTERT, H. (1976) J. Inst. Brewing, 82, 322. 

WARNAKULASURIYA, S., HARRIS, C., GELBIER, S., KEATING, J. and PETERS, T. (2002) Clin. Chim. Acta, 320, 

1. 

WEISS, A., SCHOÈ NBERGER, CH., MITTER, W., BIENDL, M., BACK, W. and KROTTENTHALER, M. (2002) J. Inst. 

Brewing, 108, 236. 


WILLIAMS, K. M., FOX, P. and MARSHALL, T. (1995) J. Inst. Brewing, 101, 365. 

YANG, J.-I. and SIEBERT, K. J. (2001) J. Amer. Soc. Brew. Chem., 59, 172. 

YBAÂ NÄ EZ, N., NAVARRO, A. and MONTORO, R. (1989) J. Inst. Brewing, 95, 257. 


20 

Beer flavour and sensory assessment 


The final arbiter of beer quality is the palate of the consumer and this can show wide 

variation between individuals, between geographical areas, and even from one time to 

another. Quality is defined as `degree of excellence, relative nature, or kind, or character' 

and accordingly the brewer refers to the many varieties of ale, beer, stout, and lager 

which he brews to satisfy varied demand as different qualities. When the customer has 

chosen the quality he wishes to drink he demands that his beverage shall have the `degree 

of excellence' which he expects and this shall not change from day to day. Much of the 

brewers' art is therefore concerned with quality control, with producing a constant 

product from variable raw materials by a biological process (Hough, 1990). Today the 

concept of quality assurance is more important whereby each stage is monitored before 

the next is allowed and the raw materials and any additives can be traced back to their 

source (see EBC Monograph No. 26 Symposium on Quality Issues & HACCP (Hazard 

Analysis Critical Control Points), 1997). 

The enjoyment of a glass of beer may be received by many senses. Smythe et al. 

(2002) have shown the impact of the appearance of beer on its perception and the 

parameters discussed in the last chapter: alcoholic content, nutritive value, colour, 

clarity, usually, the absence of haze, the formation and retention of a good head of 

foam, and the absence of gushing, all contribute to the enjoyment but it is the flavour, 

the taste and aroma, which really determine the acceptability and drinkability of the 

beer. Originally, it was the palate of the head brewer that decided if the beer was 

acceptable but later this responsibility was transferred to a tasting panel and as tasting 

panels became more sophisticated the science of sensory analysis came into being 

(Amerine et al., 1965). Analytica-EBC and the ASBC have agreed International 

methods for sensory analysis and the Institute of Brewing have published a Sensory 

Analysis Manual (Institute of Brewing, 1995). More recent advances have involved the 

use of electronic `noses' or sensors (Torline et al., 1999; Given and Parades, 2002). 

Doty (1995) has edited a Handbook of Olfaction and Gustation and Acree and Teranishi 

have edited a handbook of Flavor Science (1993). 


20.2 Flavour ± taste and odour 

According to the EBC-ASBC definitions, flavour is the combination of olfactory and 

gustatory attributes perceived during tasting, including tactile, thermal, pain and kinesthetic 

effects. Kinesthetic sense is the deep pressure sense, or proprioception. Somesthetic sense is 

the tactile sense, or skin-feel, both are part of the sense of touch. Tactile senses refer more 

to solid foodstuffs but in beer are related to what is called `palate fullness', `body', or 

`mouth-feel' (Langstaff and Lewis, 1993). EBC/ASBC define mouth-feel as the tactile 

sensations perceived at the lining of the mouth, including the tongue and teeth. With regard 

to beer, taste and odour are the most important aspects of flavour. 

Taste is the sensory attribute resulting from stimulation of the gustatory receptors in 

the oral cavity by certain soluble substances. Two American Chemical Society symposia 

have discussed taste chemistry (Boudreau, 1979; Given and Parades, 2002) and Breslin 

(2001) has provided a review. Two types of chemicoreceptor are recognized; free nerve 

endings, which occur throughout the oral cavity, and taste buds. The free nerve endings 

possess no recognizable receptors and are responsible for the perception of pungency and 

astringency. Taste buds are neural complexes of 25±50 specialized cells which occur in 

localized areas of the oral cavity. On the tongue they occur on protuberances called 

papillae. Four types of papillae are recognized: the filiform papillae have no taste buds 

and the foliate papillae, which occur in folds on the sides of the back of the tongue, are 

not well developed in man. More important are the 13±400 mushroom-like fungiform 

papillae on the tip and sides of the tongue and the 6±15 large (circum)vallate papillae at 

the back of the tongue (Fig. 20.1) (See Amerine et al., 1965, for microphotographs). Taste 

buds are pear shaped (40±70 m) connected to the oral cavity via a narrow pore (2 m). 

At the top of the taste bud microvilli (0.1±0.2 m diam. 1±2 m long) are situated in the 

pore and these are probably the first point of contact with the tastants. The taste stimuli 

apparently do not penetrate the receptor membrane but interact at the outer surface. 

Taste buds 

Vallate papillae 

Foliate papillae 

Fungiform papillae 

Pore 

Nerve 

Sensory 

cells 

Fig. 20.1 Location of some oral chemosensory receptor systems. Taste buds (schematic, upper 

right) are found on specialized papillae on the tongue and scattered on the palate and posterior 

oral structures. Free nerve endings are found on all oral surfaces (After Boudreau, 1979). 


Bitter 

Sour Salt Sweet 

Fig. 20.2 Areas of the human tongue where the four tastes are most easily sensed. All four 

tastes are perceived, but less readily, over the central area. 

The taste cells are secondary receptor cells as they have no axions of their own but are 

connected by basal synapses to taste fibres running to the central nervous system. Three 

nerves are involved; the nervus glossopharyngeus (IX) for the back third of the tongue 

including the circumvallate papillae, the nervus vagus (X) for the throat and the larynx, 

and the chorda tympani (CT) part of the nervus facialus (VII) for the front two-thirds of 

the tongue with the fungiform papillae. It is estimated that the circumvallate papillae 

contain 1,000±1,500 taste buds and the fungiform papillae 300±400 (Van der Heijden, 

1993). 

Althoughthefourbasictastes,sweet,sour,salt,andbitterareperceivedthroughoutthe 

oral cavity, they are perceived more strongly in specific areas (Fig. 20.2). Inspection of 

this figure shows that the bitter taste of beer can only be evaluated satisfactorily if the 

beerisswallowedandallowed toflow overthecircumvallatetastebudsatthebackofthe 

tongue. The sweet taste, perceived by taste buds on the fungiform papillae, has received 

the most study (Van der Heijden, 1993). The transduction of the sweet taste appears to 

involve specific membrane receptors and Teeter and Gold (1988) have proposed the 

transduction pathway shown in Fig. 20.3. Akabas et al. (1988) concluded that 

transduction of bitter taste may occur via a receptor-second messenger mechanism 

leading to neurotransmitter release and may not involve depolarization-mediated calcium 

entry. The bitter principle denatonium chloride (Fig. 20.4) is apotent blocker of outward 

potassium currents in taste cells and causes asecond messenger release of Ca 2+ from 

intracellular stores in asubset of taste cells. 

Many substances are known to taste bitter (Rouseff, 1990) and in view of their diverse 

structures (Fig. 20.4) Delwiche et al. (2001) have proposed there must be multiple 

receptor/transduction mechanisms. They found that the ratings and rankings of 26 

subjects placed bitter substances in two general clusters: (i) urea, phenylalanine, 

tryptophan, and epicatechin, and (ii) quinine (20.1), caffeine (20.2), sucrose octa-acetate 

(20.3), denatonium benzoate (20.5), tetralone Õ (tetrahydroiso- -acids, 8.47) and 

magnesium sulphate. Neither of these groups included propylthiouracil (PROP) (20.4) 

to which tasters show wide variations in sensitivity. Approximately 25% of tasters find 

this antithyroid drug extremely bitter (`supertasters'), 50% find it bitter (`tasters') but 

25%are non-tasters. It is thought that `supertasters' perceive all tastes, not just bitterness, 

more intensely. In `supertasters' the fungiform papillae on the tongue are denser (>35/ 

38.5mm 2 )comparedwithanaverageof15±35/38.5mm 2 ;non-tastershave

Chemical stimulus 

activates GTP-binding protein 

stimulates adenylate cyclase 

increases intracellular cAMP (or cGMP) 

closes apical K + channels 

depolarizes taste cell 

opens voltage-dependent Ca 2+ channels 

allows influx of Ca 2+ near synapses 

causes neurotransmitter release at synapses 

Fig. 20.3 Transduction pathway for the sweet taste (Teeter and Gold, 1988). 

CH3O 

N 

CH(OH) 

CH2OAc 

AcO·CH2 

O 

O 

OAc 

AcO 

OAc 

O 

OAc 

N 

CH 

AcO 

CH2OAc 

CH3 

CH3 

CH2 

C2H5 

C2H5 

H3C 

O 

Pr n 

N 

O 

N 

CH3 

(20.1) Quinine (20.2) Caffeine 

H 

N 

O 

CH3 

(20.3) Sucrose octaacetate (20.4) Propylthiouracil 

NH·CO·CH2·N + ·CH2 

(20.5) Denatonium benzoate 

Fig. 20.4 Bitter compounds. 


N 

NH 

N 

S 

C6H5·CO2 –

men (Goode, 2003). Delwiche et al. (2001) found that although PROP sensitive and 

PROP insensitive tasters rated the above chemicals differently, they ranked them in a 

similar order. Montmayeur and Matsunami (2002) have discussed receptors for bitter and 

sweet tastes. The concentration-response curve for most tastant-receptor reactions is 

sigmoid in shape (Breslin, 2001). 

Other basic tastes include salty which is produced by relatively high concentrations of 

inorganic ions, in particular Na + , K + , and Li + , on the fungiform papillae but it is seldom a 

dominant taste in beer. The sour taste is evoked by various Bronsted acids mainly on the 

foliate papillae on the sides of the tongue. In most beers it is regarded as an off-flavour but is 

an important character in some Belgian beers produced by spontaneous fermentation. 

Pleasant is associated with the small fibre geniculate ganglion system and is evoked by 

lactones and similar carbon-oxygen compounds. Ùmani' is a Japanese word (meaning 

`deliciousness') used to describe the sensation elicited by the amino acid monosodium 

glutamate and the nucleotides sodium inosinate and sodium guanylate. The metallic 

sensation is produced by certain salts such as silver nitrate and by oct-1-en-3-one. As 

mentioned above pungency and astringency are sensations produced at the free nerve 

endings. 

Odour is more complicated than taste (see, for example, Ohloff, 1994). Orthonasal 

olfaction occurs when an odour is sniffed through the nostrils into the nasal cavity where 

the receptors are located on the olfactory epithelium in the upper respiratory passages 

(Fig. 20.5). In man the olfactory epithelium occupies 2±4 cm 2 and contains about 9 

million neurons; it is more extensive in other animals. The axions from these cells, many 

of which cannot be seen with a light microscope, are grouped together in bundles and 

pass through the cribiform plate into the olfactory bulb, where they terminate in small 

bodies known as glomeruli. From the glomeruli, mitral cells pass directly into the 

olfactory lobe of the brain. No other stimuli are received by the brain in such a direct 

manner. Retronasal olfaction occurs when odours released during eating or drinking are 

forced behind the palate into the nasal cavity. However, it is likely that the two forms of 

olfactory input are analysed in different parts of the brain. In beer, and other beverages 

and foodstuffs, the strength of the odour impression is partly governed by the volatility of 

the molecules from water, i.e., by the air-water partition coefficient. 

Frontal sinus 

Openings to: 

frontal sinus 

maxillary sinus 

naso-lacrimal duct 

Air passages 

OLFACTORY 

BULB 

Middle 

turbinate 

Inferior 

turbinate 

OLFACTORY 

CENTRES 

INCUS 

Superior 

turbinate 

Lined with 

COLUMNAR CILIATED 

EPITHELIUM 

Opening of 

eustachian tube 

To respiratory 

system 

Fig. 20.5 Vertical section of the nasal region of the head. 


Table 20.1 Olfactometric properties of eight primary odourants (Amoore, 1991) 

It is thought that the odour perceived is made up from a number of primary odours. Many of 

these primary odours have been detected from persons with specific anosmias, `smell 

blindness' which is the olfactory analogue of colour blindness (Amoore, 1991). In the 

examples studied so far 3±47% of the population have shown specific anosmia for some 

compound. Conversely, some people exhibit hyperosmia, when their sensitivity to certain 

odours may be 1,000-fold greater than normal. Thirty-six per cent of a population had a 

specific anosmia for the malty isobutyraldehyde. These people could still detect 

isobutyraldehyde after 16 twofold dilution steps whereas the normal population could detect 

this compound after 24 dilution steps (a 500-fold deficit of sensitivity for the anosmics). When 

the same panels investigated isobutyl alcohol there was some overlap between the panels and 

the anosmic defect was only 4.1 steps In contrast with isobutyl isobutyrate, the thresholds of 

the two groups overlapped. Eight primary odorants, which have been examined in this way, are 

listed in Table 20.1 but 76 compounds have been observed for which specific anosmia has been 

reported. Several systems of odour classification have been proposed containing 4±44 classes 

and these have been reviewed by Amoore (1991) together with specific anosmia analyses. 

Goodenough (1998) has reviewed the molecular biology of olfactory perception. 

Taste and odour can be perceived separately but more often than not the two senses are 

integrated to produce the sensation of flavour. The measurement of taste, odour, or 

flavour intensity is the subject of at least two different approaches which use different 

mathematics (Meilgaard, 1975). Those working with strong flavours are concerned with 

suprathreshold effects and describe the perceived intensity R as a power factor n of the 

concentration S so that: 

R ˆ constant S n 

For example, with butanol in air the equation becomes: 

R ˆ 0:261 S 0:66 

On the other hand those working with more delicate flavours such as food, vegetables, 

whisky, wine and beer have assumed that the perceived intensity R is proportional to the 

concentration S and inversely proportional to the threshold concentration T so that: 

R ˆ …constant† S=T 

Normal threshold 

Primary odorant Primary In air l/l In water Anosmic Anosmic 

odour (v/v) mg/l (w/v) occurrence defect 

(%) factor 

Isovaleric acid Sweaty 0.0010 0.12 3 42 

1-Pyrroline Spermous 0.0018 0.020 16 39 

Trimethylamine Fishy 0.0010 0.00047 6 830 

Isobutyraldehyde Malty 0.0050 0.0018 36 340 

5 -Androst-16-en-3-one Urinous 0.00019 0.00018 47 770 

!-Pentadecalactone Musky 0.018 0.0018 12 13 

l-Carvone Minty 0.0056 0.04 8 13 

1,8-Cineole Camphorous 0.011 0.020 33 56 

The power factor n is assumed to be 1.00 and the constant is often omitted so that R is 

measured as S/T. This ratio has been given different names including `flavour units' (see 

p. 285) (Meilgaard, 1975). The power function n in the equation: 


R ˆ …S=T† n 

has been estimated for ethanol (n ˆ 1.54), dimethyl sulphide (n ˆ 1.12), diacetyl (n ˆ 

0.89) and isoamyl acetate (n ˆ 0.82) (Meilgaard and Reid, 1979). Thus, except for 

ethanol, the error in assuming n ˆ 1.00 is less than 12%. 

The EBC/ASBC define four different types of threshold and the term should not be 

used without qualification: 

1. Detection threshold (stimulus threshold, absolute threshold). The minimum value 

of a sensory stimulus needed to give rise to a sensation. 

2. Difference threshold (just noticeable difference). Value of the smallest perceptable 

change in the physical intensity of a stimulus. 

3. Recognition threshold. The minimum value of a sensory stimulus permitting 

identification of the sensation received. 

4. Terminal threshold. Maximum value of a sensory stimulus permitting identification 

of the sensation perceived. 

Usually detection thresholds are lower than recognition thresholds but the literature does 

not always indicate which is being measured especially when the assessors know the 

identity of the test substance. When substances are added to beer we are concerned with 

difference thresholds but many of the methods used for estimation involve recognition of 

the odd sample. In practice, with trained assessors the difference between difference and 

recognition thresholds is negligible (Brown et al., 1978). Threshold values are subject to 

considerable biological variation and it is therefore desirable to include statistical limits 

in any estimation. Criteria which have been used in estimating thresholds include (Brown 

et al., 1978): 

1. The concentration that can just be perceived by 50% of the population. 

2. The geometric mean of the individual thresholds (maximum likelihood threshold). 

3. The lowest concentration that can be detected with a ** statistical significance (P ˆ 

0.01). 

4. The concentration producing 50% correct choices (ASTM). Since in paired sample 

tests 50% correct choices can occur by chance alone, this may be amended to: 

5. That concentration producing a frequency of 50% correct above chance. 

Many factors influence the measurement of thresholds. For example, the influence of 

temperature on taste is not uniform and the buffering action of saliva (pH 7.0) may 

influence perception. The mode of presentation of the sample is also important; thus the 

average sensitivity threshold for sodium chloride varied from 0.135% for three drops 

placed on the tongue to 0.047% for 10 ml swallowed and 0.016% when unlimited 

amounts of the salt solution and distilled water could be compared (Richter and Maclean, 

1939). Similarly, repeated practice appears to lower the threshold at which tastes can be 

perceived. The following thresholds (percentages) were for the first and sixth 

determinations: sucrose, 0.753, 0.274; caffeine, 0.0272, 0.0078; citric acid, 0.0223, 

0.00096; and sodium chloride, 0.123, 0.047 (Pangborn, 1959). 

There can be large differences in sensitivity from person to person. For the addition of 

dimethyl sulphide to beer, tasted by an international panel of 44 persons, 37 panellists 

(85%) had thresholds in the range of 12±87 g/l, but the remaining seven panellists were 

much less sensitive and had thresholds in the range 150±2000 g/l (Brown et al., 1978). 

Similar results were obtained with diacetyl; with a panel of 16 assessors the geometric 

mean threshold was 0.080 mg/l but one assessor had a threshold of 2.26 mg/l (Meilgaard 


Table 20.2 Detection thresholds for various common taste substances (after Breslin, 2001) 

Sensation Compound [M] mg/l 

Bitter Caffeine 

Magnesium Sulphate 

Quinine hydrochloride 

Sucrose octa-acetate 

Urea 

PROP (taster) 

PROP (nontaster) 

Salty NH4Cl 

CaCl2 

LiCl 

NaCl 

KCl 

Monosodium glutamate 

L-Arginine 

L-Glutamine 

Sweet Aspartame 

Fructose 

Glucose 

Glycine 

Saccharin Na 

Sucrose 

Sour Acetic acid 

Citric acid 

Hydrochloric acid 

Malic acid 

Tartaric acid 

PROP = Propylthiouracil 

5 10 4 

3.85 10 4 

1.4 10 6 

3.58 10 6 

(1.07±1.72) 10 2 

2 x 10 5 

6 10 4 

8.39 10 4 

8 10 6 

(0.9±4) 10 2 

1.02 10 3 

(6.31±6.49) 10 3 

5 10 4 

1.23 10 3 

9.77 10 3 

(1.76±2.08) 10 5 

8.9 10 4 

7.33 10 3 

3.09 10 2 

(8.58±10.1) 10 6 

6.5 10 4 

(1.07±1.12) 10 4 

7 10 5 

1.6 10 4 

7.3 10 5 

4.78 10 5 

97 

46.35 

0.505 

2.43 

642±1033 

4.04 

121.2 

44.89 

0.888 

381±1696 

59.62 

470±484 

84.5 

214.3 

1428 

5.17±6.12 

160 

1321 

2317 

1.76±2.07 

222.5 

6.43±6.73 

13.45 

5.84 

9.79 

7.17 

and Reid, 1979). Thus, when measuring thresholds it is desirable to have apanel of at 

least 25 persons and to calculate individual thresholds so that insensitive persons do not 

over-influence the group result (Meilgaard and Reid, 1979; Brown et al., 1978). 

Analytica-EBC/ASBC give methods to determine the threshold of an added substance by 

the ascending method of limits (see later). 

Detection thresholds for representative compounds are collected in Table 20.2. Other 

values are given byMeilgaard (1975) and some can be found inthe Tables inChapter19. 

Quinine (20.1) is often considered the standard bitter taste and is the most bitter 

compound listed in Table 20.1 but denatonium salts are more bitter. Quinine tonic water 

contains 56mg/l (0.5 grain/pint) of quinine sulphate together with sugar (45g/l) and/or 

permitted artificial sweetening agents. Caffeine (20.2) is much less bitter and is found in 

proprietary soft drinks such as Coca-Cola and Lucozade as well as tea and coffee. 

Magnesium sulphate has similar bitterness to caffeine and it is noteworthy that with 

propylthiouracil (20.4) the threshold for tasters is 30 times less than that for non-tasters. 

Sodiumchloridehasthecharacteristic saltytasteabove0.05 M,butweakersolutions taste 

sweet. Potassium chloride also tastes sweet in dilute solutions and salty above 0.05 Mbut 

between 0.02 Mand 0.03 Monly bitterness is perceived and this note persists at higher 

concentration. Calcium chloride has the lowest threshold amongst the salts quoted. The 

artificial sweeteners aspartame and saccharin have much lower thresholds than the 

natural sugars. The relative sweetness of anumber of other sugars is given in Table 20.3. 


Table 20.3 Relative sweetness of sugars (Nieman, 1960) 

Lactose 39 

Maltose 46 

D-Xylose 67 

, -D-Glucose 69 

Glycerol 79 

Invert sugar 95 

Sucrose 100 

Fructose 114 

Calcium cyclamate 3,380 

Saccharin 30,000 

With mineral acids, such as hydrochloric acid, the sourness is proportional to the 

hydrogen ion concentration but this is not the case with organic acids that are largely 

undissociated. At equimolecular concentration, hydrochloric acid tastes more sour than 

acetic acid but at the same pH, acetic acid tastes sourer than the mineral acid. 

Presumably,astheH + ofthedissociatedaceticacidreactswiththetastereceptor,someof 

the undissociated acid ionizes to restore the equilibrium. 

Reviewing the primary tastes inbeeritssourness will bemeasured either asitspH(3.8± 

4.7) or as its titratable acidity. The level of acetic acid reported for normal beers (57± 

145ppm) is below the taste threshold and the same is true for lactic acid (Table 19.7). 

Infectionofbeerwithmicro-organismssuchasAcetobacterspp.or Lactobacillusspp.may 

reduce the pH and produce asour taste. Thus with lambic and gueuze beer the pH may 

drop to 3.2. The level of acetic acid in these beers is 2.6±6.9 times the taste threshold and 

the level of lactic acid is 5.8±8.6 times the taste threshold (Van Oevelen et al., 1976). 

Comparison of the level of the inorganic constituents of beer (Table 19.2) with the 

thresholds give in Table 20.2 suggests that the thresholds of potassium chloride, sodium 

chloride,calciumchloride andmagnesium sulphate couldbeexceeded, butbeers are rarely 

classed as salty. Similarly, the levels of fructose and glucose in the primed beers 2±4 in 

Table 19.5 exceed the taste thresholds and the level of glucose in the ale (sample 10) and 

the lagers (samples 13 and 14) exceeds the threshold. Acomparable threshold for maltose 

is not available; that quoted (1.36%) well exceeds the concentrations of maltose in the 

beers mentioned in Table 19.5 but, if the comparable threshold was similar to that of 

lactose (0.16%), as suggested in Table 20.3, the level of maltose in beers no. 1, 5, 6, 13, 

and 15 would exceed the threshold and influence the taste. 

Thetastethresholdforisohumulone(8.40)isreportedtobe5.6mg/l(1.5 10 5 M)and 

that of hulupone (8.85) 7.7mg/l (2.3 10 5 M)(Gienapp and SchroÈder, 1975). Similarly, 

Weiss et al. (2002) found the best estimated threshold for the dicyclohexylamine salt of 

trans-iso- -acids (66.6% iso- -acids) in tap water was 4.54mg/l (1.25 10 5 M). Thus, 

with one exception, the level of iso- -acids (bitterness units) in all the beers analysed in 

Table19.1exceedsthetastethreshold,intheextremecaseby17times.However,thelevel 

ofhulupones reported(1.1±4.3mg/l) does not exceed thethreshold.By traditional scaling 

methods quinine hydrochloride was some six times more bitter than an iso- -acids 

preparation which, in turn, was about ten times more bitter than caffeine (Lewis et al., 

1980). The character of the bitterness of these compounds is also different; iso- -acids 

especially, and quinine to aminor extent, were perceived as alingering bitterness on the 

back of the throat while caffeine gives ashort-lived bitterness on the tongue. 

Weiss et al. (2002) also found the best estimated threshold for tetrahydroiso- -acids 

(8.47)intapwaterwas1.61mg/l(4.3 10 6 M)butinunhoppedbeerwas8.50mg/l(2.32 


10 5 M). Similarly, the best estimated threshold for the rho-iso- -acids (8.49) was 6.03mg/l 

(1.65 10 5 M) in tap water and 11.22mg/l (3.03 10 5 M) in commercial beer in 

agreement with Table 8.3. These values show the ability of beer to mask bitterness. 

The levels of glutamic acid (Tables 9.3 and 19.12) do not exceed the taste threshold 

given in Table 20.2 and the levels of 5 0 -inosine monophosphate and 5 0 -guanosine 

monophosphate (Table 19.13) in beer are below the taste thresholds but these compounds 

are reported to be flavour modifiers rather than primary flavours. Additions of 5 0 - 

guanosine monophosphate do modify beer flavour but the lowest level of addition of 5 0 - 

GMP required to alter beer flavour is greatly in excess of the amount naturally present 

(Clapperton, 1974). 

Despite their widespread adoption the use of thresholds to give flavour units has been 

criticized: `Thresholds are but one point on dynamic concentration continum' (Pangborn, 

1980). There is no evidence that intensity/concentration curves for all substances are 

parallel differing only in the point where they cross the abscissa. Further, taste is not a 

single instantaneous sensation but has atemporal element. Tasters have been trained to 

record the intensity on ascale between 0(none) and 100 (extreme) on amoving recorder 

chart whereby time-intensity curves such as Fig. 20.6 are obtained (Lewis et al., 1980). 

Normally the sample is held in the mouth for 10 seconds and then expectorated or, if 

beer, swallowed. As would be expected when asucrose gelatine was expectorated the 

intensity of the sweet sensation immediately started to fall and declined to zero in about 

10s. In contrast, when asample of beer was swallowed the intensity of the bitterness 

continued to rise for afurther 8sbefore starting to fall (Fig. 20.6). 

The bitter sensation appears to persist longer (60±90s) than the sweet sensation. 

Hughes and Bolshaw (1995) compared the time-intensity curves for trans-isohumulone, 

andpreparationsofdihydro-,tetrahydro-,andhexahydro-iso- -acids(Fig.20.7).Thetwo 

tasters ranked the compounds in the same order; tetrahydroiso- -acids, hexahydroiso- - 

acids, trans-isohumulone, and dihydroiso- -acids, but the shape of the curves was 

different. Taster Bfound the tetrahydroiso- -acids much more bitter than the other 

compounds and the aftertaste persisted for over three minutes. 

Average intensity (n = 69) 

40 

30 

20 

10 

0 

30 ppm 

20 

10 

0 

0 10 20 30 40 50 60 70 80 90 

Time (s) 

Fig. 20.6 Average time-intensity curve for bitterness of four levels of iso- -acids in 

commercial lager. Samples were placed on the mouth at zero times and swallowed at 10 s. The 

judge continued to record intensity of bitterness until disappearance or for a maximum of two 

minutes. (after Lewis et al., 1980). 


Fig. 20.7 Mean time-intensity curves of trans-isohumulone and mixtures of the chemically 

modified hop bitter acids. Samples were tested in 10 mM sodium phosphate buffer (pH4.15) 

with 0.05% (v/v) ethanol in black opaque glasses under red light. The solutions were 33.1 M 

with respect to the hop bitter acid (Hughes and Bolshaw, 1995). 

In general the olfactory threshold of acompound is several orders lower than the taste 

threshold (Table 20.4). This may represent the probability of acompound taken into the 

mouth reaching the olfactory epithelium. In atriangular test (see later) the assessors may 

be required to distinguish the similar beers first on the basis of odour. Indeed with one 

brewery taste panel it was found that most members differentiated between two beers on 

the basis of aroma rather than taste 90±95% of the time (Hoff et al., 1978). The olfactory 

threshold (OT) can be calculated from the simplified equation: 

log…OT† ˆ logK L=A 0:1Ao ‡…22:13 0:5† 

where KL/A is the absorption constant for molecules passing from air to the aqueous-lipid 

interface (usually between 6.0 and 8.5) and Ao is the cross-sectional area of the molecule 

(usually between 10±60 A Ê2 ).Thus the olfactory threshold of apure compound can be 

calculated to the first approximation from the partial pressure of the compound above an 

aqueous solution of the compound, its partition coefficient between water and light 

petroleum or octanol (substituting for the lipid membrane), and its cross-sectional area, 

which can be calculated from models. 

Foramoleculetoelicitaflavour itmustreachandreactwithaspecificreceptorandto 

do so it will probably have to pass through alipid membrane, thus the lipophilicity of the 

molecule will govern this approach. Gardner (1979) showed highly significant 


Table 20.4 Odour and taste thresholds of various compounds 

correlations between taste threshold and lipophilicity. Thus for ahomologous series of 

compounds (alcohols, esters, ketones, aldehydes and acids) in beers: 

log…1=T†ˆ log…P†‡b 

Odour threshold in water Taste threshold in beer 

(ppb) (Guadagni, 1970) (ppb) (Meilgaard, 1975) 

Ethanol 100,000 14,000,000 

Butyric acid 250 2,000 

Nootkatone 170 - 

Humulene 160 - 

Butanal 70 1 000 

Myrcene 15 ± 

n-Amyl acetate 5 5,000 

Dimethyl sulphide 0.3 50 

n-Decanal 0.1 6 

Methyl mercaptan 0.02 2.0 

-Ionone 0.007 1.3 

2-Methoxy-3-isobutylpyrazine 0.002 ± 

where Tˆthreshold in mol/l and Pôctanol/water partition coefficient representing 

lipophilicity. This relationship breaks down when Pis greater than 3.0. Similarly, in 

many series of compounds with bitter taste, the bitterness increases with increasing 

lipophilicity (Gardner, 1978). Accordingly, Kaneda et al. (2001, 2003) found that the 

absorption/desorption of beer on to alipid coated quartz crystal microbalance was related 

to the sensory bitterness. When amolecule has reached areceptor whether or not it 

initiatesasignaltothebraindependsuponitssize,shape,degreeofionization andcharge 

pattern etc. The size and shape of molecules have also been expressed in terms of 

molecular connectivity (Kier and Hall, 1976). 

On the basis of threshold values and flavour units (FU), Meilgaard (1975) outlined the 

flavour chemistry of beer as illustrated in Table 20.5. Removal of any of the primary 

flavour constituents would produce adecisive change in flavour. Later work has not 

confirmed the importance of humuladienone in the hop aroma compounds but Goiris et 

al. (2002) have confirmed that an oxygenated sesquiterpenoid fraction is responsible for 

the spicy hop character of beer. Mackie and Slaughter (2002) have shown the importance 

of 2,5-dimethyl-4-hydroxy-3(2H)-furanone and related compounds among the caramelflavoured 

compounds. Removal of any one of the secondary constituents will produce a 

small change in flavour. Together the secondary flavour constituents form the bulk of a 

beer's flavour. Any differences between one beer and another of the same type is mostly 

determinedbyvariationsinthisclass.Tertiaryflavourconstituentsaddsubsidiaryflavour 

notes. Removal of any one of this class produces no perceptible change in flavour. 

Similarly it is not possible to say whether the numerous compounds, which individually 

contribute less than 0.1FU to the background flavour, are together of importance in beer 

flavour. 

Consideration of the concentration and threshold data in Table 19.8 will show which 

alcohols, acids and esters may contribute over 2FU in special beers and between 0.5 and 

2.0FU in regular beers. Octanoic acid (difference threshold, 4.5mg/l), decanoic acid 

(1.5mg/l), dodecanoic acid (0.5mg/l), and to alesser extent hexanoic acid contribute to 

the caprylic (goaty) flavour in beer (Clapperton, 1978). The effect of the acids is additive 


Table 20.5 Tentative scheme for role of constituents in determining the flavour of beer 

(Meilgaard, 1975) 

1. Primary flavour constituents (above 2 FU*) 

Ethanol 

Hop bitter compounds (e.g., isohumulone) 

Carbon dioxide 

Speciality beers 

Hop aroma compounds (e.g., humuladienone) 

Caramel-flavoured compounds 

Several esters and alcohols (high gravity beers) 

Short-chain acids 

Defective beers 

2-trans-Nonenal (oxidized, stale) 

Diacetyl and 2,3-pentanedione (fermentation) 

Hydrogen sulphide, dimethyl sulphide and other compounds (fermentation) 

Acetic acid (fermentation) 

2-Methylbut-2-enylthiol (light struck-hops) 

Others (microbial infection etc.) 

2. Secondary flavour constituents (0.5±2.0 FU) 

Volatiles 

Banana esters (e.g., isoamyl acetate) 

Apple esters (e.g., ethyl hexanoate) 

Fusel alcohols (e.g., isoamyl alcohol) 

C6, C 8, C 10 aliphatic acids 

Ethyl acetate 

Butyric and isovaleric acids 

Phenylacetic acid 

Non-volatiles 

Polyphenols 

Various acids, sugars, hop compounds 

3. Tertiary flavour constituents (0.1±0.5 FU) 

2-Phenethyl acetate, o-aminoacetophenone 

Isovaleraldehyde, methional, acetoin 

4-Ethylguaiacol, gamma-valerolactone 

4. Background flavour constituents (below 0.1 FU) 

Remaining flavour compounds 

* Flavour Units (FU) = concentration/threshold 

and there is a linear relationship between the panel score for caprylic flavour and 

concentration of octanoic + decanoic acids. During fermentation lager yeasts produce 

larger amounts of these acids than ale yeasts. Thus the caprylic flavour was observed in 

most of the lagers and 25% of the ales examined (Clapperton and Brown, 1978). 

20.3 Flavour stability 

Beer flavour is not static but in a continual state of change. The point where maturation 

ends and deterioration begins is undoubtedly different for different beers and probably 

different for different consumers. The off-flavour in one beer may be an essential 


Intensity 

Bitter taste 

Ribes aroma 

Sweet taste, 

toffee-like aroma 

and flavour 

Time 

Sweet aroma 

Cardboard flavour 

Fig. 20.8 Diagrammatic representation of sensory changes in beer flavour during ageing 

(Dalgliesh, 1977). 

character of another. Flavour stability has been reviewed by Dalgliesh (1977) and Tressl 

et al. (1980) have reviewed off-flavours. Since brewers can now largely control 

biological instability and non-biological haze, it is probably flavour instability that 

determines the shelf-life of the product. Some of the changes in flavour that occur during 

beer ageing are illustrated in Fig. 20.8. 

Many of the chemical changes during storage involve oxidation and will be 

accelerated if the beer is allowed to come in contact with oxygen after it leaves the 

fermentation vessel. The various electronic states of oxygen are illustrated in Fig. 20.9 

(Lacan et al., 2000). The ground state is adiradical with two unpaired electrons (having 

the same spin) in two * antibonding orbitals and is comparatively unreactive. 

Photoexcitation of the ground state gives `singlet oxygen' (with two electrons of opposite 

spin) but reduction gives asuperoxide radical ion (O2 ). This superoxide can be formed 

biochemically from oxygen by reactions involving ubiquinones, catecholamines or thiols 

and is normally converted to hydrogen peroxide (H2O2) by the enzyme superoxide 

dimutase (SOD). However, in acidic media it forms the hydroperoxy radical (HOO ) 

which spontaneously disproportionates into hydrogen peroxide and oxygen. In the cell 

hydrogen peroxide is usually destroyed by the enzyme catalase but it can lead to even 

more `reactive oxygen species' (ROS), notably the hydroxy radical (HO ). Thus, 

hydrogen peroxide can react with superoxide (Haber-Weiss reaction) or with ferrous ions 

(Fenton reaction) (Fig. 20.10) The hydroxyl radical is probably responsible for initiating 

the autoxidation of lipids and the role of the Fenton reaction shows the importance of 

traces of heavy metals on beer deterioration. Kaneda et al. (1992) using electron spin 

Singlet 

1 O2 

hν 

Ground state 

O2 

Superoxide 

O 2– • 

radical anion 

Peroxide 

O 2– 

Fig. 20.9 Electronic configurations of various oxygen species (Lacan et al., 2000). 


Haber-Weiss reaction 

O2 – 

+ H2O2 O2 + + HO – 

• 

HO• 

Superoxide Hydroxyl Hydroxyl 

radical ion 

Fenton’s reaction 

Fe 2+ H2O2 Fe 3+ 

resonance (ESR) showed that the signal due to non-heme Fe 3‡ increased during beer 

storage. After the addition of potassium hydroxide another signal in the ESR spectrum 

indicated the presence of free radicals but the half-life of the hydroxy radical is reported 

as only 10 9 s.Uchida and Ono (1999) measured the increase in level of hydrogen 

peroxide in beer and found it paralleled the formation of hydroxy radicals. 

Unsaturated fatty acids, either free or esterified, are particularly susceptible to 

autoxidation. Compared with oleic acid (cis-octadec-9-enoic acid), linoleic acid (20.6, 

cis, cis-octadeca-9,12-dienoic acid) reacts with oxygen 12 times faster and linolenic acid 

(cis, cis, cis-octadeca-9,12,15-trienoic acid) reacts 25 times faster. 

13 12 11 10 9 

R1·CH CH·CH2·CH CH·R2 

R1 CH3(CH2)4· 

(20.6) 

(20.7) 

R1·CH•·CH CH·CH CH·R2 +O2/H• / R1·CH(OOH)·CH CH·CH CH·R2 

R1·CH CH·CH CH·CH•R2 +O2/H• / R1·CH CH·CH CH·CH(OOH)·R2 

·H• 

R1·CH CH·CH•·CH CH·R2 

R1·CH•·CH CH·CH CH·R2 

R1·CH CH·CH CH·CH•·R2 

+ + HO – 

HO• 

Fig. 20.10 Production of the hydroxyl radical. 

R2 ·(CH2)7COOR 

(20.8) 

(20.9) 

Autoxidation involves abstraction of ahydrogen radical from acarbon atom adjacent to a 

double bond, e.g., positions 8and 11 in oleic acid and, particularly, position 11, between 

two double bonds, in linoleic acid (20.6). This forms amesomeric radical (20.7) which 

then reacts with oxygen to form a mixture of hydroperoxy radicals (ROO ) which 

abstract another proton to form hydroperoxides (ROOH), e.g. (20.8) and (20.9). Thus, 

with linoleic acid, hydroperoxides at positions 9and 13 predominate. Double bonds that 

migrate usually adopt trans-geometry. Reduction of the hydroperoxide gives ahydroxyl 

group when hydroxylation of the double bonds gives isomers of trihydroxyoctadecenoic 

acids (Table 19.7). These acids are the precursor of trans-2-nonenal, which is responsible 

for the cardboard flavour in stale beer (Lermusieau et al., 1999; NoeÈl et al., 1999). It 

appears that this compound is not formed from oxygen in the head space of bottled beer 

but during wort preparation when it is bound, for example by proteins or sulphites, and 

released during storage (NoeÈl et al., 1999). 


Thethresholdoftrans-2-nonenalisabout0.1 g/landithasbeensuggestedthatitslevel 

provides aindication of the degree of staling. Other carbonyl compounds formed from the 

lipids in beer by irradiation with light include the C9, C10, and C11-alka-2,4-dienals 

(thresholds 0.5, 0.3 and 0.01ppb respectively) (Tressl et al., 1980). Using GC-O Evans et 

al. (1999) confirmed that during ageing the level of most aldehydes increased. As well as 

trans-2-nonenal, the levels of phenylacetaldehyde, methional, 4-methoxybenzaldehyde and 

heptanal increased but the level of octanal fell. The level of diacetyl and pentane-2,3-dione 

in arange of commercial beers is given in Table 19.11. Quantities in excess of 0.15ppm 

impart abuttery flavour more noticeable in lagers than in ales. Bacterial contamination and 

petite mutants of yeast result in high levels of these diketones. 

Although 18 Oin the headspace was not incorporated into the carbonyl fraction during 

storage of bottled beer, it was involved in the oxidation of sulphites, polyphenols and 

isohumulones. Roughly 1ml of air in a300ml bottle will give an oxygen content of 1ppm, 

which is probably enough to oxidize all the reductones present in alight lager beer. The 

dissolved oxygen inbeer rapidly disappears, usually withoutthe immediate formationof an 

off-flavour, but the damage may have been done as beer contains compounds such as 

melanoidins and reductones, which act as oxygen carriers, and produce off-flavours at a 

later date. Chemically, reductones contain the grouping -C(OH)ˆC(OH)-CÔ- and the 

best characterized reductone is ascorbic acid (vitamin C)(4.96). Ascorbic acid has been 

detected in green malt and the leaves of green hops but is destroyed during kilning. 

Ascorbicacid,andotherreductones,readilycombinewithoxygenandaccordinglyascorbic 

acid finds use as achill-proofing agent in beer (Chapter 15). Ascorbic acid (4.96) is 

reversibly oxidized to dehydroascorbic acid (4.97) but more extensive decomposition 

occurs under quite mild conditions. In model experiments designed to assess the efficiency 

of various substances to degrade valine to isobutyraldehyde by the Strecker mechanism, 

dehydroascorbic acid was 5 10 times more active than fructose which, in turn, was 2 3 

times more active than glucose or sucrose or ascorbic, pyruvic or chlorogenic acids (Swain 

and Casey, 1963). Ascorbic acid is usually estimated colorimetrically with the oxidationreduction 

indicator, 2,6-dichlorophenolindophenol (20.10), but other reductones will 

interfere. The Indicator Time Test (ITT) (Gray and Stone, 1939), which measures the 

decolorization of 2,6-dichlorophenolindophenol, gives an indication of the oxidationreduction 

or redox level of abeer. 

OH 

N 

Cl Cl 

O 

2,6-Dichlorophenolindophenol 

(20.10) 

The bitterness of beer declines during storage (Fig. 20.8). De Cooman et al. (2000) 

found that both in lagers and top-fermented beers the trans-iso- -acids deteriorated at a 


Me3C 

OH 

Me 

(20.11) BHT 

CMe3 

+ R·OO• 

much faster rate than the cis-iso- -acids. During the first 15 months the loss of iso- - 

acids was mainly due to decomposition of the trans-isomers. In the trans-iso- -acids the 

double bonds in the 3-methyl-2-butenyl- and the 4-methyl-3-pentenoyl- side chains lie 

close together and it is suggested that this increases their susceptibility to autoxidation. 

Therefore the cis-trans ratio of the iso- -acids gives ameasure of abeer's deterioration. 

As discussed in Chapter 19 the flavours due to 4-vinylphenol and 4-vinylguaicol (4.134, 

threshold, 0.3 mg/l) are regarded as off-flavours in most beers but, with orcinol, are 

responsible for the clove-like character of Weizenbier. 

Beer also contains many antioxidants derived mainly from the polyphenols present in 

the malt and hops. Phenols react readily with free radicals but form mesomeric radicals 

which are not sufficiently energetic to propagate the free radical chain further (Lacan et 

al., 2000). For example, one molecule of the synthetic antioxidant, butylatedhydroxytoluene 

(BHT) (20.11) can destroy two hydroperoxy radicals (Fig. 20.11). 

Thelevelofsomevolatile sulphurcompoundsincreases during storage.We saw inthe 

last chapter that low levels of hydrogen sulphide are acceptable in ales and dimethyl 

sulphide is characteristic of some lagers (Table 19.20). As mentioned earlier, beers 

bottled in clear glass bottles and exposed to sunlight develop skunky `sunstruck' flavours 

due to 3-methyl-2-butenyl thiol (8.48, prenyl mercaptan) which has avery low threshold 

of 0.005ppb in water and 0.05ppb in beer. Other beers acquire aflavour described as 

`catty' or Ribes, as asimilar aroma is given off by the leaves and stems of flowering 

currants (Ribes spp.). The development of this flavour is closely correlated with the 

amount of headspace air (Clapperton, 1976). In beers bottled with high volumes of 

O 

SH 

Me3C 

8-Mercapto-p-menthan-3-one 

(20.12) 

O• 

O O 

Me 

Me OOR Me • 

CMe3 

Me3C CMe3 R·OO• 

Me3C CMe3 

Fig. 20.11 Antioxidant activity of Butylatedhydroxytoluene (BHT). 


headspace air the flavour develops rapidly over six weeks but thereafter slowly declines. 

One substance responsible for the catty flavour is 4-mercaptopentan-2-one (threshold, 

0.005ppb in water, 0.05ppb in beer). Beers with strong catty odours contained 1.5ppb of 

the mercaptopentanone but there may be other beer constituents which contribute to this 

off-flavour. Elsewhere (see Table 20.11 on page 750) 8-mercapto-p-menthan-3-one 

(20.12, p-menthane-8-thiol-3-one) has been proposed as a reference standard for the catty 

(Ribes) flavour. 

Gijs et al. (2002) applied Aroma Extract Dilution Analysis to fresh and aged beers 

(five days at 40 ëC, pH4.2) and found that the Flavour Dilution (FD) values were 

increased in the aged beer for ethyl butyrate, dimethyl trisulphide, 2-acetylpyrazine, 

methional, 2-methoxypyrazine, maltol (9.11), -nonalactone, -damascenone (8.165), 

and ethyl cinnamate. -Damascenone and an unknown compound, with a `dentist', 

smoked, vanilla odour with the same FD (243) are probably the most important odours in 

aged beer and more important than trans-2-nonenal (FD 81). Strecker degradation of the 

amino acid methionine gives methional, (3-(methylthio)propionaldehyde), which is 

responsible for the worty flavour of alcohol-free beers (PerpeÁte and Collin, 1999). It is 

also the precursor of dimethyl trisulphide which develops as beer ages (Gijs et al., 2000; 

Gijs and Collin, 2002). 

-Damascenone (8E-megastigma-3, 5, 8-trien-7-one, (8.165)), a degradation product 

of the carotenoid neoxanthin, is a key odour compound in a number of fruits and has an 

extremely low threshold (0.02 0.09 ng/g (ppb) in water). It is present in hops but is also 

found in unhopped wort. Hopped wort contained 450 ng/g but this was reduced during 

fermentation so fresh beers contained only low levels of -damascenone (6±25 ng/g). 

However, during ageing (five days at 40 ëC) the level increased to as much as 210 ng/g 

(Chevance et al., 2002). Experiments with -glucosidase suggest that the production of 

-damascenone during beer ageing can be partially explained by the hydrolysis of 

glucosides. This was confirmed as the production of -damascenone fell in beers aged at 

higher pH values (Gijs et al., 2002). The production of dimethyl trisulphide also fell at 

higher pH values but that of 3-(methylthio)propionaldehyde increased. 

20.4 Sensory analysis 

Sensory analysis uses the human senses to assess flavours. It has been discussed in books 

by Amerine et al. (1965), the International Organization for Standardization (1983), 

Meilgaard et al. (1987), Piggott (1988) and Lawless and Heymann (1999). Analytica- 

EBC/ASBC give details for paired comparison tests, triangular tests, duo trio tests, 

determining the threshold of added substances, description analysis, ranking tests and 

provide a flavour terminology. For other tests, for example, the À' or `not A' test, the 2out-of-5 

test, the Scheffe paired comparison test and the multiple paired comparison test, 

reference should be made to the above texts. 

Analytica-EBC/ASBC first give a glossary of terms and definitions, then detailed 

instructions for carrying out sensory analysis. They suggest a layout for a medium-sized 

sensory evaluation area for brewery control work; consumer preference testing requires a 

much larger panel. The tasting room should be situated in a quiet area free from any 

smells. They suggest six booths 60±80 cm wide with a counter top 40±50 cm deep and 

90 cm high. Dividers between the booths should project c. 46 cm and extend from the 

floor to the ceiling. Walls, floors and ceiling should be of smooth non-absorbing material 

of a pale neutral colour. The booths may be equipped with a hatch (sliding door, bread 


ox or carousel) so the samples can be delivered from an adjacent preparation room. The 

booths should be odour free with a slight overpressure of air that has been filtered through 

charcoal. The air should be at 22 ëC and at 45±55% relative humidity. Shadow-free 

illumination at 70±80 foot candles should be provided. Odourless water and salt-free 

crackers should be available to cleanse the palate. A small sink may be provided but, if 

so, the drains should be readily dismantled to allow for cleaning. The booth may be 

equipped with a computer station so that the panellists' observations can be analysed 

quickly and automatically without resource to the forms detailed below. 

Liquid samples are best provided in straight-sided cylindrical 250 ml (8 oz.) glasses 

which may be deeply coloured to mask differences in colour, clarity or foaming. The 

glasses should be washed in an odour-free detergent such as sodium hexametaphosphate 

not more than 12 hours before the test. 50±100 ml of sample should be served at the 

designated temperature. A sample temperature of 12 ëC is suitable for full perception of 

flavour, such as detecting faults or small differences. Mouth-feel and drinkability are best 

evaluated at a lower temperature, e.g., 8 ëC. For preference tasting, the beer should be 

tasted at the usual drinking temperature, making due allowance for warming which may 

occur between pouring and drinking. The number of samples presented to each assessor, 

their order and coding should be carefully monitored. The order of presentation should be 

balanced so that each sample appears in a given position an equal number of times. For 

example, the possible positions for three products A, B, and C to be compared in a 

ranking test are: 

ABC ACB BCA BAC CBA CAB 

so a multiple of six assessors should be chosen so that the six possible combinations can 

be presented an equal number of times. 

The order of presentation should be randomized, e.g., by drawing sample cards from a 

bag or using a table of random numbers. Similarly, the samples should be coded with 

three-digit random numbers. Too many samples should not be presented at a single 

session to avoid sensory fatigue. Panel sessions should be held before meals preferably 

between 10 and 12 a.m. The assessors should be carefully instructed what is required of 

them. The amount of sample to be tasted, how long it is to be held in the mouth and 

whether it should then be swallowed or expectorated should be clearly stated. The use of 

the scoresheet, any terminology and the interpretation of the scales used should be 

explained. Finally, the information sought in the test and the type of judgement/ 

evaluation required, e.g., difference, descriptive, preference, acceptance should be stated. 

At the end of the test, at a location away from the tasting area, the codes may be disclosed 

and the assessors allowed to discuss the results among themselves and/or with the panel 

leader. It is the job of the panel leader to keep the panellists motivated and give them 

regular reports on their results. 

Sensory tests have two main applications: (i) those in which the primary aim is to 

describe the product and (ii) those in which the aim is to distinguish between two or more 

products. With regard to the latter it is important to distinguish between the need to know 

if there is a difference at all, the magnitude of that difference, the direction (or quality) of 

that difference, the effect of that difference, for example, with regard to preference, and 

whether all or only part of the population detects a difference. Analytica-EBC/ASBC 

provide a simplified key to decide which tests are relevant to a given problem. 

For the selection and training of assessors, EBC/ASBC suggest that you interview and 

screen two to three times the number of assessors required. The general health of the 


candidates should be checked, whether they are on regular medication, whether they 

smoke, and their ability to communicate verbally. Smokers are not automatically 

disqualified. Candidates can then be subjected to matching tests. For example, they are 

presented with four to six coded (A±F) but unidentified common flavours (sucrose, 

tartaric acid, caffeine, sodium chloride, tannic acid and ferrous sulphate) and asked to 

familiarize themselves with the tastes using as much odourless water as they like to 

cleanse their palates. They are then presented with the same samples labelled with threedigit 

numbers and asked to say which of the standards each resembles. Candidates 

scoring less than 80% matches should be rejected. For detection tests, candidates are 

presented with three samples containing beer to which additions have been made versus 

control samples in a Triangular test (see later). The additions can be the above flavour 

standards, first at four to six times the threshold, and then at two to three times the 

threshold. Candidates scoring less than 60% at the higher level should be rejected and 

preference given to those scoring 100% at the higher level and over 60% at the lower 

level. Sequential triangular tests can be used to find the taste acuity of the candidates but 

a more important skill is the ability to detect (and grade) individual flavour notes in a 

`fog' of other impressions. 

For ranking tests, the candidates are asked to discriminate a set of beer samples to 

which additions have been made and which are presented in a random order. For 

example; 0, 0.6, 1.2, and 1.8 mg/l geraniol. Other suggested materials are sucrose, sodium 

chloride, isoamyl acetate, dimethyl sulphide and acetic acid. Only candidates that rank 

samples correctly, or invert only adjacent pairs, should be accepted. Finally, to test for 

descriptive ability, present sets of five to ten stimuli that are typical of the samples to be 

evaluated. Present the samples one at a time and ask the candidate to describe his or her 

response. Suitable substance and concentrations are given in Analytica-EBC/ASBC. 

However, most brewing companies use their own employees as tasters and this leads to 

little or no scope for selecting people with good sensory ability. In addition, employees 

often miss panel sessions and cannot devote time for training because of other duties. 

Accordingly, one brewery company has recruited an external expert sensory panel. From 

70 applicants, ten were selected and employed for three three-hour sessions per week The 

members of the panel were trained as tasters, not just `beer' tasters, and were exposed to a 

wide range of foods and drinks. After training, the 95% confidence limits for this new 

panel were less than 5% whereas those for the old in-house panel were 20±30% (Hegarty 

et al., 2001). 

For training, Analytica-EBC/ASBC suggest that the assessors should be instructed to 

be objective and ignore likes and dislikes unless specifically asked for preference 

information. The assessors should normally proceed in the order appearance, odour, taste, 

and aftertaste. When assessing odour, the assessors should take short rather than long 

sniffs and not sniff too many times so they become fatigued and confused. As well as 

training in recognition and detection of basic tastes and odours, the panel should be 

trained in the use of descriptive language and scales. Samples should be presented to 

illustrate grainy, dry hop, kettle hop and overage flavours. The panel leader should 

analyse daily results and note for each assessor any obvious defects such as drift, lack of 

interest, or failure to detect obvious product variations. Such assessors should be called 

for special training. Most tasters will require attention at least three or four times a year. 

Because of the many opportunities for variability and bias resulting from the use of 

human subjects, reports of sensory tests should contain more detail than the reports of 

physical or chemical measurements. The report should begin with a summary, of not 

more than 110 words, answering the following: 


· What was the objective? 

· What were the results? 

· What was done? 

· What can be concluded? 

Theobjectiveofthetestshouldbeclearlystatedwithacleardefinitionoftheproblemand 

theapproachtakentosolveit.Thetestobjectivesshouldbeagreedbeforetheexperiment. 

Sufficient experimental detail should be given to allow the study to be repeated. The 

experimental design should be given showing how it meets the objectives, as should the 

sensory tests used, the measurements made, and the make up and previous experience of 

thepanel.Finally,detailsofsamplepreparationandpresentation,theinformationgivento 

the panel and the statistical techniques used to analyse the results should be reported. The 

resultsshouldbepresentedintheformoftablesorfigures(notboth)anddiscussedbriefly 

stating whether they support or fail to support any original hypothesis. The report should 

end with clear-cut conclusions. 

The paired comparison test may be either (i) adirectional difference test or (ii) a 

paired preference test; it can also be used in assessor training. The former is used to 

determine in what way a particular sensory characteristic differs between the two 

samples, e.g., more sweet or less sweet. As arule at least seven assessors are required but 

up to 30 can be used (more for consumer tests). They should be familiar with the 

characteristic to be examined. For example, if the test concerns detection of an offflavour, 

the panel should first be allowed to taste asample free of any off-flavour and, if 

possible, a demonstration of the off-flavour. In general the inclusion of controls 

(reference substances) is advisable. Paired samples are offered simultaneously, an equal 

number of AB and BA with random three-digit codes. The assessors should be instructed 

to examine sets in aspecified order, e.g., always from left to right, however, they may 

make repeated tests of any sample while tasting is in progress. Specimen answer forms 

are given in Fig. 20.12. The test supervisor may use one of the following two 

possibilities: 

1. adopt the `forced choice' technique in which the assessor is asked to choose one 

sample or the other (by guessing if no difference is perceived) or 

2. allow the answer `no difference'. 

In most test situations (two-sided) the test question does not distinguish between the 

samples and the reply may favour one or the other sample. One-sided tests are 

occasionally used when the characteristic can vary only in one direction. The results are 

collected and interpreted by reference to Fig. 20.13 and Tables 20.6 and 20.7. With the 

`forced choice' technique asignificant difference or preference is established if the 

number found is equal to or larger than that given in the table. The `no difference' 

technique is not amenable to formal statistical analysis but two approaches can be used. 

In one the `no difference' replies are ignored and in the other half the `no difference' 

replies are allocated to each category. The test report should allow full identification of 

the samples examined, the characteristic studied, whether or not reference substances 

were used, and the results with their probability levels. 

The triangular test is used to determine whether asensory difference is apparent 

between two samples. The assessor is presented with aset of three samples, two of which 

are identical. After tasting the assessors complete aform similar to Fig. 20.14 showing 

the sample perceived to be different and the results are interpreted by reference to Table 

20.8. `No difference' results should be considered invalid. In the duo trio test the 


Directional Difference Test 

Name ______________________ Date __________________ 

day/month/year 

Object of Test _______________________________________ 

Test Criterion _______________________________________ 

Test Pairs Which sample is more __________________? 

Sample No. Sample No. 

_________ _________ __________________ 

_________ _________ __________________ 

_________ _________ __________________ 

Comments 

Paired Preference Test 

Name ______________________ Date __________________ 


Object of Test _______________________________________ 

Test Criterion _______________________________________ 

Test Pairs Which sample is more __________________? 

Sample No. Sample No. 

_________ _________ __________________ 

_________ _________ __________________ 

_________ _________ __________________ 

Comments 

Fig. 20.12 Specimen answer forms for the directional difference test and the paired preference 

test (Analytica-EBC). 

Directional difference test 

Paired preference test 

Two-sided test 

Question: which sample has 

the stronger intensity of the 

charcteristics studied? 

Count the number of replies 

citing one of the two samples 

the more frequently. 

Conclude that the intensity for 

this sample is significantly 

stronger than for the other if 

the number obtained is greater 

than or equal to that shown in 

Table 20.7. 

Question: Which sample do 

you prefer? 


citing one of the two samples 

the more frequently. 

Conclude that this sample is 

significantly preferred to the 

other if the number obtained 

is greater than or equal to that 

shown in Table 20.7. 

One-sided test 

Question: which sample has 

the stronger intensity of the 

charcteristics studied? 


choosing the sample of interest. 

Conclude that this stronger 

intensity is significantly 

apparent if the number of 

positive replies is greater than 

or equal to the number shown 

in Table 20.6. 

Question: Which sample do 

you prefer? 


choosing the sample of interest. 

Conclude that there is a preference 

for the sample of interest 

if the number of positive replies 

is greater than or equal to the 

number shown in Table 20.6. 

Fig. 20.13 Questions and interpretations of the directional difference test and the paired 

preference test (Analytica-EBC). 


Table 20.6 One-sided test (P =0.50 with nreplies) 

Number of Minimum number of positive replies for significance level of 

replies 

0.05 0.01 0.001 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

30 

35 

40 

45 

50 

60 

70 

80 

90 

100 

7 7 ± 

7 8 ± 

8 9 ± 

9 10 10 

9 10 11 

10 11 12 

10 12 13 

11 12 13 

12 13 14 

12 14 15 

13 14 16 

13 15 16 

14 15 17 

15 16 18 

15 17 18 

16 17 19 

16 18 20 

17 19 20 

18 19 21 

20 22 24 

23 25 27 

26 28 31 

29 31 34 

32 34 37 

37 40 43 

43 46 49 

48 51 55 

54 57 61 

59 63 66 

The values given in the tables were calculated from the exact formula binomial law for parameter P=0.50 with 

n repetitions (replies). 

When the number of replies is higher than 100 use the following formula based on the approximation of the 

binomial law by the normal law which gives the actual minimum number of assessments to be obtained with a 

maximum error equal to at most one unit. Minimum number of replies: nearest whole value to (n + 1)/2 + k p n 

in which k is chosen from the table below. 

Level of significance K 

One-sided Two-sided 

0.05 0.82 0.98 

0.01 1.16 1.29 

0.001 1.55 1.65 

Tables for calculating other significance levels may be found in Roessler et al. (1978) and Fernandus et al. 

assessors are first presented with the identified reference sample. This is followed by two 

coded samples, one of which isidentical to the reference sample. The assessor is asked to 

identify the odd sample and complete the form Fig. 20.15 and the results are interpreted 

by reference to Table 20.9. 

The measurement of the threshold of added substances is not carried out routinely but 

is required when anew substance is found which may or may not influence the flavour of 

beer. The experimental design used is known in psychophysics as the forced choice 


Table 20.7 Two-sided test* (P =0.50 with nreplies) 

Number of replies Minimum number of positive replies for significance level of 

0.05 0.01 0.001 

7 7 ± ± 

8 8 8 ± 

9 8 9 ± 

10 9 10 ± 

11 10 11 11 

12 10 11 12 

13 11 12 13 

14 12 13 14 

15 12 13 14 

16 13 14 15 

17 13 15 16 

18 14 15 17 

19 15 16 17 

20 15 17 18 

21 16 18 19 

22 17 18 19 

23 17 19 20 

24 18 19 21 

25 18 20 21 

30 21 23 25 

35 24 26 28 

40 27 29 31 

45 30 32 34 

50 33 35 37 

60 39 41 44 

70 44 47 50 

80 50 52 56 

90 55 58 61 

100 61 64 67 

*Refer to footnotes of Table 20.6 (from Analytica-EBC (1998)). 

Name ____________________ Date ___________________ 


Product submitted to test ______________________________ 

Problem: 3 samples are presented to you; circle the number of 

that which is different from the other 2. 

Comments 

Set of 3 samples 

Fig. 20.14 Specimen answer form for the triangular test (Analytica-EBC). 

modification of the ascending method of limits test. Sixteen or more assessors receive six 

sets of three beers each consisting of two controls and one test sample. Test samples 

increase in concentration by aconstant factor, usually 2.0. An approximate threshold (t a) 

may be determined first using five to ten assessors and increasing the concentration by a 


Table 20.8 Minimum number of correct replies to establish significance at various probability 

levels for the triangular test (one-sided p = 1/3)* 

Number 

of replies 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

32 

33 

34 

35 

36 

37 

38 

39 

40 

41 

42 

43 

44 

45 

46 

47 

48 

49 

50 

51 

52 

Minimum number of correct replies 

for a significance level of 

Number 

of replies 

Minimum number of correct replies for 

a significance level of 

0.05 0.01 0.001 0.05 0.01 0.001 

4 5 ± 

5 6 ± 

5 6 7 

6 7 8 

6 7 8 

7 8 9 

7 8 10 

8 9 10 

8 9 11 

9 10 11 

9 10 12 

9 11 12 

10 11 13 

10 12 13 

11 12 14 

11 13 14 

12 13 15 

12 14 15 

12 14 16 

13 15 16 

13 15 17 

14 15 17 

14 16 18 

15 16 18 

15 17 19 

15 17 19 

16 18 20 

16 18 20 

17 18 21 

17 19 21 

17 19 22 

18 20 22 

18 20 22 

19 21 23 

19 21 23 

19 21 24 

20 22 24 

20 22 25 

20 23 25 

21 23 26 

21 24 26 

22 24 27 

22 24 27 

22 25 27 

23 25 28 

23 26 28 

24 26 29 

24 27 30 

53 

54 

55 

56 

57 

58 

59 

60 

61 

62 

63 

64 

65 

66 

67 

68 

69 

70 

71 

72 

73 

74 

75 

76 

77 

78 

79 

80 

81 

82 

83 

84 

85 

86 

87 

88 

89 

90 

91 

92 

93 

94 

95 

96 

97 

98 

99 

100 

24 27 30 

25 27 30 

25 28 30 

26 28 31 

26 28 31 

26 29 32 

27 29 32 

27 30 33 

27 30 33 

28 30 33 

28 31 34 

29 31 34 

29 32 35 

29 32 35 

30 33 36 

30 33 36 

31 33 36 

31 34 37 

31 34 37 

32 34 38 

32 35 38 

32 35 39 

33 36 39 

33 36 39 

34 36 40 

34 37 40 

34 37 41 

35 38 41 

35 38 41 

35 38 42 

36 39 42 

36 39 43 

37 40 43 

37 40 44 

37 40 44 

38 41 44 

38 41 45 

38 42 45 

39 42 46 

39 42 46 

40 43 46 

40 43 47 

40 44 47 

41 44 48 

41 44 48 

41 45 48 

42 45 49 

42 46 49 

The values in this table were calculated from the exact formula: binomial law for parameter p = 1/3 with n 

repetitions (replies). When the number of replies is larger than 100, numbers of required correct replies 

may be obtained from the following formula based on the approximation binomial law by the normal law, 

with a maximum error equal to one unit: x ˆ 0:4714z n 

p ‡ ‰…2n ‡ 3†=6Š where z ˆ 1:64 for 0:05, 2.3 

for 0:01, and 3.10 for 0:001. The minimum number of correct replies is x if x is a whole number 

or the next higher integer if x is not a whole number. Tables for significance levels other than those listed 

here may be found in Amer. Soc. Testing and Materials (1979) and Jones (1956) (from Analytica-EBC 

(1988)). 


Table 20.9 Significance of results in òne-sided test' (P ˆ 0.50 with n replies) 

Number of replies Minimum number of positive replies for significance level of 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

30 

35 

40 

45 

50 

60 

70 

80 

90 

100 

0.05 0.01 0.001 

7 7 ± 

7 8 ± 

8 9 ± 

9 10 10 

9 10 11 

10 11 12 

10 12 13 

11 12 13 

12 13 14 

12 14 15 

13 14 16 

13 15 16 

14 15 17 

15 16 18 

15 17 18 

16 17 19 

16 18 20 

17 19 20 

18 19 21 

20 22 24 

23 25 27 

26 28 31 

29 31 34 

32 34 37 

37 40 43 

43 46 49 

48 51 55 

54 57 61 

59 63 66 

The values given in the tables were calculated from the exact formula binomial law for the parameter P = 0.50 

with n repetitions (replies). When the number of replies is larger than 100, numbers of correct replies may be 

obtained from the following formula based on the approximation of the binomial law by the normal law, with a 

maximum error equal to one unit: x = (n + 1)/2 + k n 

p where k = 0.82 for 0.05: and k = 1.16 for 0.01; 

and 1.55 for 0.001. Tables for significance levels other than those listed here may be found in Roessler et al. 

(1978) and Fernandus et al. (1970) (from EBC-Analytica (1998)). 

Name ____________________ Date ___________________ 


Product submitted to test ______________________________ 

Problem: The sample on the left is a control. Of the other 2 

samples, one is the same as the control and the other is different. 

Indicate the different sample. 

Comments 

Set of 3 samples 

Fig. 20.15 Specimen answer form for the triangular test (duo trio) (Analytica-EBC). 


0.25ta 

0.5ta 

Step no: 1 2 3 4 5 6 

ta 

Fig. 20.16 Threshold of added substances: example of the presentation of the six triangles t a= 

approximate threshold (Analytica-EBC). 

factor of 3.0. The dilution series in the main test is then: step 1, 0.25 ta; step 2, 0.5 ta;step 

3, ta; step 4, 2ta; step 5, 4ta; and step 6, 8ta set out as illustrated in Fig. 20.16. The 

assessors are asked to indicate the position of the test sample in each set of three beers. 

An example of the questionnaire is given in Fig. 20.17. For each assessor the best 

estimated threshold (BET) is calculated as the geometric mean of the highest 

concentration missed and the next highest adjacent concentration. Ahistogram of the 

individualBETsofthegroup isproducedandfromitthegroupthresholdisthegeometric 

mean of the BETs. In the extended form of the test, critical sets of three beers are 

repeated two to four times until both the assessor and the test supervisor agree that the 

threshold has been successfully bracketed by the assessor. For assessors at the top and 

bottom of the range, extra concentration steps may be required. 

For description analysis 15±30 trained assessors are required. In apreliminary step, 

assessors agree which attributes (usually 10 to 40) will be used, and ascale is defined for 

each attribute, if possible with the use of reference standards. In the test itself, after 

tasting the sample, assessors award an intensity score for each attribute. Results may be 

used to form asensory profile of the sample, and profiles of two or more samples may be 

compared using statistical techniques. The attributes to be rated may be chosen from the 

122termsand the14 flavour classesofthe standardterminology(Table20.11). Tento20 

attributes are chosen for the simple descriptive test and up to 50 or more for the full test. 

2ta 

Ascending method of limits 

4ta 

Date ____________________ Assessor __________________ 

8ta 

You have received 6 sets of beer samples. Each set is a triangle 

consisting of 2 identical controls and 1 test sample containing an 

added substance. Concentrations of the added substances increase 

from left to right. Please locate as many as you can on the test 

samples, indicating their position with a check mark in the 

corresponding box in each column. Avoid sensory fatigue: locate 

strong samples by smell or by taking very small sips, conserving 

your discriminatory power for those sets of 3 beers near your 

threshold. Review your results with a test supervisor. 

Triangle no.: 1 2 3 4 5 6 

Describe the flavour of the added substance _________________ 

____________________________________________________ 

Fig. 20.17 Threshold of added substances; example of questionnaire (Analytica-EBC). 


Examples of intensity scales of proven usefulness are: 

Scale A 

or 

Scale C 

0 1 2 3 4 5 6 8 7 9 

0 

not 

present 

1 

2 

just slight 

recognizable 

3 

moderate 

A 15 cm line with decriptive terms 1.5 cm from each end 

4 

strong 

5 

very 

strong 

Scale B 

weak 0 0 0 0 0 0 0 strong 

weak strong 

Assessors place a mark on the line to indicate intensity. A numerical score can be 

obtained by measuring the distance, in cm, from the left hand end of the line to the mark. 

Reference standards should be used where possible to anchor two or three points on the 

scale. 

Panel scores 

Panel scores 

Panel scores 

30 

25 

20 

15 

10 

5 

0 

25 

20 

15 

10 

5 

0 

25 

20 

15 

10 

5 

0 

(i) Lager beer 

Aroma Flavour After 

palate 

2 3 6 12 15 17 40 1 2 3 12 15 17 23 25 26 29 30 33 34 39 40 2 39 

(ii) Pale ale 

2 3 6 12 1 2 3 6 12 

(iii) Draught bitter 

2 6 12 17 40 1 2 3 4 12 23 

23 25 26 29 30 31 33 34 39 2 39 

29 31 33 34 39 40 2 4 39 40 

Fig. 20.18 Description analysis; profile analysis of the effects on flavour of the addition of 0.2, 0.6 

and 1.8 mg/l of diacetyl to three different beers. Histograms show the scoring of intensity of aroma 

and flavour qualities by a panel of eight assessors. Scores are entered from left to right for the 

control beer and the corresponding beers resulting from the three increasing levels of addition of 

diacetyl. Continuous lines depict the profile descriptions of the flavours of the unadulterated beers. 

Key to qualities: 1 liveliness (CO 2 tingle), 2 sweet, 3 sickly, 4 toffee-like, 6 estery, 12 diacetyl, 

15 cabbagy, vegetable water, 17 sulphury, 23 smooth, 25 acidic (sharp), 26 sour, 29 mouth 

coating, 30 astringent, 31 drying, 33 body, 34 watery, 39 bitter, and 40 hoppy (Analytica-EBC). 


Body 

Brand 1 

Brand 2 

Aroma 

strength 

Aftertaste Fruity, estery 

Bitterness 

Diacetyl 

Sulphidic/ 

Tic 

Hop 

character 

Malty/grainy 

Fig. 20.19 Description analysis: example of spider web plot (Analytica-EBC). Spider web plot 

showing two different beers. No. of assessors, n = 26. Mean scores are shown as distance from 

the centre. The width of the line is the last significant difference about the mean calculated as 

the studentized range, SR. For example, the following ANOVA table was produced for the hop 

character means of two brands. 

Source df SS MSE F-ratio 

Total 51 664.32 ± ± 

Between brands 1 64.32 64.32 5.63 

Error 50 600.00 12.00 ± 

The error bands would be calculated by finding 

SR ˆ Q MSE 

r 

ˆ 2:85 

N 

12 

r 

ˆ 1:933 

26 

Where Q is the upper 5 percentage points for two treatments and 50 degrees of freedom (from 

Malek et al., 1982). 

For each attribute the average rating is calculated and the results presented either as a 

Table or as ahistogram, for example as in Fig. 20.18. An alternative is aspider web plot 

(Fig. 20.19) where mean scores are shown as the distance from the centre. The ranking 

test is used to place aseries of test samples (usually from 3to 6) in rank order according 

to agiven characteristic (criterion). The criterion may be the intensity of asingle sensory 

attribute, or agroup of related attributes, or atotal impression. The test is especially 

suitable in those situations where scale estimates are not meaningful and it is convenient 

to rank aseries of samples according to preference or some other criteria. Assessors (n) 

receive the test samples (k) simultaneously in random order and rank them according to 

the specified criterion. The rank sums (R) are calculated and evaluated statistically with 

the aid of Friedman's test. Specimen answer forms are given in Figs 20.20 and 20.21. In 

preference tests, assessors are instructed to assign rank 1to the preferred sample, rank 2 

to the next preferred, etc. For intensity tests, assessors are instructed to assign rank 1to 


the lowest intensity, rank 2 to the next lowest etc. Friedman's F can be calculated by the 

formula: 

F ˆ ‰…R1 R† 2 ‡ . . . …Rk R† 2 Š 

kR=6 A 

where R is the mean rank sum calculated as 

R ˆ …R1 . . . ‡ Rk†=k ˆ n…k ‡ 1† 

Name ___________________________ Date _____________ 

(day/month/year) 

Products submitted to test _____________________________ 

Problem: You have received 5 samples placed as follows: 

left mid left mid mid right right 

Taste the samples from left to right and write ‘1’ in the box of the 

sample you like best, ‘2’ in the box of the sample you like next best, 

and so on. 

Comments: 

Fig. 20.20 Specimen answer for a ranking test (preference) (Analytica-EBC). 

Ranking Test 

Name ______________________________ Date ______________ 

(day/month/year) 

Products submitted to test __________________________________ 

Problem: You have received 4 samples labelled with the three-digit 

numbers shown in the column marked ‘Samples presented’. Taste the 

samples and place them in rank order according to bitterness, listing 

the most bitter in the column marked 4, the next most bitter in the 

column marked 3, etc. If two samples appear the same, preferably make 

a ‘best guess’ as to their rank order, or if you cannot guess, indicate 

under “Comments” the sample numbers that could not be differentiated. 

Samples presented Order of rank Comments 

1. 2. 3. 4. 

set *1 

149, 251, 347, 428, 347, 428, 251, 149 428 = ~ 251 

set *2 

014, 017, 146, 155, 017, 146, 155, 014 

set *3 

098, 123, 233, 473, 233, 123, 473, 098 123 = ~ 473 = ~ 098 

Comments: 

4 products were compared in set *1, then presented again in sets 2 and 

3 with different codes. Text items which are underlined were filled in 

by the panel leader before the test. Text in italics is the assessor’s response. 

The codes used for this assessor were: Sample A = 347, 146, 233; 

Sample B = 251, 017, 473; Sample C = 428, 014, 098; Sample D = 149, 

155, 123. 

Fig. 20.21 Specimen answer for a ranking test (bitterness) (Analytica-EBC). 


Table 20.10 Ranking test; upper probability points for the 

No. of samples (k) No. of degrees of freedom of 

(df = n 1) 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

From EBC-Analytica (1998). 

2 with 

when nis the number of assessments. If Fexceeds the upper critical value of 

k 1degrees of freedom (Table 20.10) it can be concluded that there is asignificant 

difference between the samples. If practical it is best to prohibit ties as the statistics 

become cumbersome. In the equation above Ais an adjustment for ties; if no ties are 

present A ˆ 0. If ties are present consult Analytica-EBC/ASBC for the statistical 

treatment. The multiple comparison procedure according to Friedman is also given. The 

Least Significant Difference (LSD) within the set of rank sums is given by the formula: 

… 

p kR=3† 

LSDrank ˆt =2:00 

2 -distribution 

where t /2.00 is Student's t, which equals 1.96 at the 5% level and 2.58 at the 1% level of 

significance. Any two rank sums which differ by more than the LSD are significantly 

different. 

Finally, Analytica-EBC/ASBC describe an internationally accepted flavour terminologyfor 

beer.It namesand defines each of122 separately identifiableflavour notes which 

can occur in beer (Table 20.11) The terminology was based on the principles that: 

1. Each separately identifiable flavour characteristic has its own name. 

2. Similar flavours are placed together. 


2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

2 

level of significance, % 

0.05 0.01 

5.99 9.21 

7.81 11.34 

9.49 13.28 

11.07 15.09 

12.59 16.81 

14.07 18.47 

15.51 20.09 

16.92 21.67 

18.31 23.21 

19.67 24.72 

21.03 26.22 

22.36 27.69 

23.68 29.14 

25.00 30.58 

26.30 32.00 

27.59 33.41 

28.87 34.80 

30.14 36.19 

31.41 37.57 

32.67 38.93 

33.92 40.29 

35.17 41.64 

36.41 42.98 

37.65 44.31 

38.88 45.64 

40.11 46.96 

41.34 48.28 

42.56 49.59 

43.77 50.89

Table 20.11 Description of the terminology system 

Particular relevance: O ˆ Odour T ˆ Taste M ˆ Mouth-feel W ˆ Warming Af ˆ Afterflavour 

Class term First tier Second tier Relevance Comments, synonyms, definitions Reference standard 

Class 1: aromatic, fragrant, fruity, floral 

0110 Alcoholic OTW General effect of ethanol and higher alcohols Ethanol, 50 g/1 

0111 Spicy OTW Allspice, nutmeg, peppery, eugenol: see also 1003 Vanilla Eugenol, 120 g/1 

0112 Vinous OTW Bouquet, fusely, wine-like (white wine) 

0120 Solvent-like OT Like chemical solvents 

0121 Plastics OT Plasticizers 

0122 Can-liner OT Lacquer-like 

0123 Acetone OT (Acetone) 

0130 Estery OT Like aliphatic esters 

0131 Isoamyl acetate OT Banana, pear drop (Isoamyl acetate) 

0132 Ethyl hexanoate OT Apple-like with note of aniseed: see also 0142 Apple (Ethyl hexanoate) 

0133 Ethyl acetate OT Light fruity solvent-like: see also 0120 Solvent-like (Ethyl acetate) 

0140 Fruity OT Of specific fruits or mixtures of fruits 

0141 Citrus OT Citral, grapefruit, lemony, orange rind 

0142 Apple OT 

0143 Banana OT 

0144 Blackcurrant OT Blackcurrant fruit; for blackcurrant leaves use 0810 Catty 

0145 Melony OT (6-Nonenal, cis-or trans-) 

0146 Pear OT 

0147 Raspberry OT 

0148 Strawberry OT 

0150 Acetaldehyde OT Green apples, raw apple skin, bruised apples (Acetaldehyde) 

0160 Floral OT Like flowers, fragrant 

0161 2-Phenylethanol OT Rose-like (2-Phenylethanol) 

0162 Geraniol OT Rose-like, different from 0161; taster should compare pure chemicals (Geraniol) 

0163 Perfumy OT Scented (Exaltolide musk) 

0170 Hoppy OT Fresh hop aroma; use with other terms to describe stale hop aroma; 

does not include hop bitterness (see also 1200 Bitter) 

0171 Kettle hop OT Flavour imparted by aroma hops boiled in kettle 

0172 Dry-hop OT Flavour imparted by dry hops added in tank or cask 

0173 Hop oil OT Favour imparted by addition of distilled hop oil 




Class 2: resinous, nutty, green, grassy 

0210 Resinous OT Fresh sawdust, resin, cedarwood, pinewood, sprucy, terpenoid 

0211 Woody OT Seasoned wood (uncut) 

0220 Nutty OT As in brazil nut, hazelnut, `sherry-like' 

0221 Walnut OT Fresh (not rancid) walnut 

0222 Coconut OT 

0223 Beany OT Bean soup (2,4,7-Decatrienal) 

0224 Almond OT Marzipan (Benzaldehyde) 

0230 

0231 

0232 

Grassy 

Freshy cut grass 

Straw-like 

OT 

OT 

OT 

Green, crushed green leaves, leafy, alfalfa } 

Hay-like 

cis-3-Hexanol 

Class 3: cereal 

0310 Grainy OT Raw grain flavour 

0311 Husky OT Husk-like, chaff, Glattwasser 

0312 Corn grits OT Maize grits, adjuncty 

0313 Mealy OT Like flour 

0320 Malty OT 

0330 Worty OT Fresh wort aroma; use with other terms to describe infected wort 

(e.g. 0731 Parsnip/celery) 

Class 4: caramelized, roasted 

0410 Caramel OT Burnt sugar, toffee-like 

0411 Molasses OT Black treacle, treacly 

0412 Licorice OT 

0420 Burnt OTM Scorched aroma, dry mouth-feel, sharp, acrid taste 

0421 Bread crust OTM Charred toast 

0422 Roast barley OTM Chocolate malt 

0423 Smoky OT 

Class 5: phenolic 

0500 Phenolic 

0501 Tarry OT Pitch, faulty pitching of containers 

0502 Bakelite OT 

0503 Carbolic OT Phenol, C 6 H 5OH 

0504 Chlorophenol OT Trichlorophenol (TCP), hospital-like 

0505 Iodoform OT Iodophores, hospital-like, pharmaceutical 


Class 6: soapy, fatty, diacetyl, oily, rancid 

0610 Fatty acid OT 

0611 Caprylic OT Soapy, fatty, goaty, tallowy (Octanoic acid) 

0612 Cheesy OT 

Dry, stale cheese, old hops 

0613 Isovaleric OT Hydrolytic (Isovaleric acid) 

0614 Butyric OT Rancid butter rancidity Butyric acid 3 mg/1 

0620 Diacetyl OT Butterscotch, buttermilk Diacetyl, 0.2±0.4 mg/1 

0630 Rancid OT 

Oxidative rancidity 

0631 Rancid oil OTM 

0640 Oily OTM 

0641 Vegetable oil OTM As in refined vegetable oil 

0642 Mineral oil OTM Gasoline (petrol) kerosene (paraffin), machine oil 

Class 7: sulphury 

} 

} 

0700 Sulphury OT 

0710 Sulphitic OT Sulphur dioxide, striking match, choking, sulphurous-SO 2 (KMS) 

0720 Sulphidic OT Rotten egg, sulphury-reduced, sulphurous- RSH 

0721 H 2S OT Rotten egg (H 2S) 

0722 Mercaptan OT Lower mercaptans, drains, stench (Ethyl mercaptan) 

0723 Garlic OT 

0724 Lightstruck OT Skunky, sunstruck 

0725 Autolysed OT Rotting yeast; see also 0740 Yeasty 

0726 Burnt rubber OT Higher mercaptans 

0727 Shrimp-like OT Water in which shrimp have been cooked 

0730 Cooked OT Mainly dialkyl sulphides, sulphurous-RSR 

vegetable 

0731 Parsnip/celery OT An effect of wort infection 

0732 DMS OT (Dimethyl sulphide) DMS, 100 g/l 

0733 Cooked cabbage OT Over-cooked green vegetables 

0734 Cooked sweet corn OT Cooked maize, canned sweet corn 

0735 Cooked tomato OT Tomato juice (processed), tomato ketchup 

0736 Cooked onion OT 

0740 Yeasty OT Fresh yeast, flavour of heated thiamine (see also 0725 autolysed) 

0741 Meaty OT Broth, cooked meat, meat extract, peptone, yeast broth 


}



Class 8: oxidized, stale, musty 

0800 Stale OTM Old beer, overaged, overpasteurized (Heat with air) 

0810 Catty OT Blackcurrant leaves, ribes, tomato plants, oxidized beer (p-Methane-8-thiol-3one) 

0820 Papery OT Initial stage of staling, bready (stale bread crumb), cardboard, old (5 Methylfurfural, 

beer, oxidized 25 mg/1) 

0830 Leathery OTM Later stage of staling, often used in conjunction with 0211 Woody 

0840 Mouldy OT Cellar-like, leaf mould, woodsy 

0841 Earthy OT Actinomycetes, damp soil, freshly dug soil, diatomaceous earth (Geosmin) 

0842 Musty OT Fusty 

Class 9: sour, acidic 

0900 Acidic OT Pungent aroma, sharpness of taste, mineral acid 

0910 Acetic OT Vinegar (Acetic acid) 

0920 Sour OT Lactic, sour milk: use with 0141 citrus for citrus-sour 

Class 10: sweet 

1000 Sweet OT Sucrose 7.5 g/1 

1001 Honey OT Can occur as effect of beer staling (e.g. odour of stale beer in glass), 

oxidized (stale) honey 

1002 Jam-like OT May be qualified by subclasses of 0140 Fruity 

1003 Vanilla OT Custard powder, vanillin (Vanillin) 

1004 Primings OT 

1005 Syrupy OTM Clear (golden) syrup 

1006 Oversweet OT Sickly sweet, cloying 

Class 11: salty 

1100 Salty T Sodium chloride, 1.8 g/1 

Class 12: bitter 

1200 Bitter TAf (Isohumulone) 



Class 13: mouth-feel 

1310 Alkaline TMAf Flavour imparted by accidental admixture of alkaline detergent (Sodium bicarbonate) 

1320 Mouthcoating MAf Creamy, onctueux (Fr) 

1330 Metallic OTMAf Iron, rusty water, coins, tinny, inky (Ferrous ammonium 

sulphate) 

1340 Astringent MAf Mouth puckering, puckery, tannin-like, tart Quercitrin, 240 mg/1* 

1341 Drying MAf Unsweet 

1350 Powdery OTM O-Dusty cushion, irritating, (with 0310 Grainy) mill room smell 

TM-Chalky, particulate, scratchy, silicate-like, siliceous 

1360 Carbonation M CO 2 content 

1361 Flat M Undercarbonated 60% of normal C0 2 

content for the product 

1362 Gassy M Overcarbonated 140% of normal C0 2 

content for the product 

1370 Warming WMAf See 0110 Alcoholic and 0111 Spicy 

Class 14: fullness 

1410 Body OTM Fullness of flavour and mouth-feel 

1411 Watery TM Thin, seemingly diluted 

1412 Characterless OTM Bland, empty, flavourless 

1413 Satiating OTM Extra full, filling 

1414 Thick TM Viscous, eÂpais (Fr) 

* Quercitrin is both astringent and bitter 


3. No terms are duplicated for the same flavour characteristic. (In five cases, 

overlapping pairs of chemical name terms and generally descriptive terms had to be 

permitted. The five pairs are: 0131 isoamyl acetate and 0143 banana; 0132 ethyl 

hexanoate and 0142 apple; 0133 ethyl acetate and 0120 solvent like; 0613 isovaleric 

and 0612 cheesy; and 0732 DMS and 0734 cooked sweet corn.) 

4. The system is compatible with the EBC Thesaurus for the Brewing Industry. 

5. Subjective terms such as good/bad, young/mature, balanced/unbalanced are not 

included. 

6. As far as possible the meaning of each term is illustrated with readily available 

reference standards. 

The system (Table 20.11) consists of 14 classes given general names to indicate the area 

in which any given type of flavour should be sought. Descriptors carry afour-digit 

number. Some classes have abroader term (e.g. 0700 sulphury) that serves as acommon 

descriptor for all the terms in the class; in other classes asuitable term is not available. 

There are three kinds of descriptors: class terms, first-tier terms and second-tier terms. 

In general the first two contain common terms familiar to most people, and together they 

provide avocabulary designed to fill nonspecialist needs. The flavour wheel (Fig. 20.22) 

is presented to facilitate the location of terms within the system. It is a memory aid and 

not a new system of classification. Despite the diversity of terms, a logical sequence is 

obtained in most cases, but certain discontinuities appear, as where 0700 sulphury follows 

0640 oily. The second tier of terms, together with the reference standards, form the 

theoretical backbone of the system and also serve to define those first-tier terms for which 

a reference is not available, e.g., 0220 nutty comprises a group of flavour notes 

exemplified by walnut like, coconut like, beany, and almond like. The column 

`Relevance' shows that most terms may be used to describe sensations of both odour (O) 

and taste (T). The letters M, W, and Af indicate that the terms may be used to describe 

mouth-feel effects, warming and after flavour. A number of terms that have been used in 

the past are given under `Comments, synonyms, definitions' but their use should be 

discouraged in favour of the more precise description given in the Table. Thus, 0910 

acetic is preferred to `vinegar'. The flavour caused by caprylic and/or capric acids should 

be referred to as 0611 caprylic. The term 0630 rancid is used only for oxidative rancidity 

(carbonyl compounds) and is no longer used for a butyric flavour. 

Analytica-EBC-ASBC provide a list of 27 compounds recommended for use as flavour 

reference standards together with methods of purification, difference thresholds and the 

range of values found in beer. In addition they list 15 compounds that may be suitable 

flavour reference standards after further study. 

It is recognized that terminology will change with usage and the results of research 

and so the system should be brought up to date every few years. Individual breweries may 

well use other terms which help to characterize their beers. Lee et al. (2001) have 

presented a revised flavour wheel for use with whiskies. 

Brown and Clapperton (1978b) examined the terms used to describe ale flavours by 

multi-dimensional scaling ± a technique of grouping like characters together and 

arranging these groups relative to each other by their degree of `similarity'. Such 

correlations cannot be perfect and deviations from the model are expressed as `stress'. 

Thus the seven after-flavour terms (sweet, toffee-like, caprylic, burnt, astringent, and 

mouthcoating) can be represented in a two-dimensional model with only 0.3% stress. In 

contrast for odour and flavour terms a three-dimensional model is required and the stress 

value is approximately 13%. These terms fall roughly on the surface of a sphere so that 


1350 Powdery 

1340 Astringent 

1330 Metallic 

1320 Mouthcoating 

1310 Alkaline 

1200 Bitter 

1100 Salty 

1000 Sweet 

0920 Sour 

0910 Acetic 

0900 Acidic 

0840 Mouldy 

0830 Leathery 

12. Bitter 

11. Salty 

10. Sweet 

9. Sour, acidic 

0820 Papery 

0810 Catty 

1360 Carbonation 

1370 Warming 

13. Mouthfeel 

8. Oxidized, 

stale, musty 

0800 Stale 

0740 Yeasty 

Taste 

0730 Cooked veg. 

1410 Body 

14. Fullness 

0110 Alcoholic 

7. Sulphury 

0720 Sulfidic 

0710 Sulphitic 

0700 Sulphury 

0120 Solvent-like 

0640 Oily 

0130 Estery 

1. Aromatic, fragrant, 

fruity, floral 

Odour 

Odour 

0140 Fruity 

0150 Acetaldehyde 

0160 Floral 

2. Resinous, 

nutty, green, grassy 

6. Soapy, 

fatty, discetyl, 

oily, rancid 

3. Cereal 

4. Caramelized, 

roasted 

5. Phenolic 

0170 Hoppy 

0210 Resinous 

0220 Nutty 

0230 Grassy 

0310 Grainy 

0320 Malty 

0330 Worty 

0410 Caramel 

0420 Burnt 

0500 Phenolic 

0610 Fatty acid 

0620 Diacetyl 

0630 Rancid 

Fig. 20.22 Flavour wheel showing class terms and first-tier terms (Analytica-EBC). 

Toffee-like 

Mouthcoating 

Viscous 

Estery 

Body 

Warming 

Nutty 

H.G. fullness 

Spicy 

Burnt 

Malty 

Smooth 

Worty 

Diacetyl 

Lively 

Sweet 

Cloying 

Fruity 

Smoky 

Bitter 

Aldehydic 

Oily 

Grainy 

Caprylic 

Cooked 

vegetable 

Cardboard 

Watery 

Astringent 

Musty 

Metallic 

Sour 

Sulphury 

Ribes 

Acidic 

Rancid 

Rubbery 

Fig. 20.23 Flavour terms by multi-dimensional scaling (after Brown and Clapperton, 1978b). 


those that are close together on the model (Fig. 20.23) have aclose flavour relationship, 

e.g. sour and acidic, whereas widely different or opposite characteristics such as sweet 

and bitter or body (full) and watery (thin) are further apart or at opposite points on the 

surface. The flavours associated with strong ales are seen on the left-hand side of the 

diagram (Fig. 20.23) while many of those on the right-hand side are associated with 

oxidative deterioration or staling of beer. Terms in the central section of the diagram, 

including diacetyl, caprylic, worty, grainy, lively, bitter, burnt and nutty, may be regarded 

as belonging to an intermediate category of flavours that are pleasant to some and 

unpleasant to others at their normal levels of perceived intensity in commercial beers 

High levels of hop bitter substances, particularly in beers that are fermented to dryness, 

can impart astringency as well as bitterness to the flavour so these terms are found close 

together on the model. 

Brown et al. (1974) used discriminant (cluster) analysis to examine national and 

regional differences in lager beers. In this technique each beer is represented as a point in 

multi-dimensional space, the coordinates of which are determined either by the individual 

flavour characteristics, determined by profile analysis, or by physicochemical parameters, 

determined by chemical analysis. Twenty-seven terms gave significant scores with lagers 

and the pattern of points in 27-dimensional space is simplified by a computer program to 

produce eigenvectors (mathematical devices to convert a pattern of points in multidimensional 

space into an equivalent pattern of points in a smaller number of 

dimensions). This has the advantage of bringing things down to a level nonmathematicians 

can visualize but has the disadvantage that the axes only represent 

mathematical abstractions and not brewing parameters. Thus, the two-dimensional 

pattern of North American, Continental European and British lagers gave three discrete 

tight clusters of points. When only the 12 highest scoring sensory characteristics were 

used the beers still fell into three groups but the clusters were more diffuse (Fig. 20.24). 

This result can be interpreted in terms of the perceived differences in bitterness, dimethyl 

sulphide (DMS) flavour, and palate fullness (OG). Of the flavour terms scored: (i) 

dimethyl sulphide and cabbagy-vegetable water both relate to the DMS factor; (ii) body, 

warming, and high gravity fullness, and viscous relate to differences in original gravity; 

B 

A 

x 

(a) (b) 

C 

y 

o 

B 

Low OG 

and low 

DMS 

content 

Low 

bitterness 

A 

High 

bitterness 

High OG and 

high DMS 

content 

Fig. 20.24 Discriminant analysis of sensory data on thirty-three lager beers. (a) Result; (b) 

Interpretation of result. Code: A = North American beers, B = British beers, and C = 

Continental European beers (after Brown and Clapperton, 1978a). 


C

Table 20.12 Results of cluster analysis of sensory and physiochemical data on beer (after 

Clapperton, 1979) 

Odour items Physiochemical parameters Flavour terms 

Original gravity 

Isocamyl alcohol Body 

2-Methylbutanol Estery 1 

Total carbohydrates Fruity 

Dextrins Viscous 

Potassium 

Propanol 

Mouth coating 

Warming 2 

4 Estery Ethyl acetate Spicey 

Isoamyl acetate * Watery 

Phosphate 

12 Ribes Air content 

16 Hoppy Dodecanoic acid 

Tetradeavoic acid 

Total fatty acids 

Octanoic acid Caprylic 4 

Decanoic acid 

Sweet 

Present gravity Cloying 

Toffee-like 5 

* Astringent 

Bitterness Bitter 9 

Sulphur Metallic 17 

* Negatively correlated with other terms and parameters in the same cluster 

and (iii) bitter, drying, and possibly smooth (mouth-feel) relate to differences in 

bitterness. 

Brown and Clapperton (1978a) also examined 46 ales (OG 1030±1050) from five 

brewing companies by sensory profile analysis and by instrumental analysis. The most 

important variables in the discriminant analysis were: (i) isoamyl alcohol content 

(instrumental), (ii) caprylic flavour (sensory), (iii) sodium content (instrumental); (iv) 

meaty aroma (sensory), (v) ethyl acetate content (instrumental), (vi) bitter after-flavour 

(sensory), (vii) dextrin content (instrumental), (viii) smoky flavour (sensory), and (ix) 

DMS content (instrumental). On the basis of these parameters 87% of the ales were 

correctly assigned to their brewing company. Similarly, the ales could be assigned to their 

gravity band either instrumentally on the basis of alcohol or dextrin content or by sensory 

analysis using 13 parameters. The sensory parameters in order of importance were: (i) 

body, (ii) aldehyde (odour), (iii) high gravity fullness, (iv) viscous (thick), (v) estery, (vi) 

meaty (odour), (vii) cooked vegetable (odour), (viii) rubbery, (ix) caprylic, (x) DMS 

(odour), (xi) fruity, (xii) cloying, and (xii) sour. 

Clapperton (1979) also carried out another cluster analysis using both sensory terms and 


physicochemical parameters and the results are given in Table 20.12, where the numbers 

refer to the order of formation of the clusters, i.e., the lower the number the better the 

correlation. Clusters which only contained sensory terms were omitted. The first two 

clusters isolate terms that relate to esters, alcohols and original gravity. Although there is a 

curious reversal of ester and alcohol contents and the corresponding flavour effect between 

clusters 1 and 2, estery (odour) is grouped, as expected, with ethyl and isoamyl acetate 

content in cluster 4. Apart from cluster 16 the clusters indicate causative relationships 

between the physicochemical parameters and the corresponding sensory terms. 

Principal component analysis is another statistical technique that Clapperton and 

Piggott (1979) applied to the results of profile analysis. Ales and lagers were examined 

and two-dimensional plots of the results using the first two principal components as axes 

showed resolution of the ales from the lagers and the close proximity of duplicate 

samples. Piggott and Jardine (1979) used principal component analysis to differentiate 

various brands of whisky. Most workers in this field have excluded hedonic expressions 

but Moll et al. (1978) used principal component analysis to classify Continental European 

beers as good, average or poor on the basis of nine physicochemical parameters: colloidal 

stability (7 days at 40 ëC/1 day at 0 ëC), cold sensitivity (24 h at 0 ëC), brightness at 12 ëC, 

six months test, the content of -phenylethanol, ethyl caprylate, isoamyl acetate and 

isobutanol and foam stability. 

Hoff et al. (1978) used headspace analysis to determine the levels of the isoamyl 

alcohols, isobutanol, ethyl acetate and isoamyl acetate in beers ( -phenylethanol and ethyl 

caprylate are not sufficiently volatile to be measured by headspace analysis). To compare 

two beers, the peak areas on the chromatogram (excluding ethanol and the internal 

standard xylene) were expressed as a percentage of the total peak area. A chronologically 

updated data base was used to calculate the standard deviation for each peak and a twotailed 

t-test was performed on the mean values from duplicate determinations on the two 

beers. These results were used to predict the results of triangular taste tests: 

1. If none of the peaks is significantly different between samples at 0.005 risk, one 

predicts that the tasting panel results will be insignificant at 0.05 risk. 

2. If one or more peaks are significantly different between samples at 0.001 risk, one 

predicts that the panel results will be significant at a risk of 0.05 or less. 

3. If one or more peaks are significantly different between samples at 0.005 risk but 

insignificant at 0.001 risk, no prediction is made and the sample number increased. 

The taste panel found 200 significant results out of the 234 predicted and 70 

insignificant results out of the 76 predicted and it was suggested that tastings could be 

reduced by eliminating samples that were similar by headspace analysis. However, it was 

acknowledged that differences due to sulphur compounds, staling or certain hop 

compounds may go undetected. Nevertheless the headspace/statistical method predicted 

and the tasting panel found significant differences between beers produced at three 

branch plants. 

Compared to chemical assays, flavour results from human assessors have been 

regarded as unreliable but Hegarty and White (1993) have applied two-way analysis of 

variance (ANOVA) statistics to the results of a flavour profile panel. This can establish 

the degree of variability due to assessor differences, the degree of variability due to 

differences between samples and the amount of variability that cannot be explained 

(`noise'). Results can be expressed as the F-ratio (Fisher ratio) which, if large, shows a 

significant difference. Thus, ideally, the F-ratio for the assessors should show a low value 

indicating that the score attributed to a given sample does not vary greatly between 


tasters. This approach can be expanded to monitor the performance of individual 

assessors. The mean score can show whether the assessor is scoring high or low relative 

to the rest of the group. The standard deviation is a measure of scoring consistency and 

the F-ratio indicates whether the assessor can distinguish samples consistently. The 

statistics showed that tasters do not perform equally on all characters. The key to reliable 

flavour results is on-going performance monitoring and on-going flavour training. 


ACREE, T. E. and TERANISHI, R. (eds) (1993) Flavor Science ± Sensible Principles and Techniques. 

American Chemical Society, Washington DC. pp. xvi + 352. 

AKABAS, M. H., DODD, J. and AL-AWQATI, Q. (1988) Science, 242, 1047. 

AMERICAN SOCIETY FOR TESTING AND MATERIALS (1979) Standard practice for Determination of Odour 

and Taste Thresholds by Forced Choice Ascending Concentration Series. E 67979. ASTM, 

Philadelphia, PA. 

AMERINE, M. A., PANGBORN, R. M. and ROESSLER, E. B. (1965) Principles of Sensory Evaluation of Food. 

Academic Press, New York. 

AMOORE, J. E. (1991) in Smell and Taste in Health and Disease (ed. Getchell, T. V. et al.) p. 655. 

BOUDREAU, J. C. (ed.) (1979) Food Taste Chemistry. ACS Symposium No. 115. American Chemical 

Society, Washington DC, 262pp. 

BRESLIN, P. A. S. (2001) Flavour Fragr. J. 16, 439. 

BROWN, D. G. W. and CLAPPERTON, J. F. (1978a) J. Inst. Brewing, 84, 318. 

BROWN, D. G. W. and CLAPPERTON, J. F. (1978b) J. Inst. Brewing, 84, 324. 

BROWN, D. G. W., CLAPPERTON, J. F. and DALGLIESH, C. E. (1974) Proc. Annu. Meet. Am. Soc. Brew. 

Chem., p. 1. 

BROWN, D. G. W., CLAPPERTON, J. F., MEILGAARD, M. C. and MOLL, M. (1978) J. Amer. Soc. Brew. Chem., 

36, 73. 

CHEVANCE, F., GUYOT-DECLERCK, C., DUPONT, J. and COLLIN, S. (2002) J. Agric. Food Chem., 50, 3818. 

CLAPPERTON, J. F. (1974) J. Inst. Brewing, 80, 164. 

CLAPPERTON, J. F. (1976) J. Inst. Brewing, 82, 175 

CLAPPERTON, J. F. (1978) J. Inst. Brewing, 84, 107. 

CLAPPERTON, J. F. (1979) in Land, D. G. and Nursten, H. E. Progress in Flavour Research. Applied 

Science Publishers, London, p. 1. 

CLAPPERTON, J. F. and BROWN, D. L. W. (1978) J. Inst. Brewing, 84, 90. 

CLAPPERTON, J. F. and PIGGOTT, J. R. (1979) J. Inst. Brewing, 85, 271. 

DALGLIESH, C. E. (1977) Proc. 16th Congr. Eur. Brew. Convn. Amsterdam, p. 623. 

DE COOMAN, L., AERTS, G., OVERMEIRE, H. and DE KEUKELEIRE, D. (2000) J. Inst. Brewing, 106, 169. 

DELWICHE, J. F., BULETIC, Z. and BRESLIN, P. A. S. (2001) Perception and Psychophysics, 63, 761. 

DOTY, R. L. (ed.) (1995) Handbook of Olfaction and Gustation. Marcel Dekker, New York. 

EVANS, D. J., SCHMEDDING, D. J. M., BRUIJNJE, A., HEIDEMAN, T. and KING, B. M. (1999) J. Inst. Brewing, 

105, 301. 

FERNANDUS, A., OOSERAM-KLEIJNGELD, I. and RUNNEBOOM, A. J. M. (1970) Tech. Quart. MBBA, 7, 210. 

GARDNER, R. J. (1978) J. Pharm. Pharmacol., 30, 351. 

GARDNER, R. J. (1979) Tech. Quart. MBAA, 16, 106, 148, 204. 

GIENAPP, E. and SCHROÈ DER, K. L. (1975) Die Nahrung, 19, 697. 

GIJS, L. and COLLIN, S. (2002) J. Amer. Soc. Brew. Chem., 60, 68. 

GIJS, L., PERPEÁ TE, P., TIMMERMANS, A. and COLLIN, S. (2000) J. Agric. Food Chem., 48, 6196. 

GIJS, L., CHEVANCE, F., JERKOVIC, V. and COLLIN, S. (2002) J. Agric. Food Chem., 50, 5612. 

GIVEN, P. and PARADES, D. (eds.) (2002) Chemistry of taste ± Mechanisms, Behaviors and Mimics. ACS 

Symposium No. 825. American Chemical Society, Washington, DC. 

GOIRIS, K., DE RIDDER, M., DE ROUCK, G., BOEYKENS, A. VAN OPSTAELE, F., AERTS, G., DE COOMAN, L. and 

DE KEUKELEIRE, D (2002) J. Inst. Brewing, 108, 86. 

GOODE, J. (2003) Wine Magazine, April, 2003, p. 44. 

GOODENOUGH, P. W. (1998) Int. J. Food Sci., 33, 63. 

GRAY, P. P. and STONE, I. (1939) J. Inst. Brewing, 45, 253. 

GUADAGNI, D. G. (1970) quoted by Teranishi, R. in Ohloff, G. and Thomas, A. F. (eds) (1971) Gustation 

and Olfaction Academic Press, New York, p. 170 

HEGARTY, P. K. and WHITE, F. H. (1993) Proc. 24th Congr. Eur. Brew. Convn. Oslo, p. 429 

HEGARTY, P., CHILVER, J. and THREAPLETON, L. (2001) Proc. 28th Congr. Eur. Brew. Convn. Budapest 

Paper 89. 

HOFF, J. T., CHICOYE, E., HERWIG, W. C. and HELBERT, J. R. (1978) in Charalambous, G. (ed.) (1978) 


Analysis of Foods and Beverages Headspace Techniques, Academic Press, New York, p. 187. 

HOUGH, J. S. (1990) in An Introduction to Brewing Science and Technology. Series II, Volume 3. Quality. 

Institute of Brewing, London. 

HUGHES, P. S. and BOLSHAW, L. H. (1995) Proc. 25th Congr. Eur. Brew. Conv. Brussels, p. 151. 

INSTITUTE OF BREWING (1995) Sensory Analysis Manual, 44 pp. 

INTERNATIONAL ORGANIZATION FOR STANDARDIZATION (1983) International Standard 6658. Sensory 

Analysis, Methodology, General guidance. ISO, Paris. 

JONES, F. N. (1956) Amer. J. Psychology, 69, 672. 

KANEDA, H., KANO, Y., KOSHINO, S. and OHYA-NISHIGUCHI, H. (1992) J. Agric. Food Chem., 40, 2102. 

KANEDA, H., SHINOTZUKA, K., KOBAYAKAWA, T., SAITO, S. and OKAHATA, Y. (2001) J. Amer. Soc. Brew. 

Chem., 49, 167. 

KANEDA, H., WATARI, J., TAKASHIO, M. and OKAHATA, Y. (2003) J. Inst. Brewing, 109, 27. 

KIER, L. B. and HALL, L. H. (1976) Molecular Connectivity in Chemistry and Drug Research. Academic 

Press, New York. 

LACAN, F., SOULET, S., ARNAUDINAUD, V., NAY, B., VERGEÂ , S., CASTAGNINO, C., DELAUNAY, J.-C., CHEÁ ZE, C. 

and VERCAUTEREN, J. (2000) Cerevisia, 25 (4), 35. 

LANGSTAFF, S. A. and LEWIS, M. J. (1993) J. Inst. Brewing, 99, 31. 

LAWLESS, H. and HEYMANN, H. (1999) Sensory Analysis of Foods: Principles and Practice. Kluwer/ 

Plenum, New York. 

LEE, K.-Y. M., PATTERSON, A., PIGGOTT, J. R. and RICHARDSON, G. D. (2001) J. Inst. Brewing, 107, 287. 

LERMUSIEAU, G., NOEÈ L, S., LIEÂ GEIS, C. and COLLIN, S. (1999) J. Amer. Soc. Brew. Chem., 57, 29. 

LEWIS, M. J., PANGBORN, R. M. and FUJII-YAMASHITA, J. (1980) Proc. 16th Conv. Australian and New 

Zealand section of the Institue of Brewing, Sydney, p. 165. 

MACKIE, A. E. and SLAUGHTER, J. C. (2002) J. Inst. Brewing, 108, 336. 

MALEK, D. M., SCHMIDT, D. J. and MUNROE, J. H. J. (1982) J. Amer. Soc. Brew. Chem., 40, 133. 

MEILGAARD, M. C. (1975) Tech. Quart. MBAA, 12, 107, 151. 

MEILGAARD, M. C. and REID, D. S. (1979) in Land, D. G. and Nursten, H. E. (eds) Progress in Flavour 

Research. Applied Science Publishers, London, p. 67. 

MEILGAARD, M. C., CIVILLE, G. V. and CARR, B. T. (1987) Sensory Evaluation Techniques. CRC Press, 

Boca Raton FL. 

MOLL, M., VINH, T. and FLAYEUX, R. (1978) in Charalambous, G. (ed.) Flavours of Foods and Beverages ± 

Chemistry and Technology. Academic Press, New York, p. 329. 

MONTMAYEUR, J. P. and MATSUNAMI, H. (2002) Current Opinion in Neurobiology, 12, 366. 

NIEMAN, C. (1960) quoted in Amerine, M. A., Pangborn, R.M. and Roessler, E. B. (1965) Principles of 

Sensory Evaluation of Food. Academic Press, New York, p. 95. 

NOEÈ L, S., METAIS, N., BONTE, S., BOPART, E., PELADAN, F., DUPIRE, S. and COLLIN, S. (1999) J. Inst. Brewing, 

105, 269. 

OHLOFF, G. (1994) Scent and Fragrances ± The Fascination of Odours and their Chemical Perspectives. 

Springer Verlag, Berlin, pp. xii + 238. Translated from the German Riechstoffe und Geruchssinn 

(1990) by Pickenhagen, W. and Lawrence, B. M. 

PANGBORN, R. M. (1959) Amer. J. Clin. Nutrition, 7, 280. 

PANGBORN, R. M. (1980) in Koivistoinen, P. and HyvoÈnen. L. Carbohydrate Sweeteners in Food and 

Nutrition, Academic Press, New York, p. 87. 

PERPEÁ TE, P. and COLLIN, S. (1999) J. Agric. Food Chem., 47, 2374. 

PIGGOTT, J. R. (ed.) (1988) Sensory Analysis of Foods. 2nd edn, pp. x + 422. Elsevier Applied Science, 

London. 

PIGGOTT, J. R. and JARDINE, S. P. (1979) J. Inst. Brewing, 85, 82. 

RICHTER, C. P. and MACLEAN, A. (1939) Amer. J. Physiol., 126, 1. 

ROESSLER, E. B., PANGBORN, R. M., SIDEL, J. L. and STONE, H. J. (1978) Food Science, 43, 940. 

ROUSEFF, R. L. (ed.) (1990) Bitterness in Foods and Beverages. Elsevier, Amsterdam, pp xviii + 356. 

SMYTHE, J. E., O'MAHONY, M. A. and BAMFORTH, C. W. (2002) J. Inst. Brewing, 108, 37. 

SWAIN, T. and CASEY, J. C. unpublished results quoted in Reynolds, T. M (1963) Adv. Food Research, 12, 

1. 

TEETER, J. H. and GOLD, G. H. (1988) Nature, 331, 298. 

TORLINE, P. A., DERCKSEN, A. W., AXCELL, B. C. and JOHNSTONE, W. (1999) Proc. 7th Conv. Inst. Brewing 

Africa Section, Nairobi, p. 50. 

TRESSL, R., BAHRI, D. and KOSSA, M. (1980) in Charalambous, G (ed.) The Analysis and Control of Less 

Desirable Flavours in Foods and Beverages. Academic Press, New York, p. 293. 

UCHIDA, M. and ONO, M. (1999) J. Amer. Soc. Brew. Chem., 57, 145. 

VAN DER HEIJDEN, A. (1993) in Acree, T. A. and Teranishi, R. (eds) Flavor Science ± Sensible Principles 

and Techniques. American Chemical Society, Washington, DC, p. 67. 

VAN OEVELEN, D., DE L'ESCAILLE, F. and VERACHTERT, H. (1976) J. Inst. Brewing, 82, 322. 

WEISS, A., SCHOÈ NBERGER, CH., MITTER, W., BIENDL, M., BACK, W. and KROTTENTHALER, M. (2002) J. Inst. 

Brewing, 108, 236. 


21 

Packaging 


Beermustbepackagedbeforeitissold.Toensurethebestpossiblequalityoftheproduct, 

packaging must be carried out with skill and care. Only if packaging is effectively 

performedwilltheproductbeacceptable.Beercanbeputintoanumberofpackages.The 

most important world-wide is the bottle. Bottles are of two types: returnable and nonreturnable. 

The most used isthe returnable bottle but indeveloped markets in Europe and 

theUSAthenon-returnablebottleisprevalent.Beerisalsofilledintocans,kegsandcasks. 

Usually adistinction is made between draught beer, i.e., in kegs or casks and `small-pack 

beer' in bottles and cans. There are differences between countries in the relative 

proportionsof different packages(Table21.1). The UK and Ireland are unusualin having 

most of their beer on draught. The UK is further unusual in that of the 64% of its beer sold 

on draught in 1998, 11% was conditioned in the cask and not filtered in the brewery. Cask 

conditioned beer demands very different packaging from keg beer. 

In the mature beer markets of Western Europe, Australia and the USA sales of beer in 

recent years have shown little total growth. In the UK there has been a decline of about 

1% per annum for the last ten years. This has resulted in product differentiation efforts 

being focused on packaging, particularly on small pack beers. This has coincided with an 

increase in beer consumed at home with the consequent domination of this trade by 

supermarket groups. 

Packaging is influenced by environmental issues, which are stronger in some countries 

than others. In some cases a revival in the use of returnable glass bottles and the 

outlawing of selling beer in cans has occurred. Most countries now have packaging 

legislation, which seeks to control the use of packaging material and to reduce waste. 

This sometimes leads to conflict with marketing where packaging plays such a huge role 

in product attractiveness. Packaging is the most labour-intensive part of the brewing 

process. The machinery for packaging beer has become progressively more complex with 

the object of reducing labour costs and preserving product quality. Capital employed in 

packaging is usually the highest of the brewing operations. The efficiency of operation of 

packaging machinery is of critical importance to a profitable brewery. 


Table 21.1 Relative proportion of beer sold in draught and small-pack form for important beerproducing 

countries in 2000 (BLRA, 2002) 

Country Production ('000hl) Draught sales (%) Small-pack sales (%) 

USA 233,521 9 91 

China 220,485 5 95 

Germany 110,429 19 81 

Brazil 82,600 1 99 

Japan 71,727 16 84 

UK 55,729 62 38 

The Netherlands 25,072 30 70 

Czech Republic 17,924 46 54 

Australia 17,326 24 76 

Ireland 8,710 78 22 

Denmark 7,460 11 89 

It is essential to keep records of packaging operations. These relate to the strength and 

type of the beer packaged and to the volume of the beer. Everywhere there is legislation 

governing the contents of the beer in the package for sale. This may relate to average or 

to aguaranteed minimum content. The records will be audited by officials. In most 

countries atax (excise duty) is taken relating to the strength of the beer. The packaging 

department or warehouse keeps the records on which this is based (Chapter 22). 

Packaging is thus of fundamental importance in the supply of beer and is pivotal in 

ensuring the customer is satisfied in terms of quality, quantity and legality. The 

preparation of beer for packaging was discussed in Chapter 15. The packaging options 

available, particularly for bottles and cans, are now numerous and involve different types 

of multi-pack presentation using cardboard and plastic. Brands can establish an identity 

based on the package alone and this is frequently as important as the identity created by 

the taste of the beer. This chapter deals with the underlying principles of successful 

modern packaging operations. 

21.2 General overview of packaging operations 

Apackaging line is aseries of machines designed to fill containers with beer and present 

those containers (packages) to the warehouse. The detailed design of the line will depend 

on the type of package (bottle, can, keg, or cask), the required rate of packaging and the 

types of beer to be packaged. There will also be machines to deal with any secondary 

packaging required which is usually specified by the customer. 

Modernsmall-packbeerfillersoperateatveryhighrates,bottlingatover1000bottles/ 

min. and canning at 2000 cans/min. is common. As there is little storage space in 

breweries for empty cans and bottles aconstant stream of bottles or cans to the site is 

needed. This could mean around 30 vehicles/24hcarrying 26 pallets of empty containers 

arriving at the brewery. Manufacturers of packages are frequently located near to the 

packaging plant. The pressure to supply is not so intense with returnable bottles and kegs 

butthe recovery of empties from the trade must be arranged. These logistics (Chapter 22) 

are very important in the management of packaging. 

Twoseparateflowsmustbedealtwithinanypackagingplant:theflowofthebeerand 

the flow of the containers both empty and full. Thus the mechanical engineering of 

machinery with large moving parts and its proximity to aperishable foodstuff must be 

considered. The handling of the container as well as the handling of the beer must be 


optimized. Three main aspects characterize all successful packaging operations: 

· Preventing air getting into the beer is essential. All the precautions followed in 

producing bright beer must be maintained. Probably the final product specification for 

dissolved oxygen in the beer will be < 0.2 mg/l or, in some cases, < 0.1 mg/l, so during 

filling operations the pick-up of oxygen must not exceed 0.02±0.03 mg/l. If this 

oxygen level is not achieved serious flavour deterioration will result. This control of 

oxygen is therefore a major feature of good packaging. Control of oxygen on filling is 

not important with casks. Cask beer contains live yeast (1±4 million cells/ml) for 

secondary fermentation and any oxygen present in the beer is rapidly scavenged by the 

yeast. 

· The temperature of the beer after conditioning and on filling the bright beer tank is 

likely to be 1 to 0 ëC (30±32 ëF) and the carbon dioxide content may be 2.1±2.7 

volumes depending on whether it is destined for keg or small pack. The pressure on 

the beer must be maintained and the temperature rise controlled to < 2 ëC (3.5 ëF) to 

keep carbon dioxide in solution. Carbon dioxide loss is a serious problem, which can 

disrupt beer supply as beer is held in the warehouse pending re-processing. This is 

expensive and potentially damaging to customer service. 

· The final major factor common to all filling plants is cleanliness. All the plant, not just 

that in contact with the beer, must be regularly and thoroughly cleaned. 

21.3 Bottling 

Worldwide most beer is drunk from bottles, either returnable or non-returnable. The 

filling of these bottles is essentially the same. The difference in equipment needed relates 

to the handling of the used and returned empty bottles and washing them prior to filling. 

Many of the principles of successful bottling apply equally to canning, and to some 

extent, to kegging. Details will be discussed in this section on bottling and differences 

highlighted in the sections on canning and kegging. 

A successful bottling line should allow the brewer to: 

· maintain the dissolved oxygen level in the beer to at least < 0.2 mg/l, although there 

are now reports of plant able to meet a specification of < 0.05 mg/l (Parsons, 2000) 

· ensure the beer is supplied to the customer containing no viable micro-organisms 

· operate with the minimum number of stoppages to keep losses to < 1.5% of the total 

brewery loss and lower the headspace air content to < 2ml/l and provide the highest 

efficiency. 

The bottling line consists of a series of machines and processes: 

· depalletizer 

· decrater 

· washer 

· empty bottle inspection 

· flash pasteurization or sterile filtration 

· filler 

· crowner 

· tunnel pasteurization 

· full bottle inspection 

· labeller 


· crater 

· palletizer 

· cleaning 

If the beer is flash pasteurized or sterile filtered then the subsequent operations must be 

aseptic. The flow of the beer relates to filling and, if in use, sterile filling or flash 

pasteurization. All the other machines are associated with the flow of containers both full 

and empty. Normally there is asystem for in-place cleaning (CIP) that is naturally more 

rigorouswhensterilefillingisbeingused.Themostimportant machineisthefillerandto 

satisfy the criteria above this must operate with the minimum of stops and on agiven run 

of product, operate continuously. Storage capacities before and after the filler must be 

designed to allow continuous filling. This can take up alot of space. If the efficiency of 

thefilleristakenas100%thenthemachinesoneithersideassociatedwithcontainerflow 

must have operational capacities of 110±140% to ensure uninterrupted filling. 

21.3.1 Managing the bottle flow 

The flow of containers has to be managed before and after filling and labelling. We are 

concernedwith:depalletizing;decrating; washingorrinsing;emptybottleinspection;full 

bottle inspection; labelling; crating or other secondary packaging; palletizing. Packaging 

is labour intensive. The numbers of people employed have been reduced by automation 

and by careful planning of the layout of the machinery. This must take account of the 

organization of the warehouse for the receipt of full and empty goods. The number of 

work stations on the packaging line should be minimized and hence the numbers 

employed in the packaging department should be reduced. The machines involved in 

palletizing and crating of full containers and depalletizing and decrating of empty 

containers are similar in principle. To reduce labour involvement they must be located 

together to form one work station (Fig. 21.1). This type of organization has allowed a 

productivity of 0.2 million hl/year/person in a Japanese bottling hall operating at 72,000 

bottles/h (Yokoi et al., 1991). Improvements on this figure are now being achieved. 

Bottling of returnable bottles involves receiving crates of dirty bottles on pallets 

separating the crates from the pallets and the bottles from the crates and then washing the 

empty bottles. Bottling into non-returnable bottles involves receiving new glass bottles, 

which only require rinsing not the thorough cleaning associated with washing dirty 

bottles. 

Depalletizing and palletizing 

Palletizers and depalletizers are normally closely situated and operated by one man. The 

machines must interface with forklift trucks in the warehouse. The efficient operation of 

these machines is critical to overall efficiency of the line. They are normally capable of 

working at rates of 1.4 filling rate (filling ˆ 1). Dirty returnable bottles are usually 

received at the brewery in crates. New glass bottles for the non-returnable trade are 

received on cardboard trays. These crates or trays are most frequently formed onto 

pallets, which can be wooden in which case there are layers of crates built onto the pallet 

or there can be plastic spacer boards separating layers of crates or trays. Plastic spacer 

boards are less robust than wooden pallets but are lighter and take up less space. 

The depalletizer removes the crates or trays and presents them to the decrating 

machine. This is achieved by a lifting device fixed to a frame holding a loading head. The 

loading head loads or unloads the whole layer at once. Lifting heads can be complex and 


New bottles 

Returned 

bottles 

Filled bottles 

in warehouse 

Depalletizer Decrater Sorter Washer 

Repalletizer 

Recrater 

Labeller 

Check 

fill 

Inspector 

can use clamps or hooks to effect the lift. A mixture of pneumatic or hydraulic rams and 

electric motors provides motive force. Position sensing uses micro-switches or photoelectric 

cells. The machine is complex and requires regular maintenance to ensure 

efficient operation. These machines must be working to design criteria to ensure overall 

line efficiency. Removal of the spacer board can be effected mechanically but these 

machines have to locate the position accurately and repetitively and this does not always 

happen resulting in much frustration. It is often decided, therefore, to remove the spacer 

board or the pallet manually. Obviously the machines involved in palletizing and 

depalletizing are identical but operating in the alternative mode. 

Decrating and crating 

Having disassembled the crate or tray the bottles must now be removed and presented to 

the next machine in the line, the washer. The crating operation at the other end of the line 

will need a machine but depending on the detail of the type of package it may operate 

differently from the decrater. Removal or packing of bottles is effected by `gripper heads' 

mounted on a frame. For efficiency these machines should function at 1.25 filling 

(filling ˆ 1). The machines can operate in a batch format or continuously. The choice 

should be made with reference to the efficiency obtained. These machines have many 

moving parts and need considerable maintenance. Machines that work continuously are 

less widely used in breweries because different bottle sizes require setting changes. In 

continuous machines the gripper heads run along a track in a synchronous movement and 

are constantly lowered or raised to deal with the constantly moving crates. In the batch 

process the bottles are removed or placed in the crate in a discrete step, as the crate is 

stationary. Either system will work effectively. 

Secondary packaging 

Returnable bottles are generally sold for consumption in bars and other licensed premises. 

As such the bottles are supplied in crates. The crate is not displayed to the customer and 

therefore has no marketing significance. Non-returnable bottles are sold for consumption 

at home. These are frequently bought through supermarkets where there is scope for 

elaborate secondary packaging using cardboard containers, which can display the brand 

logo and colours to the purchaser. Secondary packaging is arranged to present the bottles 

in a variety of `multi-packs' to satisfy the customer and the marketing department. Thus 

bottles can be packed in sleeves of 1 4, 2 2, 2 3, or 2 5 combinations (Parsons, 

2000; Wainwright, 1999). These clusters can then be packed into final packages of 8, 12, 

Filler 

Crowner 

Pasteurizer 

Fig. 21.1 Arrangement of equipment in a bottling hall (Hough et al., 1982). 


15, 18, 20 or 24 bottles. Corrugated or plain cardboard is used and shrink-wrapping can 

beaddedforfinalprotection.Thepossibilitiesofmulti-packingarealmostendless.Itisin 

thisareaofthebusinessthatmostexpenditureonresearch anddevelopmentisnowmade. 

This is avery important part of the fight for brand supremacy. 

Washing 

This is acritical part of returnable bottling. Bottles return to breweries in various states, 

they will be dirty and will still have labels attached. The labels may be made of paper or 

metallized foils, which may contain tin or aluminium. Bottle cleaning must remove all 

thelabels,andcleanandsterilizethebottles,whicharethenpresentedtothefiller.Thisis 

an aggressive operation and over time the bottles will become scuffed and unattractive. 

This has been aproblem in developed markets where the appearance of the package is so 

important. This has led to the rise in popularity of the one-trip non-returnable bottle, 

which can be decorated to avery high standard. This type of package, unless re-cycled, is 

less environmentally acceptable than the returnable bottle. 

Important factors in all bottle washers are: 

· soaking to remove dirt and labels 

· jetting to rinse 

· temperature 

· strength of detergent, which is normally alkaline. 

The total time of the operation is normally 10±15 minutes. The sequence of operations 

varieswithdifferentwashers.Atypicalsystemistosoakthedirtybottlesinhotwaterand 

then pass them through ahot caustic soda solution. Bottles are then successively rinsed 

with hot caustic solution, hot water, and finally cold fresh water. Bottles are transported 

through the machine in rows of perhaps 50±70. The bottles are assembled in lines so that 

they cannot fall over. Efforts are made to keep the noise as low as possible (

Fresh 

water 

Bottles out 

Bottles in 

Sewer 

Sewer 

Crate Sewer Sieve cleaning 

washer 

mechanism 


m 

2 

1 

8 

7 6 5 4 

3 

Sewer 

Fig. 21.2 Bottle-washing machine (by courtesy of Krones). 

To 

sedimentation 

tank 

1 

2 

3 

4 

5 

6 

7 

8 

Pre-soak 

Pre-jetting 

Main caustic 

Post caustic 

Warm water 1 

Warm water 2 

Cold water 

Fresh water

must be used to prevent the deposit of calcium carbonate scale. Acid scale preventing 

additives and chelating agents such as gluconates are added to the rinsing zones to 

keep calcium in solution and phosphates are added as wetting agents. Aluminium in 

some foil labels releases hydrogen gas from the caustic solution. This must be vented 

from the washer. About 90% of the caustic solution can be used again without 

treatment as it is only lightly contaminated. However, about 10% of the solution is 

heavily contaminated with paper, colloids, colour pigments, label adhesives, metal 

salts and oils. The solids are usually removed by settling in an insulated tank. If 

bottling is being performed on a two-shift basis (up to 16 h/day) this settling can be 

carried out overnight. If settling is carefully performed then the detergent can last for 

very long times in the bottle washer. 

Caustic bottle-washing solutions are slowly corrosive to glass bottles. This causes 

dissolution of soda lime glass by etching. Scuffing also occurs by the physical abrasion of 

the bottles as they rub together on the line, causing visible wear rings on the shoulder and 

base of each bottle. This results in a population of etched, scuffed bottles, which cannot 

display the beer to its best effect. This also applies to the ceramic labels, which are 

sometimes applied directly to the glass. Detergents have been formulated to reduce the 

etching and so to enhance the life and appearance of bottles (Rouillard, 1999; Rouillard 

and Howell, 1999). It is anticipated that reduced etching will result in reduced scuffing, 

as the glass is stronger. 

EDTA and phosphates present in caustic bottle-washing solutions accelerate corrosion 

of the glass (Rouillard and Howell, 1999). A corrosion inhibitor (Divobrite Integra) has 

been shown to reduce etching to an extent that the appearance of a bottle after 30 trips 

was equivalent to that after 15 trips following the use of a conventional bottle-washing 

detergent containing EDTA and phosphate. This work is important because the returnable 

bottle is environmentally friendly but its use is limited by the intense demand for a 

`perfect' package, which the returnable bottle cannot be. 

Rinsing 

Retailing of small pack beer is intensely competitive. Marketing departments have sought 

to gain competitive advantage for their beers with extreme differentiation of the package 

(see `secondary packaging' above). The non-returnable one-trip bottle is well suited for 

this. The brand image can be protected and enhanced by a `perfect' package. Much of this 

beer is sold in large retail chains for consumption at home. Recognition of the product on 

the supermarket shelf is vital to success. Non-returnable bottles are displayed in a wide 

variety of attractive secondary packaging for this purpose. 

Non-returnable bottles are delivered new to the brewery. They are normally rinsed by 

spraying inside and outside several times. For rinsing the bottles are turned upside down 

and then returned to the upright position for filling. Rinsers are now almost integral with 

the filling machine. The process is designed to wet the bottles prior to filling and to 

ensure sterility by killing micro-organisms. Steam is usually jetted into the bottle for this 

purpose followed by a purge of sterile air. This results in 0.1 to 0.2 ml of condensate 

remaining in the bottle. 

Empty bottle inspection 

A bottler of beer must demonstrate due diligence in providing a quality and wholesome 

product. He must also run his bottling plant at the highest efficiency possible consistent 

with supplying that quality product. The bottles presented to the filler must be `fit to fill'. 


The bottles must not contain any foreign bodies or in the case of returnable bottles, any 

residual caustic solution from the washer. Defective bottles with damaged necks must be 

sorted and removed. They may be fillable but could cause damage to the filling machine 

or result in the beer being presented in an unacceptable package. Inspection of empty 

bottles prior to filling is therefore an important process. Originally this inspection was 

done by eye but with the very high rate of modern bottle fillers (> 1,000 bottles/min) the 

process is now performed electronically. Frequently used systems (Kunze, 1999) employ 

very high speed cameras (picture-taking speeds of 1/250,000 s) to look for defects. 

Examinations are made of internal and external sidewalls and the base and neck of the 

bottle. Residual liquid is detected by infra-red radiation from beneath the bottle. 

Defective bottles are rejected. For returnable bottles this can mean rejection rates of 2% 

of the total bottles returned. In an ageing bottle population this rate could be higher. 

Full bottle inspection 

There are two reasons for inspecting full bottles: to check the volume of beer in the bottle, 

and to check for foreign particles. These processes are necessary no matter how careful is 

the empty bottle inspection and the filling. Usually there is a legal requirement to 

guarantee minimum or average contents of the bottle. Inspection systems normally 

involve passing a beam of white light or radiation through the bottle at a defined level. 

Bottles not meeting the fill level required are rejected. This process is often repeated after 

tunnel pasteurization. Off-line checks are also required and records must be kept for 

inspection by government agents. 

Some beers can now be described as global brands. The brand can be brewed and 

packaged in many countries. In this situation the quality and consistency of the beer is vital 

for continuing success. Sometimes it is necessary to introduce on-line checks for foreign 

particles including glass fragments in the bottled beer (Landman, 1999) to further safeguard 

product integrity and wholesomeness. The principle of one machine is to spin the bottle thus 

suspending any foreign bodies then quickly to stop the bottle leaving the beer and any 

unwanted particles spinning. The spinning beer is then examined optically using a 

computerized video camera. If differences between consecutive images are found, 

indicating contaminating particles, the bottle is rejected. The bottle inspecting device is a 

rotating carousel holding 36 bottles. Each bottle has a dedicated camera that moves with the 

bottle and takes multiple pictures as the carousel rotates. The computer detects moving 

particles against a stationary bottle. Particles of below 1 mm in size can be detected. This 

type of full bottle inspection provides the consumer with near absolute protection against 

foreign bodies and the brewer is provided with further enhancement of brand image. 

Labelling 

Managing the flow of bottles is needed before and after filling. Managing the flow of beer 

is concerned with the filling and closing operation and rendering of the beer free from 

micro-organisms. Labelling of the bottle can be considered as part of the management of 

the bottle flow. In the sequence of operations labelling follows full bottle inspection, 

which may itself have been preceded by tunnel pasteurization. Labelling is of major 

significance in the presentation of the beer brand. The label not only tells the drinker what 

the beer is but also conveys an image associated with advertising that adds to the overall 

appeal of the brand. This is particularly the case with international beer brands where 

instant identity of the brand in different countries is important. The application and the 

quality of the label must now be of the highest standard. Poor quality or poorly applied 

labels will imply a poor beer. 


· The label. Every bottle of beer has at least one label, but frequently several labels are 

applied. These can be applied to the body, back and neck of the bottle. In addition foil 

or plastic capsule tops can be applied over the crown. The function of the label is 

twofold: the legislative information required in the country in which the beer is being 

sold must be displayed and the brewer must display the logo and colour detail 

associated with the brand. The legal data may include a statement of the volume of the 

beer in the bottle and its strength and may in some countries include a statement of the 

ingredients. Frequently the shelf-life of the beer must be shown. There are many 

different types of paper used for labels; this number has increased with the popularity 

of metallized paper labels. There are some basic properties of label paper that can be 

described in terms of stiffness, weight, smoothness, density, behaviour in caustic soda 

solutions, curl characteristics, etc. (Schwartz, 1997). 

Originally label paper was resistant to the caustic soda solution used in the bottle 

washer for returnable bottles. This allowed the label to be removed without forming a 

pulp that is difficult to dispose of and, indeed, in some countries its disposal is 

prohibited. In other countries (USA and UK) alkali soluble paper is now used and pulp 

disposal is allowed. Label paper should be resistant to curling and creasing and often 

the reverse side is treated with a pigment to prevent this happening when the front side 

is coated to receive the pigments and metallized effects. The orientation of the grain of 

the paper in relation to the bottle surface is important. The fibres in the bottle label 

should run at right-angles to the longitudinal axis of the bottle. If the fibres run parallel 

then the labels will tend to come away from the bottles at the edges, a phenomenon 

known as `flagging'. 

The different types of paper in use can be divided into three categories: paper, 

metallized paper, and aluminium foil. The paper is usually designated by its weight 

per ream, which in the USA contains 480 or 500 sheets. Paper for paper labels is 

normally of 40 to 50 lb. (18±23 kg). The surface of an aluminium foil label is 99.5% 

pure aluminium at a thickness of 0.009 mm. When applied to paper the overall 

thickness is only 0.09 mm, hence very specialized machinery is needed to produce 

these labels for application to the bottle. Aluminium foils are often used for neck 

labels. The metallized paper label is formed by the vacuum deposition of an ultra-thin 

layer of metal on a paper to achieve a `metal' appearance with a much thinner label 

than the conventional aluminium foil. There is now a huge choice available to the 

brewer and this whole area of packaging is subject to continuous development as 

brewers strive for brand differentiation and product enhancement. Cut labels, ready for 

use are supplied in storable stacks, which must be kept at a relative humidity of 60 to 

70% and a temperature of 20 ëC (68 ëF). This prevents curling which otherwise renders 

the labels useless for application. 

A number of different adhesives have been successfully used to attach labels to beer 

bottles (Schwartz, 1997). The most common and successful are those based on casein. 

These adhesives apply to cold or wet bottles and will bond rapidly. They provide good 

resistance to condensate water and to ice water if bottles are submerged. Easy label 

removal in the bottle washer at low caustic strength is achieved. Alternatives have 

been based on starch or dextrin. These are cheaper but do not offer such good 

properties of, e.g., water resistance. To minimize costs as little adhesive as possible 

should be used and a rate of 10 g of adhesive/m 2 should be aimed for. 

· The labelling machine. Labels are now applied by rotating machines, which can 

operate at the speed of fillers. The machine contains a label holder and a labeltransporting 

device. Modern machines have improved the reproducibility of the 


Label 

magazine 

Gluing pallet 

carousel 

Data coding 

unit 

Glue 

roller 

Gripper cylinder 

Label 

brush-on 

unit 

Bottle 

carousel 

Fig. 21.3 Bottle-labelling station (by courtesy of Krones). 

Plan 

view 

removal of the label from the stack in the holder and its application to the bottle. 

Labels not squarely applied to the bottle are not tolerated. A glued pallet with a thin 

film of glue applied removes the label from the stack in the holder. This results in 

simultaneous gluing of the back of the label. The machines now usually employ 

oscillating glued pallets with a smooth motion capable of achieving speeds of 70,000 

bottles per hour. The overall design relating to the speed of movement of the bottles 

and the label length is critical. The speed and distance between the bottles must be 

carefully set when commissioning labelling machines. Careful design and commissioning 

will allow the use of a label length that is greater than the bottle pitch (distance 

between the bottles) with no loss of efficiency. 

The labels are normally removed from the pallets and applied to the bottles by 

gripper cylinders (Fig. 21.3). Body labels and neck labels can be transferred in this 

way. The body label application occurs when the bottle is smoothly pushed into the 

sponge section of the gripper cylinder. If a neck label is to be applied the sponge pads 

in the cylinder must be moved out by cams to make contact with the bottle. The 

conveying of the bottle through the labeller is critical to ensure that the labels are 

correctly aligned. Rotary labellers operate with an infeed star wheel that passes the 

bottles to a centring bell which firmly positions the bottle against a plate to receive the 

label. To eliminate skewing the bottles are positively clamped between the bottle 

plates and the centring bells. After application of the body and neck labels the bottles 

are rotated through 90ë and pass a brushing station, which ensures firm contact 

between the label and the bottle. If a back label is to be applied, the bottle is turned a 

further 90ë to meet the back label gripper cylinder and the application process is 

repeated. After all labels have been applied the bottles are discharged from the labeller 

via a star wheel to the discharge conveyor. 

Usually a date stamp is now put on the bottle. This might indicate the `best-by' date 

or the date of packaging of the beer known as the `born-on' date by one manufacturer. 

These dates are sometimes required by law or by the customer but in any event give 


the drinker aview of the shelf-life of the beer. The date can be put on the label but is 

often applied to the bottle itself. The brewer is also likely to include areference mark, 

which will be important for product traceability. These reference marks were 

traditionally cut into the edge of the label and this can be done at manufacture of the 

label. Recently other methods have been developed, which include stamping and 

embossing or perforating machines that can operate at up to 60,000 bottles per hour. 

These devices have largely been superseded by ink-jet or laser printers. The ink-jet 

system uses acharged stream of ink droplets in acontrolled trajectory onto the bottle 

surface. The laser directs the energy beam through a metal mask that has the 

appropriateinformationcutout.Thebeamisthenfocusedonthelabel,whichsuffersa 

change of colour by evaporation to leave the date stamp clearly marked. 

Labelling is an important part of the bottling process, vital for the maintenance of 

the brand image. Developments in label application have now caught up with that of 

filling and crowning to guarantee the presentation of avery high-class package. This 

has been important in re-establishing the bottle against the can as the preferred 

package for small pack beer. 

21.3.2 Managing beer flow 

Managing the beer flow involves bringing the beer to the package in the most efficient 

way, consistent with the highest quality being obtained. Almost all bottled beers are now 

sterile, filtered products. Sterility is traditionally achieved by pasteurizing the beer in the 

bottle after filling and crowning. However the beer can be sterilized before filling by 

flash pasteurization or sterile filtration. This is what happens with keg beer, which cannot 

be tunnel pasteurized because of the huge size of the machine which would be required. 

Flash pasteurization is not widely used with bottled beers, although new developments 

are occurring (Hyde, 2000), and this technique will therefore be considered in the section 

on kegging. In managing the beer flow we are concerned with: 

· flash pasteurization or sterile filtration 

· standard filling and aseptic filling 

· crowning 

· tunnel pasteurization. 

Untilrecentlybeerwasnotmovedlongdistancesforsaleandsomicrobiologicalstability 

of the product was not an issue. Modern microbiological stabilization processes began with 

the work of Louis Pasteur (Pasteur, 1876), who demonstrated that heating beer to a 

sufficientlyhightemperaturewoulddestroybeerspoilagemicrobes.Fortunatelybeerisnota 

goodgrowthmediumandonlysupportsslowgrowthofarelativelysmallnumberofmicrobes 

(Rainbow,1971),whichdoesnotincludepathogens(Bunker,1955;Chapter17)).Thetaskof 

providing the customer with amicrobiologically stable product is therefore simplified. 

Absolute security in the sterility of bottled products is required. Bottles may be 

distributed over very long distances, and often have shelf-lives of 40 to 52 weeks. The 

traditional practice has been to pasteurize the beer after filling in its final package (see 

later). However this process is very energy intensive, requiring at least 1,000 MJ/1,000 

bottles. Beer flavour is adversely affected by pasteurization and probably suffers even if 

oxygen contents are kept to levels < 0.1 mg/l. This has led to the use of filtration to 

sterilize the beer prior to filling. Sterile filtration physically removes organisms from the 

beer. The technique demands subsequent aseptic filling and the application of rigorous 

standards to ensure no contamination enters the bottle. 


Sterile filtration 

Some ofthe techniquesused toachieve non-biologicalstability (Chapter 15)ofbeerhave 

been used and claimed to yield sterile beer. So plate and frame filters using pulps of 

kieselguhr or perlite or sheet filters using asbestos/cellulose sheets are said to be effective 

for sterile filtration as well (Wilson, 1997). This type of sterile filtration was originally 

described in the 1930s by the Seitz Company that developed the EK sheet (Entkeimung ˆ 

sterilization). These sheets were made of a mixture of cellulose and asbestos and were 

4.5 mm thick with a pore size of 5±20 m. Flow rate achievable was 1.5 hl/m 2 /h at a 

maximum pressure of 1.5 bar. The quality and composition of the sheets has changed 

much in recent years (Chapter 15) and beer stabilizers can now be incorporated into the 

sheet such as PVPP (polyvinylpolypyrrolidone) or silica hydrogel. 

These pulp and sheet methods act by depth filtration. The pad or sheet contains 

millions of flow channels. Micro-organisms are mechanically entrapped or absorbed as 

the beer flows through the filter. There is also an adsorptive effect as microbes are fixed 

by electric charge. The organisms bear a negative charge, which attracts to the positive 

charge of the filter matrix. These methods have been used for sterile filtration of beer. 

However, since the carcinogenic properties of asbestos were recognized, asbestos was 

replaced in sheet filters with kieselguhr or perlite. Some brewers lost confidence in depth 

filtration as a method to guarantee beer sterility. The situation was changed with the 

availability of membrane filters. 

Membrane filters can be classified as surface filters which operate on a sieving method 

for the removal of organisms (Wilson, 1997). The membrane is a uniform continuous 

structure with regularly spaced uniformly sized pores. The membranes are made of 

cellulose esters and are normally 150 m thick (Bush, 1964). As beer passes through the 

filter all organisms larger than the pore size of the filter are trapped and retained on its 

surface. For brewery use a pore size of 0.45 m is necessary to retain all potential 

spoilage microbes. Membrane filtration is the only sterile filtration method that will 

provide absolute sterility. However, membranes are prone to blockage and it is essential 

that the beer presented to the membrane has received satisfactory primary filtration. The 

beer must be free of particles that will blind the sterilizing filter, the sole aim of which is 

to achieve sterility. This is now achieved by a sequence of filters after the primary 

kieselguhr filtration. These are frequently cartridge filters. There can be two or three in 

line with reducing pore sizes, e.g., 

· kieselguhr filtration 

· cartridge filter 1, 5 m pore size 

· cartridge filter 2, 1 m pore size 

· sterilizing membrane filter, 0.45 m pore size. 

Using a system of this type, high throughput and sterility can be achieved. A nondestructive 

test to indicate the suitability of beer for membrane filtration has been 

described (Pall, 1975). It is also important to have sound microbiological control 

throughout the brewery so that the effectiveness of the sterile filter is further enhanced. 

Other types of filter have been described (Moll, 1994). Ceramic candles have been 

used in Japan. Flow rates of 10 hl/m 2 /h were achieved with ceramic candles of 25 mm 

wall thickness and a pore size of 25 m (Beer, 1989). Cross-flow filtration methods have 

also been tested, where the liquid flow is tangential to the filter medium (Atkinson, 

1988). Tubular membranes and sand have been used as the filter medium. These systems 

are not fully developed for beer and have been prone to blockage with nonmicrobiological 

polymers present in the beer such as -glucans. If the pore size is 


educed to below 0.45 m, which has been the case in some systems, bittering substances 

are removed from the beer (Donhauser and Jacob, 1988). Absolute sterility has not been 

consistently achieved on the industrial scale using cross-flow methods. 

Sterile filtration has the advantage over pasteurization of giving very gentle treatment 

to the beer; with no heating and cooling there is no potential for flavour changes. The 

additional filtration can also improve non-biological stability and clarity of the beer. The 

technique has lower capital and operating costs than tunnel pasteurization. Sterile 

filtration must be operated with sterile filling and so this obviously adds to the difficulty 

of the filling operation. There must be abuffer tank between the sterilizing filter and the 

filler and aconstant flow must be obtained to avoid pressure shocks. The pressure drop 

across the filter should be monitored. A drop in pressure of 2 bar (30lb./in. 2 ) is 

acceptable. Care must be exercised when changing tanks to be filtered or when filtering 

tanks at low beer levels. Gas bubbles can break out at the filter pump which can give a 

sudden pressure drop and hence aloss of sterility in the system. This can be avoided with 

proper attention to the monitoring of pressures and flows. 

Sterile filtration must be operated in asanitary system. Effective CIP is essential. The 

system must be cleaned, sanitized, and back flushed using hot and cold de-aerated water. 

Normally water at 80ëC (175ëF) is circulated for 60 minutes. The CIP system is an 

integral part of the whole sterile filter plant. 

One source of competitive advantage to the brewer is the delivery of the freshest 

tasting beer possible. This is most likely to be achieved using sterile filtration and filling. 

Consequently there is great interest in perfecting these techniques. 

Standard filling 

The filler is the most important piece of equipment in the bottling line. For optimum 

efficiencythefillermustruncontinuouslythroughouttheshift.Machinesbeforeandafter 

thefillermustbedesignedtosupplybottlesandtakethemawayatleastasfastastheyare 

filled and sealed. The beer supply system must keep carbon dioxide in solution and 

exclude oxygen. The whole filling operation must not add more than 0.03mg/l dissolved 

oxygen to the beer. To maintain the carbon dioxide content achieved in the bright beer 

tank the pressure must be maintained during filling at one bar above the CO2 saturation 

pressure. Contamination of the beer must be avoided. Microbiological sterility will be 

achieved by tunnel pasteurization or by aseptic filling if the beer has been sterile filtered 

or flash pasteurized. 

The objectives in filling bottles are to preserve the quality of the beer and get the right 

volume of liquid into the bottle. As beer contains carbon dioxide, beer fillers always 

operate at ahigh pressure and are generally called counter-pressure fillers. To deliver the 

required volume of beer, beer fillers are designed either to fill to alevel in the bottle or to 

fill by displacement. Bottle fillers are rotary machines and can rotate clockwise or anticlockwise. 

Machines can now contain up to 200 filling valves and can fill at 1,000 to 

1,600bottlesper minute.Most fillers perform inasimilarbasicway (Fig.21.4, plan view 

and Fig. 21.6, sequence of operation): 

· beer is received 

· bottles are positioned in predetermined spacing under the filler heads on the filler 

platform 

· bottles are lifted up to the filler head 

· bottles are evacuated and counter-pressured with carbon dioxide 

· bottles are filled 


· corrections are made to fill height 

· pressure is released 

· bottles are lowered and moved onto the crowner, where they are sealed. 

Therefore in sequence we have: 

7 

6 

8 

Fig. 21.4 Bottle filling, plan view: 1, 1st evacuation; 2, carbon dioxide flushing; 3, 2nd evacuation; 

4, pressurizing; 5, filling; 6, filling completed and settling; 7, correction; 8, snifting 

(by courtesy of Krones). 

· Beer supply. Beers are almost always cold filled into bottles at 0to 5ëC (32±40ëF). 

To maintain the carbon dioxide in solution the beer should be at one bar pressure (15 

lb./in. 2 )in the supply to the filler bowl. This is normally achieved by maintaining a 

carbon dioxide top pressure in the headspace of the bright beer tank. There must be no 

dropinpressureatthetankoutletandsothebeerflowrateshouldnotexceed 225cm/s 

uptothispoint.Thebeershouldnothavetotraverseright-angle bends(Spargo,1997). 

The volume of beer supplied to the fillercan be measured in line with amagnetic flow 

meter provided there is enough straight pipe-work in the flow line for the beer to 

achieve laminar flow. All fillers have a`bowl' or tank to receive beer from the bright 

beer tank and to distribute the beer to the filler heads. On large machines (>500 

bottles/min.) the bowl is annular and the beer in the bowl is counter-pressured from 

abovewithcarbondioxide.Asensor,whichisusuallyafloatvalve,controlstheheight 

ofthebeerinthebowl.Fillingvalvesareusuallymountedatthebottom ofthebowlor 

attached to the outside of the bowl. In this way beer distribution pipe-work to the 

filling valve can be eliminated. 

· Bottle positioning and lifting (Fig. 21.5). The lifting platforms place the bottles 

against the filling heads to provide asecure connection. This is done by compressed 

air. The same device using aroller lowers bottles. Compressed air can be forced back 

in the lowering process into the supply pipe and so is not lost. The device to position 

the bottle is called acentring bell or tulip. This is carried up by the bottle being raised. 

The centring tulip ensures that the bottle is centred in the filling head and sealed. 

· Bottle evacuation, counter-pressure and filling (Fig. 21.6). There are anumber of 

different beer filling heads. Beer can be introduced into the bottle via afilling tube. 

The filling tube reaches to the bottom of the bottle. Obviously the tube must have a 

smaller diameter than the bottle opening and this limits filling speed. Filling tubes 

5 

1 2 


3 

4

c 

b 

a 

d 


Enlargement 

Fig. 21.5 Bottle filling, section view; basic position before the bottle is lifted onto the filling 

head (air in bottle); a, vent tube; b, siphon type gas lock; c, correction channel; d, control cylinder 

for liquid valve and pressurizing; e, gas needle; f, vacuum/CIP channel; g, snift channel 

(by courtesy of Krones). 

introduce beer just above the bottom of the bottle. There is very little uptake of 

oxygen. The filling tube is made up of atube through which the beer passes and a 

return gas vent at the level of the required filling height. Some high-speed fillers 

(>1,000 bottles/min.) operate without filling tubes. The beer is introduced down the 

side of the bottle neck. There is adanger of ingress of oxygen with this system but the 

system will operate at higher speed than the filling tube system. Asequence of 

operation for anon-filling tube machine is shown (Fig. 21.6). After lifting the bottle is 

evacuated (21.6a) and about 90% vacuum is achieved. The next phase is aflushing 

with carbon dioxide (21.6b), which enters the bottle from the filling bowl. There is 

then normally asecond evacuation (21.6c). The vacuum allows the displacement of 

the carbon dioxide and some air, which will still be present. The secondary evacuation 

stage would not be carried out with afilling tube bottle filler. Counter-pressurization 

with carbon dioxide then takes place (21.6d). The process is similar to the carbon 

dioxide flush (21.6b) but takes longer and ahigh concentration of the gas is achieved 

in the bottle. The pressure in the bottle is now the same as that in the filler bowl. Beer 

can now flow downwards against the bottle sides in athin film displacing the carbon 

dioxide(21.6e).Assoonasthebeerreachestheendoftheventtubeitrisesinthistube 

(21.6f) and the gas above cannot escape. There is an overfill in this stage which must 

be corrected (21.6g). The beer valve is closed and the gas valve remains open, carbon 

e 

Beer 


f 

g

1st evacuation 

Vacuum in bottle 


Beer 

Fig. 21.6 (a) 1st evacuation of bottle (vacuum in bottle) (by courtesy of Krones). 

CO 2 flushing 



Fig. 21.6 (b) Carbon dioxide flushing (carbon dioxide in bottle) (by courtesy of Krones). 

Beer 


2nd evacuation 

Vacuum in bottle 


Beer 

Fig. 21.6 (c) 2nd evacuation (vacuum in bottle) (by courtesy of Krones). 

Pressurization 



Beer 

Fig. 21.6 (d) Pressurization (carbon dioxide in bottle) (by courtesy of Krones). 


Filling 

Beer displaces 

carbon dioxide 

from bottle 


Fig. 21.6 (e) Filling (beer displaces carbon dioxide from bottle) (by courtesy of Krones). 

Filling completed 

Beer 

Beer 


Beer 

Fig. 21.6 (f) Filling completed (beer in bottle) (by courtesy of Krones). 


Correction 

Beer 


Fig. 21.6 (g) Correction of bottle contents (excess beer displaced back to filling bowl) (by 

courtesy of Krones). 

Snifting 

Beer 

Beer 


Beer 

Fig. 21.6 (h) Snifting (controlled pressure release) (by courtesy of Krones). 


dioxide at asmall overpressure enters the neck of the bottle and the beer above the 

ventpipe isforced back intothefiller bowl.Anexact fillingheightcan sobeobtained. 

ThereisfinallyacontrolledpressurereleaseknownintheUKasa`snift'(21.6h).This 

takes place before breaking the seal with the valve and the centring tulip. Avent valve 

is opened and excess pressure is released so atmospheric pressure is achieved and 

foaming of the contents is avoided. The gas movements in stage hwere formerly 

controlled using cams, which were very precisely cut; cams have now been replaced 

by electronic systems. After filling the bottle is lowered from the filling head and 

conveyed to the crowner. 

It is essential en route to the crowner to eliminate air from the head space of the bottle 

to avoid subsequent oxidation of the beer. This is now usually done by water-jetting. A 

high-pressurestreamofsterilizedwaterissprayedontoeachopenbottle.Onlyafew lof 

water enter the bottle but this causes an effective beer foaming, which rises in the neck 

and dispels oxygen and prevents any further entry. This process is carefully adjusted to 

minimize beer loss. Liquid nitrogen jetting may also be used (Donovan et al., 1999). This 

technique reduces beer losses and, perhaps more importantly, reduces waste. The foam is 

formed initially around the nitrogen gas and it is not necessary to expel foam containing 

air and losses are lowered. 

Aseptic filling 

Sterile filtration is apowerful technique, which can provide beer in its freshest form. 

However the technique must be combined with aseptic filling. Aseptic filling requires 

considerable expertise to achieve success. Anumber of fundamental requirements must 

be met. All employees involved in sterile filling must be well trained and be determined 

to make it work. The beer to be packaged must be free of organisms and therefore must 

be presented from awell managed sterile filtration plant or from flash pasteurization. If 

sterile filtered this has probably involved afinal pass through a0.45 membrane filter. 

The line must be mechanically reliable. Generally the most satisfactory operation is 

achieved from long runs. If there are frequent breakdowns then re-sanitization is required 

with consequent risk to the product and poor efficiency. There must be clearly defined 

and written-down sanitation procedures. These must include the philosophy on the 

method of sanitation be it hot water, steam, or chemical. 

There must be an effective microbiological control system in place (see also Chapter 

17). This is best achieved by the operators on the line being properly trained in 

microbiological sampling and analysis. In this way the operator comes to òwn' the 

problem of the line and problems are more likely to be solved. The main focus of the 

sampling plan is normally the packaged beer. Two samples of beer should be taken off 

each filler head for every two hours of filling. These can be analysed for contamination 

by forcing and membrane filtration. Attention should also be paid to the plant, the CIP 

system and the water in use. 

These basic principles and operating philosophy of the equipment are more important 

in achieving success with sterile filling than the detailed design of the plant. The design 

must be kept simple and open. The main objective in pipework is to avoid sharp bends 

likely to restrict the flow of cleaning materials. Pipework and tanks should be made of 

stainless steel and should be capable of withstanding a flow of hot water at 80 ëC (175 ëF) 

for 30 minutes. Steam can be used for re-circulatory cleansing but generally is not as 

effective as hot water as it lacks an attrition effect. The filler must be sanitized within one 

hour of bottling start-up. The best systems have a dedicated CIP system. The water 


should flow in the reverse direction to the beer. A final chilled water rinse is usually used 

and to this water can be added a 25 ppm iodophor solution. This can be left in the system 

and drained immediately prior to the start of a filling run. 

There has been discussion of the desirability of enclosing sterile filling plant in a 

sterile room. Some systems work successfully in the open in a packaging hall. However, 

there are advantages in enclosure. One is to show operators and visitors, including 

customers, that something special is taking place. A carefully designed enclosure allows 

the control of a sterile air supply in both temperature and humidity. It also facilitates the 

external cleaning of the plant. Sterile filling runs are usually successful over periods of 12 

to 36 h. During this time a continuous sanitation programme is necessary. General 

housekeeping must include the continuous removal of broken bottles and the spraying of 

the floor to wash away spilt beer. Foam cleaning the plant with iodophor solution during 

operation is beneficial to hygiene. Close attention should be given to the filler tubes and 

the crowner platforms (see later). If production stops during a bottling run then iodophor 

treatment should continue. If the stoppage exceeds one hour then the plant should be shut 

down and a full sanitation carried out. The filler must be cleaned at the end of the run by a 

hot water flush prior to full sanitation one hour before the next run. 

The capital cost of a sterile filtration/sterile filling plant can be 10% higher than the 

costs for a tunnel pasteurizer. However, operating costs are much lower and when 

assessing all the costs, sterile filling is about half the cost of tunnel pasteurization (Hyde, 

2000). Sterile filtration avoids the potential flavour deterioration effects of pasteurization 

and gives the consumer the freshest tasting beer. This can create potential competitive 

advantage. However, training and education of operators and acute attention to detail is 

essential for success. In addition the throughput of the line is not likely to be as good as 

with normal filling because of the need to sanitize after stoppages. 

Crowning 

The bottles must be quickly closed after filling. The closure machine is therefore an 

integral part of the filling machine and comes in the line immediately after the point when 

air has been dispelled from the bottle neck by water-jetting. Most beer bottles are closed 

with crown corks. The crown cork evolved from the natural cork stopper. Cork stoppers 

were effective closures but did not meet the need to be applied at speed. The automatic 

production of glass bottles was perfected in the USA in 1903 (Everett, 1997). This 

provided bottles with uniform dimensions at the neck that could take a standard closure. 

William Painter had already filed US Patent 468226 in 1892 for the `crown closure'. 

Painter's crown was lined with cork and characteristically had corrugations in the `skirt'. 

Subsequently a crown with no corrugations was designed along with a machine which 

applied the closure to the bottle and formed the familiar corrugations. 

The cork was mostly obtained from oak trees in the western Mediterranean and during 

the Second World War supplies ceased and prices rose. At this time plastic alternatives to 

the cork lining were developed. The first PVC (polyvinylchloride) lining was introduced 

in 1955. Nowadays, the crown cork is made of tinned or chromed steel or stainless steel 

plate 0.235 mm thick. The brewery logo is usually displayed on the outer surface. The 

standard crown cork has 21 corrugations and an outer diameter of 32.1 mm, an internal 

diameter of 26.75 mm and a height of 6.0 mm. The normal lining is plasticized PVC. A 

lubricant is included in the lining for `twist-off' crowns so they can be removed by hand. 

The lining does not affect flavour and is approved for use in most countries. The crown 

must retain the pressure in the bottle until opened and there must be no gaseous exchange 

with the atmosphere. 


Fig. 21.7 Turning tube for orientating crown corks for bottle closure, incorrectly positioned 

corks are turned up to 180ë to gain the correct orientation for bottle closure (Kunze, 1999). 

Crown corks from stock are conveyed to a storage hopper on the closure machine and 

a sorting device brings them into the correct orientation for closure onto the bottle. 

Incorrectly orientated crowns are re-positioned in a turning tube (Fig. 21.7) containing 

fixed inserts before falling downwards through a groove. This device is simple and 

effective but is subject to heavy wear and so it must be replaced regularly to avoid 

frequent stoppages as crowns become jammed in the tube. Crowns are delivered to the 

closure device either mechanically or pneumatically, preferably using carbon dioxide or, 

less desirably, air. A magnet in the ejector then holds the crown cork. The bottles are 

delivered to the crowner by a star-wheel and the crowner head, containing the correctly 

orientated crown, descends onto the bottle. The crown then makes contact with the crown 

ring of the bottle. A spring forces closure onto the bottle and the 21 corrugations of the 

crown cork are bent down onto the bottle neck. A gas tight seal is obtained. The bottles 

are usually sprayed with water to blow off any beer residues arising from the removal of 

head-space air. 

Inevitably, some returnable bottles are damaged and this damage is sometimes to the 

neck of the bottle. This can affect the security of the crown cork closure. In one case this 

resulted in as many as 800 defective closures per 1 million bottles by one Company 

(Duffy and du Toit, 2000). Development work on the crowning machine followed by 

improved preventative maintenance led to a reduction in defective closures to five to 

eight per million and a consequent enhancement of the presentation of the brand whilst 

retaining the environmental friendliness of the returnable bottle. 

Sometimes other bottle closures are used. A `rip-off' closure has been developed, 

made from light gauge aluminium that can be applied in modified crown cork closure 

machines. This type of cap has also been used on wide-mouth bottles that are easier to 

drink from. Roll-on closures have also been used on 32 oz. bottles in the USA. These are 

normally pilfer proof where the cap is held to a ring by uncut metal `bridges' which 

fracture when the cap is turned by hand to open the bottle. The ring remains on the bottle. 

Largely because of waste disposal reasons these closures have not been popular in Europe 

except on two-litre PET (see later) bottles. 


Tunnel pasteurization 

The surest way to provide the customer with beer containing no viable micro-organisms 

is to treat it at the last possible moment. That is, the beer is treated in its package after 

closure. The beer is not exposed to the atmosphere until it is consumed. This is achieved 

by pasteurization (Pasteur, 1876). Pasteurization is the killing of micro-organisms in 

aqueous solutions by heat. Beer can be pasteurized in bulk while flowing, which is known 

as flash pasteurization, or in the package, which is known as tunnel pasteurization. Flash 

pasteurization occurs before the beer is put into its final container and so is an alternative 

to sterile filtration. Flash pasteurization is usually associated with the preparation of keg 

beer. Tunnel pasteurization is most commonly used with small-pack beer, either bottles 

or cans. The theory of pasteurization is common to both systems. 

· Theory of pasteurization. The basis of pasteurization is establishing the minimum 

time and temperature required to destroy all expected biological contaminants at the 

highest concentrations at which they may occur in filtered beer. Different food products 

have different requirements for pasteurization, and those that can contain sporeforming 

bacteria require much higher heat treatment than beer. Mixed populations of 

common brewery contaminating organisms were subject to a range of times and 

temperatures in beer (Fig. 21.8, known as a lethal rate curve) and were examined for 

subsequent viability. Typically at temperatures of over 50 ëC (122 ëF) an increase in 

temperature of 7 ëC (12.5 ëF) accelerated the rate of cell kill by ten times. Therefore: 

· 53 ëC: minimum time to kill population 56 min. 

· 60 ëC: minimum time to kill population 5.6 min. 

· 67 ëC: minimum time to kill population 0.56 min. 

One pasteurization unit (PU) for beer has been arbitrarily defined as the biological 

destruction obtained by holding a beer for one minute at 60 ëC (140 ëF) (Del Vecchio 

et al., 1951). Therefore in Fig. 21.8 the point at which the line crosses the 60 ëC line 

Fig. 21.8 The effect of time and temperature on the viability of a mixed population of yeasts 

and brewery bacteria. The hatched area shows the range of conditions where all cells are killed 



gives the thermal resistance of the particular suspension of organisms, this is 5.6 min 

and so to achieve effective pasteurization the holding time at 60ëC (140ëF) must 

exceed 5.6min. The slope of the line in Fig. 21.8 is known as the Zvalue. The lethal 

effect (PU) is simply the product of the lethal rate and the time of application. The 

lethal effect at various temperatures in aprocess is additive, therefore the sum of the 

lethal effect is the quantity of sterilization achieved: 

Lethal Effect ˆLxt…PU† 

where LˆLethal Rate and tˆtime held at temperature TëC and 

Lˆ1=Log 1 …60 T/Z† 

Under laboratory conditions beer can be sterilized by treatment with 5±6 PUs when 

cell numbers are 0.2mg/l. The most heat resistant organisms are the lactic acid 

bacteria and some Saccharomyces species such as S. pastorianus. The pasteurizationofreturnedbeer 

ismoredifficultbecause thiscan containveryhigh numbersof 

organisms. It is usually only blended into fresh beer at low rates (50PU) 

are minimized. 

· Practice of tunnel pasteurization. Atunnel pasteurizer comprises alarge metal-cased 

enclosure through which the bottles are passed by aconveying system (Fig. 21.9). The 

conveying system can use aslat continuous conveyor chain on which the bottles are 

slowly moved or can use awalking beam conveyor. In awalking beam the bottles 

stand on bars and are slowly moved forward in cyclical steps (Fig. 21.10). The 

walking beam has the advantage that the moving elements remain in the same 

temperature zone whereas the continuous conveyor inevitably carries some heat and 

water away to another zone. The pasteurizer operates as a series of zones. The bottles 

are loaded at one end and then conveyed under a series of water sprays. The sprays are 

arranged such that the bottles are subject to increasingly hot water until the beer in the 

bottles reaches the pasteurization temperature. In tunnel pasteurizers this temperature 

is usually 60 ëC (140 ëF), which is held for 20 min. This delivers 20 PU to the beer. The 

bottles then move to a cooling zone where they are subjected to cooling by cold water 

sprays. The bottles then leave the pasteurizer. 

Most modern tunnel pasteurizers are double-deck machines (Fig. 21.9). The bottles 

travel to the end of the top deck of the pasteurizer and then go down to the lower deck. 

In this system the water in the sprays on the top deck pre-heats the bottles and passes 

to the lower deck to be used for cooling. Exact control of temperature is important. 

The objective is the effective kill of all organisms with the minimum use of energy. 

Temperatures in all the zones are recorded and often computed with the time to show 

pasteurization units supplied. Passage through the machine normally takes one hour. 

Tunnel pasteurizers are the biggest piece of equipment in the packaging line and a 

surface area of 3.5 m 2 /1000 bottles/h is needed to ensure effective pasteurization. 

Modern machines can achieve outputs of at least 150,000 bottles/h. The water used in 

the pasteurizer must be clean and kept at a pH value of around 8. If not the pasteurizer 

will become grossly infected with a variety of bacteria and moulds and the sprays will 


Load end 

1st 

pre-heat 2nd pre-heat Pasteurizing zone Pre-cool Cooling 

Pre-cool 

reservoir 

2nd pre-heat 

reservoir Pasteurizing reservoir 

1st 

pre-heat 

reservoir 

Overflows 

Cooling 

reservoir 

Discharge end 

Service 

water 

Fig. 21.9 General arrangement of a double-deck tunnel pasteurizer. Times and temperatures for the various zones: 1st pre-heat, 5 min. at 

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, 

2 min. at 30±20 ëC. For can pasteurization the pre-heat and cooling zones may be shortened in length and therefore the beer spends less 

time in these zones, because of the lower structural strength of the can (Hough et al., 1982). 


Fig. 21.10 Walking beam principle for moving bottles through a tunnel pasteurizer. Left-hand 

sequence represents side elevations showing successive movements of shaded grid beams in relation 

to unshaded grid. Right-hand sequence represents corresponding plan views (D indicates 

downward movement and U upward movement), (Coleman, 1976 and Hough et al.,1982). 

be blocked and the bottles will emerge dirty. Water hardness must be removed or at 

the very least calcium ions must be sequestered to keep them in solution so that the 

bottles do not dry with a coating of calcium salts. Sometimes bacteriocides containing 

chlorine are added to the water. 

During pasteurization pressure builds up in the headspaces of the bottles. The exact 

pressure build-up depends on the volume of the headspace, which must be accurately 

controlled by fill height to limit possible bottle breakage. Carbon dioxide can also be 

released from its supersaturated state in the bottle if the bottles are knocked during 

pasteurization. If the bottle is defective the bottle may burst or carbon dioxide may 

leak allowing a reduction in the gas content of the beer. For example, if the carbon 

dioxide content of the beer is 0.38% and the percentage headspace in the bottle is 

1.7%, the pressure would exceed 10 bar during pasteurization if supersaturation did 

not hold. A walking beam conveyor is generally better at avoiding bottle knocks than a 

slat conveyor. If effective controls of fill height are in place bottle breakage should not 

exceed 0.2% of the total bottles processed. 

Tunnel pasteurization is a very safe method of assuring sterile beer. It is, however, 

the most expensive method of assuring sterility in both capital and operating costs 

(Hyde, 2000), which are double those of sterile filtration and at least five times more 

than for flash pasteurization (see later). It is also a fault of tunnel pasteurizers that it is 

easy to overpasteurize the beer. The beer will then not taste fresh and, particularly if 

the dissolved oxygen level was > 0.2 mg/l, will very quickly develop a cardboard 

flavour. This has stimulated interest in sterile filtration and flash pasteurization for 

small pack beer. 

21.3.3 Managing plant cleaning 

All areas of a packaging plant must be kept clean, using a rigorous housekeeping regime. 

This is particularly important in a bottling plant, because breakages of bottles will occur 

and the broken glass is a threat to employees and the product. It must be removed at once. 


U 

U 

U 

U 

U 

U 

D 

D 

D 

D 

D 

D

The floor area around the filler should be disinfected daily. Thorough cleaning of the 

filler and crowner is essential. Infection can develop in the filler; usually as aresult of 

Acetobacter infection, which develops on residues of beer and foam. Slime caps are 

formed which can provide encouragement to other organisms particularly Lactobacillus 

sp., which will grow in beer at very low oxygen levels. This may result in aneed for 

increased pasteurization with aconsequent adverse effect on beer flavour. 

Thorough mechanical cleaning of the filler and the provision of ahot water flush unit 

avoids infection. After bottles have left the filler this device can provide an overflow of 

hot water at 90ëC (195ëF) in the filling machine, at the point of delivery of the crowns 

and over the star wheels. The sequence can be arranged so this overflow occurs every 2± 

3hours, during 2±3 filler rotations at half speed. This should be supported with foam 

cleaning of the conveyors and spraying with iodophor at the end of each bottling run. 

The key to successful plant hygiene is management and training. Even if tunnel 

pasteurization is to take place after filling this is not an excuse for poor hygiene 

management. 

21.3.4 Materials for making bottles 

The most important material for both returnable and non-returnable beer bottles is glass. 

Glass exists at room temperature as asuper-cooled liquid. It is chemically inert and will 

not add or take away properties from the product it contains. Glass will only break under 

tensile load and the compressive strength of glass fibres is such that they will withstand 

34,000 bar. Glass is recyclable and cullet, broken glass fragments, is an important 

constituent of new-made glass (Moll, 1997). 

Glass bottles can be made in several different colours but the choice of colour is 

importantforthe flavourstability of the beer. White lightcan act onthe iso -acids inbeer 

to form acompound, 3-methyl-2-butene-1-thiol. This imparts an aroma and flavour known 

as light-struck or `skunky', and renders the beer very unpleasant. It can be detected by 

some people at aconcentration of 0.4 parts per trillion (0.4ng/l)! Beer must therefore be 

protected from light. Brown glass is the best option and is much better than green glass in 

this respect. The worst type of glass is clear (flint) glass in which the flavour of beer 

deteriorates very rapidly when exposed to light. Marketing departments frequently want to 

use green or flint glass for beer considering that it gives the beer amore attractive 

appearance. This is resisted by brewers concerned about flavour risks. This is one reason 

why reduced isohumulones are now added to some beers. These compounds do not break 

down under light to the thiol. The compounds also have increased hydrophobicity 

compared to isohumulone and so they also enhance foam (Chapter 19). 

Bottles can be made to very exacting specifications, which are often controlled by 

statute. Bottles can be assessed by a number of quality control methods to ensure 

satisfactory performance on the line (Moll, 1997). Consumers prefer glass for the 

packaging of small pack beer believing it gives a more upmarket image and better flavour 

to the beer (Yeo, 2000). In the UK in 1998 the non-returnable market was split: cans 

69%, glass bottles 30%, plastic bottles (PET) 1%. Polyethylene terephthalate (PET) has 

been the preferred plastic used for containers of soft drinks since c. 1993. This product is 

gas permeable and carbon dioxide seeps out and oxygen seeps in. This has been 

overcome for beer by using a gas-proof coating on the PET or by the insertion of a nonpermeable 

barrier between two layers of PET (Nelson, 2000a). These bottles can be 

recycled and new bottles can be made of 40% recycled material. These new PET 

containers have greatly reduced oxygen ingress to the bottle and a shelf-life of six months 


has been claimed for beer (Nelson, 2000a). Carbon dioxide escape from the bottle 

remains a problem. The most-used coatings involve a silicon oxide layer on the outside of 

the bottle, which means the beer is in contact only with the PET. 

There is now an alternative to PET, polyethylene naphthalate (PEN). This polymer has 

10±15 times greater barrier properties than PET (Nelson, 2000a). This gives the potential 

to refill the bottles as well as recycle the plastic. Bottles made of PEN can be washed at 

85 ëC (185 ëF) and so can be tunnel pasteurized containing the beer. PEN is much less gas 

permeable than PET, but is at least four times as expensive. A number of major brewers 

are now using PET and PEN bottles for beer, claiming the beer is easier to chill in these 

bottles and that drinking without a glass is also easier. There is no breakage in the supply 

chain and beer in plastic bottles is more acceptable at sporting events, where broken glass 

can be dangerous. There are likely to be further innovations in this area. 

21.4 Canning 

The arrangement of a canning line has much in common with that for non-returnable 

bottles (see above). This section will concentrate on the differences from bottling, and 

will focus on the package itself and how it is filled. Beer, in cans, was first test-marketed 

in the USA in 1935. In 1962 the `rip-off' aluminium can end opening was introduced, 

again in the USA. Further developments occurred with various side seam techniques 

using tin-free steel on cans, which were at that time three-piece. The two-piece 

aluminium can was introduced in 1958, closely followed by other developments. The 

American domestic market has been dominated by beer in the can. In 1993 the market 

was split: can, 60%; non-returnable bottle, 25%; returnable bottle, 5%; draught, 10%. 

Other countries, particularly in Europe, concerned with the problem of empty can 

disposal, have not encouraged canning and the bottle dominates the market. However, the 

UK is unique among European countries and again the can is prevalent in the take-home 

trade. In 1998 the split of the market was: can, 24%; non-returnable bottle, 11%; 

returnable bottle, 2%; draught, 63%. The canning of beer, therefore, is clearly important 

in world terms but trends are certainly now towards bottles. 

The unique aspects of beer canning relate to the can itself, the unloading and rinsing of 

the cans prior to filing, the application of the date information and the filling and closing 

operations. All other aspects of palletizing and de-palletizing, secondary packaging and 

pasteurization are very similar to the operations associated with bottling. 

21.4.1 The beer can 

Cans have some advantages to the brewer compared to bottles. They are lighter and 

unbreakable. They can easily be stacked both in the brewery and in the refrigerator. The 

can has a large surface area on which to display information about the brand and it can 

be opened without a tool. Light cannot penetrate the can and so light-struck flavour 

cannot develop. Disadvantages relate almost entirely to disposal. Domestic disposal is 

usually to a waste tip mixed with other waste materials. This is not satisfactory. 

Attempts should be made to compress and recycle cans and to separate the two metals 

used in can manufacture, aluminium and steel. Making aluminium cans from aluminium 

ore takes considerable electrical energy. Only 10% of this energy is needed when new 

cans are made from recycled material. The need for active recycling programmes is 

clear. 


Beer cans are now almost entirely of two-piece construction whether made from 

aluminium or steel. The can end, reflecting the original American invention, is always 

made of aluminium. The volume of the can is usually, 33, 44, or 50 cl. The metal used to 

make the can is largely governed by the conditions prevailing in the industries of the 

country of origin. Aluminium is almost always used in the USA whilst steel dominates in 

Germany. Both metals are used in the UK. Beer cans, both aluminium and steel, are 

characterized as 206 or 202 cans relating to the internal diameter of the mouth of the can. 

The newer 202 can was introduced to save metal and hence weight: 

diameters: 206, 57.4 Ô 0.3 mm; 202, 52.4 Ô 0.3 mm 

weights for 50 cl can (aluminium): 206, 19.32 g; 202, 18.38 g (Kunze, 1999). 

The tendency is for most beer canners to use the 202 can. 

The two-piece can is manufactured from a coil of plate, which is unwound, lubricated 

and fed into a blanking press (Scruggs, 1997), which forms the sheet into shallow cups. 

These cups proceed to an ironing press to form the side wall of the can. More metal is 

retained at the top and bottom of the can for strength but in the side wall the thickness 

will be only 0.09 mm. The subsequent strength of the filled can relies on the internal 

pressure of the beer in the can. Trimming follows ironing to ensure uniform height. 

During these processes the can is covered with lubricant that is removed with sprays. 

External decoration on the can is applied by a rotary machine using inks and/or varnishes, 

which are stabilized by baking. Internal coatings are applied by an airless stationary 

spray-gun and the cans are then baked again to cure the coating and remove solvent. The 

final process is the die necking to achieve the 206 or 202 diameter. The cans are supplied 

to the brewery on pallets. For 33 cl cans 8,280 cans can be put on one pallet in 23 layers. 

For canning lines operating at 2,000 cans/min. the constant supply of cans to the canning 

line is critical for success. Manufacturers of cans often have plants close to major 

brewery sites. 

Can ends are made from 0.27 mm thick aluminium sheet. The end contains the ringpull, 

which nowadays is almost entirely of the `stay-on-tab' type which means that 

separated, and discarded rings do not pollute the environment. Lids are supplied with a 

diameter of 64.75 mm which, after fitting to the body of the can, is reduced to the 

appropriate size for the 206 or 202 can. 

21.4.2 Preparing cans at the brewery for filling 

The main concern with handling cans, compared to bottles, is the delicate nature of the 

can and that it arrives pre-decorated. This decoration must not be damaged, as it cannot 

be repaired. Cans are pushed off the pallets in layers onto a flat conveyor chain. There 

must be no gaps between the cans so they cannot fall over. The spacer packaging and the 

steel frame of the pallet is normally returned to the supplier. Can lids are also supplied on 

pallets, which can contain 250,000 lids packed in bags of 600. The `best-by' date is 

usually displayed on the base of the can. This is usually applied while the can is dry, i.e., 

before rinsing and filling. The normal method is to use an ink-jet printer, which can 

match the speed of can filling. Cans are always rinsed before filling by spraying with 

water whilst in the inverted position. The motive force for this act is usually gravity and 

hence cans are unloaded at a height of 3 m and then pass down through the rinser to the 

filler. Cans are delicate and have little structural strength when empty. They will not, 

however, break like glass bottles and so concerns about glass fragments demanding 

complex inspection machinery are not relevant to canning. 


21.4.3 Can filling 

The most important machine on the canning line is the filler. There are similarities in the 

machines used for filling bottles and cans. Some machines are dual purpose and have the 

change parts to fill bottles and cans. Dedicated machines are used for choice because the 

speeds at which a can filler can run are about twice those of a bottle filler. A 100 head can 

filler will operate at least at 2,000 cans per minute. The main technical differences 

between bottle and can filling derive from the properties and size of the empty can. Thus, 

cans have a very wide neck and so provide a large surface area for gaseous exchange 

during filling. This can lead to a loss of carbon dioxide and ingress of oxygen and cans 

are very light and will not resist a vacuum so they are not evacuated during filling. The 

filling tube has similarities with the short bottle filling tube. A quiet fill down the side of 

the can is achieved by lowering the filling assembly to achieve an air-tight connection 

with the can, i.e., the can is not lifted in this process. 

The technique of centring the delicate can is important. Some very clever designs 

allow operation at high speed (> 1,500 cans/min.) with no oxygen entry. Between the 

lowering centring tulip and the fixed filling tube is a differential pressure chamber that 

connects to the internal space of the can. During counter-pressuring and filling this 

chamber maintains the filling pressure. There is only a small difference between the 

diameter of the pressure chamber and the diameter of the can and so there is only a small 

force on the can during filling. After filling the volume in the differential pressure 

chamber is reduced and pressure increases and the can is thus released from the seal. 

The beer is held in an annular filling bowl under a blanket of carbon dioxide. The 

filling valve is supported by the annular bowl (Fig. 21.11). The sequence of filling is: 

· carbon dioxide purging 

· counter-pressuring 

· filling 

· venting 

· lifting the valve. 

Stationary 

neutral cam 

Rotation 

Stationary 

closing cam Stationary 

snift cam 

Rotating 

filler bowl 

Infeed 

Valve tulip 

operating cam Rotation 

Diverter 

guide 

Purge cam 

Infeed conveyor 

Air operated 

vacuum cam 

Air operated 

valve trip 

Infeed worm 

Infeed star 

Discharge 

conveyor 

To 

seamer 

Fig. 21.11 Can filling, schematic. Cans are fed to the infeed star, which places them on to the 

filler bowl platforms, one can beneath each filling valve. The can filling sequence then operates; 

as the bowl rotates, cams remove air from the can, charge with gas, fill with beer, close the 

valve and remove residual gas. (Heins and Heuer, 1997). 


During filling the filling valve is lowered along with the filling tube, the return air-pipe 

and the differential pressure chamber. The airtight connection is so formed. The can is 

then counter-pressured with carbon dioxide and the beer valve opens introducing the beer 

quietly down the side of the can from, usually, 15 tubes. In some can fillers these small 

tubesarereplacedbyavalve,whichdeliversthebeerinathinfilmdownthewholeofthe 

can side to the bottom. The evacuated air flows up the return air pipe until the beer level 

reaches the pipe. The pipe contains aball, which rises with the beer and closes the inlet. 

This controls the height of liquid in the can as with bottle filling. The position of the 

return air-pipe can be adjusted. The centring tulip is then raised, the differential pressure 

is increased, and the can is released from the seal and proceeds to the can seamer for 

closing. There is little fob in this process and it is not economic to attempt recovery from 

the evacuated air. In some systems after filling, jetting carbon dioxide over the beer 

surface breaks any bubbles of carbon dioxide or air. 

Controllingfillingheightbythepositionofthereturnair-pipeisanestablishedprocedure 

for bottles and cans. This does, though, require slowing the filling process towards the end. 

For this reason there has been adesire to develop procedures of volumetric fill where the 

exact filling volume is pre-determined (Fig. 21.12). Avolumetric can filling system has 

been described (Kunze, 1999) in which the volume is measured in a thin measuring cylinder 

by a floating ultrasonic sensor. The cylinder is filled from below without turbulence and is 

hence charged to deliver the appropriate volume at the beginning of the filling process. This 

allows a speeding and standardizing of the filling process. 

21.4.4 Can closing (seaming) 

There is a large surface of beer in the can and so the lid must be secured as soon as 

possible to keep out oxygen and to minimize dissolved oxygen levels ( 0.2 mg/l). Cans 

must be hermetically sealed to be impervious to air, liquids and bacteria. The closed can 

must withstand the internal pressure and be robust enough to survive distribution and 

retail display. 

The can lid is usually placed on the can whilst the can is on the filler carousel. The can 

is raised and pressed together with the loosely placed lid against a firm sealing head. A 

double seaming process follows. 

· First operation, interlocking: the outer part of the lid is bent under by a pre-roller with 

a defined profile. 

· Second operation, tightening: an air-tight seal is obtained by pressing with a seaming 

roller. 

Carbon dioxide is blown over the surface of the beer as the can end is being slid into 

position. Most of the air is blown out and replaced with carbon dioxide (under-cover 

gassing). Air contents in the head space are < 1.5 ml. A sealing compound incorporated 

in the curl end of the lid effects the hermetic seal following tightening. The compound 

may be water-based latex emulsion or synthetic rubber material. 

The success of the seaming operation depends on smooth coordinated operation of all 

moving parts of the seaming machine. Regular, planned maintenance is essential guided 

by the history of the machine. There should be regular corrective action to ensure the 

machine is always set precisely for perfect closing of the can. An efficiency of operation 

of 95% (actual time run versus time planned) should be achievable through proactive 

planned maintenance. This type of maintenance is best achieved through the team 

approach with multi-disciplined teams of operators able to work together on the line, 


Cam follower for 

valve lowering 

Differential 

pressure 

chamber 

Filling valve 

Centring unit 

Level probe 

building up a history of operation and utilizing this knowledge to effect the maintenance 

plan. Programmes of regular testing for seam integrity should be in place. Standard visual 

tests have been developed for this as well as destructive tests involving examination of 

seam tolerances with micrometer gauges (Heuer, 1997). 

Since cans are not transparent fill height is usually checked using radiation which 

penetrates the can and checks whether the beer fill height meets specification. Finally 

almost all canned beer is tunnel pasteurized. Typically beer will receive 20 PU: 

· pre-heat 46 ëC (115 ëF) 1 min. 

· superheat 62 ëC (144 ëF) 3±4 min. 

· pasteurize 60 ëC (140 ëF) 15 min. 

· pre-cool 46 ëC (115 ëF) 2 min. 

· cool 32 ëC (90 ëF) 2 min. 

Product channel 

Automatically 

applicable CIP cup 

1 Flushing, pressurizing 

and return gas valve 

2 Upper snift valve 

3 Lower snift valve 

4 Flushing and CIP 

return valve 

Metering chamber 

Beer supply valve 

Following pasteurization the cans are dried with jets of compressed air and the fill height 

is often checked again by radiation. Secondary packaging then takes place often to the 

customer's specification. 

1 

2 

3 

4 

Snift channel 

Pressurizing channel 

Fig. 21.12 Can filling valve (by courtesy of Krones). 


Microbiological checks focus on the beer supply prior to filling and the filling and 

seaming operations. 

21.4.5 Widgets in cans 

Traditional English ales were, and sometimes still are fermented in open vessels. The 

naturallevelofcarbondioxidethusentrainedinbeerattheendoffermentationwasabout 

1.2 vol/vol or 2.4g/l (Lindsay et al., 1995). This was therefore the normal carbon dioxide 

level at which the ale was drunk. Canned ale was frequently packaged with acarbon 

dioxide content of 2.6 to 3.0 vol/vol. This obviously had a very different taste to 

traditional cask ale. The beer was much more effervescent and more like alager beer in 

mouth-feel. It had been discovered in the 1940s that small amounts of nitrogen gas in 

beer had aprofound effect on head creaminess and durability (Carrol, 1979). This was 

partly as a result of the low solubility of nitrogen gas in beer (20mg/l at room 

temperature and pressure). Nitrogen is colourless, tasteless, odourless and chemically 

inert. These properties and discoveries suggested the use of nitrogen in the packaging of 

ale with the objective of providing the drinker with ataste sensation akin to that of cask 

beer, i.e., low carbon dioxide effervescence and thick creamy head. The problem was to 

introduce nitrogen into the beer in areproducible way that could then consistently give 

the consumer the appropriate taste sensation. Considerable research and development 

work carried out in the early 1980s led to the development of the `widget'. 

The concept was to introduce aplastic capsule with asmall hole and containing 

nitrogen into the can and to pressurize it during canning. The pressure would be released 

when the can was opened and the nitrogen would be forced into the beer giving the 

characteristic creamy head and reflux of bubble formation (Lindsay et al., 1995). Many 

widgets have been described (Brown, 1997). Some have been more successful than 

others. Originally widgets were inserted into the can before filling but this did not allow 

the exclusion of oxygen and hence flavour stability was poor. Now widgets can be 

supplied already attached to the base of the can. Frequently the floating widget is used 

since it is less likely to trap oxygen in the can during filling. Liquid nitrogen is often 

added during filling. Before the development of the widget some companies used solely 

liquid nitrogen to promote the reflux effect. This technique is still used in the keg 

packaging of `smooth flow' ales. In any event oxygen ingress must be kept as low as 

possible during canning and nitrogen can be used as the undercover gas. Cans often 

proceed through atunnel between the filler and the seamer to further prevent oxygen 

pick-up (Brown, 1997). 

Many companies introduced smooth flow ales in cans in the 1980s and 1990s. Some 

thought that this type of beer would replace standard canned ales with carbon dioxide 

contents of around 2.5 vol/vol. This has not happened and many consumers returned to 

the more effervescent product. 

21.5 Kegging 

Sales of beer in draught form are greater than in small-pack in only the UK and Ireland 

(Table 21.1). In these countries the sale of beer in kegs is important. Kegging is about 

filling carbonated, pressurized, pasteurized beer into sterile containers. These containers 

usually contain 25, 30, 50, or 100 litres of beer. In the UK volumes of 9 imp. gal. (firkin), 

18 imp. gal. (kilderkin or kil), or 36 imp. gal. (barrel) are still common. All kegs are 


eturnable. The collection of empty kegs from depots, bars and public houses is an 

important part of the overall management of keg packaging (Chapter 22). Kegging has 

similarities with the packaging of beer in returnable bottles. The major differences 

concern the handling and cleaning of the much bigger container and the sterilization of 

the beer. As noted beer inkegs cannot be tunnelpasteurized. The theoryof pasteurization 

(Section 21.3.2) shows how beer can be pasteurized in bulk. 

21.5.1 The keg 

Kegs are normally made of stainless steel or aluminium. Stainless steel kegs are made of 

achrome/nickelalloy. Theyareheavy;a50litrekegwillweigh 12±15kg(around30lb.). 

Aluminiumkegs aremadeofanalloyalsocontaining magnesium andsilicon.Thesekegs 

are lighter than stainless steel kegs and were originally more popular. However 

aluminium kegs are more frequently stolen than stainless steel kegs because of the ease 

with which aluminium can be melted down and sold. Aluminium kegs cannot be cleaned 

with caustic alkali-based detergents because hydrogen gas is formed. Cleaning is with 

acid or dilute alkalis. Stainless steel kegs can be cleaned with acid or alkaline detergents 

and generally are more robust in use, aproperty that is particularly important as the 

container ages. Generally, therefore, stainless steel kegs are preferred. 

All kegs have aneck containing athreaded bush (Barnes neck) into which fits akeg 

valve fitting (Fig. 21.13). This fitting is called the spear or extractor tube and through it 

the filling, emptying, cleaning and automatic closing of the keg is achieved. Kegs have 

advantages over bottles. They allow the partial dispense of the product and they operate 

as closed vessels with in-built leakage detection. Keg extractors should not be withdrawn 

outside the brewery and kegs are returned containing excess gas pressure, which prevents 

contamination entering. On modern keg filling lines there is a pressure test to 

demonstrate internal pressure and any kegs not having such apressure will be rejected 

and not filled. 

Kegs are delivered with the extractor installed and protected from dirt during delivery 

withaplastickegcap.Todispensethebeerabayonet-typedispenseheadisclampedontothe 

extractoratthebar.Thisallowstheingressofthetoppressuregasandtheoutletofthebeerto 

the dispense tap (Chapter 23). There are several types of fitting which means that kegs from 

different brewers are often not interchangeable on the dispense equipment in the bar. 

Protective 

top chime 

472 

∅ 382 

Protective 

bottom chime 

Spear for 

filling and 

emptying 

under top 

pressure 

of CO 2 

Fig. 21.13 Vertical section of a50l beer keg, height 472mm; diameter 382mm (by courtesy of 

Alumasc Ltd). 


Well-fittings contain two valves, one for the beer, normally aball valve and one for 

the top pressure gas. Aflat-top fitting has one valve which seals separately the gas and 

the beer. It is easy to use but is bigger than the well-fitting. The `combi' fitting is a 

combination of the well and flat top fitting, but has two valves. Great efforts have been 

made to make these fittings safe and tamper-proof. It is now an integral part of the filling 

operation to test for extractor tightness as well as internal pressure in the keg before 

filling. 

21.5.2 Treatment of beer for kegging 

Beer for kegging is normally conditioned to yield 1.5 to 2.5 vol/vol carbon dioxide. 

This applies to ale and lager (Chapter 15). Some `smooth-flow' ales in the UK are 

now packaged at carbon dioxide levels of 1.0 volume and with nitrogen contents of 

30±40mg/l. Nearly all beer for kegging is bulk (flash) pasteurized in acontinuous 

flow pasteurizer at high pressure, say 10 bar (150lb./in. 2 ),against aback pressure of 

1bar (15lb./in. 2 ). 

Flash pasteurization 

Flash pasteurization is carried out in aplate heat exchanger in which there are four 

sections (Fig. 21.14): 

· regeneration section 

· heating section 

· holding tube 

· cooling section. 

Beer is pumped to the regeneration section where it flows counter-current to hot beer and 

is therefore pre-heated. In the heating section it is brought up to pasteurization 

temperature by passing counter-current to hot water or steam. It is then held for apredetermined 

period in the holding tube. The beer then passes back to the regeneration 

section where it loses heat to the incoming beer. It subsequently runs counter-current to 

cold brine or alcohol in the cooling section. The maximum temperature achieved is 

between 71 and 79ëC (160±175ëF) and the holding period is usually between 15 and 60 

seconds. 

It is very important that all the beer flowing through the heat exchanger receives the 

same pasteurization treatment. Turbulent flow ensures that this is the case. The Reynolds 

number can be used to ensure that the condition of turbulent flow is achieved. This 

number is the product of liquid density, velocity and tube diameter divided by liquid 

viscosity. At Reynolds number values below 2,000 flow is laminar but above 3,000 flow 

is increasingly turbulent and this should be aimed for. 

To change the number of pasteurization units given to beer the temperature is altered. 

The heat exchanger is designed for a particular flow rate for maximum efficiency. There is 

a substantial pressure drop through the pasteurizer and to keep carbon dioxide in solution 

beer is pumped in at 8.5 to 10 bar gauge pressure against a back pressure of 1 bar. The use 

of buffer tanks before and after the pasteurizer is essential to prevent interruptions of flow 

and pressure surges on the beer in the bright beer tanks and the keg racker. 

Heat transfer in the exchanger is achieved mainly by convection. An important design 

factor is the surface area required for efficient heat transfer: 

A ˆ Q/HT 


Zephyr valve 

To filler 

or racker 

Rhodes flow indicator 

Zephyr valve 

Drain 

Heater 

Detergent–sterilizing tank 

Steam 

supply 

Drain 

Paraflow 

b 

Brine/ 

alcohol 

Cooling 

Beer pump 

Control 

cock 

Duplex 

temperature 

recorder 

Regeneration Heating Holding 

Pump 

Temperature 

controller 

Hot water set 

b a 

Oil-free air supply 

Diaphragm 

valve 

Steam 

reducing 

valve 

Steam 

supply 

Steam 

injector 

Fig. 21.14 Flash pasteuriser; (a) raw unpasteurized beer, (b) pasteurized beer (by courtesy of APV Co. Ltd). 


Unpasteurized 

carbonated 

beer tank

where Qis the heat load in joules/h, Ais the area in m 2 ,His the overall heat transfer 

coefficient in watts/m 2 /K, and Tis the logarithmic temperature difference (Hough et al., 

1982). 

Advantages of flash pasteurization compared to tunnel pasteurization are: 

· less space required 

· lower capital cost of equipment 

· lower operating costs (only 15% of the cost of tunnel pasteurization (Hyde, 2000)) 

· shorter periods of exposure of the beer to temperatures where chemical changes are 

rapid but pasteurization is slow. 

There are dangers for flavour stability of beer with flash pasteurization. If dissolved 

oxygen levels are >0.3mg/l more marked flavour changes may occur in flash 

pasteurization than in tunnel pasteurization. This is owing to higher temperatures and 

turbulent flow but more importantly, from the recycling of beer back to the buffer tank 

when the process conditions have not been met or packaging has been interrupted. This 

leadstoexcessivepasteurization.Itisgoodpracticetocirculatewater,andnotbeer,when 

flow is stopped but this is difficult to control and can lead to beer losses and dilution. 

Asthebeerissterilizedbeforefilling,thekegsthatreceivethebeermustbesterileand 

filling must be an aseptic operation. Strict microbiological checks downstream of the 

pasteurizer are essential. Atypical pasteurizer may hold the beer for 20 seconds at 75ëC 

(167ëF) and this is equivalent to 50PU. This high PU value illustrates the need for close 

control of the pasteurizer to avoid excess pasteurization and consequent flavour 

deterioration of the beer. 

21.5.3 Handling of kegs 

Kegs are transported in stacks, which can be arranged horizontally using pallets or 

vertically using spacer boards. Horizontally configured cradle pallets are very heavy and 

contribute considerably to the overall weight on adistribution vehicle. Plastic spacer 

boards are light and allow the vertical stacking of containers up to five stacks high. This 

system is now preferred in the UK and Ireland. 

Empty kegs returning to the brewery first have to be destacked or depalletized. The 

principleofoperation ofthesemachinesresemblesthatofthe machines usedfor handling 

crates of returned bottles (Section 21.3.1). Destacking is performed layer by layer by 

pushing the kegs together and lifting by pneumatic grippers. The spacer board can be 

removed manually or by machine. Stacking the full containers at the end of the line and 

after labelling is performed by similar machines. The spacer boards can be placed in the 

stacks manually or by machine. 

Efficiency of operation is critical to any kegging line. It depends as much on the 

supply of the containers as it does on the supply of the beer. It is essential that kegs are 

`fit to fill' and that kegs that are not fit do not proceed to the racking machine 

(International Bottler and Packer, 2000). After destacking, empty kegs are tested for 

internal pressure and tightness of the extractor tube in the Barnes neck. Kegs that fail 

are removed from the line for subsequent inspection and repair. These kegs may have 

been tampered with and thus be heavily infected or unsafe. Plastic protective caps will 

have been removed at the dispense site. The next requirement is to wash the exterior of 

the keg. Most returned kegs display a self-adhesive label, which will show the details 

of the previous filling, i.e., the beer quality and `best before' data. These old labels 

must be removed. This is achieved by pre-soaking, scrubbing and spraying with hot 


detergent at 70ëC (160ëF). Detergent is re-circulated. Hot water sprays remove 

detergent before the kegs leave the washer. External keg washers take up alot of space 

and use much energy and water (at least 10 hl/hourat 1.4 bar). The supply ofhot water 

is frequently waste process water and condensate, which may already have been used 

several times in the brewery. Roller or flat top chain conveyors convey kegs through 

these machines. 

21.5.4 Keg internal cleaning and filling 

Kegs are filled with pasteurized beer. Kegs must, therefore, be sterile as well as clean 

prior to receiving the beer. For this reason cleaning, sterilizing and filling are carried out 

on one machine called akeg racker. The extractor (spear) remains in place during this 

operation. On most machines the cleaning takes place with the keg inverted, when the 

drainage is quicker and more complete and water forced up the spear cleans the side and 

base of the keg more effectively. In most systems filling also takes place with kegs 

inverted but any misalignment of the filler and the neck can lead to large beer losses and 

so sometimes each keg is returned to the upright position for filling (Fig. 21.15). These 

tasks demand the inclusion in the line of 180ë turning machines, which must operate 

reliably and be of robust construction. 

Automatic keg cleaning and filling machines are generally of two basic types, inline 

or rotary. In-line machines have anumber of `lanes' at which cleaning, sterilizing 

and filling takes place in sequence at aseries of `heads'. The number of lanes (typically 

8±24) governs the capacity of the machine. The lanes are at right-angles to the 

conveyor bringing the empty kegs to be filled. In-line machines are often used in the 

UK and they have to cope with three sizes of keg, e.g., 50l(11 imp. gal.), 18 imp. gal. 

and 36 imp. gal. In these machines kegs are fed from one side and leave from the other 

and proceed in astraight line to the labeller and capper (Figs 21.15a and b). These 

diagrams show a typical overall layout and the detail of the cleaning and filling 

operation. Asimple two-head machine is shown at which washing takes place on the 

first head and filling on the second. Recently machines with four heads have been 

developed to provide alonger cleaning cycle (Carter, 2001). These machines have two 

washer heads, a sterilization station, a pre-filling head and a filler head. Stringent 

washing using water and detergent can be carried out. In-line machines can handle 

frequent size changes and short runs of different beer qualities, though this will 

considerably reduce efficiency. Downtime on an individual lane will not affect the 

other lanes and individual lanes can be taken out for maintenance during normal 

production. Lanes can also be added to an existing machine to increase production. 

Production rates of 1,200 kegs/h are achievable. 

On rotary machines kegs are handled simultaneously on aseries of stations in a 

circular motion around a central core. A typical rotary machine would comprise a 

washing machine with 24 stations and afilling machine with 12 or perhaps 16 stations 

(Fig.21.16a).Rotationtimesare65±125seconds.Kegsarebroachedonceandthenrotate 

with services switching on and off as the kegs move around the circle. Services of water, 

detergent,steam,carbondioxidegasandbeerandcondensatereturnhavetobeconnected 

into the centre of the machine. This requires complex sealing with two discs to prevent 

mixing (Figs 21.16b and c). These machines require less space than in-line machines and 

are capable of production rates of up to 2,000 kegs/h. 

The mechanical design of rotary machines is simpler than in-line machines and 

downtime is usually less, however, if one part of the machine fails the whole line is 


(b) 

(a) 

Lateral 

discharge 

conveyor 

Depalletizer 

Locator 

boards 

Unitizer 

Decant 

Capper 

Filler 

head 

Racking 

(filling) 

station 

Fit for 

fill 

Labeller 

Control 

systems 

Manifolds for 

steam, water, 

detergent, 

beer, etc. 

Empties 

turner 

Tipping 

station 

Washing 

station 

Washer 

head 

Inverted 

keg 

Lateral 

entry 

conveyor 

Keg 

positioning station 

External washer 

Washer/racker 

Checkweigher 

Container flow 

Fulls 

turner 

Conveyor for 

re-circulating 

Fob 

tank 

Sterile 

beer 

tank 

Pasteurizer 

Buffer 

tank 

Beer 

flow 

Beer 

ex BBT 

Fig. 21.15 Automatic linear (lane) internal keg washing and filling machine, equipment shown 

in (a) is found in the washer/racker area of (b) (Eaton, 2002; Hough et. al., 1982). 


stopped. Rotary machines are not suitable for frequent container size and beer quality 

changes. In the UK the cleaning agent is usually acid (e.g. 2% phosphoric acid). This 

allows mixed populations of aluminium and stainless steel kegs to be processed on the 

same machine. Constant use of acid will lead to build up of protein residues on the inside 

of the keg and from time to time kegs should be sorted and stainless steel kegs cleaned 

with caustic alkali. Aluminium kegs can be cleaned with dilute alkali. On the mainland of 

Europe caustic soda is normally used as stainless steel kegs predominate. On some 

machines there is the facility to use both acid and alkali in the same cleaning sequence. 

This is not common practice in the UK. 

Table 21.2 Time sequences for automatic cleaning, sterilizing and filling of 50 l and 100l kegs on 

a four-head racking machine (Carter, 2001). 

Sequence operation Time (s) 50 litre Time (s) 100 litre 

Head 1 

Deullage 1 

3 5 

Wash 1, recovered water 70 ëC 8 15 

Air purge 6 12 

Wash 2, detergent 18 28 

Air purge 7 15 

Vent head 1 1 

Transfer delays 12 14 

Time at head 1 55 90 

Head 2 

Wash 3, water 70 ëC 20 28 

Steam purge 105 ëC 7 17 

Steam scavenge 8 16 

Steam pressurize 1.5 bar 2 3 

Head cool 1 1 



Steam hold station 

Steam hold 46 81 

Transfer to head 3 9 9 

Time at steam hold station 55 90 

Head 3 (pre-filler) 

Steam hold 28 28 

Steam head 3 3 

CO2 purge 8 15 

CO2 pressurize 2 bar 2 3 

Vent head 2 2 



Head 4 

Steam head 3 3 

Beer fill 4±5 bar 35 70 

Scavenge 3 3 

Neck wash 3 3 



1 The process of allowing spent beer in the keg to drain. 


Each C-frame 

incorporates the 11 

components for 

two treatment 10 

stations 

9 

8 

12 

13 

14 

Starwheels are 

exchangeable to 

suit kegs of 

different diameters 

Circular 

motion 

Sequences of cleaning, sterilizing and filling vary considerably in practice and for in-line 

machineswilldependonthenumberofheadsonthelane.Atypicalsequenceforafour-head 

in-line machine is shown in Table 21.2. The washes can be pulsed to give more reliable 

cleaning and each wash except the final one is purged from the keg with sterile air. The final 

wash is purged with steam. Effective sterilization and prevention of oxygen pick-up is 

essential. Steam is used for sterilization. Counter-pressuring with an inert gas such as carbon 

dioxide is used to keep out oxygen. A temperature of 105 ëC (190 ëF) must be recorded inside 

the keg for it to be effectively sterilized. Counter-pressures can be varied but are usually in the 

range 0.7±3.5 bar (10±50 lb./in. 2 ). Gas is removed down the spear as filling proceeds through 

the gas ports of the keg. The beer flow is modulated to avoid fobbing, starting and ending at 

about 10% of full flow. This procedure also allows more precise volume control (see below). 

The sequences (Table 21.2) demonstrate the longer time for processing the larger keg. 

The time sequence will be shorter on a two-head machine but the cleaning and 

sterilization may not be as effective (Hough et al., 1982). It is advantageous to process 

long runs of beer into one keg size; frequent changes of size during a shift must be 

1 

2 3 

Fig. 21.16(a) A rotary keg racking machine shown in plan from the top. Kegs move around 

the central core at the series of stations which comprise: 1 cylinder down; 2 leak test; 3 rinse 

head; 4 blow out head; 5 pressure test; 6 counter-pressure; 7 fill; 8 disconnect; 9 rinse head; 10 

pressure relief; 11 cylinder up; 12 close keg; 13 release keg clamp; 14 discharge keg (Eaton, 

2002 and courtesy of KHS Till). 


7 

4 

5 

6

Fig. 21.16(b) and (c) Central distributor of the rotary keg racking machine shown as (b) a 

schematic block diagram and (c) in vertical section. Each seal for the supply and return of services 

comprises two discs: a carbon ring and a metal distributor ring. When the holes of the two 

rings in the rotating section are aligned liquid will flow to the head (Courtesy of KHS Till and 

Eaton, 2002). 

avoided. Washer heads and filler heads have manifolds for detergent, water, steam, 

carbon dioxide and beer. All these mains must be cleaned and steam sterilized before beer 

is passed through. Keg plants require frequent cleaning. This can take up to six hours 

twice a week depending on the volumes of beer being processed. After filling, the Barnes 

neck is usually air-dried, though this step is sometimes omitted, and the keg proceeds to 

be capped and labelled. These tasks were traditionally manual but machines are now 

available for automatic capping and labelling. 

A major factor in filling is an accurate determination of the contents of the keg. This is 

required by statute in most countries and is subject to audit. There are two requirements: 

the contents must meet a prescribed amount so that the customer is assured of receiving 

the appropriate volume and is not defrauded, and containers must not be consistently 

overfilled or the correct amount of excise duty will not be paid. Control of volume to 


meet these criteria is not easy. A gross and tare weighing system can be used with success 

on smaller volumes but this system has been unsuccessful with barrels (36 imp. gal. 

containers) still in use in the UK. A volumetric control is preferred although recently 

other devices have been described (Carter, 1998; Brewer and Carter, 2000). 

An on/off beer filling valve can give a keg brim fill to Ô0.5%. This can be improved to 

Ô0.25% with an electro-magnetic flow meter (Carter, 1998). More advanced systems 

(Brewer and Carter, 2000) incorporate a modulating back pressure valve with the flow 

meter. This system delivers pressure profile filling with a precision of Ô0.02%. No 

mechanical valve is required to regulate flow and 100 l/min. is achievable. A slow initial 

fill is possible, which reduces carbon dioxide breakout and the potential for oxygen pickup. 

After the speeding-up of the flow a quiet cut-off ensures low carry-over into the 

extractor tube and the retention of internal pressure. Other developments incorporate an 

inductive flow meter in a direct flow control filling head (International Bottler and 

Packer, 2000). The keg is counter-pressured with 1.4 bar (20 lb./in. 2 ) carbon dioxide or 

mixed gas and the flow meter coupled with a valve allows an exact pre-set amount of 

beer into the keg. A sensor in the return gas pipe keeps the counter pressure gas, and so 

the fill rate, constant. 

Filling by weight can be achieved using load cells incorporated into the line. Many keg 

lines incorporate a check-weigher to determine satisfactory operation of the volumetric 

system. A weighing platform fits within the conveyor and under-weight kegs are rejected, 

usually by a pneumatic ram onto a reject line. 

21.5.5 Keg capping and labelling 

A plastic cap is applied to the Barnes neck to protect the filling and dispense valve from 

dirt and to deter tampering. A number of designs are available. Some caps are formed in 

the applicator machine whilst other cap types are supplied with an appropriate logo 

affixed by the manufacturer. Automatic machines must be capable of 125% of the rate of 

the filler so as not to be rate limiting in the overall process. Some caps attach by shrinking 

and some by clipping to the neck of the keg. The most important aspect is the degree of 

security it provides against interference. 

Labels on kegs are important in providing the customer with information as to the 

beer quality in the keg and its `best-before' date, and the brewer and the customer 

with complete traceability of the container. Barcodes applied to paper labels are 

essential to contain this information. Machines are now available which, when 

programmed, print and subsequently apply the labels to the containers. Quality of 

printing and application is now high. (International Bottler and Packer, 2000). 

Barcodes can contain the sequential number of the container in the packaging run as 

well as the quality and best-before date. These barcodes can be scanned at various 

points in the supply chain, e.g., at despatch, on the delivery vehicle, at the selling 

point and on return to the brewery. These labels must be removed prior to reuse of the 

keg. This means that the labels can be damaged and made unreadable during the 

supply chain history of the container. 

Containers frequently disappear and the paper label will not protect against this. 

Brewers have sought more innovative solutions to the problems of container management 

and traceability. In some systems the entire population of kegs is handed over to a third 

party who, for a fee, manages the population, thus reducing the capital employed by the 

brewer (Nelson, 2000b). Radio frequency identification tags can be embedded into the 

keg, which identify the keg as a unique container throughout its whole life history. This 


could be an extremely valuable system if it could be coordinated on anational (or even 

international!) scale between competing brewers. Progress on this in the UK has been 

slow. Probably further developments will occur in ways of tracking containers, because 

of the desire for complete traceability for quality assurance and because of the high cost 

of containers (approx £40/container). 

21.5.6 Smooth flow ale in kegs 

As stated earlier, beer for kegging is normally conditioned prior to pasteurization to a 

carbon dioxide content of 1.5 to 2.5 vol/vol. Pasteurization and filling must then be 

managed to maintain this carbon dioxide concentration. In an analogous way to the 

development of the widget in canning (Section 21.4.5) there is adesire in the UK and 

Ireland to produce keg beers with the drinking characteristics of cask ales. This demands 

acarbon dioxide content of around 1.0 vol/vol and athick creamy head. The use of 

nitrogen gas provides asolution (Carrol, 1979). The problem with keg beer is getting the 

nitrogen intothe beer andkeeping itin solutionthrough packagingandup to thepointof 

dispense. Nitrogen can be introduced into aroused bright beer tank at the inlet through a 

sinter. The tank will be top pressured with nitrogen and considerable manual 

involvement will be needed to achieve asatisfactory result (Fitch, 1997). The carbon 

dioxide content of the beer must be known and, since the nitrogen is being added before 

the flash pasteurizer, the pasteurizer must have the pressure capability to keep nitrogen 

gas in solution as well. 

An improvement is to use amass flow system. Ameasurement is made of the initial 

carbon dioxide and nitrogen contents of the beer (which should be below specification) 

and both gases are added. Turbulent flow in the pasteurizer ensures good mixing and the 

boost pump of the pasteurizer system provides sufficient pressure to prevent gas breakout. 

This system demands very careful control of the operating pressures to achieve the 

desired level of nitrogen consistently. In the `Nitroset' system (Lindsay et al., 1995) 

nitrogen is injected directly into the beer after the pasteurizer. Aboost pump can achieve 

15 bar (220 lb./in. 2 )pressure to keep the gas in solution (Fig. 21.17). Accelerator mixers 

create further turbulence to aid solution. Variable flow rates of between 150 and 450hl/h 

are achievable. Nitrogen concentration can be determined pre- and post-addition with 

Òrbisphere' thermal conductivity nitrogen monitors. 

Further techniques have been described using hydrophobic membranes (Gill and 

Meneer, 1997). The membrane is in the form of ahollow fibre, which allows gas to 

diffuse into or out of aliquid without the need for intimate mixing. Each membrane 

comprises abundle of fibres with the gas on the inside and the liquid on the outside. 

Gases can be exchanged into or out of the liquid by varying the partial pressure and gas 

composition on the inside of the fibres. Pilot scale work has been carried outand scale up 

to 500hl/hour has been claimed to be feasible. This technique could be valuable for both 

the addition of nitrogen and the removal of carbon dioxide, particularly if this could be 

achieved by asingle pass through the gas exchanger. At present direct injection systems 

are the most widely used. 

Nitrogenated beer has both carbon dioxide and nitrogen in solution and these gases 

must be kept in solution at the appropriate concentrations up to the point of dispense of 

thebeer(Chapter23).Dalton'sLawdemandsthatthesametwogasesmustbeinthehead 

space of the tank or the keg in the same balanced proportions. The beer can therefore be 

filled into the keg using the mixed gas proportion of dispense or, for very low carbon 

dioxide beers (< 1 volume), 100% nitrogen can be used. 


Press. 

reg. 

Coalescing 

filter to 

dry CO 2 

Beer supply 

Press. 

switch 

CO 2 

purge 

N 2 

monitor 

Infeed 

N 2 sensor 

Pasteurizer 

system 

Beer pressure 

boost pump 

FT 

SC 

PT 

Needle 

control 

valve 

Drain 

Flow 

indicator 

NRV 

Gas 

filter 

Filter 

NRV 

Press. 

switch Press. 

Coalescing 

filter to 

dry N 2 

Mass flow 

controller 

reg. 

15 bar 

N 2 supply 

3 bar 

steam 

PLC 

PI PI 

Accelerator – mixers 

Press. 

reg. 

Coalescing 

filter to 

dry CO2 Press. 

switch 

CO 2 

purge 

To SBT 

N 2 

monitor 

Outfeed 

N 2 sensor 

Fig. 21.17 Flowsheet for the `Nitroset' beer nitrogenation system: FT flow transmitter; SC speed controller; PT pressure transmitter; PI pressure indicator; 

NRV non-return valve; PLC programmable logic controller; SBT sterile beer tank (Lindsay et al., 1995). 


Atypical `smooth-flow' English ale would have adissolved gas analysis of: 

· nitrogen (mg/l) 35 (Range 32 to 38) 

· carbon dioxide (vol/vol) 1.1 (Range 1.0 to 1.2). 

In the UK virtually the whole of the keg ale trade is in high nitrogen, smooth flow beer. 

There is virtually no interest in this type of beer in other parts of the world. There have 

been lagers packaged at high nitrogen levels in the UK. These have not been entirely 

successful because of the need to maintain asignificant carbon dioxide content (say 2.2 

vols) to achieve the sparkle typical of good lager. The nitrogen `softens' the flavour so 

that the pleasure of drinking asharp, effervescent lager is lost. 

21.6 Cask beer 

The final type of packaging to consider is the packaging of naturally conditioned beer 

intocasks.Thisbeerisnotfilteredandstabilizedinthebrewery.Itcontainsliveyeastand 

depends on asecondary fermentation in the cask to provide condition (carbon dioxide) to 

thebeerbefore itisdrunk.Thebeerisnotservedunder pressureandisdispensedbyhand 

pumps (Chapter 23). This type of beer is produced in large volumes only in the UK. Of 

the 59 million hl of beer produced in the UK in 1998, 38 million hl were sold in draught 

form and of this volume 11%, about 4.2 million hl was cask beer. Originally beer was 

filled (racked) into casks made of wood. These are now rare and almost all casks are 

stainless steel or aluminium. 

Cask beers tend to be rich in flavour and aroma, particularly those flavours 

associated with hops and this makes them unique and much sought after by some 

drinkers. The advantages to the brewer of producing cask beer relate to the relatively 

low cost of the equipment needed and the low energy input. Disadvantages include the 

inherent variability of the product, its proneness to infection if consumed slowly and the 

skill required of the publican to dispense it properly (Chapter 23). Some brewery 

marketing departments do not like these attributes of cask beer and are more 

comfortable with the predictable behaviour of keg, bottled and canned beer. As a 

consequence most ale brands in the UK are sold in keg as well as cask form and for 

some brands the relative volumes are much more in keg. This trend has been resisted by 

a consumer organisation called CAMRA ± the Campaign for Real Ale. This is a 

powerful pressure group and has done much to persuade brewers to continue to produce 

high-quality cask beers. Most brewers seek to develop good relationships with their 

local CAMRA branch. CAMRA operates in all areas of the UK. 

The issues involved in producing sound cask beer revolve around the handling of the 

casks and the handling of the beer. 

21.6.1 The cask 

Casks are not pressurized containers and are simpler in construction than kegs (Fig. 

21.18). Casks contain ahole on the top of the belly called the shive hole that is protected 

by the shive boss. Through this hole the cask is filled and then closed with the shive, 

which can be made of wood or plastic. A second hole on the end of the cask is stoppered 

with a plug called a keystone prior to filling. This plug is usually made of wood and 

through this plug is driven, where the beer is to be served, the tap from which the beer 

leaves the cask to be drunk (Chapter 23). Casks are usually of 9, 18, or 36 imp. gal. 


Protective 

bottom chime 

∅ 470 

Shive boss 

Protective 

top chime 

Keystone 

boss 

Fig. 21.18 Vertical section of 18 imp. gallon beer cask, height, 520mm; diameter, 470mm (by 

courtesy of Alumasc Ltd). 

capacity. In the 1970s much beer was sold in hogsheads (54 imp. gal.) but these are now 

very rarely seen. 

21.6.2 Handling casks 

Equipment for handling empty and full casks and removing them and stacking them on 

pallets or spacer boards is identical to the equipment used for kegs. In small breweries 

(

Stop/feed 

Lifter 

Receiving 

Bung find and 

pre-rinse 

Pre-rinse 

To 

drain 

Ext-clean and 

bung find 

HL/det. 

HL 

21.6.3 Preparing beer for cask filling 

Cask beer contains live yeast and an important aspect of preparing beer for the cask is 

controlling the yeast count at racking. Beer can be racked directly from the fermenting 

vessel after skimming and an appropriate settling time, but with this arrangement it is 

difficult to get aconsistent control of yeast count. After fermentation beer is usually run 

into racking tanks or backs. In these tanks the beer is given atime to settle, 16±48h, 

depending on the state of the beer and demand from trade. The yeast will partly settle. 

The objective is to achieve ayeast count at rack of around 1million cells/ml of beer. The 

range of yeast count at which cask beer can be successfully packaged is from 0.25 to 4 

million cells/ml, but nearer to 1million is to be preferred. Too much yeast suspended in 

the beer will result in aviolent secondary fermentation and when the casks are vented 

prior to sale beer and foam will gush from the cask and will be lost (Chapter 23). The 

remaining beer will be dull and lifeless and will have little to no head retention or foam 

character. If too little yeast is present the secondary fermentation will be too slow and 

there will be insufficient carbon dioxide inthe beer at dispense with the consequence ofa 

flat, lifeless beer. 

Settling controls yeast count but to aid this process, finings are used (Chapter 15). 

Isinglass finings are added at the rate of 1 to 4 pints/imp. brl (0.36±1.44 l/hl). These 

finings can be added in the racking tank or at any point up to when the beer is dispensed. 

The usual point of addition is at rack with perhaps a prior addition in the racking tank. In 

any event the beer will require from 12±48 h and possibly up to 72 h to fine and settle 

before it is sold. The fining of cask beer is one of the most difficult of all brewery 

operations to control consistently. Often brewers experience periods of poor fining which 

are difficult to explain. Isinglass finings bear a positive charge because of the rich 

collagen content and interact with the negative charge on the yeast cell wall. In most 

circumstances this interaction is sufficient to achieve effective clarity. 

Some beers, sometimes will not fine with isinglass alone. The yeast may have a too 

low negative charge or the concentration may be too high (say > 2 million cells/ml), or 

there may be too high a concentration of positively charged colloids in the beer. In this 

HL 

To 

To reclaim 

reclaim 

or 

re-circulation 

Steam 

To 

drain 

Delivery 

Lowerer 

Rolling rail 

discharge 

Fig. 21.19 Internal cask washing equipment: nine station chain machine with moving centre 

beam; HL, hot liquor (water); det, detergent applied to outside of cask (not used in all installations) 

(by courtesy of Porter Lancastrian Ltd.). 


situation auxiliary finings derived from alginates, carrageenan or silicic acid, and having 

a negative charge, can be added to the beer before isinglass finings to precipitate the 

positively charged colloids (Vickers and Ballard, 1974). An effective method is often to 

add the auxiliary finings in the racking tank and separate the flocs thus formed in this 

vessel and then to add the isinglass at the rack of the beer. Priming sugars are also added 

to some beers at this stage. These are normally solutions at 1150 ëSacch (37 ëP) and are 

added at rates of 1 to 5 pints/barrel (0.35±1.75 l/hl). The priming sugar provides a small 

quantity of fermentable carbohydrate (often sucrose) to assist the yeast to achieve 

effective secondary fermentation in the cask. 

The pH value of cask beer is usually in the range 3.90±4.20. Some brewers add 

potassium metabisulphite to beer, which has a bacteriostatic action at pH values below 

4.20. This can give some protection against infection but is not a substitute for good 

practice in cask washing and filling. 

21.6.4 Cask filling 

The size of breweries producing cask beer varies considerably from those producing 

national brands at over 5,000 imp. brl (8,000 hl)/week to those producing less than 100 

imp. brl (160 hl)/week. Further, micro-breweries or pub breweries may produce only 

enough beer to be sold in their own premises and this may be only 1±10 brl (2±150 hl)/ 

week. This means that cask filling operations can vary from single head manual fillers to 

large, multi-head racking machines. 

Cask beer has a shelf-life from packaging to consumption of a maximum of four 

weeks. It contains considerable quantities of yeast, which provides protection against 

flavour defects which might be attributable to excess oxygen. While the elimination of 

oxygen ingress during packaging is not as critical as with the packaging of brewery 

conditioned beers, it is minimized. Modern cask racking machines usually incorporate a 

stage of counter-pressure with carbon dioxide gas. 

After washing, casks are conveyed by roller or flat chain conveyor to the racking 

machine. Prior to racking a keystone is driven into the cask. The casks are rotated bellyup 

so that the shive hole is vertically uppermost. The cask is located beneath the filling 

tube at a `head' on the racker (Fig. 21.20). There may be up to eight heads on the racker 

and therefore up to eight casks can be filled simultaneously. The filling tube will 

Fob return pipe 

Valve 

Seal 

Sight glass 

Beer 

Detergent pipe 

Fob return vessel 

Beer return pipe valve 

Beer main 

Fig. 21.20 Traditional cask ale racking back (Hough et al., 1982). 


compriseatubeforthebeerandatubeforcarbondioxidegasandthisisloweredontothe 

shive hole to make contact with the cask, which is so sealed from outside air. The cask is 

first counter-pressured with carbon dioxide, the bottom valve is opened and the beer then 

flows at atmospheric pressure into the cask. Air/carbon dioxide in the cask flows through 

the return air-pipe, which usually contains asight glass. The cask is deemed to be full 

when beer can be seen in the return air-pipe. 

Fillers controlled by avolumetric meter are also available. The beer supply is then 

shutoffandthefillingheadisraised.Beer inthereturnair-pipe isevacuatedtoafobtank 

andcanbeaddedtoafollowingcask.Thisrequiresscrupulousattentiontothecleanliness 

ofthemains,asthisprocedureisafrequentsourceofinfectiontothecaskbeer.However, 

if this beer is not returned losses will be high. After raising the filling tube the cask is 

closedbymanuallydrivingashiveintotheboss.Isinglassfinings(1±4pints/imp.brl)can 

also be added at this stage when the beer is said to have been `fined at rack'. This is now 

usually the case in the UK because tax (duty) on the beer is paid on the volume and 

strength of the beer leaving the brewery and this cannot be controlled if the cask is 

broached for fining in adepot or at the point of sale. 

Cask beers are renowned for hop character and this derives from the particular hops 

used in the copper and also from the practice of dry-hopping beers on the racker so that 

hop aroma can develop in the cask in its period of storage prior to dispense. Dry hops are 

usually added inthe form of pressed pelletsat rates of 0.5±6ounces (14±84g)/barrel.The 

pellets are formed from whole cone hops and are lightly pressed to avoid rupture of 

lupulin glands (see also Chapters 7and 8). They are supplied in weights of Ü, and Ý 

ounce (7 and 14g). The pellets have ashort shelf-life and should be kept at below 15ëC 

(60ëF) and abatch should be used in four months. The pellets are added by hand prior to 

closing the cask with the shive. The weight of the pellets and hence the essential oil 

content is very variable. This has led to the development of various extracts in which 

more consistent levels of essential oil can be added (Chapter 7). These products are, 

however, difficult to handle on the line, are expensive and have not enjoyed wide favour 

with brewers particularly as the sales of cask beer are in decline. Many different hop 

varieties have been used for dry-hopping and particular varieties are favoured for 

particular beers. These can be as diverse as East Kent Goldings, Fuggles, Wye 

Northdown and Styrian Goldings. Brewers should pay particular attention to securing 

adequate supplies of hops for several crop years ahead to ensure beer flavour is not 

compromised. 

After filling the cask is labelled. The comments made on the labelling of beer in kegs 

(21.5.5) equally apply to beer in casks. To ensure that the beer is in optimum condition 

when drunk requires considerable effort in storage and dispense (Chapters 22 and 23). 

In the UK cask beer remains a favourite choice of experienced drinkers but its wider 

appeal is limited by inherent inconsistencies and the desire to drink beers at lower 

temperatures (< 8 ëC; 46 ëF), which seriously limit the flavour experience of the true 

cask product. 

21.7 Summary 

Packaging is a vital part of brewery operations. The rise in the drinking of beer at home 

and the influence of retail supermarkets has meant that effective packaging of beer in 

bottles and cans is essential to catch the eye of the purchaser. Huge amounts of money are 


spent by global brewers on packaging developments, certainly much more than is spent 

on research into processing or raw materials. 

In many parts of the world the major package remains the returnable bottle. But local 

brewers are in competition with the brewers of international brands putting their 

traditional markets under threat by more sophisticated packaging. This has resulted in a 

raising of standards of packaging to preserve the market for local beers. 

Even with draught products there is a greater demand by customers for cleaner 

containers and for labels providing more information about the quality of the beer and its 

shelf-life. Concurrently the brewer has made efforts to improve the traceability of beer in 

kegs or casks, to protect his investment in the container and to ensure that the customer is 

protected in any cases of poor quality. 

Brewing becomes increasingly a trans-national business and it is likely that as markets 

develop the trend will continue towards packaging in non-returnable bottles, probably 

made of plastic. This will increase the pressure on finding effective ways of recycling the 

empty package. 


ATKINSON, B. (1988) J. Inst. Brewing, 94, 261. 

BEER, C. (1989) Tech. Quart. MBAA. 26, 89. 

BLRA (2002) Statistical Handbook, Brewers' and Licensed Retailers' Association, London. 

BREWER, A. J. and CARTER, A. (2000) Tech. Quart. MBAA. 37, 105. 

BROWN, D. (1997) The Brewer, 83, 25. 

BUNKER, H. J. (1955) Proc. 5th Congr. Eur. Brew. Conv., Baden-Baden, 330. 

BUSH, J.H. (1964) Brewers' Digest, 39, 48. 

CARROL, T.C.N. (1979) Tech. Quart. MBAA. 16, 116. 

CARTER, A. (1998) Proc. 25th Conv. Inst. Brew. (Asia Pacific Section), Perth, 114. 

CARTER, A. (2001) Personal communication. 

COLEMAN, M. (1976) Brewers' Guard. 105 (10), 51. 

DEL VECCHIO, H. W., DAYHARSH, C. A. and BASELT, F. C. (1951) Proc. Ann. Meet. Amer. Soc. Brewing 

Chemists, 45. 

DONHAUSER, S. and JACOB, F. (1988) Brauwelt, 128, 1452. 

DONOVAN, P., CURRIER, R., BLANTON, R. and ROSS, J. (1999) Tech. Quart. MBAA. 36, 247. 

DUFFY, G. and DU TOIT, C. (2000) Tech. Quart. MBAA, 37, 285. 

EATON, J. B. (2002) Personal communication. 

EVERETT, J. F. (1997) MBAA. Beer Packaging Manual, 167. 

FITCH, N. (1997) Ferment, 10 (1), 41. 

GILL, C. B. and MENEER, I. D. (1997) The Brewer, 83, 77. 

HEINS, H. and HEUER, J. F. (1997) MBAA. Beer Packaging Manual, 249. 

HEUER, J. F. (1997) MBAA. Beer Packaging Manual, 274. 

HOUGH, J. S., BRIGGS, D. E., STEVENS, R. and YOUNG, T. W. (1982) Malting and Brewing Science, Chapman 

and Hall, New York, 721. 

HYDE, A. (2000) The Brewer, 86, 248. 

International Bottler and Packer, (2000) February, 40. 

KUNZE, W. (1999) Technology Brewing and Malting, VLB Berlin, 460. 

LANDMAN, B. C. J. (1999) Tech. Quart. MBAA. 36, 329. 

LINDSAY, R. F., LARSEN, E. and SMITH, I. B. (1995) Proc. 25th Congr. Eur. Brew. Conv. Brussels, 705. 

MOLL, M. (1994) Beers and Coolers, Intercept, Andover, Hampshire. 

MOLL, W. A. (1997) MBAA. Beer Packaging Manual, 83. 

NELSON, L. (2000a) Brewers' Guard. 129 (2), 25. 

NELSON, L. (2000b) Brewers' Guard. 129 (6), 26. 

PALL, D. (1975) Brygmesteren, 32, 197. 

PARSONS, M. (2000) The Brewer, 86, 347. 

PASTEUR, L. (1876) EÂ tudes sur la bieÁre, Gauthiers Villars, Paris. 

RAINBOW, C. (1971) Process Biochemistry, April. 

ROUILLARD, C. (1999) Tech. Quart. MBAA. 36, 435. 

ROUILLARD, C., and HOWELL, M. (1999) Tech. Quart. MBAA. 36, 243. 

SCHWARTZ, V. (1997) MBAA. Beer Packaging Manual, 192. 


SCRUGGS, C. E. (1997) MBAA. Beer Packaging Manual, 226. 

SPARGO, W. (1997) MBAA. Beer Packaging Manual, 149. 

VICKERS, J. C. and BALLARD, G. (1974) The Brewer, 60, 19. 

WAINWRIGHT, T. (1999) Ferment, 12 (6), 4. 

WILLOX, I. C. (1966) J. Inst. Brewing, 72, 236. 

WILSON, J. R. (1997) MBAA. Beer Packaging Manual, 352. 

YEO, A. (2000) Brewers' Guard. 129 (6), 22. 

YOKOI, T., SASAKI, T. and KURAHSHI, M. (1991) Tech. Quart. MBAA. 28, 12. 


22 

Storage and distribution 


It was commonly the practice to deliver beer from breweries directly to customers. This 

was the case in the UK and parts of Europe from the time of the Industrial Revolution 

until well into the 20th century. This was not the case in North America where the 

distances to be covered were much greater. The principles of the storage and distribution 

of beer are governed by the nature of the customer, the distances involved in the delivery 

and the type of beer. In large breweries the organization of storage and distribution is 

frequently the responsibility of an entire department, which may even be located away 

from the brewery site. As the brewing industry has become more competitive so the 

storage and distribution of beer has to be carried out with the utmost efficiency and at 

lowest cost. The likely long-term winners will be the companies who pay most attention 

to getting good quality beer to the customer when he wants it, not when it is best for the 

brewery to deliver. This has resulted in companies adopting the principles of logistics, 

derived from military experience, to ensure optimum deployment of stock and vehicle 

movement. 

22.2 Warehousing 

Brewing is a capital-intensive business, which involves both fixed assets and working 

capital. A major part of the working capital is the stock of finished goods held at the 

brewery. Beer deteriorates with age and so stock levels should be held as low as is 

possible consistent with delivery to meet customer requirements. In this way the customer 

gets fresh beer and the return on capital employed is enhanced. 

In most countries, governments take an excise tax on beer so the value of the stock 

depends to a large extent on whether it is held as tax paid or tax suspended. In duty 

systems where the beer is taxed àt the brewery gate' the beer in the brewery store is held 

duty suspended where it might be held duty paid in a store when it has exited the brewery. 

It makes financial sense to minimize overall stock and maximize duty un-paid stock. 


Brewing companieshave become expert atthisbutthedemandsofcustomerservicehave 

become more important as breweries seek to establish competitive advantage. This has 

tended to result in higher stocks of some products being held in order to avoid any 

possibility of stock-outs. The demands on the effective management of the brewery 

warehouse have, therefore, increased. Breweries will have awarehouse situated as close 

as possible to the end of the packaging lines. Frequently there is aseamless transition 

from packaging to warehouse and these parts of the brewery may be under the same 

management. This avoids disputes as to the rate-limiting steps in the sequence of 

operations from packager to customer. 

Warehouses vary in complexity in relation to the number and types of products 

stocked. This can include beer in casks and kegs as well as numerous sizes of bottles and 

cans in various package types. This complexity is judged in terms of stock keeping units 

(SKUs). Each type of package is one SKU. Beer of the same specification in bottles of 

25cl, 33cl, and 50cl is 3SKUs. This would further proliferate if those bottles were 

packed into boxes containing 15, 18, or 24 bottles, which might be needed for different 

customers. Soft drinks might also be handled as well as beer. The number of SKUs can 

reach very high levels very quickly. This puts great emphasis on stock control and the 

organization of production to meet diverse demands. 

Warehouses also have to deal with empty containers returned from trade. These will 

include kegs and bottles in most cases. These will require sorting and cleaning (Chapter 

21). Warehousing, therefore, relates to the storage and despatch of beer, and the 

reception, storage and issue of packaging materials. These packaging materials can be 

quite simple in a warehouse dealing only with large pack beer in cask or keg but 

extremely diverse in a warehouse storing small pack beer in bottles or cans in various 

types of shrink wrapping and boxes, cardboard cases and crates. 

22.2.1 Principles of warehouse operation 

Warehousing requires a large amount of space and the multi-handling of packages. No 

matter how low the stock holding, warehousing is expensive, so the area required, though 

large, must be kept to a minimum by stacking containers and packages on robust pallets 

or by using undamaged spacer boards when handling containers palletless. Space must be 

utilized carefully to minimize the extent of mechanical handling by forklift truck. 

Computer systems are widely used to manage the inventory in the warehouse. 

Information can be fed to the computer from radio data terminals on the forklift trucks 

and entries made as beer comes off the packaging line into store and leaving the 

warehouse floor onto a vehicle. 

Stock control 

Beer deteriorates with age and will have a defined shelf-life that can vary from four 

weeks for cask beer to 52 weeks for highly stabilized beer in bottle or can. It is 

essential that stock must be rotated on a `first in first out' basis. Most breweries have 

strict rules about the age of beer that may be released to a customer, to ensure optimum 

quality when the beer is drunk. The exact age on release will depend on how much 

stock is held in depots or regional distribution centres. This illustrates just how critical 

careful stock control is to the success of the brewery. Beer that goes over age in the 

warehouse must either be destroyed (which has implications for excise duty) or 

recovered and blended back at, say, 5% into mainstream beer. These are both 

expensive operations. 


Storage conditions 

Conditions in the warehouse are critical for beer quality. Filtered beer must be stored 

above freezing point and at < 22 ëC (< 72 ëF). At low temperatures a chill haze may 

develop and at temperatures greater than about 22 ëC (72 ëF) flavour stability is 

threatened. The storage of cask beer is more critical because of the presence of live yeast 

and the required secondary fermentation, which continues in the pub cellar. Cask beer is 

best stored between 10 and 17 ëC (50 and 63 ëF). Relative humidity must be kept low by 

adequate ventilation. Moist conditions promote the deterioration of secondary packaging 

on cases of bottles and cans and render the beer unsaleable. There must be high standards 

of housekeeping. There should be written hygiene procedures that must be adhered to. In 

this way pests such as mice, pigeons and cockroaches will be deterred. 

Record keeping 

In most countries the tax (duty) on beer is paid when the beer leaves the warehouse. It is a 

legal requirement to maintain records of the volume and strength of the beer being made 

and subsequently sold. These records must be available for inspection and audit at any 

time. Governments also usually have legal requirements relating to the volume of beer 

sold in packages, which may be related to minimum or average contents. This again 

requires record keeping. 

Product traceability in trade is now important. Customers are more demanding of 

product quality and are frequently encouraged to voice their opinion of the beer by 

telephoning help lines. It follows that there must be very sound systems in place for 

labelling containers and subsequently tracing the beer's history in the brewery, the 

warehouse and in the trade. 

22.2.2 Safety in the warehouse 

The warehouse is one of the most dangerous places in the brewery, where frequently the 

highest numbers of accidents occur. This is a result of the juxtaposition of forklift trucks and 

the manual handling needed to get the beer in the right place. In the UK the highest incidence 

of `lost-time accidents' is in warehouse and distribution work (BLRA, 1987±1995). In the 

period 1987±1991 there occurred an average of 1,000 accidents each year resulting in lost 

time of more than three days. Of these over half were associated with handling, lifting or 

carrying beer in its various containers. This resulted in the introduction of `Manual Handling 

Regulations' in 1993. There are similar regulations in place in other countries. The basic 

principle of these regulations is to control risk so far as is reasonably possible: 

· avoid hazardous manual handling where possible 

· assess any hazardous operations that cannot be avoided 

· remove or reduce the risk of injury so far as is reasonably practicable using the 

assessment as a basis for action. 

These regulations place a duty on employers and employees with the objective of 

reducing the number of accidents and creating a safer workplace. The employer must 

have a safe system of work, which is written down, monitored, audited and improved. 

Training discharges the responsibility of employers for much of this. The training must be 

ongoing and this can place a considerable strain on resources of both trainer and 

employee. The rewards are, of course, considerable as the `lost time' of the accidents can 

be reduced with big savings in cost and improvements in customer service and greater 

well-being of employees. 


22.3 Distribution 

Distribution is one activity that has been subjected to intense scrutiny and change in 

recent years. This is particularly the case in countries with mature beer markets in Europe 

and North America. In these areas beer consumption in the early 21st century is falling or 

at best is static. Competition between companies is intense and all companies seek to call 

themselves to the attention of the customer to gain a competitive advantage, however 

small. Product quality is of paramount importance. It is simply not possible to sell beer 

that is not of the highest standard. It is very difficult, therefore, to gain advantage on the 

basis of product quality. There will be differences in beers between brewers but the 

quality will be very similar. Likewise, low-cost production has become almost a given 

requirement for the national and international brewer. The definition of low cost will vary 

from country to country but competitors know the targets in any particular country and 

will measure their performance against those targets by benchmarking to ensure that costs 

are under control. It is therefore very difficult to gain advantage on the basis of the cost of 

production. 

The brewing industry was slow to understand the significance of customer service. In 

the UK this was certainly a result of the ownership of the retail outlets for beer. These 

outlets were, in the main, in the hands of the brewers of the beer until the report of the 

Monopolies and Mergers Commission into restrictive practice in the brewing industry 

was published in 1989. Before this time there was little incentive to devote too much time 

to the needs of the customer when that customer was a public house owned by the 

brewery. Everything changed when restrictions were placed on the number of retail 

outlets that could be owned by beer producers. Competition intensified overnight. 

Different pressures applied in other countries of the developed world. One of the main 

factors was a major increase in interest in `healthy life-styles' that occurred in the USA 

and Europe from the early 1980s. Beer sales fell and brewers realized they were not only 

in competition with each other but with manufacturers of soft drinks. A competitive 

advantage was needed and brewers looked to customer service to provide it. Much effort 

was therefore put into distribution and to understanding the needs of the customer. In 

countries where beer consumption is still growing rapidly, such as China, and parts of 

South-East Asia and parts of South America, distribution is still important but it is 

unlikely to be the main agent of competitive advantage. 

22.3.1 Logistics 

Beer supply in breweries was formerly in the hands of a transport department that then 

became known as distribution. The controlling interest was frequently the maintenance of 

the vehicles rather than the requirements of the customer. Management was mainly 

concerned with load planning, i.e., ensuring the optimum weight of beer on the vehicle 

and ensuring the most efficient route to conserve fuel. The use of logistics is concerned 

with optimum deployment of stock and vehicle movement. This implies the integration of 

planning and distribution. The driving force can now be the customer because the 

optimization can start with the customer needs and stock deployment can be built around 

this. Thus the `pull' style of planning develops, starting with the customer requirement 

and finishing with scheduling the number of brews per week in the brewhouse. 

There has generally been a big increase in the demand for small pack beer for 

consumption at home (BLRA, 1999), usually bought through supermarkets. This has 

introduced a new customer service dimension and one that has brought the producer 


much closer to the consumer. To meet this diverse demand regional distribution centres 

away from breweries have been set up in the UK and have developed in other parts of 

Europe. Regional distribution was always a feature of North American distribution 

because of the great distances involved and customer service was part of the North 

American brewer's way of life earlier than it was in Europe. 

Planning 

The core of successful logistics is planning. The requirements of the customer must be 

interpreted into plans that can deliver the right amount of stock at the right time and 

efficient productionofthatstockbythebrewery.Theplanisderived fromhistorical sales 

data and forecasting. The former is amatter of fact and the latter is frequently amatter of 

conjecture! Software systems are, however, now available that can be used to telling 

effect in forecasting customer needs. This is important because for draught beers 

customers often run on very low stocks so that their working capital is reduced. The 

customer expects the supplier to deliver the beer immediately he requires it and in the 

appropriate state for dispense. In other words the customer wants the brewer to hold the 

stock. Good planning systems can turn this to advantage by using the principle of `just in 

time' delivery so that brewery stock is also minimized. `Just in time', must not become 

`just too late'! If cask beer is being supplied then this is more difficult because of the 

finite time required for adequate secondary fermentation (Chapter 23). 

Customer demand is the driving force behind most logistic systems. However, for 

national big brands (> 800,000 hl pa, approx. 500,000 imp. brl) a system of `vendor 

managed inventory' can be used. Here the supplier controls the stock in the supply chain 

and delivers to the customer depending on the supplier's view of the customer's 

requirements. This requires having software to monitor stock at the customer's premises 

and delivering when that stock requires replenishment. This means that the beer should 

be in the right place to meet demand and should guard against the customer who orders 

beer when, based on his sales, he does not really need it. Vendor managed inventory can 

thus prevent spurious shortages in the supply chain. 

Delivery 

Traditionally two types of delivery from breweries were recognized, primary and retail or 

radial. Radial delivery was very common in the UK where pubs were situated close to 

breweries and a retail fleet of vehicles would be based at the brewery and would make 

several trips to customers probably five or six days a week returning to the brewery each 

time with empty containers. This is still a major feature of regional breweries in Britain 

and the rest of the world. 

Primary delivery constitutes delivery of beer from the brewery to a depot or regional 

distribution centre where stock is held. This is usually by heavy articulated vehicles with 

gross weights of up to 40 t. Draught beer is most often handled palletless, with groups of 

containers separated by plastic spacer boards. This reduces weight and allows a greater 

payload of beer. 

As brewers have sought to differentiate their service to gain commercial advantage, 

the composite delivery has become common. In this system the brewer offers to deliver a 

composite mixture of beer in all package types and soft drinks and wine and spirits. This 

requires having the facility to do this at the brewery or depot. A retail or radial fleet must 

operate from the depot to the customer but frequently loads can be `made-up' for the 

`primary' vehicle and merely transferred to the retail vehicle and delivered to the 

customer. 


Afurther complication is the delivery of beer by road tanker to asecondary location 

for further packaging. This might be beer for kegging or bottling or canning. In this 

situation the supplying and receiving departments must agree on the rules for delivery 

and acceptance to avoid the costly business of returning unemptied tankers to the source, 

which is to nobody's benefit. This is all the more relevant when the supplier and the 

receiver are part of the same company. 

This whole business is about delighting the customer whilst operating at the lowest 

cost. Organization has been helped by seeking the registration of the business to quality 

standards such as ISO 9002. These systems do not, of course, result in improved quality 

to the customer's benefit per se but they do provide awritten guide to asystem of work 

that can be audited and subjected to continuous improvement. Another benefit of these 

systems resides in the fact that if they are not adhered to the registration can be removed 

bytheauditingauthorityandthatwouldbeaconsiderablesourceofembarrassmenttothe 

brewer and could put him at severe disadvantage. There is, therefore, every incentive to 

maintain and improve the system. 

22.3.2 Quality assurance 

The second major aspect of modern distribution along with logistics is the assurance of 

quality. This is made much easier if the beer in the brewery is produced to sound 

standards. Customers rightly expect their beer to be delivered to the highest quality 

standards and, irrespective of any nationally approved quality management system, 

reserve the right to inspect the brewers' premises at any time. Quality is assured most 

effectively by the brewer having in place strict standards for the release of beer to 

delivery vehicles, whether for primary or retail supply. There should be aprocedure in 

place the operation of which prevents sub-standard beer being delivered. This requires 

setting specifications for the beer in the warehouse and ensuring that the beer meetsthose 

specifications before it is delivered. This should avoid the requirement for recall of beer 

from the supply chain. No matter how good the system, however, some beer will be 

released which will not be of the right quality. It follows that product recall procedures 

must also be in place to assure the customer that the brewer is in charge of quality 

throughout the supply chain. 

There will need to be some tolerance in the analytical values of the beer so as to 

prevent unnecessary hold-ups in supply. The system must also allow for `beer on hold' 

whilst further analysis is carried out. Amost important aspect of this release of beer to 

trade is the flavour of the beer. Aproperly trained panel, as part of the quality assurance 

beer release process, must taste all beers. Typical beer release parameters for alager beer 

are shown (Table 22.1). Analyses will be determined on samples drawn from each brew. 

Beer is likely to be released to primary delivery throughout the day. It is very 

important therefore that people on every shift in the warehouse or packaging department 

are trained in beer release procedures. They must have an understanding of the analytical 

values in Table 22.1. Of course, the beer should not be packaged if it does not meet the 

required specification. The required specification is often achieved by blending the 

contents of different bright beer tanks of the same beer quality. The accurate setting of the 

blend and the computation of the weighted average analysis are critical for success. 

It is clear that quality assurance for distribution demands effective quality assurance 

throughout the supply chain in the brewery. This in turn demands effective 

communication from the brewhouse through to fermentation and to beer processing 

and packaging. Beer should not move along the supply chain unless it meets the in- 


Table 22.1 Release parameters for a lager beer 

Parameter Target value Permitted range 

OG (ëSaach) 40 38.5±41.5 

PG (ëSaach) 8 6.5±9.5 

ABV (%) 4.1 3.9±4.3 

pH 4.0 3.8±4.2 

Colour (ëEBC) 8.5 7.0±10.0 

Bitterness (ëEBU) 19 16.5±21.5 

CO2 (vol.) 2.1 2.0±2.2 

Dissolved oxygen in BBT (mg/l) 6 5 min. 

Haze (ëEBC, Monitek)

23 

Beer in the trade 


In recent years there have been large increases in the consumption of beer at home in 

bottles and cans (BLRA, 1999a). This has been particularly the case in Europe but has 

also occurred in North America. There has also been an increase in the consumption of 

beer in bottles in licensed premises. This has been associated with a greater emphasis on 

food being served along with beer. Frequently this beer is drunk directly from the bottle, 

which is usually of dark brown or green glass. In this situation the beer cannot be seen, 

which denies the drinker one of the pleasures of drinking beer, that of contemplating its 

appearance in the glass. Despite these trends the quality of draught beer drunk in the 

trade, i.e., in public houses and the like, is very important. Quality is influenced not only 

by the quality of the beer delivered but also by the conditions of storage in the public 

house and the conditions of the dispense of the beer. This applies equally to cask beer 

where a controlled secondary fermentation takes place in the container and to keg beer 

dispensed from a sterile container by gas top pressure. 

Advertising of national and international beers centres on creating an ìmage' of 

brand values which must be consistent. Brewers simply cannot leave the control of the 

quality of their beer to chance. They must ensure that good practice is applied in storage 

and at dispense, so that when the beer is drunk the intended quality is attained. This can 

be done by employing teams of people to advise customers on best practice and to 

ensure it is effected or to contract out this operation to a specialist company. Both 

systems are satisfactory. The owner of a strong brand can insist on the quality standards 

of storage and dispense being a prerequisite to allowing the customer to sell the 

product. 

Beer is consumed in a wide variety of premises. These can vary from large city 

centre public houses and bars with extensive cellar facilities to beach bars where beer is 

served in ambient temperatures of over 30 ëC (86 ëF) with a simple on-line cooler. It 

follows that there must be some basic principles of management of beer in the trade if 

the brewer is going to satisfy the maximum number of customers in this variety of 

outlets. 


23.2 History 

Home brewing diversified into acraft industry to serve local communities. At this time 

all beer was dispensed on draught. Frequently the beer would be collected from the 

brewery in pots or jugs and then dispensed at home. Taxes also had an influence. In the 

UK there was atax on glass in the 19th century, which was removed at the start of the 

20th century, and this led to brewers examining the potential of bottling beer. In the UK 

thetraditionalcontainerforbeerwasthecask,originallymadefromwoodbycoopersand 

now usually made of stainless steel or aluminium (Chapter 21). There is still ademand 

for cask beer but by 1998 this type of beer represented about 10% of total draught beer 

sales in the UK (BLRA, 1999b). In spite of intense interest in this type of beer by microbrewers 

all over the world cask beer is much less significant in other countries and 

certainly so in the countries of largest production: USA, China and Germany. In these 

countries, and in most others, the most important type of draught beer is that dispensed 

from apressurized container by gas top pressure in astabilized and filtered form. This 

applies to both lager and ale. 

The development of keg beer in the UK was stimulated by the presence of US airmen 

in the Second World War. Apparently they did not take to traditional British cask beer. 

This prompted development work by the brewer Greens of Luton who refined the process 

of putting sediment free, carbonated beer into metal containers and so keg beer was 

created (Bamforth, 1998). 

23.3 Beer cellars 

Traditionally beer was served from cellars. In the main this allowed for better 

temperature control and provided a space in which the beer containers could be organized 

and proper stock control carried out. For bars serving large volumes of beer in a week, 

say > 15 hl (9 imp. brl), an organized space is essential or it will be impossible to manage 

stock rotation and ensure that the beer is served on a `first in first out' basis. The `cellar' 

will not be below ground level in all bars but the principles of cellar design will still 

apply. 

In most countries beer is classified as a foodstuff (Hunter, 1993) and so is subject to 

some food hygiene regulations. This means regular inspection by agents of local or 

national government. It follows that companies responsible for the management of the bar 

in which the beer is sold must have standards for the beer cellar which must be written 

down, continually assessed, and improved. This will be to the ultimate advantage of the 

customer and should be part of the way the brewer seeks to gain competitive advantage. 

23.3.1 Hygiene 

A beer cellar should be kept clean, and cleaning is greatly aided if the original design is 

sound. It is therefore surprising how infrequently very clean cellars are encountered. 

Floor surfaces should be hard, impervious, and smooth and should slope to a drain or 

sump. This makes regular hosing of the floor easy to manage. The sump should be 

cleaned weekly. Walls and the ceiling should have a smooth finish and can be tiled or 

treated with an anti-fungicidal paint. This allows for frequent washing and disinfection 

and the inhibition of mould growth. All equipment should be stored carefully and in a 

clean place away from the main area of beer dispense. Beer cellars should not be used for 


storing other foodstuffs. Bacteria and wild yeast can be present on foods such as cheese 

and these can infect and will affect particularly cask beer. 

23.3.2 Temperature 

Temperature control in the brewery warehouse has been discussed (Chapter 22). 

Temperature control in the beer cellar is equally important if the beer is going to be at its 

bestwhenserved.Kegbeerwillalmostcertainlybepassedthroughabeercoolerenroute 

to the glass. Nevertheless, the workload of that cooler will be reduced if the temperature 

of the cellar is controlled. Beers will deteriorate more slowly in acool cellar. 

Cask beer requires more defined temperature control to allow effective secondary 

fermentation to take place and to prevent the development of flavour defects and to guard 

against possible infection as yeast activity subsides. The ideal cellar temperature to 

satisfy all these requirements is between 10 and 17 ëC (50 and 63 ëF). Temperatures were 

originally controlled by ventilation alone but now a cellar refrigeration unit is used. This 

should be fan assisted and must be properly sized for the cellar volume. Heat generating 

units of any sort (e.g. ice makers) should not be located in the cellar so as not to compete 

with the temperature control unit. 

Ventilation of the cellar or storage area is important in its own right. Air, which is not 

changed, becomes stale, humid and musty and can promote the growth of moulds. There 

is also the safety hazard for the dispense of keg beers associated with the use of carbon 

dioxide and nitrogen. A rise in concentration of these gases can be fatal for persons 

working in cellars and they must be removed by ventilation and there must be training for 

staff to cover all aspects of the dangers of these gases. 

23.3.3 Lighting 

Cellars must be adequately lit and the level of lighting required is frequently laid down by 

statute. The best lighting is provided by fluorescent tubes with splash-proof covers 

(Hunter, 1993). In larger cellars emergency lighting is now often installed. 

23.4 Beer dispense 

Beers are judged visually as well as by flavour and aroma. How a beer appears in the 

trade is greatly influenced by how it is dispensed. This applies to both cask and keg beer 

and it is controlled by both the dispense equipment design and how it is used. This can be 

subject to historical regional influences. In the north of England drinkers generally prefer 

cask beer with a tight creamy head. In the south of England a loose open head is often 

preferred. As brewers seek to differentiate their products and gain competitive advantage 

greater emphasis has been placed on the quality of dispense. This is a key factor in 

defining the brand and major brands usually have unique dispense equipment and defined 

conditions of use. 

Many of the widely differing beer selling premises are not owned by brewers, indeed, 

the beer may be supplied by a general wholesaler. Often staff serving beer are employed 

on a temporary basis or may not have received brewery training. It follows that the beer 

dispense equipment must be robust and tamper-proof if the brewer is to be satisfied with 

the presentation of his beer. It is also important that waste is minimized because fierce 

competition has led to a general reduction in the price the brewer obtains for his beer. In 


turn this has led to a demand for a more robust cask beer that is easier to handle in the bar 

and less prone to wastage and dispense problems. 

All recent developments in the dispense systems of keg and cask beers have sought to 

guarantee the quality in all kinds of places selling beer. However, there are aspects of the 

equipment that if neglected will lead to poor beer quality in the trade with a consequent 

loss of immediate sales and an adverse effect on the long-term market share of the brand. 

23.4.1 Keg beer 

Keg beers, ales or lagers, are packaged into sterile metal containers of size 25±100 hl, 

although in the UK barrels are still used (36 imp. gal or 163 hl). The beer has a defined 

content of carbon dioxide and will normally be dispensed by applying a top pressure of 

carbon dioxide gas or by using a mixed gas usually containing carbon dioxide and 

nitrogen. This is commonly called the `free-flow' system. Pumps can also be used to 

assist the flow. A further refinement of the pumped system is to use a metered dispense 

where the beer is dispensed in pulses of a set volume, often 250 ml or, in the UK half an 

imperial pint (284 ml). Lagers are usually dispensed in a free-flow system with carbon 

dioxide alone whilst ales can be dispensed with carbon dioxide or with a carbon dioxide/ 

nitrogen mixture. This presents the first challenge in assuring the quality of dispense. 


Every keg beer has a CO2 specification. There must be careful adjustment of the pressure 

of CO2 entering the keg to ensure perfect dispense of the product. The beer will be 

susceptible to picking up carbon dioxide or losing it. In either case dispense problems 

will follow and poor quality will result. If there is insufficient top pressure gas will break 

out of the beer into the head space of the keg and into the beer lines up to the dispense 

tap; a flat beer will result. If the top pressure is too high then more CO2 will dissolve into 

the beer than is breaking out and the beer will over carbonate with consequent fobbing 

problems at the dispense tap. 

Temperature is critical when using 100% carbon dioxide for dispense. If the temperature 

of the beer rises 1 ëC (1.8 ëF) then an additional 0.1 bar (1.47 lb./in. 2 ) of pressure is needed. 

If the temperature falls by 1 ëC then the pressure needs to be reduced by 0.1 bar. 

Temperature fluctuation is a common cause of beer dispense problems and one of the most 

frequent causes of complaints to the brewery. It is not a brewery problem provided that the 

carbon dioxide content of the beer was in specification on delivery. It is easy to eradicate in 

the trade by attention to the details of temperature control in the cellar. 

Problems also occur with beer dispense in high ambient temperatures (Stanley, 1999). 

The temperature of the beer line to the tap is often at least 3 ëC (5 ëF) higher than the cool 

cabinet or cellar where the kegs are stored. Carbon dioxide will therefore break out of the 

beer in the dispense line and excessive fobbing at the tap will be the result. In this 

situation the CO2 pressure should be set to be appropriate for the lower temperature. If the 

cool cabinet is at 2 ëC (36 ëF) and the beer line at 5 ëC (41 ëF) then the top pressure should 

be increased from 0.7 bar to 0.9 bar (10 to 13 lb./in. 2 ). This would apply to a beer with, 

say, a CO 2 content of 2.6 vol/vol. A further complication is that over a period of a few 

days and as a consequence of the higher top pressure of the gas the carbonation in the 

beer will increase to about 2.9 vol/vol and fobbing at the tap will be the result. Pressure 

compensations are also required when forcing beer by top pressure along long runs of 

horizontal or vertical pipe. For every horizontal metre an additional 0.011 bar (0.16 lb./ 

in. 2 ) is required and for every vertical metre an extra 0.108 bar (1.59 lb./in. 2 ). 


Thus it is important to pay attention to the correct dispense pressures for the 

appropriate temperatures and lengths of installation pipe when using 100% carbon 

dioxide for dispense. It is highly desirable to maintain the beer at as constant a 

temperature as possible and to avoid major temperature differences between the beer in 

the store and the dispense lines to the tap. 

Mixed gases 

If the beer temperature cannot be held constant then dispense problems will always occur 

when using 100% carbon dioxide as the top pressure gas. Aresolution to this problem is 

to introduce another gas to the headspace of the keg thus increasing the system pressure 

without the potential for increasing the carbon dioxide content of the beer. This can be 

done by using nitrogen and by using the correct blend of nitrogen and carbon dioxide. 

The partial pressure of the CO 2is enough to maintain the required carbonation level of 

the beer. Nitrogen is not as soluble as CO2 and will not affect the taste of the beer unless 

used at excessively high pressures. This system of mixed gas dispense can be used to 

dispense avarietyofpale beersofdelicateflavour invarying ambientconditions, without 

over carbonation. If the beer temperature cannot be controlled to better than Ô2ëC 

( 3.6ëF) then ablend of 80% CO2 and 20% nitrogen at 1.03 bar (15 lb./in. 2 )will 

eliminate dispense problems on a2.6 vol/vol CO2 beer. 

Acommon system for using mixed gas is to use ablender to accurately proportion the 

gases from cylinders. These must be very carefully set and not tampered with. (The best 

systems will not operate when one of the gases has run out!) Pre-mixed gas can also be 

used, but this is often much more expensive than blending on site. The blends are 

sometimes not precise and are often low in carbon dioxide. They must therefore be 

checked with agas analyser before use. 

Adifferent way of exploiting the use of mixed gas has been developed in the UK. 

Small amounts of nitrogen (10±50mg/l) have amajor positive effect on beer foam 

stability (Bishop et al., 1975) and stability of draught stout is much improved by the 

inclusion ofnitrogen in the beer. This principle has been extended to ale. This has largely 

been amarket-led initiative to improve foam quality on keg ale and so provide aproduct 

similar in appearance to cask beer drawn by hand pulling. Nitrogen is added to filtered 

beer prior to packaging (Chapter 21) usually at rates of 15±40mg/l. Normally this beer 

has a carbon dioxide content of around 1.1 vol/vol which is much lower than keg beers 

not containing nitrogen (up to 2.6 vol/vol). 

Nitrogen-containing draught ales are dispensed with a mixed gas blend of carbon 

dioxide and nitrogen. The nitrogen now has the role of avoiding the loss of nitrogen in the 

beer during its dispense life as well as providing the motive force to drive the beer to the 

dispense tap without using excessively high levels of carbon dioxide (Lindsay et al., 

1996). Ale is not drunk in large volumes outside the UK and so this technique is of local 

interest. The introduction of nitrogen to lager beer has been tried but, because of the 

softening of the palate of the beer, has not always been commercially successful. 

Dispense problems with nitrogenated lager can be severe because of the much higher 

carbon dioxide level in the beer (2.6 vol/vol) which, if lowered, produces an insipid 

mouth-feel. 

Beer pumps 

Beer pumps can also be used to protect against the effects of temperature changes 

between the beer store and the dispense tap. Mechanical pressure is used to push the beer 

to the tap rather than the gas top pressure on the keg. This avoids the use of any gas other 


than carbon dioxide. Pressure on the keg can be set at 0.8 bar (12lb./in. 2 )and can be 

easily increased to 1.4 bar (20lb./in. 2 )in the beer line by apump. Systems using pumps 

require more maintenance, and hence are more costly to operate than systems relying on 

gas pressure alone. Adevelopment of the pump particularly used in the UK was the 

positive displacement meter. Beer was dispensed at high speed, in fixed pulses usually of 

one half imperial pint (284ml). These systems are not being widely used. They are 

expensive to maintain. Further, in countries where the head on the beer is ruled as being 

part of the beer measure then the meter is difficult to operate to customer satisfaction. 

The properly balanced mixed gas system offers the most flexible solution to providing 

excellence in dispense at lowest cost without the danger of over carbonation. 

Beer lines 

Abeer line is the route of supply from the storage container in the cellar to the dispense 

tap on the bar. The line itself is usually made of polythene or PVC tubing of appropriate 

diameter. Poordesign of the beer line or poor cleaning will adversely affect quality ofthe 

dispensed beer. The system should be designed and adjusted to give aflow rate of about 

12.5sec for 500ml (14 sec/imp. pint) of beer (Heron, 1992). This demands balancing the 

top pressure used against the restriction provided by the line. Although this time period is 

a good `rule of thumb' competitive advantage is sometimes sought by brewers by 

deviating widely from this norm. This particularly applies to nitrogenated ales where 

longer dispense times are demanded to heighten anticipation of the drink. 

Long beer lines (>30m; 98ft.) should be cleaned using an installed system involving 

re-circulation of the cleaning fluid. Acid based cleaners should be used frequently to 

avoid the accumulation of beer stone. Transparent pipe is best so that accumulation of 

soil can be seen easily and gas `break out' observed. There have been recent 

improvements to line cooling systems. Originally line cooler units were situated below 

the bar. Cooling used achilled water bath with the beer passing through in coils of tubing 

immersed in the chilled water. Further developments used ice-bank coolers with the beer 

subject to rapid cooling by ice to supply outlets with high demand. Modern units have 

beer re-circulation pumps to allow variable temperature control on ales and lagers. 

Amajor developmenthasbeen the use ofremotecoolers, which have the advantage of 

removing the heat source from below the bar. They are usually situated immediately 

outside the cooled beer cellar. Beer is supplied to the bar through an insulated pipe 

containing cooling water known as a`python'. Large systems can accommodate up to 16 

separateroutesforbeer,knownasproductcoils(Fig.23.1).Theprimarylagerpythoncan 

be held at a lower temperature than the secondary ale python, which can take recirculation 

water from the lager line to save energy (Fig. 23.2). Pythons are now 

sometimes used with under-counter cooling modules, which can further trim beer 

temperature, and they will take chilled water from the python. 

Manufacturers of keg beer dispense systems usually provide advice on the operation 

and cleaning of their equipment. Follow that advice! Many problems of beer quality 

encountered in bars are simply a result of poor cleaning or of operating systems at the 

wrong pressures or with poor temperature control. 

23.4.2 Cask beer 

The cask beer market in the UK still amounts to the considerable annual volume of about 

6 million hl (3.7 m imp. brl). Cask beer requires very different treatment in the trade from 

keg beer. The major factor affecting cask beer dispense is that cask beer contains live 


Water bath 

temperature 

gauge 

Control panel 

Thermostats 

Water pump 

Filler cap 

Condenser 

Agitator 

Fan 

Pump water return 

Compressor 

Product coil 

Fig 23.1 Large remote beer dispense cooler (Lindsay, 2002). 

C 

B 

A 

ALE LAGER 

D 

F 

G 

RETURN 

FLOW 

H 

E 

Filler cap 

Evaporator 

Product coil 

Compressor 

electrics 

Fig 23.2 Secondary keg ale python installed with a remote beer cooler and lager python. A, 

kegs; B, keg connector; C, PVC or polythene tubing; D, ale python; E, lager python; F, ale 

python water re-circulation line (return); G, ale python water re-circulation line (flow); H, 

python driver; I, remote cooler; J, lager python water re-circulation line (return); K, lager python 

water re-circulation line (flow); L, lager product line. The ale python takes re-circulation water 

from the lager line (Lindsay, 2002). 


L J 

K 

I

yeast to effect the secondary fermentation to give condition to the beer. Cask beer, 

therefore, has ashelf-life of only 3±4 weeks and everything must be geared to ensuring 

that the beer is drunk in perfect condition within this time period. 

Delivery of beer 

An important point is to check that the beer being delivered is within the age profile laid 

down by the brewery so that, given time to settle, it will be dispensed within the shelf-life. 

This requires knowledge of the throughput of the bar. Containers should not be leaking and 

the shives and keystones (Chapter 21) should be intact and clean. The casks should be 

located by the draymen in the position where they are to be dispensed. The more they are 

moved the less satisfactory will be the fining action on which beer clarification depends. 

Stillaging 

Casks should be positioned on the level on the stillage (a wooden or metal frame on 

which the casks are raised off the floor of the cellar) with the shive uppermost and the 

keystone centrally located at the bottom. The casks should then be secured with wooden 

wedges (scotches). The sediment will now collect in the belly of the cask. Sometimes a 

tilting stillage is used where the cask can be progressively tilted forward as the beer is 

dispensed. This has to be done with care or the sediment will block the dispense tap. 

An alternative system is to stillage vertically with the keystone uppermost (Hunter, 

1993). The beer is subsequently dispensed through ahollow rod and so the cask must be 

tilted by positioning awedge directly under the location of the keystone. The sediment 

then collects away from the rod inlet. Purists would argue this is not genuine cask beer! 

Pegging (spiling) 

The secondary fermentation in the cask produces carbon dioxide, which causes arise in 

pressure in the container. This pressure should be released gently or the sediment in the 

cask can be excessively disturbed when the cask is tapped for dispense. 

The shive should first be scrubbed clean and then a soft porous peg is driven carefully 

into the depression in the shive (tut). This will allow the beer to `work' gently as the carbon 

dioxide escapes through the porous peg. This operation should be carried out between three 

and six hours from delivery. The soft pegs should normally be changed twice a day. The 

beer may continue to give off gas for 12±24 hours. When this working has stopped the soft 

peg should be removed and replaced with a non-porous, hard peg. The hard pegs should be 

eased daily to prevent further pressure build up. Best practice demands the insertion of a 

soft peg whilst the beer is being dispensed as this acts as an air filter. Between dispense 

sessions a hard peg should be inserted to maintain beer condition. 

Tapping 

The keystone must first be scrubbed clean and then a clean tap should be smartly driven into 

the already pegged cask with a single hammer blow. A tap should never be inserted into an 

un-pegged cask. The beer should be tapped about 12±24 hours before it is required for sale. 

This can, of course, vary but most cask beers will have dropped bright (yeast and sediment 

will have settled) and have stopped working in 12±48 hours from delivery although on 

occasions 72 hours may be needed. Before connection to the dispense line about 250 ml ( 

Ý imp. pint) of beer should be drawn off through the tap to clear any sediment around the 

tap. This beer should be discarded. A further 250 ml ( Ý imp. pint) should then be drawn 

and checked for clarity, aroma, and taste. If the beer is found to be true to type the container 

can be connected to the dispense line so the beer can be sold. 


Tilting 

Stillaged casks should be tilted forward when about one-third of the contents of the cask 

have been sold and not before. Afall of about 7cm (2.75in.) is needed. This allows the 

maximum yield of beer from the cask. The task should be done carefully so as not to 

disturb sediment. 

Dispense 

Originally beer was dispensed into ajug through the tap from the stillaged cask at the 

back of the bar. In Victorian times the casks were placed in acellar but the beer was 

still drawn from taps into jugsandbroughtupto the bar by`pot-boys'.The hand pulled 

pump, the beer engine, was developed in the early 20th century so that beer could be 

pumped up from the cellar to the bar. Beer engines were originally made of brass but 

are now almost always made of stainless steel or plastic to avoid any possibility of lead 

dissolvingintothebeer(Fig.23.3).Thesizeofthecylinderisusuallyaquarterorahalf 

imperialpint(150or300ml).Thetaponthecaskisconnectedtothecylinderbytubing 

which is usually aclear plastic (often PVC) and in this way yeast build up is easily 

seen. 

Typically, cask beers were dispensed at 10±15ëC (50±59ëF) and this was achieved by 

controlling cellar temperature. There has recently been ademand to drink cask beer at 

lower temperatures sometimes down to 7ëC (45ëF). Whether this is aresult of consumer 

demand or has been led by marketing departments seeking to promote keg beer is 

uncertain! This has led to cooling cask beer en route from cask to bar. Using awatercooled 

cylinder in the beer engine, which usually means it will have the smaller volume 

of aquarter pint, can do this. Sometimes cask beer pythons are used (Fig. 23.4) in outlets 

serving a range of cask beers. There have been systems for introducing a gas mixture of 

carbon dioxide and nitrogen into the cask beer before dispense to increase the evenness of 

the foam and decrease bubble size (Watts, 2000). This is moving cask beer towards keg 

beer in sensory experience and would not find favour with beer traditionalists. 

The spout from the beer engine that delivers beer to the glass can be of various 

designs. It usually ends in a plastic removable tip, the sparkler, which can have a varying 

number of holes. Thus beer can be delivered through a `swan-neck' or straight spout to a 

tight or slack sparkler or even to no sparkler! This has a profound effect on the physical 

state of the beer in the glass, which can have a tight creamy head with much foam cling, 

or a slack open head with little foam cling as the beer is drunk. The variations are almost 

endless and are subject to much regional influence in the UK and much debate wherever 

cask beer drinkers meet. 

Hygiene 

Cleanliness is absolutely essential when dealing with cask beer. All the implements used 

must be scrupulously clean. Shives, pegs and keystones should be kept in a sealed 

container to avoid the pick-up of airborne moulds. There should be a permanent wall 

mounting for mallet, dipstick, taps and brushes. Cask beer delivery pipes must be cleaned 

at least weekly. They should be flushed with cold water and filled with detergent to the 

recommended concentration; an effective system would use 1% sodium hypochlorite 

solution. Soaking should be for 10±15 minutes and then a further 10±15 minutes with a 

fresh batch of detergent (Heron, 1992) this ensures all deposited and suspended materials 

are removed. Finally there must be a thorough rinse with cold water. 

Clean glasses are important. This applies equally to beer dispensed from kegs. Most 

bars use a glass washing machine and this must be correctly used and maintained. It must 


Outer wall of 

cylinder, made 

of brass, stainless 

steel or plastic 

Beer in 

Piston connected 

to handle on bar 

Beer out 

Piston valve 

Cylinder 

containing beer 

Cylinder 

non-return 

valve 

Fig 23.3 Beer engine cylinder. This operates as asimple lift pump. When the handle on the 

bar is pulled down the piston rises. The non-return valve at the beer inlet opens. Beer is drawn 

in beneath the piston. The piston valve remains closed. The handle is then returned to the 

upright position. The cylinder valve now closes and the piston valve opens. The piston passes 

through the beer in the cylinder and the beer is lifted out of the cylinder to the bar counter and 

dispensed. This is often through a`swan-neck' device (see Fig. 23.4). The process is repeated 

several times to deliver a full glass of beer at the bar. 

have adequate water pressure and flow rate and the detergent temperature must be 

accurately controlled. Replacement of the salt must be carried out according to the 

number of washing cycles actually used not simply on aweekly basis. Washed glasses 

must be left on aclean, odourless surface to cool completely before use. Glasses should 

be inspected by acompetent member of the bar staff to ensure they are absolutely clean 

and free of taint. 


E 

Empty casks 

Empty casks should be removed immediately from the stillage. The tap should be 

removed and thoroughly washed. Ahard peg should be driven into the shive and acork 

should be driven into the keystone hole. The cask should then be put in asuitable place 

for collection by the drayman. 

Throughput 

The final factor affecting the quality of cask beer is the throughput of the product. This is 

much morecritical than for the chilled and filtered kegbeer. The minimum throughput of 

any one quality should be two containers per week (Hunter, 1993) of whatever size is 

appropriatetotheoutlet.Acaskbeercontainerwhenbroachedshouldbesoldinnotmore 

than 72 hours and preferably within 48 hours. The frequent cause of poor cask beer 

quality is the failure to observe these guidelines. 

There is atrend in the UK to offer customers avery wide choice of cask beers in `real 

ale houses'. This has often resulted in the flouting of the throughput guidelines, with the 

result poor beer has been put on sale. This is not good for the image of cask beer and is a 

reason for the increase in sales ofthe nitrogenated keg ales (Section 23.4.1). There isalso 

thetrendtoservethecaskbeeratlowertemperaturesinanattempttoextenditsshelf-life. 

D 

O 

N 

M J 

C 

H 

B A 

Fig 23.4 Cask ale python to allow dispense of cask beer to temperatures 7ëC, often used in 

outlets serving a range of cask beers in low volume. A, stillage; B, cask tap; C, clear PVC 

tubing (0.375 in., 10 mm, internal diameter); D, pump; E, polythene tubing (0.375 in., 10 mm, 

external diameter); F, ale python; G, lager (main python); H, ale python driver; I, remote cooler; 

J, polythene tubing (0.375 in., 10 mm internal diameter); K, check valve; L, polythene tubing 

(0.375 in., 10 mm external diameter); M, Ü pint hand-pull with swan neck spout; N, agitator 

(sparkler) to suit product on dispense; O, pump clip for product on dispense (Lindsay, 2002). 

K L 


I 

F 

G

These attempts are not always successful and the true flavour of the beer often suffers. At 

the present time the trend to keg beer continues and it seems likely that cask beer sales in 

the UK will continue to fall. 

23.4.3 Bottled and canned beer 

Drinking beer from bottles and cans is associated with drinking beer at home. Indeed in 

the UK off-trade sales of beer have risen from 10% of the total to over 30% of the total in 

the last 30 years (BLRA, 1999b). On-trade sales of small-pack beer were traditionally in 

returnable bottles but as a result of the influence of the package types sold in 

supermarkets, non-returnable bottles and even cans are now sold in the on-trade. Total 

sales of beer in cans in the UK have increased from 5% in 1974 to 25% in 2000 and in 

non-returnable bottles from < 1% in 1974 to 12% in 2000. Much of this beer is now 

drunk in pubs. 

International beer brands are frequently sold in several package types, usually in 

bottles, cans and kegs. To preserve the identity and integrity of the brand it is essential 

that the quality is maintained in all package types. In the trade, therefore, it is important 

that beer served from a bottle or can is served correctly. This is not, of course, as difficult 

as with keg or cask beer. The main requirement is to serve clean bottles and to make sure 

the bottles are displayed attractively with the labels visible to the public. Almost always 

the bottle contents will be served cold, in some countries as low as 3 ëC (37 ëF). The 

bottles will be stored cold behind the bar in a refrigerator or in a cooled cabinet with clear 

glass doors. 

The pleasure of drinking some beers is enhanced by the use of branded glasses. 

However, there is a marked trend amongst young drinkers to drink from bottles when 

some of the pleasures of drinking beer are lost. 

23.5 Quality control 

Most brewing companies operate to the principles of quality assurance so that sound and 

in-specification beer is delivered to their customers. The control of quality in the trade 

depends on the strict observance of all the principles of storage and dispense of beer 

discussed in this chapter. If these principles are followed then the customer should 

receive a perfect beer to drink. 

Many brewing companies employ inspectors to investigate premises where their beer 

is consumed. The focus is on the condition of the beer cellar, the storage of the beer and 

on the condition of the beer dispense systems. Throughout the world these inspection 

systems have come under pressure as brewers have sought to save costs and less trade 

quality control work is now carried out. This is dangerous for the competitive position of 

beer related to other drinks. 

23.6 New developments in trade quality 

Conventional beer analyses do not always indicate how the beer will perform in the trade 

under the variety of conditions now prevalent (Watts, 2000). Foam quality and head 

retention on a beer, for example, has always been an important attribute. Head retention 

has been measured by a number of techniques amongst the most robust of which is the 


Rudin method (Rudin, 1958), which measures foam collapse under standard and 

reproducible conditions. Recent studies have revealed beers with very similar Rudin 

values and yet having markedly different foam stabilities (Watts, 2000). This could 

representproblemsforabrewer.Foam stability isdefined asthe depthoffoamremaining 

on abeer after atwo-minute interval following astandard pour (Watts, 2000). A`visual 

assessment profile' has been developed to complement traditional beer analysis and so 

provide amore complete picture of the behaviour of the beer under trade conditions. 

Beers are assessed both quantitatively and qualitatively for foam quality, lacing, 

clarity, colour, `reflux' and `seeding' (see also Chapters 19 and 20). Reflux is a 

combination of the wave-like surge motion and overall bubble settling time on a draught 

dispensed beer. Seeding refers to the bubbles that rise up in the body of the beer. A 

trained panel of assessors examines each beer five times after a standard pour. This 

technique has been used to assess the effect of dispense head nozzle design on reflux. It 

was demonstrated that on a nitro-keg beer a nozzle pore size of 0.6 mm (0.024 in.) gave a 

significant increase in reflux compared to nozzle sizes of 0.7 mm (0.027 in.) and 0.8 mm 

(0.031 in.). This was ascribed to the greater shear when the beer was forced through the 

smaller nozzle. This provides considerable scope for changing the physical character of 

the beer where it is dispensed. 

Another factor of recent significance is the use of toughened glasses. There is 

considerably less seeding in lager beers poured into toughened glasses (Watts, 2000). 

This is presumably a result of the smoother surface not providing nucleation points for 

bubble formation. 

23.7 Summary 

Beer is drunk in a wide variety of premises, where the consumer forms his opinion of the 

beer. If the opinion is favourable then it is likely that the consumer will return and drink 

the beer again, as the brewer desires ± he seeks the delighted consumer. The brewer starts 

this process by assuring the highest quality of his beer to his customer, the vendor. It is 

then essential to ensure that the beer is stored and dispensed according to good practice at 

the premises where it is sold. This means scrupulous attention to the procedures of 

dispense for keg and cask beer, and detailed attention to hygiene of the cellar, dispense 

equipment and the glass. 


BAMFORTH, C. W. (1998) Tap into the Art and Science of Brewing, New York, Insight Books. 

BLRA (1999a) Beer and Pub Facts, Brewers' and Licensed Retailers' Association, London. 

BLRA (1999b) Statistical Handbook, Brewers' and Licensed Retailers' Association, London. 

BISHOP, L. R., WHITEAR, A. L. and INMAN, W. R. (1975) J. Inst. Brewing, 81, 131. 

HERON, P. C. (1992) Brewers' Guard. 121 (9), 29. 

HUNTER, A. R. (1993) The Brewer, 79, 159. 

LINDSAY, R. F. (2002) Personal communication. 

LINDSAY, R. F., LARSSON, E. and SMITH, I. B. (1996) Tech. Quart. MBAA, 33, 181. 

RUDIN, A. D. (1958) J. Inst. Brewing, 64, 238. 

STANLEY, J. M. (1999) Tech. Quart. MBAA. 36, 293. 

WATTS, E. (2000) Brewers' Guard. 129 (2), 20. 


Appendix: units and some data of use in 

brewing 

Table A1 SI derived units 


Table A3 Comparison of thermometer scales 



Table A6 Equivalence between Institute of Brewing units of hot water extract 

Table A7 Solution divisors of some sugars 

Table A8 Some properties of water 

Table A9 The density and viscosity of water at various temperatures 

Table A10 Some properties of water 

Table A11 The relationship between pressure and the temperature of steam 

Table A12 The solubility of pure gases in water at different temperatures 

Table A13 Salts in brewing liquors 

Table A14 Units of degrees of water hardness 

Table A15 Characteristics of some brewing materials 

Table A16 Pasteurization units (PUs) 

Fig. A1 The relationship between ethanol/water mixtures and the 

densities of the solutions. 

Introduction 

Although there is agradual trend towards the use of metric, or the derived SI units of 

measurement, their use is not universal and, of course, they were used less in the past. 

The nightmare muddle into which older British unit `systems' had degenerated is noted 

elsewhere,particularlyastheyrelatetocerealgrainsandmalt(Briggs,D.E.(1998)Malts 

and Malting, p. 742, London, Blackie Academic and Professional, or Aspen Publishers, 

Gaithersberg). American units, while having the same names as British units, often differ 

significantly in size. Thus, while the American gallon (gal. US) contains only 8pounds 

(lb.) of water or 8pints each weighing 1lb., the British gallon (gal. UK, or imp. gal.) 


contains 10lb., and so the British pint contains 1.25lb. of water. To have access to older 

and modern technical literature it is imperative that all units be understood and that 

readers have the information to allow them to convert reported values into units they are 

familiar with. In this book, wherever practical, metric or SI units have been used but 

equivalents in other units are given where this may be helpful even though this may be 

irritating to some readers, as it is to the authors. 

This appendix contains values for derived SI units and prefixes for them (Tables A1 

and A2), equivalents for thermometer scales (ëC, ëF and ëR; Table A3) and 

interconversion factors for metric, British (imp. or UK) and American (US) units (Table 

A4). Table A5 gives the relationships between SG and extract values, while Table A6 

gives the approximate equivalents between hot water extracts as determined by the older 

and the modern methods of the Institute and Guild of Brewing. Table A7 gives some 

solution divisors for sugars. Various properties of water are given in Tables A8, A9 and 

A10, including the pKw values and the amounts of oxygen dissolved from air or oxygen 

at different temperatures, the specific heats, latent heats, and surface tension at several 

temperatures. The relationship between steam pressure and temperature is summarized in 

Table A11. Table A12 gives the solubility of oxygen, nitrogen and carbon dioxide in 

water at different temperatures. Tables A13 and A14 give some data regarding salts in 

brewing liquors and units of water hardness. Table A15 contains information on the 

storage characteristics of some grist materials. Table A16 gives data on beer treatment 

temperatures and pasteurization units (PUs). Figure A.1 shows the relationship between 

the composition of ethanol/pure water, mixtures and their densities. 

As explained in the text, there are cases where there are no valid interconversion 

factors between analytical values. Thus the laboratory mashes specified in the 

Recommended Methods of the Institute and Guild of Brewing and Analytica-EBC 

(Section 1.15.1, p.9) aredifferent andcan yielddifferent extract recoveries andyields of 

soluble nitrogen. In consequence there are no valid factors for interconverting HWE and 

E, or the SNR and the Kolbach Index. 

Table A1 SI derived units. The base units are shown below 

Physical quantity Name Symbol Definition 

Energy joule J kg m 2 /s 2 

Force newton N kg m/s 2 ˆ J/m 

Power watt W kg m 2 /s 3 ˆ J/s 

Electrical charge coulomb C A s 

Electrical potential difference volt V kg m 2 /s 3 per A ˆ J/A per s 

Electrical resistance ohm kg m 2 /s 3 per A 2 ˆ V/A 

Inductance henry H kg m 2 /s 2 per A ˆ V/A per s 

Luminous flux lumen lm cd sr 

Illumination lux lx cd sr/m 2 

Frequency hertz Hz per s 

Pressure pascal Pa N/m 2 

Where m is the metre; kg is the kilogram; s is the second; A is the ampere; cd is the candela (luminous intensity); 

sr is the steridian (solid angle). 

Other base units are: Kelvin (K), thermodynamic temperature and temperature interval; mole (mol) molecular 

(or atomic) mass. 



Fraction Prefix Symbol 

10 15 

10 12 

10 9 

10 6 

10 3 

10 2 

10 1 

10 1 

10 2 

10 3 

10 6 

10 9 

10 12 


femto f 

pico p 

nano 

micro 

n 

milli m 

centi c 

deci d 

deka da 

hecto h 

kilo k 

mega M 

giga G 

tera T

Table A3 Comparison of thermometers showing the relative indications of the Fahrenheit, 

centigrade and ReÂaumur scales* 

Scale Boiling point Freezing point 

Fahrenheit (ëF) 212 32 

Centigrade (ëC) 100 0 

ReÂaumur (ëR)* 80 0 

Conversion of thermometer degrees 

1 ëC ˆ 1.8 ëF ˆ 0.8 ëR. ëC to ëR, multiply by 4 and divide by 5. ëC to ëF, multiply by 9. divide by 5, 

then add 32. ëR to ëC, multiply by 5 and divide by 4. ëR to ëF, multiply by 9, divide by 4, then add 

32. ëF to ëR, first subtract 32, then multiply by 4, and divide by 9. ëF to ëC, first subtract 32, then 

multiply by 5, and divide by 9. 

ëF ëC ëR ëF ëC ëR 

302 150 120 159.8 71 56.8 

284 140 112 158 70 56 

266 130 104 156.2 69 55.2 

257 125 100 154.4 68 54.4 

248 120 96 152.6 67 53.6 

239 115 92 150.8 66 52.8 

230 110 88 149 65 52 

221 105 84 147.2 64 51.2 

212 100 80 145.4 63 50.4 

210.2 99 79.2 143.6 62 49.6 

208.4 98 78.4 141.8 61 48.8 

206.6 97 77.6 140 60 48 

204.8 96 76.8 138.2 59 47.2 

203 95 76 136.4 58 46.4 

201.2 94 75.2 134.6 57 45.6 

199.4 93 74.4 132.8 56 44.8 

197.6 92 73.6 131 55 44 

195.8 91 72.8 129.2 54 43.2 

194 90 72 127.4 53 42.4 

192.2 89 71.2 125.6 52 41.6 

190.4 88 70.4 123.8 51 40.8 

188.6 87 69.6 122 50 40 

186.8 86 68.8 120.2 49 39.2 

185 85 68 118.4 48 38.4 

183.2 84 67.2 116.6 47 37.6 

181.4 83 66.4 114.8 46 36.8 

179.6 82 65.6 113 45 36 

177.8 81 64.8 111.2 44 35.2 

176 80 64 109.4 43 34.4 

174.2 79 63.2 107.6 42 33.6 

172.4 78 62.4 105.8 41 32.8 

170.6 77 61.6 104 40 32 

168.8 76 60.8 102.2 39 31.2 

167 75 60 100.4 38 30.4 

165.2 74 59.2 98.6 37 29.6 

163.4 73 58.4 96.8 36 28.8 

161.6 72 57.6 95 35 28 


Table A3 continued 

ëF ëC ëR ëF ëC ëR 

93.2 34 27.2 57.2 14 11.2 

91.4 33 26.4 55.4 13 10.4 

89.6 32 25.6 53.6 12 9.6 

87.8 31 24.8 51.8 11 8.8 

86 30 24 50 10 8 

84.2 29 23.2 48.2 9 7.2 

82.4 28 22.4 46.4 8 6.4 

80.6 27 21.6 44.6 7 5.6 

78.8 26 20.8 42.8 6 4.8 

77 25 20 41 5 4 

75.2 24 19.2 39.2 4 3.2 

73.4 23 18.4 37.4 3 2.4 

71.6 22 17.6 35.6 2 1.6 

69.8 21 16.8 33.8 1 0.8 

68 20 16 32 0 0 

66.2 19 15.2 30.2 1 0.8 

64.4 18 14.4 28.4 2 1.6 

62.6 17 13.6 23 5 4 

60.8 16 12.8 14 10 8 

59 15 12 4 20 16 

* In this table ëR is the ReÂaumur (not the Rankine) scale. 



m ˆ 1.0936 yard ˆ 3.280 ft.; cm ˆ 0.39370 in.; 

hectare ˆ 2.471 acre; m 2 ˆ 10.764 ft. 2 ; cm 2 ˆ 0.1550 in. 2 ; 

m 3 ˆ 1000 dm 3 (or litre) ˆ 33.315 ft. 3 ˆ 61024 in. 3 ˆ 1.30795 yd. 3 ; 

hl ˆ 100 dm 3 (or litre) ˆ 21.998 gal. (British) ˆ 26.418 (US) ˆ 0.6111 barrel (British) ˆ 0.8387 

barrel (US) ˆ 0.8522 beer barrel (US) ˆ 0.1 m 3 

l ˆ 35.196 fl. oz. (British) ˆ 33.815 fl. oz. (US) ˆ 0.21998 gal. (British) ˆ 0.26418 gal. (US) ˆ 

0.035315 ft. 3 

tonne ˆ 1000 kg ˆ short ton ˆ 0.984207 long ton ˆ 2204.62 lb. ˆ 10 doppelzentner ˆ 20 zentner; 

zentner ˆ 50 kg ˆ 0.984 cwt ˆ 110,321 lb.; 1 kg ˆ 2.20462 lb. 

g ˆ 0.03527 oz. ˆ 15.432 grain; kg/m 3 ˆ g/dm 3 ˆ 0.062428 lb./ft. 3 

British measures 

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 

in. ˆ 2.540 cm ˆ 1000 thou (thousandth of an inch); lb./ft. 3 ˆ 16.0185 kg/m 3 

thou ˆ 25.4 micron or micrometer ( m); 

acre ˆ 4840 yd. 2 ˆ 0.4047 hectare (ha); 

yd. 2 ˆ 0.8361 m 2 ; ft. 2 ˆ 9.290 dm 2 ; in. 2 ˆ 6.4516 cm 2 ; 

yd. 3 ˆ 0.7646 m 3 ; ft. 3 ˆ 28.317 dm 3 ; in. 3 ˆ 16.3871 cm 3 ; 

ton (long) ˆ 20 cwt ˆ 2240 lb. ˆ 1016 kg ˆ 1.01605 tonne (short ton; metric); 

lb. ˆ 16 oz. ˆ 256 dram ˆ 7000 grains ˆ 0.45359 kg; 

oz. ˆ 28.35 g; grain ˆ 64.80 mg; 

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); 

pint ˆ 0.5682 litre; fl. oz. ˆ 28.412 ml; 

butt ˆ 2 hogshead ˆ 3 barrel ˆ 108 gal. ˆ 4.9096 hl; 

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); 

bushel (bu, British) ˆ 8 gal. (UK) ˆ 36.3687 dm 3 ˆ 0.9690 bu (US); 8 bu ˆ 1 Qr. 

100% proof spirit (British) ˆ 57.10% ethyl alcohol (v/v) ˆ 49.28% ethyl alcohol (m/m), relative 

density 0.91976 at 60 ëF. 

US measures 

beer brl ˆ 31 gal. (US) ˆ 25.81 gal. (British) ˆ 1.734 hl ˆ 0.717 brl British; 

standard brl ˆ 31.5 gal. (US) ˆ 26.23 gal. (British) ˆ 1.1924 hl ˆ 0.729 brl (British); 

gal ˆ 8 pint ˆ 128 fl. oz. ˆ 3.7853 litre ˆ 0.8327 gal. (British) ˆ 231.0 in. 3 

bushel (bu, US) ˆ 35.23907 litre ˆ 1.03203 bu (British). 

Barley and malt measures 

Britain and South Africa: barley bushel ˆ 56 lb. ˆ 75.401 kg; 

barley quarter ˆ 448 lb. ˆ 4 cwt ˆ 0.2 ton ˆ 203.209 kg; 

malt bushel ˆ 42 lb. ˆ 19.051 kg; 

malt quarter ˆ 336 lb. ˆ 3 cwt ˆ 0.15 ton ˆ 152.407 kg 

Australia and New Zealand: barley bushel ˆ 50 lb.; malt bushel ˆ 40 lb. 

US and Canada: barley bushel ˆ 48 lb.; malt bushel ˆ 34 lb. 

Useful data (some equivalents are approximate) 

kcal ˆ 4.186 kJ ˆ 3.968 btu (BTU) ˆ 1.1628 Wh ˆ 3088 ft. lb. 

btu ˆ 1.055 kJ ˆ 0.252 kcal ˆ 0.2931 Wh ˆ 778.2 ft. lb. 

Wh ˆ 3.6 kJ ˆ 0.860 kcal ˆ 3.412 btu ˆ 2655 ft. lb. 

therm ˆ 105.506 MJ ˆ 29.307 kWh; 

standard ton refrigeration per 24 h ˆ 12000 btu/h ˆ 3024 kcal/h 

atm ˆ 14.6959 lb./in. 2 ˆ 760 mm Hg (at 0 ëC and 45ë latitude) ˆ 101.325 kPa ˆ 1.01325 bar (ˆ 

33.899 ft. water ˆ 9.935 m of 1040 wort) 

lb./in. 2 ˆ 6896.76 N/m 2 ˆ 0.06895 bar ˆ 703 kg/m 3 ˆ 27.7 inches water 

lb./gal. (British) ˆ 99.76 g/l; lb./gal. (US) ˆ 119.8 g/l 

lb./brl (British) ˆ 3.336 g/l; lb./brl (US) ˆ 3.865 g/l 

grain/gal (British) ˆ 14.25 mg/l; grain/gal (US) ˆ 17.12 mg/l 

CO 2 in beer: g/100 ml ˆ 5.06 v/v in beer; v/v beer ˆ 0.198 g/100ml 



The following table is based on those compiled by Dr Plato for the German Imperial Commission (Normal-Eichungskommission) and refers to the apparent 

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 and % w/w 

represent g per 100 ml and g per 100 g of solution, respectively. The percentages by weight in column 6, corresponding with the specific gravities at 60 ëF given in 

column 1, were computed by interpolation from Plato's table for true specific gravities at 15ë/15 ëC and 16ë/15 ëC corrected to 60ë/60 ëF and then brought to 60 ëF/ 

60 ëF in air by adding (SG ± 1) 0.00121. The cane sugar weight percentages were converted to volume percentages and the solution divisors calculated. The 

column headed Plato gives the specific gravities in air at 20 ë/20 ëC related to the cane sugar weight percentages and, with the latter, corresponds with the Plato 

table commonly used in breweries and laboratories where 20 ëC is the standard temperature. The column headed Balling similarly gives the specific gravities at 

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 

and 20ë/20 ëC on account of the errors in Balling's table. The following densities were used in the calculations: 

Water at 15 ëC/4 ëC 0.999126 

60 ëF/4 ëC 0.999035 

20 ëC/4 ëC 0.998234 

Specific gravity conversion table for cane sugar solutions 

British units Plato 

SG Brewers' Cane sugar Solution SG Cane sugar Balling BaumeÂ 

60 ëF pounds (% w/v) divisor 20 ëC (% w/w, Brix) SG 17.5 ëC Modulus 145 

1002.5 0.9 0.643 3.888 1.00250 0.641 1.00256 0.36 

1005.0 1.8 1.287 3.885 1.00499 1.281 1.00513 0.72 

1007.5 2.7 1.932 3.882 1.00748 1.918 1.00767 1.08 

1010.0 3.6 2.578 3.879 1.00998 2.552 1.01021 1.43 

1012.5 4.5 3.225 3.876 1.01247 3.185 1.01274 1.78 

1015.0 5.4 3.871 3.875 1.01496 3.814 1.01528 2.14 

1017.5 6.3 4.517 3.874 1.01745 4.439 1.01776 2.48 

1020.0 7.2 5.164 3.873 1.01993 5.063 1.02025 2.83 

1022.5 8.1 5.810 3.872 1.02242 5.682 1.02273 3.17 

1025.0 9.0 6.458 3.871 1.02490 6.300 1.02523 3.52 

1027.5 9.9 7.107 3.869 1.02740 6.917 1.02776 3.86 

1030.0 10.8 7.755 3.868 1.02989 7.529 1.03027 4.20 

1032.5 11.7 8.405 3.867 1.03238 8.140 1.03277 4.54 

1035.0 12.6 9.054 3.866 1.03486 8.748 1.03527 4.88 

1037.5 13.5 9.703 3.865 1.03736 9.352 1.03775 5.22 


1040.0 14.4 10.354 3.863 1.03985 9.956 1.04024 5.55 

1042.5 15.3 11.003 3.862 1.04234 10.554 1.04273 5.88 

1045.0 16.2 11.652 3.862 1.04481 11.150 1.04523 6.21 

1047.5 17.1 12.303 3.861 1.04731 11.745 1.04773 6.54 

1050.0 18.0 12.953 3.860 1.04979 12.336 1.05022 6.87 

1052.5 18.9 13.604 3.859 1.05227 12.925 1.05269 7.20 

1055.0 19.8 14.255 3.858 1.05476 13.512 1.05515 7.52 

1057.5 20.7 14.907 3.857 1.05726 14.097 1.05760 7.84 

1060.0 21.6 15.560 3.856 1.05975 14.679 1.06005 8.16 

1062.5 22.5 16.213 3.855 1.06224 15.259 1.06252 8.48 

1065.0 23.4 16.866 3.854 1.06472 15.837 1.06500 8.80 

1067.5 24.3 17.519 3.853 1.06720 16.411 1.06747 9.12 

1070.0 25.2 18.173 3.852 1.06970 16.984 1.06995 9.44 

1072.5 26.1 18.827 3.851 1.07218 17.554 1.07244 9.75 

1075.0 27.0 19.482 3.850 1.07467 18.122 1.07494 10.06 

1077.5 27.9 20.135 3.849 1.07717 18.687 1.07743 10.37 

1080.0 28.8 20.791 3.848 1.07965 19.251 1.07990 10.69 

1082.5 29.7 21.446 3.847 1.08213 19.812 1.08237 11.00 

1085.0 30.6 22.101 3.846 1.08462 20.370 1.08486 11.30 

1087.5 31.5 22.758 3.845 1.08712 20.927 1.08737 11.61 

1090.0 32.4 23.414 3.844 1.08960 21.481 1.08986 11.91 

1092.5 33.3 24.071 3.843 1.09209 22.033 1.09235 12.21 

1095.0 34.2 24.726 3.842 1.09457 22.581 1.09481 12.51 

1097.5 35.1 25.384 3.841 1.09707 23.129 1.09730 12.81 

1100.0 36.0 26.041 3.840 1.09956 23.674 1.09980 13.11 

1102.5 36.9 26.700 3.839 1.10204 24.218 1.10230 13.41 

1105.0 37.8 27.360 3.838 1.10454 24.760 1.10480 13.71 

1107.5 38.7 28.019 3.837 1.10703 25.299 1.10730 14.00 

1110.0 39.6 28.679 3.836 1.10952 25.837 1.10983 14.30 

1112.5 40.5 29.339 3.834 1.11200 26.372 1.11235 14.59 

1115.0 41.4 30.000 3.833 1.11450 26.906 1.11486 14.88 

1117.5 42.3 30.660 3.832 1.11698 27.436 1.11735 15.17 

1120.0 43.2 31.321 3.831 1.11947 27.965 1.11984 15.46 

1122.5 44.1 31.981 3.830 1.12195 28.491 1.12231 15.74 

1125.0 45.0 32.643 3.829 1.12445 29.016 1.12478 16.03 


Table A5 Continued 

British units Plato 

SG Brewers' Cane sugar Solution SG Cane sugar Balling BaumeÂ 

60 ëF pounds (% w/v) divisor 20 ëC (% w/w, Brix) SG 17.5 ëC Modulus 145 

1127.5 45.9 33.305 3.828 1.12694 29.539 1.12729 16.31 

1130.0 46.8 33.970 3.827 1.12944 30.062 1.12980 16.60 

1132.5 47.7 34.632 3.826 1.13191 30.580 1.13228 16.88 

1135.0 48.6 35.295 3.825 1.13441 31.097 1.13477 17.16 

1137.5 49.5 35.958 3.824 1.13689 31.611 1.13723 17.44 

1140.0 50.4 36.621 3.823 1.13938 32.124 1.13971 17.72 

1142.5 51.3 37.285 3.822 1.14186 32.635 1.14221 18.00 

1145.0 52.2 37.951 3.821 1.14435 33.145 1.14473 18.27 

1147.5 53.1 38.617 3.820 1.14685 33.653 1.14727 18.55 

1150.0 54.0 39.284 3.818 1.14934 34.160 1.14980 18.82 


Table A6 Equivalence (approximate) between Institute of Brewing (UK) units of hot water extract 

Litre ë/kg (20 ëC) against lb/Qr (15.5 ëC, 60ëF) in the body of the table. Calculated from the Institute of Brewing Recommended Methods of Analysis (1993), on the 

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 

(15.5 ëC); 251 lë/kg (15.5 ëC) is equivalent to 83.6 lb/Qr (15.5 ëC). 

240 250 260 270 280 290 300 310 320 330 

0 80.2 83.7 87.2 90.8 94.2 97.8 101.3 104.8 108.3 111.9 

1 80.5 84.1 87.5 91.1 94.6 98.1 101.7 105.2 108.6 112.2 

2 80.8 84.4 87.9 91.4 94.9 98.5 102.0 105.5 109.0 112.6 

3 81.2 84.8 88.3 91.8 95.3 98.8 102.4 105.9 109.3 112.9 

4 81.5 85.1 88.6 92.1 95.7 99.2 102.7 106.2 109.7 113.2 

5 81.9 85.5 89.0 92.5 96.0 99.5 103.1 106.6 110.1 113.6 

6 82.3 85.8 89.4 92.8 96.4 99.9 103.4 106.9 110.4 113.9 

7 82.7 86.1 89.7 93.2 96.8 100.2 103.8 107.3 110.8 114.3 

8 83.0 86.5 90.1 93.5 97.1 100.6 104.1 107.6 111.2 114.6 

9 83.4 86.8 90.4 93.9 97.5 100.9 104.5 107.9 111.5 115.0 


Table A7 Solution divisors of some sugars at 20 ëC (68 ëF) (Pauls' Malt Brewing Room Book 

1995±7, p. 231). The apparent solids content of a solution (g/100 cm 3 ) = G/D where G is the excess 

specific gravity and D is the solution divisor 

SG (approx) D (sucrose) D (invert sugar) D (fructose) D (glucose) Wort solids 

1016.0 3.872 3.875 3.907 3.805 ± 

1020.0 3.869 3.872 3.904 3.805 ± 

1028.0 3.864 3.866 3.900 3.805 3.92 

1036.0 3.860 3.861 3.896 3.803 ± 

1061.0 3.849 3.845 3.882 3.794 3.90 

1083.0 3.840 3.832 3.869 3.784 ± 

1106.0 3.831 3.819 3.853 3.772 ± 

Table A8 Some properties of water at various temperatures (various sources) 

Temperature log Kw (pKw) Oxygen content of air-saturated water 

ëC ëF (mg/l) (mg/l) (cm 3 at NTP/kg) 

0 32 14.9435 ± 14.35 10.19 

5 41 ± ± 12.73 8.9 

10 50 14.5346 ± 11.25 7.9 

15 59 ± 9.8 10.06 7.0 

20 68 14.1669 8.8 9.08 6.4 

24 75.2 14.0000 ± ± ± 

25 77 13.9965 8.1 8.25 5.8 

30 86 13.8330 7.5 ± 5.3 

35 95 ± 7.0 ± ± 

50 122 13.2617 ± ± ± 

60 140 13.0171 ± ± ± 

Table A9 The density and viscosity ( ) at various temperatures of pure air-free water* 

Temperature 

ëC ëF (cP) Density (g/ml) 

10 14 ± 0.99812 

Ice, 0 32 ± 0.91700 

0 32 1.787 0.99987 

3.98 39.16 ± 1.00000 

5 41.00 ± 0.99999 

10 50.00 ± 0.99973 

15 59.00 ± 0.99913 

20 68.00 1.002 0.99823 

30 86.00 ± 0.99567 

50 122.00 0.5468 0.98807 

60 140.00 ± 0.98324 

65 149.00 ± 0.98059 

75 167.00 0.3781 ± 

80 176.00 ± 0.97183 

100 212.00 0.2818 0.95838 

*Data of Weast (1977), Moll (1979) 


Table A10 Some properties of water (Bak et al., 2001; Moll, 1979) 

Temperature Specific heat Latent heat Latent heat Surface tension 

(cal/g/K) of vaporization of fusion (dynes/cm) 

(ëC) (ëF) (cal/g) (cal/g) 

6.62 20.1 ± ± 76 ± 

0 32 1.00738 594.8 79.7 (approx) 75.6 

10 50 1.00129 ± ± ± 

20 68 0.99883 ± ± 72.75 

100 212 ± 539 ± 58.90 

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 

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 

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 

20 ëC, is 2595 MJ. 

Table A11 The relationship between the absolute pressure and the 

temperature of water-saturated steam 

Pressure 

(bar) 

0.5780 

0.7011 

0.8453 

1.000 

1.01325 

1.1 

1.2 

1.3 

1.5 

1.8 

2.0 

2.5 

3.0 

3.5 

4.0 

4.5 

5.0 

5.5 

6.0 

8.0 

10.0 

15.0 

20.0 

30.0 

40.0 

50.0 

Temperature 

(ëC) (ëF) 

85 

90 

95 

99.6 

100 

102.3 

104.8 

107.1 

111.4 

116.9 

120.2 

127.4 

133.5 

138.9 

143.6 

147.9 

151.8 

155.5 

158.8 

170.4 

179.9 

198.3 

212.4 

233.8 

250.3 

263.9 


185 

194 

203 

211.3 

212 

216.1 

220.6 

224.8 

232.5 

242.4 

248.4 

261.3 

272.3 

282.0 

290.5 

298.2 

305.2 

311.9 

317.8 

338.7 

355.8 

388.9 

414.3 

452.8 

482.5 

507.0

Table A12 The solubility of pure gases in water under 1 atmosphere pressure and at the 

temperatures shown. The volumeof gasdissolved is given in ml,reduced to STP, that is at 0ëC and 

1atmosphere pressure (760 mm of mercury at 0ëC, at latitude 45ë). (After Dawson et al., 1987) 

Temperature Gas solubility (ml at STP/litre water) 

(ëC) (ëF) O2 N2 CO2 

0 

5 

10 

15 

20 

25 

30 

35 

40 

45 

50 

32 

41 

50 

59 

68 

77 

86 

95 

104 

113 

122 

49 

43 

38 

34 

31 

28 

26 

24 

23 

22 

21 

24 

21 

19 

17 

15 

14 

13 

13 

12 

11 

11 

1713 

1424 

1194 

1019 

878 

759 

665 

592 

530 

479 

436 

The gram molecular volumeof an ideal gas at STP is 22.414 litres. The molecular weights of the gases are O2 = 

32.0; N2 =28.02; CO2 =44.01. Consequently 1ml each (non-ideal) gas, at STP, will contain approximately 

oxygen, 1.428 mg; nitrogen, 1.250 mg and carbon dioxide, 1.964 mg (compare with Table 10.5). 

Table A13 Salts in brewing liquors. Salt concentrations in brewing liquor are expressed in several 

different ways. This may be as the unit weight/unit volume, as in molarity (M, i.e. the gram 

molecular weight per litre) or as the equivalents per litre. Sometimes no account is taken of the fact 

that in dilute solution salts are nearly completely ionized, that is their component ions (cations, 

positively charged; anions, negatively charged) are separated in solution. Consequently it makes 

more sense to give the concentrations of the ions. At present these are usually expressed as mg (or 

g)/litre. In the past concentrations were often expressed as millvals, milligram equivalents/litre. 

When this is done the total concentration of the anions should equal the total concentration of the 

cations. The characteristics of some commonly encountered ions are indicated below. By ppm 

(parts/million) mg/litre are usually understood, but this is ambiguous and mg/kg may be intended 

Ion Symbol Atomic/molecular 

weight 

Calcium 

Magnesium 

Sodium 

Potassium 

Bicarbonate 

Carbonate 

Sulphate 

Chloride 

Nitrate 

Nitrite 

Ca 2+ 

Mg 2+ 

Na + 

K + 

HCO3 

CO 3 2 

SO4 2 

Cl 

NO3 

NO 2 

40.08 

24.32 

23.00 

39.10 

61.01 

60.01 

96.07 

35.48 

62.01 

46.01 


Equivalent weight 

20.04 

12.16 

23.00 

39.10 

61.01 

30.01 

48.04 

35.48 

62.01 

46.01

Table A14 Units of degrees of water hardness. Equivalences between various units (degrees) of 

hardness in water. Total hardness usually means the sum of the calcium and magnesium expressed 

as equivalent amounts of CaO or CaCO3. This convention is historically interesting, but irrational, 

since calcium carbonate has limited solubility and calcium oxide (`quicklime') reacts violently with 

water to give calcium hydroxide (After Moll, 1979; Benson et al., 1997) 

French English German Ca Ca 

degree degree degree (mg/litre) (mM/litre) 

1 French degree a 

1.00 0.70 0.56 4.004 0.100 

1 English degree b 

1 German degree 

1.43 1.00 0.80 5.73 0.143 

c 

1 USA degree 

1.78 1.25 1.00 7.17 0.179 

d 

0.10 0.07 0.056 0.40 0.01 

1 Calcium (mg/l) 0.25 0.175 0.140 1.00 0.025 

Ca (millimoles/l) 0.00625 0.00438 0.0035 0.025 0.00063 

a CaCO3, 10 mg/l; b 1 Clark degree, CaCO3, 14.3 mg/l (1 grain/imp. gallon); c CaO, 10 mg/l; 

d CaCO3, 1 mg/l. 

Table A15 Some characteristics of raw materials and offal relevant to storage. The angle of 

repose is the angle from the horizontal adopted by materials when poured onto a flat surface. These 

values are somewhat variable depending on, for example, how well grain has been dressed, its 

moisture content and the variety. The valley angle is the included angle in the cone at the base of a 

silo or a store, which must not be exceeded if material is to run out freely (Briggs, 1998, Sugden et 

al., 1999). 

Material Bulk density 

(kg/m 3 )* 

Barley 

Malt 

Malt grist, 

roller-milled 

Malt grist, 

hammer milled 

Rice (polished) 

Rice (broken) 

Rye 

Sorghum 

Oats 

Wheat 

Maize grits 

Pelletized malt 

culms 

Malt culms 

Cereal dust 

610±618 

520±750 

510±550 

450±550 

300±370 

680±730 

780±850 

840±890 

680±740 

720±770 

350±520 

780±830 

630±700 

560±640 

224 

± 

Angle of repose 

(degrees) 

32 

23±35 

30±45 

26 

30-45 

± 

30 

30 

26 

30±45 

28±32 

30 

30±45 

± 

50 

60 

Maximum valley angle 

(degrees) 

*The reciprocals of these values indicate the amount of storage space needed for each kilogram of material (that 

is m 3 /kg) 


116 

± 

90 

± 

90 

± 

120 

120 

128 

90 

116 

120 

90 

± 

± 

±

Table A16 Beer pasteurization; product temperature and lethal rate in Pasteurization Units, PUs. 

One PU is the lethal effect of holding for one minute at 60 ëC (140 ëF). Data selected from Paul's 

Brewing Room Book, 1998±2000, p. 271, where some practical considerations are considered 

Product temperature PU/min. 

(ëC) (ëF) 

53.0 

55.0 

56.5 

57.5 

59.0 

60.0 

61.5 

63.0 

64.0 

65.0 

66.5 

67.5 

69.0 

70.0 

71.5 

72.5 

74.0 

75.0 

76.5 

77.5 

79.0 

80.0 

127.4 

131.0 

133.7 

135.5 

138.2 

140.0 

142.7 

145.4 

147.2 

149.0 

151.7 

153.5 

156,2 

158.0 

160.7 

162.5 

165.2 

167.0 

169.7 

171.5 

174.2 

176.0 


0.10 

0.19 

0.32 

0.45 

0.72 

1.0 

1.7 

2.7 

3.7 

5.2 

8.6 

12 

19 

27 

45 

62 

100 

139 

231 

320 

519 

720

Ethanol (%) 

13 

12 

11 

10 

9 

8 

7 

6 

5 

4 

3 

2 

1 

Composition by 

weight (in vacuo) 

Composition by volume 

(in air at 20°C) 

Pure ethanol has a density, 

in air at 20°C, of 788.16 kg/m 3 . 

Under these conditions water has 

a density of 997.15 kg/m 3 

0 980 1 2 3 4 5 6 7 8 9 990 1 2 3 4 5 6 7 998 

Density (kg/m 3 ) in air at 20°C 

Fig. A1 The relationships between the densities (kg/m 3 ) of alcohol (ethanol)/water mixtures, by 

volume in air at 20 ëC, or by mass in vacuo. Data from the Brewing Room Book, 1998±2000, p. 281, 

Pauls Malt, Ipswich and Kentford, Suffolk, UK. 

References 

BAK, S. N., EKENGREN, OÈ ., EKSTAM, K., HAÈ RNULV, G., PAJUNEN, E., PRUCHA, P. and RASI, J. (2001) Water in 

Brewing: a European Brewery Convention Manual of Good Practice. 128 pp. NuÈrnberg. 


BENSON, J. T., COLEMAN, A. R., DUE, J. E. B., HENHAM, A. W., TWAALFLLOVEN, J. G. P. and VINCKX, W. (1997) 

Brewery Utilities: a European Brewery Convention Manual of Good Practice. 200 pp. NuÈrnberg. 


BRIGGS, D. E. (1998) Malts and Malting. p. 750. London: Blackie Academic & Professional. Gaithersburg: 

Aspen Publishing. 

DAWSON, R. M. C., ELLIOTT, D. C., ELLIOTT, W. H. and JONES, K. M. (1987) Data for Biochemical Research. 

(3rd edn). 580 pp. Oxford. The Clarendon Press. 

MOLL, M. (1979) in Brewing Science, 1, (J. R. A. Pollock, ed.) p. l. London. Academic Press. 

SUGDEN, T. D., WEBB, C., BYRNE, H., VAN WAESBERGHE, J. and WULFF, T. (1999) Milling: a European 

Brewery Convention Manual of Good Practice. 102 pp. NuÈrnberg. Fachverlag Hans Carl. 

WEAST, R. C. (ed.) (1977) CRC Handbook of Chemistry and Physics. (58th edn). Cleveland. CRC Press.

Title: Brewing science and practice

Create successful ePaper yourself

Delete template?

Save as template?