FDOT BDV25 977 03 RPT

Distribution of Chloride, pH, Resistivity, and
Sulfate Levels in Backfill for Mechanically-

Stabilized Earth Walls and Implications for
Corrosion Testing
BDV25-977-03
Mr. Ivan Sokolic, Project Manager
Construction Manager
FDOT Ft. Myers Operations Center—District One
2981 N. E. Pine Island Road
Cape Coral, Florida 33909
Dr. Manjriker Gunaratne, Principal Investigator

Chair
Department of Civil & Environmental Engineering
University of South Florida
4202 E. Fowler Avenue
Tampa, Florida 33620
Final Report
Performance Period: June 3, 2013 through May 31, 2015
May 2015
Prepared by
Noreen Poor, Manjriker Gunaratne, and Thilanki Rajaguru
University of South Florida, Tampa, Florida
Disclaimer
The opinions, findings, and conclusions expressed in this publication are those of the authors and
not necessarily those of the State of Florida Department of Transportation.
ii
Approximate Conversions to SI Units
SYMBOL WHEN YOU KNOW MULTIPLY BY TO FIND SYMBOL

LENGTH
mm millimeters 0.039 inches in
m meters 3.28 feet ft
m meters 1.09 yards yd
km kilometers 0.621 miles mi

AREA
mm 2
square millimeters 0.0016 square inches in2
m2 square meters 10.764 square feet ft2
m2 square meters 1.195 square yards yd2
ha hectares 2.47 acres ac
km 2
square kilometers 0.386 square miles mi2

VOLUME
mL milliliters 0.034 fluid ounces fl oz
L liters 0.264 gallons gal
m 3
cubic meters 35.314 cubic feet ft3
m3 cubic meters 1.307 cubic yards yd3

MASS
g grams 0.035 ounces oz
kg kilograms 2.202 pounds lb
Mg (or "t") megagrams (or "metric 1.103 short tons (2000 lb) T
ton")

TEMPERATURE (exact degrees)
o o
C Celsius 1.8C+32 Fahrenheit F
*SI is the symbol for the International System of Units. Appropriate rounding should be made to comply with
Section 4 of ASTM E380.
(Revised March 2003)
(Downloaded from http://www.fhwa.dot.gov/aaa/metricp.cfm January 2015)
iii
Technical Report Documentation Page
1. Report No. 2. Government Accession No. 3. Recipient's Catalog No.
4. Title and Subtitle 5. Report Date

Distribution of Chloride, pH, Resistivity, and Sulfate Levels in
Backfill for Mechanically-Stabilized Earth Walls and
Implications for Corrosion Testing 6. Performing Organization Code
7. Author(s) 8. Performing Organization Report No.

Noreen Poor, Manjriker Gunaratne, Thilanki Rajaguru
9. Performing Organization Name and Address 10. Work Unit No. (TRAIS)
Department of Civil & Environmental Engineering University of South Florida
4202 E. Fowler Avenue 11. Contract or Grant No.
Tampa, Florida 33620 BDV25-977-03
12. Sponsoring Agency Name and Address 13. Type of Report and Period Covered
Florida Department of Transportation Final Report
605 Suwannee Street, MS-30 06/03/2013 to 05/31/2015
Tallahassee, Florida 32399
14. Sponsoring Agency Code
15. Supplementary Notes
16. Abstract
The ultimate goals of this research were to improve quality, speed completion, and reduce risk in mechanically-stabilized
earth (MSE) wall projects. Research objectives were to assure (1) that variability in the corrosion properties of soil (pH,
minimum resistivity, chloride, and sulfate levels) due to sampling and analytical technique was much lower than variability
in these levels within and between soil sources and types and thus did not inflate the risk of emplacing a corrosive soil as
MSE wall backfill, (2) that the number of soil type samples analyzed prior to acceptance of a backfill was appropriate, based
on the expected distribution of corrosion properties within the backfill, and (3) that the corrosion properties of backfill
material did not change appreciably over time, especially after emplacement and over the design lifetime of an MSE wall.
Corrosion properties of soil were tested with Florida Methods (FMs) 5-550, 5-551, 5-552, and 5-553 for pH, minimum
resistivity, water-soluble chloride, and water-soluble sulfate, respectively. Changes to the quality assurance (QA) plan for
acceptance quality characteristics (AQC) of backfill corrosion properties were recommended based on research outcomes.
These changes had as their focus an improvement in the buyer’s statistical power to accept good backfill material through
reductions in test error associated with material properties, sample processing, and laboratory procedures. Recommended
changes included (1) proposed revisions to the FMs for pH, minimum resistivity, chloride, and sulfate, (2) an increase in the
number of independent samples tested for pH and minimum resistivity, (3) a revised acceptance limit for minimum
resistivity, (4) method operator training and independent audits, (5) an inter-laboratory study post-implementation of the
revised methods, and (6) re-evaluation of the need for chloride and sulfate testing for backfill above a pre-determined
minimum resistivity. Geochemical modeling with the U. S. Geological Survey (USGS) model PHREEQC for a
conservative case of a low ionic strength and poorly buffered sandy backfill revealed that in a few years’ time the pore water
of emplaced backfill could equilibrate with infiltrating rainfall. The model results suggested that a soil’s buffering capacity
might be important consideration when metal is used as reinforcement in MSE wall backfill. Field and laboratory testing of
candidate sands to calibrate the model were recommended ahead of any proposed changes to the QA plan based on model
results.
17. Key Word 18. Distribution Statement
Mechanically-stabilized earth (MSE) wall, corrosion, pH, No restrictions.
resistivity, chloride, sulfate, select backfill, ruggedness
study, inter-laboratory study
19. Security Classif. (of this report) 20. Security Classif. (of this page) 21. No. of Pages 22. Price
Unclassified Unclassified 378
Form DOT F 1700.7 (8-72) Reproduction of completed page authorized
iv
Acknowledgments
The authors of this report would like to express their gratitude for access to mines from which we
collected soil samples, and we would like to thank mine owners and operators that gave us that
access: Angelo’s Recycled Materials, Dade City; C. C. Calhoun, Haines City; Cemex Alico Quarry,
Ft. Myers; Hickey Excavation, Sebring; Jahna Industries, Haines City; Titan Florida Center Sand,
Clermont; Transcor Dirt Services, Wimauma; and Youngquist Brothers Rock Mine, Ft. Myers. We
would also like to thank the commercial laboratories that opened their laboratories for a visit or
participated with a split-sample analysis. These laboratories included Environmental &
Geotechnical Specialists (EGS), Tallahassee; Ellis & Associates, Jacksonville; Universal
Engineering Services (UES), Jacksonville; Universal Engineering Services (UES), Orlando;
Professional Service Industries (PSI), Orlando; Madrid Engineering, Bartow; Professional Service
Industries (PSI), Tampa; Tally Engineering, Tampa; Tierra, Tampa; Universal Engineering
Services (UES), Tampa; Ardaman & Associates, Ft. Myers; and Professional Services Industries
(PSI), Ft. Lauderdale. We would like to extend our sincere appreciation for the help received from
the FDOT facilities that allowed two laboratory visits and participated in the inter-laboratory study:
Districts 1 and 7 Materials Office Laboratory, Bartow; District 2 Materials Office Laboratory,
Lake City; District 3 Materials Office Laboratory, Chipley; Districts 4 and 6 Materials Office
Laboratory, Davie; District 5 Materials Office, Deland; and the State Materials Office (SMO)
Corrosion Laboratory and Environmental Laboratory, Gainesville. The authors would like to thank
Ingrid Eversley, Barbara Johnson, Melissa Lopez, Carol Marrero-Placeres, Julie McCoy and
Shivali Vyas, USF College of Engineering, for their assistance with purchasing and travel. Finally,
we would like to thank project team members—formal and informal—for their enthusiasm and
support of this project. These team members included Project Manager Ivan Sokolic, Sam Joseph,
Timothy Meeks, Xiaoyan Zheng, Teresa Puckett, and Angela Koloc, FDOT Districts 1 and 7
Materials Office; Mario Paredes (retired from FDOT), David Horhota, Ronald Simmons, and
Nikita Reed, State Materials Office; Patti Brannon and Latashi Kitchen, FDOT Research Office;
and Corinne Walters and Stephanie Rios, USF Sponsored Research. This research would not have
been possible without the assistance provided by courteous and professional staff of mine owners
and operators, commercial laboratories, USF departments, and FDOT offices.
v
Executive Summary
The ultimate goals of this research were to improve quality, speed completion, and reduce risk in
mechanically-stabilized earth (MSE) wall projects. Research objectives were (1) to assure that
variability in the corrosion properties of soil (pH, minimum resistivity, chloride, and sulfate levels)
due to sampling and analytical technique was much lower than variability in these levels within and
between soil sources and types and thus did not inflate the risk of emplacing a corrosive soil as
MSE wall backfill, (2) that the number of soil type samples analyzed prior to acceptance of a
backfill was appropriate, based on the expected distribution of corrosion properties within the
backfill, and (3) that the corrosion properties of backfill material did not change appreciably over
time, especially after emplacement and over the design lifetime of an MSE wall. Corrosion
properties of soil were tested with Florida Methods (FMs) 5-550, 5-551, 5-552, and 5-553 for pH,
minimum resistivity, water-soluble chloride, and water-soluble sulfate, respectively.
Divergence in corrosion test results between laboratories was seen for pH, minimum resistivity,
chloride, and sulfate for one or more of the eight sands collected from mines as part of this study.
Changes to the FMs for minimum resistivity, chloride, and sulfate that were tested by USF showed
promise for improving between-laboratory agreement. Split sample analyses revealed that four of
the eight sands failed to meet an acceptance limit for one or more of the corrosion properties in one
or more of the partnered laboratories. If the split sample analyses between two laboratories
represented quality control and verification tests, at least three out of the eight sands would have
required resolution testing. Reduction in a relatively high rate of resolution testing may offset
additional costs associated with method improvements.
Replicate studies of mined soils for each corrosion FM produced an estimate of test error that
included contributions from material properties, environmental influences, and sampling,
processing, and single-operator testing (method repeatability) of soils. For each FM, its test error
estimate was adjusted to include the contribution from multi-laboratory testing (method
reproducibility). The adjusted estimate was incorporated into a method operating characteristic
(OC) curve, which had a Type I () error of 0.01 and a Type II () error of 0.05; the OC curve
related the buyer’s statistical power (1-) to accept a good backfill to the number of samples, test
error, and acceptance limit(s).
Changes to the quality assurance (QA) plan for acceptance quality characteristics (AQC) of backfill
corrosion properties were recommended based on research outcomes. These changes had as their
focus an improvement in the buyer’s statistical power to accept good backfill material through
reductions in test error associated with material properties, sample processing, and laboratory
procedures. Recommended changes included (1) proposed revisions to the FMs for pH, minimum
resistivity, chloride, and sulfate, (2) an increase in the number of independent samples tested for pH
and minimum resistivity, (3) a revised acceptance limit for minimum resistivity, (4) method
operator training and independent audits, (5) an inter-laboratory study post-implementation of the
vi
revised methods, and (6) re-evaluation of the need for chloride and sulfate testing for backfill above
a pre-determined minimum resistivity. Recommended changes are outlined below:
 Revise the FMs for pH, minimum resistivity, chloride, and sulfate. Proposed revisions to the
FMs included (1) minor edits to method grammar, organization, and units of measure, (2)
updates to the apparatus and reagents sections, with a hazardous materials note as
appropriate, (3) step-by-step instructions to facilitate method oversight, (4) one or more
QA/QC procedures, and (5) a new section for method precision and bias.
 Increase the number of independent samples per soil type for pH and minimum resistivity.
Recommended for pH were at least three independent pH samples per soil type tested both
pre-construction and post-emplacement. Recommended for minimum resistivity were at
least two minimum resistivity samples per soil type tested both pre-construction and post-
emplacement and an increase in the acceptance limit to 4,000 ohm-cm.
 Conduct operator training and routine on-site laboratory audits of the FMs. Operator
training was recommended for all FMs, for example, training through a video demonstration
of each method as well as method outcomes for both well-executed and poorly-executed
procedures. Also recommended for FMs were routine independent audits, which provide
oversight on method performance and put the method and operator in a context that includes
environmental influences (temperature, noise, dust, crowding), laboratory practices, and
pressure for time.
 Perform a Florida-wide inter-laboratory study of the revised FMs. A Florida-wide inter-

laboratory study of the revised FMs within a year of adoption was recommended to assure
that FDOT and commercial laboratories were proficient with the revised methods and to
establish for each FM the method reproducibility.
 Re-evaluate the FMs for chloride and sulfate. A review of Florida-wide post-revision
chloride and sulfate results for MSE wall backfill after a two-year data collection period
was recommended to assess if further testing of chloride and sulfate were necessary for
select backfill above a pre-determined minimum resistivity.
Geochemical modeling with the U. S. Geological Survey (USGS) model PHREEQC for a
conservative case of a low ionic strength and poorly buffered sandy (quartz) backfill revealed that
in a few years’ time the pore water of emplaced backfill could equilibrate with infiltrating rainfall.
The model results suggested that a soil’s buffering capacity might be an important consideration
when metal is used as reinforcement in MSE wall backfill. Field and laboratory testing of
candidate sands to calibrate the model were recommended ahead of any proposed changes to the
QA plan based on model results.
vii
Table of Contents
Disclaimer ........................................................................................................................................... ii
Approximate Conversions to SI Units ............................................................................................... iii
Technical Report Documentation Page ............................................................................................. iv
Acknowledgments............................................................................................................................... v
Executive Summary ........................................................................................................................... vi
List of Figures ................................................................................................................................... xv
List of Tables .................................................................................................................................. xvii
List of Abbreviations ...................................................................................................................... xxii
1 Introduction ................................................................................................................................. 1
1.1 Research Statement ........................................................................................................................... 1
1.2 Scope and Objectives ........................................................................................................................ 1
1.2.1 Literature review ....................................................................................................................... 2
1.2.2 Trends in backfill properties ..................................................................................................... 2
1.2.3 Single-laboratory contributions to method reproducibility ....................................................... 2
1.2.4 Replicate studies........................................................................................................................ 2
1.2.5 Multi-laboratory contributions to method reproducibility ........................................................ 2
1.2.6 Geochemical modeling.............................................................................................................. 3
1.2.7 Mine Information ...................................................................................................................... 3
1.3 Organization of Report ...................................................................................................................... 4
2 Literature Review ........................................................................................................................ 5
2.1 Mechanically-Stabilized Earth (MSE) Walls .................................................................................... 5
2.1.1 Features of an MSE wall and properties of select backfill ........................................................ 5
2.1.2 Corrosion of MSE wall reinforcement ...................................................................................... 7
2.1.3 Susceptibility of concrete to damage from environmental exposures ..................................... 10
2.2 Quality Assurance (QA) Plans for Construction Materials ............................................................. 16
2.2.1 Acceptance quality characteristics (AQC) for corrosion tests ................................................ 17
2.2.2 FDOT’s QA plan for select backfill corrosion AQC .............................................................. 21
viii
2.2.3 Sampling of select backfill ...................................................................................................... 22
2.2.4 Good laboratory practices ....................................................................................................... 24
2.2.5 Other QA/QC activities ........................................................................................................... 25
3 Measurement of Soil pH, Minimum Resistivity, Chloride, and Sulfate.................................... 27
3.1 Overview ......................................................................................................................................... 27
3.2 FM for pH ....................................................................................................................................... 27
3.2.1 Theory of measurement ........................................................................................................... 27
3.2.2 Description of method ............................................................................................................. 34
3.2.3 Method comparisons ............................................................................................................... 35
3.2.4 Estimate of method precision and bias .................................................................................... 36
3.3 Minimum Resistivity ....................................................................................................................... 36
3.3.3 Methods comparisons.............................................................................................................. 39
3.4 Chloride ........................................................................................................................................... 40
3.5 Sulfate ............................................................................................................................................. 43
3.6 Summary ......................................................................................................................................... 47
4 Trends in Select Backfill ........................................................................................................... 49
4.1 Spatial Trends.................................................................................................................................. 49
4.1.1 Methods ................................................................................................................................... 49
4.1.2 Results and discussion............................................................................................................. 53
ix
4.2 Failure Trends ................................................................................................................................. 53
4.2.1 Methods ................................................................................................................................... 53
4.2.2 Results and discussion............................................................................................................. 54
4.3 Summary ......................................................................................................................................... 58
5 Single-Laboratory Precision and Bias ....................................................................................... 59
5.1 Definitions of Precision, Accuracy, and Minimum Detection Level .............................................. 59
5.1.1 Approaches to calculating accuracy and precision.................................................................. 59
5.1.2 Method limit of detection (LOD) and limit of quantitation (LOQ)......................................... 60
5.2 Single Laboratory, Aqueous Standards ........................................................................................... 61
5.2.1 2011 version of FMs for pH, minimum resistivity, chloride, and sulfate ............................... 61
5.2.2 Proposed revised FMs for pH, minimum resistivity, chloride, and sulfate ............................. 62
5.3 Single Laboratory, Proficiency Test Soil ........................................................................................ 64
5.4 Multiple Laboratories, Proficiency Test Soil .................................................................................. 66
5.5 Summary ......................................................................................................................................... 67
6 Laboratory Visits ....................................................................................................................... 68
6.1 Introduction ..................................................................................................................................... 68
6.2 Protocol ........................................................................................................................................... 68
6.3 Laboratory Environment ................................................................................................................. 68
6.4 Results ............................................................................................................................................. 69
6.4.1 FM for pH ............................................................................................................................... 69
6.4.2 FM for minimum resistivity .................................................................................................... 71
6.4.3 FM for chloride ....................................................................................................................... 75
6.4.4 FM for sulfate.......................................................................................................................... 77
6.5 Summary ......................................................................................................................................... 79
7 Soil Sampling of Select Backfill ............................................................................................... 80
7.1 Introduction ..................................................................................................................................... 80
7.2 Field Sampling of Select Backfill ................................................................................................... 80
7.3 Geotechnical Properties of Select Backfill...................................................................................... 81
7.4 Split Sample Analyses..................................................................................................................... 82
7.5 Summary ......................................................................................................................................... 83
x
8 Ruggedness Studies ................................................................................................................... 85
8.1 Overview ......................................................................................................................................... 85
8.2 Ruggedness Study for Santa Fe River Soil ..................................................................................... 85
8.2.1 Summary results ...................................................................................................................... 85
8.2.2 FM for pH ............................................................................................................................... 86
8.2.4 FM for chloride ....................................................................................................................... 91
8.2.5 FM for sulfate.......................................................................................................................... 94
8.3 Ruggedness Studies for Mined Soils............................................................................................... 96
8.3.1 Summary statistics .................................................................................................................. 96
8.3.2 FM for pH ............................................................................................................................... 98
8.3.4 FM for sulfate........................................................................................................................ 101
8.4 Summary ....................................................................................................................................... 102
9 Treatment and Other Studies ................................................................................................... 104
9.1 Treatment Studies.......................................................................................................................... 104
9.1.1 Summary statistics ................................................................................................................ 104
9.1.2 FM for pH ............................................................................................................................. 106
9.1.3 FM for minimum resistivity .................................................................................................. 108
9.1.4 FM for chloride ..................................................................................................................... 111
9.1.5 FM for sulfate........................................................................................................................ 112
9.2 Other Factor Analysis Studies ....................................................................................................... 113
9.2.1 Effect of ionic strength and pH electrode on pH measurement............................................. 113
9.2.2 Effect of incremental water volume and pre-wetting on minimum resistivity ...................... 115
9.2.3 Effects of extraction method and soil type on chloride measurements ................................. 117
9.2.4 Effect of extraction method on sulfate measurement ............................................................ 120
9.2.5 Influence of meter/electrode system on pH measurement .................................................... 122
9.2.6 Effect of ionic strength adjustment on pH measurement ...................................................... 125
9.2.7 Potassium chloride versus calcium chloride to adjust ionic strength .................................... 126
9.2.8 Effect of dilution water and slurry water content on minimum resistivity in soils ............... 128
xi
9.2.9 Choice of range selection on sulfate measurements with a test kit ....................................... 131
9.2.10 Soil conductivity ................................................................................................................... 132
9.2.11 Contribution to pH by minerals retained on a No. 10 sieve .................................................. 132
9.3 Summary ....................................................................................................................................... 133
10 Inter-Laboratory Study......................................................................................................... 135
10.1 Scope and Purpose of Study .......................................................................................................... 135
10.2 Sample Preparation ....................................................................................................................... 135
10.3 Data Management ......................................................................................................................... 136
10.4 Results and Discussion.................................................................................................................. 136
10.5 Method Precision Statements ........................................................................................................ 139
10.5.1 Precision statement FM for pH ............................................................................................. 139
10.5.2 Precision statement FM for minimum resistivity .................................................................. 139
10.5.3 Precision statement FM for chloride ..................................................................................... 140
10.5.4 Precision statement FM for sulfate........................................................................................ 140
10.6 Summary ....................................................................................................................................... 140
11 Replicate Study and Implications for Quality Assurance .................................................... 141
11.1 Introduction ................................................................................................................................... 141
11.2 Methods ......................................................................................................................................... 141
11.2.1 Field sampling ....................................................................................................................... 141
11.2.2 Sample processing ................................................................................................................. 142
11.2.3 Sample analysis ..................................................................................................................... 142
11.2.4 Data analysis ......................................................................................................................... 143
11.3 Results ........................................................................................................................................... 144
11.3.1 FM for pH ............................................................................................................................. 144
11.3.2 FM for minimum resistivity .................................................................................................. 148
11.3.3 FM for chloride ..................................................................................................................... 153
11.3.4 FM for sulfate........................................................................................................................ 155
11.4 Summary ....................................................................................................................................... 159
12 Geochemical Modeling of Ion Transport in Sandy Soil ...................................................... 161
12.1 Problem Statement ........................................................................................................................ 161
12.2 Model Description......................................................................................................................... 161
xii
12.3 Model Inputs and Parameters ........................................................................................................ 164
12.4 Model Results and Discussion ...................................................................................................... 166
12.5 Summary ....................................................................................................................................... 167
13 Revised Florida Methods for pH, Resistivity, Chloride and Sulfate ................................... 168
13.1 Introduction ................................................................................................................................... 168
13.2 FM for pH ..................................................................................................................................... 168
13.3 FM for Minimum Resistivity ........................................................................................................ 169
13.4 FM for Chloride ............................................................................................................................ 170
13.5 FM for Sulfate ............................................................................................................................... 171
13.6 Proposed Revisions to the FM ...................................................................................................... 172
14 Summary and Recommendations ........................................................................................ 174
References ....................................................................................................................................... 178
Appendix A Ruggedness Study Results ......................................................................................... 190
A.1 Mine 1 Wimauma Sand ...................................................................................................................... 190
A.2 Mine 2 Jahna Sand ............................................................................................................................. 196
A.3 Mine 3 Youngquist Sand .................................................................................................................... 200
A.4 Mine 4 Calhoun Sand ......................................................................................................................... 206
A.5 Mine 5 Angelo’s Sand ........................................................................................................................ 210
A.6 Mine 6 Sebring Sand .......................................................................................................................... 219
A.7 Mine 7 Clermont Sand ....................................................................................................................... 228
A.8 Mine 8 Alico Road Sand .................................................................................................................... 235
Appendix B Treatment Study Results............................................................................................. 242
B.1 Mine 1 Wimauma Sand ...................................................................................................................... 242
B.2 Mine 2 Jahna Sand.............................................................................................................................. 246
B.3 Mine 3 Youngquist Sand .................................................................................................................... 250
B.4 Mine 4 Calhoun Sand ......................................................................................................................... 254
B.5 Mine 5 Angelo’s Sand ........................................................................................................................ 258
B.6 Mine 6 Sebring Sand .......................................................................................................................... 262
B.7 Mine 7 Clermont Sand........................................................................................................................ 266
B.8 Mine 8 Alico Road Sand .................................................................................................................... 270
B.9 ANOVA All Sands ............................................................................................................................. 274
xiii
B.9.1 ANOVA for pH—all sands ......................................................................................................... 274
B.9.2 ANOVA for minimum resistivity—mines 5-8 ............................................................................ 275
B.9.3 ANOVA for sulfate—mines 1, 3, 5, and 6 .................................................................................. 276
Appendix C Inter-Laboratory Study Plan and Data Sheet .............................................................. 277
C.1 First Inter-Laboratory Study ............................................................................................................... 277
C.2 Second Inter-Laboratory Study .......................................................................................................... 285
Appendix D Analysis of Variance (ANOVA) for Replicate Study ................................................ 292
D.1 ANOVA for pH .................................................................................................................................. 292
D.2 ANOVA for Minimum Resistivity ..................................................................................................... 293
D.3 ANOVA for Sulfate ........................................................................................................................... 294
Appendix E PHREEQ Input and Output Files ................................................................................ 295
E.1 Input File............................................................................................................................................. 295
E.2 Output File .......................................................................................................................................... 296
Appendix F Proposed Revised Florida Methods ............................................................................ 313
F.1 Florida Method of Test for pH of Soil and Water, FM 5-550............................................................. 313
F.2 Florida Method of Test for Minimum Resistivity of Soil and Water, FM 5-551 ............................... 318
F.3 Florida Method of Test for Chloride in Soil and Water, FM 5-552 .................................................... 324
F.4 Florida Method of Test for Sulfate in Soil and Water, FM 5-553 ...................................................... 331
Appendix G Audit Checklists ......................................................................................................... 338
G.1 Checklist for Florida Method of Test for pH of Soil and Water ........................................................ 338
G.2 Checklist for Florida Method of Test for Minimum Resistivity of Soil and Water ........................... 341
G.3 Checklist for Florida Method of Test for Chloride in Soil and Water ............................................... 345
G.4 Checklist for Florida Method of Test for Sulfate in Soil and Water .................................................. 351
xiv
List of Figures
Figure 2-1. Schematic of a mechanically-stabilized earth (MSE) wall (FDOT, 2011). ..................... 6
Figure 2-2. Permanent retaining wall selection process flowchart. .................................................. 15
Figure 2-3. Process control for pH of select backfill for MSE walls with metallic
reinforcement. ................................................................................................................................... 18
Figure 2-4. Example operating characteristic (OC) for pH N = 2. .................................................. 20
Figure 3-1. Schematic of a pH measurement system. ....................................................................... 29
Figure 3-2. Calculated influence of carbon dioxide on an open-system pH measurement of a
weak sulfuric acid solution with and without KCl added. ................................................................ 33
Figure 3-3. Resistivity versus water content for sands from mines 5 through 8. ............................. 37
Figure 3-4. Diagram of soil box for measurement of soil resistivity. ............................................... 38
Figure 3-5. Hach Pocket Colorimeter II Analysis System for Sulfate pre-programmed
sulfate calibration curves by range selection and reagent lot. .......................................................... 46
Figure 4-1. FDOT’s District 1 sources tested for MSE wall backfill (circles: small = 1
record, large = 10 records) and construction sand and gravel mines that in 2002 produced
more than 45,000 metric tons (pushpin). .......................................................................................... 50
Figure 4-2. FDOT’s District 7 sources tested for MSE wall backfill (circles: small = 1
record, large = 100 records) and construction sand and gravel mines that in 2002 produced
more than 45,000 metric tons (pushpins). ......................................................................................... 51
Figure 4-3. FDOT’s District 1 and District 7 sources passed for MSE wall backfill (circles:
small = 1 record, large = 10 records). ............................................................................................... 52
Figure 4-4. Normality plots for pH and log-transformed minimum resistivity, chloride, and
sulfate. ............................................................................................................................................... 57
Figure 6-1. Modeled and observed soil minimum resistivity versus temperature. ........................... 74
Figure 6-2. Observed soil minimum resistivity versus water resistivity. ......................................... 75
Figure 9-1. Variability plot of pH by soil type, pH electrode, and CaCl2 amendment. .................. 114
Figure 9-2. Variability plot of soil resistivity (ohm-cm) by soil type, incremental water
volume (ml), and pre-wetted versus dry soil. ................................................................................. 116
Figure 9-3. Variability plot of chloride concentration (ppm) by soil type and extraction
method............................................................................................................................................. 119
Figure 9-4. Variability plot of sulfate concentration (ppm) by soil type and extraction
method............................................................................................................................................. 121
Figure 9-5. Comparison of two different pH meter/electrode systems for pH measurements
in Calhoun sand............................................................................................................................... 124
Figure 9-6. Comparison of two pH meter/electrode systems for pH measurements in
Youngquist sand.............................................................................................................................. 125
Figure 9-7. Effect of potassium chloride (KCl) addition on pH. .................................................... 126
Figure 9-8. Variation in minimum resistivity by sand across two levels of dilution water
resistivity and for slurry water content in the soil box that is either mostly sand or mostly
water. ............................................................................................................................................... 129
Figure 9-9. Variation in minimum resistivity by dilution water resistivity for six sands and
for slurry water content in the soil box that is either mostly sand or mostly water. ....................... 130
xv
Figure 9-10. Variation in minimum resistivity by slurry water content for six sands and for
dilution water resistivity between 200,000 and 1,000,000 ohm-cm. .............................................. 131
Figure 11-1. Box plots of soil pH by soil mine............................................................................... 146
Figure 11-2. Operating characteristic (OC) curve for pH for sample size N = 3. ........................... 147
Figure 11-3. Box plots of soil minimum resistivity by soil mine. .................................................. 150
Figure 11-4. Operating characteristic (OC) curve for minimum resistivity for sample size N
= 2. .................................................................................................................................................. 151
Figure 11-5. Operating characteristic (OC) curve for chloride for sample size N = 1. .................. 154
Figure 11-6. Box plots of soil sulfate concentration by soil mine. ................................................. 157
Figure 11-7. Operating characteristic (OC) curve for sulfate for sample size N = 1. ..................... 158
Figure 12-1 User’s input screen for PHREEQC. ............................................................................ 162
Figure 12-2. Cumulative probability for weekly rainfall amount at NADP’s FL41 site in
Sarasota County (2000 through 2012). ........................................................................................... 164
Figure 12-3. Modeled groundwater ion concentrations and pH at soil column exit....................... 167
xvi
List of Tables
Table 1-1 Mine Information................................................................................................................ 3
Table 2-1 FDOT Gradation Limits for Backfill Used in Walls with Soil Reinforcements ................ 6
Table 2-2 Acceptance Limits for Corrosion Properties of Compacted Select Backfill ...................... 7
Table 2-3 Recommended Sampling Protocol for Electrochemical Testing of MSE Wall
Backfillǂ ............................................................................................................................................... 9
Table 2-4 Sulfate Exposure Categories and Classes (ACI, 2011) .................................................... 13
Table 2-5 Ion Concentrations (ppm) in Weekly Rainfall Collected between 2000 and 2012
at NADP’s NTN FL 41 Site in Sarasota County, Florida ................................................................. 13
Table 2-6 FDOT MSE Retaining Wall Classification ...................................................................... 16
Table 2-7 Suggested Risk Levels (AASHTO R-9) ........................................................................... 19
Table 2-8 Typical Laboratory QA/QC Activities ............................................................................. 25
Table 3-1 Selected Test Methods for Soil Electrochemical Properties ............................................ 27
Table 3-2 Influence of Temperature on the pH of Pure Water ......................................................... 30
Table 4-1 Statistics for FDOT’s District 1 and 7 MSE Wall Backfill Corrosion Tests.................... 55
Table 5-1 Accuracy and Precision of FM for pH, Minimum Resistivity, Chloride, and
Sulfate Based on Aqueous Standards (2011 Version of FMs) ......................................................... 62
Sulfate Based on Aqueous Standards (Proposed Revised FMs) ....................................................... 63
Table 5-3 Accuracy and Precision of Florida Methods for pH, Minimum Resistivity,
Chloride, and Sulfate with Santa Fe River Soil ................................................................................ 65
Table 5-4 Level of Detection (LOD) and Method Detection Level (MDL) of Florida
Methods for pH, Minimum Resistivity, Chloride, and Sulfate for Santa Fe River Soil ................... 66
Table 5-5 Multiple Laboratory Results of Florida Methods for pH, Minimum Resistivity,
Chloride, and Sulfate for Starvation Hill Soil................................................................................... 67
Table 6-1 Summary Statistics for pH Measurements for a Sub-Sample of Air-Dried, Sieved,
and Mechanically-Split Santa Fe River Soil ..................................................................................... 70
Table 6-2 Sources of Variability in pH between Laboratories ......................................................... 71
Table 6-3 Summary Statistics for Minimum Resistivity Measurements for a Sub-Sample of
Air-Dried, Sieved, and Mechanically-Split Santa Fe River Soil ...................................................... 72
Table 6-4 Sources of Variability in Minimum Resistivity between Laboratories ............................ 73
Table 6-5 Summary Statistics for Measurements of a 30-ppm Chloride Standard* ......................... 76
Table 6-6 Sources of Variability in Chloride Concentration between Laboratories ........................ 77
Table 6-7 Summary Statistics for Measurements of a 30-ppm Sulfate Standard* ............................ 78
Table 6-8 Sources of Variability in Sulfate Concentrations between Laboratories.......................... 79
Table 7-1 Geotechnical Properties of Select Backfill Samples ........................................................ 81
Table 7-2 Results for Laboratory Analysis of Split Samples ........................................................... 82
Table 8-1 Summary Statistics for Ruggedness Studies with Santa Fe River Soil ............................ 86
Table 8-2 Ruggedness Study for pH ................................................................................................. 87
Table 8-3 Effect of Factors in pH Ruggedness Study for Unmodified Santa Fe River Soil ............ 88
Table 8-4 Effect of Factors in pH Ruggedness Study for Salt-Amended Santa Fe River Soil......... 88
Table 8-5 Ruggedness Study for Minimum Resistivity.................................................................... 90
xvii
Table 8-6 Effect of Factors in Minimum Resistivity Ruggedness Study for Salt-Amended
Santa Fe River Soil ........................................................................................................................... 91
Table 8-7 Ruggedness Study for Soil Chloride Concentration ......................................................... 92
Table 8-8 Effect of Factors in Chloride Ruggedness Study for Salt-Amended Santa Fe River
Soil .................................................................................................................................................... 93
Table 8-9 Ruggedness Study for Soil Sulfate Concentration ........................................................... 94
Table 8-10 Effect of Factors in Sulfate Ruggedness Study for Salt-Amended Santa Fe River
Soil .................................................................................................................................................... 95
Table 8-11 Summary Statistics for Ruggedness Study Results ........................................................ 97
Table 9-1 Summary Statistics for Treatment Study Results ........................................................... 105
Table 9-2 Comparison of pH Means by Soil Type, pH Electrode, and Calcium Chloride
Amendment ..................................................................................................................................... 115
Table 9-3 ANOVA for pH by Soil Type, pH Electrode, and CaCl2 Amendment .......................... 115
Table 9-4 Comparison of Soil Resistivity Means by Soil Type, Incremental Water Volume,
and Pre-Wetted versus Dry Soil ...................................................................................................... 117
Table 9-5 ANOVA for Soil Resistivity Means by Soil Type, Incremental Water Volume,
and Pre-Wetted versus Dry Soil ...................................................................................................... 117
Table 9-6 Comparison of Soil Chloride Means by Soil Type and Extraction Method................... 119
Table 9-7 ANOVA for Soil Chloride by Soil Type and Extraction Method .................................. 120
Table 9-8 Comparison of Soil Sulfate Means by Soil Type and Extraction Method ..................... 122
Table 9-9 ANOVA for Soil Sulfate by Soil Type and Extraction Method ..................................... 122
Table 9-10 Summary Statistics for pH in Calhoun Sand Using Two Different pH
Meter/Electrode Systems ................................................................................................................ 123
Table 9-11 Summary Statistics for pH in Youngquist Sand Using Two pH Meter/Electrode
Systems ........................................................................................................................................... 125
Table 9-12 Summary Statistics for Minimum Resistivity by Sand ................................................ 129
Table 9-13 Comparison of Conductivity (Field) and Minimum Resistivity (Lab) ......................... 132
Table 9-14 Contribution of Mineral Fraction Retained on a No. 10 Sieve to Total Mass and
pH.................................................................................................................................................... 133
Table 10-1 Inter-Laboratory Study Results for Proposed Revised FMs for pH, Minimum
Resistivity, Chloride and Sulfate .................................................................................................... 137
Table 10-2. Material C Results Revised FM for pH ....................................................................... 138
Table 10-3. Material D Results Revised FM for pH ....................................................................... 138
Table 10-4 pH Averages, Standard Deviations, and %RSD ........................................................... 139
Table 10-5 Multi-Laboratory Precision for FM for pH .................................................................. 139
Table 10-6 Multi-Laboratory Precision for FM for Minimum Resistivity ..................................... 140
Table 11-1 Descriptive Statistics for pH ......................................................................................... 145
Table 11-2 Comparison of Replicate and Treatment Study Statistics for pH................................. 148
Table 11-3 Descriptive Statistics for Minimum Resistivity ........................................................... 149
Table 11-4 Comparison of Replicate and Treatment Study Statistics for Minimum
Resistivity ....................................................................................................................................... 152
Table 11-5 Descriptive Statistics for Soil Sulfate Concentration ................................................... 156
Table 11-6 Comparison of Replicate and Treatment Study Statistics for Sulfate .......................... 159
xviii
Table 12-1 Selected Aqueous-Phase or Precipitation Reactions .................................................... 163
Table 12-2 Modeled ionic composition of rainwater and groundwater .......................................... 165
Table 12-3 Candler Soil Properties ................................................................................................. 165
Table A-1 Wimauma Sand Ruggedness Study Results for pH....................................................... 190
Table A-2 Wimauma Sand Significance of Ruggedness Study for pH .......................................... 191
Table A-3 Wimauma Sand Ruggedness Study Results for Minimum Resistivity ......................... 192
Table A-4 Wimauma Sand Significance of Ruggedness Study for Minimum Resistivity ............. 193
Table A-5 Wimauma Sand Ruggedness Study Results for Sulfate ................................................ 194
Table A-6 Wimauma Sand Significance of Ruggedness Study for Sulfate.................................... 195
Table A-7 Jahna Sand Ruggedness Study Results for pH .............................................................. 196
Table A-8 Jahna Sand Significance of Ruggedness Study for pH .................................................. 197
Table A-9 Jahna Sand Ruggedness Study Results for Minimum Resistivity ................................. 198
Table A-10 Jahna Sand Significance of Ruggedness Study for Minimum Resistivity .................. 199
Table A-11 Youngquist Sand Ruggedness Study Results for pH................................................... 200
Table A-12 Youngquist Sand Significance of Ruggedness Study for pH ...................................... 201
Table A-13 Youngquist Sand Ruggedness Study Results for Minimum Resistivity ..................... 202
Table A-14 Youngquist Sand Significance of Ruggedness Study for Minimum Resistivity ......... 203
Table A-15 Youngquist Sand Ruggedness Study Results for Sulfate ............................................ 204
Table A-16 Youngquist Sand Significance of Ruggedness Study for Sulfate................................ 205
Table A-17 Calhoun Sand Ruggedness Study Results for pH ........................................................ 206
Table A-18 Calhoun Sand Significance of Ruggedness Study Factors for pH .............................. 207
Table A-19 Calhoun Sand Ruggedness Study Results for Minimum Resistivity .......................... 208
Table A-20 Calhoun Sand Significance of Ruggedness Study for Minimum Resistivity .............. 209
Table A-21 Angelo’s Sand Ruggedness Study Results for pH (Replicate 1) ................................. 210
Table A-22 Angelo’s Sand Significance of Ruggedness Study for pH (Replicate 1) .................... 211
Table A-23 Angelo’s Sand Ruggedness Study Results for pH (Replicate 2) ................................. 212
Table A-24 Angelo’s Sand Significance of Ruggedness Study for pH (Replicate 2) .................... 213
Table A-25 Angelo’s Sand Significance of Ruggedness Study Factors for pH Obtained from
the Combined Results of Replicate Studies .................................................................................... 214
Table A-26 Angelo’s Sand Ruggedness Study Results for Minimum Resistivity ......................... 215
Table A-27 Angelo’s Sand Significance of Ruggedness Study for Minimum Resistivity ............. 216
Table A-28 Angelo’s Sand Ruggedness Study Results for Sulfate ................................................ 217
Table A-29 Angelo’s Sand Significance of Ruggedness Study for Sulfate.................................... 218
Table A-30 Sebring Sand Ruggedness Study Results for pH (Replicate 1) ................................... 219
Table A-31 Sebring Sand Significance of Ruggedness Study for pH (Replicate 1) ...................... 220
Table A-32 Sebring Sand Ruggedness Study Results for pH (Replicate 2) ................................... 221
Table A-33 Sebring Sand Significance of Ruggedness Study for pH (Replicate 2) ...................... 222
TableA-34 Sebring Sand Significance of Ruggedness Study Factors for pH Obtained from
the Combined Results of Replicate Studies .................................................................................... 223
Table A-35 Sebring Sand Ruggedness Study Results for Minimum Resistivity............................ 224
Table A-36 Sebring Sand Significance of Ruggedness Study for Minimum Resistivity ............... 225
Table A-37 Ruggedness Study Results for Sulfate ......................................................................... 226
Table A-38 Sebring Sand Significance of Ruggedness Study for Sulfate ...................................... 227
xix
Table A-39 Clermont Sand Ruggedness Study Results for pH (Replicate 1) ................................ 228
Table A-40 Clermont Sand Significance of Ruggedness Study Factors for pH (Replicate 1) ....... 229
Table A-41 Clermont Sand Ruggedness Study Results for pH (Replicate 2) ................................ 230
Table A-42 Clermont Sand Significance of Ruggedness Study Factors for pH (Replicate 2) ....... 231
Table A-43 Clermont Sand Significance of Ruggedness Study Factors for pH Obtained
from the Combined Results of Replicate Studies ........................................................................... 232
Table A-44 Clermont Sand Ruggedness Study Results for Minimum Resistivity ......................... 233
Table A-45 Clermont Sand Significance of Ruggedness Study for Minimum Resistivity ............ 234
Table A-46 Alico Road Sand Ruggedness Study Results for pH (Replicate 1) ............................. 235
Table A-47 Alico Road Sand Significance of Ruggedness Study Factors for pH (Replicate
1) ..................................................................................................................................................... 236
Table A-48 Alico Road Sand Ruggedness Study Results for pH (Replicate 2) ............................. 237
Table A-49 Alico Road Sand Significance of Ruggedness Study Factors for pH (Replicate
2) ..................................................................................................................................................... 238
Table A-50 Alico Road Sand Significance of Ruggedness Study Factors for pH Obtained
from the Combined Results of Replicate Studies ........................................................................... 239
Table A-51 Alico Road Sand Ruggedness Study Results for Minimum Resistivity...................... 240
Table A-52 Alico Road Sand Significance of Ruggedness Study for Minimum Resistivity ......... 241
Table B-1 Wimauma Sand Treatment Study Results for pH .......................................................... 242
Table B-2 Wimauma Sand Treatment Study Results for Minimum Resistivity ............................ 243
Table B-3 Wimauma Sand Treatment Study Results for Chloride................................................. 244
Table B-4 Wimauma Sand Treatment Study Results for Sulfate ................................................... 245
Table B-5 Jahna Sand Treatment Study Results for pH ................................................................. 246
Table B-6 Jahna Sand Treatment Study Results for Minimum Resistivity .................................... 247
Table B-7 Jahna Sand Treatment Study Results for Chloride ........................................................ 248
Table B-8 Jahna Sand Treatment Study Results for Sulfate ........................................................... 249
Table B-9 Youngquist Sand Treatment Study Results for pH ........................................................ 250
Table B-10 Youngquist Sand Treatment Study Results for Minimum Resistivity ........................ 251
Table B-11 Youngquist Sand Treatment Study Results for Chloride............................................. 252
Table B-12 Youngquist Sand Treatment Study Results for Sulfate ............................................... 253
Table B-13 Calhoun Sand Treatment Study Results for pH ........................................................... 254
Table B-14 Calhoun Sand Treatment Study Results for Minimum Resistivity.............................. 255
Table B-15 Calhoun Sand Treatment Study Results for Chloride .................................................. 256
Table B-16 Calhoun Sand Treatment Study Results for Sulfate .................................................... 257
Table C-17 Angelo’s Sand Treatment Study Results for pH .......................................................... 258
Table B-18 Angelo’s Sand Treatment Study Results for Minimum Resistivity ............................ 259
Table B-19 Angelo’s Sand Treatment Study Results for Chloride................................................. 260
Table B-20Angelo’s Sand Treatment Study Results for Sulfate .................................................... 261
Table B-21 Sebring Sand Treatment Study Results for pH ............................................................ 262
Table B-22 Sebring Sand Treatment Study Results for Minimum Resistivity............................... 263
Table B-23 Sebring Sand Treatment Study Results for Chloride ................................................... 264
Table B-24 Sebring Sand Treatment Study Results for Sulfate ..................................................... 265
Table B-25 Clermont Sand Treatment Study Results for pH ......................................................... 266
xx
Table B-26 Clermont Sand Treatment Study Results for Minimum Resistivity ............................ 267
Table B-27 Clermont Sand Treatment Study Results for Chloride ................................................ 268
Table B-28 Clermont Sand Treatment Study Results for Sulfate ................................................... 269
Table B-29 Alico Road Sand Treatment Study Results for pH ...................................................... 270
Table B-30 Alico Road Sand Treatment Study Results for Minimum Resistivity ......................... 271
Table B-31 Alico Road Sand Treatment Study Results for Chloride ............................................. 272
Table B-32 Alico Road Sand Treatment Study Results for Sulfate................................................ 273
Table B-33 Input Data for pH ANOVA ......................................................................................... 274
Table B-34 Treatment Study ANOVA for pH ................................................................................ 274
Table B-35 ANOVA Input Data for Minimum Resistivity (ohm-cm) ........................................... 275
Table B-36 Treatment Study ANOVA for Minimum Resistivity (ohm-cm) ................................. 275
Table B-37 ANOVA Input Data for Sulfate (ppm) ........................................................................ 276
Table B-38 Treatment Study ANOVA for Sulfate (ppm) .............................................................. 276
Table D-1 ANOVA for Replicate Study pH ................................................................................... 292
Table D-2 Replicate Study ANOVA for Minimum Resistivity (ohm-cm)..................................... 293
Table D-3 Replicate Study ANOVA for Sulfate (ppm).................................................................. 294
xxi
List of Abbreviations
200% fraction of aggregate passing through a 200 mesh sieve
Ag/AgCl silver/silver chloride
AASHTO American Association of State Highway and Transportation Officials
ASTM ASTM International, formerly the American Society for Testing and Materials
AQC acceptance quality characteristic
C-22 sampling datasheet
C-S-H calcium silicate hydrates
CEI Construction Engineer and Inspector
cm centimeter
CO2 carbon dioxide
CR corrosion rate
CV coefficient of variation
D2S maximum acceptable difference between test results
FDOT Florida Department of Transportation
F-test Used to test if variances from two populations are equal
F Faraday’s constant (Equation 3-1)
FM Florida Method
H+, H3O+ hydronium, proton
hr hr, hr
Hz hertz
I current
LL liquid limit
LIMS laboratory information management system
LOD limit of detection
M molar; moles/liter
MDL minimum detection level
min minute
MSE mechanically-stabilized earth
N number of samples
n number of electrons participating in an electrochemical reaction (Equation 3-1)
NADP National Atmospheric Deposition Program
NH4+ ammonium
NO3- nitrate
NIST National Institute for Standards and Technology
nm nanometer
NTN NADP National Trends Network
OC operating characteristic
%OM % organic matter
PHREEQC pH redox equilibrium model coded in ‘C’
PI plasticity index
xxii
ppm parts per million
QA/QC quality assurance/quality control
QST Qualified Sampling Technician
R gas constant (Equation 3-1)
R surface recession (Equation 2-3)
R, RT, R15.5 resistance, resistance at observed temperature, resistance at 15.5 oC (Equation 3-4)
%RE % relative error
%RSD % relative standard deviation
s sample standard deviation
sr repeatability standard deviation
sR reproducibility standard deviation
SMEWW Standard Methods of Evaluation for Water and Wastewater Analysis
SMO FDOT State Material Office
SO2 sulfur dioxide
SO42- sulfate
t-test Student’s t-test, use making inferences about the means for a small sample size
Type I () false positive; incorrect rejection of the null hypothesis
Type II () false negative, failure to reject a false hypothesis
USEPA United States Environmental Protection Agency
USF University of South Florida
USGS United States Geological Survey
W/C water to concrete ratio
V volts
yr year
µ mean, average
µg m-3 micrograms per cubic meter
µm micrometer
 population standard deviation, test error
2
√𝜎𝑡𝑜𝑡𝑎𝑙 total test error
 soil resistivity
xxiii
1 Introduction
1.1 Research Statement
Construction of mechanically-stabilized earth (MSE) walls in metropolitan areas within FDOT’s

Districts 1 and 7 has put pressure on local mines for soil that meets the requirements for select
backfill (material 092L). MSE walls are designed with reinforcement that consists of wire mesh,
metal strips, or structural geosynthetics; such reinforcement is imbedded in select backfill.
Select backfill must be non-corrosive, non-plastic, have a low liquid limit, and have low organic
matter and fine particle content (Scheer, 2013). Acceptance testing of select backfill occurs after
emplacement and compaction. The purpose of this research was to increase the probability that
after emplacement and compaction the acceptance quality characteristics (AQC) for corrosion
properties of select backfill were correctly met. Variability in measured corrosion properties was
seen not only because of soil heterogeneity, but also because of sampling strategies and
analytical methods. Improvements in sampling strategies and analytical methods were proposed
with the intent to improve the decision to accept or reject a select backfill.
1.2 Scope and Objectives
The ultimate goals of this research were to improve quality, speed completion, and reduce risk of
mechanically-stabilized earth (MSE) wall projects. Research objectives were to assure (1) that
variability in the corrosion properties of soil (pH, minimum resistivity, chloride, and sulfate
levels) due to sampling and analytical technique was much lower than variability in these levels
within and between soil sources and types and thus did not inflate the risk of emplacing a
corrosive soil as MSE wall backfill, (2) that the number of soil type samples analyzed prior to
acceptance of a backfill was appropriate, based on the expected distribution of corrosion
properties within the backfill, and (3) that the corrosion properties of backfill material did not
change appreciably over time, especially after emplacement and over the design lifetime of the
MSE wall. Corrosion properties of soil were tested with Florida Methods (FMs) 5-550, 5-551, 5-
552, and 5-553 for pH, minimum resistivity, water-soluble chloride, and water-soluble sulfate,
respectively.
The scope of this research included tasks to (1) review relevant literature, (2) examine trends in
corrosion properties of Districts 1 and 7 MSE backfill mines based on existing data, (3)
determine the single-laboratory contributions to method reproducibility of factors that relate to
field sampling and laboratory analysis and improve method reproducibility through changes in
the methods, (4) discover the distribution of corrosion properties within and between mines for a
representative sample of mines as suggested by objective (2), (5) model analytically the potential
for the corrosion properties of MSE backfill to change over time with an initial focus on changes
due to the ionic composition of rainfall, (6) determine multi-laboratory contributions to method
reproducibility, and (7) disseminate the project results as a final report. Tasks 1 through 6 are
briefly described in the following sections. This research was limited to 10 soils that were
1
candidate materials for select backfill and thus low in clay, organic matter, and salt content; these
soils were typical of MSE wall backfill, but not likely representative of the broader variability of
Florida waters and soils.
1.2.1 Literature review
In support of the overall project effort, literature surveyed under this effort included publications
on MSE wall design and performance, mechanisms of concrete damage from direct exposure to
air pollution and acidic rainfall, the structure and basis of quality assurance plans, and
descriptions and comparisons of FDOT’s methods (FMs) for corrosion properties with those of
similar national methods. Also included in the literature review was an expanded discussion of
pH measurement, with an emphasis on practical considerations for evaluating pH electrode
performance in soil suspensions.
1.2.2 Trends in backfill properties
Trends in backfill properties were investigated to (1) identify sources and source regions of
acceptable MSE wall backfill within FDOT’s Districts 1 and 7, and for these districts (2) review
databased results of MSE wall backfill (material 092L) for failure trends in pH, resistivity,
chloride, and sulfate acceptance tests.
1.2.3 Single-laboratory contributions to method reproducibility
Single-laboratory contributions to method reproducibility of factors that relate to method field

sampling and laboratory analysis were explored to improve method reproducibility through
changes in sampling strategies, sampling processing, and analytical methods. Field sampling
was used in the broadest sense to include sample collection, transport, and storage. Evidence-
based recommendations for revisions to the FMs for pH, resistivity, chloride, and sulfate levels
were proposed.
1.2.4 Replicate studies
Distributions of pH, resistivity, chloride, and sulfate levels within and between mines were
estimated for select backfill based on replicate samples of mined soils. From these distributions
were derived the method operating characteristic (OC) curves that form the basis of acceptance
limits. Twelve replicate soil samples were collected at each of eight mines and these samples
were characterized for geotechnical and corrosion properties by the USF soils laboratory.
1.2.5 Multi-laboratory contributions to method reproducibility
Multi-laboratory contributions to method reproducibility were estimated pre- and post-method

revision to assess improvement in pH, minimum resistivity, chloride and sulfate measurements.
Prior to method development, on-site testing of a select backfill was conducted at six FDOT and
nine commercial laboratories, with representative laboratories in each FDOT district. Post-
2
revision, six FDOT and the USF soils laboratory tested four materials that were candidates for
select backfill as part of two consecutive inter-laboratory studies. From the inter-laboratory
study results were prepared updated OC curves and a precision statement for each of the revised
methods.
1.2.6 Geochemical modeling
Modeling under this task investigated the time it would take for infiltrating rainwater to alter
significantly the corrosion properties of select backfill. Calculations were done with PHREEQC
version 3, which is a publically-available model that was used to simulate one-dimensional
transport of groundwater with multi-phase chemical equilibrium reactions.
1.2.7 Mine Information
Early in the project, two soils that met the requirements for select backfill—Starvation Hill and
Santa Fe River— were collected from storage bins at FDOT’s State Materials Office (SMO) in
Gainesville. These soils served as proficiency test soils to gain experience with the FMs for pH,
minimum resistivity, chloride and sulfate levels prior to collecting soil samples from mines.
Over the project period, eight soils from as many mines were collected and transported back to
the USF soils laboratory for geotechnical and corrosion property testing. The names and
locations of mines that participated in this study were listed in Table 1-1.
Table 1-1 Mine Information

Mine ID Date Sampled FDOT District Mine
Transcor Dirt Services, LLC
1 05/08/2014 7
16410 Balm-Wimauma Road, Wimauma, FL 33598
Jahna Industries, Inc.
2 05/29/2014 1
4910 SR 544, Haines City, FL 33844
Youngquist Bros Rock Mine
3 06/19/2014 1
15401 Alico Road, Ft. Myers, FL 33913
C C Calhoun Mine, Pit # 7
4 07/14/2014 1
3000 Bannon Island Road, Haines City, FL 33844
Angelo’s Recycled Materials
5 09/17/2014 7
41111 Enterprise Road, Dade City, FL 33525
Hickey Excavation, Inc.
6 10/19/2014 1
7877 County Road 17S, Sebring, FL 33876
Titan Florida Center Sand
7 10/29/2014 5
16375 Hartwood Marsh Road, Clermont, FL 34711
Cemex Alico Quarry
8 11/12/2014 1
11840 Alico Road, Ft. Myers, FL 33913
3
1.3 Organization of Report
This report was organized along the lines of the contractual tasks, although not necessarily in the
same temporal sequence. Tasks were mapped into chapters as follows:
 Literature review was divided into Chapter 2 Literature Review and Chapter 3
Measurement of Soil pH, Minimum Resistivity, Chloride and Sulfate;
 Trends in backfill properties were placed in Chapter 4 Trends in Select Backfill;
 Single-laboratory contributions to method reproducibility were split into four chapters,

Chapter 5 Single Laboratory Precision and Bias, Chapter 7 Soil Sampling of Select
Backfill, Chapter 8 Ruggedness Studies, and Chapter 9 Treatment and Other Studies;
 Replicate studies were placed in Chapter 11 Replicate Studies and Implications for
Quality Assurance;
 Multi-laboratory contributions to method reproducibility were grouped into Chapter 6

Laboratory Visits and Chapter 10 Inter-Laboratory Study;
 Geochemical modeling was organized as Chapter 12 Geotechnical Modeling of Ion

Transport in Sandy Soil; and
 Integration of project results occurred in Chapter 13 Revised Florida Methods for pH,
Minimum Resistivity, Chloride, and Sulfate and Chapter 14 Summary and
Recommendations
Proposed revisions were based on 2011 versions of FDOT FM 5-550 Florida Method of Test for
pH of Soil and Water, FDOT FM 5-551 Florida Method of Test for Resistivity of Soil and
Water, FDOT FM 5-552 Florida Method of Test for Chloride of Soil and Water, and FDOT FM
5-553 Florida Method of Test for Sulfate of Soil and Water. In this report further references to
these methods are as FM for pH, FM for minimum resistivity, FM for chloride, and FM for
sulfate, respectively.
4
2 Literature Review
2.1 Mechanically-Stabilized Earth (MSE) Walls
2.1.1 Features of an MSE wall and properties of select backfill
Mechanically-stabilized earth (MSE) walls have evolved over the past 40 years as a competitive
replacement for reinforced concrete walls. MSE walls can be built to heights that exceed 25 m
without the need of rigid foundation support and are advantageous where land availability is
limited or at high cost, and where the ground is subject to slope instability or foundation soils are
poor (Elias et al., 2001).
The features of an MSE wall are its pre-cast concrete facing panels built on a leveling pad and
topped with a coping, and reinforcing elements attached to panels that extend behind the front
face of the wall into select granular backfill or flowable fill, as shown in Figure 2-1 (Castellanos,
2012; FDOT, 2011). Construction of MSE walls with dry cast modular blocks is an option for
lower retaining walls (Elias et al., 2001). Filter fabric installed behind the wall panels mitigates
soil extrusion through the wall joints (Castellanos, 2012). Reinforcing elements are typically
galvanized steel strips, but may also be grids, bar mats, welded or woven wire mesh of either
galvanized steel or plain steel, or synthetic mesh of polypropylene, polyethylene, or polyester
(Elias et al., 2001). Select granular backfill or flowable fill and reinforcing elements represent
much of the construction cost (Elias et al., 2001).
Acceptable MSE wall backfill is characterized by high friction between the reinforcing element
and backfill, good drainage, ease of emplacement and compaction, and low risk of
microbiologically-induced corrosion (Elias et al., 2001). As such, backfill criteria generally
exclude soils with significant clay or organic matter content. The Florida Department of
Transportation (FDOT) requires that MSE wall backfill be a non-plastic soil, with a liquid limit
less than 15, organic matter content not more than 2%, and gradation limits as shown in Table 2-
1, with testing done per AASHTO T90, T89, T267, and T27, respectively (Scheer, 2013).
Organic matter content is calculated as the average for three samples collected at each pit stratum
or within each stockpile; if an individual sample exceeds 3% of the average, the pit stratum or
stockpile is unsuitable (Scheer, 2013). In comparison, AASHTO (2012) limits the organic
matter content to 1%.
5
Figure 2-1. Schematic of a mechanically-stabilized earth (MSE) wall (FDOT, 2011).
Table 2-1 FDOT Gradation Limits for Backfill Used in Walls with Soil Reinforcements
Sieve Size Percent Passing
88.9 mm (3-1/2 inches) 100
19 mm (3/4 inch) 70-100
No. 4 30-100
No. 40 15-100
No. 100 0-65
No. 200 0-12
The design life of an MSE wall is the period of time that the wall is expected to be in service as
defined by service and strength conditions (AASHTO, 2012). The design life of MSE walls is
typically 50, 75, or 100 yr (Fishman and Withiam, 2011); the FDOT mandates a minimum
design life of 75 yr or 100 yr (FDOT, 2013). Choices for select granular backfill and reinforcing
elements are critical in achieving the design life of MSE walls, and these choices are determined
6
in large part by the potential for a pull-out failure of the reinforcing elements (AASHTO, 2012;
Fishman and Withiam, 2011; Gladstone et al., 2006).
2.1.2 Corrosion of MSE wall reinforcement
Reinforcing elements made of metal, whether galvanized metal strips or plain steel mesh, once
emplaced in soil are subject to corrosion. Variations in the soil environment can establish
voltage potentials between adjacent metallic sites while soil pore water acts as an electrolyte for
current flow between the sites (Jack and Wilmott, 2011). Key environmental factors that affect
metal corrosion in soil include temperature, soil porosity, reduction/oxidation potential, water
content, oxygen level, acidity or alkalinity, salt content, and microbial activity (Jack and
Wilmott, 2011; Sagüés et al., 2009). To reduce the corrosion risk of metal reinforcing elements,
elements are placed in backfill that is engineered to minimize corrosion. For example, backfill:
 Gradation limits provide for rapid drainage, low water retention, and soil aeration;
 Low organic matter content minimizes nutrient availability for microbial growth;
 Low water content reduces the opportunity for variations in oxygen levels within the soil
that can lead to differential voltages and thus active corrosion sites;
 Low levels of salts and high soil resistivity keep low the corrosion currents; and
 Low levels of acids and bases minimize the potential attack of these chemicals on
protective metallic oxide layers that in time coat the emplaced metal.
In Table 2-2 are shown the AASHTO and FDOT acceptance limits for soil pH, minimum
resistivity, chloride and sulfate levels in compacted select backfill (Scheer, 2013).
Table 2-2 Acceptance Limits for Corrosion Properties of Compacted Select Backfill
Agency pH Minimum Resistivity, ohm-cm Chloride, ppm Sulfate, ppm
AASHTO 5 to10 3,000 <100 <200
FDOT *5 to 9 3,000 <100 <200
*Exceptions allow for use of select backfill material with pH 4.5 to 5.0 per Scheer (2013).
Exceptions for use of select backfill material with pH 4.5 to 5.0 are as follows (Scheer, 2013):
 The interior face of MSE wall panels have 3 inches (760 mm) of concrete cover over the
reinforcement; and
 For concrete used in the panels, the quantity of cement replaced with Type F fly ash is
10% to 20% by weight;
 The quantity of cement replaced with slag is 50% to 60% by weight;
 Portland cement is 30% by weight of total cementitious material;
 The total weight of Type F fly ash and slag does not exceed 70% of total cementitious
material; and
7
 Metallic pipe must not be placed in backfill materials with pH < 5.
AASHTO’s (2012) corrosion model for galvanized steel in MSE wall service is shown in
Equation 2-1, where Ec is the thickness of metal at the end of service life, En is the nominal
thickness of steel reinforcement at construction, and Es is the sacrificial thickness of metal that is
expected to be lost by uniform corrosion during the service life.
𝐸𝑐 = 𝐸𝑛 − 𝐸𝑠 (2-1)
AASHTO applies to its model a 75-yr service life, an initial zinc thickness of 86 µm, a time of 2
yr for the initial zinc corrosion rate, and rates of 15 µm yr-1, 4 µm yr-1, and 12 µm yr-1 for the
initial zinc, final zinc, and steel corrosion rates, respectively. This model yields 16 yr to loss of
the zinc coating and a steel loss of 708 µm per side (AASHTO, 2012; Fishman and Withiam,
2011). FDOT’s model is the same, except that FDOT’s corrosion rates in mils yr-1 give a closer
approximation than the AASHTO’s corrosion rates in mils yr-1 to the above rates expressed in
µm yr-1; moreover, FDOT includes a steel corrosion rate of 7 µm yr-1 applied between 75 and
100 years for those walls with a 100-yr service life (FDOT, 2013).
In-situ and laboratory testing of in-service MSE wall reinforcing elements indicate that
AASHTO’s model is conservative, that is, it tends to over-predict corrosion losses of zinc in
galvanized steel, when the select granular backfill material meets AASHTO’s criteria for
electrochemical parameters (Table 2-2) (Fishman and Withiam, 2011; Gladstone et al., 2006;
Berke et al., 2008).
Fishman and Withiam (2011) investigated corrosion rates of reinforcing elements for 53 MSE
walls in six states including Florida and applied a best-fit power law model (Equation 2-2) to
describe the corrosion rate (CR) in µm yr-1 as a function of soil resistivity () in ohm-cm.
Fishman and Withiam (2011) did not make clear for Equation 2-2 that the modeled data were for
minimum resistivity, a distinction they made for Equation 2-1. MSE walls ranged in age from
one to 30 yr with an average age of 13 yr; thus, corrosion rates largely reflected corrosion of the
zinc coating. They found on the average, corrosion rates were 10 times higher for observations
at sites with soil resistivity less than 3000 ohm-cm than for observations at sites with soil
resistivity greater than 3000 ohm-cm.
𝐶𝑅 ≈ 1400𝜌−0.75 (2-2)
In simple terms, the presence of salt and water in soils provides a pathway for corrosion-inducing
currents. All four of the soil electrochemical tests (Table 2-2) relate to soil salt content but Elias
et al. (2009) found that the most reliable indicator of a soil’s corrosion potential was minimum
resistivity. Fishman and Withiam (2011) found that in reinforced fills, the condition of minimum
resistivity < 3000 ohm-cm and pH < 5.0 significantly affected steel corrosion rates. Nevada’s
Department of Transportation (NDOT), for example, found that in two MSE walls, aged 9 yr and
25 yr, metal reinforcing elements were corroded such that the older wall required expensive
8
mitigation (Thornley et al., 2010). NDOT attributed the cause to over-predicted soil resistivity
based on Nevada’s T235B conductivity method (NDOT, 1999) as compared to AASHTO’s T288
minimum resistivity method.
Elias et al. (2009) reported that wide scatter in electrochemical test results were evidence of poor
backfill conditions and recommended that sampling protocols include not only a minimum
number of samples but also a maximum standard deviation in test results (Table 2-3). Moreover,
their recommendation emphasized that samples for electrochemical testing be acquired with the
help of an excavator to reach the interior top, mid, and bottom levels of a stockpile.
Table 2-3 Recommended Sampling Protocol for Electrochemical Testing of MSE Wall
Backfillǂ
Range of During
Preconstruction
Minimum General Construction
Comments
Resistivity Description Number of resistivity, Sample
min, ohm-cm Samples ohm-cm Interval, yd3*
Crushed (1) pH outside the
rock and specified limits is not
1 for
gravel, < allowed for any sample.
>10,000 resistivity; NA NA
10% (2) Backfill sources shall
3 for pH
passing No. be rejected if min
10 sieve measured for any sample
Sandy 3 for 4,000 for is less than 700 ohm-cm,
5,000 to
gravel and resistivity; 6 <2,000 resistivity; chloride is > 500 ppm, or
10,000
sands for pH 2,000 for pH sulfate is greater than
1,000 ppm.(3) For
Silty sands materials with min < 5,000
5 for 1,000 for
and clayey ohm-cm,  for chloride
<5,000 resistivity; <1,000 resistivity;
sand, and sulfate shall be less
10 for pH 2,000 for pH
screenings than 100 ppm and 200
ppm, respectively.
ǂ
Elias et al. (2009); *1 yd3 = 0.7646 m3
Reinforcing elements made of synthetic polymeric materials are also subject to degradation over
time and the degradation rate, albeit slow for typical MSE wall construction, is influenced by
material properties, reinforcement design, and environmental factors such as temperature,
mechanical damage, stress levels, and chemical exposure (AASHTO, 2012). FDOT provides
guidance on the suitability of synthetic reinforcing products for MSE walls (Scheer, 2013).
9
2.1.3 Susceptibility of concrete to damage from environmental exposures
Air pollution has been associated with damage to stone, mortar, cement, and concrete in both
ancient and modern construction, especially in urban or industrial areas where acidic gases and
catalytic particles may be present in relatively high concentrations (Lipfert and Daum, 1992;
Hamilton and Crabbe, 2009; Roots, 2008; Sabbioni et al., 2001). Air pollutants commonly
linked to materials degradation include sulfur dioxide, sulfuric acid, nitrogen oxides, nitric acid,
ozone and its precursors, and aerosols (Hamilton and Crabbe, 2009). Concrete degradation is
also caused by acid rain: rain, sleet, snow, ice, fog, and cloud water that have interacted with
atmospheric carbon dioxide to form carbonic acid and with atmospheric pollutants such as
sulfuric and nitric acids. Carbon dioxide although a natural constituent of the atmosphere, may
be present at elevated concentrations in urban areas as a consequence of closely packed sources
of fossil fuel combustion (Grimmond et al., 2002; Neville, 1996; Idso et al., 2002; Rice and
Bostrom, 2011). Alkali dust, sea salt, and ammonia can neutralize acids present in the
atmosphere and in rainfall (Hamilton and Crabbe, 2009).
In-service 20th-century concrete structures exposed to relatively high levels of air pollution have
weathered crusts with evidence of carbonation, sulfur-attack, alkali-aggregate reactions, and
cracking (Marinoni, et al., 2003; Ozga et al., 2011). Damage mechanisms to concrete include (1)
dissolution and leaching of soluble components, (2) alterations that soften or weaken the
concrete, and (3) formation of expansive constituents that can cause cracking. Studies that
explore the interaction of dry- or wet-depositing pollutants on the properties of construction
materials include one or more of these three research components: examination of in-service or
retired structures, field or laboratory testing of specimens, and mathematical modeling of attack
processes. Models tend to be calibrated against accelerated testing of concrete specimens in
aggressive media on the assumption that this type of testing can adequately simulate long-term
behavior (Marques and Costa, 2010; Rozière et al., 2009; Samson and Marchand, 2007).
Summarized here were the general trends that appeared across these modes of study and that
were relevant to the service life of MSE walls, with an emphasis on acid attack, carbonation, and
sulfur attack.
Properties of concrete that are diminished over time by acid attack include loss of mass, loss of
tensile and compressive strength, loss of flexibility, and elastic and dynamic modulus of
elasticity, as examples (Fan et al., 2010; Zivica and Bajza, 2001). Acids that can severely
damage concrete include inorganic acids: carbonic, hydrochloric, hydrofluoric, nitric,
phosphoric, and sulfuric, and organic acids: acetic, citric, formic, humic, lactic, and tannic
(Neville, 1996). Attack can be severe below pH 5.5, but the continued availability of the acid is
as or more important than the pH in causing damage, because it takes time for the acid to
penetrate the reacted product to the undamaged concrete (Beddoe and Dorner, 2005; Neville,
1996).
The general mechanism of acid attack in ordinary Portland concrete is summarized as follows
(Beddoe and Dorner, 2005):
10
 Movement of protons (more precisely, hydronium ions H3O+) into the concrete;
 Dissolution of calcium hydroxide and calcium from calcium silicate hydrate (C-S-H) or
calcium aluminum hydrate;
 Migration of the calcium towards the concrete surface; and
 Development of a concrete damage layer of predominantly hydrated silicates.
The solids that dissolve and precipitate as the acid attack proceeds depend upon the counter
anion (carbonate, sulfate, and nitrate as examples), the salts and minerals present in the concrete,
and the concrete pore solution pH (Beddoe and Dorner, 2005). Under some conditions,
precipitating solids such as calcium carbonate and calcium sulfate can reduce the pore volume
and slow acid diffusion; some solids, including calcium chloride and calcium nitrate, have
relatively high water solubility and readily leach in pore water (Beddoe and Dorner, 2005; Zivica
and Bajza, 2001).
Acidic gases—carbon dioxide, sulfur dioxide, and nitric acid, as examples—can diffuse faster
and further into concrete than their aqueous counterparts, but ultimately require water for the
attack to proceed (Okochi et al., 2000). Concrete pore water can help or hinder acid attack: too
much pore water and diffusion rates slow; too little water and reaction rates slow; attack rates are
optimized when the relative humidity is between 50% and 75% (Zivica and Bajza, 2001). Wet
and dry cycling may damage concrete faster than conditions that are consistently either wet or
dry (Beddoe and Dorner, 2005; Gruyaert et al., 2012; Sersale et al., 1998; Xie et al., 2004);
Okochi et al. (2000), however, found that neutralization depth was greater in sheltered than in
unsheltered concrete after a 2-yr exposure to ambient air pollution. Neutralization depth (Okochi
et al., 2000) and strength loss (Cao et al., 1997; Xie et al., 2004) was proportional to proton
concentration in test specimens.
Properties of concrete such as the ratio of water to cement (W/C), alkali and sulfur content,
aggregate material, and additives may enhance or retard acid attack. For example, higher W/C
ratios tend to increase attack rates due to the increased availability of water (Sersale et al., 1998),
but larger pore volumes may mitigate the damage due to expansive mineral formation (Jacques et
al., 2010). The presence of sodium or potassium hydroxides may increase attack rates as the
dissolved alkali diffuse readily to the concrete surface. The use of acid-resistant aggregate and
aggregate with a greater content of small particles that can reduce concrete pore volume tend to
reduce attack rates (Zivica and Bajza, 2002). Pozzolans such as fly ash or silica fume can
convert calcium hydroxide to C-S-H, which can increase the concrete density and slow acid
diffusion and thus reduce a concrete’s vulnerability to acid attack, or pozzolans can by lowering
the calcium hydroxide content reduce the amount of acid needed to lower the pore water pH and
thus increase the damage (Neville, 1996). Replacement of cement with blast furnace slag in
Portland concrete resulted in a concrete with lower calcium oxide but higher silica dioxide
content that was more resistant to acid attack (Gruyaert et al., 2012).
11
Carbonation of ordinary Portland concrete occurs when water-dissolved carbonic acid reacts
with calcium hydroxide in concrete to form calcium carbonate and water (Rozière et al., 2009;
Steffens et al., 2002: Zivica and Bajza, 2001). Calcium hydroxide is present in concrete paste at
concentrations of 20% to 25% by volume (Mindess et al., 2003). Once calcium hydroxide has
been carbonated, the pH of the concrete pore water drops from ~13 to ~9. At pH 9, the
passivation layer on reinforcement steel is removed. If oxygen and moisture are present,
removal of the passivation layer allows corrosion of the steel and formation of expansive
corrosion products that can lead to cracking (Mindess et al., 2003; Neville, 1996). Carbonation
progresses at a rate proportional to yr0.5, but the rate depends on internal factors such as concrete
porosity and moisture content, and external factors such as atmospheric carbon dioxide
concentration and temperature (Duffó et al., 2012; Neville, 1996). Over a 75-yr lifetime in good
quality concrete the carbonation depth is on the order of 2 to 9 mm (Duffó et al., 2012); a
carbonation of depth of 15 mm in 16 yr has been observed in low quality concrete (Neville,
1996). Observed carbonation rates may increase over the next century if global trends in
atmospheric carbon dioxide concentrations and temperatures continue (Yoon et al., 2007).
Carbonation tends to increase both the tensile and compressive strength of concrete but it also
increases shrinkage, which can lead to crack formation (Mindess et al., 2003); precipitates from
the carbonation process may also clog concrete pores and form a protective zone (Glasser et al.,
2008).
Concrete structures are exposed to sulfate from atmospheric sulfur dioxide, from sulfuric acid on
aerosols or dissolved in rainwater, or from sulfuric acid or sodium sulfate in groundwater, as
examples. The American Concrete Institute (ACI, 2011) has divided sulfate exposure into
categories and classes (Table 2-4). When exposed to sulfate, concrete can undergo a reaction
that precipitates gypsum, ettringite, and thaumasite, or mixtures of these solids (Cao et al., 1997;
Glasser et al., 2008; Neville, 1996). Thaumasite formation requires not only sulfate and C-S-H,
but carbonate in solution (Glasser et al., 2008). The presence of gypsum and thaumasite is
associated with loss of material strength and adhesion, and the presence of ettringite with
localized pressure that leads to concrete cracking (Cao et al., 1997; Collepardi, 2003).
Collepardi (2003) emphasizes that three conditions must be met for ettringite formation: high
permeability concrete, a sulfate-rich environment, and the presence of water. The concentration
and species of dissolved ions such as protons, sodium, potassium, calcium, and magnesium
influence mineral formation and leaching (Fan et al., 2010; Glasser et al., 2008; Kanazu et al.,
2001). (Sulfur may also be present in the starting materials of concrete and under some
conditions during the lifetime of the concrete may dissolve and re-precipitate as ettringite.)
12
Table 2-4 Sulfate Exposure Categories and Classes (ACI, 2011)
Condition
Severity Class Water-soluble sulfate in soil, % Dissolved sulfate in water,
by mass ppm
Not applicable S0 sulfate < 0.10 sulfate < 150
Moderate S1 0.10 ≤ sulfate < 0.20 150 ≤ sulfate <1500
Severe S2 0.20 ≤ sulfate ≤ 2.00 1500 ≤ sulfate ≤10,000
Very severe S3 sulfate > 2.00 sulfate > 10,000
Stark (2002) found that low ratios of W/C provided the greatest resistant by concrete to sulfate
attack. Concretes that are prepared with pozzolanic cements such as metakaolin, silica fume,
blast furnace slag, and fly ash have lower permeability and offer resistance to sulfate attack (Cao
et al., 1997; Djuric et al., 1996), except that fly ash amendment may not protect against
thaumasite formation (Glasser et al., 2008). Results conflict for blast furnace slag and silica
fume, which is a possible consequence of varied approaches and scenarios for accelerated testing
(Cao et al., 1997; Gruyaert et al., 2012; Stark, 2002). Cement replaced with blast furnace slag in
Portland concrete was more susceptible to carbonation during an accelerated test for sulfate
attack, and damage from sulfate attack was more severe with high replacement levels of blast
furnace slag (Gruyaert et al., 2012).
One approach to assessing the impact of environmental exposure on concrete durability is from a
dose-response relationship, where dose refers to the exposure of the concrete to an air
pollutant(s) and is proportional to air pollutant concentration x exposure time, and response is a
corresponding loss of concrete mass or function. Dose and response relationships can be
established from experiments in a controlled setting (Fan et al., 2010; Jacques et al., 2010) or
measured under natural conditions (Tidblad, et al., 1998; Roots, 2008). Whether simulated or
natural, the temporal pattern of weather phenomena such as temperature, rainfall rate and
amount, and relatively humidity plays an important role in concrete degradation.
Table 2-5 shows ion concentrations in weekly rainfall that was collected at a National
Atmospheric Deposition Program (NADP) National Trends Network (NTN) FL 41 site in
Sarasota County, Florida (27.38 N, 82.28 W). For the period 2000 to 2012, the median,
minimum, and maximum pH values at this site were 5.01, 3.99, and 6.59, respectively; pH 7 is
neutral, so this rainfall was acidic (NADP, 2013). The median pH of rainfall observed at this site
was typical for NADP NTN sites across Florida (NADP, 2013).
Table 2-5 Ion Concentrations (ppm) in Weekly Rainfall Collected between 2000 and
2012 at NADP’s NTN FL 41 Site in Sarasota County, Florida
Statistic pH Lab Ca2+ Cl- K+ Mg2+ Na+ NH4+ NO3- SO42-
Median 5.01 0.125 0.626 0.026 0.049 0.345 0.123 0.684 0.820
Minimum 3.99 0.010 0.027 0.003 0.003 0.010 0.006 0.020 0.120
Maximum 6.59 1.29 8.8 0.300 0.639 5.83 1.23 5.60 5.27
13
NADP’s NTN sites were intentionally located outside the direct influence of one or more air
pollution sources such as a power plant plume, but in urban areas concentrations of pollutants in
rainfall may be higher. For example, Strayer et al. (2007) observed lower pH and higher
chloride and sulfate concentrations in rainfall measured at the eastern end of the Gandy Bridge in
Tampa, Florida (27.85 N, 82.55 W) between 1996 and 2004. For comparison, at the Gandy
Bridge site the median, minimum and maximum values were, respectively, 4.5, 3.4, and 6.0 for
pH; respectively; 1.48, 0.19, and 14.3 ppm for chloride concentrations, respectively; and 2.09,
0.19, 15.8 ppm for sulfate concentrations, respectively. Even for urban Tampa, however, sulfate
concentrations did not pose a threat to concrete of sulfate attack (Table 2-4).
Tidblad et al. (1998) developed a dose-response curve for unsheltered Portland limestone
exposed to air pollution (Equation 2-3), where R is the surface recession (depth to undamaged
surface) in µm, t is time in yr (1 to 8 yr), SO2 is the sulfur dioxide ambient air concentration in
µg m-3 (1 to 83 µg m-3), T is temperature in oC (2-19 oC), Rain is rainfall in mm (327-2144 mm),
and H+ is the proton concentration in ppm (0.0006 to 0.13 ppm).
R = t0.96(2.7[SO2]0.48exp{-0.018T}+0.019Rain[H+]) (2-3)
Average values for urban Tampa were substituted into Equation 2-3: t = 8 yr, SO2 = 50 µg m-3
(USEPA, 2013), T = 19 oC, Rain = 1200 mm, and H+ = 0.032 ppm (Strayer et al., 2007) for a
surface recession R of 98 µm or  0.1 mm. An updated formulation of the dose-response curve
for Portland limestone published by Roots (2008) included the combined effects of relative
humidity, sulfur dioxide, nitric acid, aerosols, and rainfall acidity and yielded a surface recession
R of  0.3 mm for Tampa conditions. From the dose-response relationship, average urban
Tampa air pollution does not appear to pose a threat to MSE wall concrete, with the caveat that
air pollutant concentrations in the vicinity of power plants or industrial facilities such as sulfuric
acid or nitric acid manufacturing plants may pose a significant risk. To address this risk, FDOT
(2013) specifies amendment of structural concrete with pozzolans (silica fume, metakaolin, or
ultrafine fly ash) (Figure 2-2, Table 2-6).
14
Figure 2-2. Permanent retaining wall selection process flowchart.
15
Table 2-6 FDOT MSE Retaining Wall Classification
Other Allowable FDOT Wall
Durability Requirements
Types
Applicable Concrete Concrete Pozzolan Soil
FDOT Cove Class for Additions? Reinforcement 2A 2B 2C 2D 2E 2F
Wall Type (in)* Panels **
Type
Type 2A 2 II No Metal     
Type 2B 2 IV No Metal    
Type 2C 3 IV No Metal   
Type 2D 3 IV Yes Metal  
Type 2E 3 IV No Plastic 
Type 2F 3 IV Yes Plastic
*
1 in = 25.4 mm; **Silica fume, metakaolin, or ultrafine fly ash
2.2 Quality Assurance (QA) Plans for Construction Materials
Quality assurance (QA) plans devised for construction materials support an overall project goal
to build a structure of known integrity and service life. Burati et al. (2003) outlined five basic
steps to creating or modifying a QA plan:
 Assimilate data and information to determine relevant construction parameters;

 Establish the desired level of quality for construction materials;
 Design a QA plan, including acceptance quality characteristics (AQC), statistical quality
measures, buyer’s and seller’s risks, lot size, number of samples, acceptance limits, and
payment-adjustment provision;
 Monitor how the plan performs; and
 Make adjusts as necessary.
The ultimate goal of a change to a QA plan is to improve product quality, speed project
completion, reduce risk, or lower bid cost (Burati et al., 2003).
Quality assurance (QA) plans for MSE wall materials are intended to reduce the risk of an early
wall failure. Select backfill is a critical component of an MSE wall and thus properties of select
backfill such as particle size gradation, organic matter content, plasticity, liquid limit, and
corrosion have been identified as important acceptance quality characteristics (AQC) (Scheer,
2013). Our research had as its focus the AQC for corrosion—pH, minimum resistivity, chloride
and sulfate levels—and was motivated by a material rejection frequency that was too high for
one or more of these AQC. Under consideration were the AQC, statistical measures, risk, lot
size or soil type, number of samples, and acceptance limits but not payment-adjustment
provisions, which were beyond the research scope.
16
2.2.1 Acceptance quality characteristics (AQC) for corrosion tests
Acceptance quality characteristics (AQC) are embedded in a QA plan that includes both process
(quality) control and product acceptance (Burati et al., 2003). Figure 2-3 illustrates FDOT’s
process control for pH of select backfill, as an example. Test results for material process control
and product acceptance depend not only on the written procedures of a test method but also on a
broader suite of influences such as material properties, physical infrastructure, laboratory
practices, operator training, costs versus payments, and so forth (Burati et al., 2003). These
influences manifest themselves as bias and (im)precision in test results. Process control and
product acceptance limits for select backfill corrosion AQC such pH, minimum resistivity,
chloride, and sulfate levels include a margin of safety, which was initially established based on
statistical inference drawn from test data that capture the relative (im)precision due to these
influences. Equation 2-4 offers a perspective on how variability (or error) in test results is
partitioned (Burati et al., 2003; Mason, 1992). Sampling error, however, is often the largest
contributor to total error (Mason, 1992). Standards such as AASHTO T2 and T248 intend to
reduce systematic error associated with sampling and process, respectively; certification of
Qualified Sampling Technicians (QSTs) also aims to reduce systematic error associated with
sampling. Test methods such as the FM for pH help to minimize systematic errors due to testing;
use of qualified or certified operators and/or laboratories, reference materials or split samples,
and control charts are additional strategies for mitigating systematic error due to testing (Burati
et al., 2003).
2 2 2 2 2 2
𝜎𝑡𝑜𝑡𝑎𝑙 = 𝜎𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙 + 𝜎𝑒𝑛𝑣𝑖𝑟𝑜𝑛𝑚𝑒𝑛𝑡 + 𝜎𝑠𝑎𝑚𝑝𝑙𝑖𝑛𝑔 + 𝜎𝑝𝑟𝑜𝑐𝑒𝑠𝑠 + 𝜎𝑡𝑒𝑠𝑡𝑖𝑛𝑔 (2-4)
Verification is a key aspect of process control and product acceptance. Burati et al. (2003)
distinguishes between verification of test method and product, where the former is done through
a comparison of a split samples—that is, subsamples of the same material—and the latter is
done through a comparison of independent samples. For verification, one simple approach is to
apply the ‘D2S limits’ or the maximum acceptable difference between test results, where the
D2S limits represent the 95% confidence interval for individual results and may be derived from
test method inter-laboratory studies (ASTM C670). If the difference between agency or engineer
and contractor test results is within the D2S limits, for example, the contractor’s test is verified.
Where replicate tests of independent samples are available, a difference between agency or
engineer and contractor results can be inferred from a t-test on the averages and an F-test on the
variances.
17
Process Control
for Backfill: pH
AASHTO T-2,
T-246; QST
Certification,
FM5 - 550
Agency or Engineer
samples and tests
stockpile or pit material
Yes No
5 ≤ pH ≤ 9
Certify material Reject material

for pH
Transport material to MSE

wall construction site
Figure 2-3. Process control for pH of select backfill for MSE walls with metallic
reinforcement.
The basis of a process control or acceptance test limit is an operating characteristic (OC) curve,
2
which in some way relates test error, ideally √𝜎𝑡𝑜𝑡𝑎𝑙 (Equation 2-4), to risk. Burati et al. (2003)
referred to a Type I () error as the seller’s risk—the risk the contractor has that a product is
rejected—and a Type II () error as the buyer’s risk—the risk the agency has in accepting a
faulty product. Burati et al. (2003) provided a summary table of suggested risk levels based on
AASHTO R9 (Table 2-7).
18
Table 2-7 Suggested Risk Levels (AASHTO R-9)
Degree of Importance Recommended Type I () error Recommended Type II () error
Critical: Essential to the
0.050 0.005
preservation of life
Major: Necessary for the
prevention of substantial 0.010 0.050
financial loss
Minor: Does not materially affect
0.005 0.100
performance
Contractual: Provides uniform
0.001 0.200
standards for bidding
Although not explicitly stated in FDOT guidance documents, the process control and product
acceptance limits for minimum resistivity, chloride, and sulfate (Table 2-2) likely have a margin
of safety. The criteria for substructure environmental classification as it applies to steel exposed
to minimum resistivity in soil are as follows: <1,000 ohm-cm, extremely aggressive; >5,000
ohm-cm, slightly aggressive; and between 1,000 and 5,000 ohm-cm, moderately aggressive
(FDOT, 2013). A 3,000 ohm-cm minimum resistivity for MSE wall backfill falls within the
range of moderately aggressive (Scheer, 2013). The criteria for substructure environmental
classification as it applies to steel exposed to chloride concentrations in soil are as follows:
>2,000 ppm, extremely aggressive; <500 ppm, slightly aggressive; and between 500 ppm and
2,000 ppm, moderately aggressive (FDOT, 2013). A 100-ppm chloride concentration falls
within the range of slightly aggressive (Scheer, 2013). Finally, the criteria for substructure
environmental classification as it applies concrete exposed to sulfate concentrations in soil are as
follows: >2,000 ppm, extremely aggressive; <1000 ppm, slightly aggressive; and between 1,000
ppm and 2,000 ppm, moderately aggressive (FDOT, 2013). A 200 ppm sulfate concentration
falls within the range of slightly aggressive (Scheer, 2013). Thus, in the development of OC
curves for minimum resistivity, chloride, and sulfate levels, the assumed critical (rejection) limits
were 1,000 ohm-cm, 500 ppm, and 1,000 ppm, respectively.
Environmental classification for soil pH between steel substructures and MSE wall backfill are
incongruous. The criteria for substructure environmental classification as it applies to steel
exposed to pH in soil are as follows: <6 pH units, extremely aggressive; >7 pH units, slightly
aggressive; and between 6 and 7 pH units, moderately aggressive (FDOT, 2013). For MSE wall
backfill, pH must fall between pH 5 and pH 9 for metal reinforcement and between pH 4.5 and
pH 9 for geosynthetic reinforcement (Scheer, 2013). According to Elias et al. (2009), “corrosion
literature suggests that for bare steel and pH between 4 and 10, the corrosion rate is independent of
pH and depends only on how rapidly oxygen diffuses on the metal surface.” Soils that have a pH 4
or less are considered extremely acidic and soils that have a pH 10 or greater are considered
extremely alkaline. Thus, in the development of an OC curve for pH, the assumed lower and
upper critical (rejection) limits (for metal reinforcement) were pH 4 and pH 10, respectively.
19
A method operating characteristic (OC) curve can be constructed to show the probability of
accepting a good material (1-) versus rejection and acceptance limits, as was shown for pH in
Figure 2-4. Development of this OC curve assumed sampling and analysis for pH of two
independent and random samples (N = 2), a Type I () error of 0.010 and a Type II () error of
0.05 (Table 2-7), rejection limits of pH 4 and pH 10, and acceptance limits of pH 5 and pH 9
(Table 2-2). Also shown in Figure 2-4 was the impact of test error  on the OC curve. Refer to
Chapter 11, Section 11.2.4, for an explanation of the calculations behind the OC curve. For the
conditions modeled, the OC curve in Figure 2-4 indicated that for  > 0.4, the buyer’s statistical
power for accepting a good backfill was lower than the target probability of 0.95 or 95% (solid
horizontal line). In this example, several options were available to improve the buyer’s risk: (1)
change the acceptance limits, (2) increase the number of samples, (3) reduce the test error, (4)
shift more risk to the seller, (5) or modify the reinforcement design or material.
Figure 2-4. Example operating characteristic (OC) for pH N = 2.
20
2.2.2 FDOT’s QA plan for select backfill corrosion AQC
FMs for pH, minimum resistivity, chloride, and sulfate began as field tests for environmental
characterization at bridge sites but over time evolved to laboratory-based tests, especially as
specifications co-evolved to use the results for AQC acceptance tests. Laboratories that run the
FMs for pH, resistivity, chloride, and sulfate are at present neither certified by FDOT nor
accredited through AASHTO, although a laboratory may be accredited by AASHTO to run
similar AASHTO tests. CMEC, which is a company that specializes in construction
accreditation, education, and certification, may upon request inspect a laboratory for soil
corrosion tests, but this inspection is not recognized by FDOT. Commercial laboratories that do
soil corrosion testing are typically well-known in the construction industry and have offices
convenient to district construction activities. FDOT may assist with preliminary sampling of a
potential MSE wall backfill source but neither certifies the backfill nor accepts the emplaced
material.
2.2.2.1 Process control for corrosion AQC
Sources of MSE wall backfill include borrow pits or stockpiles from construction sites,
excavations by individual land owners or non-mining businesses, commercial yards that sell
construction materials, and commercial mines. A stockpile, portion of a stockpile, or stratum in
a borrow pit is identified as candidate material for MSE wall backfill in response to a project
need. Upon request by the Construction Engineering and Inspection (CEI) personnel, a qualified
sampling technician (QST) collects a sample of the material at the stockpile or pit per AASHTO
T2 and T248. This sample is returned to a laboratory and processed and tested in accordance
with AASHTO T248 and the FMs for pH, minimum resistivity, chloride, and sulfate. The QST
may be an independent contractor or may be an FDOT employee. If testing indicates that the
material meets the requirements for MSE wall backfill, the test results are certified, that is,
signed and sealed by a Florida Professional Engineer, and the source certification is provided to
the CEI. Ideally, these test results are entered into FDOT’s Laboratory Information Management
System (LIMS). For corrosion AQC, certification is based on analysis of one sample and on test
results that are within the acceptance limits shown in Table 2-2. In FDOT’s Districts 1 and 7,
certified test results are transferable between projects if the test results are less than five years
old; in other districts, separate testing and certification may be required for each project. The
laboratory that does the testing for stockpile or borrow pit certification must be different from the
laboratory that does the product acceptance testing. Refer to Figure 2-3 for a simplified
schematic of process control for pH.
2.2.2.2 Product Acceptance for Corrosion AQC
When the certified MSE backfill is trucked from its source to the construction site, a separate on-
site stockpile is created for the material. Upon arrival of the backfill at the site a delivery ticket
is given to the contractor, and the ticket includes information about the source. An on-site CEI
may keep a sample of the backfill material by source in a clear plastic bag to compare with the
21
delivered material or to help identify soil type for testing purposes. Key construction personnel
maintain the knowledge of on-site stockpiles by source. On-site stockpiles of MSE wall backfill
are consumed typically within days to weeks. Backfill is emplaced and compacted in lifts in the
MSE wall, and water may be added during the compaction process. Water quality must meet the
same requirements as water for concrete (Scheer, 2013). Acceptance samples are taken after
backfill has been emplaced (Scheer, 2013). A sample of backfill is shoveled into a bucket or bag
from a compacted lift, usually a lower lift and early in the emplacement process. One bag
contains backfill for many tests, among which are the corrosion tests. In Districts 1 and 7,
however, a subsample is placed in a plastic bag and transferred in a cooler to the laboratory for
pH testing, a practice that is consistent with the FM for pH. Samples for corrosion and other
testing are accompanied by a completed C-22 form that describes the sampling details.
The contractor is compelled to obtain one backfill sample per soil type; this is a quality control
sample, which is coded in LIMS as sample type ‘Q’. The CEI is compelled to obtain an
independent verification sample, which is coded ‘V’ in LIMS. Both ‘Q’ and ‘V’ samples are
part of product acceptance. The ‘V’ sample is obtained at the same sampling frequency, but
ideally at a different time and location than the ‘Q’ sample. The ‘Q’ and ‘V’ samples are tested
by different laboratories. If there is a failure of either the ‘Q’ or the ‘V’ sample, a split sample of
the ‘Q’ sample is re-tested by a third laboratory as a resolution sample, which is coded ‘R’ in
LIMS. The CEI may request a FDOT laboratory to test a ‘V’ or an ‘R’ sample. Laboratories
enter the corrosion test results and data provided on the C-22 form directly into LIMS.
2.2.3 Sampling of select backfill
Sampling is a source of systematic error in corrosion test results (Equation 2-1). Intrinsic to the
material sampled and the equipment used for sampling is “sample correctness,” which Mason
(1992) describes as “when all particles in a randomly chosen sampling unit have the same
probability of being selected for inclusion in the sample.” Failure to observe sample correctness
introduces systematic error or bias in test results. This bias may be small for homogeneous soils
or strata but large for heterogeneous soils. Other sampling errors relate to the heterogeneity of
the material, fluctuations and segregation of material properties, and extraction error (Mason,
1992). Error associated with the inherent heterogeneity of a soil, for example, can only be
diminished by grinding the material to a finer particle size or testing a larger sample. Errors
associated with fluctuations and segregation of material property, for example, particle size
segregation or chemical gradients, can be reduced by combining and mixing random subsamples
to form a representative composite sample. Extraction error relates the geometry of the sampling
unit to the sampling device; to properly sample a 2-dimensional soil stratum, for example, a
cylindrical core should penetrate the entire depth of the stratum (Mason, 1992).
FDOT has developed a training and certification program to minimize sampling errors. A person
sampling soil for MSE wall backfill must be an FDOT Qualified Sampling Technician (QST),
trained in the performance of AASHTO T2 (ASTM D75), modified by FDOT FM1-T002, and T
248 (ASTM C702). QST training includes much more detail with regard to strategies for
sampling of stockpiles than is given in AASHTO T2. AASHTO T2 is a standard practice for
22
sampling aggregate and AASHTO T248 is a standard method of test for reducing samples of
aggregate to testing size. AASHTO T2 is broadly applicable for aggregate from 2.36 mm to 90
mm in diameter; its implicit focus is sampling to characterize the aggregate size gradation.
Procedures in AASHTO T2 in common with the FMs for pH, minimum resistivity, chloride, and
sulfate are
 Obtain a representative sample;

 Avoid the outer surface or disturbed surface (overburden for a pit or bank); and
 Transport the sample to preclude loss or contamination of sample or damage to contents
from mishandling.
For sampling stockpiles, pits or banks, and truck loads of coarse or coarse plus fine aggregate,
the AASHTO T2 practice include in general
 Design a sampling plan for specific case under consideration;

 Define the number of samples to represent lots and sub lots of specific sizes;
 Determine the number of field samples to give the desired confidence in test results;
 Collect a minimum sample mass is 10 kg for aggregate 2.36, 4.75, and 9.5 mm in
diameter;
 Label individual shipping containers with sample ID to facilitate sample field and
laboratory reporting; and
 Document the supply location, approximate quantity of supply, quantity and quality of
overburden, length of haul to worksite, character of the haul, and details of the extent and
location of material represented by each sample.
For stockpiles:
 Use power equipment to separate into a small sampling pile materials drawn from various
locations and levels of the main pile after which several increments may be combined to
compose the sample – or – draw several samples from separate areas of the pile; if no
power equipment are available take three increments from the top third, the middle third,
and the bottom third of the pile;
 Use sampling tubes 30-mm diameter x 2-m long; and
 Take a minimum of 5 increments from random locations.
For pits or banks:
 Identify different stratum and select samples from each stratum;

 Drill or excavate holes to look for variation in quality ;
23
 Channel the face vertically to obtain samples; and
 Mix and quarter samples to obtain at least 12 kg (of sand).
Field sampling instructions in common to FMs for pH, minimum resistivity, chloride, and sulfate
address not only sampling error but also systematic error introduced from sample processing and
from environmental conditions (Equation 2-1):
 Obtain a representative sample;

 Use clean tools;
 Avoid or reduce excessive moisture;
 Avoid the weathered or vegetated surface;
 Place a minimum of 1 kg in a plastic or plastic-lined bag;
 Minimize the contact of collected soil with air;
 Store and transport the soil in cool, dark conditions; and
 Store up to 7 days at 4C; do not freeze (pH, chloride, and sulfate methods).
2.2.4 Good laboratory practices
Once a sample reaches the laboratory, general laboratory practices are among those influences
that contribute either directly or indirectly to systematic error of test results. Good laboratory
practices (ASTM D3856) include
 A workspace that is clean, comfortable, and safe;

 Staff that have education and training appropriate for assigned tasks;
 Written procedures that provide clear direction for sample collection, handling and
storage, sample analysis, and calibration and maintenance of the measurement system;
 Supervision of laboratory staff to promote uniform and compliant performance of
procedures;
 Recordkeeping to document equipment calibration, repair, and maintenance;
 A chemical hygiene plan;
 A system for tracking sample disposition and analyses;
 Recordkeeping to document QA/QC activities; and
 Communication and troubleshooting protocols for investigating out-of-control results.
Additional quality assurance and quality control (QA/QC) measures are often part of materials
testing. These measures may be implicit—part of good laboratory practices such as checking the
calibration of an analytical balance or the conductivity of the water generated by a deionization
system—or explicit in a test method. Typical laboratory QA/QC activities are shown in Table 2-
7 (ASTM D3856; FDEP, 2009).
24
Table 2-8 Typical Laboratory QA/QC Activities
Activity Description Purpose
Calibration check Check the calibration of the analytical Assures that the analytical balance is in
(analytical balance) balance with Class S or better weight. good working order.
Verify instrumentation is working using Assures that equipment response is
traceable stock standard solutions. The linear or curvilinear as expected and
Calibration curve supplier of this standard must be establishes the relationship between the
different from the supplier for the check instrument reading and the media
standard. concentration.
Test a blank made from deionized Identifies contamination from reagent
Blank
(reagent) water. water.
Test analyte content in a check standard
Calibration (check) Assesses accuracy and precision.
prepared from a stock standard solution.
standard Analyte recovery should be within 95%
Prepare standard solution from ACS-
to 105%.
grade or NIST-traceable reagents.
Laboratory Provides an on-going check of method
Analyze sample in duplicate or triplicate.
replicates precision.
Checks for matrix interferences.
Add a known amount of analyte to
Analyte recovery should be within 85%
sample media and take media through
Matrix spike to 115%, unless historical data indicate
measurement. Determine the recovery of
tighter limits can be routinely
the spiked analyte.
maintained.
Assesses method accuracy and
precision and is used to calculate
Internal media of known analyte
Laboratory control method LOD. Serves as a method blank
concentration that is taken through the
sample to identify contamination from
entire measurement process.
reagents, glassware, and materials
handling.
External media of known analyte Assesses method accuracy and
Independent or concentration, preferably a media that precision. Identifies systematic errors.
standard reference has been tested by several laboratories, Precision and recovery should not be
material that is taken through the entire statistically different that the certified
measurement process. value.
2.2.5 Other QA/QC activities
Operator (technician) training and method oversight through independent audits are two more
QA/QC activities to serve not only reduce systematic error in test results but also to avoid costly
rework that stem from such error. Operator training has many forms: demonstrations versus
practice, one-on-one in the laboratory versus a classroom lecture, and in favor today internet-
accessible video clips, webinars, and cartoons. For potentially life-threatening situations such as
hazardous materials handling, training may be required on a recurring basis. Certification
represents a formal documentation of a training program that was successfully completed.
Independent audits of standard methods not only provide oversight on method performance, but
put the method and operator in its proper context, a context that includes environmental
25
influences (temperature, noise, dust, crowding), laboratory practices, and pressure for time. With
competent auditors, independent audits are an excellent way to maintain method performance
through time and across laboratories.
26
3 Measurement of Soil pH, Minimum Resistivity, Chloride, and Sulfate
3.1 Overview
The FMs for pH, minimum resistivity, chloride and sulfate were reviewed and compared with
AASHTO, ASTM, and USEPA standard methods (Table 3-1). Theory of measurement for each
method was also discussed to introduce interferences associated with the test methods and to
identify method steps that have large leverage in test results.
Table 3-1 Selected Test Methods for Soil Electrochemical Properties

Agency pH Min Resistivity Chloride Sulfate
FDOT FM5-550 FM5-551 FM5-552 FM5-553
AASHTO T289 T288 T291A T290B
ASTM G51 G187 D512B C1580
USEPA Method 9045D -- Method 9253 Method 9038
3.2 FM for pH
3.2.1 Theory of measurement
3.2.1.1 Acidity, activity, and ionic strength
The difference between soil acidity and pH, the basis of potentiometric measurements of pH, and
the design of a typical pH measurement system were reviewed to understand the sources and
nature of errors associated with electrochemical measurement of soil pH. To discuss soil acidity,
a broad definition of an acid is needed. An acid is defined as a substance that when dissolved in
water increases the concentrations of protons (H+, or more accurately H3O+), or that donates H+
to a chemical reaction, or that can accept an electron pair for a covalent bond (Snoeyink and
Jenkins, 1980). Soils may contain not only strong mineral acids such as hydrochloric, nitric, and
sulfuric acids, but also weak acids such as alcohols and amines, organic acids, and ferric (Fe3+) or
aluminum (Al3+) ions. Titration of a soil extract yields a higher acidity than does an
electrochemical measurement of pH, as the titration includes exchangeable H+ that are made
available as pH increases with titration, while an electrochemical measurement is sensitive to the
H+ activity at the soil pH. Soil pH is “the H+ activity in the system when the solid and liquid
phases are in equilibrium” (Peech and Bradfield, 1948), where pH = -log{H+} and the
relationship of H+ activity (unitless) to concentration is {H+} = [H+], the H+ concentration is in
molarity M (moles/liter) and the activity coefficient  is in liters/mole. Activity represents the
influence of charged ions or particles in solution on the analyte of interest. The activity
coefficient  is related in part to the ionic strength of the solution, which in turn is related to the
concentration and valence of dissolved ions. In an aqueous extract of soil where the solution
composition is unknown, solution conductivity (up to ~1,200 µS/cm) is a useful surrogate for
ionic strength. For a solution conductivity less than ~100 µS/cm (resistivity greater than 10,000
27
ohm-cm), ionic strength is less than ~0.001 and  for H+ can be approximated as 1; such a
solution is a low ionic strength solution.
3.2.1.2 pH measurement system and electrode design
At the core of the electrochemical pH measurement is the simplified Nernst equation (Equations
3-1 and 3-2), which applies to the change in free energy of a substance between two states, for
example, its current state and a standard state, where E and E0 are the current and standard state
voltages (V), T is the temperature (K), R is the gas constant (CV/moleK), n is the number of
electrons transferred in a reaction, F is Faraday’s constant (C/mole), and 2.303 a conversion
from the natural (ln) to base 10 logarithm (log) scales (Kohlmann, 2003). (In these simplified
equations, the initial state voltage E0 includes the reference electrode voltage.)
2.303𝑅𝑇
𝐸 = 𝐸0 + 𝑙𝑜𝑔{𝐻 + } (3-1)
𝑛𝐹
𝐸 = 𝐸 𝑜 − 0.059𝑝𝐻 @ 25 oC (3-2)
The electrochemical pH measurement system consists minimally of two solution half-cells that
are joined through a high impedance voltmeter and also by the sample across an H+-sensing
surface, for example, a glass bulb on the indicating cell, and a liquid junction on the reference
cell (Figure 3-1). Ideally, in this system changes in voltage E are in response only to the {H+} of
the sample, with all other system voltages Eo held constant. A silver/silver chloride reference
electrode combined in one unit with an H+-sensing glass membrane is commonly used in the
laboratory. As an example design, the reference electrode consists of a silver wire with it
terminus coated in silver chloride suspended in silver-saturated 4 M KCl. The half-cell reaction
is silver chloride (solid) + e-  silver (solid) + chloride (aqueous). The indicating electrode has
a silver wire coated with silver chloride that extends into the thin-walled glass bulb terminus,
which is filled with 0.1 M hydrochloric acid solution. The half-cell reaction is silver (solid) 
silver chloride (solid) + e-. The activity of protons adsorbed to the inside the glass bulb is fixed,
but the protons adsorbed to the outside of the glass bulb depend upon the sample {H+} or pH.
Adsorption and desorption of H+ occurs in a thin (0.001 mm) hydrated gel layer on the glass bulb
outer surface. At pH < 7.00, H+ ions move into the gel layer; at pH > 7.00, H+ ions move out of
the gel layer. A change in H+ between the inside and outside gel layers of the glass bulb creates
a differential voltage (galvanic potential), which induces electron flow. Electron flow through
the voltmeter produces a voltage drop proportional to {H+} in the sample; ion flow through the
liquid junction maintains solution electro-neutrality. The total voltage measured by the pH meter
includes the net voltages from the glass bulb interface, the electrolyte and wire of the indicating
electrode, the electrolyte and wire of the reference electrode, and diffusion potentials at
solid/solution (liquid junction) interface. Ideally, when measuring a pH 7.00 solution, these
voltages sum to zero; in practice, the voltages do not sum to zero and the voltage displacement
must be offset during calibration. The magnitude of the offset voltage (asymmetry potential)
grows over time with the deterioration of either the indicating or reference electrodes.
28
Voltmeter
Indicating electrode Reference electrode
Glass bulb Liquid junction
Sample
Figure 3-1. Schematic of a pH measurement system.
3.2.1.3 Liquid junction
A key component of the typical reference electrode design is the liquid junction. The purpose of
this junction is to allow a small but steady leakage of the electrolyte into the sample solution
during the measurement process so as to provide ion flow to maintain solution electro-neutrality.
Many combination electrodes allow the reference electrode electrolyte to be replenished over
time. Junctions are commonly made of a non-reactive porous material such as ceramic or fiber;
the number and geometry of the junctions vary with the pH measurement application. An equi-
transferent electrolyte at high ionic strength such as 4 M potassium chloride keeps low the
voltage drop across the liquid junction (Brezinski, 1985; Kadis and Lieto, 2010). Equi-
transferent for potassium chloride means that both the potassium and chloride ions diffuse at
roughly the same rates, which evenly distributes the charge through the liquid junction. The
flow rate of the electrolyte through the liquid junction dictates the responsiveness—stability and
speed—of the electrode (Davison and Woof, 1985). For routine pH measurements in low ionic
strength waters, a liquid junction designed for a higher electrolyte flow rate is recommended
(Davison and Woof, 1985; Illingworth, 1981). A double-junction electrode is better suited for
dirty samples. In a double-junction electrode the potassium chloride electrolyte that contains
silver chloride is retained by the first junction in a chamber containing the wire, and a potassium
chloride electrolyte (with trace silver chloride) is retained by the second junction in an adjoining
29
chamber, where the second junction bridges the electrolyte and the sample. This design reduces
contamination of the first junction and chamber by the sample.
3.2.1.4 Temperature error
The temperature-dependence of pH measurements has at least three sources: (1) the

electrochemical reaction as expressed by the Nernst equation, (2) the change in resistance of the
glass used in the H+-sensing glass bulb, and (3) the dissociation of ions in solution. From
Equation 3-1, pH measurement is sensitive to temperature. The contribution of this error to the
total temperature error is 0.003 pH error/pH unit/°C, where pH unit and oC are referenced
to pH 7 and 25oC, respectively, or ~0.05 pH units for a sample solution at pH 4.00 and 30 oC.
Glass resistivity decreases with increasing temperature, which changes the voltage drop across
the glass and thus the measured pH. The relationship of glass resistance to temperature depends
on the type of glass, its size, shape, and thickness, and the condition of hydrated layer. Modern
instruments automatically compensate for the shift in voltage due to a change in glass resistance
provided that the solution temperature is either sensed by the electrode or otherwise measured
and input into the pH meter. Finally, the dissociation of ions in solution is temperature-
dependent. For example, in pure water at 25 oC, 10-7 moles/liter of H2O disassociates into H+
and OH-; more water dissociates at higher temperatures and less at lower temperatures; thus, at
20oC, pH 7.08 is neutral and pH 7.00 is acidic (Table 3-2) (Skoog et al., 1996). The temperature
of the sample affects the H+ concentration on both sides of the glass bulb, but the net effect will
depend on sample composition. The temperature effects on pH measurements highlight the
importance of measuring calibration buffer and samples at the same temperature and at or near
25 oC.
Table 3-2 Influence of Temperature on the pH of Pure Water

Temperature (oC) pH
10 7.27
20 7.08
25 7.00
30 6.92
40 6.77
3.2.1.5 Sodium (alkaline) and acid errors
Sodium ions if present in the sample at much greater concentrations than H+ ions can penetrate
into the glass and displace H+ ions with a commensurate shift the charge balance. The actual
conditions under which the shift occurs depends upon electrode design, temperature, and
precedent exposure, but may be significant at pH >11. Under extremely acidic conditions (pH <
1), the high acid content relative to water content diminishes the gel layer thickness; the
consequence is a higher measured versus actual pH.
30
3.2.1.6 Soil-to-water ratios
Soil-to-water ratios used for pH measurement vary widely, for example, from a saturated paste to
dilutions of 1:5; a 1:1 soil-to-water ratio using distilled or deionized water is typical (Miller and
Kissel, 2010). Miller and Kissel (2010) found that pH increased with water dilution. For their
dataset of North American soils, the non-linear increase was from pH 4.50 for a saturated paste
to pH 4.87 for a 1:1 soil-to-water dilution to pH 5.62 for a 1:5 soil-to-water dilution. Puri and
Asghar (1938) proposed that salt present in soils depressed the soil pH. Their experimental work
was with soils leached free of base cations, to which select base cations were added at soil-to-
water dilutions of 1:5 to 1:25. They found that in the absence of base cations, the soil pH did not
change with water dilution; in the presence of even small amounts of added neutral salts,
however, the pH decreased. Conversely, leached soils had a higher pH than their unleached
counterparts, but leached and unleached soils extracted in potassium chloride had the same pH.
Their explanations for these phenomena were that (1) the presence of a salt reduced the
hydrolysis of an exchangeable base, and (2) surface ionization of colloidal particles was not
affected by dilution. In other words, adding water to the soil diluted the soil’s salt content, which
raised the pH. Puri and Asghar (1938) recommended that soils be extracted in a potassium
chloride solution to assure uniformity in pH results.
3.2.1.7 Stirring error
A difference in pH is sometimes observed between a stirred and quiescent solution, and this
difference is termed a stirring error. Brezinski (1983) reported that the root cause of the stirring
error was a change in the electrode electrolyte concentration (e.g., 4 M potassium chloride) at the
interface of the liquid junction and the sample. Errors were minimal for a well-designed
electrode when the sample was similar in ionic strength and transference to the electrolyte, for
example, with standard buffers used to calibrate pH electrodes. He postulated that for dissimilar
solutions, mixing of sample solution and electrolyte solution occurred within the liquid junction
and caused a shift in voltage drop. Moreover, between two dissimilar samples the liquid junction
“remembered” the last sample (consistent with a trace of the last sample retained in the liquid
junction), and this effect—while transient—significantly slowed the electrode response. For
example, Brezinski (1983) found that transferring an electrode from a 1 M potassium chloride to
a 10-4 M hydrochloric acid sample resulted in a relatively stable pH reading that was -0.6 pH
units in error. Stirring effects were exacerbated for no-flow (non-refillable gel electrodes) or
low-flow (i.e., clogged) electrodes, and for low-ionic strength samples. Davison and Woof
(1985) recommended monitoring the stirring error as a measure of electrode performance and
stated that for low ionic strength solutions, pH measured without stirring were closest to the
expected pH.
3.2.1.8 Ionic strength
An electrochemical pH measurement is a challenge in low ionic strength solutions. In pure water

that is equilibrated with carbon dioxide at atmospheric concentrations, the addition of a neutral
salt such as 1 M potassium chloride can shift the pH downward by ~0.10 pH units due to the
31
dissimilarities in ionic strength and transference between the electrolyte and sample (Bier, 2009;
Brezinksi, 1983; Kadis and Leito, 2010). Illingworth (1981) reported a 0.2 pH unit error for each
10-fold difference in salt concentration between the calibration buffer and an aqueous sample.
Increasing the ionic strength of the pure water sample improves the stability and speed of the pH
measurement (Bier, 2009). In 1:1 soil-to-water extracts of soil, however, the addition of a
neutral salt such as 1 M potassium chloride or 0.01 M calcium chloride causes a shift in pH of
~0.5 pH units (Miller and Kissel, 2010) to 2 pH units (Peech and Bradfield, 1948; Yuan, 1963),
and explanations for this shift were more complicated. In the agricultural sector, amendment of
samples with 0.01 M calcium chloride is often recommended not only to reduce the sensitivity of
pH measurement to salt concentration in low ionic strength solutions but also to better estimate
lime requirements in the presence of a seasonal change in soil pH measurements due to rainwater
leaching of salts of agricultural soils (Kissel et al., 2009).
Statistical models have been developed to relate the pH of 1:1 soil-to-water extracts with that of
1:1 soil-to-0.01 M calcium chloride extracts, as these pH values are strongly correlated (Brennan
and Bolland, 1998; Miller and Kissel, 2010; Minasny et al., 2011). The observed downward
shift in pH with the addition of neutral salts to soil extracts is due to a combination of factors that
include the displacement by added metal cations (potassium, calcium) of (1) exchangeable H+
and (2) exchangeable aluminum, iron, and manganese, which may be subsequently hydrolyzed
and act as H+ sources, and (3) the presence of a voltage bias at the liquid junction due to the
difference ionic strength between the electrode electrolyte and soil extract (Miller and Kissel,
2010; Peech and Bradfield, 1948;Yuan, 1963). Miller and Kissel (2010) argue that because the
downward shift in pH is relatively constant for both acid and alkaline soils, the voltage bias at
the liquid junction dominates the effect. Miller and Kissel (2010) made an argument for
extracting a soil with a neutral salt solution but Busenberg and Plummer (1987) argued the
opposite because of the potential for pH-altering contamination in purchased salts.
3.2.1.9 Carbonate Equilibria
Measurement systems for pH can be open or closed; closed systems are more often found in
process monitoring than on a laboratory bench. In an open system such as a 1:1 soil-to-water
suspension in an open beaker, solution and atmospheric gases will exchange and trend to
equilibrium at the solution pH, temperature, ionic strength, and with stirring. Carbon dioxide is
present in the atmosphere at an average global concentration approaching 400 ppm; local
atmospheres (urban and laboratory as examples) and in situ soils may see higher carbon dioxide
concentrations (Macpherson, 2009; Rice and Bostrom, 2011). At a room temperature of 25 oC, a
stirred beaker of pure water tends towards a carbonate equilibrium of pH 5.6 or an {H+} of 2.5 x
10-6 M. The interference of carbon dioxide on pH measurements on soil pH will be the greatest
for a near-neutral, low ionic strength soil with little buffering capacity (Figure 3-2). In this case,
similar to pure water, the pH electrode response will be slow, noisy, and unstable, and with
significant stirring effects.
32
Figure 3-2. Calculated influence of carbon dioxide on an open -system pH
measurement of a weak sulfuric acid solution with and without KCl added.
3.2.1.10 Suspension error
Suspended colloidal particles—particles in the size range of 0.001 to 0.10 µm—interfere with
electrochemical pH measurements (Keller and Matlack, 1990; Yang et al., 1989). Literature
defined the suspension effect in two ways: (1) the change in measured pH with the addition of
suspended particles, and (2) the change in pH for indicating and reference electrodes immersed
in the sediment suspension versus both electrodes immersed in a solution equilibrated with the
sediment. Colloidal particles are difficult to filter or settle and if present in the soil will be
suspended in the soil extract. Keller and Matlack (1990) wrote that pH paper worked better than
electrochemical measurement for determining pH in clay soils, even after the soil extract was
filtered through a 0.10 µm filter to produce an optically-clear solution. Yang (1989) and Oman
(2000) agreed that the suspension effect is the combined response of the indicating and reference
electrodes to particles suspended in the sample. In general, the charge fields of indicating
electrode and particles proximate to or adsorbed on the indicating electrode interact to alter the
33
electrode voltage. Moreover, the electrolyte effusing from the reference electrode liquid junction
interacts with suspended particles to create an ‘anomalous liquid junction potential.’
3.2.1.11 Damage to the electrode
A pH electrode is a fragile and sensitive instrument and is easily damaged. Rough handling or
abrasive samples can break or crack the glass bulb. An unresponsive reading, usually a pH 7.00
or 0 mV reading in all buffers and samples, is a symptom of a broken glass bulb or wire. This
may go undiagnosed if the electrode is calibrated with only a pH 7.00 buffer. Colloidal particles
suspended in soil extracts can migrate into and clog the liquid junction as can silver chloride or
insoluble silver precipitates. A slow electrode response or steep stirring potentials indicate a
clogged liquid junction. Colloidal particles or dissolved sample constituents can migrate through
the liquid junction and into the electrolyte when the height of the sample is above the height of
electrolyte, as evidenced by a change in the electrolyte opacity or color or electrode performance.
A dry glass bulb affects electrode response. Proper hydration of the glass bulb is necessary for a
stable reading; storage in distilled or deionized water can deplete the hydrated gel layer. Oil,
enzymes, or particles adhered to the glass bulb affect electrode performance but can be removed
with an appropriate cleaner. Many but not all of these adverse effects can be reversed with
proper electrode maintenance.
3.2.2 Description of method
FDOT specifies FM 5-550 for testing the pH of soil in candidate MSE wall backfill material
(Chapter 2, Section 2.1.2). As presented in Section 3.2.1, the basis of this measurement is a
potentiometric measurement using an indicating electrode and a reference electrode, where the
voltage produced in this system is linear with pH. The voltage response to pH is calibrated with
buffers of known pH at the measured sample temperature.
The first step towards analysis of pH in soil is preparation of the soil sample. Per the FM, these
steps are:
 Place 100 mL of soil sample in a beaker and add an equal volume of distilled or
deionized (dilution) water;
 Break any soil clumps and stir the mixture at 10 min intervals for 30 min;
 Allow sample to reach room temperature;
 Calibrate the pH meter and electrode every 30 min or if a difference of one pH unit is
encountered between tests; and
 Calibrate with a standard buffered solution in the range of the sample to be tested or with
a pH 7 standard buffer if the sample pH is not known.
34
The steps to measure pH in soil are given by either the manufacturer’s instructions or by
Standard Methods for the Examination of Water and Wastewater (SMEWW) 4500-H+ B (Rice et
al., 2012). Typical steps are:
 Set meter mode to pH;

 Rinse the pH electrode with distilled water;
 Dip the pH electrode and temperature sensor into the sample;
 Allow time for the reading to stabilize; and
 Record the pH reading.
SMEWW 4500-H+ B recommends that the pH meter be re-calibrated if the buffer solution pH
after calibration measures more than 0.1 pH unit from its expected value, which may indicate a
faulty electrode.
3.2.3 Method comparisons
3.2.3.1 Soil processing
The FM for pH requires that the soil be transported in a cool dark area and if stored, stored at
4oC for no more than seven days. Soil is tested in the “as received” condition. Per AASHTO
T289 soil is selected according to AASHTO T248, dried to a moist condition either in air or in a
drying apparatus not to exceed 60oC (which AASHTO T289 states is equivalent to air drying),
and sieved with a 2.00-mm (No. 10) sieve. ASTM G51 is for in situ field measurements of pH.
Calibration of the pH meter is generally consistent with other methods discussed herein;
however, the pH electrode is lowered into the surface or subsurface through an augured hole to
make direct, undiluted contact with the soil (with the caveat that some soils may be too dry).
The pH electrode can be installed in a protective casing for ease of handling.
3.2.3.2 pH meter and electrode calibration
For pH meter calibration, the FM calls for use of at least one standard buffer solution in the
range of the pH to be tested or pH 7 if the range is unknown; AASHTO T289 and USEPA
9045D use at least two buffer solutions that are 3 pH units apart and bracket the sample pH. Re-
calibration is required after every 30 min or if a difference in one pH unit is encountered; both of
these methods require temperature compensation, but with USEPA 9045D, after calibration the
measured buffer pH should be within 0.05 pH units of the stated value.
3.2.3.3 Soil testing
The FM for pH calls for 100 mL of “as received” soil, an amount that for moist sand typically
weighs from 150 to 170 g, to which is added 100 mL of dilution water. AASHTO T289
specifies
35
30 g of soil and to which is added an equal mass of dilution water. With the FM for pH, soil and
water are equilibrated for 30 min with stirring at 10-min intervals; for AASHTO T289 soil and
water are equilibrated for 60 min with stirring at 10-min intervals. For USEPA 9045D, soil with
less than 20% moisture is tested; 20 mL of dilution water is added to 20 g of soil, and the slurry
is continuously stirred for 5 min and allowed to settle for 1 hr. The water portion of the slurry is
then decanted, filtered, or centrifuged to separate the water from clay particles before its pH is
measured. In short, for these three methods the soil is either “as received” or dried and sieved,
with soil-to-water ratios between 1:1 and ~1.5, sample sizes from 20 to 150 g, and sample
extracts either filtered or unfiltered. SMEWW 4500-H+ B recommends that for dilute, poorly
buffered samples, the pH electrode be equilibrated with three to four successive portions of the
sample. SMEWW 4500-H+ B and USEPA 9045D both mention that a sodium error is likely at
pH >10, in which case a low-sodium error electrode should be used. Another variation on soil
pH measurements is to use a 0.01 M calcium chloride (Miller and Kissel, 2010; ASTM D4972)
solution in lieu of distilled or deionized water. According to ASTM D4972, the added calcium
displaces exchangeable aluminum in the soil and the elevated salt content better represents the
in-situ soil ionic strength.
3.2.4 Estimate of method precision and bias
According to Fisher Scientific (2012), for an AB150 benchtop meter, the pH range is -2.000 to
20.000 pH, with a resolution of 0.001 pH units and an accuracy of ± 0.002 + 1 LSD pH units
(LSD is not defined in the manual). For SMEWW 4500-H+ B method precision and accuracy
are ±0.02 pH and ±0.1 pH, respectively, under normal conditions for water and poorly buffered
solutions (Rice et al., 2012).
3.3 Minimum Resistivity
Soils with a higher concentration of salts are more corrosive. A measure of a soil’s salt content
is resistivity, where salt content is inversely proportional to resistivity. Soil resistivity
measurements can be made in-situ or on a laboratory bench and while the latter approach of
interest here, equipment design (Section 3.3.2) likely evolved from field techniques. Ohm’s Law
relates resistance R to voltage V and current I, where R = V/I in compatible units. In concept,
impressing I across a block of soil generates a V, which is measured and from which R is
calculated. Resistance R is a property of the soil block; to express as a property of the soil, R in
ohms is converted to resistivity  in ohm-cm according to the geometry of the block and
Equation 3-3 (Samouëlian et al., 2005):
𝐴𝑟𝑒𝑎
𝜌 = 𝑅 𝐿𝑒𝑛𝑔𝑡ℎ (3-3)
Resistance of a soil to its conduct of ions depends upon its liquids, solids, and voids (Samouëlian
et al., 2005). Water content, amount and type of dissolved ions, and the presence of other fluids
36
such as oil affect soil resistivity. For example, as water content in soil is increased from near
zero to 15%, a soil’s resistivity rapidly decreases, as illustrated in Figure 3-3 for sands from
Mines 5 through 8 (Chapter 1, Section 1.2.7, Table 1-1). Resistivity is also affected by a soil’s
mineral composition and particle size distribution and tends to decrease with an increase in clay
content, as clay particles have a high surface charge density. Finally, the number, size, and
connectivity of voids in a soil’s matrix affect the movement of ions through a soil and thus its
resistivity. These same soil constituents—liquids, solids, and voids—have insulating or
dielectric properties that respond to the frequency of the applied current (White and Zegelin,
1995). The ability to differentiate the earth’s subsurface features by varying the frequency of
applied current has led to advances in electrical prospecting, environmental surveys, and water
content mapping (White and Zegelin, 1995).
Figure 3-3. Resistivity versus water content for sands from mines 5 through 8.
Ion motion increases with temperature, thus soil resistivity decreases as temperature increases.
ASTM G187 requires that laboratory measurements be made at or near 20C, but offers Equation
3-4 to correct resistance at field temperature RT to resistance at a standard temperature of 15.5C
R15.5, where T is the field temperature in C.
37
𝑅15.5 = 𝑅𝑇 (24.5 + 𝑇)/40 (3-4)
For similar soil resistivity meters (Miller 400A, Nilsson 400, and Tinker and Rasor SR-2, as
examples), an alternating current at ~97 Hz is applied to pins C1 and C2 of a four-pin soil box
and the voltage drop is measured between pins P1 and P2 (Figure 3-4). A two-pin measurement
is possible with the voltage drop measured between pins C1 and C2; the four-pin configuration,
however, eliminates the problem of contact resistance at the electrodes (White and Zegelin,
1995). The soil resistance in ohms is read from the meter and multiplied by the geometry-
dependent distance (cm) between P1 and P2 to get resistivity (ohm-cm). The FM for minimum
resistivity calls for a soil box area-to-length ratio of 1 cm. AASHTO T288 and ASTM G157 both
use a two-pin measurement and provide guidance for construction of soil box and calculation of
the area-to-length ratio based on the box geometry.
Figure 3-4. Diagram of soil box for measurement of soil resistivity.
FDOT specifies FM 5-551 for testing the minimum resistivity of soil in candidate MSE wall
backfill material (Chapter 2, Section 2.1.2). As presented in Section 3.3.1, soil resistivity is
based on a measured voltage drop due to an impressed current across wetted soil that is
contained in a box of standard dimensions.
According to the FM for minimum resistivity, the method steps are:
 Rinse the soil box with deionized water before and after completion of each test;
 Follow the manufacturer’s instructions for connecting the leads to the soil box;
 Place one kg of as-received soil sample in a large round pan; remove debris and mix soil
sample;
38
 Fill the soil box, compact, and level; connect the soil box to the meter and determine the
resistance; place the soil sample back into the pan;
 Record the volume of water used and the corresponding resistance reading;
 Add 100 mL of distilled/deionized water to the soil sample, mix thoroughly, and repeat
the above procedure until the lowest resistance is obtained; and
 Report results of minimum resistivity with units of ohm-cm.
3.3.3 Methods comparisons
The FM for minimum resistivity does not require that the soil be dried or sieved, nor are soil
storage instructions provided. AASHTO T288 requires that the soil sample is dried in air or in a
drying apparatus not to exceed 60C, and that the sample is selected per AASHTO T248 and
sieved with a 2.00-mm (No. 10) sieve, such that 1,500 g is available for the test. Moreover, 150
mL of dilution water is added to the 1,500-g sample of dried soil and the wetted sample is
covered and allowed to equilibrate or cured for a minimum of 12 hr. Dilution water must have a
resistivity greater than 20,000 ohm-cm. ASTM G187 does not require either drying or sieving of
the soil sample, but requires laboratory testing at room temperature and cautions against field-
testing soils that are frozen or partially frozen, and gives an equation to convert resistances
measured in the field to a standard temperature of 15.5C (Equation 3-4). Dilution water must
have a resistivity greater than 200,000 ohm-cm.
3.3.3.2 Method calibration
Calibration of the soil resistivity meter is established by checking the zero of the meter, verifying
the resistance of resistors across a range of known values, and measuring the resistivity of
standard salt solutions. The FM for minimum resistivity does not include a calibration step, but
AASHTO T288 requires calibration of the soil resistivity meter with 100-, 200-, 500- and 900-
ohm resistors; the meter readings must be within 10% of the resistor value. ASTM G187
recommends (1) periodically calibrating the soil resistivity meter with a commercial resistance
decade box or with a series of resistors of different values; the meter errors must be within 5%
over the range of the instrument, and (2) twice a year checking the meter and box with standard
solutions in the range of 1,000, 5,000, and 10,000 ohm-cm.
Steps for soil testing in the FM for minimum resistivity are nearly identical to those for
AASHTO T288, with one important exception. For AASHTO T288, as water is added and soil
slurry develops, the method operator is directed to fill the soil box with slurry water first, and if
the water cannot fill the box to the top, to continue to fill the box with slurry soil. Thus, the
minimum resistivity may occur when the box is completely full of slurry water. ASTM G187
39
measures soil resistivity when the soil is just saturated with water. Soil is placed in the box layer
by layer, and each layer is wetted and then densely compacted by hand. When the box is full and
leveled, a thin sheen of water is present on the soil surface. The resistivity of soil at water
saturation is not the same as minimum resistivity.
The resistance meter’s range is typically 0.01 ohm to 1,100,000 ohm, and the accuracy of the
measurement is ± 10% at full scale and ± 5% at 85% of full scale (Corrpro, 2013; MCM, 2011;
Tinker and Rasor, 2013). ASTM G187 reported the results for a 2003 inter-laboratory study in
Tampa, Florida: across three soils the reproducibility was 10.6%; therefore at the 95%
confidence level, results of two properly-conducted tests in different laboratories on the same
material were not expected to differ by more than 29.7%.
3.4 Chloride
The basis of the wet chemistry chloride test is the titration of dissolved chloride with silver
nitrate to form relatively insoluble silver chloride in neutral or slightly alkaline solution. The
titration endpoint is indicated by the reaction of potassium chromate with unreacted silver to
form red silver chromate. According to Rice et al. (2012), sulfide, thiosulfate, and sulfite ions
can interfere but can be oxidized by treatment with hydrogen peroxide. Elevated levels of
orthophosphate or iron may also interfere. Bromide, iodide, and cyanide titrate as chloride.
With soil extracts, the presence of color and suspended particles interfere with diagnosis of the
titration endpoint. The measurement range is 5 to 250 ppm chloride and stated range varies with
method, or more specifically, with the volume of sample (mg of chloride) and concentration of
silver nitrate. The advantages of this test are its relative simplicity, portability, speed, and low
cost; its disadvantages are interferences from suspended soil particles and the use of hazardous
materials including silver nitrate and potassium chromate. Other methods for water-soluble
chloride include titration with mercuric nitrate, detection of chloride ions with an ion-specific
electrode, and ion chromatography, in order of increasing equipment cost and complexity.
FDOT specifies FM 5-552 for testing water-soluble chloride in candidate MSE wall backfill
(Chapter 2). The sample analysis steps in the FM for chloride are determined either by Hach
method 8207 (Hach, 2003) or by Standard Methods for the Examination of Water and
Wastewater (SMEWW) 4500-Cl- B, argentometric method (Rice et al., 2012).
40
The first step towards analysis of chloride is preparation of the soil sample. Per the FM, these
steps are:
 Dry sample to constant mass for ~four hr in an oven at 110C;

 Sieve through a No. 10 (2.0 mm) mesh sieve; and
 Split the sample per AASHTO T248 (ASTM C702) to obtain ~400 g.
The second step towards analysis is an aqueous extraction of chloride from soil. Per the FM,
these steps are:
 Combine in a 500-mL flask 300 mL of water with 100 g of dried soil;

 Shake mixture for 20 s, then allow to settle for one hr and shake for another 20 s;
 Settle the soil suspension for a minimum of 12 hr;
 Gravity filter through a fast, coarse filter the soil suspension to produce a clear extract;
and
 Allow the test sample to reach room temperature.
The sample analysis steps for Hach 8207 are as follows:
 Select the appropriate silver nitrate cartridge (concentration);

 Insert a clean delivery tube into the titration cartridge and attach the cartridge to the
digital titrator;
 Hold the digital titrator with the cartridge tip pointing up, turn the delivery knob until a
few drops of titrant are expelled, then reset the counter to zero;
 Use a graduated cylinder or pipet to measure sample volume and bring volume up to 100
mL;
 Transfer sample to a clean 250-mL flask;
 Add the contents of one Chloride 2 (potassium dichromate and sodium bicarbonate)
indicator powder pillow and swirl to mix;
 Place delivery tube tip into the sample solution and swirl the flask while titrating with
silver nitrate from a yellow to a red-brown color;
 Record the number of digits required;
 Calculate the soil chloride concentration from the number of digits, multiplication factor,
and dilution factor; and
 Report chloride results in ppm.
SMEWW 4500-Cl- B is very similar to Hach 8207. Both of these methods note that samples
should be adjusted to a near neutral pH with sulfuric acid or sodium hydroxide and treated with
41
hydrogen peroxide to remove sulfite interference; however, SMEWW 4500-Cl- B recommends
treating the sample with an aluminum hydroxide suspension if sample is highly colored.
ASTM D512B and USEPA 9253 methods assume water samples and thus give no guidance on
soil processing. For AASHTO T291A, soil samples are dried in air or in a drying apparatus not
to exceed 60oC, soil is sieved through a No. 10 (2 mm) mesh, and a 250-mg sample is selected
based on AASHTO T248. With the lower drying temperature, some moisture is retained in the
soil. AASHTO T291A has as a final step adjustment of the chloride concentration to the
moisture content of a soil subsample dried to 110C. For both the FM for chloride and
AASHTO T291A, 100 g of soil is diluted with 300 mL of reagent water, shaken to mix, allowed
to settle for one hr, and then shaken to mix once more. Reagent water is ASTM 1193 Type III,
which has a minimum resistivity of 4,000,000 ohm-cm and a maximum chloride concentration of
0.010 ppm. The AASHTO T291A method separates without delay the suspended soil particles
by centrifugation (50 mL centrifuge tubes), followed by filtration through a 0.45-µm filter if
turbidity remains.
3.4.3.2 Method Calibration
For an accuracy check, Hach method 8207 recommends titrating increments of standard solution
added directly to the sample; a linear response indicates no interference. Both SMEWW 4500-Cl-
B and AASHTO T291A include steps to standardize the silver nitrate titrant against a sodium
chloride standard and to establish the reagent blank value, while ASTM D512B tests two levels
of the same sample to screen for interferences.
USEPA 9253 and ASTM D512B are nearly identical. AASHTO T291 and ASTM D512B are
similar to Hach method 8207 and SMEWW 4500-Cl- B, but ASTM D512B has the below-listed
key variations:
 Dilution water is ASTM D1193 Type IV, with a minimum resistivity of 200,000 ohm-cm
and a maximum chloride concentration of 0.050 ppm;
 Use of a sample size less than 50 mL, diluted to 50 mL;
 Adjustment of the sample pH to the phenolphthalein endpoint of 8.3;
 Titration of the sample in a white porcelain dish;
 Sample testing under yellow light or with yellow goggles; and
 Repetition of the analysis with one-half the sample size to screen for interferences.
42
ASTM D512B reported the precision and bias results from an inter-laboratory study. Precision
and bias were concentration-dependent. The reproducibility standard deviation was sR = 0.013X
+0.70, where X is the chloride concentration; for an 80 ppm chloride concentration, sR was 1.74
ppm; therefore at the 95% confidence level, results of two properly-conducted tests in different
laboratories on the same material were not expected to differ by more than 4.9 ppm. For an 80
ppm chloride concentration, the method bias was -1.13 ppm.
3.5 Sulfate
FDOT specifies FM 5-553 for testing water-soluble sulfate in candidate MSE wall backfill
material (Chapter 2, Section 2.1.2). The basis of the test is a reaction of barium chloride with
dissolved sulfate to form relatively insoluble barium sulfate. The reaction is allowed to proceed
for a few minutes, after which time the visible light absorption of barium sulfate precipitate is
read with a photometer at ~450 nm and over a path length of 2 cm. The test range is 2 to 70
ppm, and minimum detection level is ~2 ppm. Interferences include sulfite and sulfides, which
oxidize to sulfate, suspended particles, and color. The advantages of this test are its relative
simplicity, portability, speed, and low cost; its disadvantages are interferences from suspended
soil particles and use of a hazardous material barium chloride. Other methods for water-soluble
sulfate include gravimetric, colorimetric, and ion chromatographic methods, in order of
increasing equipment cost and complexity. The gravimetric and colorimetric methods also use
barium chloride as a reagent. Hayes (2007) completed a comparison of four methods for water-
soluble sulfate from soil extracts and recommended the ASTM C1580 gravimetric method over
the ASTM C1580 turbidimetric method due to instrument drift. Hayes (2007) used a 2005
version of ASTM C1580; the 2009 version, however, does not include a gravimetric method.
The sample analysis steps in the FM for sulfate are determined either by Hach method 8051
(Hach, 2012) or by Standard Methods for the Examination of Water and Wastewater (SMEWW)
4500-SO42- E argentometric method (Rice et al., 2012). The first step towards analysis of sulfate
in soil is preparation of the soil sample. Per the FM, these steps are:
 Dry sample to constant mass ~four hr in an oven at 110C;

 Sieve through a No. 10 (2.0 mm) mesh sieve; and
 Split the sample per AASHTO T248 (ASTM C702) to obtain ~400 g.
The second step towards analysis of sulfate is an aqueous extraction of sulfate from soil. Per the
FM, these steps are:
43
 Combine in a 500-mL flask 300 mL of water with 100 g of dried soil;
 Shake mixture for 20 s, then allow to settle for one hr and shake for another 20 s;
 Settle the soil suspension for a minimum of 12 hr:
 Gravity filter the soil suspension through a fast, coarse filter to produce a clear extract;
 Add hydrochloric acid to clarify the extract, if cloudy; and
 Allow the test sample to reach room temperature.
Both Hach Pocket Colorimeter II Analysis System for Sulfate (Hach sulfate) and Standard
Methods for the Examination of Water and Wastewater (SMEWW) Method 4500-SO4= E (Rice
et al., 2012) are acceptable for the FM for sulfate. The procedure for Hach sulfate is:
 Fill a cell with 10 mL of filtered soil extract (sample);

 Add SulfaVer 4 reagent (powder of barium chloride and citric acid) to cell;
 Swirl vigorously to dissolve powder;
 React barium chloride in sample for 5 min;
 Prepare a blank containing sample but no SulfaVer 4 reagent;
 At the end of the 5-min reaction period, insert the blank in the photometer and zero the
meter;
 Insert sample cell into photometer and read sulfate concentration in ppm (calibration
curve is programmed into the photometer);
 Multiply the photometer value by the dilution factor to get the soil sulfate concentration;
and
 Report the sulfate results in ppm.
SMEWW 4500-SO4= E takes a different approach:
 Measure 100 mL of sample into a 250-mL flask;

 Add 20 mL of buffer solution and mix in a stirring apparatus at constant speed;
 Add a spoonful of barium chloride crystals and continue to stir for 60 ± 2 s;
 Pour solution into a reaction cell and measure light absorption at 5 ± 0.5 min;
 Subtract from this the light absorption for a sample blank containing no barium chloride;
 Determine the sulfate concentration from the calibration curve;
 Multiply the sample sulfate concentration by the dilution factor to get the soil sulfate
concentration; and
 Report the sulfate results in ppm.
44
The buffer solution contains distilled water, magnesium chloride, sodium acetate, potassium
nitrate, and acetic acid, and is intended to stabilize the barium chloride suspension.
AASHTO T290B, ASTM C1580, and USEPA 9038 have the same basis as the FM for sulfate;
USEPA 9038, however, assumes a water sample. In AASHTO T290B, the soil preparation
procedure calls for drying the sample in air or in a drying apparatus not to exceed 60C, selection
of 250 g soil by AASHTO T248, and sieving through a No. 10 (2 mm) mesh. With the lower
drying temperature some moisture is retained in the soil, and AASHTO T290B has as a final step
adjustment of the sulfate concentration to the moisture content of a soil subsample dried to
110C. ASTM C1580 requires drying 100 g of the sample at 60C but for 18-24 hr, and
crushing and sieving the sample through a No. 30 (0.60 mm) mesh.
AASHTO T290B mixes 300 mL of water and 100 g of soil but separates the suspended soil
particles without delay by centrifugation, followed by filtration through a 0.45-µm filter if
turbidity remains. No details are given on the centrifugation volume or speed. As with the FM
for sulfate, a drop of concentrated acid is added as needed to precipitate any remaining finely
divided suspended matter after centrifugation and filtration. ASTM C1580 mixes 3 g and 30 g of
the soil each with 250 mL of dilution water; both samples are stirred for an hr and then filtered,
twice if needed, to get a clear extract. Dilution water is ASTM D1193 Type I (18,000,000 ohm-
cm).
3.5.3.2 Method calibration
Two sulfate calibration curves in the range of 0 to 70 ppm are pre-programmed in the photometer
for the Hach sulfate method. One curve has a steeper slope than the other; the choice of curves
may be needed to adjust for either the age or the manufacturer of barium chloride reagent, as is
suggested by Figure 3-5. SMEWW 4500-SO4= E and ASTM C1580 are similar to the Hach
method except for these methods standards are prepared at 5 ppm intervals from 0 to 40 ppm, as
barium sulfate suspensions above 40 ppm are unstable. SMEWW 4500-SO4= E recommends
checking the reliability of the curve by running a standard after every 3 or 4 samples. The
AASHTO T290B method requires a standard curve prepared at concentrations of 0.0, 4.0, 10.0,
20.0, 30.0, 40.0, 60.0, 80.0, and 100.0 ppm; both AASHTO T290B and ASTM C1580 state that
a separate calibration curve must be prepared for each photometer and if any change is made to
the photometer or its cell, lamp, or filter, or to the reagent. The curve is checked with each series
of sample tests by running two or more solutions of known sulfate concentrations. USEPA 9038
requires that a calibration curve with a minimum of one blank and three standards be prepared
every hr; that the calibration curve be checked with an independently prepared standard every 15
samples, and that a duplicate spike sample be run every 10 samples.
45
Figure 3-5. Hach Pocket Colorimeter II Analysis System for Sulfate pre -
programmed sulfate calibration curves by range selection and reagent lot.
The Hach sulfate method uses pre-packaged barium chloride and citric acid, which are added
directly to sample in the test cell; all other mentioned methods adjust the pH sample into the acid
range and stir barium chloride crystals with an aliquot of sample and then decant the sample into
a test cell for an absorption reading. AASHTO T290B is similar to SMEWW 4500-SO4= E, with
up to 50 mL of sample volume and a 4-cm instead of a 2-cm path length. Aliquots of a distilled
water/glycerin solution and a sodium chloride/hydrochloric acid solution are added to the sample
as a conditioner. Likewise, ASTM C1580 and USEPA 9038 are similar to SMEWW 4500-SO4=
E, except that in ASTM C1580 for each sample, four subsamples are prepared, each with
different dilutions, with the expectation that these subsamples will bracket the sulfate
concentration. Both ASTM C1580 and USEPA 9038 add to the sample a conditioner made of
distilled water, hydrochloric acid, ethanol, sodium chloride, and glycerol.
46
For a multi-laboratory study the USEPA 9083 reported for a mixed ion water sample with 259
ppm sulfate a multi-laboratory %RSD and %RE of 9.1% and 1.2%, respectively. Multi-
laboratory precision for ASTM C1580 was 21.2%; therefore, at the 95% confidence level, results
of two properly-conducted tests in different laboratories on the same material were not expected
to differ by more than 60%.
3.6 Summary
For AASHTO, ASTM, and USEPA methods that in many ways compare closely with the FMs
for corrosion AQC, many differences exist in the procedural details. Differences that merit
further attention include but are not limited to
For pH:
 in-situ, on-site, as-received, and air-dried pH;

 Transport temperature;
 Soil volume and water-to-soil ratio;
 Equilibration time and technique;
 Measurement temperature; and
 Use of 1 M potassium chloride or 0.01 M calcium chloride for dilution water.
For minimum resistivity:
 Soil drying, sieving, and water equilibration;

 Initial soil volume;
 Water quality;
 Meter calibration; and
 Soil versus water fraction of slurry in soil box.
For chloride:
 Drying time and temperature;

 Extraction time and method;
 Technique(s) to clarify sample;
 Use of check standards;
 Enhancement of the titration end point color change; and
 A check for significant interferences.
47
For sulfate:
 Drying time and temperature;

 Extraction time and method;
 Technique(s) to clarify sample;
 Calibration curve intervals and range;
 Use of check standards; and
 A check for significant interferences.
48
4 Trends in Select Backfill
Trends in backfill properties were investigated to (1) identify sources and source regions of
acceptable MSE wall backfill with FDOT’s Districts 1 and 7, and for these districts (2) review
databased results of MSE wall backfill (material 092L) for failure trends in pH, minimum
resistivity, chloride, and sulfate acceptance tests.
4.1 Spatial Trends
4.1.1 Methods
FDOT’s District 1 and District 7 Materials Office maintains a local database to document by
county the preliminary sampling conducted on potential MSE wall backfill sources in support of
district construction projects. This database includes fields for the project identification number,
project transfer and date if a source was sampled for one project and test results transferred to
another project, stockpile or borrow pit number, survey date, test results for stratum sampled, the
fraction of soil passing a number 200 sieve (200%), % organic matter (%OM), liquid limit (LL),
plasticity index (PI), AASHTO soil group, plus the address and geo-coordinates of material
source and remarks. The remarks field often contains the mine name and/or number, additional
location information, and whether or not the material passed for embankment or MSE wall
backfill. Multiple samples are typically obtained during preliminary sampling of a stockpile or
borrow pit, and test results for 200% and %OM are input into the database as a range.
The districts’ database was re-configured into a form that could be imported into Microsoft’s
Access database and queried with search terms such as district, county, project number, survey
date, or mine name. For records with limited location information, the street address, road
intersection or mine name were researched to get an approximate latitude and longitude for the
source material. Not all records contained enough information to identify the source or source
location; these records were removed from the Access database. The Access database was
queried for separate listings by district of all sources tested and for those sources that were
accepted for MSE wall backfill. In the latter case, the database query included passing test
results for the upper limits of 200% and %OM, passing results for LL and PI, and excluded
records for which the remarks indicated the material failed for MSE wall backfill. With ESRI’s
ArcMap 10.1 geographical information system (GIS) were plotted the location of each source
along with a symbol proportional to the number of times the source appeared as a record in the
database (Figures 4-1, 4-2, and 4-3). Also included in Figures 4-1 and 4-2 were the locations of
commercial sand and gravel mines that produced more than 45,000 metric tons of material in
2001 within FDOT’s District 1 and District 7 (USGS, 2002).
49
Figure 4-1. FDOT’s District 1 sources tested for MSE wall backfill (circles: small =
1 record, large = 10 records) and construction sand and gravel mines that in 2002
produced more than 45,000 metric tons (pushpin).
50
Figure 4-2. FDOT’s District 7 sources tested for MSE wall backfill (circles: small =
1 record, large = 100 records) and construction sand and gravel mines that in 2002
produced more than 45,000 metric tons (pushp ins).
51
Figure 4-3. FDOT’s District 1 and District 7 sources passed for MSE wall backfill
(circles: small = 1 record, large = 10 records).
52
4.1.2 Results and discussion
Figure 4-1 and 4-2 show the dispersed locations of sources tested for MSE wall backfill within
FDOT’s District 1 and 7; more so for District 1 than for District 7. Only 30 of the 73 mapped
sources had material that passed for MSE wall backfill (Figure 4-3); many of the sources were
not commercial mines. Approximately 50% of the 43 sources that had material fail for MSE
backfill, had material fail to meet gradation requirements (Chapter 2, Section 2.1.1, Table 2.1).
Tampa Groves Mine (and neighboring mines) and CC Calhoun Pit were important backfill
sources for Hillsborough and Pinellas Counties and for Polk County, respectively. MSE wall
backfill sources were relatively scarce in some counties including the more metropolitan
Hillsborough, Pinellas, and Polk Counties (Figure 4-3).
The links between the borrow pit or stockpile source, certification data, and corrosion AQC
acceptance data as downloaded from Laboratory Information Management System (LIMS) were
for the most part broken. Although the C-22 card does ask for plant or pit number, this
information was largely missing from the LIMS records, possibly because material was obtained
from on-site or because no plant or pit number was included in the material certification. A
recommendation is that future documentation include the geo-coordinates (latitude and
longitude, for example) of the borrow pit or stockpile of certified MSE wall backfill and that
these geo-coordinates accompany the certification data as well as the acceptance data into LIMS.
4.2 Failure Trends
4.2.1 Methods
The LIMS database is an FDOT-resource for laboratory data input, storage, and retrieval, with
features that assist with QA/QC and with data tracking, mining, and analysis. LIMS was
accessed to retrieve FM test results and related project information. For this project, LIMS was
entered via an “ad hoc” report, which could be modified to select or reject fields and to query for
selected tables. The fields chosen for this analysis were District, County Code, Project ID, Test
Code, Date Sample Taken, Sample ID, Sample Number, Lab ID, Assay Results, Sample Type,
and Plant or Pit Number. The query was written to select only those records with Date Sample
Taken after 2/20/2002; with Material ID (Specification Table) = 092L, which is the material for
MSE wall backfill; and for Test Code (Test Table) = FM5-550 or FM5-550A or FM5-550B or
FM5-551 or FM5-552 or FM5-553, which corresponded to pH, pH for MSE walls with metal
strips, pH for MSE walls with polymer strips, minimum resistivity, chloride, and sulfate,
respectively. In most cases, Sample ID and Sample Number corresponded with a suite of
corrosion tests from a common test date. Lab ID referred to the independent or FDOT laboratory
that tested the material. Sample types were Q, V, or R, for the contractor’s quality assurance
sample, the CEI’s verification sample, and infrequently the resolution sample, respectively.
From LIMS, data were retrieved as a table with 16,283 records and copied into Microsoft’s
Excel 2010 spreadsheet for further editing. The unedited table containing FDOT’s District 1
53
and District 7 records were imported into the Access database. Only FDOT’s District 1 and
District 7 AQC acceptance test data were included in the analysis of statistical trends; these data
were edited. FM5-550 A was renamed to FM5-550A, and the below records were removed:
 Duplicate records;
 Records with blank test results;
 Records with either an ‘<’ or an ‘>’ alongside a number; and
 Resistivity records before August 2010, as before this date units on the test result
appeared to both ohm-cm and ohm-m, but after this date with two exceptions, units
appeared to be ohm-m.
Four additional records were removed as outliers: one record with a sulfate concentration >5000
ppm and three records with a minimum resistivity less than 1 ohm-m.
StatSoft’s Statistica 12 is a commercially-available statistical software application that features

advanced query features, a visual analytic workspace, and reporting tables. Statistica 12 was
employed to transform data, compile summary statistics, perform inference tests, generate
frequency tables, and create normal probability graphs, as examples.
4.2.2 Results and discussion
Summary statistics revealed important information and potential trends about the underlying data
(Table 4-1). For normally-distributed data, as an example, the mean, median, and mode are the
same. These conditions were not met for any of the corrosion AQC test data, and a
Kolmogorov-Smirnov (K-S) test for normality confirmed that the distributions were not normal
(p<0.05).
54
Table 4-1 Statistics for FDOT’s District 1 and 7 MSE Wall Backfill Corrosion Tests
Frequency
Percentile
Percentile
Percentile
Percentile
of Mode
Valid N
Median
St Dev
Group
Mode
Mean
Max
Min
10th
25th
75th
90th
Variable
All 536 6.9 7.0 8.2 19 4.0 9.9 5.3 6.1 7.9 8.4 1.1
pH 1 124 7.4 7.5 7.2 8 4.9 8.8 6.1 7.0 8.2 8.4 0.9
7 412 6.8 6.8 7.3 14 4.0 9.9 5.2 5.9 7.8 8.2 1.1
Minimum All 528 250 130 120 19 10 1,500 30 59 360 620 280
Resistivity 1 82 120 90 Multi 4 15 380 40 56 170 260 88
(ohm-m) 7 446 272 150 120 15 10 1,500 28 60 400 680 300
All 706 60 56 60 266 1.0 780 23 30 60 75 78
Chloride
1 106 38 30 30 39 1.0 90 15 30 60 60 19
(ppm)
7 600 64 60 60 242 2.0 780 23 30 60 90 84
All 673 61 39 0.0 67 0.0 1,080 0.5 12 81 141 78
Sulfate
1 105 57 44 4.8 8 0.0 204 4.8 14 77 137 51
(ppm)
7 568 62 38 0.0 65 0.0 1,080 0.0 11 81 143 82
The minimum, maximum, and percentile statistics revealed that one or more samples failed to
meet the acceptance criteria. From Table 4-1, at least two backfill samples but not more than
10% of the samples in District 1 and 7 were below the acceptance limit of pH 5; likewise, at least
one backfill sample but not more than 10% of the samples is District 7 were above the
acceptance limit of pH 9. From the frequency tables generated in Statistica 12:
 For pH, 3% of the samples were below pH 5 or above pH 9; the failure rate was 1% for
District 1 and 4% for District 7 (zero and three samples were below pH 4.5 for District 1
and District 7, respectively);
 For minimum resistivity, 10% of the data were below the acceptance limit of 3000 ohm-
cm (30 ohm-m); the failure rate was 6% for District 1 but 11% for District 7;
 For chloride, 4% of the data were above the acceptance limit of 100 ppm, all of these
from District 7; and
 For sulfate, 4% of the data were above the acceptance limit of 200 ppm; the failure rate
was 1% for District 1 but 5% for District 7.
These results suggested that except for minimum resistivity, overall failure rates were relatively
low, but MSE wall backfill used in District 7 projects had a higher failure rate than did backfill
used in District 1 projects.
Hypotheses were tested that between District 1 and District 7 the means and variances within
corrosion AQC tests were not different. To do so, a t-test and F-test, respectively, were applied
on the log-transformed data distributions for minimum resistivity, chloride, and sulfate; pH is
55
already log-transformed (Figure 4-4). The log-transformed mean or standard deviations were
significantly different at the 95% confidence level (p<0.05), results that suggested the properties
of the MSE wall backfill between the districts were different.
For inference testing, ideally the sample data are independent, normally distributed, and have
equal variances. If the data are not normally distributed or the data sets do not have equal
variance, the t-test results are approximately correct when the data set is large (Ott, 1993). For
inference testing, at least two other options are available: data transformation or a non-parametric
inference test. A log transformation is often applied to data sets with a positive skew (Ott, 1993).
Except for log-transformed minimum resistivity, log-transformed data distributions were not
normally distributed but were a better match to a normal distribution than the un-transformed
distribution (Figure 4-4); thus, for pH, chloride, and sulfate, the probabilities returned by the
inference tests were only approximate.
Hypothesis 1: mean (District 1) = mean (District 7)

Hypothesis 2: variance (District 1) = variance (District 7)
 pH: t-test, p<<0.05; F-test, p<<0.05; reject both hypotheses

 log Resistivity: t-test, p<<0.05; F-test, p<<0.05; reject both hypotheses
 log chloride: t-test, p<<0.05; F-test, p=0.07; reject hypothesis 1; accept hypothesis 2
 log sulfate: t-test, p=0. 53; F-test, p<<0.05; accept hypothesis 1; reject hypothesis 2
Although the differences could be due to MSE wall backfill properties between the two districts,
they could also be due to differences in sampling and analysis techniques. Proximity of District
7 backfill sources to the Gulf of Mexico may explain the higher average and standard deviation
chloride concentrations; another explanation, however, is that a difference exists in method
procedures used by major laboratories in each district. For example, for the same Hach Chloride
Low Range Test Kit Model 8-P, 5-400 ppm, one laboratory may use a sample size of 23 mL and
a soil chloride resolution of 15 ppm and another laboratory may use a sample size of 10 mL and
a soil chloride resolution of 60 ppm.
Hypotheses were also tested that between Q and V samples, the means and variances within
corrosion tests were not different. Since across both districts the same laboratory can on one
project test the contractor’s Q samples and on another project the CEI’s V samples, differences
were not expected, yet significant differences were found.
Hypothesis 3: mean (Q) = mean (V)

Hypothesis 4: variance (Q) = variance (V)
 pH: t-test, p=0.10; F-test, p=0.06; accept both hypotheses

 log minimum resistivity: t-test, p=0.34; F-test, p<<0.05; accept hypothesis 3, reject
hypothesis 4
56
 log chloride: t-test, p<<0.05; F-test, p<<0.05; reject both hypotheses
 log sulfate: t-test, p<<0.05; F-test, p=0.18; reject hypothesis 3, accept hypothesis 4
In most but not every case Q and V samples came from the same backfill, albeit sampled at a
different time or from a different location within the MSE wall during the emplacement process.
The exception to this case is for borrow pit or stockpile sampling as those samples were likely
coded as V. Evidence for this is that of 2443 records of edited LIMS data, 1570 or 64% were
coded V. Thus, one explanation for these test results is that there is some segregation between
laboratories that perform Q and V tests; for example, FDOT’s Districts 1 and 7 Materials Office
may perform only V tests.
Figure 4-4. Normality plots for pH and log-transformed minimum resistivity,

chloride, and sulfate.
57
4.3 Summary
In summary, of the 73 potential MSE wall backfill sources databased by FDOT’s Districts 1 and
7 Materials Office laboratory, backfill from only 30 sources passed the geotechnical and
corrosion requirements. Approximately 50% of the 43 sources that were screened for MSE
backfill had material fail to meet gradation requirements. MSE wall backfill sources were widely
dispersed across Districts 1 and 7, and were relatively scarce in some counties including in the
more metropolitan Pinellas, Hillsborough and Polk Counties.
The frequency of failure rates for corrosion properties of MSE wall backfill based on the LIMS
data set suggested that except for minimum resistivity with a 10% overall failure rate, failure
rates were relatively low. MSE wall backfill used in District 7 projects had a higher failure rate
than backfill used in District 1 projects.
MSE wall backfill corrosion properties varied significantly between District 1 and 7, results that
could be explained by differences in the backfill or by differences in sampling and analysis
techniques by the major laboratories in each district. The quality control or Q samples and the
verification or V samples varied significantly for some but not all corrosion properties, which
suggested that some segregation was present between laboratories that performed Q versus V
samples.
acceptance data as downloaded from LIMS were for the most part broken. A recommendation is
that future documentation include the geo-coordinates (latitude and longitude, for example) of
the borrow pit or stockpile of certified MSE wall backfill, and that these geo-coordinates
accompany the certification data as well as the acceptance data into LIMS.
58
5 Single-Laboratory Precision and Bias
5.1 Definitions of Precision, Accuracy, and Minimum Detection Level
5.1.1 Approaches to calculating accuracy and precision
A measured value y is the sum of its unknown true value µ plus an unknown measurement error
 (Montgomery, 2005) (Equation 5-1), where  consists of both a fixed component (bias) and a
random component.
𝑦 =𝜇+𝜀 (5-1)
In many situations, unknown µ may be represented as known, for example, when µ is a certified,
traceable, or consensus value or when µ was determined by a method with  that is low in
comparison to the method that produced y (Rice et al., 2012). If the random component of  is
negligible, the fixed component or bias of  can be estimated from Equation 5-2.
𝜀 = 𝑦 − 𝜇 = 𝜀𝑟𝑎𝑛𝑑𝑜𝑚 + 𝜀𝑏𝑖𝑎𝑠 (5-2)
A generalized expression of relative error %RE is presented in Equation 5-3 for the case where µ
is treated as known and yi is the ith observation of y. The sign is retained to indicate the direction
of bias. The lower is the %RE the more accurate is the method.
1 (𝑦𝑖 −𝜇)
%𝑅𝐸 = 100 ∙ 𝑛 ∑𝑛𝑖=1 (5-3)
𝜇
For the case of unknown µ, however, Equation 5-4 is more appropriate.
(𝑦𝑚𝑎𝑥 −𝑦𝑚𝑖𝑛 ) (𝑦𝑚𝑎𝑥 −𝑦𝑚𝑖𝑛 )

%𝑅𝐸 = 100 ∙ 1 𝑛 = 100 ∙ (5-4)
∑ 𝑦 𝑦̅
𝑛 𝑖=1 𝑖
Another and related measure of accuracy is the amount of analyte added to a sample that is
recovered by the method. Recovery is calculated using Equation 5-5, where Cs+ms is the analyte
concentration in the sample with a matrix spike, Cs is the analyte concentration in the sample
without a matrix spike, and Cms is target concentration of the matrix spike.
𝐶(𝑠+𝑚𝑠) −𝐶𝑠
%𝑅𝑒𝑐𝑜𝑣𝑒𝑟𝑦 = 100 ∙ (5-5)
𝐶𝑚𝑠
The usual measure of data scatter is standard deviation  and estimates of  are represented at s,
where s has both a random and a systematic component. Youden (1975) explained that the
random component is determined from replicate measurements in the same laboratory under the
same conditions, while the systematic component is introduced under differing conditions in the
same laboratory, for example, another operator, or between laboratories. Youden (1975) noted
59
that if other laboratories were not following the same procedures for reasons such as ambiguous
instructions or undefined environmental factors, then random error would vary between
laboratories as well. In general, however, the terms random and systematic are often used
interchangeably with single-operator and multi-laboratory, respectively, in reference to
components of variance.
Similar to Youden (1975), Rice et al. (2012) referred to the random component of method
standard deviation as the repeatability of a method sr, which is estimated by pooling sample
standard deviations sri of measurements of N different specimens as shown in Equation 5-6 (Rice
et al., 2012). A standard deviation that characterizes both random and systematic errors is the
reproducibility of a method sR. A method with lower sr or sR is a more precise method.
1
𝑠𝑟 = √𝑁 ∑𝑁 2
𝑖=1 𝑠𝑟𝑖 (5-6)
Another and related measure of precision is the relative standard deviation %RSD, as defined in
Equation 5-7, and applies generally to either  or s; that is, to either population or sample
statistics (Skoog et al., 1996); %RSD is synonymous with coefficient of variation (CV).
𝑠
%𝑅𝑆𝐷 = 100 ∙ 𝑦̅ = 𝐶𝑉 (5-7)
5.1.2 Method limit of detection (LOD) and limit of quantitation (LOQ)
5.1.2.1 Definition of LOD and MDL
The limit of detection (LOD) is defined by Ellison et al. (2009) as the “minimum concentration
of the analyte that can reliably be detected with a specified level of confidence.” The general
features of an LOD experiment or a similar method detection level (MDL) experiment are
(Ellison et al., 2009; Rice et al., 2012):
 7 to 20 replicate samples containing either no analyte (blank samples) or a low-level of

analyte are analyzed over a period of days;
 For analysis of a low-level of analyte, the analyte concentration should be in the range of
1 to 5 times the LOD; and
 For a method LOD (or method detection level MDL), replicate samples must be
processed through the entire method.
LOD is calculated according to Equation 5-8 using the average blank concentration x0 and
standard deviation s, and the one-tailed Student’s t value for the experimental degrees of freedom
 and at the specified error levels, for example, 5% for both Type I () and Type II () errors
(Ellison et al., 2009). The values for x0 and k1 are 0 and 1, respectively, if observations have
independent baseline corrections; xblank and 1, respectively, with no baseline correction, and 0
60
1
and k1 = √1 + 𝑁 , respectively, where Nbl are the number of observations averaged to obtain the
𝑏𝑙
baseline correction, for observations without independent baseline corrections (Ellison et al.,
2009). For x0 = 0, k1 = 1,  =  =0.05, and number of samples N = 7, Equation 5-8 simplifies to
Equation 5-9.
LOD  x0  kl st ,   kl st ,  (5-8)
𝐿𝑂𝐷 = 3.88 ∙ 𝑠 (5-9)
For comparison, MDL is calculated according to Equation 5-10, where the specified level of a
Type I error is 99% and for N = 7 simplifies to Equation 5-11 (Rice et al., 2012; FDEP, 2009).
MDL  st ,  (5-10)
𝑀𝐷𝐿 = 3.14 ∙ 𝑠 (5-11)

5.1.2.2 Definition of LOQ
The limit of quantitation (LOQ), or similarly the practical quantitation limit (PQL), is defined as
the “lowest concentration of analyte that can be determined with an acceptable level of
uncertainty” (Ellison et al., 2009) and takes into account variations in LOD or MDL from
laboratory to laboratory. Ellison et al. (2009) describes LOQ as 10  s; Rice et al. (2012) as ~5 
MDL.
5.1.2.3 Calculation of Standard Deviation
The standard deviation associated with the baseline-corrected analyte concentration is given in
Equation 5-12, where ss is the sample standard deviation and sbl is the blank standard deviation.
2
𝑠 = √𝑠𝑠2 + 𝑠𝑏𝑙 (5-12)
If ss = sbl, then Equation 5-12 is simplified to 𝑠 = √2 𝑠𝑠 .
5.2 Single Laboratory, Aqueous Standards
5.2.1 2011 version of FMs for pH, minimum resistivity, chloride, and sulfate
Aqueous standards traceable to the National Institute for Standards and Technology (NIST) for
pH, conductivity, chloride, and sulfate were obtained from reputable sources and tested in USF’s
soil laboratory. For pH, standards were buffer solutions of pH 5.00 ± 0.01 (25C) and pH 9.00 ±
0.01 (25 oC); for minimum resistivity, the standard was a sodium chloride conductivity solution
61
prepared at 250 ± 1 µS/cm (25C) or 4000 ± 16 ohm-cm (25C); for chloride, the standard was a
sodium chloride solution prepared at 100.0 ± 0.3 ppm as chloride; and for sulfate, the standard
was a potassium sulfate solution prepared at 202.0 ± 0.2 ppm as sulfate. Resistivity and pH
standard concentrations corrected to an ambient temperature of 21 oC were resistivity 4,167 ± 17
ohm-cm, pH 4.98 ± 0.05, and pH 9.02 ± 0.05. The pH and conductivity standards were tested
without dilution; however, the chloride and sulfate standards were diluted by a factor of three,
and the diluted solutions decanted through a coarse filter as is done in the extraction process.
The sulfate standard required further dilution by a factor of two. Results were summarized in
Table 5-1. A one-sided z-test was applied to test the hypothesis that the standard concentration
was higher than the observed concentration. The z-test was significant for chloride and sulfate
concentrations with biases of 5 ppm and 10 ppm, respectively.
Sulfate Based on Aqueous Standards (2011 Version of FMs)
FM N Average St Dev, sr %RSD %RE
pH 5 5.00 0.01 0.23 0.32
pH 5 9.02 0.00 0.00 -0.22
Resistivity, ohm-cm 5 4,160 198 4.76 -0.17
Chloride, ppm 5 95.1 5.34 5.62 -4.90
Sulfate, ppm 3 192 6.00 3.13 -4.95
In USF’s soil laboratory, pH measurements with NIST-traceable buffers at pH 5.00 and pH 9.00
were made with no significant bias and with precision comparable to those reported for the
buffers. Likewise, resistivity of a 4,000 ohm-cm solution was made with no significant bias and
with a %RSD within the 5% mid-scale precision reported by the equipment manufacturer.
Measurements of both chloride and sulfate, however, had biases of -5%, which were significant.
For chloride, such a bias could have been caused by error in the blank determination; for sulfate
the bias could have been caused by error in the calibration curve. For both, error in volumetric
measurements could have been a source of bias. The single laboratory or repeatability sr and
%RSD were somewhat high because of the sample dilution. For example, a sample of 200 ppm
sulfate must be diluted by a factor of ~6 to obtain a consistent photometer reading, thus a change
of one ppm on the photometer reading equates to a change of six ppm in the sample.
5.2.2 Proposed revised FMs for pH, minimum resistivity, chloride, and sulfate
Aqueous standards traceable to the NIST for pH, conductivity, chloride, and sulfate were
obtained from reputable sources and tested in USF’s soil laboratory and results were summarized
in Table 5-2. Refer to Table 5-2 for pH and resistivity of temperature-corrected standard
concentrations. For pH, the Accumet meter/electrode AB150/13-620-631 was calibrated with
three standard buffers, nominally pH 4, 7, and 10. Test buffers included two purchased normal-
ionic strength buffers (pH 5.00 and pH 9.00 at 25C), two purchased low-ionic strength buffers
(pH 4.10 and pH 6.97 at 25C), and two very low-ionic strength buffers (pH 4.00 and pH 5.00)
prepared in the USF soils laboratory from NIST-traceable 0.02 N hydrochloric acid. For pH, the
62
test temperature was 20C and all measurements were made on the same day. Minimum
resistivity was tested using a Nilsson Model 400 analog meter and a McMiller 280-mL soil box
with both a 250-µS/cm (4,000 ohm-cm) and a 500-µS/cm (2,000 ohm-cm) sodium chloride
conductivity standard. Three measurements were made with each standard one day apart, with
measurement temperatures of 23C and 21C on the first and second day, respectively. Chloride
and sulfate were tested using their respective Hach kits with 100 ppm standards diluted by a
factor of three to be within the range of the test kits. For chloride and sulfate, three
measurements were made with each standard two days apart, and different lot numbers for the
Chloride 2 indicator (potassium dichromate and sodium bicarbonate) and SulfaVer 4 reagent
(barium chloride and citric acid) were used on consecutive days.
Sulfate Based on Aqueous Standards (Proposed Revised FMs)
FM Standard N Avg St Dev %RSD %RE


pH 4.01 3 4.00 0.012 0.29 -0.33
pH *4.10 3 4.07 0.010 0.24 -0.73
pH 4.98 3 5.03 0.010 0.20 1.00

pH 5.00 3 4.99 0.012 0.23 -0.27
pH *6.98 3 6.88 0.006 0.08 -1.38
pH 9.05 3 9.00 0.006 0.06 -0.52
Min Res, ohm-cm 4,170 3 4,200 0 0.00 0.72
Min Res, ohm-cm 2083 3 2,050 50 2.40 -1.58
Min Res, ohm-cm 4,350 3 4,383 58 1.33 0.77
Min Res, ohm-cm 2,174 3 2,150 50 2.30 -1.10
Chloride, ppm 100 6 97.5 8.22 8.22 -2.50
Sulfate, ppm 100 6 99.5 10.6 10.6 -0.50

Low-ionic strength buffers prepared from 0.02 N HCl; *Commercially-available low-ionic
strength buffers
Overall the results were not too different for the proposed revised FMs than for the 2011
versions. For the proposed revised FMs, pH precisions were within those given for the standards;
the only significant (z-test, p = 0.04) bias was seen for the regular-ionic strength pH 5.00 buffer
(1 %RE). Minimum resistivity precisions trended lower and biases higher, but for chloride and
sulfate these trends were reversed. Bias statements were developed for each proposed revised
FM from the results shown in Table 5-2.
63
5.2.2.1 Bias statement FM for pH
Single-operator, single laboratory bias for this method using an Accumet meter/electrode
AB150/13-620-631 was evaluated at 20C with repeated measures of each six aqueous standards
ranging from pH 5.00 to pH 9.00. The average bias was -0.01 pH units.
5.2.2.2 Bias statement FM for minimum resistivity
Single-operator, single laboratory bias for this method was evaluated using a Nilsson Model 400
analog meter and a McMiller 280-mL soil box and repeated measures of two aqueous standards
with resistivities at 25C of 2,000 ohm-cm and 4,000 ohm-cm. After correcting for measurement
temperatures in the range of 21C to 23C, average biases were -30 ohm-cm (-1.3%) for the
2,000 ohm-cm standard and 30 ohm-cm (0.7%) for the 4,000 ohm-cm standard. With no
correction for temperature, average biases were 100 ohm-cm (5%) for the 2,000 ohm-cm
standard and 300 ohm-cm (7%) for the 4,000 ohm-cm standard.
5.2.2.3 Bias statement FM for chloride
Single-operator, single laboratory bias for this method was evaluated using a Hach Chloride Low
Range Test Kit Model 8-P, 5-400 ppm, and repeated measures of a 100-ppm aqueous standard,
which was diluted by a factor of three to be in the target range of the test kit. Six replicate
standards were tested, three with one lot number and three with a different lot number of
Chloride 2 indicator. The average bias was -2.5 ppm (-2.5%).
5.2.2.4 Bias statement FM for sulfate
Single-operator, single laboratory bias for this method was evaluated using a Hach Sulfate,
Pocket Colorimeter II Test Kit and repeated measures of a 100-ppm aqueous standard, which
was diluted by a factor of three to be in the target range of the test kit. Six replicate standards
were tested, three with one lot number and three with a different lot number of SulfaVer 4
reagent. The average bias was -0.5 ppm (-0.5%).
5.3 Single Laboratory, Proficiency Test Soil
Santa Fe River soil (Chapter 7, Section 7.3, Table 7-1) was air-dried and sieved through a No. 10
(2 mm) mesh and replicate samples were prepared according to AASHTO T248. With this soil,
repetitive measurements were made according to the FMs for pH, minimum resistivity, chloride,
and sulfate in USF’s soil laboratory. Chloride and sulfate concentrations were low in the Santa
Fe River soil, at or below the LOD or MDL. Thus, chloride and sulfate were added to this soil in
the form of solid sodium chloride or sodium sulfate such that the minimum resistivity was near
3,000 ohm-cm. Typical amendments were 0.1 g/kg of sodium chloride and 0.1 g/kg of sodium
sulfate with and without 0.01 g/kg sodium bisulfate, or 0.25 g/kg sodium sulfate. Salts were
added to the soil and then mechanically mixed for 10 min in 3-kg batches. As necessary,
64
multiple batches were mixed and mechanically split per AASHTO T248 to obtain about 1,200 g
per replicate, which was sufficient soil for all four corrosion tests. Amendment with sodium
bisulfate was tried as a way to lower pH but at 0.2 g/kg sodium bisulfate the soil resistivity
reached ~3,000 ohm-cm with little change in pH. Dilution water was deionized water from
USF’s Nanotechnology Research and Education Center, which typically had a resistivity greater
1,000,000 ohm-cm. Nine replicates of Santa Fe River soil were tested with and without salt-
amendment; average, standard deviation, %RSD and %RE were reported in Table 5-3.
Table 5-3 Accuracy and Precision of Florida Methods for pH, Minimum Resistivity,
Chloride, and Sulfate with Santa Fe River Soil
FM N Average St Dev, sr %RSD %RE
pH
No amendment 9 7.97 0.06 0.73 2.13
Salt amendment* 9 7.95 0.07 0.88 2.39
Minimum Resistivity, ohm-cm
No amendment 9 15,900 782 4.92 12.6
Salt amendment* 9 2,828 227 8.01 28.3
Chloride, ppm
No amendment 9 -1 2 -193 837
Salt amendment* 9 65 12 18.9 61.0
Sulfate, ppm
No amendment 9 37 27 71.4 240
Salt amendment* 9 99 8 8.19 90.9
* 0.1 g/kg sodium chloride, 0.1 g/kg sodium sulfate, and 0.01 g/kg sodium bisulfate
For pH measurement, %RE and %RSD were reasonably low, with or without sodium salts
added. For minimum resistivity, %RE and %RSD increased when conducting salts were added,
which may reflect a greater degree of heterogeneity between replicates rather than method
factors. Note that for the unmodified soil, the %RSD was within the 5% mid-scale precision
reported by the equipment manufacturer. For chloride, with no salts added the %RE and %RSD
were very high, a strong indication that the chloride concentration in the sample was below the
detection level. (A negative concentration for the average is possible with a blank subtraction,
when the blank concentration is higher than the sample concentration). With salts added, the
%RE and %RSD were still rather high, a result that suggested there was room for improvement
in the method. The salt amendment increased the soil chloride concentration by 60 ppm
compared with an observed average concentration of 65 ppm; this represented a chloride
recovery of 107% and suggested that the average concentration of chloride in the unmodified soil
may have been on the order of 5 ppm. For sulfate, with no salts added %RE and %RSD were
high, an indication that the sulfate concentration in the sample was below the detection level.
With salts added, %RE and %RSD were still rather high, a result that suggested there was room
for improvement in the method. The salt amendment increased the soil sulfate concentration by
75 ppm compared with an observed average concentration of 99 ppm; this represents a sulfate
65
recovery of 130% and suggested that the average concentration of sulfate in unmodified soil may
have been on the order of 25 ppm.
Both LOD and MDL are common approaches to identify the level at which a signal emerges
from noise. For a probability of Type I and Type II error of 5% and N = 9, the LOD for pH,
minimum resistivity, and sulfate was 3.72  s, with no baseline or independent baseline
corrections, and 5.26  s for chloride, with one blank concentration measurement for several
sample measurements. The MDL was 2.896  s, given a Type I error of 1% and N = 9. LODs and
MDLs were reported in Table 5-4.
Table 5-4 Level of Detection (LOD) and Method Detection Level (MDL) of Florida
Methods for pH, Minimum Resistivity, Chloride, and Sulfate for Santa Fe River Soil
FM N LOD MDL
pH
No amendment 9 0.22 0.17
Salt amendment* 9 0.26 0.20
Minimum Resistivity, ohm-cm
No amendment 9 2,910 2,260
Salt amendment* 9 842 656
Chloride, ppm
No amendment 9 8 5
Sulfate, ppm
No amendment 9 99 77
*0.1 g/kg sodium chloride, 0.1 g/kg sodium sulfate, and 0.01 g/kg sodium bisulfate
The LOD or MDL for pH and for minimum resistivity in salt-amended soil were far enough
away from acceptance limits of pH 5.00 and 3,000 ohm-cm, respectively, and posed no problem
for the current method. Likewise the LOD or MDL for sulfate, while high at 30 ppm or 25 ppm,
respectively, was relatively distant from the acceptance limit of 200 ppm and therefore did not
pose a problem for the current method. For chloride, however, the LOD approached 2/3 of the
acceptance limit of 100 ppm, thus the LOD would benefit from method improvement.
5.4 Multiple Laboratories, Proficiency Test Soil
Starvation Hill soil (Chapter 7, Section 7.3, Table 7-1) was air-dried, sieved through a No. 10 (2
mm) mesh, and quartered according to AASHTO T248 into ~3 kg replicate samples. These
samples were distributed to four laboratories, including USF’s soil laboratory, and tested
according to the FMs for pH, minimum resistivity, chloride, and sulfate analysis. Results were
summarized in Table 5-5.
66
Table 5-5 Multiple Laboratory Results of Florida Methods for pH, Minimum
Resistivity, Chloride, and Sulfate for Starvation Hill Soil
Minimum Resistivity,
Laboratory pH Chloride, ppm Sulfate, ppm
ohm-cm
1 6.60 4,100 26 24
2 7.75 23,810 8 3
3 5.60 5,300 60 3
4 7.57 23,500 3 14
For pH, multi-laboratory results reflected the range of pH measurement technique,

instrumentation, and electrode condition seen in laboratory visits (Chapter 6). For minimum
resistivity, the large disparities in readings pointed to differences in how the soil slurry was
handled during testing: minimum resistivity would be lower if the water portion of the slurry was
decanted into the soil box first and then topped with the soil portion of the slurry versus filling
the box with mostly the soil slurry. Refer to Chapter 9, Section 9.2.7, for further discussion. For
chloride and sulfate, results were consistent with levels near or below detection (Table 5-4).
5.5 Summary
Estimates of single laboratory precision (%RSD) and bias (%RE) were obtained with aqueous
standards and with a proficiency test soil with FMs for pH, minimum resistivity, chloride, and
sulfate. Across all four FMs, estimates of %RSD and %RE were higher for a soil sample than
for aqueous standards as expected, but much higher for chloride and sulfate concentrations
especially for chloride and sulfate concentrations below ~5 ppm and ~25 ppm, respectively. A
preliminary multi-laboratory study with a proficiency test soil yielded disparate results for both
pH and minimum resistivity, which suggested that significant differences in procedures or
equipment existed between laboratories and merited further investigation.
67
6 Laboratory Visits
6.1 Introduction
Multi-laboratory contributions to method reproducibility were estimated pre- and post-method

revision to assess improvement in pH, minimum resistivity, chloride and sulfate measurements.
Prior to method development, on-site testing of a select backfill was conducted at six FDOT and
nine commercial laboratories, with representative laboratories in each FDOT district.
6.2 Protocol
Dr. Noreen Poor conducted the visits. The protocol was similar between laboratories. Method
operators were asked a standard set of questions regarding sample processing, soil extraction,
and each of the FMs. Tests were done in the same order: minimum resistivity, pH, chloride, and
sulfate. For resistivity and pH, operators were asked to test Santa Fe River soil that had been air-
dried, sieved, and mechanically-split into 1,500-g subsamples. For chloride and sulfate,
operators were asked to test 30 ppm chloride and 30 ppm sulfate standard solutions, respectively.
These solutions were prepared in advance through dilution of NIST-traceable 1000-ppm stock
solutions. Operators were also asked to do a few additional quality assurance/quality control
(QA/QC) measurements along with the FMs. For resistivity, these included measurement of
resistivity for (1) a 3,300-ohm resistor, (2) a 3,333 ohm-cm sodium chloride conductivity
standard, and (3) the laboratory’s dilution water. An audit measurement was also made with
USF’s resistivity meter of the soil box plus soil at the minimum resistivity. For pH, slope and
offset were checked for working electrodes; and after completion of the operator’s measurement,
an audit measurement was made with USF’s pH meter and electrode. For chloride, the use of a
blank to check and adjust for chloride content in the dilution water was demonstrated. Finally,
for sulfate, for those laboratories that were not using an instrument with a built-in conversion of
light attenuation to sulfate concentration, a machine-specific calibration curve was prepared.
6.3 Laboratory Environment
For the participating laboratories, laboratory environments were somewhat different between
commercial laboratories and FDOT laboratories. For commercial laboratories, MSE wall
backfill samples were processed either as quality control or as verification tests. Facilities were
those of a geotechnical rather than a chemistry laboratory: space was likely as not to be
controlled for climate, noise, dust, or vapor; MSE wall backfill samples typically arrived in
sufficient quantity for all required geotechnical and corrosion tests, and dilution water was
typically purchased. Commercial laboratories processed anywhere from 1 to 30 samples per
week, and estimated that on the average it took about two hr of labor to do all four corrosion
tests. Although not required by FDOT, oversight for the corrosion methods came either from
CMEC or AASHTO Materials Reference Laboratory (AMRL).
For FDOT laboratories, MSE wall backfill samples were typically processed as resolution tests.
Facilities were more a combination of geotechnical and chemistry laboratories and likely to have
68
ventilated and climate-controlled space, a fume hood, and on-site production of dilution water.
MSE wall backfill samples typically arrived in sufficient quantity for all required geotechnical
and corrosion tests or in a smaller quantity sufficient for all four corrosion tests. FDOT
laboratories processed anywhere from 0 to 15 samples per week and estimated that on the
average it took about three hr of labor to do all four corrosion tests. FDOT laboratories reported
no oversight of the four corrosion methods, with the exception of FDOT’s SMO Environmental
Laboratory, which is accredited by the National Environmental Laboratory Accreditation
Program (NELAP). FDOT’s SMO Environmental Laboratory routinely tests chloride and sulfate
levels in both water and soil samples.
Both FDOT and commercial laboratories tested core samples of soil and sediment, which at
times did not contain enough soil for all four corrosion tests. In this case soil was typically
removed for pH, chloride, and sulfate analyses and the remaining soil used for a resistivity
measurement, where the volume of the soil box (270 mL or 80 mL) limited the minimum sample
size. Water was added in 10% increments, for example, 50 mL increments for 500 g of soil or
25 mL for 250 g of soil.
6.4 Results
6.4.1 FM for pH
In general, pH measurements were inconsistent across all 15 laboratories. With one exception,
no two pH meter/electrode systems were matched. Accumet, which is a pH meter marketed by
Thermo Fisher Scientific and Cole Palmer, was the most common brand. The capabilities of pH
meter/electrode systems also varied to include one or more combinations of pH, millivolts (mV),
temperature (C), conductivity, and ion specific electrode (ISE); most were bench top systems
and all but one pH meter/electrode system had a glass electrode. Electrodes ranged in age from
new to 10 years old; electrodes at three laboratories did not function properly and electrodes at
four laboratories were suspect (high offset voltages, erratic readings, sensitivity to vibration,
slow response time). Calibration of pH measurements ranged from a single pH 7 buffer
calibration to a three-point calibration with pH 4, 7, and 10 buffers. Summary statistics from the
laboratory visits (Table 6-1) were compared with those generated in the USF laboratory (Chapter
5, Section 5.3, Table 5-2), and audit measurements in each laboratory. Although the data sets
summarized in Table 6-1 were based on Santa Fe River soil collected from FDOT’s State
Materials Office (SMO), soil sampling dates were November 11, 2013 for the USF laboratory
results and March 10, 2014 for the inter-laboratory results.
69
Table 6-1 Summary Statistics for pH Measurements for a Sub-Sample of Air-Dried,
Sieved, and Mechanically-Split Santa Fe River Soil
FM for pH N Average St Dev (s) %RSD %RE
USF Laboratory 9 7.97 0.06 0.73 2.13
USF On-Site Audit 14 8.15 0.15 1.83 6.62
Laboratory Visits 15 7.85 0.80 10.1 32.4
For pH, the average, standard deviation s, %RSD, and %RE were higher for the USF on-site
audit than for USF laboratory measurements and measurements between laboratories had a lower
average but a higher standard deviation, %RSD, and %RE than for USF laboratory and on-site
audit results (Table 6-1). Bias and thus accuracy remain undefined for a method in the absence
of an accepted reference value, but a large %RE suggested the presence of systematic error.
Refer to Table 6-2 for sources of variability in the FM for pH measurement between laboratories.
Following the guidance of ASTM E177, example method statements of precision generated from
the between-laboratory statistics were:
Soil pH measurements were made by 15 laboratories with randomly selected subsamples

of an air-dried and sieved sandy soil. For a like material, approximately 95% of
individual test results from laboratories similar to those in the inter-laboratory study can
be expected to differ in absolute value from their average by less than 2.0 x s = 1.6 pH
units.
Soil pH measurements were made by 15 laboratories with randomly selected subsamples

of an air-dried and sieved sandy soil. The average test value was 7.85 pH units, and the
reproducibility limit (between laboratory) was 2.8 x s = 2.2 pH units.
In practical terms, if the pH average and standard deviation (laboratory visit s from Table 6-1) of
the quality control and verification samples were 5.0 ± 0.8 pH units, the null hypothesis (H0) that
the soil pH was equal to pH 4 could not be rejected at the 95% confidence level. Likewise, if the
pH average and standard deviation (laboratory visit s from Table 6-1) of the quality control
sample and verification samples were 9.0 ± 0.8 pH units, the null hypothesis (H0) that the soil pH
was equal to pH 10 could not be rejected at the 95% confidence level. These results suggested
that for pH, the multi-laboratory precision was not adequate. Note that s for pH was
underestimated, as statistics generated by this study were limited to one soil and did not include
variability from soil heterogeneity or from field sampling, transport, and storage of soil.
70
Table 6-2 Sources of Variability in pH between Laboratories
Transported soil at ambient vs cool temperatures
Stored soil under ambient conditions or in the refrigerator
Used 100 g or 100 mL of soil
Tested “as is” or air dried
Used soil from resistivity measurement
Waited zero to 30 min or more to test sample
Stirred once or up to three times
Put soil/water mixture on shaker table for three 10-min intervals
Stirred or did not stir sample during pH measurement
Measured pH as soon as stable light came on or waited until reading was stable for 1 min
Prepared soil water/mixture in a beaker, bottle, or disposable cup
Calibrated with one, two, or three buffers
Used fresh buffers or re-used buffers for calibration
Stored pH electrode in distilled water, tap water, buffer, KCl, or dry
Used glass electrode or solid state electrode
Refilled glass electrode or used disposable electrode
Kept electrode for up to 10 yr
In summary, for an average of two samples the multi-laboratory precision was not adequate to
assure that a pH 5 or pH 9 measurement was different from critical or rejection limits of pH 4
and pH 10, respectively. Note that s for pH was underestimated, as statistics generated in this
study were limited to one soil and did not include variability from soil heterogeneity or from
field sampling, transport, and storage of soil. The accuracy of the method could be improved by
reducing apparent systematic errors.
6.4.2 FM for minimum resistivity
In general, soil minimum resistivity measurements were relatively consistent across all 15
laboratories. All but two laboratories had an analog Nilsson 400 model resistivity meter and all
but one laboratory used a four-pin measurement technique. Calibration checks with a resistor
and with a conductivity standard were within 5% and 10%, respectively; likewise, little or no
differences were seen when minimum resistivity measurements were checked with an audit
meter. Summary statistics for laboratory visit minimum resistivity measurements were
compared with those generated in the USF laboratory (Chapter 5, Section 5.3, Table 5-2).
Although the data sets summarized in Table 6-3 were based on Santa Fe River soil collected
from FDOT’s State Materials Office (SMO), soil sampling dates were November 11, 2013 for
the USF laboratory results and March 10, 2014 for the laboratory visit results.
71
Table 6-3 Summary Statistics for Minimum Resistivity Measurements for a Sub-
Sample of Air-Dried, Sieved, and Mechanically-Split Santa Fe River Soil
FM N Average, ohm-cm St Dev (s), ohm-cm %RSD %RE
USF Laboratory 9 15,900 782 4.92 12.6
Laboratory Visits 15 24,000 2,840 11.9 41.7
For minimum resistivity, the average, standard deviation s, %RSD, and %RE were higher for
laboratory visits than for the USF laboratory (Table 6-3). The higher average minimum
resistivity may reflect in part heterogeneity of the stockpiled soil. Bias and thus accuracy remain
undefined for a method in the absence of an accepted reference value, but a large %RE suggested
the presence of systematic error. Refer to Table 6-4 for sources of variability in minimum
resistivity measurements between laboratories.
the between-laboratory statistics were:
Soil minimum resistivity measurements were made by 15 laboratories with randomly

selected subsamples of an air-dried and sieved sandy soil. For a similar material,
approximately 95% of individual test results from laboratories similar to those in the
inter-laboratory study can be expected to differ in absolute value from their average by
less than 2.0 x %RSD = 24%.
Soil minimum resistivity measurements were made by 15 laboratories with randomly

selected subsamples of an air-dried and sieved sandy soil. The average test value was
24,000 ohm-cm, and the reproducibility limit (between laboratory) was 2.8 x %RSD =
33%.
In practical terms, if the minimum resistivity average and standard deviation (laboratory visit s
from Table 6-3) of the quality control and verification samples were 3,000 ± 360 (or ± 12%)
ohm-cm, the null hypothesis (H0) that the minimum resistivity was equal to 1,000 ohm-cm could
be rejected at the 95% confidence level. These results suggested that the multi-laboratory
precision for minimum resistivity was adequate with the caveat that statistics generated by this
study were limited to one soil with a high minimum resistivity and did not include variability
from soil heterogeneity or from field sampling, transport, and storage of soil.
72
Table 6-4 Sources of Variability in Minimum Resistivity between Laboratories
Soil stored at ambient temperatures or refrigerated
Soil tested at ambient temperatures or in a climate-controlled space (19 oC to 30 oC)
Used 250 g, 500 g, or 1000 g of soil and 10% water
Switched from 100 mL to 50 mL water when soil became saturated or used 100 mL for entire test
Tested soil “as is” or air dried
Used large or small soil box or homemade box
Mixed in containers of varying composition, size, and shape
Prepared soil with deionized or distilled water
Rinsed, partially rinsed, or did not rinse box between measurements
Added water with a resistivity ranging from 130,000 to >1,400,000 ohm-cm
Mixed with utensils or with hands (gloved and ungloved)
Filled box to a different level
An important factor in minimum resistivity measurements is soil temperature. ASTM G187

provides a model for correcting observed soil resistivity to a resistivity at a standard temperature
(Chapter 3, Section 3.3.1, Equation 3-4). Soil minimum resistivity and corresponding
temperature observations from laboratory visits were plotted with a least squares fit of the model
to the data, which occurred for a standard temperature resistivity of 28,500 ohm-cm (Figure 6-1).
Much of the data variability was not explained by the temperature model.
73
Figure 6-1. Modeled and observed soil minimum resistivity versus temperature .
Dilution water was a possible source of variability and its resistivity should be tested prior to
making soil minimum resistivity measurements. In Figure 6-2 is shown soil minimum resistivity
versus water resistivity. No clear trend was seen and the resistivity of most if not all of the
laboratory water samples that were tested exceeded 200,000 ohm-cm.
74
Figure 6-2. Observed soil minimum resistivity versus water resistivity .
In summary, for the average of two samples the multi-laboratory precision appeared adequate to
assure that a measured minimum resistivity of 3,000 ohm-cm exceeded a rejection limit of 1000
ohm-cm, with the caveat that statistics generated in this study were limited to one soil with a
high minimum resistivity and did not include variability from soil heterogeneity or from field
sampling, transport, and storage of soil. The accuracy of the method could be improved by
reducing apparent systematic errors.
6.4.3 FM for chloride
In general, measured chloride concentrations were in poor agreement across all 15 laboratories.
Ten laboratories employed the Hach chloride low range test kit (5-400 ppm); four laboratories,
the Hach chloride test kit with a digital titrator (10-10,000 ppm); and one laboratory a buret
titration in accordance with SMEWW (Rice et al., 2012). Method procedures included steps to
extract chloride from the soil with a 3:1 dilution using distilled water, followed by titration of
chloride with silver nitrate in the presence of potassium dichromate. Silver chloride is relatively
75
insoluble and so the chloride titration completes first and the dichromate titration second. As
dichromate is titrated, the solution turns from a bright yellow to an orange-red. This color
indicates the end point of the chloride titration. Lack of agreement in test results between
laboratories were due in part to a difference in opinion regarding the intensity of the color at
which the titration was completed and in part to the amount of sample titrated. The FDOT SMO
laboratory was the only laboratory that tested the dilution water or a known chloride
concentration to check their techniques and reagents. Summary statistics for chloride
concentration measurements from the laboratory visits were compared with those generated in
the USF laboratory through routine measurements of a 30-ppm check standard (Table 6-5). Test
results were multiplied by three to mimic dilution of a soil extract.
Table 6-5 Summary Statistics for Measurements of a 30-ppm Chloride Standard *

FM 5-552 Chloride N Average, ppm St Dev (s), ppm %RSD %RE
USF Laboratory 4 92 5 4.9 9.8
Laboratory Visits 15 140 48 35 120
*
Test results were multiplied by 3 to mimic dilution of a soil extract
For chloride concentration, the average, standard deviation s, %RSD, and %RE were higher for
laboratory visits than for the USF laboratory (Table 6-5). For the entire method, which includes
soil sampling, transport, storage, extraction, and analysis, the bias and thus accuracy remain
undefined in the absence of an accepted reference value, but a large %RE suggests the presence
of systematic error. Refer to Table 6-6 for sources of variability in chloride concentrations
between laboratories. Also included in this table were sources of variability from the soil
extraction process, as the extraction process was described by method operators.
the inter-laboratory statistics were:
Chloride concentration measurements were made by 15 laboratories with a 30-ppm

standard chloride solution. Results were multiplied by three to mimic results for a diluted
soil extract. For a soil with a 90-ppm chloride concentration, approximately 95% of
individual test results from laboratories similar to those in the inter-laboratory study can
be expected to differ in absolute value from their average by less than 2.0 x s = 95 ppm.
Chloride concentration measurements were made by 15 laboratories with a 30-ppm

standard chloride solution. Results were multiplied by three to mimic results for a diluted
soil extract. The average test value was 136 ppm, and the reproducibility limit (between
laboratory) was 2.8 x s = 140 ppm.
In practical terms, if the chloride concentration average and standard deviation of the quality
control and verification samples were 100 ± 48 ppm, the null hypothesis (H0) that the chloride
concentration was equal to 500 ppm could be rejected at the 95% confidence level. These results
suggested that the multi-laboratory precision for chloride is adequate, with the caveat that
76
statistics generated by this study were limited to one 30-ppm check standard and did not include
variability from soil extraction, soil heterogeneity or field sampling, transport, and storage of
soil.
Table 6-6 Sources of Variability in Chloride Concentration between Laboratories

Mechanically split, quartered or thieved from original sample
Dried in oven at 60 oC or at 110 oC or in air at room temperature
Waited overnight to 8 days before filtering
Waited a few minutes to up to 5 days after filtering and before testing
Filtered with Fisher Q8, Fisher P9, Hach 132, Whatman 4, Whatman 41, or Whatman 42
Used two filters in sequence: Whatman 4 followed by Whatman 42
Filtered all of the sample or filtered water plus slurry or filtered water only
Filtered by gravity or filtered by suction
Filtered with a large or small funnel, or with a funnel made of plastic or glass
Removed filtrate by pipetting or by decanting
Filtered a split sample to speed filtration time
Based color change on a chart or on experience
Tested a soil extract volume of 10 mL or 23 mL or 40 mL or 100 mL (based on titrant concentration)
assure that an observed chloride concentration of 100 ppm did not exceed a rejection limit of 500
ppm, with the caveat that statistics generated in this study were limited to a chloride standard
solution and did not include variability from soil extraction, soil heterogeneity, or field sampling,
transport, and storage of soil. The accuracy of the method could be improved by reducing
apparent systematic errors.
6.4.4 FM for sulfate
In general, measured sulfate concentrations were in poor agreement across all 15 laboratories.
Five laboratories used a Hach photometer or spectrophotometer with a built-in sulfate calibration
curve, another laboratory used a turbidimeter with a calibration curve prepared over a sulfate
range of 10 to 120 ppm (Rice, 2012), and the nine remaining laboratories used a
spectrophotometer such as but not limited to the Spectronics 20D+. For all but one of these nine
laboratories, the conversion of % light transmission to sulfate concentration was based on a chart
from an older version of FM for sulfate. Method procedures included steps to extract sulfate
from the soil with a 3:1 dilution using distilled water, followed by reaction of a small volume of
sample with barium chloride to form insoluble barium sulfate. The turbidity or light attenuation
of the sample is proportional to its sulfate concentration. Lack of agreement between
laboratories was due in part to a mismatch between the % transmission versus sulfate
concentration relationship given by the chart and the actual response of the spectrophotometer.
Other issues included measuring at a different wavelength, not using any calibration curve, and
using the wrong reagent or out-of-date reagents. Only the FDOT SMO laboratory routinely
tested a known sulfate concentration to check their techniques and reagents. Summary statistics
for sulfate concentrations from the laboratory visits were compared with those generated in the
77
USF laboratory through routine measurements of a 30-ppm check standard (Table 6-7). The test
results were multiplied by three to mimic dilution of a soil extract.
Table 6-7 Summary Statistics for Measurements of a 30-ppm Sulfate Standard *

St Dev (s),
FM N Average, ppm %RSD %RE
ppm
USF Laboratory 5 88 8 9.5 24
Inter-Laboratory 15 100 35 34 150
*
Test results were multiplied by 3 to mimic dilution of a soil extract.
For sulfate concentration, the average, standard deviation s, %RSD, and %RE were higher for
laboratory visits than for the USF laboratory (Table 6-7). For the entire method, which includes
soil sampling, transport, storage, extraction, and analysis, the bias and thus accuracy remain
undefined in the absence of an accepted reference value, but a large %RE suggests the presence
of systematic error. Refer to Table 6-8 for sources of variability in sulfate concentrations
between laboratories. Refer to Table 6-6 for sources of variability from the soil extraction
process, as the extraction process was described by method operators.
the inter-laboratory statistics were:
Sulfate concentration measurements were made by 15 laboratories with a 30-ppm

standard sulfate solution. Results were multiplied by three to mimic results of a diluted
soil extract. For a soil of similar concentration, approximately 95% of individual test
results from laboratories similar to those in the inter-laboratory study could be expected
to differ in absolute value from their average by less than 2.0  s = 70 ppm.
Sulfate concentration measurements were made by 15 laboratories with a 30-ppm

standard sulfate solution. Results were multiplied by three to mimic results of a diluted
soil extract. The average test value was 103 ppm, and the reproducibility limit (between
laboratory) was 2.8  s = 99 ppm.
In practical terms, if the sulfate concentration average and standard deviation of the quality
control and verification samples were 100 ± 35 ppm, the null hypothesis (H0) that the sulfate
concentration was equal to the rejection limit of 1000 ppm could be rejected at the 95%
confidence level. These results suggested that the multi-laboratory precision for sulfate was
adequate, with the caveat that statistics generated by this study were limited to one 30 ppm-check
standard and did not include variability from soil extraction, soil heterogeneity, or field
sampling, transport, and storage of soil.
78
Table 6-8 Sources of Variability in Sulfate Concentrations between Laboratories
Added HCl different amounts, concentrations, and to different sample volumes, if added
Chose a different Hach photometer range (range 1 or range 2) for built-in curve
Used a turbidity (light scattering) instead of a light attenuation measurement
Added a buffer to stabilize barium sulfate precipitate
Checked and adjusted pH before testing
Shook reaction tube either gently or vigorously
Reacted and measured sulfate in same tube or reacted in one tube and transferred to a cuvette for
measurement
Used distilled water for a blank or used an unreacted sample
Converted from light attenuation to sulfate by % transmission vs sulfate chart in older version of FM or
by built-in sulfate curves prepared by Hach or by other calculation
Made measurement at 360 nm or 420 nm or 470 nm
Diluted sample by 1:9 or 1:99 when needed; diluted in a flask or in a reaction tube
assure that an observed sulfate concentration of 200 ppm did not exceed a rejection limit of 1000
ppm, with the caveat that statistics generated in this study were limited to a sulfate standard
solution and did not include variability from soil extraction, soil heterogeneity, or field sampling,
transport, and storage of soil. The accuracy of the method could be improved by reducing
apparent systematic errors.
6.5 Summary
Results of laboratory visits suggested that for pH, analysis of more than two independent
samples may be necessary to distinguish an observed pH at the acceptance limits of pH 5 or 9
from the corresponding pH rejection limits of pH 4 or 10. For minimum resistivity, chloride, and
sulfate, two independent samples were adequate to distinguish an observation at the acceptance
limit from the rejection limit. The caveats for these results were (1) hypothesis testing included
Type I () error but not Type II () error; and (2) statistics generated in this study were limited to
a test soil and standard solutions and did not include variability from soil extraction, soil
heterogeneity, or field sampling, transport, and storage of soils. The accuracy of all methods
could be improved by reducing apparent systematic errors.
79
7 Soil Sampling of Select Backfill
7.1 Introduction
In this chapter were described field sampling protocols for soils collected for ruggedness,
treatment, and other factor analysis studies (Chapters 8 and 9), where field sampling was used in
the broadest sense to include sample collection, transport, and storage. Also in this chapter were
summarized the geotechnical properties of the all soils tested and the results of split sample
analyses.
7.2 Field Sampling of Select Backfill
In addition to select backfill collected from bins at FDOT’s SMO, samples of backfill material
were also collected from eight mines, seven of which were in FDOT’s Districts 1 and 7 and one
of which was in District 5. Refer to Table 1-1 (Chapter 1, Section 1.2.7) for summary of the
mine owners, locations, and sampling dates.
The Transcor Dirt Services mine (Mine 1) was an open pit with fine sands that ranged in color
from white to reddish yellow to dark brown. A backhoe operator dug 12 holes about 3-m deep in
a grid pattern with holes spaced roughly 50 m apart. Sand samples of ~25 kg were collected in
buckets from each of the excavated dirt piles, and these buckets were transported to a nearby
nylon tarp. Sands in the buckets were emptied onto the tarp and mixed by rolling the tarp. Once
the sands were mixed, subsamples were acquired and divided into buckets or sealable plastic
bags for transport under ambient conditions or in a cooler on ice, respectively, to the USF soils
laboratory. One ~3-kg sub-sample was placed in sealable plastic bag for on-site analyses and
two ~3-kg subsamples were placed in sealable plastic bags and transported in coolers on ice for
split-sample analyses by local laboratories.
The Jahna Industries (Mine 2), Youngquist Brothers Rock (Mine 3), Angelo’s Recycled Material
(Mine 5), Titan (Mine 7) and Cemex (Mine 8) mines had stockpiles of backfill; the Calhoun
(Mine 4) and Hickey Excavating (Mine 6) mines were sand pits but the sampling protocol was
the same. Mine 2 and Mine 7 samples were classified by the mine owners as “mason’s sand”;
Mine 8 samples came from a tailings stockpile. At each of these mines an equipment operator
removed one front loader bucket of sand from three separate locations near the base of the
stockpile or pit. The three loads were kept separate but placed one next to the other. Four sand
samples of ~12 kg were collected in buckets from each of the excavated piles, and these buckets
were carried to a nearby nylon tarp. Sands in the buckets were emptied onto the tarp and mixed
by rolling the tarp. Once the sands were mixed, subsamples were acquired and divided into
buckets or sealable plastic bags for transport under ambient conditions or in a cooler on ice,
respectively, to the USF soils laboratory. One ~3-kg sub-sample was placed in a sealable plastic
bag for on-site analyses and one or two ~3-kg subsamples were placed in sealable plastic bags
and transported in a cooler on ice for split-sample analyses by an FDOT or commercial
laboratory. For samples from Mines 1 through 4, on-site analyses included measurement in
triplicate of backfill pH and resistivity; for Mines 5 through 8, pH and conductivity. Samples
80
brought back to the USF soils laboratory were processed for geotechnical testing and ruggedness,
treatment and other factor analysis studies, with subsamples separated for storage in a
refrigerator or at room temperature and air- or oven-dried, as necessary, for each study.
7.3 Geotechnical Properties of Select Backfill
Geotechnical testing was performed in the USF soils laboratory. The methods employed were
AASHTO T88 for particle size analysis (not including the hydrometer test), AASHTO T89 for
liquid limit, AASHTO T90 for plasticity limit (hand rolling technique), AASHTO T265 for
moisture content, and AASHTO T267 for organic matter content. Refer to Table 7-1 for a
summary of the geotechnical properties of the sand mine samples. The names given to backfill
samples were for project use only and were not related to U. S. Geological Survey or U. S.
Department of Agriculture soil names. All of the listed samples were classified as sand
(AASHTO soil classification A-3) and all samples met the geotechnical requirements for select
backfill (Chapter 2, Section 2.1.2).
Table 7-1 Geotechnical Properties of Select Backfill Samples

Mine ID Backfill Name Liquid Limit Plasticity Index % Passing #200 Sieve % Organic Matter
SMO Starvation Hill NP NP 2.00 0.37
SMO Santa Fe NP NP 3.03 0.67
1 Wimauma NP NP 2.10 0.53
2 Jahna NP NP 2.21 0.08
3 Youngquist NP NP 5.81 0.60
4 Calhoun NP NP 3.41 0.34
5 Angelo’s NP NP 6.10 1.07
6 Sebring NP NP 4.64 0.47
7 Clermont NP NP 0.74 0.07
8 Alico Road NP NP 3.40 0.81
81
7.4 Split Sample Analyses
Results from split sample analyses were shown in Table 7-2. The USF results were the averages
obtained across two or more treatments (Chapter 9, Section 9.1).
Table 7-2 Results for Laboratory Analysis of Split Samples

Laboratory pH Minimum Resistivity, ohm-cm Soil Chloride, ppm Soil Sulfate, ppm
Mine 1Wimauma
Laboratory 2 4.6 42,000 30 21
Laboratory 5 4.9 42,000 15 6
USF* 4.7 44,700 BD 10
Mine 2 Jahna
Laboratory 4 7.8 19,000 85 73
USF* 5.2 116,000 BD BD
Mine 3 Youngquist
Laboratory 1 8.1 6,100 30 118
Laboratory 2 8.0 5,500 31 81
USF* 8.2 9,150 BD 36
Mine 4 Calhoun
Laboratory 3 5.1 31,000 9 BD
USF* 4.6 104,000 BD BD
Mine 5 Angelo’s
Laboratory 6 5.6 27,000 113 30
USF* 4.6 17,800 BD 24
Mine 6 Sebring
Laboratory 3 5.0 38,000 6 12
USF* 4.4 36,100 BD 16
Mine 7 Clermont
Laboratory 6 8.0 19,700 34 21
USF* 7.5 28,200 BD BD
Mine 8 Alico Road
Laboratory 1 8.9 24,000 15 79
USF* 9.0 11,200 BD BD
*Results were averaged across all treatments (Chapter 9, Table 9-1) and rounded for ease of comparison.
BD = below detection
For pH, agreement between laboratories was within 0.5 pH units for Wimauma, Youngquist,
Calhoun, Clermont, and Alico Road sands, within 0.6 pH units for Sebring sand, 1 pH unit for
Angelo’s sand, and 2.6 pH units for Jahna sand (Table 7-2). For divergent pH values, USF
results were lower than partnered laboratories (Table 7-2). For samples for Mines 5 through 8,
USF trialed the use of KCl addition to soil solutions prepared for pH measurement. The addition
of KCl tends to lower the sample pH by as much as 0.5 pH units. Among the eight soils, four
soils had at least one pH result below the acceptance limit of pH 5.
82
For minimum resistivity, between-laboratory agreement was reasonable for three sands:
Wimauma, Youngquist, and Sebring (Table 7-2). For Jahna and Calhoun sands, USF reported
much higher values than the partnered laboratory (Table 7-2). USF trialed a change to the FM
between Mines 1 through 4 and Mines 5 through 8; for samples from the latter four mines the
water was decanted first into soil box from the soil slurry that formed with the addition of water.
This change brought the between-laboratory results at least to the same measurement scale. No
reported minimum resistivity approached the acceptance limit of 3000 ohm-cm.
For chloride, USF reported that all concentrations were below detection. For the commonly-
used Hach chloride test kits, the stated method detection level is 5 ppm (15 ppm if multiplied by
a dilution factor of 3 to obtain soil concentration). For these chloride measurements, USF trialed
two changes to the FM: the first change was to remove suspended particles and most, if not all,
of the color in the sand extract prior to testing for chloride; the second change was to test the
dilution water and to subtract the dilution water chloride concentration from the chloride
concentration of the sample extract. Reducing the potential for interference may explain in part
the below-detection results obtained by USF. From Table 7-2, between-laboratory chloride
concentrations diverged for Jahna and Angelo’s sands, and one result for Angelo’s sand
exceeded the acceptance limit of 100 ppm.
For sulfate, USF reported that four concentrations were below detection. For the Hach
photometer kit, which was not in routine use, the stated detection level is 2 ppm (6 ppm if
multiplied by a dilution factor of 3 to obtain soil concentration). For these sulfate measurements,
USF trialed a change to the FM to remove suspended particles and most, if not all, of the color in
the sand extract prior to testing for sulfate. Reducing the potential for interference may explain
in part the below-detection results obtained by USF. From Table 7-2, agreement in sulfate
concentration between laboratories was reasonable for Wimauma, Calhoun, Angelo’s, Sebring,
and Clermont sands but diverged for Alico Road, Youngquist, and Jahna sands. No reported
sulfate concentration exceeded the acceptance limit of 200 ppm.
Differences in method procedures between laboratories aside, across eight sands, four sands—
Wimauma, Calhoun, Angelo’s, and Sebring—failed to meet the acceptance limit for one or more
of the corrosion properties. For Wimauma sand, three laboratories agreed on the failed
acceptance limit, but for the other three sands, the failure was in dispute. If the split sample
analyses between two laboratories represented quality control and verification tests, at least three
out of the eight sands would require resolution testing. Reduction in a relatively high rate of
resolution testing may offset additional costs associated with method improvements.
7.5 Summary
Changes to the FMs for minimum resistivity, chloride, and sulfate that were trialed by USF
showed promise for improving between-laboratory agreement. Split sample analyses revealed
that four of the eight sands failed to meet an acceptance limit for one more of the corrosion
83
properties in one or more of the partnered laboratories. If the split sample analyses between two
laboratories represented quality control and verification tests, at least three out of the eight sands
would require resolution testing. Reduction in a relatively high rate of resolution testing may
offset additional costs associated with method improvements.
84
8 Ruggedness Studies
8.1 Overview
A ruggedness study is a tool to examine how robust a method is to small changes in equipment,
procedures, materials, and environmental factors, as examples, which simulates differences in
method performance that might be expected between laboratories and anticipates the method
precision or reproducibility. Ideally, a ruggedness study precedes release of a new or revised
method for an inter-laboratory study; in this case, however, the ruggedness study also serves as a
baseline for evaluating potential changes to a current method.
Ruggedness studies were first conducted with Santa Fe River soil prior to laboratory visits
(Chapter 6) to explore sensitivity of results to anticipated variations in method procedures
(Section 8.2). Soil samples were dried, sieved, and stored prior to use and as such did not
represent fully soils that were freshly acquired from a pit but serve better as examples for later
studies.
Transport, storage, and processing conditions were included as factors for the second round of
ruggedness studies (Section 8.3). Refer to Chapter 7, Section 7.2, for a description of the
sampling protocols. These studies followed the laboratory visits (Chapter 6) and were expanded
to evaluate procedural differences between laboratories, between AASHTO and FMs, and
between current and proposed FMs. The ruggedness study worked well for identifying changes
in procedural steps that have a strong influence on the results, but at the same time, the presence
of a strong influence worsened the method’s relative precision and bias, and in effect, swamped
the roles of less influential factors. Factor analysis studies such as the treatment studies (Chapter
9) explored further changes to procedural steps that had a strong influence.
For all ruggedness studies, statistical significance of factor effects was determined by t-tests for
averages and F-tests for standard deviations between related factors at the 95% confidence level
(p < 0.05).
8.2 Ruggedness Study for Santa Fe River Soil
8.2.1 Summary results
Summary statistics for ruggedness studies with Santa Fe River soil are presented in Table 8-1
and are discussed in more detail with respect to each method. Sodium chloride and sodium
sulfate were added to the soil to achieve detectable chloride and sulfate concentrations but still
have a minimum resistivity above 3,000 ohm-cm; the pH ruggedness studies were done for soils
with and without added salt.
85
Table 8-1 Summary Statistics for Ruggedness Studies with Santa Fe River Soil
Unmodified Amended with 0.10 g/kg NaCl and 0.10 g/kg Na2SO4
Statistic Min Resistivity Chloride Sulfate
pH pH
(ohm-cm) (ppm) (ppm)
Average 8.09 7.91 3,540 50 142
St Dev 0.18 0.13 290 12 62
%RSD 2.27 1.70 8.20 23.5 44.0
%RE 6.18 4.30 22.6 63.0 145
8.2.2 FM for pH
The pH ruggedness study included both refrigerated and non-refrigerated samples to evaluate the
effect of temperature change. To evaluate the effect of ionic strength, the equilibration time was
varied, as was the water-to-soil ratio, the use of distilled versus deionized water (distilled water
had a higher conductivity), and the degree of sample stirring; moreover, the study was repeated
with salt amendment, which raised the sample ionic strength. To evaluate the effect of electrode
condition, two electrodes with acceptable but different offsets were tested. Also tested was the
measurement geometry (height of water above the sample) between a beaker and flask. Factors
and pH determinations are presented in Table 8-2.
To assess the effect of a change from the original, for each factor the average of the change is
subtracted from the average of the original (Youden, 1975). With this approach and according to
the test matrix, all other factors appear twice on each side of the calculation and are thus
canceled out. For example, in the measurement of pH (Table 8-2), the effect of using 100 g
instead of 100 mL of unmodified soil is calculated as ((7.99 + 8.01 + 7.90 + 8.18) / 4) - ((7.94 +
8.40 + 8.00 + 83.1) / 4) = - 0.14; interestingly, the effect is lessened with the addition of salt to
the sample: ((7.86 + 7.98 + 7.75 + 8.00 + 8.09) / 4) - (8.09 + 7.75 + 8.04 + 7.81 )/4) = -0.025.
To estimate the method reproducibility sR, the standard deviation of the eight results was
calculated. For pH, the standard deviations were 0.18 pH unit and 0.13 pH unit for unmodified
and salt-amended soils, respectively (Table 8-1).
86
Table 8-2 Ruggedness Study for pH
Factors
Determination
Original Change 1 2 3 4 5 6 7 8
A a ML ML ML ML G G G G
B b RT RT C C RT RT C C
C c 30 0 30 0 30 0 30 0
D d DS DS DI DI DI DI DS DS
E e 1 2 1 2 2 1 2 1
F f GS NS NS GS GS NS NS GS
G g wet dry dry wet dry wet wet dry
pH (unmodified;
7.99 8.01 7.90 8.18 7.94 8.40 8.00 8.31
wetted/dry soil)
pH (salt-amended;
7.86 7.98 7.75 8.00 8.09 7.75 8.04 7.81
beaker/flask)
A 100 mL soil (ML)
a 100 g soil (G)
B Room temperature sample (RT)
b Cold sample (taken directly from the refrigerator at 4 oC) (C)
C 30-min equilibration (30)
c No equilibration (0)
D Mixed with distilled water (DS)
d Mixed with deionized water (DI)
E pH electrode 1 (1)
e pH electrode 2 (2)
F Gently stirred (GS)
f Not stirred (NS)
G Wetted overnight (wet) or tested in beaker
g Kept dry (dry) or tested in flask
For the unmodified soil, the pH average, standard deviation, %RSD, and %RE were 8.09, 0.18,
2.27%, and 6.18%, respectively (Table 8-1); for the salt-amended soil they were 7.91, 0.13,
1.70%, and 4.30%. Note that the addition of sodium salts lowered the average pH by 0.18 pH
units. The effect of factors on pH measurement was summarized in Table 8-3 and Table 8-4 for
unmodified and salt-amended Santa Fe River soils, respectively. For the unmodified soil, the pH
average was significantly higher when pH was measured without delay than after a 30-min
equilibration (Table 8-3). One explanation is that a 30-min equilibrium period raised the
solution ionic strength compared with no equilibrium period. For the unmodified soil, no other
factors had a significant effect either on the average or standard deviation soil pH (Table 8-3).
For the salt-amended soil, the pH average was significantly higher when pH was measured with
electrode 2 rather than electrode 1, possibly due to differing responses of the electrode liquid
87
junctions to ionic strength. For the salt-amended soil, no other factors had a significant effect
either on the average or standard deviation soil pH (Table 8-4).
Table 8-3 Effect of Factors in pH Ruggedness Study for Unmodified Santa Fe River
Soil
Factors Avg St Dev Diff F-tests t-tests pH was higher if sample
A 8.020 0.117
Water-to-soil ratio was higher
a 8.163 0.227 -0.143 0.306 0.306
B 8.085 0.212
Was colder
b 8.098 0.183 -0.012 0.814 0.932
C 7.958 0.046
Measured without delay
c 8.225 0.169 -0.267 0.061 0.023
D 8.078 0.155
Diluted with deionized water
d 8.105 0.232 -0.027 0.525 0.850
E 8.150 0.242
Measured with electrode 1
e 8.033 0.103 0.118 0.194 0.407
F 8.105 0.171
Was gently stirred
f 8.078 0.221 0.027 0.688 0.850
G 8.143 0.193
Wetted overnight
g 8.040 0.186 0.102 0.953 0.473
Table 8-4 Effect of Factors in pH Ruggedness Study for Salt-Amended Santa Fe River
Soil
Factors Avg St Dev Diff F-tests t-tests pH was higher if sample
A 7.898 0.116
Water-to-soil ratio was higher
a 7.923 0.168 -0.025 0.562 0.814
B 7.920 0.147
Was warmer
b 7.900 0.142 0.020 0.951 0.851
C 7.935 0.158
Measured after 30 min
c 7.885 0.124 0.050 0.700 0.636
D 7.923 0.106
Diluted with distilled water
d 7.898 0.174 0.025 0.436 0.814
E 7.793 0.053
Measured with electrode 2
e 8.028 0.049 -0.235 0.886 0.001
F 7.940 0.128
Was gently stirred
f 7.880 0.152 0.060 0.787 0.569
G 7.913 0.133
Was tested in a beaker
g 7.908 0.156 0.005 0.801 0.963
88
Recall that for Santa Fe River sand tested during laboratory visits the %RSD and %RE for pH
were 10.1% and 32.4%, respectively (Chapter 6, Section 6.4.1, Table 6-1) compared with the
ruggedness study results shown in Table 8-3 of 2.27% and 6.18% for the same statistics,
respectively. Thus, this ruggedness study for pH did not introduce the variability in pH seen in
the laboratory visits, which was likely dominated by electrode failure.
In summary, the FM for pH was sensitive to a 30-min equilibration of the soil versus no
equilibration period; when the soil was amended with salt this sensitivity disappeared and the
improvement in precision and bias revealed a significantly different pH response by electrode.
These results suggested that soil ionic strength, equilibration period, and electrode condition
were factors important method bias and precision.
The ruggedness study for resistivity included both refrigerated and non-refrigerated samples to
evaluate the effect of temperature change. To evaluate the effect of ionic strength, an overnight
wetting period was introduced, as were changes in the water-to-soil ratio and the use of distilled
versus deionized water (distilled water had a higher conductivity). Matched and unmatched test
leads were compared as unmatched leads likely add resistance to a balanced bridge resistance
measurement; a change in incremental water volume was compared as smaller volumes might
better locate the minimum resistance; and small and large soil boxes were compared as a larger
box might better describe soil heterogeneity. Factors and minimum resistivity determinations
were presented in Table 8-5.
89
Table 8-5 Ruggedness Study for Minimum Resistivity
Factors Determination
A a 1000 1000 1000 1000 750 750 750 750
B b RT RT C C RT RT C C
C c 100 75 100 75 100 75 100 75
E e SM LG SM LG LG SM LG SM
F f dry wet wet dry dry wet wet dry
G g M UM UM M UM M M UM
Minimum Resistivity
3500 3250 3900 3400 3250 3500 4050 3450
(ohm-cm)
A 1000 g soil (1000)
a 750 g soil (750)
B Room temperature sample (RT)
b Cold sample (taken directly from the refrigerator at 4 oC) (C)
C 100 mL water increments (100)
c 75 mL water increments (75)
D Mixed with distilled water (DS)
d Mixed with deionized water (DI)
E Small soil box (SM)
e Large soil box (LG)
F Dry soil (dry)
f Soil wetted overnight (wet)
G Matched wire (M)
g Unmatched wires (UM)
For this ruggedness study, the soil minimum resistivity average, standard deviation, %RSD, and
%RE were 3,540 ohm-cm, 290 ohm-cm, 8.20%, and 22.6%, respectively (Table 8-1). From the
resistivity ruggedness study, the effects of factor perturbations were seen in the differences of
average resistivity between the original and change, which were summarized in Table 8-6 for
salt-amended Santa Fe River soil. No factors had a significant effect either on the average or
standard deviation soil minimum resistivity (Table 8-6).
90
Table 8-6 Effect of Factors in Minimum Resistivity Ruggedness Study for Salt-
Amended Santa Fe River Soil
Factors Avg St Dev Diff F-tests t-tests Minimum resistivity was higher if
A 3,513 278
Mass (weight) of sample was lower
a 3,563 342 -50 0.740 0.828
B 3,375 144
Sample was colder
b 3,700 324 -325 0.217 0.117
C 3,675 366
For 100 mL increments of water
c 3,400 108 275 0.075 0.200
D 3,563 342
Diluted with distilled water
d 3,513 278 50 0.740 0.828
E 3,588 210
Measured with small soil box
e 3,488 382 100 0.352 0.662
F 3,400 108
Sample was wetted overnight
f 3,675 366 -275 0.075 0.200
G 3,613 295
If test leads were matched
g 3,463 307 150 0.953 0.507
Recall that from the laboratory visits (Chapter 6, Section 6.4.2, Table 6-3) for Santa Fe River
sand the %RSD and %RE for minimum resistivity were 11.9%, and 41.7%, respectively,
compared with the ruggedness study results shown in Table 8-5 of 8.2% and 22.6% for the same
statistics, respectively. This ruggedness study for minimum resistivity did not capture the
variability seen in laboratory visits.
Factors for a ruggedness study for chloride included changes from the original FM in the amount
of chloride tested (soil mass, amount of sample), sample clarity (settling time, use of membrane
filtration, portion of extract that was filtered), and potential interferences (distilled versus
deionized water and stirring technique). Factors and chloride concentration determinations are
presented in Table 8-7.
91
Table 8-7 Ruggedness Study for Soil Chloride Concentration
A a 100 100 100 100 75 75 75 75
B b 24 24 48 48 24 24 48 48
C c HI LO HI LO HI LO HI LO
E e N Y N Y Y N Y N
F f ALL 100 100 ALL ALL 100 100 ALL
G g HS MS MS HS MS HS HS MS
Chloride Concentration
68 57 50 65 42 43 41 36
(ppm)
A 100 g dried, sieved soil (100)
a 75 g dried, sieved soil (75)
B 24 hr of settling (24)
b 48 hr of settling (48)
C High range for 0.22 N AgNO3 test (40 mL sample) (HI)
c Low range for 0.22 N AgNO3 test (100 mL sample) (LO)
D Distilled water (DS)
d Deionized water (DI)
E No membrane filtration (N)
e Filtration with a 0.45 micron mixed cellulose ester filter (Y)
F Filter all of the extract (ALL)
f Filter the first 100 mL of extract (100)
G Hand swirl sample (HS)
g Machine stir sample (MS)
For this ruggedness study, the soil chloride concentration average, standard deviation, %RSD,
and %RE were 50 ppm, 12 ppm, 23.5%, and 63.0%, respectively (Table 8-1). Only one factor—
a change in soil mass—had a significant effect on average chloride concentration (Table 8-8).
The sensitivity of the method to soil mass suggests that the accuracy of chloride analysis would
be worsened for measurements made in the field versus in a laboratory, because in the field soil
mass would vary with volumetric measures and field moisture content. No factors had a
significant effect on the standard deviation of chloride concentration (Table 8-8).
92
Table 8-8 Effect of Factors in Chloride Ruggedness Study for Salt-Amended Santa Fe
River Soil
Factors Avg St Dev Diff F-tests t-tests Chloride concentration was higher if
A 60 8
Mass (weight) of sample was higher
a 40 3 19 0.141 0.004
B 52 12
Settling time was lower
b 48 13 5 0.967 0.613
C 50 12
(No difference based on sample size)
c 50 13 0 0.938 0.984
D 50 15
(No difference based on water type)
d 50 10 1 0.594 0.952
E 49 14
Filtered through a membrane
e 51 12 -2 0.814 0.825
F 53 16
All the extract was filtered
f 47 7 5 0.246 0.584
G 54 14
Titrated sample was hand-swirled
g 46 9 8 0.488 0.396
In the execution of the ruggedness study for chloride concentration, a few problems with the
digital titrator were noted. First, the titrator mechanism that facilitates a quick removal of the
titrant cartridge did not function, which made cartridge removal a very time-consuming task.
The titrant degrades faster if exposed to light, and the titrant cartridge should be capped and
returned to its box after each use to prevent this exposure. In lieu of this, the installed cartridge
was covered with aluminum foil. Second, at times when dialing the digits from 9 to 10 or 99 to
100, the digit in the tenth or 100th place, respectively, did not turn over, causing an
undercounting of digits by at least 10 digits. Finally, the delivery tube seemed to leak a small but
continuous stream of titrant, which grew stronger with stirring. This too, would cause an
undercounting of digits and results in a lower estimate of the soil chloride concentration.
Combined, these problems contributed to noise in the chloride analyses.
Recall that from the laboratory visits (Chapter 6, Section 6.4.3, Table 6-5) for a 30-ppm aqueous
chloride standard, the %RSD and %RE for chloride concentration were 35.1%, and 115%,
respectively, compared with the ruggedness study results shown in Table 8-7 of 23.5% and
63.0% for the same statistics, respectively. Thus, this ruggedness study for chloride did not
capture the variability seen in the laboratory visits.
In summary, the FM for chloride was sensitive to soil mass tested, which suggested that drying
the soil to a constant mass is an important method step. Also noted were issues with the digital
titrator that make it unsuitable for a standard test method.
93
Factors for a ruggedness study for sulfate included to changes from the original FM in the
amount of sulfate tested (soil mass), sample clarity (settling time, use of membrane filtration,
acid addition, portion of extract that was filtered), and potential interferences (distilled versus
deionized water and timing of analysis). Factors and sulfate concentration determinations were
presented in Table 8-9. Soil samples were diluted by a factor of three (3:1 water-to-soil ratio),
but an additional factor of two dilution of the filtered soil extract was necessary to obtain a
reading on the Hach photometer, as the range on the custom calibration curve was from 0 to 40
ppm as sulfate. Thus, any error in the photometer reading was inflated by a factor of six.
Table 8-9 Ruggedness Study for Soil Sulfate Concentration

A a 100 100 100 100 75 75 75 75
B b 48 48 24 24 48 48 24 24
C c N Y N Y N Y N Y
E e N Y N Y Y N Y N
F f ALL 100 100 ALL ALL 100 100 ALL
G g 5 10 10 5 10 5 5 10
Sulfate Concentration (ppm) 228 132 204 132 138 168 132 114
A 100 g dried, sieved soil (100)
a 75 g dried, sieved soil (75)
B 24 hr of settling (24)
b 48 hr of settling (48)
C No membrane filtration (N)
c Filtration with a 0.45 micron mixed cellulose ester filter (Y)
D Distilled water (DS)
d Deionized water (DI)
E No added acid (N)
e Added acid (Y)
F Filter all of the extract (ALL)
f Filter the first 100 mL of extract (100)
G Test at 5 min (5)
g Test at 10 min (10)
For this ruggedness study, the soil sulfate concentration average, standard deviation, %RSD, and
%RE were 142 ppm, 62 ppm, 44.0%, and 145%, respectively (Table 8-1). Compounded by
dilution error, large swings in concentration seen for sample mass (Factors A, a), acid addition
(Factors E, e), and test interval (Factors G, g) contributed to high %RSD and %RE. Average
sulfate concentrations were not significantly higher for any factor, but sulfate concentration
94
standard deviation was significantly higher when soil extracts were not filtered through a
membrane (Table 8-10). These results emphasized the need to further clarify sample extracts
before sulfate testing. Although not statistically significant, another relatively large but expected
effect was noted for soil mass: sulfate concentration was 65 ppm higher when 100 g versus 75 g
were extracted (Table 8-10). The sensitivity of the method to soil mass suggested that the
accuracy of sulfate analysis would be worsened for measurements made in the field versus in a
laboratory, because in the field soil mass would vary with volumetric measures and field
moisture content.
Table 8-10 Effect of Factors in Sulfate Ruggedness Study for Salt-Amended Santa Fe
River Soil
Factors Avg St Dev Diff F-tests t-tests Sulfate concentration was higher if
A 174 49
Mass (weight) of sample was higher
a 109 62 65 0.725 0.153
B 138 86
Settling time was longer
b 146 40 -8 0.238 0.876
C 147 92
Not filtered through a membrane
c 137 23 10 0.046 0.836
D 152 52
Extracted with distilled water
d 132 78 20 0.514 0.688
E 179 50
No acid was added
e 105 55 74 0.880 0.092
F 124 84
First 100 mL of extract was filtered
f 159 34 -35 0.177 0.473
G 165 45
Tested at 5 min
g 118 74 47 0.436 0.325
With the Hach photometer, a custom calibration curve proved difficult to prepare and maintain
because (1) sulfate concentrations prepared in 5 ppm increments often had the same absorbance
reading, and (2) even with 10-ppm increments, curves were not linear over the target range.
Performance of the photometer’s built-in calibration curve is discussed in Chapter 9 (Section
9.2).
sulfate standard the %RSD and %RE for sulfate were 34.1%, and 146%, respectively, compared
with the ruggedness study results shown in Table 8-9 of 44% and 145% for the same statistics,
respectively. Thus, this ruggedness study for sulfate mimicked the variability seen in the
laboratory visits.
95
In summary, for the FM for sulfate the soil mass and extract clarity were factors that potentially
but not conclusively contributed to measurement bias and precision. Moreover, the accuracy of
the photometer-based measurement was limited due to the non-linearity of the calibration curve.
8.3 Ruggedness Studies for Mined Soils
8.3.1 Summary statistics
Ruggedness studies were accomplished for pH and resistivity on all eight sands, sulfate
concentration on four sands, but chloride concentration on no sands. Refer to Appendix A for
ruggedness study factors, determinations, and statistical significance, which were tabulated by
sand and by method. Transport, storage, and processing conditions were included as factors in
the ruggedness studies; note, however, that for pH and resistivity, refrigerated samples were
brought to room temperature before testing. Samples processed for sulfate analyses were oven-
dried and stored at room temperature. The overall average, standard deviation, % relative
standard deviation (%RSD), and % relative error (%RE) for each study were summarized in
Table 8-11. Note that these statistics were suggestive of the method reproducibility to the extent
that the chosen factors and their variability capture the variability that occurs between
laboratories.
96
Table 8-11 Summary Statistics for Ruggedness Study Results
Minimum
Chloride Sulfate
Statistic pH Resistivity, ohm-
Concentration, ppm Concentration, ppm
cm
Mine 1 Wimauma Sand
Avg 4.65 52,400 24
St Dev 0.19 7,790 13
Below detection
% RSD 4.05 14.9 55
% RE 11.0 42.0 163
Mine 2 Jahna Sand
Avg 4.72 147,000
St Dev 0.12 19,300
Not tested Not tested
% RSD 2.64 13.1
% RE 8.27 40.9
Mine 3 Youngquist Sand
Avg 8.12 6,530 45
St Dev 0.07 2,060 14
Not tested
% RSD 0.87 31.6 30
% RE 2.22 85.1 80
Mine 4 Calhoun Sand
Avg 4.48 72,300
St Dev 0.28 35,000
% RSD 6.24 48.4
% RE 14.1 107
Mine 5 Angelo’s Sand
Avg 4.64 17,500 31
St Dev 0.21 2,550 22
Not tested
% RSD 4.49 14.6 71
% RE 15.7 40.0 194
Mine 6 Sebring Sand
Avg 4.38 35,300 8
St Dev 0.09 7,030 2
Not tested
% RSD 1.98 19.9 30
% RE 8.4 58.1 80
Mine 7 Clermont Sand
Avg 7.40 22,700
St Dev 0.27 3,260
% RSD 3.72 14.4
% RE 11.6 35.3
Mine 8 Alico Road Sand
Avg 8.51 9,940
St Dev 0.37 1,510
% RSD 4.31 15.1
% RE 11.4 47.8
97
8.3.2 FM for pH
In the ruggedness study for pH, factors such as transport temperature, storage temperature, soil
moisture content, sample mass, equilibration techniques, electrode differences, and salt addition
were varied (Appendix A). Factor perturbations were tried on a minimum of two soils. For sands
from mines 5 through 8, each ruggedness study for pH was repeated, thus expanding the total
number of pH determinations from 8 to 16 and improving the sensitivity of the ruggedness study
with regard to factor effects (ASTM C1067). Except where amended with calcium chloride, pH
samples from mines 5 through 8 were amended with 0.1 g potassium chloride per 100 mL or 100
g of sample.
Recall that for Santa Fe River sand tested during laboratory visits the %RSD and %RE for pH
were 10.1% and 32.4%, respectively (Chapter 6, Section 6.4.1, Table 6-1) compared with the
ruggedness study results shown in Table 8-11, which ranged from 0.87% to 6.24% (average
3.54%) and from 2.22% to 15.7% (average 10.3%) for the same statistics, respectively. None of
the ruggedness studies for pH across all four sands mimicked the variability seen in the
laboratory visits, which was likely dominated by electrode failure.
From the pH ruggedness study results for eight sands (Appendix A), no significant differences
were detected in pH averages for samples that were
 Transported in a cooler (10-20C) versus at ambient temperatures (~30C);
 Either 100-mL versus 100-g or 100-mL versus 30-g, except for Alico Road sand;
 Stirred briefly at 10-min intervals over a 30-min versus a 60-min period, if stirred or not
stirred immediately before testing, if stirred vigorously or gently, or if stirred briefly at
10-min intervals over 30 min versus stirred once then allowed to sit for 30 min; and
 Tested with different electrodes, where each electrode was the same model 3-in-1
combination double-junction electrode but had a slightly different calibration slope and
offset, or in the case Youngquist sand between two refillable glass bulb electrodes, both
with temperature compensation, with differences in slopes and offset of 0.2% and 2.8
mV.
No significant differences were found in the standard deviations of all factors for Alico Road,
Calhoun, Clermont, Jahna, and Youngquist sands.
Significant differences were found in the standard deviations of all factors for Angelo’s sand and
Sebring sand (one replicate) (Appendix A, A-22, A-24, A-31) and these results may be due to
one or more outlier pH values in the datasets, for example, determination 3 in Table A-21
(Appendix A). What these two soils have in common is that they were tested with the same
model but different serial number electrodes, and each electrode had its own calibration. One
98
explanation is that in switching back and forth between electrodes, at some point the right
calibration was used with the wrong electrode. Another explanation is inadvertent carry-over of
pH 5 buffer into the sample. The consequence of these outliers was to dull the sensitivity of the
study to factor effects; this consequence was mitigated, however, because the experiments were
duplicated.
A significant difference was detected in pH average but not standard deviation in Calhoun sand
with and without the addition of 0.1 g of potassium chloride as an ionic strength adjustment
(Appendix A, A-18). No significant difference was detected in pH average or standard deviation
in Jahna sand with and without the addition of 0.1 g of potassium chloride
A significant difference was detected in pH average but not standard deviation in Wimauma sand
when comparing two different electrodes (Appendix A, Table A-2). Both electrodes had glass
bulbs but one was refillable and temperature-compensated and the other was disposable and
without temperature compensation; between the electrodes the differences in slope and offset
were at least 1.7% and 6.5 mV, respectively.
For Alico Road and Clermont sands, samples amended with calcium chloride had significantly
lower pH averages than those samples amended with potassium chloride (Appendix A, Tables A-
40, A-42, A-43, A-50). For Alico Road sand, this effect was seen only with 16 rather than 8
determinations.
For Alico Road, Clermont, and Sebring sands, storage temperature had a significant effect on
average pH, with a higher pH for cold storage of Alico Road and Sebring sands and a higher pH
for warm storage of Clermont sand (Appendix A, Tables A-33, A-34, A-43, A-50). For Alico
Road and Sebring sand, storage temperature effects were seen only with 16 rather than 8
determinations.
In summary, these results suggested that the FM for pH was most sensitive to the electrode
condition, to the ionic strength of the soil solution, to the choice of salt, if salt is used to increase
the ionic strength, and to storage temperature. The impact of these factors on pH is discussed
further in Chapter 9, Section 9.2.
In the ruggedness study for resistivity, factors considered included transport temperature, storage
temperature, “as-received” versus dried samples, sample mass, soil box size, increments of water
addition, dilution water conductivity, equilibration period, rinsing and topping off the box with
sample, and the amount of water of sample transferred to the box (Appendix A). Factor
perturbations were tried on a minimum of two soils.
Recall that from the laboratory visits (Chapter 6, Section 6.4.2, Table 6-3) for Santa Fe River
sand the %RSD and %RE for minimum resistivity were 11.9% and 41.7%, respectively,
compared with the ruggedness study ranges of 13.1% to 48.4% (average 21.5%) and 35.5% to
99
107% (average 57.0%) for the same statistics, respectively (Table 8-11). The ruggedness studies
for minimum resistivity across all four sands mimicked the variability seen in laboratory visits
and revealed an additional source of variability.
No significant differences were detected in either minimum resistivity averages or standard

deviations for samples that were
 Transported in a cooler (~10 - 20C) versus at ambient temperatures (~30C);
 Stored in a refrigerator (~4C) or at room temperature (~23C);
 Tested “as-received” or when air-dried and sieved;
 Not equilibrated or equilibrated overnight with 10% dilution water;
 Not equilibrated or equilibrated for 30 min with 20% dilution water;
 Placed in a soil box that was rinsed or not rinsed with dilution water between tests, or
placed at or just below the rim of the soil box;
 Either 1.0-kg or 0.5- kg samples or 1.0-kg and 1.5-kg samples;
 Altered in wiring with the P1 and C1 and P2 and C2 terminals on the resistivity meter
shorted; and
 Tested with water added in increments of either 100 mL or 50 mL.
Significantly lower minimum resistivity averages but not standard deviations were detected in
Calhoun and Youngquist sands when the soil slurry created by sequential water additions was
decanted to the soil box as mostly soil or as mostly water for studies (Appendix A, Tables A-14,
A-20). This effect was explored further in a factor analysis study and is discussed in Chapter 9
(Section 9.2).
A significantly higher standard deviation of minimum resistivity was detected in Angelo’s sand
(Appendix A, Table A-27) for testing with dilution water of >1,000,000 ohm-cm versus with
dilution water amended with sodium chloride such that the resistivity was 175,000 ohm-cm.
Significantly higher minimum resistivity averages were seen for Alico Road and Clermont sands
(Appendix A, Tables A-45, A-52) when these sands were tested with large soil box versus a
small soil box. A significantly higher standard deviation was seen in Clermont sand (Appendix
A, Table A-45) for tests using a small soil box but for Angelo’s sand (Appendix A, Table A-27)
these results were reversed. Both soil boxes were manufactured by McMiller and had an area-to-
100
length ratio of 1 cm; the capacities were 80 mL and 270 mL for the small and large box,
respectively.
In summary, these results suggested that the FM for minimum resistivity was most sensitive to
the amount of water included in soil slurry during the test procedure, to the size of the soil box,
and to the resistivity of the dilution water. The impact of these factors on minimum resistivity is
further explored in Chapter 9 (Section 9.2).
In the ruggedness study for sulfate, factors considered included transport temperature, storage
temperature, drying temperature, acid addition, extraction and filtration techniques, changes in
the calibration curve, and the use of blanks and sulfate spikes (Appendix A).
sulfate standard the %RSD and %RE for sulfate were 34.1% and 146%, respectively, compared
with the ruggedness study ranges of 30% to 70% (average 46.5%) and 80% to 194% (average
129%) for the same statistics, respectively (Table 8-11). The ruggedness studies for sulfate
across both sands mimicked the variability seen in the laboratory visits and revealed an
additional source of variability. For Wimauma and Youngquist sands, all sample extracts were
filtered through a 0.45-µm membrane prior to analysis for sulfate.
For Wimauma and Youngquist sands, no significant differences were detected in either
minimum resistivity averages or standard deviations for samples that were
 Transported in a cooler (~10-20C) and stored in a refrigerator (~4C) versus transported

at ambient temperatures (~30C) and stored at room temperature (~23C);
 Extracted with a 3:1 water-to-soil ratio versus a 1:1 water-to-soil ratio followed by post-
filtration dilution by a factor of three; and
 Filtered through a fast, coarse paper or centrifuged to remove suspended particles.
 Not spiked or spiked with a 20 ppm sulfate standard. Adding a spike can increase the
total sulfate concentration to a concentration above the method detection level; the
concentration of the spike is then subtracted from the total concentration to get the
sample sulfate concentration.
 Zeroed with a reagent blank versus an unreacted sample for a blank. The purpose of the
unreacted sample blank is to zero out any remaining color or suspended particles; that no
differences were measured between these blanks means that steps taken to clarify the
sample were successful.
101
A significant difference was detected in sulfate concentration average but not standard deviation
in Youngquist sand (Appendix A, Table A-16) for samples tested on either Range 1 or Range 2
using the manufacturer’s default sulfate calibration curve: Range 2 results were higher.
A significantly lower standard deviation but not average in sulfate concentration was detected in
Wimauma sand (Appendix A, Table A-6) for samples dried either at 60C or at 110C.
For Angelo’s and Sebring sands, no significant differences were detected in sulfate concentration
averages and standard deviations for samples that were
 Transported in a cooler (~10-20C) and stored in a refrigerator (~4C) versus transported

at ambient temperatures (~30C) and stored at room temperature (~23C);
 Dried at 60C versus 110C;
 Amended with or without a few drops of acid added to help settle suspended particles in
the soil extract;
 Filtered with a 0.45-micron membrane or not filtered after filtration with a coarse, fast
filter; and
 Not spiked or spiked with a 20-ppm sulfate standard.
Significant differences in both sulfate concentration average and standard deviation were seen
for Sebring sand (Appendix A, Table A-38) tested with different lot numbers. A significantly
higher standard deviation was seen for Angelo’s sand (Appendix A, Table A-29) when tested
with the photometer calibration curve on Range 2 versus Range 1.
In summary, these results suggested that the FM for sulfate was sensitive to the sulfate
calibration curve, reagent lot number, and possibly to the drying temperature. The impact of
these factors on sulfate concentration is further explored in Chapter 9 (Sections 9.1 and 9.2).
8.4 Summary
In summary, ruggedness study results suggested that factors important to method bias and
precision were for
 FM for pH: soil ionic strength, equilibration period, electrode condition, choice of salt if
amended with salt, and storage temperature;
 FM for minimum resistivity: amount of water versus soil decanted into the soil box
during the test procedure, volume of soil box, and resistivity of dilution water;
102
 FM for chloride: soil mass; and
 FM for sulfate: sulfate calibration curve, reagent lot number, and possibly drying
temperature.
For the FM for chloride, also noted were issues with the digital titrator that makes it unsuitable
for a standard test method.
103
9 Treatment and Other Studies
9.1 Treatment Studies
Treatment studies were conducted to assess the impact of transport, storage, and processing of
backfill samples on sample pH, minimum resistivity, chloride, and sulfate levels. Backfill
samples from mines 1 through 4 were tested with the current versions and mines 5 through 8
with proposed revised versions of the FMs. Refer to Chapter 11, Section 11.2.3, for a summary
of changes between these versions.
9.1.1 Summary statistics
In Table 9-1 were presented for each method and soil the average, standard deviation, %RSD,
and %RE across all treatments. These averages were included for USF in Chapter 7, Section 7.4,
Table 7-2. Refer to Appendix B for treatment details, method calibration data, and results. For
pH and minimum resistivity, three treatments were investigated: treatment 1 (on-site), on-site
analysis immediately after sampling; treatment 2 (ambient), transport under ambient conditions
(~30C), storage at room temperature (~23C), and testing at “as-received” soil moisture content;
and treatment 3 (cold), transport in a cooler on ice (~20C), storage in a refrigerator (~4C), and
testing at “as-received” soil moisture content. For some but not all soils, additional treatments
were investigated such “as-received” versus air-dried soils, or for soils dried at different
temperatures, as examples. For sulfate and chloride the primary interest was in whether the
drying temperature, either 60C or 110C, affected measured concentrations.
104
Table 9-1 Summary Statistics for Treatment Study Results
Minimum
Soil Chloride Soil Sulfate
Statistic pH Resistivity, ohm-
Concentration, ppm Concentration, ppm
cm
Mine 1 Wimauma
Avg 4.65 44,700 10
St Dev 0.11 3,010 2
Below detection
%RSD 2.26 6.74 16
%RE 9.67 24.6 30
Mine 2 Jahna
Avg 5.16 116,000
St Dev 0.17 8,210
Below detection Below detection
%RSD 3.21 7.09
%RE 11.8 25.9
Mine 3 Youngquist
Avg 8.23 9,150 36
St Dev 0.12 975 6
Below detection
%RSD 1.42 10.7 16
%RE 4.74 35.5 50
Mine 4 Calhoun
Avg 4.58 104,000
St Dev 0.08 12,000
%RSD 1.78 11.5
% RE 5.90 38.4
Mine 5 Angelo’s Sand
Avg 4.59 17,830 24
St Dev 0.03 1030 4
Below detection
% RSD 0.76 5.78 17
% RE 3.27 14.0 62
Mine 6 Sebring Sand
Avg 4.38 36,060 16
St Dev 0.06 3750 5
Below detection
% RSD 1.38 10.4 33
%RE 4.57 26.4 75
Mine 7 Clermont Sand
Avg 7.54 28,170
St Dev 0.17 3650
%RSD 2.23 13.0
%RE 8.62 32.0
Mine 8 Alico Road
Avg 8.99 11,170
St Dev 0.05 430
Below detection Below Detection
%RSD 0.58 3.88
%RE 2.00 13.4
105
9.1.2 FM for pH
For all mines, on-site pH measurements were made in triplicate, took ~one hr, and required at a
minimum a portable pH meter (AP85/13-62-AP55, with ATC), a flat surface, 2 L of dilution
water, and beakers, buffers, stirring rod, tissues, rinse bottle, and a 100-mL scoop. The portable
pH meter gave erratic readings during two site visits when the mid-morning temperature quickly
reached 30 oC, but readings stabilized once the instrument was cooled by shade.
9.1.2.1 Mines 1 through 4
For samples from mines 1 through 4, three treatments were investigated: treatment 1 (on-site),
on-site analysis with 1 hr after sampling; treatment 2 (ambient), transport under ambient
conditions (~30C), storage at room temperature (~23C), selection by quartering, and testing at
“as-received” soil moisture content; and treatment 3 (cold), transport in a cooler on ice (~20C),
storage in a refrigerator (~4C), selection by quartering, and testing at “as-received” soil
moisture content. An additional treatment (air-dried) was investigated for mines 1 through 3:
transport under ambient conditions (~30C), storage at room temperature (~23C), air-dried,
sieved through a No. 10 (2 mm) mesh, and sample selection by mechanically splitting. Although
the soil processing steps were varied between treatments, the analytical test protocol remained
the same: 100 mL of soil were diluted with 100 mL of deionized water, stirred for ~20 s at 10-
min intervals over 30 min, and tested with the same portable pH meter and compatible electrode
(AP85/13-62-AP55). Soils that were stored cold were brought to room temperature before
testing.
The pH repeatability across all treatments was the best for Youngquist sand with an average,
standard deviation, %RSD, and %RE of 8.23, 0.12, 1.42%, and 4.74%, respectively, and worst
for Jahna sand with an average, standard deviation, %RSD, and %RE of 5.16 pH units, 0.17 pH
units, 3.21%, and 11.8%, respectively (Table 9-1). Refer to Appendix B for the treatment
details, method calibration data, and results. Treatment study results for pH are summarized
below.
 For Wimauma, Jahna, and Calhoun, no significant differences were detected in pH

averages or standard deviations between on-site, ambient, and cold treatments.
 For Jahna sand, the average pH was significantly higher for air-dried treatment than for
ambient treatment; for Wimauma sand, pH was significantly higher for air-dried than for
on-site or ambient treatments. No significant differences were detected in pH standard
deviations.
 For Youngquist sand, average pH was significantly higher for cold treatment than for on-
site or ambient treatment; no significant differences were detected in pH standard
deviations.
106
 For Youngquist sand, average pH was significantly lower in air-dried sand than for
ambient or cold treatments. No significant differences were detected in pH standard
deviations.
These results suggested that the FM for pH may be sensitive to air-dried versus “as-received”
samples and possibly to samples transported and stored at cold versus at ambient temperatures.
For samples from mines 5 through 8, four treatments were investigated: treatment 1 (on-site), on-
site analysis within 1 hr after sampling; treatment 2 (ambient), transport under ambient
conditions (~30C), storage at room temperature (~23C), and testing at “as-received” soil
moisture content; treatment 3 (cold), transport in a cooler on ice (~10 - 20C), storage in a
refrigerator (~4C), and testing at “as-received” soil moisture content, and 4 (air dried) transport
under ambient conditions, storage at room temperature, dry at 60C, and sieved through a No. 10
mesh. Although the soil processing steps were varied between treatments the analytical test
protocol remained the same: 100 g of soil were amended with 0.1 g potassium chloride, diluted
with 100 mL of deionized water, and stirred for ~20 s at 10-min intervals over 30 min. In the
field, the samples were tested with a portable pH meter and compatible electrode (AP85/13-620-
AP55); in the laboratory, however, samples were tested with a benchtop meter (AB150/13-620-
631) as well as with the portable meter to compare the two instruments. Soils that were stored
cold were brought to room temperature before testing. Refer to Appendix B for the treatment
details, method calibration data, and results.
For Angelo’s sand, the average pH was significantly lower in dry soil for a benchtop meter
versus a portable meter, a likely consequence of differences in meter/electrode calibrations and
low standard deviations for both measurements.
For Sebring sand, average pH measurements were significantly higher for cold versus warm
measurements, consistent with ruggedness study results for the same soil (Chapter 8, Section
8.3.2).
For Clermont sand, there was a factor of 10 jump in pH standard deviations between on-site and
laboratory measurements. The lowest pH average occurred for sand transported and stored cold,
as compared with on-site measurements, consistent with results for a ruggedness study (Chapter
8, Section 8.3.2). In contradiction, the average pH measurement by portable pH meter of
samples transported and stored cold was higher than averages for all other treatments and
significantly higher than for cold and dry pH measurements made with a benchtop meter.
For Alico Road sand, dried and sieved soil had a significantly lower pH than for ambient
treatments as tested with both the portable and benchtop pH meters, which can be explained by
the removal of a significant fraction of particle mass in the sieving process. Of the 10 soils
tested in this project, Alico Road was the only soil that tested positive for the presence of
107
carbonates (a few drops of 10% hydrochloric acid on ~1 gram of sample). The fraction retained
on the No. 10 sieve was rich in carbonates and its removal diminished the observed pH. Refer to
Section 9.2.10 for further discussion. For this soil, sieving shifted the pH from not acceptable to
acceptable based on FDOT’s criterion that select backfill have a pH ≤ 9.0. This finding offers a
compelling argument to test pH samples “as received.”
If the soil pH were sensitive to transport and/or storage temperature, under what conditions do
the laboratory measurements best duplicate on-site measurements? The square root of the sum
of squares for the differences between pH averages for on-site versus warm, cold, and dry soil
treatments yielded 0.22, 0.07, and 0.14 pH units, respectively; thus the least error was seen
between soil on-site pH averages and pH averages for soil that had been transported cool and
stored cold.
The pH repeatability across all treatments was the best for Alico Road sand with an average,
standard deviation, %RSD, and %RE of 8.99, 0.05, 0.58%, and 2.00%, respectively, and for
worst for Clermont sand with an average, standard deviation, %RSD, and %RE of 7.54, 0.17,
2.23%, and 8.62%, respectively (Table 9-1). In summary, for one soil the average pH
measurement was significantly lower after the soil was dried and sieved than for other
treatments, a likely consequence of carbonate minerals retained on the sieve. For one soil, the
average pH measurement was significantly lower for warm transport and storage than for cool
transport and cold storage; across all four soils the average pH observations were in reasonable
agreement between measurements made on-site and those made in the laboratory for soils
transported cool and stored cold.
On-site minimum resistivity measurements were made in triplicate, took ~two hr, and required a
portable resistivity meter with leads, a flat surface, 2 L of dilution water, soil box, large bowl,
spatula, and spoon. The resistivity meter did not have a temperature compensation feature.
For samples from mines 1 through 4, four treatments were investigated: treatment 1 (on-site), on-
site analysis within two hr after sampling; treatment 2 (ambient), transport under ambient
conditions (~30C), storage at room temperature (~23C), selection by quartering, and testing at
“as-received” soil moisture content, and treatment 3 (cold), transport in a cooler on ice (~20C),
storage in a refrigerator (~4C), selection by quartering, and testing at “as-received” soil
moisture content; a fourth treatment (air-dried) was transport under ambient conditions (~30C),
storage at room temperature (~23C), air-dried or dried at 60C, sieved through a No. 10 (2 mm)
mesh, and sample selection by mechanically splitting. Although the soil processing steps were
varied between treatments, the analytical test protocol remained the same: use of a Nilsson
Model 400 resistivity meter and leads, a large soil box, 1,000 g of soil, 100-mL increments of
added deionized water, and soil-rich slurry content in the soil box. Refer to Appendix B for the
108
treatment details, method calibration data, and results. Treatment study results for minimum
resistivity are summarized below.
 For Wimauma sand, minimum resistivity averages were higher for air-dried sand than for
on-site or cold treatments; no significant differences were detected in minimum
resistivity standard deviations.
 For Jahna sand, minimum resistivity averages were significantly lower for on-site
treatment than for ambient and cold treatments and significantly higher for ambient
treatment than for cold treatment or for air-dried sand. All three minimum resistivity
measurements for cold treatment were the same, so the standard deviation was zero and
significantly lower than for all other treatments.
 For Youngquist sand, on-site measurements were made with the resistivity meter
terminals shorted, P1 to C1 and P2 to C2, which yielded resistivity readings that were
~15% higher than when terminals were not shorted. No significant difference was
detected between the minimum resistivity average and standard deviation for on-site and
warm treatments when both treatments had the terminals shorted. Minimum resistivity
averages were significantly lower for cold treatment compared to warm treatment and for
cold treatment compared to air-dried sand, where terminals were in their normal
configuration (not shorted). Minimum resistivity standard deviation was significantly
lower for air-dried sand than for on-site, warm, or cold treatments.
 For Calhoun sand, minimum resistivity averages were significantly lower for on-site
treatment than for ambient and cold treatments, and significantly lower for air-dried sand
than for ambient and cold treatment; minimum resistivity standard deviations were
significantly lower for ambient treatment than for on-site and cold treatments.
The minimum resistivity repeatability across all treatments was the best for Wimauma sand with
an average, standard deviation, %RSD, and %RE of 44,700 ohm-cm, 3,010 ohm-cm, 6.74%, and
24.6%, respectively, and the worst for Calhoun sand with an average, standard deviation, %RSD,
and %RE of 104,000 ohm-cm, 12,000 ohm-cm, 11.5%, and 38.4%, respectively. These results
suggested that the FM for resistivity may be sensitive to on-site analysis of soil samples. For the
four sands tested, temperatures were 5 to 10C higher for on-site analyses than for laboratory
analyses; Equation 3-4 (Chapter 3, Section 3.3.1) predicted a 10% and a 17% decrease in
resistivity between 23C and 28C and between 23C and 33C, respectively, and predicted
fairly well the average difference in minimum resistivity between on-site and ambient treatments
for all four sands. No consistent pattern was seen where minimum resistivity was significantly
different between treatments other than on-site treatment.
The FM for resistivity was less compatible with field work, as the time, volume, and surface area
requirements were greater than for pH. Moreover, the method was sensitive to temperature
(Chapter 3, Section 3.3.1). An alternative approach to field measurement of resistivity is to
109
measure conductivity using portable conductivity meter with automatic temperature
compensation, as the inverse of conductivity is resistivity. The typical protocol for conductivity
measurements, however, dilutes the sample well beyond that of the soil box measurements.
Thus, field measurements of conductivity could serve as a screening tool but would not replace
the FM for resistivity.
For samples from mines 5 through 8, three treatments were investigated: treatment 1 (ambient),
transport under ambient conditions (~30C), storage at room temperature (~23C), and testing at
“as-received” soil moisture content; treatment 2 (cold), transport in a cooler on ice (~10 - 20C),
storage in a refrigerator (~4C), and testing at “as-received” soil moisture content; and treatment
3 (air dried), transport under ambient conditions, stored at room temperature, dried at 60C, and
sieved through a No. 10 mesh. Soils were brought to room temperature before testing. Although
the soil processing steps were varied between treatments, the analytical test protocol remained
the same: use of a Nilsson Model 400 resistivity meter and leads, a large soil box, 1,000 g of soil,
100-mL increments of deionized water, and water-rich slurry content in the soil box. Refer to
Appendix B for treatment details, method calibration data, and results.
Treatment study results for minimum resistivity are summarized below.
 For Alico Road and Angelo’s sands, the minimum resistivity average was significantly
higher for cold treatment than for warm or dry treatments,
 For Clermont sand, the minimum resistivity average was significantly lower for cold
treatment than for warm or dry treatment, and
 For Sebring sand, the minimum resistivity average was significantly higher for dry
treatment than for warm or cold treatments.
The minimum resistivity repeatability across all treatments was the best for Alico Road sand
with an average, standard, %RSD, and %RE of 11,190, 430, 3.88% and 13.8%, and the worst for
Clermont sand with an average, standard deviation, %RSD, and %RE of 28,170 ohm-cm, 3,650
ohm-cm, 13.0%, and 32.0%, respectively. Standard deviations were low or zero, for the most
part, which made it easier to diagnose differences between treatment means. Temperature did
not appear to be a contributing factor in the differences between treatment means. These results
suggested that variability in sample minimum resistivity between laboratories could be reduced if
laboratories processed samples in a similar manner, for example, with transport under ambient
conditions and samples stored at room temperature.
For pH measurements in sieved soils, the pH changed for soils with a significant mass retained
on the sieve; by inference in some soils under similar circumstances the minimum resistivity
could change. AASHTO T288 requires the soil be sieved for a minimum resistivity
110
measurement but provides guidance for separating and testing the fraction retained on a No. 10
sieve. Testing soils “as received” was recommended.
For chloride the primary objectives were to find the best technique for producing a clear sample
and to determine if drying temperature, either 60C or 110C, affected measured concentrations.
Although the soil processing steps were varied between treatments, the analytical test protocol
remained the same. A Hach Chloride Low Range Test Kit Model 8-P was used. A 23-mL
sample of filtered (clear) soil extract was amended with one packet of Chloride 2 reagent
(potassium dichromate and sodium bicarbonate) and titrated to a reddish color with 0.0493 N
silver nitrate using a calibrated dropper. A blank and a check standard were titrated along with
the sample. The number of drops (always one drop) required to titrate the blank was subtracted
from the number of drops required to titrate the sample and check standard. The blank-corrected
number of drops was then multiplied by 5 ppm per drop and by the dilution factors, which were
three for the soil extract and one for the check standard. Refer to Appendix B for the treatment
details, method calibration data, and results.
Wimauma sand was transported in a cooler on ice (~20C), stored in a refrigerator (~4C),
selected by quartering, dried at either 60C or 110C, and sieved through a No. 10 (2 mm) mesh.
A soil extract was prepared by diluting 100 g of soil with 100 mL of water; this extract was
shaken for 20 s allowed to settle for 1 hr, shaken again for 20 s, allowed to settle for 36 hr,
filtered through a coarse, fast filter (Fisher Q-8), and finally diluted by a factor of three. The
resulting extract was clear.
Jahna sand was transported under ambient conditions (~30C), stored at room temperature (~23
C), dried at either 60C or 110C, sieved through a No. 10 (2 mm) mesh, and selected by
mechanical splitting. A soil extract was prepared by diluting 100 g of soil with 100 mL of water;
this extract was shaken for 20 s allowed to settle for 1 hr, shaken again for 20 s, allowed to settle
for 24 hr, then diluted by a factor of three and filtered through a 0.45-micron mixed cellulose
ester membrane filter. The resulting extract was clear.
Youngquist sand was transported under ambient conditions (~30C), stored at room temperature
(~23C), dried at either 60C or 110C, sieved through a No. 10 (2 mm) mesh, and selected by
for 24 hrs, then filtered first through a coarse, fast filter (Whatman 4), second through another
coarse, fast filter (Fisher Q-8), and finally filtered through a 0.45-micron mixed cellulose ester
membrane filter. The resulting extract was clear.
Calhoun sand was transported under ambient conditions (~30C), stored at room temperature
(~23C), dried at either 60C or 110C, sieved through a No. 10 (2 mm) mesh, and selected by
111
for 24 hrs, then filtered through a coarse, fast filter (Fisher Q-8) and finally filtered through a
0.45-micron mixed cellulose ester membrane filter. The resulting extract was clear. For a third
treatment, Calhoun sand was transported in a cooler on ice (~20C), stored in a refrigerator
(~4C), dried 60C or 110C, sieved through a No. 10 (2 mm) mesh, and selected by mechanical
splitting. The rest of the soil processing was the same as for the first two treatments.
Angelo’s, Sebring, Clermont, and Alico Road sands were transported under ambient conditions
(~30C), stored at room temperature (~23C), dried at either 60C or 110C, sieved through a
No. 10 (2 mm) mesh, and selected by mechanical splitting. A soil extract was prepared by
diluting 300 g of soil with 100 mL of water; this extract was shaken for 20 s allowed to settle for
1 hr, shaken again for 20 s, allowed to settle for 24 hr, then filtered through a coarse, fast filter
(Fisher Q-8) and finally filtered through a 0.45-micron mixed cellulose ester membrane filter.
The resulting extract was clear. For Clermont sand two additional treatments included testing of
samples that were transported in a cooler on ice (~20C), stored in a refrigerator (~4C), dried
60C or 110C, sieved through a No. 10 (2 mm) mesh, and selected by mechanical splitting. The
rest of the soil processing was the same as for the first two treatments.
Chloride concentrations in all eight sands were below the stated method detection level of 5 ppm
(15 ppm if multiplied by a dilution factor of three to convert to a soil concentration). The
Youngquist sand was the most difficult of the four sands to clarify. Although a clear sample was
obtained with final filtration through a 0.45-micron membrane filter, the membrane filter was
quickly clogged and filtering enough extract to test took ~30 min.
For sulfate the primary objectives were to find the best technique for producing a clear sample
and to determine if drying temperature, either 60C or 110C, affected measured concentrations.
The soil processing steps were varied between treatments, but the analytical test protocol
remained the same. A Hach Photometer II Sulfate Test Kit was used. A 10-mL sample of
filtered (clear) soil extract in a test cell was amended one packet of SulfaVer 4 reagent (barium
chloride and citric acid), gently mixed, and allowed to sit for 5 min. A second 10-mL sample of
filtered extract was handled in like manner but no reagent was added to it. The photometer was
zeroed with an unamended sample and the sulfate concentration was read directly from the
photometer with the amended sample. Samples were tested on the built-in photometer sulfate
calibration curve labeled Range 1, except for Angelo’s sand, which was tested on both Range 1
and Range 2. This reading was multiplied by the dilution factor to get the soil sulfate
concentration. A blank of dilution water and a check standard were also prepared and tested.
Treatments by soil were as described for chloride in the previous section. Refer to Appendix B
for the treatment details, method calibration data, and results.
The sulfate repeatability across all treatments was the best for Wimauma sand with an average,
standard deviation, %RSD, and %RE of 10 ppm, 2 ppm, 16%, and 30%, respectively, and worst
112
for Sebring sand with an average, standard deviation, %RSD, and %RE of 16 ppm, 5 ppm, 33%,
and 75%, respectively (Table 9-1). Sulfate concentrations were below the stated method
detection of 2 mg/L (6 mg/L when multiplied by a dilution factor of three to get the soil
concentration) for Jahna, Calhoun, Clermont, and Alico Road sands.
 For Wimauma, Youngquist, and Angelo’s sands, no significant differences were detected
in sulfate averages or standard deviations where treatments varied only by the drying
temperature, which was either 60C or 110C.
 For Angelo’s sand, the average sulfate concentration was higher with sulfate calibration
curve Range 2 when compared with Range 1.
 For Sebring sand, the average and standard deviation sulfate concentration was
significantly higher for sand dried at 110C versus 60C.
The results for Sebring sand suggested that for some soils, the FM for sulfate is sensitive to
drying temperature. Since the moisture content is higher after drying at 60C rather than 110C,
a 100-g aliquot would contain less soil and thus less sulfate. That said, drying soil at a higher
temperature can cause gypsum present in some soils to lose its waters of hydration and to shift to
a more water-soluble mineral phase, thus increasing the observed sulfate concentration (Herrero
et al., 2009). The current FM allows drying the soil either at room temperature or at 110C. To
reduce variability in sulfate concentrations measured between laboratories, amending the FM to
allow drying the soil from room temperature up to 60C, consistent with the AASHTO T290
drying temperature, was recommended.
9.2 Other Factor Analysis Studies
9.2.1 Effect of ionic strength and pH electrode on pH measurement
To look further into the fixed effects on pH of ionic strength and electrode condition, a replicated
factor analysis experiment was conducted with two unmodified soils, Starvation Hill and Santa
Fe River, where each soil was diluted with and without 0.01 M calcium chloride (CaCl2), and
each dilution was tested with both pH electrode 1 and electrode 2. Both pH electrodes were
Accumet, with part numbers 2242017P and 8074017P for electrode 1 and electrode 2,
respectively. Offset voltages for calibration with pH 7 and pH 10 buffers were 1.8 mV for
electrode 1 and 25.4 mV for electrode 2.
Results showed that dilution of Starvation Hill and Santa Fe River soils with 0.01 M calcium
chloride shifted the average pH lower by 0.58 pH units, compared to a shift in average pH
between soils of 0.17 pH units and between electrodes of 0.14 pH units, where electrode 2 had a
higher pH (Figure 9-1, Table 9-2). Analysis of variance (ANOVA) showed that it was unlikely
that the average pH between soils, between electrodes, and between calcium chloride treatments
were equal (Table 9-3). Interactions were not significant, with the possible exception of the
113
interaction between pH electrode and amendment (Table 9-3), where the difference in average
pH was greater for calcium chloride-amended soils when pH was measured with electrode 2.
8.4
8.2
8.0
7.8
pH
7.6
7.4
7.2
7.0
Probe1 Probe2 Probe1 Probe2 Probe Amendment:
NoCaCl2
SantaFe StarvationHill SoilType YesCaCl2
Figure 9-1. Variability plot of pH by soil type, pH electrode, and CaCl 2 amendment.
114
Table 9-2 Comparison of pH Means by Soil Type, pH Electrode, and Calcium Chloride
Amendment
Descriptive Statistics (Statistica pH Spreadsheet)
Effect Level of Number of pH pH
Factor Samples Mean Std Dev
Total 16 7.72 0.33
Soil Type Santa Fe 8 7.80 0.33
Soil Type Starvation Hill 8 7.63 0.32
Electrode Electrode 1 8 7.65 0.29
Electrode Electrode 2 8 7.79 0.37
Amendment No CaCl2 8 8.01 0.17
Amendment Yes CaCl2 8 7.43 0.10
Table 9-3 ANOVA for pH by Soil Type, pH Electrode, and CaCl 2 Amendment
Fixed Effect Test for pH (Statistica pH Spreadsheet)
Restricted Maximum Likelihood (REML) Type V
decomposition
Effect Num. DF Den. DF F p
Soil Type 1 8 22.3 0.001
Electrode 1 8 16.2 0.004
Amendment 1 8 268 0.000
Soil Type*Electrode 1 8 0.005 0.946
Soil Type*Amendment 1 8 0.0199 0.891
Electrode*Amendment 1 8 5.10 0.054
Soil Type*Electrode*Amendment 1 8 0.318 0.588
To summarize, if the ionic strength of the soil solution were low, adding a small amount of salt
to the sample to increase its ionic strength could reduce the sensitivity of the pH measurements
to minor variations in technique. In such a case, however, the measured pH may shift lower by
more than 0.5 pH units, and the condition of the pH electrode may have a greater impact on the
measured pH.
9.2.2 Effect of incremental water volume and pre-wetting on minimum resistivity
To look further into the fixed effects on soil resistivity of incremental water volume and of pre-
wetted versus dry soil, a replicated factor analysis experiment was conducted with two salt-
115
amended soils: Starvation Hill and Santa Fe River, where for each a pre-wetted and dry soil
sample was tested with the addition of either 75 mL or 100 mL of deionized water volume
increments. The average soil resistivity was lowered by 156 ohm-cm with the addition of 75 mL
instead of 100 mL of water volume increments and lowered by 106 ohm-cm when the soil was
pre-wetted overnight prior to testing (Figure 9-2, Table 9-4). Analysis of variance (ANOVA)
showed that the hypothesis that the average resistivity between soils, between incremental water
volumes, and between pre-wetted versus dry soil were equal cannot be rejected (Table 9-5).
Interactions were not significant (Table 9-5).
3800
3700
3600
3500
Resistivity
3400
3300
3200
3100
3000
2900
75 100 75 100 Water_Vol_Incr Wet_Dry:
Dry
Santa Fe Starvation Hill SoilType Wet
Figure 9-2. Variability plot of soil resistivity (ohm-cm) by soil type, incremental
water volume (ml), and pre-wetted versus dry soil.
116
Table 9-4 Comparison of Soil Resistivity Means by Soil Type, Incremental Water
Volume, and Pre-Wetted versus Dry Soil
Descriptive Statistics (Statistica Resistivity Spreadsheet)
Effect Level of Number of Resistivity Resistivity
Factor Samples Mean (ohm-cm) Std Dev (ohm-cm)
Total 16 3,250 219
Soil Type Santa Fe 8 3,140 166
Soil Type Starvation Hill 8 3,350 225
Incr Water Vol 75 8 3,170 214
Incr Water Vol 100 8 3,330 207
Pre-Wetted vs Dry Dry 8 3,300 220
Pre-Wetted vs Dry Wet 8 3,190 218
Table 9-5 ANOVA for Soil Resistivity Means by Soil Type, Incremental Water Volume,
and Pre-Wetted versus Dry Soil
Fixed Effect Test for Resistivity (Statistica
Resistivity Spreadsheet) Restricted Maximum
Effect Likelihood (REML) Type V decomposition
Num. DF Den. DF F p
Soil Type 1 8 4.11 0.077
Incremental Water Volume 1 8 2.36 0.163
Pre-Wetted vs Dry Soil 1 8 1.09 0.327
Soil Type*Incr Water Vol 1 8 1.09 0.327
Soil Type*Pre-Wetted vs Dry 1 8 0.034 0.858
Incr Water Vol*Pre-Wetted vs Dry 1 8 0.457 0.518
Soil Type*Incr Water Vol*Pre-Wetted vs Dry 1 8 0.185 0.679
To summarize, a change in incremental water volume did not significantly affect measured soil
resistivity, so no change is recommended in the current practice of water volume increments of
100 mL. No significant difference was seen in soil resistivity between soil samples that were
initially dry or pre-wetted overnight, so this additional method step is not recommended for soils
that have been air-dried and sieved prior to testing.
9.2.3 Effects of extraction method and soil type on chloride measurements
To look further into the fixed effects of extraction method on chloride concentration, a replicated
factor analysis experiment was conducted with two soils: Starvation Hill and Santa Fe River.
Both soils were air-dried, sieved through a No. 10 sieve, and amended with 0.10 g/kg sodium
chloride and 0.10 g/kg sodium sulfate. The first of two extraction methods was 100 g soil
117
diluted with 100 mL of deionized water in a 500-mL flask, shaken for 20 s and allowed to settle
for 1 hr then shaken again for 20 s, allowed to settle overnight, filtered by gravity through a
coarse filter into a 500-mL flask and again allowed to settle overnight, and as a final step 50 mL
of extract was decanted into a 250-mL flask and diluted to 150 mL. The second of two
extraction methods was as described above, except that 100 g soil was diluted with 300 mL of
deionized water and as a final step 150 mL of extract was decanted and filtered through a 47-mm
diameter 0.45-micron membrane filter (Pall Metricel P/N 63069). Interestingly, the settling rate
of soil particles was faster and the resulting extract clearer for a 1:1 (mass-to-volume) soil-to-
water ratio. Both methods produced clear samples. The average chloride concentration was
higher by 2 ppm and its standard deviation was lower by 5 ppm for the 1:1 soil-to-water
extraction method (Extraction Method 1, Figure 9-3,Table 9-6) compared with membrane
filtration (Extraction Method 2), although neither a t-test for means (p = 0.65) nor an F-test for
variance (p = 0.07) were significant. Note that the variation in chloride concentrations for both
revised extraction methods were lower than the variation reported in Table 5-2 (Chapter 5,
Section 5.3) for salt-amended soil chloride concentration. Analysis of variance (ANOVA)
showed that the hypothesis that the average chloride concentration between soils and extraction
methods were equal cannot be rejected (Table 9-7). The interaction of soil type and extraction
method was not significant (Table 9-7).
118
85
80
75
70
Chloride
65
60
55
50
45 ExtractionMethod:
Santa Fe Starvation Hill 1
SoilType 2
Figure 9-3. Variability plot of chloride concentration (ppm) by soil type and
extraction method.
Table 9-6 Comparison of Soil Chloride Means by Soil Type and Extraction Method
Descriptive Statistics (Statistica Chloride Spreadsheet)
Effect Level of Number of Chloride Chloride
Factor Samples Mean (ppm) Std Dev (ppm)
Total 12 65.9 6.36
Extraction Method 1 6 66.8 3.54
119
Table 9-7 ANOVA for Soil Chloride by Soil Type and Extraction Method
Fixed Effect Test for Chloride (Statistica Chloride Spreadsheet)
Effect decomposition
Num. DF Den. DF F p
Soil Type 1 8 1.23 0.300
Soil Type*Extraction Method 1 8 1.04 0.337
In summary, laboratory rather than field testing for chloride concentration in soil is
recommended to accurately measure soil mass and to prepare a sample extract that is sufficiently
clear of particles and color that may interfere with the chloride analysis. Both extraction
methods produced clear sample extracts with no detectable difference in chloride concentration.
Further testing of these two extraction methods on other soils is recommended.
9.2.4 Effect of extraction method on sulfate measurement
To look further into the fixed effects of extraction method on sulfate concentration, a replicated
factor analysis experiment was conducted with two soils: Starvation Hill and Santa Fe River.
Soils were prepared and extracted as described in the previous section for chloride. The average
sulfate concentration was lower by 6 ppm and its standard deviation was lower by 4 ppm for the
1:1 soil-to-water extraction method (Extraction Method 1, Figure 9-4, Table 9-8) compared with
membrane filtration (Extraction Method 2), although neither a t-test for means (p = 0.29) nor an
F-test for variance (p = 0.30) were significant. Note that the variation in sulfate concentrations
for both revised extraction methods spanned the variation reported in Table 5-2 (Chapter 5,
Section 5.3) for salt-amended sulfate concentration. Analysis of variance (ANOVA) showed that
the hypothesis that the average sulfate concentration between soils and extraction methods were
equal cannot be rejected (Table 9-9). The interaction of soil type and extraction method was not
significant (Table 9-9).
120
105
100
95
90
Sulfate
85
80
75
70 ExtractionMethod:
Santa Fe Starvation Hill 1
SoilType 2
Figure 9-4. Variability plot of sulfate concentration (ppm) by soil type and
extraction method.
121
Table 9-8 Comparison of Soil Sulfate Means by Soil Type and Extraction Method
Descriptive Statistics (Statistica Sulfate Spreadsheet)
Effect Level of Number of Sulfate Sulfate
Factor Samples Mean (ppm) Std Dev (ppm)
Total 12 86.8 8.62
Table 9-9 ANOVA for Soil Sulfate by Soil Type and Extraction Method
Fixed Effect Test for Sulfate (Statistica Sulfate Spreadsheet)
Effect decomposition
Num. DF Den. DF F p
Soil Type 1 8 4.52 0.066
Soil Type*Extraction Method 1 8 2.64 0.143
In summary, laboratory rather than field testing for sulfate concentration in soil is recommended
to accurately measure soil mass and to prepare a sample extract that is sufficiently clear of
particles and color that may interfere with the sulfate analysis. Of the two extraction method
tested, both produced clear sample extracts with no detectable difference in sulfate concentration.
Further testing of these two extraction methods on other soils is recommended.
9.2.5 Influence of meter/electrode system on pH measurement
Seen in the ruggedness study results for pH (Chapter 8, Section 8.3.2) that the FM for pH was
sensitive to the electrode condition. This result also apparent from the laboratory visits (Chapter
6, Section 6.4.1). A new bench top pH meter and electrode system (Accumet AB150; 13-620-
631) was purchased to compare with the portable system in current use (Accumet AP85; 13-620-
AP55); both meters were recent mid-grade instruments with automatic temperature
compensation (ATC) and visual display of electrode calibration slope and offset; both electrodes
included indicating and reference electrodes plus a thermocouple (3-in-1 combination), were
mercury-free with a silver/silver chloride reference electrode, had glass bulbs and an epoxy
body. The bench top system had a double-junction electrode and the portable system had a
single-junction electrode. The double-junction electrode was designed to extend the life of an
electrode that measures pH in dirty solutions. The second pH meter/electrode system was
received in time to make comparative measurements for Youngquist and Calhoun sands.
122
Twelve replicates of Calhoun sand were transported cool (~20C), stored in a refrigerator (~4C)
for 5 d, quartered, brought to room temperature, and tested with the “as-received” moisture
content. For the portable AP85/single-junction electrode system, after calibration the slope was
99.8%, offset 6.7 mV, and soil slurry temperature 23.3C; for the bench top AB150, after
calibration the slope was 97.4%, offset 10.1 mV, and soil slurry temperature 23.6C. The
samples were prepared with 100 mL of sand, 100 mL of DI water, stirred briefly at 10-min
intervals over a 30-min period, and re-stirred just prior to measurement, as 12 samples took
about one hr to test. Both electrodes were suspended in the sample and paired readings were
taken simultaneously. Summary statistics are presented in Table 9-10 and data are graphed in
Figure 9-5. The average pH measured by the AB150/double-junction electrode system was
significantly higher than the average pH measured by the AP85/single-junction electrode system,
based on a two-sided, paired t-test.
Table 9-10 Summary Statistics for pH in Calhoun Sand Using Two Different pH
Meter/Electrode Systems
Statistic AP85/13-620-AP55 AB150/13-620-631
Avg 4.64 4.88
St Dev 0.21 0.20
%RSD 4.61 4.01
%RE 18.5 15.2
123
Figure 9-5. Comparison of two different pH meter/electrode systems for pH
measurements in Calhoun sand.
In Figure 9-5, the box plots show moving from the center bar outward the median, 25th and 75th
percentiles, 5th and 95th percentiles, and outliers, and revealed two important features of the data
set: first, that pH measurements with the AP85 meter/electrode system were skewed and second,
that the median values between the two systems were 0.28 pH units apart. In Figure 9-5, a plot
of the 1:1 line shows that although the slopes of the two electrodes were similar, the double-
junction electrode data points were on the average 0.24 pH units apart, while the difference of
3.4 mV in offset voltages predicted only 0.06 pH units (3.4 mV/59 mV/pH unit).
Twelve replicates of Youngquist sand were transported cool (~20C), stored in a refrigerator
(~4C) for 7-9 d, quartered, brought to room temperature, and tested with the “as-received”
moisture content. For the portable AP85/single-junction electrode system, after calibration the
slope was 99.1%, offset 3.5 mV, and soil slurry temperature ~24.0C; for the bench top
AB150/double-junction electrode, after calibration the slope was 98.3%, offset 6.8 mV, and soil
slurry temperature 24.2C. The samples were prepared with 100 mL of sand, 100 mL of DI
water, stirred briefly at 10-min intervals over a 30-min period, and re-stirred just prior to
measurement, as 12 samples took about one hr to test. Electrodes were sequentially suspended
in the sample and paired readings were taken about 5 min apart. Measurements were made on
different days, with pH of six replicates measured on one day and six replicates measured two
days later. Summary statistics were presented in Table 9-11 and data are graphed in Figure 9-6.
The average pH measured by the AB150/double-junction electrode system was significantly
higher than the average pH measured by the AP85/single-junction electrode system, based on a
two-sided, paired t-test.
124
Table 9-11 Summary Statistics for pH in Youngquist Sand Using Two pH
Meter/Electrode Systems
Statistic AP85/13-620-AP55 AB150/13-620-631
Avg 8.15 8.34
St Dev 0.20 0.24
%RSD 2.51 2.89
%RE 7.40 8.60
Figure 9-6. Comparison of two pH meter/electrode systems for pH measurements in

Youngquist sand.
In Figure 9-6, the box plots show moving from the center bar outward the median, 25th and 75th
percentile, 5th and 95th percentiles, and outliers, and revealed two important features of the data
set: first, that pH measurements with both meter and electrode systems were skewed and second,
that the median values between the two systems were 0.21 pH units apart. In Figure 9-6, a plot
of the 1:1 line shows that although the slopes of the two electrodes were similar, the double-
junction electrode data points were on the average 0.18 pH units apart, while the difference of
3.3 mV in offset voltages predicted only 0.06 pH units (3.3 mV/59 mV/pH unit).
The preliminary conclusion drawn from a comparison of pH meter/electrode systems is that even
with pH meters/electrode systems in good working order, variability between systems may make
a large contribution to the method reproducibility.
9.2.6 Effect of ionic strength adjustment on pH measurement
In low ionic strength solutions an increase in the voltage drop across the liquid junction of the
reference electrode tends to bias high the pH measurement. The addition of a salt such as
potassium chloride or calcium chloride can improve the speed and stability of the pH
measurement by improving the transference of ions through the liquid junction to the bulk
solution (Chapter 3, Section 3.2.1). In Figure 9-7 is illustrated the effect of potassium chloride
125
addition on pH measurement for Santa Fe River sand and Jahna sand, two sands with low ionic
strengths but with alkaline and acidic pH, respectively. Potassium chloride was added using a
0.05-mL spatula (and its mass monitored on an analytical balance) to 100 g of air-dried, sieved
soil. To the soil was added 100 mL of deionized water and the soil solution was stirred briefly
every 10 min for a 30-min period, after which time the pH was measured with the
AB150/double-junction electrode. The electrode calibrations were 100% slope and 5.6 mV
offset for pH 4 to pH 7 and 108% slope and 5.6 mV offset for pH 7 to pH 10. Soil solution
temperatures were ~25.0C. The pH dropped by ~0.5 pH units for Santa River sand and ~0.7 pH
units for Jahna sand as potassium chloride was increased from 0.0 g to 0.5 g in 100 g soil; most
of this drop occurred for a potassium chloride addition of 0.10 g to 100 g of soil (Figure 9-7). A
measurement shift of ~0.5 pH units would likely disqualify some of Florida’s sands as
candidates for MSE wall backfill.
Figure 9-7. Effect of potassium chloride (KCl) addition on pH.
9.2.7 Potassium chloride versus calcium chloride to adjust ionic strength
There appears to be no controversy regarding the difficulty of accurate pH measurements in a

water or soil solution with low ionic strength (Bier, 2009; Busenberg and Plummer, 1987; Miller
and Kissel, 2010; Wolt, 1994). Bier (2009) recommended adding 0.1 g of potassium chloride to
100 mL of pure or tap water sample to improve the reading stability and notes that the pH
reading will shift ~0.10 pH units in response to the salt addition. Observed in the USF
126
laboratory was a shift in pH from~ 5.60 to ~5.50 with the addition of 0.1 g of potassium chloride
to a deionized water sample; these pH readings, however, were quite noisy even with 0.10 g
potassium chloride added. Busenberg and Plummer (1987) estimated that in low ionic strength
water the error associated with the liquid junction potential was ~0.06 pH units; they also
reported that adding potassium chloride to low ionic strength water improved the stability of pH
readings. They argued against adding potassium chloride, however, because of their experience-
based concern that potassium chloride salts or solutions were often contaminated and that
contaminants may affect the pH readings.
Miller and Kissel (2010) maintained that 0.01 molar calcium chloride added to a 1:1 soil:water
extract counteracts the measurement error associated with the liquid junction potential in a glass
bulb-reference electrode system. They provided evidence for agricultural soils that a downward
shift of ~1 pH unit occurs when a 1:1 soil:pure water extract was replaced with a 1:1 soil:0.01
molar calcium chloride extract. With an exception for acid soils, they argued against the theory
that calcium chloride displaces protons on the soil surface, and based their argument on the
amount of acid titrated in samples before and after salt addition. From a careful reading of Puri
and Asghar (1938), the addition of potassium chloride to soil samples has the opposite effect as
that of adding water to dilute the soil and its soluble salts; hence, potassium chloride addition
drives the pH reading toward the in-situ pH. Observed in the laboratory was a downward shift of
~0.5 pH units with the addition of 0.1 g of potassium chloride to pH samples. Using calcium
chloride instead of potassium chloride shifted the pH even lower by an average of 0.7 pH units
for Alico Road sand and 0.4 pH units for Clermont sand (Chapter 8, Section 8.3.2). A significant
decrease in soil pH standard deviation was seen in the fixed effect study (Section 9.2.1), but in
ruggedness studies such decreases were not significant (Appendix A, Tables A-8, A-18, A-40,
A-47); neither was a significant decrease seen in the replicate study (Chapter 11, Section 11.3.1).
Wolt (1994) wrote that adjusting the ionic strength of a soil solution can mitigate the pH change
due to differences in the standard pH buffers and low ionic strength soil solutions. A standard
pH buffer solution has a conductivity of ~7,000 µS/cm compared with a Sebring soil
conductivity of ~30 µS/cm (conductivity is proportional to ionic strength). Purchased “low ionic
strength” buffer solutions of pH 4.10 and pH 6.97 had at room temperature conductivities of
1,850 µS/cm and 2,010 µS/cm, respectively. In the USF soils laboratory, biases in pH
measurements were small for low-ionic strength buffers as were differences between normal-
and low-ionic strength buffers (Chapter 5, Section 5.2.2).
In summary, least four explanations were found for a downward shift in pH in soil solutions with
the addition of potassium chloride or calcium chloride:
 The salt cations displaced protons from exchange sites on the soil surface;
 The salt cations displaced aluminum, iron, or manganese from exchange sites on the soil
surface and these metals ions were hydrolyzed and released protons;
 There was a change in the liquid junction potential; and

127
 The typical pH electrode calibration failed in a low ionic strength solution.
The use of potassium chloride or calcium chloride to adjust the ionic strength of soil solutions
during pH measurement was not recommended at this time. The observed downward shift in pH
in sandy soils with neutral salt addition, however, may indicate that these soils have significant
total acidity. Recommended was further investigation of the relationship between metal
corrosion rates in select backfill and soil acidity, where soil acidity is determined as both proton
activity (electrochemical pH measurement) and exchangeable acidity.
9.2.8 Effect of dilution water and slurry water content on minimum resistivity in soils
In this experiment the effect of dilution water resistivity and slurry water content on minimum
resistivity in six sands was investigated. Slurry water content was tested because the FM for
resistivity appeared to be sensitive to the amount of water transferred in slurry from the mixing
bowl to the soil box (Chapter 8, Section 8.3.3). The sands tested were Calhoun (CA), Jahna
(JA), Santa Fe (SF), Starvation Hill (SH), Wimauma (WI), and Youngquist (YO). Dilution
water resistivity was tested at two levels, which were nominally 1,000,000 ohm-cm (no added
salt) and 200,000 ohm-cm (added salt). The upper level represented the high end of the
resistivity meter scale and the lower level represented the minimum resistivity for Type IV
reagent water (ASTM D1193-06). Slurry water content was also tested at two levels, where the
slurry content was predominately soil (SS) or predominately water (WS). For the sands tested,
slurry typically formed when the water content exceeded 20%, but the nature of the slurry was
different for sands with a greater proportion of fine particles such as the Santa Fe River and
Youngquist sands. In these sands the boundaries between sand and water in the slurry were less
distinct than seen for the larger-grained sands. To transfer mostly soil, a spoon was used and to
transfer mostly water, water was decanted directly from the mixing bowl into the soil box and
soil was spooned in as needed to fill the box.
All soils were dried either at room temperature or in an oven at a temperature not exceeding
60C, sieved through a No. 10 sieve, and mechanically split to obtain 1000-g samples.
Deionized (DI) water was obtained from the USF Nanotechnology Research and Education
Center and at ~23oC had a resistivity greater than 1,000,000 ohm-cm. Slightly-salted water with
a measured resistivity of 175,000 ohm-cm was prepared by adding to the DI water 5.0 mg/L
(ppm) of reagent-grade sodium chloride. Testing took place over three days.
Summarized in Table 9-12 are the average, standard deviation, %RSD, and %RE minimum
resistivity across all four combinations of dilution water and slurry water content by sand;
presented in Figure 9-8 are the corresponding 25th, 50th (median), and 75th percentiles. Disparate
minimum resistivity results reported between laboratories in Table 7-2 (Chapter 7, Section 7.4)
are now seen to be within two standard deviations of the average as shown in Table 9-12, or in
other words, the between-laboratory variability can be explained by dilution water resistivity or
slurry water content.
128
Table 9-12 Summary Statistics for Minimum Resistivity by Sand
Sand Avg, ohm-cm St Dev, ohm-cm %RSD %RE
Calhoun 64,900 35,300 54.4 104
Jahna 79,600 34,300 43.0 89.8
Santa Fe 12,800 7,160 56.0 116
Starvation Hill 22,900 12,900 56.3 103
Wimauma 38,900 21,700 55.9 108
Youngquist 7,900 2,020 25.6 55.7
Figure 9-8. Variation in minimum resistivity by sand across two levels of dilution
water resistivity and for slurry water content in the soil box that is either mostly
sand or mostly water.
Tested at the 95% confidence level was the null hypothesis that the minimum resistivity mean
and standard deviation were the same for a dilution water resistivity greater than 1,000,000 ohm-
cm and a dilution water resistivity of 175,000 ohm-cm. Based on a t-test (p = 0.67) and an F-test
(p = 0.64) the null hypotheses was not rejected (Figure 9-9). These results supported a
129
recommendation to use Type IV reagent water (e.g., deionized or distilled water) with a
resistivity of 200,000 ohm-cm or greater as dilution water for the FM for resistivity.
Figure 9-9. Variation in minimum resistivity by dilution water resistivity for six
sands and for slurry water content in the soil box that is either mostly sand or
mostly water.
Tested at the 95% confidence level was the null hypothesis that the minimum resistivity mean
and standard deviation were the same for soil slurry placed in the soil box that is mostly sand or
mostly water. Based on a t-test (p = 0.021) and an F-test (p = 0.0074) the null hypothesis was
rejected (Figure 9-9). These results supported the recommendation to expand the FM for
resistivity to include testing water decanted from soil slurry, where the slurry is formed during
the test procedure. This recommendation is conservative, consistent with AASHTO T288, and
intended to reduce the variability seen between laboratories.
130
Figure 9-10. Variation in minimum resistivity by slurry water content for six sands
and for dilution water resistivity between 200,000 and 1,000,000 ohm -cm.
9.2.9 Choice of range selection on sulfate measurements with a test kit
The current FM for sulfate lists as an apparatus the Hach Sulfate Photometer Test Kit (or
equivalent), which upon purchase included not only an instruction manual and the photometer
but two glass bottles and powder pillows of barium chloride reagent (SulfaVer 4) for 100 tests.
The advantages of this kit are its affordability, small size (portability), rapid response, and ease
of use. The photometer includes built-in calibration curves for sulfate, although a user can
manually input a new curve. Even with the built-in calibration curves, however, a method
operator should check one or more calibration standards to determine which range gives the best
result. Figure 3-5 (Chapter 3, Section 3.5.3) shows that for the instrument the slope of the sulfate
calibration curves were different between Range 1 and Range 2, and Range 2 was higher.
Moreover, Figure 3-5 gives graphical evidence that the photometer response was lower for
reagent lot A3303 than for A4120; reagent lot A3303 had been purchased six months earlier than
lot A4120 and their respective expiration dates were October 2017 and April 2018, respectively.
These results suggested that reagent potency was not consistent between lot numbers and
possibly not consistent over time.
131
9.2.10 Soil conductivity
Field conductivity measurements were substituted for field minimum resistivity measurements.
Conductivity measurements were made using a portable pH/mV/conductivity meter AP85 with
automatic temperature compensation and a compatible temperature-sensing conductivity
electrode. The electrode was calibrated in the field with a 250-µS/cm (4,000 ohm-cm) sodium
chloride conductivity standard, and the minimum reading was checked using deionized water.
Approximately 75 mL of sand was diluted with 150 mL of deionized water in a 250-mL beaker
(Hanlon, 2009). Conductivity expressed as resistivity in ohm-cm was summarized for each soil
in Table 9-13. Given a dilution factor of two for the conductivity (field) measurements, the
results were in good agreement for Alico Road and Angelo’s sand (ratio of 2); the results did not
agree well, however, for Clermont and Sebring soils (ratios of ~8 and ~4, respectively). Note,
however, that for Clermont and Sebring sands, the conductivities approached the detection limit
of the conductivity meter. Further consideration of soil conductivity as a field screening method
for minimum resistivity is recommended.
Table 9-13 Comparison of Conductivity (Field) and Minimum Resistivity (Lab)

Sand Field Resistivity, ohm-cm Lab Min Resistivity, ohm-cm Ratio
Alico Road 23,600 11,200 2.11
Angelo's 38,900 17,800 2.18
Clermont 221,000 28,200 7.85
Sebring 134,000 36,100 3.88
9.2.11 Contribution to pH by minerals retained on a No. 10 sieve
For Alico Road sand, a significant fraction of sand was retained on a No. 10 sieve. For replicate
samples of this sand (Chapter 11), both passing and retained fractions were weighed, and the
relative contribution of the retained fraction was calculated, as shown in Table 9-14. (AASHTO
T288 provides guidance for separating and testing the fraction retained on a No. 10 sieve.) The
pH was also determined in both fractions; for the retained fraction, however, pH measurements
were made with 30 g of soil and 30 g of deionized water, as not all samples had 100 g. The
combined pH was predicted by apportioning the measured pH to passing and retained fractions;
for the two samples that were inadvertently discarded, an average pH of 9.10 was used instead.
Note that none of the Alico Road predicted values were greater than pH 9.00 (Table 9-14), but
treatment study samples transported under ambient conditions, stored at room temperature, and
tested “as received” were greater than pH 9.0, which suggested that some alkalinity was lost in
the drying process. Based on these results, testing “as received” soil for pH was recommended.
132
Table 9-14 Contribution of Mineral Fraction Retained on a No. 10 Sieve to Total Mass
and pH
Passing, Retained, Total, Retained, pH pH pH
Replicate
g g g % Passing Retained Predicted
1 2,962 134 3,095 4.3 8.86 9.16 8.87
2 2,854 178 3,032 5.9 8.87 9.17 8.89
3 2,904 141 3,045 4.6 8.88 9.14 8.89
4 3,103 296 3,399 8.7 8.86 9.16 8.89
5 2,775 223 2,998 7.4 8.89 9.01 8.90
6 2,930 93 3,024 3.1 8.88 9.03 8.88
7 3,183 361 3,545 10 8.88 9.11 8.90
8 3,067 145 3,212 4.5 8.81 9.08 8.82
9 2,796 185 2,981 6.2 8.84 - 8.86
10 3,009 149 3,159 4.7 8.89 9.08 8.90
11 2,366 676 3,042 22 8.85 - 8.91
12 2,836 162 2,997 5.4 8.90 9.02 8.91
Averages 2,899 229 3,127 7.3 8.87 9.10 8.88
9.3 Summary
For pH measurements in sieved soils, the pH changed for soil with a significant mass retained on
the sieve; by inference in some soils under similar circumstances the minimum resistivity could
also change. AASHTO T288 requires the soil be sieved for a minimum resistivity measurement
but provides guidance for separating and testing the fraction retained on a No. 10 sieve. For both
pH and minimum resistivity, the current practice of testing soils “as received” was
recommended. For one soil, the average pH measurement was significantly lower for warm
transport and storage than for cool transport and cold storage; across all four soils the average pH
observations were in reasonable agreement between measurements made on-site and those made
in the laboratory for soils transported cool and stored cold. Treatment study results also
suggested that variability in sample minimum resistivity between laboratories could be reduced if
laboratories processed samples in a similar manner, for example, with transport under ambient
conditions and samples stored at room temperature. To reduce variability in sulfate
concentrations measured between laboratories, amending the FM to allow drying the soil from
room temperature up to 60C, consistent with the AASHTO T290 drying temperature, was
recommended.
Some sands that met the specifications for MSE wall backfill had high minimum resistivity (low
conductivity) and thus low ionic strength, and were a challenge for pH measurements.
Potassium chloride, a neutral salt, is a possible amendment to soil samples for pH testing. Three
advantages to the addition of 0.1 g of potassium chloride to 100 g of soil sample were: the pH
133
reading was faster, more stable, and was more likely to better represent the in-situ soil pH; the
major disadvantage, however, was a drop of ~0.5 pH units in 1:1 soil:water solutions. Such a
shift would likely disqualify some of Florida’s sands as candidates for MSE wall backfill. The
use of potassium chloride or calcium chloride to adjust the ionic strength of soil solutions during
pH measurement was not recommended at this time. The observed downward shift in pH in
sandy soils with neutral salt addition, however, may indicate that these soils have significant total
acidity. Recommended was further investigation of the relationship between metal corrosion
rates in select backfill and soil acidity, where soil acidity is determined as both proton activity
(electrochemical pH measurement) and exchangeable acidity.
A comparison of pH meter/electrode systems revealed that even with pH meters/electrode

systems in good working order, variability between electrodes may make a large contribution to
the method (im)precision.
Research results supported the recommendation to use Type IV reagent water (e.g., deionized or
distilled water) with a resistivity of 200,000 ohm-cm or greater as dilution water for the FM for
resistivity and to expand the FM for resistivity to include testing water decanted from soil slurry,
where the slurry is formed during the test procedure. This recommendation was conservative,
consistent with AASHTO T288, and intended to reduce the variability seen between laboratories.
Results of soil conductivity as a field screening method were encouraging but were based on low
and very low conductivity soils. Further consideration of conductivity as a screening tool was
recommended. For both the FMs for chloride and sulfate, laboratory rather than field testing was
recommended to accurately measure soil mass; moreover, soil extracts must be clear of
suspended particles and color that may interfere with analyses. After gravity filtration through a
coarse fast filter, vacuum filtration of soil extract through a 0.45-micron membrane filter was
recommended. Finally, for the FM for sulfate (Hach test kit method), research results suggested
that reagent potency was not consistent between lot numbers and possibly not consistent over
time. The relationship between lot number and sulfate calibration curve should be checked prior
to a sample measurement.
134
10 Inter-Laboratory Study
10.1 Scope and Purpose of Study
The purpose of this study was to estimate a multi-laboratory precision (reproducibility) for the
revised methods, and to the extent practicable the study was designed in accordance with ASTM
C802. Over a three-day period, each participating laboratory analyzed two select backfill (FDOT
092L) materials for pH, minimum resistivity, chloride, and sulfate per proposed revised FMs.
Results for each material were aggregated to produce an estimate of between-laboratory
variances and, consistent with ASTM C670, a precision statement was prepared to accompany
each of the revised methods. ASTM C802 recommended that each material be tested in duplicate
for 10 or more laboratories and with higher replication for fewer laboratories to obtain an
estimate of within-laboratory variability. FDOT’s corrosion series methods require a single test
rather than replicate tests for the reported test result and routine comparisons of within-
laboratory replicate results were not made (in practice, a laboratory re-ran a sample if results
were unusual). Moreover, material properties affected test results for one or more methods; for
example, the presence of colloidal particles interferes with both pH and sulfate determinations.
For this reason, two different materials rather than replicates of one material were tested to
obtain two estimates of the between-laboratory precision at the expense of the within-laboratory
precision.
Participating laboratories included five FDOT district laboratories in Bartow, Chipley, Davie,
Deland, Lake City; the FDOT State Material Office laboratories in Gainesville, and the USF
soils laboratory in Tampa. Each laboratory was assigned at random a number between 1 and 7,
which was not shared, and results were reported by laboratory number. Laboratory personnel
were given training on the proposed revised methods ahead of the study date. Each laboratory
received in advance of the study a study plan (Appendix C), with a data sheet and the proposed
revised FMs attached. Both paper and electronic documents were provided. The proposed
revisions updated and expanded the methods’ procedural details and included changes to method
equipment, sample processing, and data management.
The results of the first inter-laboratory study indicated that further work on the FM for pH was
needed. At the USF soils laboratory, study conditions were mimicked and revealed that a
memory effect likely contributed to dispersion in the pH results; this effect was pronounced
between samples of disparate pH values, for example, pH 7 and pH 9. The method was revised
and a second inter-laboratory study was conducted for pH for two samples, each in duplicate.
Refer to Appendix C for the study plan and data sheet for the second study.
10.2 Sample Preparation
Material A was characterized by the mining operator as mine tailings. The sand fizzed when a
few drops of 10% hydrochloric acid were added to a ~1-g subsample, which indicated the
presence of carbonates. Material B was characterized by the mine operator as mason’s sand; the
presence of carbonate was not detected in this sand. Both sands had ~30% of its mass passing a
135
No. 100 (0.15 mm) sieve; with 3.4 % and 0.25% passing a No. 200 (0.075 mm) sieve for
Material A and B, respectively. For Material B, however, 3:1 water-to-soil extracts were poorly
settled even after 24 hr, which suggested the presence of colloidal particles (particles with a
diameter in the range of 0.001 to 0.10 µm). Both Material A and B as sampled had chloride and
sulfate concentrations below their respective method detection limits. Material A was not
amended with salts but to Material B was added 0.01% by weight sodium chloride for a target
chloride concentration of 61 ppm and 0.01% by weight sodium sulfate for a target sulfate
concentration of 81 ppm. The anticipated minimum resistivity of Material B was ~3000 ohm-cm
based on the added salts alone.
Soils were dried in an oven at 60C and sieved through a No. 10 (2 mm) mesh upon receipt at
USF from the mines (see Table 1-1 for sampling dates) and stored at room temperature in a
covered plastic bucket. Material A was from Mine 8 (Alico Road sand) and Material B was from
Mine 2 (Jahna sand). In mid-December 2014, Material A was mechanically split into ~1,500-g
samples; salts were added to Material B in 3-kg batches, and these batches were placed under a
mixer on low speed for 10 minutes, then combined and mechanically split to ~1,500-g samples.
For both materials, samples were placed in numbered and sealed plastic bags, stored in covered
plastic buckets, and delivered to participating laboratories by automobile. Samples likely
experienced modest temperature swings during the course of their transit.
For the second pH inter-laboratory study Material C was an acid sand from Mine 6 (Sebring
sand) and Material D was a mason’s sand from Mine 7 (Clermont sand). Soils were dried in an
oven at 60C and sieved through a No. 10 (2 mm) mesh upon receipt at USF from the mines (see
Table 1-1 for sampling dates) and stored at room temperature in a covered plastic bucket. For
both materials, samples were mechanically split to ~300 g, placed in numbered and double-
wrapped in sealed plastic bags, and were shipped by US Postal Service priority mail to arrive at
participating laboratories ~1 week before the March 2015 start date.
10.3 Data Management
Data was handled and reported as required by the revised FMs, with exceptions as noted on the
data sheet (Appendix C). Measurements that were below detection were to be flagged as such.
Upon completion of testing, the data sheets were electronically submitted to the study contact.
10.4 Results and Discussion
Study results were presented in Table 10-1. For Material A, the average ( standard deviation)
values for each method were pH, 9.19  0.07; minimum resistivity, 12,200  1,090 ohm-cm;
chloride, 7  8 ppm, and sulfate: 4  4 ppm. Consensus values (average and median) for
chloride and sulfate were below the stated method detection levels of 15 ppm and 6 ppm for
Hach-kit chloride and sulfate, respectively, when expressed as a soil concentration. For Material
B, the average ( standard deviation) values for each method were pH: 7.02  0.72; minimum
resistivity, 2,310  430 ohm-cm; chloride, 66  12 ppm; and sulfate, 67  16 ppm. For Material
136
B, the relative standard deviations (%RSD) and relative error (%RE) were much higher for pH
and resistivity and much lower for chloride and sulfate than for Material A. For chloride and
sulfate the improvement in %RSD and %RE can be explained by concentrations above their
corresponding detection levels. Recoveries for added chloride and sulfate were 99% and 83%,
respectively.
Table 10-1 Inter-Laboratory Study Results for Proposed Revised FMs for pH,
Minimum Resistivity, Chloride and Sulfate
Material A Material B
Lab Resistivity, Chloride, Sulfate, Resistivity, Chloride, Sulfate,
pH pH
ohm-cm ppm ppm ohm-cm ppm ppm
1 9.15 12,500 0 0 6.88 2,200 60 87
2 9.11 13,000 15 3 6.33 2,700 60 39
3 9.31 11,800 15 9 6.46 2,360 90 72
4 9.24 11,000 0 3 6.85 2,300 60 54
5 9.13 12,000 0 3 6.97 2,600 60 69
6 9.18 14,000 15 1 7.63 2,600 75 75
7 9.18 10,950 3 10 8.39 1,420 57 70
Avg 9.19 12,200 7 4 7.07 2,310 66 67
St Dev 0.07 1,090 8 4 0.72 433 12 16
Median 9.18 12,000 3 3 6.88 2,360 60 70
%RSD 0.8 9.0 110 93 10 19 18 23
%RE 2.2 25 220 240 29 55 50 72
For pH, if we assume that the method instrumentation, operator, and measurement temperatures
were the same for both materials, then the difference seen in reproducibility between the two
materials were likely related to material properties. Since salt was added to Material B, effects
due to low ionic strength were ruled out. Four possible explanations were:
 The liquid junction of the pH electrode trapped a portion of the previous sample or
buffer. If this occurred, the current pH reading would be distorted until the liquid junction
re-equilibrated with the current sample;
 Colloidal particles were present in the sample and interfered with the voltage drops on
both the glass bulb and in the liquid junction; and
 For an un-buffered material at pH 7.0, if the proton (H+) concentration in solution was
close to that in gel layers on the inside and outside of glass bulb surfaces, the electrode
response was unstable.
 The space between the glass bulb and thermocouple trapped a small amount of the
previous sample, buffer, or rinse water and thus affected the pH reading.
137
The FM for pH was revised to include steps to mitigate the memory effect, and included (1)
directions to disable the hold feature of the pH electrode, (2) steps to re-check electrode
calibration in a pH 7 buffer between samples, and (3) a second dilution water rinse and re-test of
pH within each sample. Results of the second pH inter-laboratory study were reported in Tables
10-2 and 10-3. The pH averages, standard deviations, and %RSDs were given in Table 10-4;
estimates of within- and between-laboratory standard deviations were calculated according to
ASTM C802.
Table 10-2. Material C Results Revised FM for pH

Material C
Laboratory
Replicate 1 Replicate 2 Avg St Dev
1 8.23 8.29 8.26 0.04
2 7.62 7.69 7.66 0.05
3 7.64 7.67 7.66 0.02
4 7.44 7.56 7.50 0.08
5 7.67 7.68 7.68 0.01
6 7.37 7.48 7.43 0.08
7 8.14 8.12 8.13 0.01
Avg 7.73 7.78 7.76 0.04
St Dev 0.33 0.30 0.32 -
Table 10-3. Material D Results Revised FM for pH

Material D
Laboratory
Replicate 1 Replicate 2 Avg St Dev
1 5.20 5.06 5.13 0.10
2 5.06 5.12 5.09 0.04
3 4.81 4.86 4.84 0.04
4 5.11 5.12 5.12 0.01
5 5.29 5.22 5.26 0.05
6 5.04 5.09 5.07 0.04
7 5.27 5.24 5.26 0.02
Avg 5.11 5.10 5.11 0.04
St Dev 0.17 0.13 0.14 -
138
Table 10-4 pH Averages, Standard Deviations, and %RSD
Standard Deviations %RSD
Material Average
Within Lab Between Labs Within Lab Between Labs
C 7.76 0.051 0.32 0.66 4.09
D 5.11 0.049 0.15 0.97 2.86
An improvement in method precision for pH was seen from the first to the second inter-
laboratory study, but the between-laboratory precision for Material C was double that of Material
D. Within-laboratory precision for both sands were ~0.05 pH units. Both soils had relatively high
minimum resistivity (Chapter 9, Section 9.1.1, Table 9-1) or low conductivity and thus low ionic
strength, and low ionic strength coupled with a lower proton concentration likely contributed to
the greater dispersion in pH values seen for Material C. Recall from Chapter 5 (Section 5.1.1)
that within-laboratory precision typically represents the random component of analytical testing
error and between-laboratory precision the systematic component of error. Overall the inter-
laboratory results suggested that progress was made in reducing the systematic component of
error.
10.5 Method Precision Statements
10.5.1 Precision statement FM for pH
For two test materials, multi-laboratory standard deviations of a single test result were dependent
on the material (Table 10-5). For materials of similar pH, the results of two properly-conducted
tests in different laboratories on the same material are not expected to differ by more than the pH
units shown in column labeled “Acceptable Range of Two Results.”
Table 10-5 Multi-Laboratory Precision for FM for pH

Material Avg St Dev Acceptable Range of Two Results
Sand (A-3) 7.76 0.32 0.91
Sand (A-3) 5.11 0.15 0.42
10.5.2 Precision statement FM for minimum resistivity
For two test materials, multi-laboratory standard deviations of a single test result were dependent
on material (Table 10-6). For materials of similar minimum resistivity, the results of two
properly-conducted tests in different laboratories on the same material are not expected to differ
by more than the ohm-cm shown in column labeled “Acceptable Range of Two Results.”
139
Table 10-6 Multi-Laboratory Precision for FM for Minimum Resistivity
Material Avg, ohm-cm St Dev, ohm-cm Acceptable Range of Two
Results
Sand (A-3) 12,200 1,090 3,090
Sand (A-3) 2,310 430 1,230
10.5.3 Precision statement FM for chloride
For a test material (sand, A-3) with an average chloride concentration of 66 ppm, the multi-
laboratory standard deviation of a single test result has been found to be 12 ppm. Therefore,
results of two properly-conducted tests in different laboratories on the same material are not
expected to differ by more than 34 ppm.
10.5.4 Precision statement FM for sulfate
For at test material (sand, A-3) with an average sulfate concentration of 67 ppm, the multi-
laboratory standard deviation of a single test result has been found to be 16 ppm. Therefore,
results of two properly-conducted tests in different laboratories on the same material are not
10.6 Summary
Precision of pH, minimum resistivity, chloride, and sulfate measurements in candidate materials
for select backfill were dependent upon the material. Relative standard deviations (%RSDs) for
pH measurements were higher for a near-neutral mason’s sand than for a pH 9 carbonate-bearing
sand or a pH 5 acid sand. Improvement in method precision for pH was seen between the first
and second inter-laboratory study, and the latter featured pH method revisions that were intended
to reduce electrode memory effects. Relative standard deviations (%RSDs) of chloride and
sulfate measurements were much higher in a material with concentrations near the detection
level than for concentrations of ~70 ppm, which was expected.
140
11 Replicate Study and Implications for Quality Assurance
11.1 Introduction
Distributions of pH, minimum resistivity, chloride, and sulfate were determined within and
between soil mines for soils that qualify as select backfill and from these statistics were derived
method operating characteristic (OC) curves. Sampling of select backfill takes place when a
material is certified for this use at its source and again upon its emplacement behind an MSE
wall. At the source, backfill from different locations in the strata or stockpile may be placed in a
smaller stockpile from which samples are collected and composited for testing. At the
construction site, independent tests of backfill are made after emplacement at least twice for each
soil type: once for quality control by the project contractor and once for verification by the
project engineer. Soil type is determined not only by visual inspection, but by other physical test
results such as optimum moisture content and particle size distribution. If a backfill fails any one
of the four corrosion tests, the backfill is tested a third time for resolution by an FDOT District
Materials Office, State Materials Office, or their representative. Refer to Chapter 2, Section
2.2.3, for further explanation.
11.2 Methods
11.2.1 Field sampling
Soil samples were collected from eight different mines as listed in Table 1-1 (Chapter 1, Section
1.2.7); all samples were classified as sand (AASHTO A-3) and all samples met the geotechnical
requirements for select backfill (FDOT material 092L) (Chapter 7, Section 7.3, Table 7-1). Of
the eight mines, five mines were located in FDOT’s District 1, two in District 7, and one in
District 5. Mines included both dry pit and dredged strata, and samples were taken from a mine
floor (one mine), mine walls (two mines), and stockpiles (five mines), two of which were mine
tailings and two of which were size-segregated as mason’s sand. Of the eight soil mines one was
in Clermont, one in Dade City, two in Ft. Myers, two in Haines City, one in Sebring, and one in
Wimauma. Each mine was identified by numerical sequence (1 to 8) in order of sampling date,
because sample processing and analysis changed between samples collected from the first four
mines and last four mines.
Mines were chosen by selecting names at random from a list of 30 sources that had one or more
samples previously accepted for select backfill according to a FDOT Districts 1 and 7 Materials
Office database. An attempt was made to contact each of these sources. Many of these mines
were defunct, had changed ownership and no longer supplied select backfill, or did not respond
to a telephone and/or e-mail inquiry. Seven of the eight mines sampled, however, were a subset
of initial list.
Sampling was subject to selection bias because of the practical limits on geographical coverage,
pre-selected characteristics of the soil, and accessibility of soil in stockpiles or soil strata in pits,
141
and the results reported herein were not fully representative of Florida’s resources. Soil sampling
was more systematic than random or stratified random. The reported soil sample standard
deviations include contributions from material, sampling, processing, environment, and analysis
(Chapter 2, Section 2.2.1, Equation 2-4), where analyses were accomplished by USF soils
laboratory.
At the Wimauma mine a backhoe operator dug 12 holes about 3-m deep in a grid pattern with
holes spaced roughly 50 m apart. From each excavated soil pile, a ~3-kg soil sample was
collected with a shovel and placed in a sealable plastic bag. The 12 samples were transported in
a cooler on ice back to the USF soils laboratory. Soil from the remaining mines came either from
a subsample of a stockpile or a pit wall and the sampling protocol was the same. At each of
these mines an equipment operator removed one front loader bucket of soil from three separate
locations near the base of the stockpile or pit. The three loads were kept separate but placed one
next to the other; the tops of the soil piles were flattened with the back of the front loader bucker.
From each of the three small soil piles were collected with a shovel four ~3-kg samples spaced
roughly equidistant from each other for a total of 12 samples; each sample was placed in a
separate sealable plastic bag. These 12 samples were transported in a cooler on ice back to the
USF soils laboratory.
11.2.2 Sample processing
Samples were processed in accordance with FMs for pH, minimum resistivity, chloride, and
sulfate. Samples were placed in new plastic bags and the bag was rolled down to reduce the air
pocket between the sample and the bag seal. Sealed bags were arranged in a cooler and kept
cool by frozen bottled water or ice packs, such that the transport temperature was typically 10 to
15C cooler than ambient midday temperatures of 26 to 33C. No samples were saturated with
water, so removal of excess water was not an issue.
Upon receipt at the USF soils laboratory, samples collected from mines 1 through 4 were stored
in a refrigerator at 4C, quartered, with a portion brought to room temperature for pH and
resistivity analysis and a portion dried for ~4 hr at 110C for chloride and sulfate analysis. Once
dried, the latter portion was stored at room temperature (~23C). For samples collected from
mines 5 through 8, samples were dried at 60C for ~ 4 hr, sieved through a No. 10 (2 mm) mesh,
and stored at room temperature prior to tests for pH, minimum resistivity, chloride, and sulfate.
11.2.3 Sample analysis
In addition to soil processing, analytical procedures also changed for pH and resistivity between
samples collected from mines 1 through 4 and mines 5 through 8. For pH, a portable pH meter
with a single-junction electrode was switched to a bench-top pH meter and a double-junction
electrode, 100 g of dried and sieved soil was tested instead of 100 mL of soil at the “as received”
moisture content, and 0.1 g of potassium chloride was added to each beaker that contained 100 g
of soil and 100 mL of water. For minimum resistivity, 1,000 g of dried and sieved soil was
142
tested instead of 1,000 g of soil at the “as received” moisture content, and the water portion
rather than the soil portion of the soil/water slurry was preferentially tested. For chloride and
sulfate, samples from mines 1 through 8 underwent an additional filtration step to clarify the
sample extracts as necessary prior to analysis by chemical methods, as otherwise the interference
from suspended particles and color was significant.
11.2.4 Data analysis
For pH, minimum resistivity, chloride, and sulfate, the central tendencies and data dispersion
were presented with descriptive statistics such as the mean, median, and mode, standard error
and standard deviation, as well as with box plots. In each box plot, the center line of the box
represented the data’s median, the box lower and upper boundaries represented the 25th and 75th
percentiles, respectively, the box lower and upper whiskers represented the 5th and 95th
percentiles, respectively, and the data points represented any outliers. The hypothesis was tested
based on a one-sample Lilliefors-corrected Kolmogorov-Smirnov (K-S) test at the 95%
confidence level (IBM SPSS Statistics 22) that the data distributions were not different from
either a normal or a uniform distribution. The latter distribution was considered because (1)
more than one mine separated material stockpiles by size, and (2) the potential for an unbroken
geological stratum to have consistent geological and chemical properties.
A single-factor analysis of variance (ANOVA) was performed on each of the four datasets to
assess the between- and within-sample variance (Appendix D). The within-sample variance was
2
used as an estimate of the total variance 𝜎𝑡𝑜𝑡𝑎𝑙 (Equation 2-4). This estimate, however, was for
2
analysis by a single laboratory; to expand 𝜎𝑡𝑜𝑡𝑎𝑙 for analysis by multiple laboratories, the single-
laboratory variance was subtracted from and the multiple-laboratory variance added to the total
variance. The single-laboratory variance was obtained from triplicate treatment study
measurements for a treatment where sample processing and analyses were similar to that for the
replicate study (Chapter 9, Section 9.3). The multi-laboratory variance was obtained from the
results of an inter-laboratory study (Chapter 10, Section 10.4).
For the operating characteristic (OC) curves, the Type II () error was calculated according to
Equations 11-1 (minimum resistivity, chloride, and sulfate) and 11-2 (pH). The inverse of the
standard normal cumulative distribution function (NORM.S.INV, Microsoft Excel 2010) was
used to determine zor z/2, where the Type I () was assigned as p = 0.01 (Table 2-7), and 
was obtained from the standard normal distribution functions at z (NORM.S.DIST, Microsoft
Excel 2010). In Equations 11-1 and 11-2, µ0 is the rejection limit, µa the acceptance limit,  the
2
method test error estimated by√𝜎𝑡𝑜𝑡𝑎𝑙 , and N the sample size (Burati et al., 1995; Ott, 1993).
|𝜇0 −𝜇𝑎 |
𝛽 = 𝑃 [𝑧 < 𝑧𝛼 − 𝜎 ] (11-1)
⁄ 𝑁
√
143
|𝜇0 −𝜇𝑎 |
𝛽 ≈ 𝑃 [𝑧 < 𝑧𝛼⁄2 − 𝜎 ] (11-2)
⁄ 𝑁
√
Between the source and the construction site, the process of mining, loading, transporting,
stockpiling, and emplacing soil likely enhances soil mixing and improves the uniformity of
emplaced material. For non-uniform soils, however, such mixing may yield a different average
property than an average of individual samples collected from a relatively undisturbed source.
To assess the influence of soil mixing on its measured properties, the mean and variance of
individual soils from replicate studies were compared with the mean and variance of the same
soils from treatment studies. For the treatment and other studies, ~100 kg of source material was
composited on site and divided into smaller subsamples for sample handling (Chapter 7, Section
7.2). The sample handling and analysis protocols for replicate studies were matched with those
of the treatment studies. To compare means, a two-tailed t-test with unequal variances was
applied at the 95% confidence level (p < 0.05), where 𝑥̅1 and 𝑥̅2 were sample means for replicate
and treatment studies, respectively. The null and alternative hypotheses were
H0: 𝑥̅1 − 𝑥̅2 = 0 (11-3)
H1: 𝑥̅1 − 𝑥̅2  0 (11-4)
To compare variances, an F-test was applied at the 95% confidence level (p<0.05), where 𝑠12
and 𝑠22 were sample variances for replicate and treatment studies, respectively. The null and
alternative hypotheses were
H0: 𝑠12 = 𝑠22 (11-5)
H1: 𝑠12 > 𝑠22 (11-6)
11.3 Results
11.3.1 FM for pH
11.3.1.1 Distribution of pH in mined soil
In Table 11-1 were presented the descriptive statistics and in Figure 11-1 the box plots for soil
pH within and between samples collected from eight mines. For MSE wall backfill reinforced
with metal, backfill must have a pH between 5.0 and 9.0; for backfill reinforced with synthetic
material, backfill must have a pH between 4.5 and 9.0 (Chapter 2, Section 2.1.2, Table 2-2). A
single-factor analysis of variance (ANOVA) revealed that the variability of pH was greater
between than within soils (p = 0.00) and the hypothesis that pH averages were equal across the
eight soils was rejected (Appendix D, D.1, Table D-1). There was no clear evidence—either for
central tendencies (t-test, p = 0.06) or dispersion (F-test, p = 0.07)—that differences in sample
144
processing and analyses between the first four and the last four mines affected pH measurements
within each soil (Figure 11-1). The pH in replicates from across all mines ranged from 4.00 to
8.90, with an average (± standard deviation) of 6.08 (± 1.68) and at the 95% confidence level, the
average pH was between 5.74 and 6.42. The hypothesis that the distribution of replicate pH
measurements was normal was not rejected for soil from all mines except mine 2; the hypothesis
that the distribution of replicate pH measurements was uniform was not rejected for soil from all
mines except mines 2 and 5. The mean (average), median, and mode (if present) pH from
samples of soils from mines 1, 4, 5 and 6 were below pH 5.0 but above pH 4.5, and no soils had
an average pH above pH 9.0. Standard deviations ranged from 0.025 to 0.63 pH units; all but the
soil from mine 6 had standard deviations at or below 0.22 pH units. Kurtosis (peakedness) and
skewness of the distributions is significant when these values exceed ± 2.0, for example, as seen
for mine 2, and may reflect an impact of geological or industrial processes such as size
segregation on the uniformity of pH in source material.
Table 11-1 Descriptive Statistics for pH

Mine Identifier
Statistic
1 2 3 4 5 6 7 8 All
Mean 4.66 5.37 8.15 4.88 4.65 4.70 7.40 8.87 6.08
Standard Error 0.025 0.064 0.059 0.057 0.059 0.183 0.022 0.007 0.170
Median 4.66 5.30 8.21 4.86 4.59 4.56 7.42 8.88 5.26
Mode 4.70 5.30 7.77 none 4.46 none 7.47 8.88 8.88
Std Deviation 0.087 0.220 0.204 0.196 0.204 0.634 0.077 0.025 1.68
Sample Variance 0.008 0.049 0.042 0.038 0.042 0.401 0.006 0.001 2.81
Kurtosis -0.64 6.79 0.32 1.70 2.02 1.76 -1.08 1.11 -1.38
Skewness 0.40 2.47 -1.16 0.93 1.38 1.26 -0.09 -1.04 0.56
Range 0.28 0.83 0.60 0.74 0.70 2.21 0.24 0.09 4.90
Minimum 4.54 5.17 7.77 4.60 4.44 4.00 7.29 8.81 4.00
Maximum 4.82 6.00 8.37 5.34 5.14 6.21 7.53 8.90 8.90
Count 12 12 12 12 12 12 12 12 96
Confidence Level (95.0%) 0.055 0.140 0.130 0.124 0.130 0.403 0.049 0.016 0.340
%RSD 1.87 4.11 2.51 4.01 4.39 13 1.05 0.28 27
145
Figure 11-1. Box plots of soil pH by soil mine.
11.3.1.2 Operating characteristic (OC) curve for pH

2
For pH, the variance 𝜎𝑡𝑜𝑡𝑎𝑙 pooled across eight soils was 0.073 (Appendix D, Section D.1, Table
D-1). The portion of this variance that was related to single-laboratory analytical testing was
0.0054 (Appendix B, Section B.9, Table B-34). The estimate of the variance component for
multi-laboratory analytical testing was 0.0225 (Chapter 10, Section 10.5, Table 10-5, Material
2
D). Thus, 𝜎𝑡𝑜𝑡𝑎𝑙 including single- and multi-laboratory analytical testing variances were 0.073
and 0.090, respectively, and the corresponding standard deviations or test errors were 0.27 and
0.30 pH units, respectively. In Figure 11-2 were plotted for pH and sample size N = 3 a series of
OC curves at test errors  of 0.20, 0.30, 0.40, and 0.50 pH units. Acceptance limits were pH 5
and pH 9 and critical (rejection) limits of pH 4 and pH 10 were assumed. Figure 11-2 suggested
that for triplicate independent, randomly-selected samples with a normal pH distribution, the
probability of accepting a good backfill (or power, 1-β) was 95% or better if test error was at or
below 0.40 pH units; for a test error of 0.50 pH units, the power dropped to 81%. For N = 2,
power was greater than 95% for a test error of 0.30 pH units but for a test error of 0.40 pH units
146
dropped to 83%. As previously noted in Chapter 2 (Section 2.2.1), several options were available
to improve the statistical power: (1) change the acceptance limits, (2) increase the number of
samples, (3) reduce the test error, (4) shift more risk to the seller, (5) or modify the reinforcement
design or material. Each and every one of these options has cost implications, but results from
the inter-laboratory study suggested there was room to reduce the test error.
Figure 11-2. Operating characteristic (OC) curve for pH for sample size N = 3.
The difference in the means and variances between replicate and treatment studies suggested the
relative influence of material mixing and subsampling at the source on the observed sample pH
(Table 11-2). Note, for example, that for mine 6 the pH decreased from pH 4.70 to pH 4.39, and
the latter pH failed to meet the lower acceptance limit for either metal and geosynthetic
reinforcement. From Table 11-2 sample means were significantly different between replicate and
treatment studies for soils from mines 3 and 4 and variances were significantly higher between
replicate and treatment studies for soils from mines 5 and 6. In other words, for soil from mines
3 and 4, material handling between the mine and the laboratory affected the observed average
147
pH; for mines 5 and 6, material handling between the mine and the laboratory diminished the
observed variability in pH.
Table 11-2 Comparison of Replicate and Treatment Study Statistics for pH

Mine Identifier
Statistic
1 2 3 4 5 6 7 8
Replicate Average 4.66 5.37 8.15 4.88 4.65 4.70 7.40 8.87
Replicate Standard Deviation 0.09 0.22 0.20 0.20 0.20 0.63 0.08 0.03
Treatment Average 4.65 5.16 8.41 4.52 4.56 4.39 7.44 8.92
Treatment Standard Deviation 0.08 0.15 0.04 0.06 0.02 0.04 0.09 0.05
t-test (p) 0.81 0.11 0.00 0.00 0.18 0.12 0.52 0.19
F-test (p) 0.99 0.69 0.08 0.19 0.02 0.01 0.65 0.09
Elias et al. (2009) recommended that for an MSE wall six pH samples of sand be tested pre-
construction and one pH sample be tested every 1,529 m3 (2,000 yd3, which is 100 to 200
truckloads, or ~70 lifts of 28 yd3 per lift) during construction. Results for pH suggested that
emphasis on pre-construction sampling was well-placed due to the potential lack of uniformity of
source material with respect to pH. Recommendations to improve the buyer’s statistical power
were:
 Revise the FM for pH;
 Test at least three independent samples for pH per soil type, both pre-construction and
post-emplacement;
 Provide operator training;
 Conduct routine on-site independent audits of the FM for pH (see Appendix G for an
audit checklist); and
 Complete a Florida-wide inter-laboratory study of the revised FM within a year of its

adoption.
11.3.2.1 Distribution of minimum resistivity in mined soil
In Table 11-3 were presented the descriptive statistics and in Figure 11-3 the box plots for
minimum resistivity of soil samples within and between samples collected from eight mines. For
MSE wall backfill, soil minimum resistivity must be greater than 3,000 ohm-cm (Chapter 2,
Section 2.1.2, Table 2-2). There was evidence—for both central tendencies (t-test, p = 0.00) and
148
dispersion (F-test, p = 0.00)—that differences in sample processing and analysis between the
first four and the last four mines affected minimum resistivity measurements within each soil
(Figure 11-3). These differences were likely due to a change in the analytical procedure that
preferentially measures the resistivity of the water rather than soil content in the soil slurry
produced as water is added to the sample during testing. The minimum resistivity of replicates
across mines 5 through 8 ranged from 9,500 to 64,000, with an average (± standard deviation) of
25,300 (± 11,700), and at the 95% confidence level, the average minimum resistivity was
between 22,000 and 28,600. A single-factor ANOVA (Appendix D, Section D.2, Table D-2)
revealed that the variability of minimum resistivity was greater between than within soils (p =
0.00); the hypothesis that the minimum resistivity averages were equal across soils from mines 5
through 8 was rejected. The hypothesis that the distribution of replicate minimum resistivity
measurements was normal was not rejected for soil from all mines except mines 3 and 6; the
hypothesis that the distribution of replicate minimum resistivity measurements was uniform was
not rejected for soil from all mines except mines 3, 6, and 8.
Table 11-3 Descriptive Statistics for Minimum Resistivity

Mines
Statistic Mines
1 2 3 4 5 6 7 8
5-8
Mean 57,600 135,000 8,780 152,000 19,100 36,400 33,000 12,500 25,300
Standard Error 7,470 5,380 1,110 22,300 1,820 2,950 1,040 441 1,690
Median 55,500 143,000 8,250 110,000 18,500 33,300 32,800 12,500 27,300
Mode none 150,000 4,100 110,000 20,000 none 30,000 13,000 13,000
Std Dev 25,900 18,600 3,840 77,400 6,300 10,200 3600 1530 11,700
Sample 6.69 3.48 1.48 5.98 3.96 1.05 1.30 2.34 1.37
Variance E+08 E+08 E+07 E+09 E+07 E+08 E+07 E+06 E+08
Kurtosis 0.26 0.67 -0.85 0.99 7.74 4.72 -0.22 0.15 0.74
Skewness 1.01 -1.13 0.45 1.41 2.54 1.98 0.88 -0.19 0.73
Range 80,500 61,000 11,900 248,000 24,500 38,000 10,500 5,500 54,500
Minimum 29,500 94,000 4,100 77,000 13,000 26,000 29,500 9,500 9,500
Maximum 110,000 155,000 16,000 325,000 37,500 64,000 40,000 15,000 64,000
Count 12 12 12 12 12 12 12 12 48
Confidence
16,400 11,800 2,440 49,000 4,000 6,500 2,290 972 3,310
Level (95.0%)
%RSD 45 14 44 51 33 28 11 12 46
149
Figure 11-3. Box plots of soil minimum resistivity by soil mine.
11.3.2.2 Operating characteristic (OC) curve for minimum resistivity
Soil resistivity was measured with scaled meter readings, for example, from 100 to 1,000 ohm-
cm, 1,000 to 10,000 ohm-cm, and 10,000 to 100,000 ohm-cm; the standard deviation of the
measurement followed the scale and method precision across the scales was better described by a
2
relative standard deviation (%RSD). For minimum resistivity, the variance 𝜎𝑡𝑜𝑡𝑎𝑙 pooled across
eight soils and expressed as a squared relative standard deviation was 0.062 (Appendix D,
Section D-2, Table D-2). The portion of this that was related to single-laboratory analytical
testing was 0.00071 (Appendix B, Section B.9, Table B-36) and the multi-laboratory
2
contribution was 0.035 (Chapter 10, Section 10.4, Table 10-1, Material B). Thus, 𝜎𝑡𝑜𝑡𝑎𝑙
including single- and multi-laboratory analytical testing squared relative standard deviations
were 0.062 and 0.096, respectively, or %RSDs of 24.9% and 31.0%, respectively. In Figure 11-
4 were plotted for minimum resistivity and for N = 2 a series of OC curves with test errors  of
300, 600, 900, and 1200 ohm-cm or 10%, 20%, 30%, and 40% of the acceptance limit of 3,000
ohm-cm. A rejection limit of 1,000 ohm-cm was assumed (Chapter 2, Section 2.2.1). Section
150
Figure 11-4 suggested that for duplicate independent and random samples, the probability of
accepting a good backfill (or power, 1-β) at 3000 ohm-cm was 95% or better if method
reproducibility was at or below 600 ohm-cm (20%); for a method reproducibility of 900 ohm-cm
(30%), the power dropped to 79%.
Figure 11-4. Operating characteristic (OC) curve for minimum resistivity for sample
size N = 2.
As noted in Chapter 2 (Section 2.2.1), several options were available to improve the statistical
power: (1) change the acceptance limits, (2) increase the number of samples, (3) reduce the test
error, (4) shift more risk to the seller, (5) or modify the reinforcement design or material. Each
and every one of these options has cost implications, but from the results of the inter-laboratory
study there was room to reduce the test error; another possibility was to increase the acceptance
limit for minimum resistivity to 4,000 ohm-cm. In doing so, for sample size N = 2 and  = 900,
the power was 99%.
151
The difference in the means and variances between replicate and treatment studies suggested an
influence of material mixing and subsampling at the source on the observed sample minimum
resistivity (Table 11-4). Note, for example, that for mine 7 the average minimum resistivity
decreased 18% from 33,300 in the replicate study to 27,300 in the treatment study. From Table
11-4 the sample mean was significantly different between replicate and treatment studies for
soils from mines 7 and 8 and variances were significantly higher between replicate and treatment
studies for soils from mines 5 and 6. In other words, for soils from mine 7 and 8, material
handling between the mine and the laboratory affected the observed average minimum
resistivity; for mines 5 and 6, material handling between the mine and the laboratory diminished
the observed variability in minimum resistivity.
Table 11-4 Comparison of Replicate and Treatment Study Statistics for Minimum
Resistivity
Mine Identifier
Statistic 5 6 7 8
Replicate Average 19,100 36,400 33,300 12,500
Replicate Standard Deviation 6,300 10,200 3,600 1,530
Treatment Average 17,300 31,300 27,300 10,800
Treatment Standard Deviation 289 764 764 289
t-test (p) 0.36 0.12 0.00 0.00
F-test (p) 0.00 0.01 0.09 0.07
Elias et al. (2009) recommended that for an MSE wall three resistivity samples of sand be tested
pre-construction and one resistivity sample be tested every 3,058 m3 (4,000 yd3, which is ~200
truckloads, or ~140 lifts of 28 yd3 per lift) during construction. Results suggested that emphasis
on pre-construction sampling was well-placed due to the potential lack of uniformity of source
material with respect to minimum resistivity. Recommendations to improve the buyer’s
statistical power were:
 Revise the FM for minimum resistivity;
 Increase the minimum resistivity acceptance limit to 4,000 ohm-cm;
 Test at least two independent samples for minimum resistivity per soil type, both pre-
construction and post-emplacement;
 Provide operator training;
 Conduct routine on-site independent audits of the FM for minimum resistivity (see
Appendix G for an audit checklist); and
152
 Complete a Florida-wide inter-laboratory study of the revised FM within a year of its
adoption.
11.3.3.1 Distribution of chloride in mined soil
Soil chloride concentrations were below the stated method detection level for soil samples from
all eight mines.
11.3.3.2 Operating characteristic (OC) curve for chloride
For chloride, an inter-laboratory study yielded an estimate of 12 ppm for method reproducibility
(Chapter 10, Section 10.4, Table 10-1), although this estimate did not include contributions from
soil sampling or processing (Chapter 2, Section 2.2.3). The method reproducibility estimate was
for an average soil concentration of 66 ppm and did not extend to higher chloride concentrations
where further dilution of the soil extract compounded the error. An OC curve for chloride was
built assuming a sample size N = 1 and test errors of 20, 40, 60, and 100 ppm (Figure 11-5).
With a soil chloride concentration rejection limit of 500 ppm and acceptance limit of 100 ppm
(Chapter 2, Section 2.2.1), a test error or method reproducibility could be as high 100 ppm and
the probability of accepting a good backfill would still exceed 95%.
153
Figure 11-5. Operating characteristic (OC) curve for chloride for sample size N = 1.
No changes were recommended for the chloride concentration acceptance limit or the sample
size based on the OC curve (Figure 11-5). Moreover, since minimum resistivity is a measure of
dissolved salts concentration, a minimum resistivity exists above which the chloride
concentration will be below 100 ppm. At a solution temperature of 25C, for example, a 165-
ppm sodium chloride (100-ppm chloride) solution has a calculated resistivity of 2,900 ohm-cm
(Lide, 2001). The constituency of dissolved salts in a soil solution is more complex and the
resistivity less predictable than for pure solutions but it may be possible to over time to establish
a relationship between soil minimum resistivity and chloride concentration and from this
determine if a test for soil chloride concentration were necessary. Recommended was an
analysis of the minimum resistivity versus chloride concentration for soils across Florida after a
two-year period of data collection under the revised Florida methods.
154
11.3.4.1 Distribution of sulfate in mined soil
Of the samples collected from eight mines, samples from only four mines had sulfate
concentrations above the method’s stated detection level. In Table 11-5 were presented the
descriptive statistics and in Figure 11-6 the box plots for sulfate concentrations within and
between samples collected from these four mines. For MSE wall backfill, soil sulfate
concentration must be less than 200 ppm (Chapter 2, Section 2.1.2, Table 2-2). A single-factor
ANOVA revealed that the variability of soil sulfate concentration was greater between than
within soils (p = 0.00) (Appendix D, D.3, Table D-3) and the hypothesis that sulfate
concentration averages were equal across the eight soils was rejected. Between samples from
mines 1 and 3 and samples from mines 5 and 6, there was no evidence of a shift in sulfate results
for central tendencies (t-test, p = 0.22) but there was a discernible shift in dispersion (F-test, p =
0.00), driven by the large dispersion seen for Youngquist sand from mine 3. Results from
ruggedness studies for sulfate (Chapter 8, Section 8.3.4) and treatment studies for sulfate
(Chapter 9, Section 9.1.5) do hint that precision was improved for drying at a lower temperature.
Soil sulfate concentrations of replicates across mines 1, 3, 5, and 6 ranged from 0 to 102 ppm,
with an average (± standard deviation) of 20 (± 18) ppm, and at the 95% confidence level the
average sulfate concentration was between 15 and 25 ppm. The mean (average), median, and
mode (if present) sulfate concentration from samples of soils from mines 1, 3, 5 and 6 were well
below 200 ppm. Sulfate concentration standard deviations ranged from 4 to 30 ppm. Kurtosis
(peakedness) and skewness of the distributions is significant when these values exceed ± 2.0, for
example, but no distributions had significant kurtosis or skewness. The hypothesis that the
distribution of replicate sulfate measurements was normal was not rejected for soil from all four
mines except mine 6; the hypothesis that the distribution of replicate sulfate measurements was
uniform was not rejected for soil from all four mines.
155
Table 11-5 Descriptive Statistics for Soil Sulfate Concentration
Mine Identifier
Statistic
1 3 5 6 All
Mean 13 34 24 10 20
Standard Error 2 9 1 1 3
Median 11 29 24 9 15
Mode 9 0 21 9 9
Standard Deviation 6 30 4 4 18
Sample Variance 39 917 16 15 324
Kurtosis 0.23 0.90 0.92 0.03 9
Skewness 0.74 0.99 0.80 -0.56 3
Range 21 102 15 12 102
Minimum 3 0 18 3 0
Maximum 24 102 33 15 102
Count 12 12 12 12 48
Confidence Level (95.0%) 4 19 3 2 5
%RSD 49 90 17 40 90
156
Figure 11-6. Box plots of soil sulfate concentration by soil mine.
11.3.4.2 Operating characteristic (OC) curve for sulfate

2
For soil sulfate, the 𝜎𝑡𝑜𝑡𝑎𝑙 pooled across four soils was 247 (Appendix D, Section D.3, Table D-
3). The portion of this variance that was related to single-laboratory analytical testing was 10
(Appendix B, Section B.9.3, Table B-38). The estimate of multi-laboratory variance was 242
2
(Chapter 10, Section 10.4, Table 10-1, Material B). Thus, 𝜎𝑡𝑜𝑡𝑎𝑙 including single- and multi-
laboratory analytical testing variances were 247 and 479, respectively, or corresponding standard
deviations were 16 ppm and 22 ppm, respectively. In Figure 11-7 were plotted for sample size N
= 1 a series of OC curves for sulfate and test errors  of 20, 40, 60, 80 and 100 ppm. With a soil
sulfate concentration rejection limit of 1,000 ppm and acceptance limit of 200 ppm, the
probability of accepting a good backfill still exceeded 95% (curve not shown) for a test error as
high 200 ppm.
157
Figure 11-7. Operating characteristic (OC) curve for sulfate for sample size N = 1.
The difference in the means and variances between replicate and treatment studies suggested an
influence of material mixing and subsampling at the source on the observed sample sulfate
concentration (Table 11-6). Note, for example, that for mine 3 sulfate concentration was 5 ppm
higher for the treatment study than for the replicate study. From Table 11-6 sample means were
significantly different between replicate and treatment studies for none of the soils tested but
variance was significantly higher between replicate and treatment studies for soils from mine 6.
In other words, for soil from 6, material handling between the mine and the laboratory
diminished the observed variability in sulfate concentration.
158
Table 11-6 Comparison of Replicate and Treatment Study Statistics for Sulfate
Mine Identifier
Statistic 1 3 5 6
Replicate Average 13 34 24 10
Replicate Std Deviation 6 30 4 4
Treatment Average 10 39 24 12
Treatment Std Deviation 2 5 3 0
t-test (p) 0.21 0.58 1.00 0.07
F-test (p) 0.14 0.06 0.82 0
No changes were recommended for the sulfate concentration acceptance limit or the sample size
based on the OC curve (Figure 11-7). Moreover, since minimum resistivity represents a measure
of dissolved salts concentration, a minimum resistivity exists above which the sulfate
concentration will be below 200 ppm. At a solution temperature of 25 oC, for example, a 445-
ppm sodium sulfate (200-ppm sulfate) solution has a calculated resistivity of 1,380 ohm-cm
(Lide, 2001), for example. The constituency of dissolved salts in a soil solution is more complex
and the resistivity less predictable than for pure solutions but it may be possible to over time to
establish a relationship between soil minimum resistivity and sulfate concentration and from this
determine if a test for soil sulfate concentration were necessary. Recommended was an analysis
of the minimum resistivity versus sulfate concentration for soils across Florida after a two-year
data collection period under the revised Florida methods.
11.4 Summary
The statistical approach taken to estimate the appropriate number of samples for pre-construction
and post-emplacement testing assumed that independent, randomly-selected samples of soil were
collected and that soil pH, minimum resistivity, chloride, or sulfate concentrations were normally
distributed. Sampling, however, was subject to selection bias because of the practical limits on
geographical coverage, pre-selected characteristics of the soil, and accessibility of soil in
stockpiles or soil strata in pits, and the results reported herein are not fully representative of
Florida’s resources. Chloride concentrations in soils from all eight mines were below the stated
method detection level of 15 ppm. For pH, minimum resistivity, and sulfate concentrations, data
distributions were in general not different from normal but they were also in general not different
from uniform, a likely consequence of the limited number of replicates tested. Reported soil
sample variances included contributions from the material, sampling, processing, environment
and analysis, where analysis was accomplished by USF soils laboratory. These variances were
adjusted by subtracting single-laboratory contributions based on treatment studies and adding
multi-laboratory contributions from the inter-laboratory study to obtain an estimate of the total
2
test error √𝜎𝑡𝑜𝑡𝑎𝑙 for the method. Operating characteristic (OC) curves were developed for each
method and for a range of test errors, inclusive of the estimated total test error.
159
Based on the method OC curves, QA plan recommendations were as follows:
 For pH, revise the FM, test at least three independent samples for pH per soil type both
pre-construction and post-emplacement; provide operator training and routine on-site
laboratory audits of the FM (see Appendix G for an audit checklist); and conduct a
Florida-wide inter-laboratory study of the revised FM within a year of its adoption.
 For minimum resistivity, revise the FM, test at least two independent samples per soil
type for minimum resistivity both pre-construction and post-emplacement; impose a
higher acceptance limit for minimum resistivity of 4000 ohm-cm; provide operator
training and routine on-site laboratory audits of the FM (see Appendix G for an audit
checklist); and conduct a Florida-wide inter-laboratory study of the revised FM within a
year of its adoption.
 For chloride and sulfate, revise the FM, test at least one independent sample per soil type
both pre-construction and post-emplacement, and investigate the relationship between
minimum resistivity versus chloride and sulfate concentrations for soils across Florida
after a two-year data collection period under the revised Florida methods to see if further
testing of chloride and sulfate were necessary for select backfill.
160
12 Geochemical Modeling of Ion Transport in Sandy Soil
12.1 Problem Statement
Investigated through geochemical modeling was the hypothesis that soil mined from the
subsurface and placed on the surface either in a stockpile or as backfill in an MSE wall over time
undergoes a change in corrosion properties due to biological, chemical, or physical processes
induced by infiltrating rainfall. The goals of this analytical modeling effort were (1) to
recommend changes to the QA plan for corrosion AQC, if any, and (2) to assess the need for
more in-depth modeling supported by field and laboratory measurements that were beyond the
scope of this project. PHREEQC—pH redox equilibrium in C language—version 3 was chosen
for the modeling effort.
12.2 Model Description
PHREEQC v 3 was a publicly-available model maintained by the U. S. Geological Survey

(USGS) and was available for download at http://wwwbrr.cr.usgs.gov/projects
/GWC_coupled/phreeqc/. PHREEQC v 3 was “a computer program for speciation, batch
reaction, one-dimensional transport, and inverse geochemical calculations” and included a large
and expandable database of multi-phase equilibria and related dependencies for chemical
reactions of importance in groundwater, rainwater, seawater, soil water, and surface water
(Parkhurst and Appelo, 2013). Example input and output files were provided with the model
download, which supplemented a User’s Manual for the model (Parkhurst and Appelo, 2013).
Published model applications such as clean-up of nuclear contamination and acid mine drainage
(Hormann and Kirchner, 2002), drinking water treatment (Rozell, 2010), subsurface nutrient
transport (Domagalski and Johnson, 2011), soil profiling (Appelo,1994), and concrete leaching
(Schiopu et al., 2009) gave a sense of the model’s broad capabilities.
The PHREEQC input files were created and output files reviewed from a user-friendly interface
(Figure 12-1); also available at this interface were error files, databases, example files, and a
glossary of code words. Program operations were directed through a sequence of key words and
data. Program operations used for this modeling effort included (1) initial solution or speciation,
(2) initial exchange, and (3) advection calculations, in that order. Reaction temperatures were set
to 25 oC. For equilibrium calculations, units for input species were converted to moles/liter and
the moles of each species were distributed between all phases and exchange sites according to
databased reactions and partitioning coefficients (Table 12-1) (Parkhurst and Appelo, 2013). In
addition to cation exchange, the model was run with the groundwater and infiltrating rainwater in
equilibrium with carbon dioxide and quartz. Quartz was assumed to be abundant. For cation
exchange, groundwater was equilibrated with soil exchanger X, based on the soil’s cation
exchange capacity (CEC); the exchanger’s composition changed but not the groundwater
composition. PHREEQC uses by default a Gaines-Thomas method, where the activities of the
equivalent fractions (as opposed to the mole fractions) were used. Equations 12-1 and 12-2
161
shows the exchange were a multivalent ion replaces a monovalent ion. The databased
partitioning coefficients were Na+ + X-  NaX, log K = 0.0; K+ + X-  KX; log K = 0.7; Ca+2 +
2X-  CaX2, log K = 0.8; Mg+2 + 2X-  MgX2, log K = 0.6. The hydronium ion (H+) was not
included in the cation exchange. According to Bohn et al. (1985), in acid soils exchangeable
aluminum can be present and appears as exchangeable H+ when salt is added to displace the
adsorbed cations; otherwise exchangeable H+ can occur in soils with pH < 4.
A2+ + BX  0.5AX2 + B+ (12-1)
{𝐴𝑋2 }0.5 {𝐵+ }

𝐾𝐴/𝐵 = {𝐵𝑋} {𝐴2+ }0.5
(12-2)
Figure 12-1 User’s input screen for PHREEQC.

162
Table 12-1 Selected Aqueous-Phase or Precipitation Reactions
Aqueous-Phase or Precipitation Reactions Log K Dissolved Ions, Minerals and Gases (g)
Ca+2 + H2O  CaOH+ + H+ -12.8 aqueous
Ca + CO3  CaCO3
+2 -2
3.22 aqueous
Ca+2 + CO3-2 + H+  CaHCO3+ 11.4 aqueous
Ca+2 + HSO4-  CaHSO4+ 1.08 aqueous
CaCO3  CO3-2 + Ca+2 -8.34 aragonite
CaCO3  CO3-2 + Ca+2 -8.48 calcite
CaMg(CO3)2  Ca+2 + Mg+2 + 2 CO3-2 -17.0 dolomite
Ca+2 + SO4-2  CaSO4 2.25 aqueous
CaSO4  Ca+2 + SO4-2 -4.28 anhydrite
CaSO4:2H2O  Ca+2 + SO4-2 + 2 H2O -4.58 gypsum
CO2  CO2 -1.46 CO2(g)
CO3 + H  HCO3
-2 + -
10.3 aqueous
CO3-2 + 2 H+  CO2 + H2O 16.7 aqueous
H2  H2 -3.10 H2(g)
2 H+ + 2 e-  H2 -3.15 aqueous
H2O  OH- + H+ -14.0 aqueous
H2O  H2O 1.50 H2O(g)
2 H2O  O2 + 4 H + 4 e
+ -
-86.1 aqueous
KCl  K+ + Cl- 0.90 sylvite
K + SO4  KSO4
+ -2 -
0.85 aqueous
Mg+2 + H+ + CO3-2  MgHCO3+ 11.4 aqueous
Mg + H2O  MgOH + H
+2 + +
-11.4 aqueous
Mg + CO3  MgCO3
+2 -2
2.98 aqueous
Mg + SO4  MgSO4
+2 -2
2.37 aqueous
Na + CO3  NaCO3
+ -2 -
1.27 aqueous
NaCl  Cl + Na - +
1.57 halite
Na + HCO3  NaHCO3
+ -
-0.25 aqueous
Na + SO4  NaSO4
+ -2 -
0.70 aqueous
O2  O2 -2.89 O2(g)
SiO2 + 2 H2O  H4SiO4 -2.71 SiO 2(a)
SiO2 + 2 H2O  H4SiO4 -3.55 chalcedony
SiO2 + 2 H2O  H4SiO4 -3.98 quartz
SO4 + H  HSO4
-2 + -
1.99 aqueous
Transport by advection was modeled without dispersion or diffusion, which simulated plug flow.
A theoretical soil column was divided into adjacent cells and the fluid in each cell was initialized
to that of groundwater in equilibrium with carbon dioxide, quartz, and exchanger X. The
infiltrating rainfall was ‘shifted’ from cell 0 to cell 1 at the first time step; simultaneously, the
groundwater in cell 1 was shifted to cell 2, and so forth. At each time step, the fluids moved
163
another step down the column. Once in a new cell, the fluid was equilibrated with carbon
dioxide, quartz, and exchanger X.
12.3 Model Inputs and Parameters
For 1-D transport modeling in PHREEQC, soil with isotropic properties was assumed to be in
the form of a column with dimensions of 6 m x 6 m x 6 m = 216 m3, and to have an available
pore volume of 30%. Flow of rainwater was assumed to be vertical and downward through the
column. To model unsaturated flow, the column was sliced horizontally into 60 equal cells, such
that each cell was water-saturated during a weekly rainfall event of 30-mm (1.2 in) or greater,
which was assumed to happen 15 times per yr based on the probabilities in Figure 12-2 (NADP,
2015). Thus, the total pore volume in the column was 0.030 m x 6 m x 6 m = 1.08 m3 x 60 cells
= 64.8 m3, and with a time step or ‘shift’ of 1/15 yr, one column’s pore volume was exchanged in
60 shifts or 4 years.
Figure 12-2. Cumulative probability for weekly rainfall amount at NADP’s FL41
site in Sarasota County (2000 through 2012).
For rainwater, annual average precipitation-weighted concentrations for major ions were
downloaded from the National Atmospheric Deposition Program (NADP) website for a National
164
Trends Network (NTN) site FL-41 in Sarasota County (NADP, 2015), and these concentrations
were averaged over the years 2000 to 2012 (Table 12-2). Note that the precipitation-weighted
average concentrations were lower than the weekly median concentrations reported in Chapter 2,
(Section 2.1.3, Table 2-5). The average rainfall amount was 1300 mm (940 mm to 1650 mm)
and the electrical conductivity (EC) of rainwater was ~10 µS/cm. The nitrogen species—
ammonium (NH4+) and nitrate (NO3-)—were not input and the chloride concentration was
adjusted upward from 0.50 to 0.68 ppm to maintain a charge balance. Groundwater median ionic
composition was based on 11 samples collected from surficial aquifer wells in Ocala National
Forest (Schiffer, 1989). The EC and alkalinity of groundwater were 27 µS/cm at 25C and 0.82
ppm, respectively (Schiffer, 1989).
Table 12-2 Modeled ionic composition of rainwater and groundwater

Units are ppm
Source pH Lab
Ca++ Cl- K+ Mg++ Na+ SO4=
Rainwater 4.9 0.1 0.5 0.02 0.04 0.3 0.7
Groundwater 5.0 0.4 4 0.1 0.5 2.1 0.6
Units are moles/liter
Source pH Lab
Ca++ Cl- K+ Mg++ Na+ SO4=
Rainwater 4.9 2.50E-6 1.92E-5 5.12E-7 1.65E-6 1.31E-5 7.29E-6
Groundwater 5.0 9.98E-6 1.13E-4 2.56E-6 2.06E-5 9.14E-5 6.25E-6
Soil properties were based on Candler soil, an uncoated hyperthermic Typic Quartzipsamments,
which is a permeable, well-drained sandy soil present in central Florida in thick deposits, and
one of the Lake series soils based on a classification by the U. S. Department of Agriculture
(Table 12-3). Candler soil pH, % organic matter (%OM), cation exchange capacity (CEC), and
EC were obtained from Elrashidi et al. (2001). Note the reasonable agreement between
groundwater and soil pH and EC.
Table 12-3 Candler Soil Properties
Soil pH %OM CEC, mol/kg EC, µS/cm
Candler 5.05 0.65 0.0405 30
Each cell had a soil volume of 3.6 m3 x 70% = 2.52 m3 and with an assumed sand density of
2,500 kg/m3, the mass of soil in each cell was 6,300 kg and the CEC per cell was 6,300 kg x 0.04
moles/kg  250 moles of exchange sites per cell. Perhaps the soil closest in composition to
Candler soil was from Mine 2 in Haines City (nicknamed Jahna sand) with a pH of 5.16, a
resistivity of 54,000 ohm-cm (19 µS/cm), and 0.08 %OM.
Refer to Appendix E for the PREEQC input and output files. Due to the file length, the output
included some but not all of the cell equilibration results. Temporal values for pH, calcium,
potassium, and sodium were printed to a separate file for plotting. Carbon dioxide concentration
was initially adjusted such that the groundwater pH did not change much from the reported pH
5.0 during the equilibration step. In the simulation carbon dioxide equilibrated with groundwater
165
at a partial pressure of ~12,000 ppm, which was in line with subsurface CO2 concentrations
reported and modeled by Macpherson (2009).
12.4 Model Results and Discussion
Modeled was a conservative case of rainwater and soil interaction, conservative in the sense that
the groundwater/soil system modeled had a low buffering capacity and a pH already close to that
of rainwater. In this simulation, at four yr the last cell shifted from properties of groundwater to
properties of rainwater, where both were equilibrated with the exchange sites, with ~12,000-ppm
carbon dioxide, and with quartz. Precipitation of solids including calcite, dolomite, gypsum, etc.,
listed in Table 12-1 was not predicted, but chalcedony (SiO2) had a saturation index of -0.43, the
closest of any of the solids to saturation. With rainwater infiltration, the soil exchange sites
picked up calcium and let go potassium and chloride; moreover, the pore water pH decreased
from 5.05 to 4.67 (Figure 12-3). (Modeled pore water pH decreases to below pH 5.00 even if
soil column carbon dioxide concentration is assumed to equilibrate with the atmospheric
concentration.) These results suggested that the possibility existed for well-leached, poorly
buffered quartz-type sands to reach a new equilibrium with rainwater within a few years after
emplacement in a stockpile or behind an MSE wall.
166
Figure 12-3. Modeled groundwater ion concentrations and pH at soil column exit.
12.5 Summary
Geochemical modeling with the USGS model PHREEQC for a conservative case of a low-ionic
strength and poorly buffered sandy (quartz) backfill revealed that in a few years’ time the pore
water of stockpiled or emplaced backfill could equilibrate with infiltrating rainfall. The model
results suggested that a soil’s buffering capacity might be important consideration when metal is
used as reinforcement in MSE wall backfill. Field and laboratory testing of candidate sands to
calibrate the model were recommended ahead of any proposed changes to the QA plan based on
model results.
167
13 Revised Florida Methods for pH, Resistivity, Chloride and Sulfate
13.1 Introduction
Results of field testing and laboratory analyses suggested that the accuracy and precision of the
Florida Methods (FMs) for pH, minimum resistivity, chloride, and sulfate could be improved
with relatively small changes in method procedures, as discussed in Sections 13.2 through 13.6.
Proposed changes to the FMs fall into one of five categories: (1) minor edits to method grammar,
organization, and units of measure; (2) updates to the apparatus and reagents sections, with a
hazardous materials note as appropriate; (3) step-by-step instructions to facilitate method
oversight; (4) one or more quality assurance/quality control procedures; and (5) a new section for
method precision and bias. Revised FMs for pH, minimum resistivity, chloride, and sulfate are
found in Appendix F. Note that the proposed changes were based on studies performed with
select backfill and may not extend to water samples or to soil samples with a significant content
of salt, clay or organic matter.
13.2 FM for pH
Changes to the scope, apparatus, reagent, samples, sample preparation, and test procedure
sections of the FM were proposed as was the inclusion of a precision and bias section in the FM.
Based on ruggedness and other factor analysis studies as well as laboratory visits (Chapters 6, 8,
and 9) these major sources of variability were identified in the FM using a calibrated pH
meter/electrode system:
 Condition of the electrode,

 Measurement temperature,
 Electrode memory effects, and
 Ionic strength of the soil solution.
The pH meter combination electrode is a delicate and sensitive sensor and with routine treatment
and use its lifetime is ~18 months. A 3-in-1 combination electrode typically encases a glass-bulb
indicating electrode, a silver/silver chloride reference electrode, and a thermocouple. Over time
and with use the reference electrode is consumed, but temporary or permanent damage to the
electrode can be caused by contamination or depletion of the filling solution, scratches or cracks
on the glass bulb surface, and clogging of the liquid junction, as examples. Electrode condition
can be monitored through calibration and performance checks. The revised method calls for a
three-buffer (pH 4.00, 7.00, and 10.00) calibration, a calibration slope between 95% and 102% of
theoretical, an offset of less than 10 mV, and a minimum 10-sec stable pH reading. Attention to
electrode performance is critical to a precise and accurate pH measurement.
Measured pH is inversely proportional to temperature and pH meter/electrode systems are

typically designed for a standard temperature of 25C (Chapter 3, Section 3.2.1). A pH meter
with automatic temperature compensation (ATC) measures and corrects for the sample
168
temperature. Ideally, the buffers and the samples are 25C, but with ATC, buffers and samples
need be at or near the same temperature. The proposed revised method specified a pH/mV meter
with ATC and equipped with a 3-in-1 combination electrode (Appendix F.1).
For pH measurement in low-ionic strength samples, the observed pH may be biased high because
of the mismatch between the ionic strength of the calibration buffers and the sample and because
of sample dilution. Select backfill, by virtue of its low fine particle content, low chloride and
sulfate content, and high resistivity has relatively low ionic strength. According the pH
meter/electrode sales literature (Bier, 2009; Thermo-Scientific, 2007), this bias can be reduced in
three ways: (1) with use of low-ionic strength buffers, (2) with use of an electrode that has a high
flow rate of electrolyte through its liquid junction, and (3) by increasing the ionic strength of the
sample. The latter approach is the easiest way to reduce measurement bias in low-ionic strength
water samples but causes a significant decrease in pH for soil samples, where the soil surface and
pore water equilibria influence proton activity (refer to Chapter 9, Section 9.2.7, for further
discussion).
Equilibration period and storage temperature were two other factors that had a significant effect
on pH in ruggedness or treatment studies. For pH, a significant difference was seen between a
~0-min and a 30-min equilibration period but no significant difference between a 30-min and a
60-min equilibration period; therefore, a 30-min equilibration period was retained in the method.
In the treatment studies with soils from mines 5-8, pH of samples stored in refrigerator and then
warmed to room temperature were closest to the pH of samples measured on site. Therefore,
instructions for cool transport and cold storage were retained in the method.
The electrode has a memory of its previous measurement and the contribution of this memory on
the current pH measurement was more pronounced for low-ionic strength (low salt content)
solutions. The memory effect on sample pH could be reduced by careful attention to rinsing and
drying the electrode between measurements and by a second or even third consecutive
measurement of sample pH.
For pH measurement in low-ionic strength solutions a stable pH reading could take 5 minutes or
more to obtain, as the reading is affected by the exchange of carbon dioxide between the sample
and the atmosphere, the effects of stirring and vibrations, and a mismatch between the ionic
strength of the sample and the reference solution in the electrode. If after ~ 5 min the pH reading
does not stabilize in a gently stirred sample, a pH reading in the unstirred sample should be
recorded.
13.3 FM for Minimum Resistivity
Changes to the scope, apparatus, reagent, samples, sample preparation, and test procedure
sections of the FM were proposed as well as the inclusion of a precision and bias section in the
FM. Based on ruggedness and other factor analysis studies as well as laboratory visits (Chapters
6, 8, and 9), these major sources of variability were identified in the FM using a Nilsson Model
400 resistivity meter:
169
 Water content of soil slurry placed in the soil box during resistivity testing, and
 Measurement temperature.
The components of soil resistivity are soil liquids, solids, and voids (Chapter 3, Section 3.3.1).
As the water content of a soil sample is increased beyond saturation, the path that ions must
travel is increasingly free of impediment and soil resistivity drops until the soil is a dilute
suspension in water. At the same time, the addition of deionized water dilutes the concentration
of dissolved ions, which raises the resistivity. Minimum resistivity occurs at the intersection of
these two effects. ASTM and AASHTO standards represented two different positions along the
soil resistivity continuum: ASTM G187 measured soil resistivity when the soil is just saturated
and AASHTO T288 measured minimum soil resistivity, with the water from soil slurry decanted
into the soil box as the water content in the slurry increases. The FM for minimum resistivity did
not provide direction on how to dispense soil slurry into the soil box, so this decision was left to
the method operator. The proposed revised FM adopted the more conservative approach of
AASHTO T288.
Resistivity measurements are sensitive to temperature (Chapter 3, Section 3.3.1); for example,
resistivity can decrease by 10% if the measurement temperature is increased from 23C to 28C.
Minimum resistivity measurements should be made at room temperature and the soil slurry
temperature recorded during the measurement procedure in case temperature played a significant
role in observed soil minimum resistivity between quality assurance, verification, and resolution
samples.
A higher ion concentration in dilution water translates into a lower resistivity. Dilution water
resistivity should be high enough not to contribute to or obscure the resistivity of the wetted soil,
and ASTM D1193 Type 4 reagent water, which has a lower resistivity bound of 200,000 ohm-
cm, was recommended. A routine check of dilution water as part of laboratory QA/QC was
recommended because on-site water treatment systems can experience break-through of ions and
bottled water—once opened—can over time absorb ions from acidic and basic gases in the
atmosphere.
The resistivity meter calibration can be checked with one or more resistors that span the range of
interest but a resistivity measurement of a conductivity standard will check not only the meter,
but the lead connections, soil box, and measurement temperature. A resistivity meter calibration
with a 250 µS/cm (4,000 ohm-cm) calibration standard at least once per quarter was proposed.
13.4 FM for Chloride
Changes to the scope, apparatus, reagents, samples, sample preparation, and test procedure
sections of the FM were proposed, as well as the inclusion of a precision and bias section in the
FM. The FM for chloride in soil and water is based on a silver nitrate titration of a sample that is
amended with potassium chromate. Dissolved silver combines first with chloride and then with
chromate in a mixture of the two. Both silver chloride and silver chromate are insoluble in water
170
but silver chromate is a brown-red crystal and its appearance in the sample marks the titration
endpoint. Hach chloride test kits (or equivalent) were recommended for laboratory-based
screening of water and soil chloride concentrations; for a greater degree of accuracy and
precision the analytical methods as described by the Standard Method for the Evaluation of
Water and Wastewater (SMEWW) 4500-Cl- or 4110 B for anions (Rice et al., 2012 ) were
recommended. The SMEWW method 4500- Cl- uses smaller increments of titrant and imposes
additional QA/QC; SMEWW method 4110 B can detect lower anion concentrations but is not
appropriate for seawater samples. Based on ruggedness and other factor analysis studies as well
as laboratory visits (Chapter 6, 8, and 9), these major sources of variability in the FM were
identified using the chloride test kit:
 Soil mass,
 Suspended solids or color in the filtered sample extract,
 Incorrect or out-of-date reagents, and
 Blank correction.
For chloride determination an accurate soil mass can be obtained if the soil is dried to a constant
mass and its mass weighed on an analytical balance; these two requirements limit the portability
of the method. In laboratory studies, suspended particles or color present in sample extracts
interfered with the color change at the titration endpoint, with the effect that extra titrant was
added and the chloride concentration in soil was thus biased high. Attempts to clarify filter
extracts included ferric sulfate addition, a 1:1 water-to-soil ratio in the soil extraction, serial
gravity filtration using Whatman 4 and Whatman 42 filters, centrifugation, and vacuum filtration
through a membrane filter (Chapters 5, 9). One effective combination for clearing the sample
extract was to add ~0.2 g of ferric sulfate dissolved in 1 mL of water to the sample extract,
followed by at least three hr of settling time, gravity filtration through a Whatman 41 filter, and
finally vacuum filtration through a 0.45-micron filter. Vacuum filtration through a 0.45-micron
membrane filter was recommended as a final filtration step. Also recommended were separate
3:1 water-to-soil extracts for chloride and sulfate, so that ferric sulfate could be added directly to
the unfiltered extract, if needed. If ferric sulfate were not needed, only one extract has to be
filtered for chloride and sulfate determinations. A chloride check standard and an associated
reagent blank were included in the proposed revised FM to help establish the sample color at the
titration endpoint, to guard against bad reagents, and to subtract out any chloride contamination
in the dilution water.
13.5 FM for Sulfate
Changes to the scope, apparatus, reagents, samples, sample preparation, and test procedure
sections of the FM were proposed, as well as the inclusion of a precision and bias section in the
FM. The FM for sulfate in soil and water in based on the reaction of water-dissolved sulfate and
barium to form an insoluble precipitate of barium sulfate. The turbidity produced by this
precipitate is proportional to the sulfate concentration in the sample. The Hach sulfate
photometer test kit (or equivalent) was proposed for laboratory-based screening of sulfate in
171
water and soil; for a greater degree of accuracy and precision the analytical methods as described
by the Standard Method for the Evaluation of Water and Wastewater (SMEWW) 4500 for sulfate
and 4110 B for anions (Rice et al., 2012) were recommended. The SMEWW method 4500 for
sulfate includes additional reagents to stabilize the barium sulfate precipitate, a user-generated
calibration curve, and additional QA/QC; SMEWW method 4110 B can detect lower anion
concentrations but is not appropriate for seawater samples. Based on ruggedness and other
factor analysis studies as well as laboratory visits (Chapters 6, 8, and 9), these major sources of
variability in the FM were identified using the sulfate photometer test kit:
 Soil mass,
 Incorrect or out-of-date reagents, and
 Sulfate calibration curve.
For sulfate determination an accurate soil mass can be obtained if the soil is dried to a constant
mass and its mass weighed on an analytical balance; these two requirements limit the portability
of the method. In laboratory studies, suspended particles or color present in sample extracts
interfered with sulfate measurements and caused either an invalid reading when the sample blank
had a higher photometer response than the reacted soil extract or an elevated sulfate
concentration when the suspended particles or color were not effectively subtracted from the
sample with the sample blank (Chapters 5, 9). Attempts to clarify the soil extract included acid
addition, a 1:1 water-to-soil ratio in the soil extraction, serial gravity filtration using Whatman 4
and Whatman 42 filters, centrifugation, and vacuum filtration through a membrane filter. One
effective combination for clearing the sample extract was acid addition followed by at least one
hr of settling time, gravity filtration through a Whatman 41 filter, and finally vacuum filtration
through a 0.45-micron filter. Vacuum filtration through a 0.45-micron membrane filter was
recommended as a final filtration step. Also recommended were separate 3:1 water-to-soil
extracts for chloride and sulfate, so that acid could be added directly to the unfiltered extract, if
needed. If acid were not needed, only one extract has to be filtered for chloride and sulfate
determinations. A photometer with factory-programmed sulfate calibration curves gave
reasonable results provided the curve was routinely checked, as the potency of the reagents used
in the method varied by lot number and possibly with the age of the reagents. A sulfate check
standard and associated reagent blank were included in the proposed revised FM to help to guard
against bad reagents, contamination of dilution water, or a mismatched photometer response.
13.6 Proposed Revisions to the FM
Proposed revisions to the FMs for pH, resistivity, chloride and sulfate were sent to project team
members, 18 commercial laboratories, and the District and State Material Offices. The revised
methods were demonstrated at FDOT’s Districts 1 and 7 Materials Office and State Material
Office. Comments were received from those laboratory managers and operators. Changes that
stemmed from these comments included
172
 For the FM for minimum resistivity, a shift in the placement of routine checks of dilution
water and a calibration standard from test procedure to a new best practice section;
 For the FMs for chloride and sulfate, addition of the ion chromatography method Section
4110 B from Standard Methods for the Examination of Water and Wastewater (Rice,
2012) as an acceptable method for sulfate and chloride analyses; procedural steps that
clearly include both water and soil extract analyses; and addition of steps for high-range
chloride tests, e.g., for seawater samples.
For the FMs for pH and minimum resistivity, changes that were made based on the results of
ruggedness and treatment study testing of Alico Road, Angelo’s, Clermont, and Sebring soils
(mines 5 through 8) were:
 Return to storing samples for pH measurements under refrigeration;
 Return to testing samples “as received” for pH and resistivity; and
 Removal of the steps for adding 0.1 g potassium chloride to pH samples.
These changes were included in the version of the FMs that were delivered to FDOT laboratories
for the inter-laboratory study, with the exception for the FM for pH, where steps were added to
address electrode memory effects ahead of the second inter-laboratory study (pH only). Proposed
revised FMs are included in Appendix F.
173
14 Summary and Recommendations
The ultimate goals of this research were to improve quality, speed completion, and reduce risk in
mechanically-stabilized earth (MSE) wall projects. Research objectives were (1) to assure that
variability in the corrosion properties of soil (pH, minimum resistivity, chloride, and sulfate
levels) due to sampling and analytical technique was much lower than variability in these levels
within and between soil sources and types and thus did not inflate the risk of emplacing a
corrosive soil as MSE wall backfill, (2) that the number of soil type samples analyzed prior to
acceptance of a backfill was appropriate, based on the expected distribution of corrosion
properties within the backfill, and (3) that the corrosion properties of backfill material did not
change appreciably over time, especially after emplacement and over the design lifetime of the
MSE wall. Corrosion properties of soil were tested with Florida Methods (FMs) 5-550, 5-551, 5-
552, and 5-553 for pH, minimum resistivity, water-soluble chloride, and water-soluble sulfate,
respectively.
The scope of this research included tasks to (1) review relevant literature, (2) examine trends in
corrosion properties of MSE wall backfill from Districts 1 and 7 mines based on existing data,
(3) determine the single-laboratory contributions to method reproducibility of factors that relate
to field sampling and laboratory analysis and improve method reproducibility through changes in
the methods, (4) discover the distribution of corrosion properties within and between mines for a
representative sample of mines as suggested by objective (2), (5) model analytically the potential
for the corrosion properties of MSE backfill to change over time with an initial focus on changes
due to the ionic composition of rainfall, (6) determine multi-laboratory contributions to method
reproducibility, and (7) disseminate the project results as a final report. This research was
limited to 10 soils that were candidate materials for select backfill and thus low in clay, organic
matter, and salt content; these soils were typical of MSE wall backfill but not likely
representative of the broader variability of Florida waters and soils.
acceptance data as downloaded from LIMS were for the most part broken. A recommendation
was that future documentation include the geo-coordinates (latitude and longitude, for example)
of the borrow pit or stockpile of certified MSE wall backfill, and that these geo-coordinates
accompany the certification data as well as the acceptance data into LIMS.
Single-operator operator studies conducted in the University of South Florida (USF) soils
laboratory yielded reasonable accuracy (%RE) and precision (%RSD) for pH and minimum
resistivity, but not for chloride and sulfate. Chloride and sulfate analyses were subject to
interferences from suspended particles and color in soil extracts, which were significantly
reduced by final filtration of the soil extract through a 0.45-micron mixed cellulose ester
membrane. In multi-laboratory studies such as split-sample analyses, soil testing during
laboratory visits, and an inter-laboratory study, differences seen in results indicated systematic
variability between laboratories. These differences were explored through ruggedness,
treatment, and other factor analysis studies, with a goal to reduce these differences in part
through changes in sample processing and analysis procedures. As an example, a proposed
174
revision to the FM for pH calls for checking the electrode condition using the slope and offset of
the calibration curve before making a pH measurement. As another example, a proposed
revision to the FM for minimum resistivity preferentially uses the water portion of the soil slurry
for the resistivity measurement.
Changes to the FMs for minimum resistivity, chloride, and sulfate that were trialed by USF
showed promise for improving between-laboratory agreement. Split sample analyses revealed
that four of the eight sands failed to meet an acceptance limit for one or more of the corrosion
properties in one or more of the partnered laboratories. If the split sample analyses between two
laboratories represented quality control and verification tests, at least three out of the eight sands
would have required resolution testing. Reduction in a relatively high rate of resolution testing
may offset additional costs associated with method improvements.
Some sands that met the specifications for MSE wall backfill had high minimum resistivity (low
conductivity) and thus low ionic strength, and were a challenge for pH measurements.
Potassium chloride, a neutral salt, is a possible amendment to soil samples for pH testing. Three
advantages to the addition of 0.1 g of potassium chloride to 100 g of soil sample were: the pH
reading was faster, more stable, and was more likely to represent the in-situ soil pH; the major
disadvantage, however, was a drop of ~0.5 pH units in 1:1 soil:water solutions. Such a shift
would likely disqualify some of Florida’s sands as candidates for MSE wall backfill. The use of
potassium chloride or calcium chloride to adjust the ionic strength of soil solutions during pH
measurement was not recommended at this time. The observed downward shift in pH in sandy
soils with neutral salt addition, however, may indicate that these soils have significant total
acidity. Recommended was further investigation of the relationship between metal corrosion
rates in select backfill and soil acidity, where soil acidity is determined as both proton activity
(electrochemical pH measurement) and exchangeable acidity.
Replicate studies of mined soils for each corrosion FM produced an estimate of test error that
included contributions from material properties, environmental influences, and sampling,
processing, and single-operator testing (method repeatability) of soils. For each FM, its test error
estimate was adjusted to include the multi-laboratory testing (method reproducibility)
contribution. The adjusted estimate was incorporated into a method operating characteristic
(OC) curve, which had a Type I () error of 0.01 and a Type II () error of 0.05; the OC curve
related the buyer’s statistical power (1-) to accept a good backfill to the number of samples, test
error, and acceptance limit(s).
Changes to the quality assurance (QA) plan for acceptance quality characteristics (AQC) of
backfill corrosion properties were recommended based on research outcomes. These changes
had as their focus an improvement in the buyer’s statistical power to accept good backfill
material through reductions in test error associated with material properties, sample processing,
and laboratory procedures. Recommended changes included (1) proposed revisions to the FMs
for pH, minimum resistivity, chloride, and sulfate, respectively; (2) an increase in the number of
175
independent samples tested for pH and minimum resistivity; (3) a revised acceptance limit for
minimum resistivity; (4) method operator training and independent audits; (5) an inter-laboratory
study post-implementation of the revised methods; and (6) re-evaluation of the need for chloride
and sulfate testing for backfill above a pre-determined minimum resistivity. Recommended
changes were outlined below:
 Revise the FMs for pH, minimum resistivity, chloride and sulfate. Proposed revisions to
the FMs included (1) minor edits to method grammar, organization, and units of measure;
(2) updates to the apparatus and reagents sections, with a hazardous materials note as
appropriate; (3) step-by-step instructions to facilitate method oversight; (4) one or more
QA/QC procedures; and (5) a new section for method precision and bias.
 Increase the number of independent samples per soil type for pH and minimum
resistivity. Recommended for pH were at least three independent pH samples per soil
type tested both pre-construction and post-emplacement. Recommended for minimum
resistivity were at least two independent minimum resistivity samples per soil type tested
both pre-construction and post-emplacement and an increase in the acceptance limit for
minimum resistivity to 4,000 ohm-cm.
 Conduct operator training and routine on-site laboratory audits of the FMs. Operator
training was recommended for all FMs, for example, training through a video
demonstration of each method as well as method outcomes for both well-executed and
poorly-executed procedures. Also recommended for FMs were routine independent
audits, which provide oversight on method performance and put the method and operator
in a context that includes environmental influences (temperature, noise, dust, crowding),
laboratory practices, and pressure for time.
 Perform a Florida-wide inter-laboratory study of the revised FMs. A Florida-wide inter-

laboratory study of the revised FMs within a year of adoption was recommended to
assure that FDOT and commercial laboratories were proficient with the revised methods
and to establish for each FM the method reproducibility.
 Re-evaluate the FMs for chloride and sulfate. A review of Florida-wide post-revision
chloride and sulfate results for MSE wall backfill after a two-year data collection period
was recommended to assess if further testing of chloride and sulfate were necessary for
select backfill above a pre-determined minimum resistivity.
Geochemical modeling with the U. S. Geological Survey (USGS) model PHREEQC for a
conservative case of a low ionic strength and poorly buffered sandy (quartz) backfill revealed
that in a few years’ time the pore water of stockpiled or emplaced backfill could equilibrate with
infiltrating rainfall. The model results suggested that a soil’s buffering capacity might be an
important consideration when metal is used as reinforcement in MSE wall backfill. Field and
176
laboratory testing of candidate sands to calibrate the model were recommended ahead of any
proposed changes to the QA plan based on model results.
177
References
American Concrete Institute (ACI). (2011). Building code requirements for structural concrete
(ACI 318-11): An ACI standard and commentary. Farmington Hills, Michigan.
American Association of State Highway and Transportation Officials (AASHTO). (2012).

AASHTO LRFD bridge design specifications, customary U. S. units (publication code LRFDUS-
6). Washington, D. C.

Standard recommended practice for acceptance sampling plans for highway construction, R9.
Washington, D. C.

Standard method of test for determination of organic content in soils by loss on ignition, T267-
86. Washington, D. C.

Standard method of test for determining pH of soil for use in corrosion testing, T289-91.
Washington, D. C.

Standard method of test for determining the plastic limit and plasticity limit index of soils, T90-
00. Washington, D. C.

Standard method of test for determining water-soluble chloride ion content in soil, T291-94.
Washington, D. C.

Standard method of test for particle size analysis of soils, T88-13. Washington, D. C.

Standard method of test for sampling of aggregates, T2-91. Washington, D. C.

Standard method of test for reducing samples of aggregate to testing size, T248-11. Washington,
D. C.

Standard method of test for sieve analysis of fine and coarse aggregates, T27-11. Washington,
D. C.
178
Standard method of test for determining minimum laboratory soil resistivity, T288-12.
Washington, D. C.

Standard method of test for determining water-soluble sulfate ion content in soil, T290-95.
Washington, D. C.

Standard method of test for laboratory determination of moisture content of soils, T265-12.
Washington, D. C.

Standard method of test for determining the liquid limit of soils, T89-13. Washington, D. C.
Appelo, C. A. J. (1994). Cation and proton exchange, pH variations, and carbonate reactions in a
freshening aquifer. Water Resources Research, 30, 2793-2805.
ASTM International (ASTM). (2003). Standard practice for preparing precision and bias
statements for test methods for construction materials, C 670-03. West Conshohocken,
Pennsylvania.
ASTM International (ASTM). (2003). Standard practice for reducing samples of aggregate to
testing size, C702-98. West Conshohocken, Pennsylvania.
ASTM International (ASTM). (2004). Standard test methods for chloride ion in water, D512-04.
West Conshohocken, Pennsylvania.
ASTM International (ASTM). (2006). Standard guide for good laboratory practices engaged in
sampling and analysis of water, D3856-95. West Conshohocken, Pennsylvania.
ASTM International (ASTM). (2006). Standard specification for reagent water, D1193-06. West
Conshohocken, Pennsylvania.
ASTM International (ASTM). (2006). Standard test method for water-soluble sulfate in soil,
C1580-06. West Conshohocken, Pennsylvania.
ASTM International (ASTM). (2007). Standard practice for conducting a ruggedness or

screening program for test methods for construction materials, C1067-00. West Conshohocken,
Pennsylvania.
ASTM International (ASTM). (2008). Standard practice for use of the terms precision and bias
in ASTM Test Methods, E177-08. West Conshohocken, Pennsylvania.
179
ASTM International (ASTM). (2009). Standard practice for sampling aggregates, D75/D75M-
09. West Conshohocken, Pennsylvania.
ASTM International (ASTM). (2009a). Standard practice for conducting an interlaboratory test
program to determine the precision of test methods for construction materials, C802-09a. West
ASTM International (ASTM). (2012). Standard test method for measuring pH of soil for use in
corrosion testing G51-95. West Conshohocken, Pennsylvania.
ASTM International (ASTM). (2012a). Standard test method for measurement of soil resistivity
using the two-electrode soil box method, G187-2012a. West Conshohocken, Pennsylvania.
ASTM International (ASTM). (2013). Standard test method for pH of soils, D4972-13. West
Beddoe, R. E., & Dorner, H. W., (2005) Modelling acid attack on concrete: Part I. The essential
mechanisms. Cement and Concrete Research, 35, 2333-2339.
Berke, B. S., Sagüés, A. A., & Powers, R. G. (2008). Long term corrosion performance of soil
reinforcement in mechanically stabilized earth walls (paper no. 08319). In NACE International
Corrosion Conference & Exposition. Retrieved October 11, 2013, from
http://www.eng.usf.edu/~sagues/Documents/
Bier, A. W. (2009). pH measurement in low-ionic strength samples (literature no. 2073).

[Brochure]. Loveland, Colorado: HACH Company.
Bohn, H. L., McNeal, B. L., & O’Connor, G. A. (1985). Soil chemistry (2nd ed.). New York, New
York: John Wiley & Sons.
Brennan, R. F., & Bolland, M. D. A. (1998). Relationship between pH measured in water and
calcium chloride for soils of Southwestern Australia. Communications in Soil Science and Plant
Analysis, 29, 2683-2689.
Brezinski, D. P. (1983). Kinetic, static and stirring errors of liquid junction reference electrodes.
The Analyst, 108, 425-442.
Burati, J. L., Weed, R. M., Hughes, C. S., & Hill, H. S. (2003). Optimal procedures for quality
assurance specifications, FHWA-RD-02-095). McLean, Virginia: Federal Highway
Administration.
Busenberg, E., & Plummer, L. N. (1987). pH measurement of low-conductivity waters, water-

resources investigations report 87-4060. Reston, Virginia: U. S. Geological Survey.
180
Cao, H. T., Bucea, L., Ray, A., & Yozghatlian, S. (1997). The effect of cement composition and
pH of environment on sulfate resistance of Portland cements and blended cements. Cement and
Concrete Composites, 19, 161-171.
Castellanos, J. F. (2012). Mechanically stabilized earth wall inspector’s handbook (revised).

Tallahassee, Florida: Florida Department of Transportation. Retrieved October 11, 2013, from
http://www.dot.state.fl.us/construction/CONSTADM/guidelist/InspectGuidelist/FY1213/MSE-Wall-
Inspectors-Handbook.pdf
Collepardi, M. (2003). A state-of-the-art review on delayed ettringite attack on concrete. Cement

and Concrete Composites, 25, 401-407.
Corrpro Companies Europe Limited (Corrpro). (2013). Nilsson model 400 soil resistance meter.
Stockton on Tees, Great Britain. Retrieved October 11, 2013, from
http://www.corrpro.co.uk/pdf/Nilsson_model_400_soil_resistance_meter.pdf
Davison, W., & Woof, C. (1985). Performance tests for the measurement of pH with glass
electrodes in low ionic strength solutions including natural waters. Analytical Chemistry, 57,
2567-2570.
Djuric, M., Ranogajec, J., Omorjan, R., & Miletic, S. (1996). Sulfate corrosion of Portland
cement—pure and blended with 30% fly ash. Cement and Concrete Research, 26, 1295-1300.
Domagalski, J. L., & Johnson, H. M. (2011). Subsurface transport of orthophosphate in five

agricultural watersheds. Journal of Hydrology, 409, 157-171.
Duffó, G. S., Reinoso, M., Ramos, C. P., & Farina, S. B. (2012). Characterization of steel rebar
embedded in a 70-year old concrete structure. Cement and Concrete Research, 42, 111-117.
Elias, V., Christopher, B. R., & Berg, R. R. (2001). Mechanically stabilized earth walls and
reinforced soil slopes design and construction guidelines, FHWA-NHI-00-043. Washington, D.
C.: National Highway Institute, Federal Highway Administration, U.S. Department of
Transportation.
Elias, V., Fishman, K. L., Christopher, B. R., & Berg, R. R. (2009). Corrosion/degradation of
soil reinforcements for mechanically stabilized earth walls and reinforced soil slopes, FHWA-
NHI-09-87. Washington, D. C.: National Highway Institute, Federal Highway Administration,
U.S. Department of Transportation.
Ellison, S. L. R., Barwick, V. J., & Farrant, T. J. D. (2009). Practical statistics for the analytical
scientist: A bench guide (2nd ed.). Cambridge, United Kingdom: Royal Society of Chemistry,
Thomas Graham House.
181
Elrashidi, M. A., Alva, A. K., Huang, Y. F., Calvert, D. V., Obreza, T. A., & He, Z. I. (2001).
Accumulation and downward transport of phosphorus in Florida soils and relationship to water
quality. Communications in Soil Science and Plant Analysis, 32, 3099-3119.
Fan, Y. F., Hu, Z. Q., Zhang, Y. Z., & Liu, J. L. (2010). Deterioration of compressive property of
concrete under simulated acid rain environment. Construction and Building Materials, 24, 1975-
1983.
Florida Department of Environmental Protection (FDEP). (2009). Quality manual for State of
Florida, Department of Environmental Protection, Chemistry Laboratory, NELAC Certification
#31640 and E31780. Tallahassee, Florida.
Florida Department of Transportation (FDOT). (2011). Permanent MSE retaining wall systems,
2010 interim design standard (index no. 6020). Tallahassee, Florida. Retrieved October 15,
2013, from http://www.dot.state.fl.us/rddesign/DS/10/Int/6020.pdf
Florida Department of Transportation (FDOT). (2013). Florida structures manual, vol. 1.

Tallahassee, Florida. Retrieved May 11, 2014, from
http://www.dot.state.fl.us/structures/StructuresManual/CurrentRelease/StructuresManual.shtm.
Florida Department of Transportation (FDOT). (2000). Florida method of test for sampling
aggregates, FM 1 T-002. Tallahassee, Florida.
Florida Department of Transportation (FDOT). (2011). Florida method of test for pH of soil and
water, FM 5-550. Tallahassee, Florida.
Florida Department of Transportation (FDOT). (2011). Florida method of test for resistivity of
soil and water, FM 5-551. Tallahassee, Florida.
Florida Department of Transportation (FDOT). (2011). Florida method of test for chloride of soil
and water, FM 5-552. Tallahassee, Florida.
Florida Department of Transportation (FDOT). (2011). Florida method of test for sulfate of soil
and water, FM 5-553. Tallahassee, Florida.
Fishman, K. L., & Withiam, J. L. (2011). LRFD metal loss and service-life strength reduction
factors for metal-reinforced systems, NCHRP Report 675. Washington, D. C.: Transportation
Research Board.
Gladstone, R. A., Anderson, P. L., Fishman, K. L., & Withiam, J. L. (2006). Durability of
galvanized soil reinforcement: More than 30 years of experience with mechanically stabilized
earth (no. 1975). In Transportation Research Record: Journal of the Transportation Research
Board. Washington, D. C.: Transportation Research Board of the National Academies.
182
Glasser, F. P., Marchand, J., & Samson, E. (2008). Durability of concrete—Degradation
phenomena involving detrimental chemical reactions. Cement and Concrete Research, 38, 226-
246.
Grimmond, C. S. B., King, T. S., Cropley, F. D., Nowak, D. J., & Souch, C. (2002). Local-scale
fluxes of carbon dioxide in urban environments: methodological challenges and results from
Chicago. Environmental Pollution, 116, S243-S254.
Gruyaert, E., Van de Heede, P., Maes, M., Belie, N. D. (2012). Investigation of the influence of
blast-furnace slag on the resistance of concrete against organic acid or sulphate attack by means
of accelerated degradation tests. Cement and Concrete Research, 42, 173-185.
Hach Company. (2003). Hach water analysis handbook: Method 8207, digital titrator.
[Brochure]. Loveland, Colorado.
Hach Company (2012). Method 8051, sulfate (DOC316.53.01135). [Brochure]. Loveland,

Colorado.
Hamilton, R., & Crabbe, H. (2009). Chapter 1 environment, pollution and effects. In J. Watt, J.
Tidblad, V. Kucera, & R. Hamilton (Eds.), The Effects of Air Pollution on Cultural Heritage (pp.
1-27). New York, New York: Springer Science + Business Media.
Hanlon, E. A. (2009). Soil pH and electrical conductivity: A county extension soil laboratory
manual (CIR1081). [Brochure]. Gainesville, Florida: University of Florida, Soil and Water
Science Department, Florida Cooperative Extension Service, Institute of Food and Agricultural
Sciences Extension. Retrieved October 26, 2013, from http://edis.ufl.edu
Hayes, C. (2007). Test methods for water-soluble sulfate in soils (PCA R&D Serial No. 3016).
[Brochure]. Skokie, Illinois: Portland Cement Association.
Herrero, J., Artieda, O., & Hudnall, W. H. (2009). Gypsum, a tricky material. Soil Science
Society of America Journal, 73, 1757-1763.
Hormann, V., & Kirchner, G. (2002). Prediction of the effects of soil-based countermeasures on
soil solution chemistry of soils contaminated with radiocesium using the hydrogeochemical code
PHREEQC. The Science of the Total Environment, 289, 83-95.
Idso, S. B., Idso, C. D., & Balling, R. C., Jr. (2002). Seasonal and diurnal variations of near-
surface atmospheric CO2 concentration within a residential sector of the urban CO2 dome of
Phoenix, AZ, USA. Atmospheric Environment, 36, 1655-1660.
Illingworth, J. A. (1981). A common source of error in pH measurements. Biochemistry Journal,

195, 259-262.
183
Jack, T. R., & Wilmott, M. J. (2011). Corrosion by soils. In R. W. Revie (Ed.), Uhlig’s
Corrosion Handbook (3rd ed, pp. 333-349). Hoboken, New Jersey: John Wiley & Sons.
Jacques, D., Wang, L., Martens, E., & Mallans, D. (2010). Modelling chemical degradation of
concrete leaching with rain and soil water types. Cement and Concrete Research, 40, 1306-1313.
Kadis, R., & Leito, I. (2010). Evaluation of the residual liquid junction potential contribution to
the uncertainty in pH measurement: a case study on low ionic strength natural waters. Analytica
Chimica Acta, 664, 129-135.
Kanazu, T., Matsumura, T., & Nishiuchi, T. (2001). Effect of simulated acid rain on
deterioration of concrete. Water, Air, and Soil Pollution, 130, 1481-1486.
Keller, W. D., & Matlack, K. (1990). The pH of clay suspensions in the field and laboratory, and
methods of measurement of their pH. Applied Clay Science, 5, 123-133.
Kissel, D., Sonon, L., Vendrell, P. F., & Isaac, R. A. (2009). Salt concentration and measurement
of soil pH. Communications in Science and Plant Analysis, 40, 179-187.
Kohlmann, F. (2003). What is pH, and how is it measured? A technical handbook for industry.
[Brochure]. Loveland, Colorado: Hach Company.
Lide, D. R. (Ed.) (2001). CRC handbook of chemistry and physics. (82nd ed.) Boca Raton,
Florida: CRC Press.
Lipfert, F. W., & Daum, M. L. (1992). The distribution of common construction materials at risk
to acid deposition in the United States. Atmospheric Environment, 26B, 217-226.
Macpherson, G. L. (2009). CO2 distribution in groundwater and the impact of groundwater

extraction on the global C cycle. Chemical Geology, 264, 328-336.
Marinoni, N., Birelli, M. P., Rostagno, C., & Pavese, A. (2003). The effects of atmospheric
multipollutants on modern concrete. Atmospheric Environment, 37, 4701-4712.
Marques, P. F., & Costa, A. (2010). Service life of RC structures: Carbonation induced
corrosion. Prescriptive vs performance-based methodologies. Construction and Building
Materials, 24, 258-265.
Mason, B. J. (1992). Preparation of soil sampling protocols: Sampling techniques and

strategies, EPA/600/R-92/128. Washington, D. C.: Office of Research and Development, U. S.
Environmental Protection Agency.
Miller Company (MCM). (2011). Miller 400A analog resistance meter part #44500 user’s
manual. [Brochure]. Sebastian, Florida.
184
Miller, R. O., & Kissel, D. E. (2010). Comparison of soil pH methods on soils of North America.
Soil Science Society of America Journal, 74, 310-316.
Mindess, S., Young, J. F., & Darwin, D. (2003). Concrete (2nd ed.). Upper Saddle River, New
Jersey: Prentice-Hall, Pearson Education, Inc.
Minasny, B., McBratney, A. B., Brough, D. M., & Jacquier, D. (2011). Models relating soil pH
measurements in water and calcium chloride that incorporate electrolyte concentration.
European Journal of Soil Science, 62, 728-732.
Montgomery, D. C. (2005). Design and analysis of experiments (6th ed.). New York, New York:
John Wiley & Sons.
National Atmospheric Deposition Program (NADP). (2013). NADP Maps and Data. Retrieved
October 23, 2013, from http://nadp.sws.uiuc.edu/data/
National Atmospheric Deposition Program (NADP). (2015). NADP Maps and Data. Retrieved
February 12, 2015 from http://nadp.sws.uiuc.edu/data/
State of Nevada, Department of Transportation, Materials Division (NDOT), (1999). Test

method Nevada T235B. Retrieved October 15, 2013, from
www.nevadadot.com/uploadedFiles/NDOT/About_NDOT/NDOT_Divisions/Operations/Materials/MTM
_T235B.pdf
Neville, A. M. (1996). Properties of Concrete (4th & final ed.). New York, New York, John
Wiley & Sons.
Okochi, H., Kameda, H., Hasegawa, S., Saito, N., Kubota, K., & Igawa, M. (2000). Deterioration
of concrete structures by acid deposition—an assessment of the role of rainwater on deterioration
by laboratory and field exposure experiments using mortar specimens. Atmospheric
Environment, 34, 2937-2945.
Oman, S. (2000). A step to uniform definition and interpretation of the suspension effect.
Talanta, 51, 21-31.
Ott, L. (1993). An introduction to statistical methods and data analysis (4th ed.). Belmont,
California: Duxbury Press.
Ozga, I., Bonazza, A., Bernardi, E., Tittarelli, F., Favoni, O., Ghedini, N., Morselli, L., &
Sabbioni, C. (2011). Diagnosis of surface damage induced by air pollution of 20th-century
concrete buildings. Atmospheric Environment, 45, 4986-4995.
185
Parkhurst, D. L., & Appelo, C. A. J. (2013). Description of input and examples for PHREEQC
version 3—A computer program for speciation, batch-reaction, one-dimensional transport, and
inverse geochemical calculations. U.S. Geological Survey Techniques and Methods, Book 6,
Chapter A43. Retrieved October 25, 2013, from http://pubs.usgs.gov/tm/06/a43/
Peech, M., & Bradfield, R. (1948). Chemical methods for estimating lime needs of soil. Soil
Science, 65, 35-56.
Puri, A. N., & Asghar, A. G. (1938). Influence of salts and soil-water ratio on pH value of soils.
Soil Science, 46, 249-257.
Rice, A., & Bostrom, G. (2011). Measurements of carbon dioxide in an Oregon metropolitan
region. Atmospheric Environment, 45, 1138-1144.
Rice, E. W., Baird, R. B., Eaton, A. D., & Clesceri, L. S., (Eds.). (2012). Standard Methods for
the Examination of Water and Wastewater (22nd ed.). Washington, D. C.: American Public
Health Association (APHA) Publication Office.
Roots, O. (2008). Materials corrosion and air pollution. Long-term studies at the Lahemaa
monitoring station, Estonia. Proceedings of the Estonian Academy of Sciences, 57, 107-116.
Rozell, D. (2010). Modeling the removal of arsenic by iron oxide coated sand. Journal of
Environmental Engineering, 136, 246-248.
Rozière, E., Loukili, A., & Cussigh, F. (2009). A performance based approach for durability of
concrete exposed to carbonation. Construction and Building Materials, 23, 190-199.
Sabbioni, C., Zappia, G., Riontino, C., Blanco-Varela, M. T., Aguilera, J., Puertas, G., Van
Balen, K., & Toumbakari, E. E. (2001). Atmospheric deterioration of ancient and modern
hydraulic mortars. Atmospheric Environment, 35, 539-548.
Sagüés, A., Poor, N., Cáseres, L., Akhoondan, M., 2009. Development of a rational method for
predicting corrosion rates of metals in soils and waters, report no. BD497. Tallahassee, Florida:
Florida Department of Transportation.
Samouëlian, A., Cousin, I., Tabbagh, A., Bruand, A., & Richard, G. (2005). Electrical resistivity
survey in soil science: a review. Soil and Tillage Research, 83, 173-193.
Samson, E., & Marchand, J. (2007). Modeling the transport of ions in unsaturated cement-based
materials. Computers and Structures, 85, 1740-1756.
Scheer, D. (2013). Standard specifications for road and bridge construction 2014, Tallahassee,
Florida: Florida Department of Transportation. Retrieved October 15, 2013, from
http://www.dot.state.fl.us/specificationsoffice/Implemented/SpecBooks/
186
Schiffer, D. (1989). Effects of three highway-runoff detention methods on water quality of the
surficial aquifer system in central Florida. U. S. Geological Survey, water resources
investigations report 88-4170. Tallahassee, Florida: Florida Department of Transportation.
Retrieved January 20, 2015, from fl.water.usgs.gov
Schiopu, N., Tiruta-Barna, L., Jayr, E., Méhu, J., & Moszkowicz, P. (2009). Modelling and
simulation of concrete leaching under outdoor exposure conditions. Science of the Total
Environment, 407, 1613-1630.
Sersale, R., Frigione, G., & Bonavita, L. (1998). Acid depositions and concrete attack: Main
influences. Cement and Concrete Research, 28, 19-24.
Skoog, D. A., West, D. M., & Holler, F. J. (1996). Fundamentals of analytical chemistry (7th
ed.). Fort Worth, Texas: Saunders College Publishing, Harcourt Brace College Publishers.
Snoeyink, V. L., & Jenkins, D. (1980). Water chemistry. New York, New York: John Wiley &
Sons.
Stark, D. (2002). Performance of concrete in sulfate environments, research and development

bulletin RD129. Skokie, Illinois: Portland Cement Association.
Steffens, A., Dinkler, D., & Ahrens, H. (2002). Modeling carbonation for corrosion risk
prediction of concrete structures. Cement and Concrete Research, 32, 935-941.
Strayer, H., Smith, R., Mizak, C., & Poor, N. (2007). Influence of air mass origin on the wet
deposition of nitrogen to Tampa Bay, Florida—An eight-year study. Atmospheric Environment,
41, 4310-4322.
Thermo Fisher Scientific (Thermo). (2012). Fisher Scientific Accumet basic (AB) benchtop
meters, instruction manual (68X613601 Rev 0). Retrieved February 2, 2015, from
www.fishersci.com/accumet
Thermo-Scientific (2007). Technical bulletin 501 – pH measurements in low ionic strength

solutions. [Brochure]. Beverly, Massachusetts: Water Analysis Instruments, Environmental
Instruments, Thermo-Scientific.
Thornley, J. D., Siddharthan, R. V., Luke, B., & Salazar, J. M. (2010). Investigation and
implications of mechanically stabilized earth wall corrosion in Nevada (no. 2186). In
Transportation Research Record: Journal of the Transportation Research Board (pp.154-160).
Washington, D. C. Transportation Research Board of the National Academies.
Tidblad, J., Kucera, V., & Mikhailov, A. A. (1998). Statistical analysis of 8 year materials
exposure and acceptable deterioration and pollution levels, report no. 30. In UN/ECE
187
International Co-operative Programme on Effects of Materials, Including Historical and
Cultural Monuments. Stockholm, Sweden: Swedish Corrosion Institute.
Tinker & Rasor Corrosion Mitigation Instrumentation (Tinker & Rasor). (2013). Product
instructions model SR-2 soil resistivity meter, 101 043. [Brochure]. San Bernardino, California.
Retrieved October 26, 2013, from http://tinker-rasor.com/instructions/
United States Environmental Protection Agency (USEPA). (2013). Air trends, sulfur dioxide.
Retrieved October 28, 2013, from http://www.epa.gov/airtrends/sulfur.html#soloc
United States Environmental Protection Agency (USEPA). (1986). Sulfate (turbidimetric)

method 9038, rev 0. Retrieved October 25, 2013, from
http://www.epa.gov/osw/hazard/testmethods/sw846/pdfs/9038.pdf
United States Environmental Protection Agency (USEPA). (2004). Soil and waste pH, method
9045D, rev. 4. Retrieved October 25, 2013, from
http://www.epa.gov/osw/hazard/testmethods/sw846/pdfs/9045d.pdf
United States Environmental Protection Agency (USEPA). (1994). Chloride (titrimetric, silver
nitrate), method 9253, Rev 0. Retrieved October 25, 2013 from
http://www.epa.gov/osw/hazard/testmethods/sw846/pdfs/9253.pdf
United States Geological Survey (USGS). (2002). Sand and gravel operations in the United
States. National atlas. Retrieved November 8, 2013, from http://nationalatlas.gov/mld/sndgrvx.html
White, I., & Zegelin, S. J., 1995. Electric and dielectric methods for monitoring soil-water
content. In L. G. Wilson, L. G. Everett, & S. J. Cullen, (Eds.), Handbook of vadose zone
characterization and Monitoring (pp. 343-386). Boca Raton, Florida, CRC Press.
Wolt, J. D. (1994). Soil solution chemistry: applications to environmental science and

agriculture. New York, New York: John Wiley & Sons.
Xie, S., Qi, L., & Zhou, D. (2004). Investigation of the effects of acid rain on the deterioration of
cement concrete using accelerated tests established in laboratory. Atmospheric Environment, 38,
4457-4466.
Yang, S. X., Cheng, K. L., Kurtz, L. T., & Peck, T. R. (1989). Suspension effect in
potentiometry. Particulate Science and Technology: An International Journal, 7, 139-152.
Yoon, I.-S., Çopuroğlu, O., & Park, K.-B. (2007). Effect of global climatic change on
carbonation progress of concrete. Atmospheric Environment, 41, 7274-7285.
188
Youden, W. J. (1975). Statistical techniques for collaborative tests. In Statistical manual of the
association of official analytical chemists (pp. 1-63). Gaithersburg, Maryland: AOAC
International.
Yuan, T. L. (1963). Some relationships among hydrogen, aluminum, and pH in soil solution and
soil systems. Soil Science, 95, 155-163.
Zivica, V., & Bajza, A. (2001). Acidic attack of cement-based materials—a review. Part 1.
Principal of acidic attack. Construction and Building Materials, 15, 331-340.
Zivica, V., & Bajza, A. (2002). Acidic attack of cement-based materials—a review. Part 2.
Factors of acidic attack and protective measures. Construction and Building Materials, 16, 215-
222.
189
Appendix A Ruggedness Study Results
Statistically-significant results are highlighted.
A.1 Mine 1 Wimauma Sand
Table A-1 Wimauma Sand Ruggedness Study Results for pH

Factors pH Determination
A a CT CT CT CT WT WT WT WT
B b CS CS WS WS CS CS WS WS
C c wet dry wet dry wet dry wet dry
D d ML ML G G G G ML ML
E e 1 2 1 2 2 1 2 1
F f 30 0 0 30 30 0 0 30
G g V G G V G V V G
pH Results 4.55 4.92 4.49 4.69 4.68 4.41 4.90 4.52
A Transport in cooler (~15C) (CT)
a Transport under ambient conditions (~30C) (WT)
B Store in refrigerator (~ 4C) (CS)
b Store at room temperature (~23C) (WS)
C Test "as is" (wet)
c Air-dried, sieved (dry)
D 100 mL of sample (ML)
d 100 g of sample (G)
E pH electrode 1 (ATC, refillable) (1)
e pH electrode 2 (No ATC, not refillable) (2)
F 30 min equilibration, stirred 3 x (30)
f Stirred once, no equilibration (0)
G Stir vigorously (V)
g Stir gently so as not to entrain air (G)
Note 1 Sampling date was 5/8/2014 and ruggedness study date was 5/20/2014
Note 2 Electrode 1: 13-62-AP55 (Accumet), 98.1% and 4.7 mV; 98.6% and 6.5 mV
Note 3 Electrode 2: WD-35801-00 (Oakton), 100.3%, 13.0 mV
190
Table A-2 Wimauma Sand Significance of Ruggedness Study for pH
Factors Avg St Dev Diff F-tests t-tests pH Was:
A 4.66 0.19
Higher if transported cool; not significant
a 4.63 0.21 0.04 0.863 0.815
B 4.64 0.22
Higher if stored warm; not significant
b 4.65 0.19 -0.01 0.824 0.947
C 4.66 0.18
Higher if tested “as is”; not significant
c 4.64 0.22 0.02 0.748 0.894
D 4.72 0.22
Higher for a 100-mL sample; not significant
d 4.57 0.14 0.16 0.488 0.275
E 4.49 0.06
Higher for electrode 2; significant
e 4.80 0.13 -0.31 0.237 0.005
F 4.61 0.09
Higher if not equilibrated; not significant
f 4.68 0.27 -0.07 0.099 0.637
G 4.64 0.21
Higher if stirred gently; not significant
g 4.65 0.20 -0.02 0.924 0.920
191
Table A-3 Wimauma Sand Ruggedness Study Results for Minimum Resistivity
Factors Minimum Resistivity Determination
D d 100 100 50 50 50 50 100 100
F f 1 0.5 0.5 1 1 0.5 0.5 1
G g no yes yes no yes no no yes
Resistivity Results (ohm-cm) 47,500 54,000 43,000 63,000 41,000 56,500 59,000 55,000
D DI water increments of 100 mL (100)
d DI water increments of 50 mL (50)
F 1.0 kg soil (1.0)
f 0.5 kg soil (0.5)
G Electrode not covered (no)
g Electrode covered (yes)
Note 2 DI water resistivity: 690,000 small soil box; 565,000 large soil box
Note 3 4000 ohm-cm conductivity standard: 4,250 ohm-cm at 24.0C
192
Table A-4 Wimauma Sand Significance of Ruggedness Study for Minimum Resistivity
Factors Avg St Dev Diff F-tests t-tests Minimum Resistivity Was:
A 51,900 8,680 Higher for transport under ambient
a 52,900 8,090 -1,000 0.910 0.872 conditions; not significant
B 49,800 6,960 Higher for storage at room
b 55,000 8,640 -5,250 0.731 0.380 temperature; not significant
C 47,600 8,060 Higher for testing a dried sample; not
c 57,100 4,050 -9,500 0.288 0.080 significant
D 53,900 4,770 Higher for water increments of 100
d 50,900 10,600 3,000 0.221 0.625 mL; not significant
E 50,500 6,360 Higher for large soil box; not
e 54,300 9,570 -3,750 0.520 0.538 significant
F 51,600 9,500
Higher for 0.5 kg soil; not significant
f 53,100 7,050 -1,500 0.637 0.808
G 56,500 6,570 Higher if electrode is not covered; not
g 48,300 7,270 8,250 0.871 0.143 significant
193
Table A-5 Wimauma Sand Ruggedness Study Results for Sulfate
Factors Sulfate Determination
C c 110 60 110 60 110 60 110 60
D d 1 1 2 2 2 2 1 1
E e 1 2 1 2 2 1 2 1
F f RB US US RB RB US US RB
Sulfate Results (ppm) 21 15 39 15 48 24 9 21
C Oven dry at 110C, sieve (110)
c Oven dry at 60C, sieve (60)
D Range 1 factory default (1)
d Range 2 factory default (2)
E Extraction method 1: 3 to 1 extraction; membrane filtration (1)
e Extraction method 2: 1 to 1 extraction; 3:1 dilution; membrane filtration (2)
F Reagent blank (RB)
f Unreacted sample blank (US)
G No sample spike (no)
g Spiked sample to 20 ppm (yes)
Note 2 20 ppm check standard: 19 ppm
194
Table A-6 Wimauma Sand Significance of Ruggedness Study for Sulfate
Factors Avg St Dev Diff F-tests t-tests Sulfate Was:
A 23 11 Higher if transported under ambient
a 26 16 -3 0.565 0.773 conditions; not significant
B 27 14
Higher if stored cold; not significant
b 21 13 6 0.859 0.560
C 29 18 Lower if dried at 60C; significantly
c 19 5 10 0.051 0.291 lower standard deviation
D 17 6 Higher if tested on Range 2; not
d 32 15 -15 0.154 0.108 significant
E 26 9 Higher if extracted with 3:1 dilution;
e 22 18 4 0.267 0.664 not significant
F 26 15 Higher if tested with a reagent blank;
f 22 13 4 0.843 0.664 not significant
G 17 7 Higher if sample spiked with 20 ppm
g 31 15 -14 0.202 0.158 sulfate; not significant
195
A.2 Mine 2 Jahna Sand
Table A-7 Jahna Sand Ruggedness Study Results for pH

E e yes no yes no no yes no yes
F f no yes yes no no yes yes no
G g V G G V G V V G
pH Results 4.79 4.66 4.71 4.70 4.98 4.60 4.59 4.69
B Store in refrigerator (~4C) (CS)
D 100 mL (ML)
d 100 g (G)
E pH measurement with temperature sensed (23C, 98.9% slope, 5.3 mV) (yes)
e pH measurement without temperature sensed (25C, 99.3% slope, 5.9 mV) (no)
F No salt amendment (no)
f Amended with 0.1 g KCl (yes)
G Stir vigorously (V)
g Stir gently so as not to entrain air (G)
Note 2 Electrode/meter AP85/13-620-AP55
196
Table A-8 Jahna Sand Significance of Ruggedness Study for pH
A 4.72 0.05 The same if transported cool or warm; not
a 4.72 0.18 0.00 0.078 1.000 significant
B 4.76 0.17
b 4.67 0.06 0.09 0.102 0.374
C 4.77 0.16
Higher if tested “as is”; not significant
c 4.66 0.05 0.11 0.062 0.263
D 4.68 0.08
Higher for a 100 g sample; not significant
d 4.75 0.16 -0.07 0.298 0.503
E 4.70 0.08 Higher if temperature is not sensed; not
e 4.73 0.17 -0.04 0.230 0.723 significant
F 4.79 0.13 Lower if amended with 0.1 g KCl; not
f 4.64 0.06 0.15 0.184 0.085 significant
G 4.67 0.09 Higher if stirred gently so as not to entrain
g 4.76 0.15 -0.09 0.477 0.345 air; not significant
197
Table A-9 Jahna Sand Ruggedness Study Results for Minimum Resistivity
D d 100 100 50 50 50 50 100 100
E e no yes no yes yes no yes no
F f 1.0 0.5 0.5 1.0 1.0 0.5 0.5 1.0
G g yes no no yes no yes yes no
Resistivity Results
125,000 185,000 150,000 140,000 160,000 130,000 150,000 135,000
(ohm-cm)
D DI water increments of 100 mL (100)
d DI water increments of 50 mL (50)
E C1 & P1 and C2 & P2 terminals not shorted (no)
e C1 & P1 and C2 & P2 terminals shorted (yes)
F 1.0 kg soil (1.0)
f 0.5 kg soil (0.5)
G Soil above electrode (yes)
g Soil not above electrode (no)
Note 2 Soil slurry temperatures were ~22C (21.7 to 22.4)
Note 3 No calibration data recorded.
198
Table A-10 Jahna Sand Significance of Ruggedness Study for Minimum Resistivity
A 150,000 25,500 Higher for cool transport; not
a 144,000 13,800 6,250 0.339 0.681 significant
B 150,000 28,000 Higher for cold storage; not
b 144,000 7,500 6,250 0.058 0.681 significant
C 146,000 14,900
Higher for dried soil; not significant
c 148,000 25,300 -1,250 0.408 0.935
D 149,000 26,300 Higher for water increments of 100
d 145,000 12,900 3,750 0.274 0.806 mL; not significant
E 135,000 10,800
Higher with terminals shorted; not
- significant
e 159,000 19,300 0.365 0.075
23,800
F 140,000 14,700
Higher with 0.5 kg soil; not
- significant
f 154,000 22,900 0.488 0.351
13,800
G 136,000 11,100
Higher with soil below electrode; not
- significant
g 158,000 21,000 0.321 0.124
21,300
199
A.3 Mine 3 Youngquist Sand
Table A-11 Youngquist Sand Ruggedness Study Results for pH

E e 1 2 1 2 2 1 2 1
F f 30 60 60 30 30 60 60 30
pH Results 8.12 8.16 8.16 8.19 8.20 8.03 8.07 8.02
D 100 mL + 100 mL water (ML)
d 30 g + 30 g water (G)
E Electrode/meter 1 (1)
e Electrode/meter 2 (2)
F 30 min equilibration with stirring every 10 min (30)
f 60 min equilibration with stirring every 10 min (60)
G No stir right before testing (no)
g Stir immediately before testing (yes)
Note 2 Meter/electrode 1: AP85/13-62-AP55; slope = 99.1%, offset = 3.0 mV
Note 3 Meter/electrode 2: AB150/13-620-631; slope = 98.9%, offset = 5.8 mV
200
Table A-12 Youngquist Sand Significance of Ruggedness Study for pH
A 8.16 0.03 Higher for cooler transport; not
a 8.08 0.08 0.08 0.115 0.128 significant
B 8.13 0.07
Higher for cold storage; not significant
b 8.11 0.08 0.02 0.899 0.755
C 8.14 0.06
Higher for “as is”; not significant
c 8.10 0.09 0.04 0.475 0.497
D 8.09 0.06
Higher for a 30-g sample; not significant
d 8.15 0.08 -0.05 0.684 0.331
E 8.08 0.07
Higher for electrode 2; not significant
e 8.16 0.06 -0.07 0.815 0.160
F 8.13 0.08 Higher for 30 min equilibration; not
f 8.11 0.07 0.03 0.708 0.622 significant
G 8.10 0.07 Higher for stir immediately before
g 8.14 0.08 -0.03 0.830 0.558 testing; not significant
201
Table A-13 Youngquist Sand Ruggedness Study Results for Minimum Resistivity
D d SS SS WS WS WS WS SS SS
F f 1.0 1.5 1.5 1.0 1.0 1.5 1.5 1.0
A Transport in cooler (~18C)
a Transport under ambient conditions (~30C)
B Store in refrigerator (~ 4C)
b Store at room temperature (~23C)
C Test “as is”
c Air-dried, sieved, 10% water added and allowed to equilibrate overnight
D Mostly soil (when sample is saturated) (SS)
d Mostly water (when sample is saturated) (WS)
F 1.0 kg soil (1.0)
f 1.5 kg soil (1.5)
G Rinse box with DI water between measurements (yes)
g Do not rinse box with DI water between measurements (no)
Note 2 Resistivity of DI (dilution) water was 810,000 ohm-cm
Note 3 Resistivity of a 4,000 ohm-cm conductivity standard was 4,000 ohm-cm
202
Table A-14 Youngquist Sand Significance of Ruggedness Study for Minimum
Resistivity
A 7,050 2,260 Resistivity was higher for cool
a 6,000 2,010 1,050 0.854 0.514 transport; not significant
B 6,600 2,690 Resistivity was higher for cold storage;
b 6,450 1,620 150 0.425 0.927 not significant
C 6,480 2,280 Resistivity was higher for dried soil;
c 6,580 2,160 -100 0.933 0.951 not significant
D 8,280 790 Resistivity was higher for soil slurry;
d 4,780 1,050 3,500 0.653 0.002 significant
E 6,300 2,720 Resistivity was higher for large soil
e 6,750 1,540 -450 0.374 0.783 box; not significant
F 7,080 1,900 Resistivity was higher for 1.0 kg of
f 5,980 2,340 1,100 0.744 0.493 soil; not significant
G 6,630 2,330 Resistivity was higher if box rinsed
g 6,430 2,110 200 0.873 0.903 between tests; not significant
203
Table A-15 Youngquist Sand Ruggedness Study Results for Sulfate
A a CC CC CC CC WW WW WW WW
B b 1 1 2 2 1 1 2 2
C c 110 60 110 60 110 60 110 60
D d 1 1 2 2 2 2 1 1
E e 1 2 1 2 2 1 2 1
F f 10 18 18 10 10 18 18 10
A Transport in cooler (~15C), cold storage (~4C) (CC)
a Warm transport (~30C), room temperature storage (WW)
B 100 mg soil/300 mL water (3:1 dilution) (1)
b 100 mg soil/100 mL water (1:1 dilution); diluted 3:1 post-filtration (2)
D Filtration through Q8, acid added, then 0.45 micron membrane (1)
d Centrifuge, acid added, then 0.45 micron membrane (2)
E Range 1 Factory Default (1)
e Range 2 Factory Default (2)
F Sieved through No. 10 (2 mm) (10)
f Sieved through No. 18 (1 mm) (18)
Note 2 A 20-ppm check standard was 18 ppm on Range 1 and 22 ppm on Range 2
Note 3 A 20-ppm check standard was 16 ppm on Range 1 and 20 ppm on Range 2
204
Table A-16 Youngquist Sand Significance of Ruggedness Study for Sulfate
A 43 9 Sulfate was higher for warm transport
a 47 18 -4 0.245 0.673 and storage; not significant
B 50 14 Sulfate was higher for extraction with
b 41 13 9 0.938 0.387 3:1 dilution; not significant
C 52 14 Sulfate was higher for soil dried at
c 38 11 14 0.702 0.173 110 C; not significant
D 47 13 Sulfate was higher for filtration; not
d 43 16 4 0.779 0.673 significant
E 36 6 Sulfate was higher on Range 2:
e 54 13 -18 0.285 0.048 significant
F 44 16 Sulfate was higher if sieved through
f 46 13 -2 0.785 0.889 No. 18 sieve; not significant
G 44 12 Sulfate was higher if spiked with 20
g 47 17 -3 0.624 0.780 ppm standard; not significant
205
A.4 Mine 4 Calhoun Sand
Table A-17 Calhoun Sand Ruggedness Study Results for pH

E e no yes no yes yes no yes no
F f 30 60 60 30 30 60 60 30
pH Results 4.79 4.19 4.82 4.23 4.26 4.66 4.20 4.65
D 100 mL sample size (ML)
d 100 g sample size (G)
E No salt amendment (no)
e Amendment with 0.1 g KCl (yes)
f 1 hr equilibration with stirring every 10 min (60)
Note 2 Electrode/meter AB150/13-620-631; slope = 100.7% and offset = 4 mV
206
Table A-18 Calhoun Sand Significance of Ruggedness Study Factors for pH
A 4.51 0.34 Higher if transported cool; not
a 4.44 0.25 0.07 0.598 0.769 significant
B 4.48 0.30 The same if stored cold or at room
b 4.48 0.31 0.00 0.943 1.000 temperature; not significant
C 4.52 0.33 Higher if tested “as is”; not
c 4.43 0.26 0.09 0.683 0.700 significant
D 4.46 0.31 Higher for a 100 g sample; not
d 4.49 0.29 -0.04 0.936 0.875 significant
E 4.73 0.09 Lower if amended with 0.1 g KCl;
e 4.22 0.03 0.51 0.128 0.000 significant
F 4.48 0.28 Higher if equilibrated for 30 min;
f 4.47 0.32 0.02 0.828 0.946 not significant
G 4.47 0.30 Higher if stirred immediately
g 4.48 0.30 -0.01 0.981 0.964 before testing; not significant
207
Table A-19 Calhoun Sand Ruggedness Study Results for Minimum Resistivity
D d SS SS WS WS WS WS SS SS
F f 1.0 1.5 1.5 1.0 1.0 1.5 1.5 1.0
c Air-dried, sieved, 10% water added and allowed to equilibrate overnight (dry)
D Mostly soil (when sample is saturated) (SS)
d Mostly water (when sample is saturated) (WS)
F 1.0 kg soil (1.0)
f 1.5 kg soil (1.5)
G Rinse box with DI water between measurements (yes)
g Do not rinse box with DI water between measurements (no)
Note 1 Sampling date was 7/14/2014; ruggedness study date was 7/23/2014
Note 2 DI water resistivity was 780,000 ohm-cm at ~24 C
Note 3 4,000 ohm-cm conductivity standard was 4,000 ohm-cm at ~24C
208
Table A-20 Calhoun Sand Significance of Ruggedness Study for Minimum Resistivity
Factors Average St Dev Diff F-tests t-tests Minimum Resistivity Was:
A 71,400 37,200 Higher if transported under
a 73,300 38,300 -1,880 0.962 0.946 ambient conditions; not significant
B 71,900 37,200 Higher if stored at room
b 72,800 38,400 -875 0.957 0.975 temperature; not significant
C 76,300 38,000 Higher if tested “as is”; not
c 68,400 37,000 7,880 0.967 0.777 significant
D 104,600 5,960 Higher if tested as a slurry of
d 40,000 6,160 64,600 0.958 0.000 mostly soil; significant
E 71,500 41,100 Higher if tested in the large soil
e 73,100 34,100 -1,630 0.766 0.953 box; not significant
F 75,500 36,600 Higher if tested with 1 kg of soil;
f 69,100 38,610 6,380 0.932 0.819 not significant
G 72,800 42,000 Higher if soil box rinsed between
g 71,900 33,100 875 0.707 0.975 tests; not significant
209
A.5 Mine 5 Angelo’s Sand
Table A-21 Angelo’s Sand Ruggedness Study Results for pH (Replicate 1)

E e 1 2 1 2 2 1 2 1
F f 30 60 60 30 30 60 60 30
pH Results 4.63 4.53 5.26 4.55 4.53 4.57 4.54 4.53
C Test "as received" (wet)
c Dried at 60C, No. 10 sieve (dry)
D 100 mL (ML)
d 100 g (G)
E Electrode 1, 13-620-631 (99.2% slope, 3.3 mV, 25.0C) (1)
e Electrode 2, 13-620-631 (100% slope, -0.8 mV, 25.4C) (2)
F 30-min equilibrium period with stirring every 10 min (30)
f 60-min equilibrium period with stirring every 10 min (60)
G No stir before testing (no)
Note 1 Sampling date was 9/17/2014 and ruggedness study dates were 9/27/2014
Note 2 pH meter AB150
210
Table A-22 Angelo’s Sand Significance of Ruggedness Study for pH (Replicate 1)
A 4.74 0.35
a 4.54 0.02 0.20 0.001 0.294
B 4.57 0.05
b 4.72 0.36 -0.16 0.007 0.426
C 4.74 0.35 Higher if tested “as received”; not
c 4.55 0.02 0.19 0.001 0.308 significant
D 4.56 0.05
d 4.73 0.36 -0.17 0.008 0.380
E 4.75 0.34
e 4.54 0.01 0.21 0.000 0.268
F 4.56 0.05 Higher for a 60-min equilibrium, not
f 4.73 0.36 -0.17 0.008 0.395 significant
G 4.57 0.04 Higher if stirred immediately before
g 4.71 0.37 -0.14 0.004 0.475 testing; not significant
211
Table A-23 Angelo’s Sand Ruggedness Study Results for pH (Replicate 2)
E e 1 2 1 2 2 1 2 1
F f 30 60 60 30 30 60 60 30
pH Results 4.61 4.55 4.64 5.05 4.53 4.58 4.54 4.58
D 100 mL (ML)
d 100 g (G)
E Electrode 1, 13-620-631 (99.5% slope, 28 mV, 24.8 C) (1)
e Electrode 2, 13-620-631 (100.5% slope, -1.0 mV, 25.0 C) (2)
212
Table A-24 Angelo’s Sand Significance of Ruggedness Study for pH (Replicate
2)
A 4.71 0.23
a 4.56 0.03 0.15 0.005 0.226
B 4.57 0.04
b 4.70 0.24 -0.13 0.011 0.300
C 4.58 0.05 Higher if tested dried and sieved; not
c 4.69 0.24 -0.11 0.034 0.406 significant
D 4.57 0.03
d 4.70 0.24 -0.13 0.008 0.320
E 4.60 0.03
e 4.67 0.26 -0.07 0.005 0.631
f 4.58 0.05 0.11 0.021 0.384 significant
G 4.70 0.24 Higher if not stirred immediately before
g 4.58 0.05 0.12 0.026 0.362 testing; not significant
213
Table A-25 Angelo’s Sand Significance of Ruggedness Study Factors for pH Obtained
from the Combined Results of Replicate Studies
Factor F Statistic1 Conclusion
A, a 4.43 Cool vs warm transport is not significant
B, b 1.97 Cold vs warm storage is not significant
C, c 0.02 “As received” vs dried, sieved is not significant
D, d 2.26 100 mL vs 100 g is not significant
E, e 0.12 Electrode 1 vs 2 is not significant
F, f 0.00 30-min vs 60-min equilibrium period is not significant
G, g 0.00 Not stirring vs stirring before testing is not significant
1
Calculated as described in ASTM C1067 (2007); an F statistic ≥ 5.59 is significant at the 95%
confidence level.
214
Table A-26 Angelo’s Sand Ruggedness Study Results for Minimum Resistivity
D d DI DI DI-S DI-S DI-S DI-S DI DI
F f 1.0 0.5 0.5 1.0 1.0 0.5 0.5 1.0
Resistivity Results
17,000 21,000 16,500 15,000 16,500 16,000 22,000 16,000
(ohm-cm)
D DI water (>1,000,000 ohm-cm; measured in large soil box) (DI)
d DI water, sodium chloride added (175,000 ohm-cm, measured in large soil box) (DI-S)
F 1.0 kg soil (1.0)
f 0.5 kg soil (0.5)
G No equilibrium period (no)
g Equilibrium period of 30 min after adding 20% dilution water (yes)
Note 1 Sampling date was 9/17/2014 and ruggedness study dates were 9/27/2014 & 9/28/2014
Note 2 4,000 ohm-cm conductivity standard was 4,000 ohm-cm at ~25.0C measured in the large box
215
Table A-27 Angelo’s Sand Significance of Ruggedness Study for Minimum Resistivity
A 17,400 2,560 Higher for transport under ambient
a 17,600 2,930 -250 0.832 0.902 conditions; not significant
b 17,400 3,150 250 0.613 0.902 significant
C 18,000 2,680 Higher for “as received” soil; not
c 17,000 2,710 1,000 0.985 0.618 significant
D 19,000 2,940
Higher for DI water; not significant
d 16,000 707 3,000 0.043 0.095
E 16,400 479 Higher for large soil box; not
e 18,600 3,400 -2,250 0.009 0.238 significant
F 16,100 854
f 18,900 3,070 -2,750 0.064 0.135
G 17,500 3,110 No difference with and without an
g 17,500 2,350 0 0.655 1.000 equilibrium period; not significant
216
Table A-28 Angelo’s Sand Ruggedness Study Results for Sulfate
B b 1 1 2 2 1 1 2 2
C c 110 60 110 60 110 60 110 60
D d no no yes yes yes yes no no
E e 1 2 1 2 2 1 2 1
A Transport in cooler, store in refrigerator (CC)
a Transport under ambient conditions, store at room temperature (WW)
B Reagent Lot 4120 (1)
b Reagent Lot 4202 (2)
D No added acid (no)
d Added acid (yes)
E Range 1 factory default (1)
e Range 2 factory default (2)
F No membrane filtration (no)
f Membrane filtration (yes)
Note 2 20 ppm check standard: 24 ppm, range 1, lot 4120; 27 ppm, range 2, lot 4204
217
Table A-29 Angelo’s Sand Significance of Ruggedness Study for Sulfate
A 28 28 Higher for transport and storage under
a 35 19 -7 0.572 0.724 ambient conditions; not significant
B 44 26 Higher for reagent lot 4120; not
b 19 10 25 0.139 0.132 significant
C 32 21 Higher if dried at 110C; not
c 31 27 1 0.730 0.922 significant
D 36 22
Higher if no added acid; not significant
d 27 25 9 0.869 0.599
E 20 4 Higher if tested on Range 2; not
e 43 29 -23 0.008 0.175 significant
F 29 24 Higher for membrane filtration; not
f 34 24 -5 0.969 0.789 significant
G 20 9 Higher with a 20 ppm sulfate spike; not
g 43 27 -23 0.104 0.164 significant
218
A.6 Mine 6 Sebring Sand
Table A-30 Sebring Sand Ruggedness Study Results for pH (Replicate 1)

E e 1 2 1 2 2 1 2 1
F f 30 60 60 30 30 60 60 30
pH Results 4.69 4.42 4.39 4.36 4.39 4.39 4.38 4.32
D 100 mL (ML)
d 100 g (G)
E Electrode 1, 13-620-631 (99.6% slope, 1.5 mV, 24.7 C) (1)
e Electrode 2, 13-620-631 (99.2% slope, 2.8 mV, 24.5 C) (2)
219
Table A-31 Sebring Sand Significance of Ruggedness Study for pH (Replicate 1)
A 4.47 0.15
a 4.37 0.03 0.10 0.034 0.268
B 4.47 0.15
b 4.36 0.03 0.11 0.030 0.190
c 4.37 0.04 0.09 0.066 0.297 significant
D 4.45 0.16 Higher for a 100 mL sample; not
d 4.38 0.02 0.07 0.003 0.427 significant
E 4.45 0.17
e 4.39 0.03 0.06 0.011 0.499
f 4.40 0.02 0.05 0.004 0.616 significant
220
Table A-32 Sebring Sand Ruggedness Study Results for pH (Replicate 2)
E e 1 2 1 2 2 1 2 1
F f 30 60 60 30 30 60 60 30
pH Results 4.39 4.35 4.34 4.34 4.37 4.35 4.32 4.33
D 100 mL (ML)
d 100 g (G)
E Electrode 1, 13-620-631 (99.5% slope, 1.3 mV, 24.5 C) (1)
e Electrode 2, 13-620-631 (99.4% slope, 1.8 mV, 24.3 C) (2)
221
Table A-33 Sebring Sand Significance of Ruggedness Study for pH (Replicate 2)
A 4.36 0.02
a 4.34 0.02 0.01 0.910 0.471
B 4.37 0.02
Higher if stored cold; significant
b 4.33 0.01 0.03 0.285 0.023
c 4.34 0.01 0.01 0.084 0.471 significant
D 4.35 0.03
d 4.35 0.01 0.00 0.231 0.888
E 4.35 0.03
e 4.35 0.02 0.01 0.710 0.670
f 4.34 0.01 0.02 0.303 0.301 significant
222
TableA-34 Sebring Sand Significance of Ruggedness Study Factors for pH Obtained
B, b 5.35 Cold vs warm storage is significant (borderline)
E, e 0.27 Electrode 1 vs Electrode 2 is not significant
1
confidence level.
223
Table A-35 Sebring Sand Ruggedness Study Results for Minimum Resistivity
F f 1.0 0.5 0.5 1.0 1.0 0.5 0.5 1.0
Resistivity Results
36,500 42,500 37,000 30,000 34,500 28,500 47,000 26,500
(ohm-cm)
d DI water, sodium chloride added (175,000 ohm-cm, measured in large soil box) (DI-S)
F 1.0 kg soil (1.0)
f 0.5 kg soil (0.5)
Note 2 4,000 ohm-cm conductivity standard was 4,000 ohm-cm at ~24.0C measured in the large box
224
Table A-36 Sebring Sand Significance of Ruggedness Study for Minimum Resistivity
A 36,500 5,120 Higher for transport under cool
a 34,100 9,230 2,380 0.358 0.668 conditions; not significant
b 35,100 9,040 375 0.481 0.947 significant
C 38,800 5,610 Higher for “as received” soil; not
c 31,900 7,230 6,880 0.686 0.183 significant
D 38,100 8,860
Higher for DI water; not significant
d 32,500 3,940 5,630 0.216 0.290
E 32,100 5,410 Higher for large soil box; not
e 38,500 7,670 -6,380 0.580 0.223 significant
F 31,900 4,500
f 38,800 7,960 -6,880 0.373 0.183
G 35,500 8,420 Higher for no equilibrium period; not
g 35,100 6,650 375 0.708 0.947 significant
225
Table A-37 Ruggedness Study Results for Sulfate
B b 1 1 2 2 1 1 2 2
C c 110 60 110 60 110 60 110 60
D d no no yes yes yes yes no no
E e 1 2 1 2 2 1 2 1
A Transport in cooler, store in refrigerator (CC)
a Transport under ambient conditions, store at room temperature (WW)
B Reagent Lot 4120 (1)
b Reagent Lot 4202 (2)
D No added acid (no)
d Added acid (yes)
E Range 1 factory default (1)
e Range 2 factory default (2)
F No membrane filtration (no)
f Membrane filtration (yes)
G No change (no)
g No change (yes)
Note 1 Sampling date was 10/15/2014 and ruggedness study date was10/30/2014
Note 2 30 ppm check standard: 35 ppm, range 1, lot 4120; 41 ppm, range 2, lot 4204
226
Table A-38 Sebring Sand Significance of Ruggedness Study for Sulfate
A 8 2 Not different for transport and storage
under ambient conditions; not
a 8 3 0 0.39 1.00 significant
B 8 0
Higher for reagent lot 4120; significant
b 6 2 2 0.00 0.05
C 8 2
Higher if dried at 110 C; not significant
c 7 3 1 0.31 0.39
D 8 3
Higher if no added acid; not significant
d 8 2 0 0.39 1.00
E 7 3 Higher if tested on Range 2; not
e 8 2 -1 0.31 0.39 significant
F 7 3 Higher for membrane filtration; not
f 8 2 -1 0.31 0.39 significant
G 8 2
Higher for no change, not significant
g 7 3 1 0.31 0.39
227
A.7 Mine 7 Clermont Sand
Table A-39 Clermont Sand Ruggedness Study Results for pH (Replicate 1)

E e KCl CaCl2 KCl CaCl2 CaCl2 KCl CaCl2 KCl
F f 30 60 60 30 30 60 60 30
pH Results 7.55 7.16 7.84 7.38 7.05 7.31 7.15 7.70
E Amendment with 0.1 g KCl (KCl)
e Amendment with 0.1 g CaCl (CaCl2)
Note 2 Electrode/meter AB150/13-620-631; slope = 99.8% and offset = 2.1 mV @ 25.3C
228
Table A-40 Clermont Sand Significance of Ruggedness Study Factors for pH (Replicate
1)
a 7.30 0.29 0.18 0.995 0.408 significant
B 7.27 0.22 Higher if stored warm; not
b 7.52 0.31 -0.25 0.565 0.236 significant
c 7.39 0.28 0.01 0.457 0.964 significant
d 7.40 0.33 -0.00 0.790 0.982 significant
E 7.60 0.23 Higher if amended with 0.1 g KCl;
e 7.19 0.14 0.41 0.444 0.021 significant
before testing; not significant
g 7.44 0.39 -0.09 0.193 0.686
229
Table A-41 Clermont Sand Ruggedness Study Results for pH (Replicate 2)
F f 30 60 60 30 30 60 60 30
pH Results 7.55 7.28 7.81 6.98 7.16 7.59 7.23 7.67
230
Table A-42 Clermont Sand Significance of Ruggedness Study Factors for pH (Replicate
2)
A 7.41 0.36 Higher if transported warm; not
a 7.41 0.26 -0.007 0.595 0.974 significant
B 7.40 0.21 Higher if stored warm; not
b 7.42 0.39 -0.027 0.341 0.904 significant
c 7.38 0.32 0.057 0.940 0.801 significant
D 7.43 0.21 Higher for a 100 mL sample; not
d 7.39 0.38 0.048 0.359 0.835 significant
e 7.16 0.13 0.492 0.830 0.001 significant
f 7.48 0.27 -0.138 0.785 0.540 not significant
g 7.48 0.31 -0.143 0.907 0.525
231
Table A-43 Clermont Sand Significance of Ruggedness Study Factors for pH Obtained
B, b 8.11 Cold vs warm storage is significant
E, e 927 0.1 g KCl vs 0.1 g CaCl2 is significant
1
confidence level.
232
Table A-44 Clermont Sand Ruggedness Study Results for Minimum Resistivity
F f 1.0 1.5 1.5 1.0 1.0 1.5 1.5 1.0
D DI water (>1,000,000 ohm-cm; measured in small soil box) (DI)
d DI water, sodium chloride added (165,000 ohm-cm, measured in small soil box) (DI-S)
F 1.0 kg soil (1.0)
f 1.5 kg soil (1.5)
Note 2 4,000 ohm-cm conductivity standard was 3,850 ohm-cm at ~24.3C
233
Table A-45 Clermont Sand Significance of Ruggedness Study for Minimum Resistivity
a 22,900 3,710 -375 0.860 0.885 ambient conditions; not significant
B 22,800 3,590 Higher if stored cold; not
b 22,600 3,450 125 0.948 0.962 significant
C 22,900 3,710 Higher if tested “as received”; not
c 22,500 3,320 375 0.860 0.885 significant
D 23,600 1,890 Higher if tested with DI water; not
d 21,800 4,350 1,880 0.204 0.459 significant
E 20,000 2,310 Higher if tested in the large soil
e 25,400 479 -5,400 0.028 0.004 box; significant
f 21,600 4,190 2,130 0.274 0.398 not significant
G 22,600 3,450 Higher for a 30-min equilibrium
g 22,800 3,590 -125 0.948 0.962 period; not significant
234
A.8 Mine 8 Alico Road Sand
Table A-46 Alico Road Sand Ruggedness Study Results for pH (Replicate 1)
F f 30 60 60 30 30 60 60 30
pH Results 8.99 8.06 8.87 8.28 8.32 8.94 8.02 8.66
235
Table A-47 Alico Road Sand Significance of Ruggedness Study Factors for pH
(Replicate 1)
a 8.49 0.40 0.07 0.851 0.836 significant
B 8.58 0.46 Higher if stored cold; not
b 8.46 0.38 0.12 0.761 0.702 significant
c 8.49 0.39 0.07 0.803 0.836 significant
d 8.60 0.35 -0.17 0.636 0.585 significant
e 8.17 0.15 0.70 0.943 0.001 significant
G 8.56 0.48 Higher if not stirred immediately
g 8.48 0.36 0.08 0.638 0.799
236
Table A-48 Alico Road Sand Ruggedness Study Results for pH (Replicate 2)
F f 30 60 60 30 30 60 60 30
pH Results 8.85 8.11 8.95 8.13 8.28 8.84 8.17 8.67
237
(Replicate 2)
a 8.49 0.32 0.02 0.574 0.945 significant
B 8.52 0.38 Higher if stored cold; not
b 8.48 0.40 0.04 0.946 0.889 significant
c 8.44 0.37 0.13 0.930 0.661 significant
d 8.55 0.41 -0.10 0.871 0.727 significant
e 8.17 0.08 0.66 0.503 0.000 significant
f 8.52 0.44 -0.04 0.666 0.903 not significant
g 8.50 0.38 -0.01 0.928 0.986
238
Obtained from the Combined Results of Replicate Studies
A,a 0.96 Cool vs warm transport is not significant
B,b 12.0 Cold vs warm storage is significant
C,c 23.9 “As received” vs dried, sieved is significant
D,d 97.3 100 mL vs 100 g is significant
E,e 61,000 0.1 g KCl vs 0.1 g CaCl2 is significant
F,f 0.17 30-min vs 60-min equilibrium period is not significant
G,g 0.58 Not stirring vs stirring before testing is not significant
1
confidence level.
239
Table A-51 Alico Road Sand Ruggedness Study Results for Minimum Resistivity
F f 1.0 1.5 1.5 1.0 1.0 1.5 1.5 1.0
d DI water, sodium chloride added (150,000 ohm-cm, measured in small soil box) (DI-S)
F 1.0 kg soil (1.0)
f 1.5 kg soil (1.5)
Note 2 4,000 ohm-cm conductivity standard was 4,150 ohm-cm at ~24.5C
240
Table A-52 Alico Road Sand Significance of Ruggedness Study for Minimum
Resistivity
a 10,300 941 -613 0.236 0.605 ambient conditions; not significant
B 9,810 922 Higher if stored at room
b 10,100 2,100 -263 0.210 0.826 temperature; not significant
C 9,640 1,630 Higher if tested dried and sieved;
c 10,300 1,540 -613 0.931 0.605 not significant
D 9,600 985 Higher if tested with 150,000 ohm-
d 10,300 2,000 -713 0.276 0.546 cm water; not significant
E 8,890 905 Higher if tested in the large soil
e 11,000 1,230 -2,110 0.632 0.032 box; significant
f 9,550 1,360 788 0.687 0.503 not significant
G 10,560 1,570 Higher for no equilibrium period;
g 9,330 1,340 1,240 0.799 0.276 not significant
241
Appendix B Treatment Study Results
Statistically-significant results are highlighted.
B.1 Mine 1 Wimauma Sand
Table B-1 Wimauma Sand Treatment Study Results for pH

Test Date 5/8/2014 5/9/2014 5/9/2014 5/10/2014 5/10/2014 5/15/2014
Treatment 1 2 3 4 5 6
Soil Slurry T, C 28.5 23.0 23.0 23.3 23.3 25.0
Replicate 1 4.50 4.60 4.70 4.62 4.41 4.77
Replicate 2 4.57 4.67 4.68 4.71 4.63 4.79
Replicate 3 4.67 4.65 4.56 4.86 4.70 4.68
Avg 4.58 4.64 4.65 4.73 4.58 4.75
St Dev 0.09 0.04 0.08 0.12 0.15 0.06
t-tests 0.33 0.37 0.15 1.00 0.05
0.90 0.29 0.54 0.05
0.37 0.53 0.14
0.25 0.84
0.15
No F-tests were significant
pH meter/electrode calibration for Accumet AP85/13-620-AP55
Slope, % Not recorded
Offset, mV Not recorded
Treatments
Treatment 1 On-site in the shade within 1 hr of sampling, and tested "as received"
Sample was transported to the lab at ambient temperature (~30C) and stored at
Treatment 2
ambient temperature (23-25C), quartered, and tested “as received”
Sample was transported to the lab in a cooler (~15C) and stored in the refrigerator (4
Treatment 3
C), brought to room temperature over several hr, quartered, and tested “as received”
Treatment 4
ambient temperature (23-25C), oven-dried at 85C, sieved, mechanically split
Same as 4, except measurement was repeated after sample was quiescent at room
Treatment 5
temperature for an hr
Treatment 6
ambient temperature (23-25C), soil was air-dried, sieved, mechanically split
242
Table B-2 Wimauma Sand Treatment Study Results for Minimum Resistivity
Test Date 5/8/2014 5/9/2014 5/9/2014 5/10/2014 5/15/2014
Treatment 1 2 3 4 6
Soil Slurry T, C 27.0 22.5 22.0 23.0 23.0
Replicate 1 39,000 42,500 42,000 48,500 50,000
Replicate 2 41,500 46,500 44,000 48,000 46,500
Replicate 3 45,000 42,000 44,000 44,000 46,500
Avg 41,800 43,700 43,300 46,800 47,700
St Dev 3,010 2,470 1,160 2,470 2,020
t-tests 0.46 0.47 0.09 0.05
0.84 0.19 0.10
0.09 0.03
0.67
Resistivity meter calibration checks
DI water, ohm-cm 310,000 850,000 850,000 1,100,000 Not recorded
Std, 4,000 ohm-
Not recorded 4,200 4,200 4,200 Not recorded
cm
Treatment
Treatment 2
Sample was transported to the lab in a cooler (~15C) and stored in the refrigerator
Treatment 3 (4C), brought to room temperature over several hr, quartered, and tested “as
received”
Sample was transported to the lab at ambient temperature (~30C and stored at
Treatment 4
Treatment 5
ambient temperature (23-25C), soil was air-dried, sieved, mechanically split
243
Table B-3 Wimauma Sand Treatment Study Results for Chloride
Test Date 6/1/2014 6/1/2014
Treatment 1 2
Replicate 1 BD BD
Replicate 2 BD BD
Replicate 3 BD BD
Avg Not calculated Not calculated
St Dev Not calculated Not calculated
Results for chloride spike, ppm
20 ppm spike 15 Not tested
10 ppm spike Not tested 5
Treatments
Samples were transported cool, stored cold, quartered, oven-dried at 60C, extracted
Treatment 1 with 100 g soil and 100 mL DI water and allowed to sit for 36 hr; samples were then
filtered through a coarse, fast filter (Fisher, Q8); 50 mL of sample extract was diluted
with 100 mL of DI water (3 x dilution).
Samples were transported cool, stored cold, quartered, oven-dried at 110C,
Treatment 2 extracted with 100 g soil and 100 mL DI water and allowed to sit for 36 hr; samples
were then filtered through a coarse, fast filter (Fisher, Q8); 50 mL of sample extract
was diluted with 100 mL of DI water (3 x dilution).
Notes
Note 1 Analysis was with Hach chloride test kit, 5 to 400 mg/L
Samples turned color with one drop of silver nitrate reagent, indicating that the
Note 2 concentration was at or below the minimum detection level of 5 ppm; less if
corrected for the blank
244
Table B-4 Wimauma Sand Treatment Study Results for Sulfate
Test Date 6/1/2014 6/1/2014
Treatment 1 2
Replicate 1 12 9
Replicate 2 9 9
Replicate 3 9 12
Avg 10 10
St Dev 2 2
Results for sulfate check standard and spike, ppm
20 ppm CS 17 Not tested
Treatments
Samples were transported cool, stored cold, quartered, oven-dried at 60C, extracted
Treatment 1 with 100 g soil and 100 mL DI water and allowed to sit for 36 hr; samples were then
filtered through a coarse, fast filter (Fisher, Q8); 50 mL of sample extract was diluted
with 100 mL of DI water (3 x dilution).
Samples were transported cool, stored cold, quartered, oven-dried at 110C,
Treatment 2 extracted with 100 g soil and 100 mL DI water and allowed to sit for 36 hr; samples
were then filtered through a coarse, fast filter (Fisher, Q8); 50 mL of sample extract
was diluted with 100 mL of DI water (3 x dilution).
Notes
Analysis was with Hach photometer test kit, 2 to 70 mg/L, on Range 1 (factory
Note 1
default)
245
B.2 Mine 2 Jahna Sand
Table B-5 Jahna Sand Treatment Study Results for pH

Test Date 5/29/2014 5/30/2014 5/30/2014 6/5/2014 6/5/2014
Treatment 1 2 3 4 5
Avg Soil Slurry T, C 27.1 23.3 23.1 23.7 23.7
Replicate 1 5.08 4.97 5.29 5.22 5.34
Replicate 2 5.08 4.76 5.00 5.35 5.37
Replicate 3 5.15 5.17 5.18 5.27 5.22
Average 5.10 4.97 5.16 5.28 5.31
St Dev 0.04 0.21 0.15 0.07 0.08
t-tests 0.32 0.58 0.02 0.02
0.26 0.07 0.05
0.25 0.19
0.64
Slope, % Not recorded 94.3 97.7 Not recorded Not recorded
Offset, mV Not recorded 4.7 8.3 Not recorded Not recorded
Treatments
Treatment 1 On-site in the shade within 2 hr of sampling, soil "as received"
Treatment 2
Sample was transported to the lab in a cooler (~18C) and stored in the
Treatment 3 refrigerator (4C), brought to room temperature over several hr, quartered, and
tested “as received”
Treatment 4
Treatment 5 ambient temperature (23-25C), oven-dried at 110C, sieved, mechanically
split
246
Table B-6 Jahna Sand Treatment Study Results for Minimum Resistivity
Test Date 5/29/2014 5/30/2014 5/30/2014 6/9/2014
Treatment 1 2 3 4
Soil Slurry T, C 27.1 22.5 23.8 23.4
Replicate 1 110,000 125,000 120,000 110,000
Replicate 2 110,000 130,000 120,000 110,000
Replicate 3 100,000 120,000 120,000 115,000
Average 107,000 125,000 120,000 112,000
St Dev 5,770 5,000 0 2,890
t-tests 0.01 0.02 0.25
0.16 0.02
0.01
Standard deviation was significantly lower for treatment 3 than treatments 1, 2, and 4
DI water, Ω-cm Not recorded 1,100,000 1,100,000 850,000
Standard, 4000 Ω-cm Not recorded 3,900 3,900 4,000
Treatments
1 On-site in the shade within 2 hr of sampling, and tested "as received"
Sample was transported to the lab at ambient temperature (~30C) and stored
2
at ambient temperature (23-25C), quartered, and tested “as received”
Sample was transported to the lab in a cooler (~18C) and stored in the
3 refrigerator (4C), brought to room temperature over several hr, quartered, and
4 at ambient temperature (23-25C), oven-dried at 60C, sieved, mechanically
split
247
Table B-7 Jahna Sand Treatment Study Results for Chloride
Test Date 6/1/2014 6/1/2014
Treatment 1 2
Replicate 1 BD BD
Replicate 2 BD BD
Replicate 3 BD BD
Average Not calculated Not calculated
20 ppm spike 15 Not tested
Treatments
Sample was transported to the lab at ambient temperature (~30C and stored at
Treatment 1 ambient temperature (23-25C), oven-dried at 60C, extracted with 1:1
soil:water, diluted x 3, filtered with 0.45 micron membrane filter
Treatment 2 at ambient temperature (23-25C), oven-dried at 110C, extracted with 1:1
Notes
248
Table B-8 Jahna Sand Treatment Study Results for Sulfate
Test Date 6/1/2014 6/1/2014
Treatment 1 2
Replicate 1 BD BD
Replicate 2 BD BD
Replicate 3 BD BD
Results for sulfate check standard and spike, ppm

Treatments
Notes
Note 1
default)
Photometer readings were 2 ppm or less (6 ppm as soil sulfate) and were below
Note 2
the method detection level
249
B.3 Mine 3 Youngquist Sand
Table B-9 Youngquist Sand Treatment Study Results for pH

Test Date 6/19/2014 6/20/2014 6/20/2014 6/22/2014 7/11/2014
Treatment 1 2 3 4 5
Soil Slurry T,
28.2 21.2 21.0 21.1 24.8
C
Replicate 1 8.20 8.25 8.42 8.11 8.05
Replicate 2 8.17 8.33 8.44 8.16 8.20
Replicate 3 8.08 8.22 8.36 8.18 8.27
Average 8.15 8.27 8.41 8.15 8.24
St Dev 0.06 0.06 0.04 0.04 0.12
t-tests 0.07 0.00 1.00 0.77
0.03 0.04 0.27
0.00 0.03
0.75
Slope, % 95.0 97.9 97.9 97.9 98.5
Offset, mV 5.3 4.1 4.1 4.7 4.8
Treatments
Sample was transported to the lab at ambient temperature (~27-30C) and stored at
Treatment 2
Sample was transported to the lab in a cooler (~16-23C) and stored in the refrigerator
Treatment 3 (4C), brought to room temperature over several hr, quartered, and tested “as
received”
Treatment 4
Same as 4, except sample was dried at 110C and tested with electrode/meter
Treatment 5
Accumet AB150/13-620-631
250
Table B-10 Youngquist Sand Treatment Study Results for Minimum Resistivity
Test Date 6/19/2014 6/20/2014 6/20/2014 6/21/2014 6/22/2014
Treatment 1 2 3 4 5
Soil Slurry T, C 29.2 22.0 21.1 22.0 21.4
Replicate 1 9,600 8,600 8,050 9,750 8,750
Replicate 2 10,200 9,200 7,850 10,500 8,650
Replicate 3 10,000 9,150 7,500 10,800 8,700
Average 9,930 8,980 7,800 10,300 8,700
St Dev 306 333 278 320 50
t-tests 0.02 0.00 0.31 0.00
0.01 0.02 0.22
0.00 0.01
0.01
F-tests revealed a significantly lower variance for treatment 5 versus treatments 1, 2, and 4
DI water, ohm-cm Not recorded >1,100,000 >1,100,000 Not recorded >1,100,000
Std, 4 000 ohm-
Not recorded 4,400 5,050 Not recorded 4,400
cm
Treatments
On-site in the shade within 2 hr of sampling, and tested "as received" (Note:
Treatment 1
terminals P1&C1 and P2&C2 were shorted)
Sample was transported to the lab at ambient temperature (~27-30 C) and stored at
Treatment 2
ambient temperature (22-24 C), quartered, and tested “as received”
Sample was transported to the lab in a cooler (~16-23 C) and stored in the
Treatment 3 refrigerator (4 C), brought to room temperature over several hr, quartered, and
Treatment 4 ambient temperature (22-24 C), quartered, and tested “as received” (Note:
terminals P1&C1 and P2&C2 were shorted)
Treatment 5
ambient temperature (22-24 C), soil dried at 60 C, sieved, mechanically split
251
Table B-11 Youngquist Sand Treatment Study Results for Chloride
Test Date 7/10/2014 7/10/2014
Treatment 1 2
Replicate 1 BD BD
Replicate 2 BD BD
Replicate 3 BD BD
Results for chloride check standard, ppm
Treatments
Treatment 1 ambient temperature (23-25C), oven-dried at 60C, extracted with 3:1 soil:water,
filtered with Whatman 4, filtered with Fisher Q8, filtered with 0.45 micron membrane
filter
filtered with Whatman 4, filtered with Fisher Q8, filtered with 0.45 micron membrane
filter
Notes
252
Table B-12 Youngquist Sand Treatment Study Results for Sulfate
Test Date 7/10/2014 7/10/2014
Treatment 1 2
Replicate 1 27 45
Replicate 2 36 36
Replicate 3 36 36
Average 33 39
St Dev 5 5
t-tests 0.23
Results for Sulfate Check Standard, ppm
Treatments
Treatment 1 ambient temperature (23-25C), oven-dried at 60 C, extracted with 3:1 soil:water,
filtered with Whatman 4, 4 drops conc HNO3 added, filtered with 0.45 micron
membrane filter
filtered with Whatman 4, 4 drops conc HNO3 added, filtered with 0.45 micron
membrane filter
Notes
Note 1
default)
253
B.4 Mine 4 Calhoun Sand
Table B-13 Calhoun Sand Treatment Study Results for pH

Test Date 7/14/2014 7/15/2014 7/15/2014
Treatment 1 2 3
Avg Soil Slurry T, C 31.5 23.5 24.0
Replicate 1 4.51 4.56 4.47
Replicate 2 4.72 4.60 4.59
Replicate 3 4.56 4.57 4.50
Replicate 4 4.53
Replicate 5 4.74
Replicate 6 4.56
Average 4.60 4.59 4.52
St Dev 0.11 0.08 0.06
t-tests 0.77 0.35
0.21
Slope, % 98.6 99.1 96.6
Offset, mV 4.1 2.4 2.4
Treatments
Sample was transported to the lab at ambient temperature (~30-32C) and
Treatment 2
stored at ambient temperature (~24C), quartered, and tested "as received"
Sample was transported to the lab in a cooler (~17-23C) and stored in the
tested "as received"
254
Table B-14 Calhoun Sand Treatment Study Results for Minimum Resistivity
7/14/201
Date 7/15/2014 7/15/2014 7/16/2014
4
Treatment 1 2 3 4
Soil Slurry T, C 32.4 22.6 24.1 23.5
Replicate 1 103,000 109,000 125,000 92,000
Replicate 2 85,000 110,000 120,000 97,000
Replicate 3 95,000 110,000 110,000 93,500
Average 94,300 110,000 118,000 94,200
St Dev 9,020 577 7,640 2,570
t-tests 0.04 0.02 0.98
0.12 0.00
0.01
Significant variance between treatment 2 and treatments 1 and 3
DI water, ohm-cm 725,000 >1,000,000 >1,000,000 >1,000,000
Treatments
Treatment 2
ambient temperature (~24C), quartered, and tested "as received"
tested "as received"
Treatment 4
ambient temperature (~24C), quartered, dried at 60C, sieved through No. 10 sieve
255
Table B-15 Calhoun Sand Treatment Study Results for Chloride
Test Date 7/18/2014 7/18/2014 7/18/2014
Treatment 1 2 3
Replicate 1 BD BD BD
Average Not calculated Not calculated Not calculated
St Dev Not calculated Not calculated Not calculated
Results for chloride check standard, ppm
20 ppm CS 20 20 25
Treatments
filtered with Fisher Q8, filtered with 0.45 micron membrane filter
Treatment 3 (4C), quartered, oven-dried at 60C, extracted with 3:1 soil:water, filtered with
Fisher Q8, filtered with 0.45 micron membrane filter
Notes
Note 2 concentration was at or below the minimum detection level of 5 ppm; less if corrected
for the blank
256
Table B-16 Calhoun Sand Treatment Study Results for Sulfate
Test Date 7/18/2014 7/18/2014 7/18/2014
Treatment 1 2 3
Replicate 1 BD BD 9
Replicate 2 BD 9 BD
Average Not calculated Not calculated Not calculated
St Dev Not calculated Not calculated Not calculated
Treatments
Treatment 3 refrigerator (4C), quartered, oven-dried at 60C, extracted with 3:1 soil:water,
Results for sulfate check standard, ppm
20 ppm CS 13 16 16
Notes
Note 1
default)
Except for two extracts with photometer readings of 3 ppm (9 ppm as soil sulfate),
Note 2 photometer sulfate readings were 2 ppm or less (6 ppm as soil sulfate) and were
below the method detection level
257
B.5 Mine 5 Angelo’s Sand
Table C-17 Angelo’s Sand Treatment Study Results for pH

Test Date 9/17/2014 9/20/2014 9/20/2014 9/18/2014 9/18/2014
Treatment 1 2 3 4 5
Soil Slurry T, C 23.5 24.6 24.3 24.5 24.5
Replicate 1 4.56 4.58 4.69 4.54 4.59
Replicate 2 4.55 4.57 4.58 4.56 4.60
Replicate 3 4.59 4.60 4.61 4.58 4.60
Average 4.57 4.58 4.63 4.56 4.60
St Dev 0.02 0.02 0.06 0.02 0.01
t-tests 0.33 0.16 0.71 0.07
0.27 0.18 0.23
0.13 0.41
0.04
F-tests 0.70 0.24 0.96 0.14
0.13 0.74 0.25
0.22 0.02
0.15
Meter/electrode Accumet AP85/13-620-AP55
slope, % 97.3 . . . 97.0
offset, mV 9.5 . . . 10.6
buffer temp, C 24.1 . . . 24.5
Meter/electrode Accumet AB150/13-620-631
slope, % . 99.5 99.0 99.1 .
offset, mV . 3.0 3.5 3.7 .
buffer temp, C . 24.9 24.8 24.5 .
Treatments
1 On-site testing, indoors
Sample was transported to the lab at ambient temperature (~24C) and stored at ambient
2
temperature (~24C), quartered, soil at "as received" moisture content
Sample was transported in a cooler (~15-21C) and stored in a refrigerator at 4C,
3
quartered, and brought to room temperature before testing "as received"
4
temperature (~24C), dried at 60C, No. 10 sieve, mechanically split
Sample was transported to the lab at ambient temperature (~24 C) and stored at ambient
5
temperature (~24C), dried at 60C, No. 10 sieve, mechanically split
258
Table B-18 Angelo’s Sand Treatment Study Results for Minimum Resistivity
Test Date 9/20/2014 9/20/2014 9/21/2014
Treatment 1 2 3
DI water, ohm-cm 780,000 780,000 1,000,000
Soil Slurry T, C 24.1 24.2 24.0
Replicate 1 17,000 19,000 17,500
Replicate 2 17,000 19,000 17,500
Replicate 3 17,000 19,500 17,000
Average 17,000 19,167 17,333
St Dev 0 289 289
t-tests 0.00 0.12
0.00
F-tests 0.00 0.00
1.00
Treatments
1
ambient temperature (~24C), quartered, soil at "as received" moisture content
Sample was transported in a cooler (~15-21C) and stored in a refrigerator at 4C,
2 quartered, and brought to room temperature before testing at "as received"
moisture content
3
ambient temperature (~24C), dried at 60C, No. 10 sieve, mechanically split
259
Table B-19 Angelo’s Sand Treatment Study Results for Chloride
Test Date 9/19/2014 9/19/2014
Treatment 1 2
Replicate 1 BD BD
Replicate 2 BD BD
Replicate 3 BD BD
Treatments
1 at ambient temperature (23-25C), oven dried at 60C, extracted with 3:1
soil:water, filtered with Fisher Q8, filtered with 0.45 micron membrane filter
2 at ambient temperature (23-25C), oven dried at 110C, extracted with 3:1
soil:water, filtered with Fisher Q8, filtered with 0.45 micron membrane filter
Notes
1 Analysis was with Hach chloride test kit, 5 to 400 mg/L
2 concentration was at or below the minimum detection level of 5 ppm; less if
260
Table B-20Angelo’s Sand Treatment Study Results for Sulfate
Test Date 9/19/2014 9/19/2014 9/19/2014 9/19/2014
Treatment 1 2 3 4
Test Range Range 1 Range 2 Range 1 Range 2
Replicate 1 24 27 21 27
Replicate 2 27 33 21 27
Replicate 3 21 21 18 24
Average 24 27 20 26
St Dev 3 6 2 2
t-tests 0.48 0.12 0.37
0.12 0.80
0.01
F-tests 0.40 0.50 0.50
0.15 0.15
1.00
Treatments
1 temperature (23-25C), oven dried at 60C, extracted with 3:1 soil:water, filtered with
Range 1 Range 2
30 ppm CS 31 40
Notes
1 Analysis was with photometer test kit, 2 to 70 mg/L, on range 1 (factory default)
Except for two extracts with photometer readings of 3 ppm (9 ppm in soil), photometer
2
readings were 2 ppm or less and were below the method detection level
261
B.6 Mine 6 Sebring Sand
Table B-21 Sebring Sand Treatment Study Results for pH

Test Date 10/15/2014 10/16/2014 10/17/2014 10/20/2014 10/16/2014 10/17/2014
Soil Slurry T,
25.8 24.5 24.7 24.5 24.1 24.8
C
Replicate 1 4.37 4.27 4.39 4.35 4.29 4.43
Replicate 2 4.43 4.33 4.42 4.39 4.33 4.46
Replicate 3 4.30 4.33 4.43 4.43 4.36 4.47
Average 4.37 4.31 4.41 4.39 4.33 4.45
St Dev 0.07 0.03 0.02 0.04 0.04 0.02
t-tests 0.25 0.30 0.62 0.40 0.09
0.01 0.06 0.59 0.00
0.42 0.02 0.08
0.11 0.07
0.01
slope, % 95.4 . . . 95.0 96.2
offset, mV 1.8 . . . 2.4 4.7
buffer, C 25.7 . . . 24.5 24.6
slope, % . 99.7 99.0 99.2 . .
offset, mV . 2.1 1.3 2.30 . .
buffer, C . 24.6 25.0 25.1 . .
Treatments
1 On-site in the shade within 2 hr of sampling, soil "as received"
2
temperature (~25C), quartered, soil tested "as received"
3
(4C), brought to room temperature, quartered, soil tested "as received"
4 temperature (~24C), dried at 60C for ~4 hr, sieved through a No. 10 mesh, and
mechanically split
5
temperature (~24C), quartered, soil tested "as received"
6
262
Table B-22 Sebring Sand Treatment Study Results for Minimum Resistivity
Test Date 10/16/2014 10/17/2014 10/19/2014
Treatment 1 2 3
DI water, ohm-cm >1,000,000 >1,000,000 810,000
Conductivity Std,
4,000 . 4,000
4,000 ohm-cm
Soil Slurry T, C 24.0 24.3 24.8
Replicate 1 40,000 36,000 32,000
Replicate 2 38,000 38,500 30,500
Replicate 3 40,000 38,000 31,500
Average 39,300 37,500 31,300
St Dev 1,160 1,320 764
t-tests 0.14 0.00
0.00
F-tests 0.86 0.61
0.50
Treatments
1 ambient temperature (~25C), quartered, soil tested at the "as received" moisture
content
2 refrigerator (4C), brought to room temperature over several hr, quartered, soil
was tested at "as received" moisture content
3 ambient temperature (~25C), dried at 60C for ~ 4 hr, sieved through a No. 10
mesh, and mechanically split
263
Table B-23 Sebring Sand Treatment Study Results for Chloride
Test Date 10/19/2014 10/19/2014
Treatment 1 2
Replicate 1 BD BD
Replicate 2 BD BD
Replicate 3 BD BD
Results for Chloride Check Standard, ppm
30 ppm CS 30
Treatments
1 ambient temperature (23-25C), oven dried at 60C, extracted with 3:1 soil:water,
Notes
1 Analysis was with Hach chloride test kit, 5 to 400 mg/L
2 concentration was at or below the minimum detection level of 5 ppm; less if
264
Table B-24 Sebring Sand Treatment Study Results for Sulfate
Test Date 10/19/2014 10/19/2014
Treatment 1 2
Replicate 1 12 21
Replicate 2 12 24
Replicate 3 12 15
Average 12 20
St Dev 0 4
t-test 0.04
F-test 0.00
Treatments
Results for Sulfate Check Standard, ppm
30 ppm CS 34
Notes
Analysis was with photometer test kit, 2 to 70 mg/L, on range 1 (factory default);
1
reagent lot was 4120
Except for two extracts with photometer readings of 3 ppm (9 ppm in soil),
2
photometer readings were 2 ppm or less and were below the method detection level
265
B.7 Mine 7 Clermont Sand
Table B-25 Clermont Sand Treatment Study Results for pH

Test Date 10/29/2014 11/1/2014 11/1/2014 11/1/2014 11/1/2014 11/2/2014
Soil Slurry T, C 28.5 23.5 23.4 23.3 23.0 23.4
Replicate 1 7.55 7.30 7.44 7.54 7.49 7.69
Replicate 2 7.57 7.28 7.80 7.39 7.46 7.93
Replicate 3 7.56 7.48 7.45 7.39 7.65 7.69
Average 7.56 7.35 7.56 7.44 7.53 7.77
St Dev 0.01 0.11 0.21 0.09 0.10 0.14
t-tests 0.03 0.98 0.08 0.68 0.06
0.19 0.34 0.11 0.02
0.39 0.83 0.22
0.29 0.02
0.08
F-tests 0.02 0.00 0.03 0.02 0.01
0.45 0.76 0.92 0.77
0.30 0.40 0.63
0.84 0.56
0.70
slope, % 97.7 . . 95.0 96.9
offset, mV 1.2 . . 2.4 7.7
buffer, C 28.0 . . 24.5 23.3
slope, % . 99.7 99.7 99.6 .
offset, mV . 2.0 2.0 1.70 .
buffer, C . 23.5 23.5 23.7 .
Treatments (Note: 0.1 g KCl was added to 100 g soil)
1 On-site in the shade within 2 hr of sampling, soil "as received" moisture content
2
temperature (~24C), quartered, soil tested at the "as received" moisture content
3 Sample was transported to the lab in a cooler (~8-19C) and stored in the refrigerator (4C),
brought to room temperature over several hr, quartered, soil was tested at "as received"
4 temperature (~24C), dried at 60C for ~4 hr, sieved through a No. 10 mesh, and
mechanically split
Sample was transported to the lab at ambient temperature (~28 C) and stored at ambient
5
temperature (~24C), quartered, soil tested at the "as received" moisture content
Sample was transported to the lab in a cooler (~8-19C) and stored in the refrigerator (4C),
6 brought to room temperature over several hr, quartered, soil was tested at "as received"
moisture content
266
Table B-26 Clermont Sand Treatment Study Results for Minimum Resistivity
Test Date 11/1/2014 11/1/2014 11/3/2014
Treatment 1 2 3
DI water, ohm-cm >1,000,000 . 930,000
Conductivity Std,
4,100 . 4,200
4,000 ohm-cm
Soil Slurry T, C 21.3 22.4 22.4
Replicate 1 33,000 24,000 26,500
Replicate 2 33,000 25,500 27,500
Replicate 3 32,000 24,000 28,000
Average 32,700 24,500 27,300
St Dev 577 866 764
t-tests 0.00 0.00
0.01
F-tests 0.62 0.73
0.88
Treatments
1 ambient temperature (~24C), quartered, soil tested at the "as received"
moisture content
2 refrigerator (4C), brought to room temperature over several hr, quartered, soil
was tested at "as received" moisture content
3 ambient temperature (~24C), dried at 60C for ~4 hr, sieved through a No. 10
mesh, and mechanically split
267
Table B-27 Clermont Sand Treatment Study Results for Chloride
Test Date 11/3/2014 11/3/2014 11/3/2014 11/3/2014
Treatment 1 2 3 4
Replicate 1 BD BD BD BD
Average Not calculated Not calculated Not calculated Not calculated
St Dev Not calculated Not calculated Not calculated Not calculated
Chloride concentrations were below method detection (BD) level
Treatments
filtered with Whatman 4, filtered with 0.45 micron membrane filter
Sample was transported to the lab in a cooler (8-19C) and stored in a refrigerator
3 (4C), oven dried at 60C, extracted with 3:1 soil:water, filtered with Whatman 4,
filtered with 0.45 micron membrane filter
Sample was transported to the lab in a cooler 8-19C) and stored in a refrigerator
Results of Chloride Check Standard, ppm
30 ppm CS 30
Notes
1 Analysis was with chloride test kit, 5 to 400 mg/L
2 concentration was at or below the minimum detection level of 5 ppm; less if corrected
for the blank
268
Table B-28 Clermont Sand Treatment Study Results for Sulfate
Test Date 11/3/2014 11/3/2014 11/3/2014 11/3/2014
Treatment 1 2 3 4
Average Not calculated Not calculated Not calculated Not calculated
St Dev Not calculated Not calculated Not calculated Not calculated
Sulfate concentrations were below method detection (BD) level
Treatments
Sample was transported to the lab in a cooler (8-19C) and stored in a refrigerator
Sample was transported to the lab in a cooler 8-19C) and stored in a refrigerator
Results of Sulfate Check Standard, ppm
30 ppm CS 36
Notes
Analysis was with photometer test kit, 2 to 70 mg/L, on range 1 (factory default);
1
reagent lot was 4120
Except for two extracts with photometer readings of 3 ppm (9 ppm in soil),
2
photometer readings were 2 ppm or less and were below the method detection level
269
B.8 Mine 8 Alico Road Sand
Table B-29 Alico Road Sand Treatment Study Results for pH

Test Date 11/12/2014 11/15/2014 11/15/2014 11/15/2014 11/15/2014 11/16/2014
Slurry T, C 22.4 23.7 23.7 23.4 23.4 24.4
Replicate 1 8.93 9.02 8.96 8.91 9.04 8.95
Replicate 2 8.98 9.06 9.04 8.98 9.06 9.02
Replicate 3 9.05 8.99 8.98 8.88 9.01 8.98
Average 8.99 9.02 8.99 8.92 9.04 8.98
Std Dev 0.06 0.04 0.04 0.05 0.03 0.04
t-tests 0.41 0.88 0.24 0.26 0.94
0.39 0.05 0.62 0.24
0.14 0.20 0.77
0.03 0.17
0.10
F-tests No F-tests were significant
slope, % 95.0 . . . 97.9 98.6
offset, mV 0 . . . 3.5 3.5
buffer, C 22.7 . . . 23.4 23.4
slope, % . 99.1 99.1 99.5 . .
offset, mV . 1.9 1.9 3.50 . .
buffer, C . 24.1 24.1 24.4 . .
Treatments (Note: 0.1 g KCl added per 100 g sample)
1 On-site in the shade within 2 hr of sampling, soil "as received"
2
ambient temperature (~24C), quartered, soil tested "as received"
3
4 ambient temperature (~24C), dried at 60C for ~4 hr, and sieved through a No. 10
mesh
5
ambient temperature (~24C), quartered, soil tested "as received"
6
270
Table B-30 Alico Road Sand Treatment Study Results for Minimum Resistivity
Test Date 11/15/2014 11/15/2014 11/16/2014
Treatment 1 2 3
DI water, ohm-cm >1,000,000 . >1,000,000
Conductivity Std, 4,000
ohm-cm 4,200 . .
Soil Slurry T, C 22.4 23.6 23.8
Replicate 1 11,000 11,500 11,000
Replicate 2 11,000 12,000 11,000
Replicate 3 11,000 11,500 10,500
Avg 11,000 11,700 10,800
St Dev 0 289 289
t-tests 0.02 0.37
0.02
F-tests 0.00 0.00
1.00
Treatments
Sample was transported to the lab at ambient temperature (~29C) and
1 stored at ambient temperature (~24C), quartered, soil tested at the "as
received" moisture content
Sample was transported to the lab in a cooler (~19-20C) and stored in
2 the refrigerator (4C), brought to room temperature over several hr,
quartered, soil was tested at "as received" moisture content
Sample was transported to the lab at ambient temperature (~29C) and
3 stored at ambient temperature (~24C), dried at 60C for 4 hr, and sieved
through a No. 10 mesh
271
Table B-31 Alico Road Sand Treatment Study Results for Chloride
Test Date 11/18/2014 11/18/2014
Treatment 1 2
Replicate 1 BD BD
Replicate 2 15 BD
Replicate 3 BD BD
30 ppm CS 35 .
Treatments
Treatment 1 ambient temperature (23-25C), oven dried at 60C, extracted with 3:1 soil:water,
Notes
Note 2 concentration was at or below the minimum detection level of 5 ppm; less if corrected
for the blank
272
Table B-32 Alico Road Sand Treatment Study Results for Sulfate
Test Date 11/18/2014 11/18/2014
Treatment 1 2
Replicate 1 BD BD
Replicate 2 BD BD
Replicate 3 BD BD
Treatments
Results for sulfate check standard, ppm
30 ppm CS 33
Notes
Note 1
default)
Except for two extracts with photometer readings of 3 ppm (9 ppm as soil sulfate),
Note 2 photometer sulfate readings were 2 ppm or less (6 ppm as soil sulfate) and were
below the method detection level
273
B.9 ANOVA All Sands
B.9.1 ANOVA for pH—all sands
Table B-33 Input Data for pH ANOVA

Mine 1 2 3 4 5 6 7 8
Replicate 1 4.70 5.29 8.42 4.47 4.54 4.35 7.54 8.95
Replicate 2 4.68 5.00 8.44 4.59 4.56 4.39 7.39 9.02
Replicate 3 4.56 5.18 8.36 4.50 4.58 4.43 7.39 8.98
Note: For each soil, treatment selected for table was matched with treatment for replicate study samples.
Table B-34 Treatment Study ANOVA for pH

ANOVA: Single Factor for pH
SUMMARY
Groups Count Sum Average Variance
Mine 1 3 13.94 4.65 0.0057
Mine 2 3 15.47 5.16 0.0214
Mine 3 3 25.22 8.41 0.0017
Mine 4 3 13.56 4.52 0.0039
Mine 5 3 13.68 4.56 0.0004
Mine 6 3 13.17 4.39 0.0016
Mine 7 3 22.32 7.44 0.0075
Mine 8 3 26.95 8.98 0.0012
ANOVA
Source of Variation SS df MS F P-value F crit
Between Groups 78.49 7 11.21 2,060 2.01E-22 2.66
Within Groups 0.09 16 0.01
Total 78.58 23
274
B.9.2 ANOVA for minimum resistivity—mines 5-8
Table B-35 ANOVA Input Data for Minimum Resistivity (ohm-cm)

Mine 5 Mine 6 Mine 7 Mine 8
17,500 32,000 26,500 11,000
17,500 30,500 27,500 11,000
17,000 31,500 28,000 10,500
Note: For each soil, treatment selected for table was matched with treatment for replicate study samples.
Table B-36 Treatment Study ANOVA for Minimum Resistivity (ohm-cm)

ANOVA: Single Factor for Minimum Resistivity
SUMMARY
Mine 5 3 52,000 17,300 83,300
Mine 6 3 94,000 31,300 583,000
Mine 7 3 82,000 27,300 583,000
Mine 8 3 32,500 10,800 83,300
ANOVA
Between Groups 7.85E+08 3 2.62E+08 785 3.23E-10 4.07
Within Groups 2,670,000 8 333,300
Total 7.88E+08 11
275
B.9.3 ANOVA for sulfate—mines 1, 3, 5, and 6
Table B-37 ANOVA Input Data for Sulfate (ppm)

Replicate Mine 1 Mine 3 Mine 5 Mine 6
1 12 27 24 12
2 9 36 27 12
3 9 36 21 12
Table B-38 Treatment Study ANOVA for Sulfate (ppm)

ANOVA: Single Factor for Sulfate
SUMMARY
Mine 1 3 30 10 3
Mine 3 3 99 33 27
Mine 5 3 72 24 9
Mine 6 3 36 12 0
ANOVA
Between Groups 1050 3 349 36 5.54E-05 4.07
Within Groups 78 8 10
Total 1120 11
276
Appendix C Inter-Laboratory Study Plan and Data Sheet
C.1 First Inter-Laboratory Study
277
FDOT Inter-Laboratory Study Contact:
Noreen Poor, Ph. D., P. E.

Associate Research Professor
(813) 956-0855 Cell
npoor@usf.edu
Hr: M-F 9:00 a.m. - 5:30 p.m.
Study Start Date: Tuesday, January 6, 2015

Study End Date: Thursday, January 8, 2015
278
Participant Instructions
1. Background
Florida Department of Transportation (FDOT) has prepared revisions to the Florida

Method of Tests (FMs) for Determining pH in Soil and Water (FM 5-550), Resistivity in
Soil and Water (FM 5-551), Chloride in Soil and Water (FM 5-552), and Sulfate in Soil
and Water (FM 5-553); revised methods are provided in Attachments A, B, C, and D,
respectively. These revisions update and expand the methods’ procedural details and
include significant changes to method equipment, sample processing, and data
management.
2. Scope and Purpose of Study
The purpose of this study is to estimate a multi-laboratory precision (reproducibility) for

the revised methods and to the extent practicable is designed in accordance with ASTM
C802.1
Over a three-day period, each participating laboratory will analyze in two select backfill
(FDOT 092L) materials for pH, resistivity, chloride, and sulfate per the revised FMs.
Results for each material will be aggregated to produce an estimates of within-laboratory
and between-laboratory variances and, consistent with ASTM C670,2 a precision
statement to accompany each of the revised methods.
3. Revised Methods
Results of field testing and laboratory analyses suggested that the accuracy and
precision of the Florida Methods (FMs) for pH, resistivity, chloride, and sulfate could be
improved with relatively small changes in method procedures. Changes to the FMs
belong in one of five categories: (1) minor edits to method grammar, organization, and
units of measure; (2) updates to the apparatus and reagents sections, with a hazardous
materials note as appropriate; (3) step-by-step instructions to facilitate method oversight;
(4) one or more quality assurance/quality control procedures; and (5) a new section for
method precision and bias, to be completed as part of this inter-laboratory study.
Materials provided have been dried at 60C, sieved through a No. 10 (2 mm) mesh, and
mechanically split into ~1,500-g samples. Chloride and sulfate has been added by
mechanical mixer to one of the two materials to assure that these quantities are above
the method detection levels.
a. FM for the Determination of pH in Soil and Water
279
Revisions were made to the scope, apparatus, reagent, samples, sample
preparation, and test procedure sections of the FM. Refer to Attachment A for the
revised method. Based on ruggedness and other factor analysis studies as well
as laboratory visits, major sources of variability in the FM were
 Electrode condition, and
A pH meter with automatic temperature compensation (ATC) measures the

sample temperature and corrects the slope of the calibration curve for the
measurement temperature. Ideally, the buffers and the samples are at 25C, but
with ATC, buffers and samples need only to be at or near the same temperature.
This is especially important if ambient temperatures are more than a few C from
standard temperature. The revised method requires a pH/mV meter with ATC
and equipped with a 3-in-1 combination electrode.
The pH meter combination electrode is a delicate and sensitive sensor and with
routine treatment and use its lifetime is ~18 months. A 3-in-1 combination
electrode typically encases a glass-bulb indicating electrode, a silver/silver
chloride reference electrode, and a thermocouple. Attention to electrode
performance is critical to a precise and accurate pH measurement. Over time
and with use the reference electrode is consumed, but temporary or permanent
damage to the electrode can be caused by contamination or depletion of the
filling solution, scratches or cracks on the glass bulb surface, and clogging of the
liquid junction, as examples. Electrode condition can be monitored through
calibration and performance checks, which are included in the revised FM.
For pH measurement in low-ionic strength (low salt content) a stable pH reading

can take 5 minutes or more to obtain, as the reading is affected by the exchange
of carbon dioxide between the sample and the atmosphere, the effects of stirring
and vibrations, and a mismatch between the ionic strength of the sample and the
reference solution in the electrode.
b. FM of Test for Resistivity in Soil and Water

preparation, and test procedure sections of the FM. Refer to Attachment B for the
revised method. Based on ruggedness and other factor analysis studies as well
as laboratory visits, major sources of variability in the FM were
280
 Measurement temperature, and
 Water content of soil slurry placed in the soil box during resistivity testing.
Resistivity measurements are sensitive to temperature. The revised FM requires

that samples be brought to room temperature and that the soil slurry temperature
be measured and recorded.
ASTM G1873 and AASHTO T2884 represent two different positions along the soil
resistivity continuum: ASTM G187 measures soil resistivity when the soil is just
saturated and AASHTO T288 measures minimum soil resistivity, where the water
from soil slurry is decanted into the soil box and tested as the water content in
the slurry increases. The revised FM adopts the approach of AASHTO T288 for
obtaining a minimum resistivity.
ASTM D11935 Type 4 reagent water, with a lower resistivity bound of 200,000
ohm-cm is recommended for resistivity measurements. The revised FM
recommends as a best practice a daily check of dilution water resistivity because
on-site water treatment systems can experience break-through of ions and
bottled water can over time absorb ions from acidic and basic gases in the
atmosphere. A higher ion concentration in dilution water translates into a lower
resistivity.
The revised method also recommends that the resistivity meter calibration be
checked at least once per quarter with a National Institute of Standards and
Technology (NIST)-traceable conductivity standard.
c. FM of Test for Chloride in Soil and Water
Revisions were made to the scope, apparatus, reagents, samples, sample

preparation, and test procedure sections of the FM. Refer to Attachment C for a
revised method. The FM offers a choice of one of two methods: (1) a Hach
chloride test kit (or equivalent) as a laboratory-based screening approach, or (2)
for a greater degree of accuracy and precision an analytical approach as
described by the Standard Method for the Evaluation of Water and Wastewater
4500-Cl- or 4110 B.6 Based on ruggedness and other factor analysis studies as
well as laboratory visits major sources of variability in the FM were:
 Soil mass,
 Reagent quality, and
281
 Blank correction.
For chloride determination an accurate soil mass can be obtained if the soil is
dried to a constant mass and its mass weighed on an analytical balance; these
two requirements limit the portability of the method. Residual suspended solids
or color, if present, can be removed by vacuum filtration of the extract through a
0.45-micron filter as a final treatment step. For some soils, ferric sulfate addition
may be required to settle suspended particles prior to filtration. A separate 3:1
water-to-soil extract is prepared for chloride and sulfate, so that if needed ferric
sulfate can be added to the unfiltered extract. A chloride check standard and an
associated reagent blank help to establish the sample color at the titration
endpoint, to guard against bad reagents and to subtract out any chloride
contamination in the glassware and dilution water.
d. FM of Test for Sulfate in Soil and Water
Revisions were made to the scope, apparatus, reagents, samples, sample

preparation, and test procedure sections of the FM. Refer to Attachment D for a
revised method. The FM offers a choice of one of two methods: (1) a Hach
sulfate photometer test kit (or equivalent) as a laboratory-based screening
approach, or (2) for a greater degree of accuracy and precision an analytical
approach as described by the Standard Method for the Evaluation of Water and
Wastewater 4500-SO42- or 4110 B.6 Based on ruggedness and other factor
analysis studies as well as laboratory visits major sources of variability in the FM
were:
 Soil mass,
 Reagent quality, and
 Sulfate calibration curve.
For sulfate determination an accurate soil mass can be obtained if the soil is
dried to a constant mass and its mass weighed on an analytical balance; these
two requirements limit the portability of the method. Residual suspended solids
or color, if present, can be removed by vacuum filtration of the extract through a
0.45-micron filter as a final treatment step. For some soils, acid addition may be
required to settle suspended particles prior to filtration. A separate 3:1 water-to-
soil extract is prepared for chloride and sulfate, so that if needed hydrochloric
acid can be added to the unfiltered extract. A photometer with factory-
programmed sulfate calibration curves give reasonable results provided the
curve is periodically checked, as the potency of the reagents used in the method
varies by lot number and possibly with the age of the reagents. A sulfate check
282
standard and associated reagent blank help to guard against bad reagents,
contamination of dilution water, or a mismatch between reagent and photometer
response.
4. Sample management
Each participating laboratory will receive a 1,500-g sample of air-dried and sieved soil of
each material, which is sufficient soil for each FM to be run once. Samples will be
delivered to participating laboratories with one month of the study start date.
5. Data management
Data should be handled and reported as required by the revised FMs. Measurements
that are below detection should be flagged as such. Upon completion of testing, scan
and submit the attached data sheet (Attachment E) electronically to the study contact.
Keep data records on file for at least one year past the study date.
References
1. ASTM C802-2009a Standard Practice for Conducting an Inter-Laboratory Test Program

to Determine the Precision of Test Methods for Construction Materials. ASTM
International, West Conshohocken, Pennsylvania. 18 pp.
2. ASTM C670-2003 Standard Practice for Preparing Precision and Bias Statements for
Test Methods for Construction Materials, ASTM International, West Conshohocken,
Pennsylvania. 9 pp.
3. ASTM G187-2012a. Standard Test Method for Measurement of Soil Resistivity Using the
Two-Electrode Soil Box Method G 187-2012a. ASTM International, West
Conshohocken, Pennsylvania. 6 pp.
4. AASHTO T288-2012. Standard Method of Test for Determining Minimum Laboratory Soil
Resistivity. American Association of State Highway and Transportation Officials,
Washington, D. C. 7 pp.
5. ASTM D1193-2006. Standard Specification for Reagent Water. ASTM International,

West Conshohocken, Pennsylvania. 5 pp.
6. Rice, E. W., Baird, R. B., Easton, A. D., Clesceri, L. S., Editors, 2012. Standard Methods
for the Examination of Water and Wastewater, 22nd Edition, American Public Health
Association (APHA), American Water Works Association (AWWA), and Water
Environment Federation (WEF), APHA Publication Office, Washington, D. C.
Attachment A Revised FM5-550

Attachment B Revised FM5-551
283
Attachment C Revised FM5-552
Attachment D Revised FM5-553
Attachment E Inter-Laboratory Study Data Sheet
Data Sheet
Select Backfill (092L)
Complete and submit data sheet by e-mail to npoor@usf.edu on or before
January 9, 2015.
Sample pH, pH units Solution Temperature, C

Example pH 7.01 25.0
Material A
Material B
Sample Minimum Resistivity, ohm-cm Solution Temperature, C

Example 19,500 25.0
Material A
Material B
Sample Soil Chloride Concentration, ppm

Example 35
Material A
Material B
Sample Soil Sulfate Concentration, ppm

Example 6 (Below detection)
Material A
Material B
Comments (type or print legibly):
284
C.2 Second Inter-Laboratory Study
285
FDOT Inter-Laboratory Study Contact:
Noreen Poor, Ph. D., P. E.

Associate Research Professor
(813) 956-0855 Cell
npoor@usf.edu
Hr: M-F 9:00 a.m. - 5:30 p.m.
Study Start Date: Wednesday, March 11, 2015

Study End Date: Thursday, March 12, 2015
286
Participant Instructions
1. Background
Florida Department of Transportation (FDOT) has prepared proposed revisions to the

Florida Method of Test (FM) for Determining pH in Soil and Water (FM 5-550) (attached).
These revisions update and expand the method’s procedural details and include
changes to method equipment, sample processing, and data management.
2. Scope and Purpose of Study
The purpose of this study is to estimate a multi-laboratory precision (reproducibility) for

the revised method and to the extent practicable the study is designed in accordance
with ASTM C802.1
Over a two-day period, each participating laboratory will analyze in duplicate two select
backfill (FDOT 092L) materials for pH. These materials are labeled Material C and
Material D. Results for each material will be aggregated to produce an estimate of
within-laboratory and between-laboratory variance and, consistent with ASTM C670,2 a
precision statement to accompany the revised method.
3. Revised Method
Results of field testing and laboratory analyses suggested that the accuracy and
precision of the Florida Method (FM) for pH could be improved with relatively small
changes in method procedures. Changes to the FM belong in one of five categories: (1)
minor edits to method grammar, organization, and units of measure; (2) updates to the
apparatus and reagents sections; (3) step-by-step instructions to facilitate method
oversight; (4) one or more quality assurance/quality control procedures; and (5) a new
section for method precision and bias, to be completed as part of this inter-laboratory
study.
Materials provided have been dried at 60C, sieved through a No. 10 (2 mm) mesh, and
mechanically split into ~200-g (Material D) and ~300-g (Material C) samples.
FM for the Determination of pH in Soil and Water

preparation, and test procedure sections of the FM (see attached method).
287
Based on ruggedness and other factor analysis studies as well as laboratory
visits and an inter-laboratory study, major sources of variability in the FM were
 Electrode condition,
 Electrode memory, and
A pH meter with automatic temperature compensation (ATC) measures the

sample temperature and corrects the slope of the calibration curve for the
measurement temperature. Ideally, the buffers and the samples are at 25C, but
with ATC, buffers and samples need only to be at or near the same temperature.
This is especially important if ambient temperatures are more than a few C from
standard temperature. The revised method requires a pH/mV meter with ATC
and equipped with a 3-in-1 combination electrode.
The pH meter combination electrode is a delicate and sensitive sensor and with
routine treatment and use its lifetime is ~18 months. A 3-in-1 combination
electrode typically encases a glass-bulb indicating electrode, a silver/silver
chloride reference electrode, and a thermocouple. Attention to electrode
performance is critical to a precise and accurate pH measurement. Over time
and with use the reference electrode is consumed, but temporary or permanent
damage to the electrode can be caused by contamination or depletion of the
filling solution, scratches or cracks on the glass bulb surface, and clogging of the
liquid junction, as examples. Electrode condition can be monitored through
calibration and performance checks, which are included in the revised FM.
The electrode has a memory of its previous measurement and the contribution of
this memory on the current pH measurement is more pronounced for low-ionic
strength (low salt content) solutions. The memory effect on sample pH can be
reduced by careful attention to rinsing and drying the electrode between
measurements and by a second or even third consecutive measurement of
sample pH.
For pH measurement in low-ionic strength solutions a stable pH reading can take

5 minutes or more to obtain, as the reading is affected by the exchange of carbon
dioxide between the sample and the atmosphere, the effects of stirring and
vibrations, and a mismatch between the ionic strength of the sample and the
reference solution in the electrode. If after ~ 5 min the pH reading does not
stabilize in a gently stirred sample, obtain a pH reading in the unstirred sample.
288
4. Sample management
Each participating laboratory will receive by mail one week ahead of the start
date sufficient air-dried and sieved soil to test each material in duplicate for pH.
The anticipated effort is ~1.5 person-hr. Treat each duplicate as an independent
sample and run samples in a random order. Fill in Table 1 with the random order
of sample analysis.
Table 1 Random Order of Sample Analysis

Sample Sample Number Order Analyzed
Material C, Replicate 1 1
Material C, Replicate 2 2
Material D, Replicate 1 3
Material D, Replicate 2 4
Visit the web site www.random.org (Figure 1). In the box labeled “True Random
Number Generator” and enter 1 for Min and 4 for Max, then select the button for
Generate. A result will appear. The first result to appear is the first sample to be
analyzed; for example, if the first result is “1”, analyzed sample 1 first. Select the
button for Generate again. If the second result is “4”, analyze sample 4 second,
and so forth. Ignore a repeated result.
289
Figure 1. Random number generator.
5. Data management
Data should be handled and reported as required by the revised FM. Upon
completion of testing, submit the attached data sheet electronically to the study
contact. Keep data records on file for at least one year past the study date.
References
1. ASTM C802-2009a Standard Practice for Conducting an Inter-Laboratory Test

Program to Determine the Precision of Test Methods for Construction Materials.
ASTM International, West Conshohocken, Pennsylvania. 18 pp.
2. ASTM C670-2003 Standard Practice for Preparing Precision and Bias

Statements for Test Methods for Construction Materials, ASTM International,
West Conshohocken, Pennsylvania. 9 pp.
290
Data Sheet
Select Backfill (092L)
Complete and submit data sheet by e-mail to npoor@usf.edu on or before
March 12, 2015.
Assigned Laboratory Number _______

Sample pH, pH units Sample Temperature, C
Example pH 7.01 25.0
Material C, Replicate 1
Material C, Replicate 2
Material D, Replicate 1
Material D, Replicate 2
Comments (type or print legibly):
291
Appendix D Analysis of Variance (ANOVA) for Replicate Study
D.1 ANOVA for pH
Table D-1 ANOVA for Replicate Study pH

ANOVA: Single Factor for pH
SUMMARY
Mine 1 12 55.92 4.66 0.0076
Mine 2 12 64.42 5.37 0.0486
Mine 3 12 97.85 8.15 0.0418
Mine 4 12 58.61 4.88 0.0384
Mine 5 12 55.74 4.65 0.0416
Mine 6 12 56.39 4.70 0.4013
Mine 7 12 88.80 7.40 0.0060
Mine 8 12 106.41 8.87 0.0006
ANOVA
Between Groups 261 7 37.3 509 2.66E-68 2.11
Within Groups 6.45 88 0.07
Total 267.25 95
292
D.2 ANOVA for Minimum Resistivity
Table D-2 Replicate Study ANOVA for Minimum Resistivity (ohm-cm)

ANOVA: Single Factor for Minimum Resistivity
SUMMARY
Mine 5 12 229,000 19,100 39,600,000
Mine 6 12 437,000 36,400 1.05E+08
Mine 7 12 400,000 33,300 13,000,000
Mine 8 12 151,000 12,500 23,400,000
ANOVA
Between Groups 4.66E+09 3 1.55E+09 38.0 1.87E-12 2.82
Within Groups 1.75E+09 44 39,900,000
Total 6.42E+09 47
293
D.3 ANOVA for Sulfate
Table D-3 Replicate Study ANOVA for Sulfate (ppm)

Anova: Single Factor for Sulfate
SUMMARY
Mine 1 12 153 13 39.5
Mine 3 12 405 34 917
Mine 5 12 288 24 16.4
Mine 6 12 117 10 14.9
ANOVA
Between Groups 4,350 3 1,450 5.88 0.00183 2.82
Within Groups 10,900 44 247
Total 15,200 47
294
Appendix E PHREEQ Input and Output Files
E.1 Input File
SOLUTION 0 Rainwater #VW annual averages from NADP FL41 2000-2012
units ppm
pH 4.9 charge
temp 25.0
Ca 0.10
Mg 0.04
Na 0.3
K 0.02
Cl 0.68 #adjusted to keep pH at 4.9 in equilibrium with CO2
S(+6) 0.7
SAVE SOLUTION 0
END
SOLUTION 1-60 Groundwater #from Shiffer, 1989

units ppm
pH 5
temp 25.0
Ca 0.4
Mg 0.5
Na 2.1
K 0.1
Cl 4.0
S(+6) 0.6
Alkalinity 0.82 as HCO3
SAVE SOLUTION 1 60
END
USE SOLUTION 0
EQUILIBRIUM_PHASES 0
CO2(g) -3.5
QUARTZ 0.0
SAVE SOLUTION 0
END
USE SOLUTION 1 60
EQUILIBRIUM_PHASES 1-60
CO2(g) -1.9
QUARTZ 0.0
SAVE SOLUTION 1 60
END
USE SOLUTION 0 60
EXCHANGE 1-60
-equilibrate 1
X 250 #from Elrashidi et al., 2001
COPY cell 1 60
ADVECTION
-cells 60
-shifts 75 #15 shifts per year
-time_step 0.06667 yr
-initial_time 0
-punch_cells 60
-punch_frequency 1
-print_cells 60
295
-print_frequency 15
SELECTED_OUTPUT
-file exRainX.sel
-reset false
-step
-pH
-totals K Ca Na
USER_PUNCH
-heading Time
10 PUNCH (STEP_NO)/15
END
E.2 Output File

(Partial printing of output file)
Input file: C:\Program Files (x86)\USGS\Phreeqc Interactive 3.0.6-

7757\examples\Rainwater and Groundwater Exchange 2.pqi
Output file: C:\Program Files (x86)\USGS\Phreeqc Interactive 3.0.6-
7757\examples\Rainwater and Groundwater Exchange 2.pqo
Database file: C:\Program Files (x86)\USGS\Phreeqc Interactive 3.0.6-
7757\database\phreeqc.dat
------------------
Reading data base.
------------------
SOLUTION_MASTER_SPECIES
SOLUTION_SPECIES
PHASES
EXCHANGE_MASTER_SPECIES
EXCHANGE_SPECIES
SURFACE_MASTER_SPECIES
SURFACE_SPECIES
RATES
END
------------------------------------
Reading input data for simulation 1.
------------------------------------
DATABASE C:\Program Files (x86)\USGS\Phreeqc Interactive 3.0.6-

7757\database\phreeqc.dat
TITLE Rainwater and Groundwater - Exchange
SOLUTION 0 Rainwater #VW annual averages from NADP FL41 2000-2012
units ppm
pH 4.9 charge
temp 25.0
Ca 0.10
Mg 0.04
Na 0.3
K 0.02
Cl 0.68 #adjusted to keep pH at 4.9 in equilibrium with CO2
S(+6) 0.7
SAVE SOLUTION 0
END
-----
TITLE
-----
296
Rainwater and Groundwater - Exchange
-------------------------------------------
Beginning of initial solution calculations.
-------------------------------------------
Initial solution 0. Rainwater
-----------------------------Solution composition------------------------------
Elements Molality Moles
Ca 2.495e-006 2.495e-006
Cl 1.918e-005 1.918e-005
K 5.115e-007 5.115e-007
Mg 1.645e-006 1.645e-006
Na 1.305e-005 1.305e-005
S(6) 7.287e-006 7.287e-006
----------------------------Description of solution----------------------------
pH = 4.928 Charge balance

pe = 4.000
Specific Conductance (uS/cm, 25 oC) = 7
Density (g/cm3) = 0.99704
Volume (L) = 1.00297
Activity of water = 1.000
Ionic strength = 4.514e-005
Mass of water (kg) = 1.000e+000
Total alkalinity (eq/kg) = -1.191e-005
Total carbon (mol/kg) = 0.000e+000
Total CO2 (mol/kg) = 0.000e+000
Temperature (deg C) = 25.00
Electrical balance (eq) = 8.272e-018
Percent error, 100*(Cat-|An|)/(Cat+|An|) = 0.00
Iterations = 3
Total H = 1.110124e+002
Total O = 5.550625e+001
----------------------------Distribution of species----------------------------
Log Log Log mole V

Species Molality Activity Molality Activity Gamma cm3/mol
H+ 1.191e-005 1.181e-005 -4.924 -4.928 -0.003 0.00

OH- 8.635e-010 8.568e-010 -9.064 -9.067 -0.003 -4.13
H2O 5.551e+001 1.000e+000 1.744 -0.000 0.000 18.07
Ca 2.495e-006
Ca+2 2.492e-006 2.415e-006 -5.603 -5.617 -0.014 -18.23
CaSO4 3.028e-009 3.028e-009 -8.519 -8.519 0.000 7.50
CaHSO4+ 2.370e-013 2.351e-013 -12.625 -12.629 -0.003 (0)
CaOH+ 3.420e-014 3.393e-014 -13.466 -13.469 -0.003 (0)
Cl 1.918e-005
Cl- 1.918e-005 1.903e-005 -4.717 -4.721 -0.003 18.05
H(0) 1.976e-021
H2 9.880e-022 9.880e-022 -21.005 -21.005 0.000 28.61
K 5.115e-007
K+ 5.115e-007 5.075e-007 -6.291 -6.295 -0.003 8.99
297
KSO4- 2.533e-011 2.513e-011 -10.596 -10.600 -0.003 (0)
Mg 1.645e-006
Mg+2 1.643e-006 1.592e-006 -5.784 -5.798 -0.014 -21.91
MgSO4 2.631e-009 2.631e-009 -8.580 -8.580 0.000 5.84
MgOH+ 4.932e-013 4.894e-013 -12.307 -12.310 -0.003 (0)
Na 1.305e-005
Na+ 1.305e-005 1.295e-005 -4.884 -4.888 -0.003 -1.41
NaSO4- 4.610e-010 4.574e-010 -9.336 -9.340 -0.003 18.39
NaOH 1.109e-024 1.109e-024 -23.955 -23.955 0.000 (0)
O(0) 0.000e+000
O2 0.000e+000 0.000e+000 -50.370 -50.370 0.000 30.40
S(6) 7.287e-006
SO4-2 7.273e-006 7.049e-006 -5.138 -5.152 -0.014 13.92
HSO4- 8.160e-009 8.096e-009 -8.088 -8.092 -0.003 40.25
CaSO4 3.028e-009 3.028e-009 -8.519 -8.519 0.000 7.50
MgSO4 2.631e-009 2.631e-009 -8.580 -8.580 0.000 5.84
NaSO4- 4.610e-010 4.574e-010 -9.336 -9.340 -0.003 18.39
KSO4- 2.533e-011 2.513e-011 -10.596 -10.600 -0.003 (0)
CaHSO4+ 2.370e-013 2.351e-013 -12.625 -12.629 -0.003 (0)
------------------------------Saturation indices-------------------------------
Phase SI log IAP log K(298 K, 1 atm)
Anhydrite -6.49 -10.77 -4.28 CaSO4

Gypsum -6.19 -10.77 -4.58 CaSO4:2H2O
H2(g) -17.90 -21.01 -3.10 H2
H2O(g) -1.50 -0.00 1.50 H2O
Halite -11.18 -9.61 1.57 NaCl
O2(g) -47.48 -50.37 -2.89 O2
Sylvite -11.91 -11.02 0.90 KCl
------------------
End of simulation.
------------------
------------------------------------
------------------------------------
SOLUTION 1-60 Groundwater #from Shiffer, 1989

units ppm
pH 5
temp 25.0
Ca 0.4
Mg 0.5
Na 2.1
K 0.1
Cl 4.0
S(+6) 0.6
Alkalinity 0.82 as HCO3
SAVE SOLUTION 1 60
END
-------------------------------------------
Beginning of initial solution calculations.
-------------------------------------------
Initial solution 1. Groundwater
298
Alkalinity 1.344e-005 1.344e-005

Ca 9.980e-006 9.980e-006
Cl 1.128e-004 1.128e-004
K 2.557e-006 2.557e-006
Mg 2.057e-005 2.057e-005
Na 9.135e-005 9.135e-005
S(6) 6.246e-006 6.246e-006
pH = 5.000
pe = 4.000
Volume (L) = 1.00301
Total carbon (mol/kg) = 5.456e-004
Total CO2 (mol/kg) = 5.456e-004
Iterations = 6
Total H = 1.110125e+002
Total O = 5.550736e+001
Log Log Log mole V

H+ 1.016e-005 1.000e-005 -4.993 -5.000 -0.007 0.00

OH- 1.029e-009 1.012e-009 -8.988 -8.995 -0.007 -4.13
H2O 5.551e+001 1.000e+000 1.744 -0.000 0.000 18.07
C(4) 5.456e-004
CO2 5.220e-004 5.220e-004 -3.282 -3.282 0.000 29.09
HCO3- 2.359e-005 2.322e-005 -4.627 -4.634 -0.007 24.69
MgHCO3+ 5.319e-009 5.235e-009 -8.274 -8.281 -0.007 5.46
CaHCO3+ 2.815e-009 2.770e-009 -8.551 -8.557 -0.007 9.66
NaHCO3 1.174e-009 1.174e-009 -8.930 -8.930 0.000 2.13
CO3-2 1.161e-010 1.089e-010 -9.935 -9.963 -0.028 -5.58
MgCO3 2.002e-012 2.002e-012 -11.699 -11.699 0.000 -17.09
CaCO3 1.710e-012 1.711e-012 -11.767 -11.767 0.000 -14.60
NaCO3- 1.852e-013 1.823e-013 -12.732 -12.739 -0.007 -0.63
Ca 9.980e-006
Ca+2 9.968e-006 9.351e-006 -5.001 -5.029 -0.028 -18.20
CaSO4 9.673e-009 9.674e-009 -8.014 -8.014 0.000 7.50
CaHCO3+ 2.815e-009 2.770e-009 -8.551 -8.557 -0.007 9.66
CaCO3 1.710e-012 1.711e-012 -11.767 -11.767 0.000 -14.60
CaHSO4+ 6.462e-013 6.359e-013 -12.190 -12.197 -0.007 (0)
CaOH+ 1.577e-013 1.552e-013 -12.802 -12.809 -0.007 (0)
Cl 1.128e-004
Cl- 1.128e-004 1.110e-004 -3.948 -3.955 -0.007 18.06
299
H(0) 1.416e-021
H2 7.079e-022 7.079e-022 -21.150 -21.150 0.000 28.61
K 2.557e-006
K+ 2.557e-006 2.517e-006 -5.592 -5.599 -0.007 8.99
KSO4- 1.045e-010 1.029e-010 -9.981 -9.988 -0.007 (0)
Mg 2.057e-005
Mg+2 2.053e-005 1.927e-005 -4.688 -4.715 -0.028 -21.89
MgSO4 2.627e-008 2.628e-008 -7.580 -7.580 0.000 5.84
MgHCO3+ 5.319e-009 5.235e-009 -8.274 -8.281 -0.007 5.46
MgOH+ 7.108e-012 6.996e-012 -11.148 -11.155 -0.007 (0)
MgCO3 2.002e-012 2.002e-012 -11.699 -11.699 0.000 -17.09
Na 9.135e-005
Na+ 9.134e-005 8.989e-005 -4.039 -4.046 -0.007 -1.41
NaSO4- 2.663e-009 2.621e-009 -8.575 -8.582 -0.007 18.40
NaHCO3 1.174e-009 1.174e-009 -8.930 -8.930 0.000 2.13
NaCO3- 1.852e-013 1.823e-013 -12.732 -12.739 -0.007 -0.63
NaOH 9.098e-024 9.098e-024 -23.041 -23.041 0.000 (0)
O(0) 0.000e+000
O2 0.000e+000 0.000e+000 -50.080 -50.080 0.000 30.40
S(6) 6.246e-006
SO4-2 6.201e-006 5.817e-006 -5.208 -5.235 -0.028 13.95
MgSO4 2.627e-008 2.628e-008 -7.580 -7.580 0.000 5.84
CaSO4 9.673e-009 9.674e-009 -8.014 -8.014 0.000 7.50
HSO4- 5.747e-009 5.656e-009 -8.241 -8.247 -0.007 40.26
NaSO4- 2.663e-009 2.621e-009 -8.575 -8.582 -0.007 18.40
KSO4- 1.045e-010 1.029e-010 -9.981 -9.988 -0.007 (0)
CaHSO4+ 6.462e-013 6.359e-013 -12.190 -12.197 -0.007 (0)
Anhydrite -5.99 -10.26 -4.28 CaSO4

Aragonite -6.66 -14.99 -8.34 CaCO3
Calcite -6.51 -14.99 -8.48 CaCO3
CO2(g) -1.82 -3.28 -1.46 CO2
Dolomite -12.58 -29.67 -17.09 CaMg(CO3)2
Gypsum -5.68 -10.26 -4.58 CaSO4:2H2O
H2(g) -18.05 -21.15 -3.10 H2
H2O(g) -1.50 -0.00 1.50 H2O
Halite -9.57 -8.00 1.57 NaCl
O2(g) -47.19 -50.08 -2.89 O2
Sylvite -10.45 -9.55 0.90 KCl
------------------
End of simulation.
------------------
------------------------------------
------------------------------------
USE SOLUTION 0
EQUILIBRIUM_PHASES 0
CO2(g) -3.5
QUARTZ 0.0
SAVE SOLUTION 0
END
300
-----------------------------------------
Beginning of batch-reaction calculations.
-----------------------------------------
Reaction step 1.
Using solution 0. Rainwater

Using pure phase assemblage 0.
-------------------------------Phase assemblage--------------------------------
Moles in assemblage
Phase SI log IAP log K(T, P) Initial Final Delta
CO2(g) -3.50 -4.96 -1.46 1.000e+001 1.000e+001 -1.133e-005

Quartz -0.00 -3.98 -3.98 1.000e+001 1.000e+001 -1.046e-004
C 1.133e-005 1.133e-005
Ca 2.495e-006 2.495e-006
Cl 1.918e-005 1.918e-005
K 5.115e-007 5.115e-007
Mg 1.645e-006 1.645e-006
Na 1.305e-005 1.305e-005
S 7.287e-006 7.287e-006
Si 1.046e-004 1.046e-004

pe = 0.438 Adjusted to redox
equilibrium
Volume (L) = 1.00297
Iterations = 11
Total H = 1.110124e+002
Total O = 5.550648e+001
Log Log Log mole V

H+ 1.231e-005 1.221e-005 -4.910 -4.913 -0.003 0.00

OH- 8.353e-010 8.288e-010 -9.078 -9.082 -0.003 -4.13
H2O 5.551e+001 1.000e+000 1.744 -0.000 0.000 18.07
C(-4) 4.144e-024
301
CH4 4.144e-024 4.144e-024 -23.383 -23.383 0.000 32.22
C(4) 1.133e-005
CO2 1.093e-005 1.093e-005 -4.961 -4.961 0.000 29.09
HCO3- 4.012e-007 3.981e-007 -6.397 -6.400 -0.003 24.69
CaHCO3+ 1.236e-011 1.227e-011 -10.908 -10.911 -0.003 9.65
MgHCO3+ 7.474e-012 7.415e-012 -11.126 -11.130 -0.003 5.45
NaHCO3 2.898e-012 2.898e-012 -11.538 -11.538 0.000 2.13
CO3-2 1.577e-012 1.529e-012 -11.802 -11.816 -0.014 -5.60
CaCO3 6.202e-015 6.202e-015 -14.207 -14.207 0.000 -14.60
MgCO3 2.322e-015 2.322e-015 -14.634 -14.634 0.000 -17.09
NaCO3- 3.714e-016 3.685e-016 -15.430 -15.434 -0.003 -0.64
Ca 2.495e-006
Ca+2 2.492e-006 2.415e-006 -5.603 -5.617 -0.014 -18.23
CaSO4 3.027e-009 3.027e-009 -8.519 -8.519 0.000 7.50
CaHCO3+ 1.236e-011 1.227e-011 -10.908 -10.911 -0.003 9.65
CaHSO4+ 2.449e-013 2.430e-013 -12.611 -12.614 -0.003 (0)
CaOH+ 3.308e-014 3.282e-014 -13.480 -13.484 -0.003 (0)
CaCO3 6.202e-015 6.202e-015 -14.207 -14.207 0.000 -14.60
Cl 1.918e-005
Cl- 1.918e-005 1.903e-005 -4.717 -4.721 -0.003 18.05
H(0) 2.806e-014
H2 1.403e-014 1.403e-014 -13.853 -13.853 0.000 28.61
K 5.115e-007
K+ 5.115e-007 5.075e-007 -6.291 -6.295 -0.003 8.99
KSO4- 2.533e-011 2.513e-011 -10.596 -10.600 -0.003 (0)
Mg 1.645e-006
Mg+2 1.643e-006 1.592e-006 -5.784 -5.798 -0.014 -21.91
MgSO4 2.630e-009 2.630e-009 -8.580 -8.580 0.000 5.84
MgHCO3+ 7.474e-012 7.415e-012 -11.126 -11.130 -0.003 5.45
MgOH+ 4.770e-013 4.733e-013 -12.321 -12.325 -0.003 (0)
MgCO3 2.322e-015 2.322e-015 -14.634 -14.634 0.000 -17.09
Na 1.305e-005
Na+ 1.305e-005 1.295e-005 -4.884 -4.888 -0.003 -1.41
NaSO4- 4.609e-010 4.573e-010 -9.336 -9.340 -0.003 18.39
NaHCO3 2.898e-012 2.898e-012 -11.538 -11.538 0.000 2.13
NaCO3- 3.714e-016 3.685e-016 -15.430 -15.434 -0.003 -0.64
NaOH 1.073e-024 1.073e-024 -23.969 -23.969 0.000 (0)
O(0) 0.000e+000
O2 0.000e+000 0.000e+000 -64.674 -64.674 0.000 30.40
S(-2) 6.396e-018
H2S 6.337e-018 6.337e-018 -17.198 -17.198 0.000 37.16
HS- 5.981e-020 5.934e-020 -19.223 -19.227 -0.003 20.58
S-2 6.056e-028 5.869e-028 -27.218 -27.231 -0.014 (0)
S(6) 7.287e-006
SO4-2 7.272e-006 7.048e-006 -5.138 -5.152 -0.014 13.92
HSO4- 8.434e-009 8.368e-009 -8.074 -8.077 -0.003 40.25
CaSO4 3.027e-009 3.027e-009 -8.519 -8.519 0.000 7.50
MgSO4 2.630e-009 2.630e-009 -8.580 -8.580 0.000 5.84
NaSO4- 4.609e-010 4.573e-010 -9.336 -9.340 -0.003 18.39
KSO4- 2.533e-011 2.513e-011 -10.596 -10.600 -0.003 (0)
CaHSO4+ 2.449e-013 2.430e-013 -12.611 -12.614 -0.003 (0)
Si 1.046e-004
H4SiO4 1.046e-004 1.046e-004 -3.980 -3.980 0.000 52.08
H3SiO4- 1.273e-009 1.263e-009 -8.895 -8.899 -0.003 27.95
H2SiO4-2 7.259e-018 7.035e-018 -17.139 -17.153 -0.014 (0)
302
Anhydrite -6.49 -10.77 -4.28 CaSO4
Aragonite -9.10 -17.43 -8.34 CaCO3
Calcite -8.95 -17.43 -8.48 CaCO3
CH4(g) -20.54 -23.38 -2.84 CH4
Chalcedony -0.43 -3.98 -3.55 SiO2
Chrysotile -28.08 4.12 32.20 Mg3Si2O5(OH)4
CO2(g) -3.50 -4.96 -1.46 CO2 Pressure 0.0 atm, phi 1.000
Dolomite -17.96 -35.05 -17.09 CaMg(CO3)2
Gypsum -6.19 -10.77 -4.58 CaSO4:2H2O
H2(g) -10.75 -13.85 -3.10 H2
H2O(g) -1.50 -0.00 1.50 H2O
H2S(g) -16.15 -24.14 -7.99 H2S
Halite -11.18 -9.61 1.57 NaCl
O2(g) -61.78 -64.67 -2.89 O2
Quartz -0.00 -3.98 -3.98 SiO2
Sepiolite -19.64 -3.88 15.76 Mg2Si3O7.5OH:3H2O
Sepiolite(d) -22.54 -3.88 18.66 Mg2Si3O7.5OH:3H2O
SiO2(a) -1.27 -3.98 -2.71 SiO2
Sulfur -11.38 -6.50 4.88 S
Sylvite -11.91 -11.02 0.90 KCl
Talc -25.24 -3.84 21.40 Mg3Si4O10(OH)2
------------------
End of simulation.
------------------
------------------------------------
------------------------------------
USE SOLUTION 1 60
EQUILIBRIUM_PHASES 1-60
CO2(g) -1.9
QUARTZ 0.0
SAVE SOLUTION 1 60
END
-----------------------------------------
-----------------------------------------
Reaction step 1.
Using solution 1. Groundwater

Using pure phase assemblage 1.
Moles in assemblage
CO2(g) -1.90 -3.36 -1.46 1.000e+001 1.000e+001 8.817e-005

Quartz 0.00 -3.98 -3.98 1.000e+001 1.000e+001 -1.046e-004
303
C 4.575e-004 4.575e-004
Ca 9.980e-006 9.980e-006
Cl 1.128e-004 1.128e-004
K 2.557e-006 2.557e-006
Mg 2.057e-005 2.057e-005
Na 9.135e-005 9.135e-005
S 6.246e-006 6.246e-006
Si 1.046e-004 1.046e-004

equilibrium
Volume (L) = 1.00301
Total alkalinity (eq/kg) = 1.344e-005
Iterations = 11
Total H = 1.110125e+002
Total O = 5.550739e+001
Log Log Log mole V

H+ 8.931e-006 8.792e-006 -5.049 -5.056 -0.007 0.00

OH- 1.170e-009 1.151e-009 -8.932 -8.939 -0.007 -4.13
H2O 5.551e+001 1.000e+000 1.744 -0.000 0.000 18.07
C(-4) 4.815e-024
CH4 4.815e-024 4.815e-024 -23.317 -23.317 0.000 32.22
C(4) 4.575e-004
CO2 4.351e-004 4.351e-004 -3.361 -3.361 0.000 29.09
HCO3- 2.236e-005 2.201e-005 -4.650 -4.657 -0.007 24.69
MgHCO3+ 5.043e-009 4.963e-009 -8.297 -8.304 -0.007 5.46
CaHCO3+ 2.668e-009 2.626e-009 -8.574 -8.581 -0.007 9.66
NaHCO3 1.113e-009 1.113e-009 -8.954 -8.954 0.000 2.13
CO3-2 1.251e-010 1.174e-010 -9.903 -9.930 -0.028 -5.58
MgCO3 2.159e-012 2.159e-012 -11.666 -11.666 0.000 -17.09
CaCO3 1.845e-012 1.845e-012 -11.734 -11.734 0.000 -14.60
NaCO3- 1.997e-013 1.965e-013 -12.700 -12.707 -0.007 -0.63
Ca 9.980e-006
Ca+2 9.968e-006 9.353e-006 -5.001 -5.029 -0.028 -18.20
CaSO4 9.678e-009 9.679e-009 -8.014 -8.014 0.000 7.50
CaHCO3+ 2.668e-009 2.626e-009 -8.574 -8.581 -0.007 9.66
CaCO3 1.845e-012 1.845e-012 -11.734 -11.734 0.000 -14.60
CaHSO4+ 5.684e-013 5.594e-013 -12.245 -12.252 -0.007 (0)
CaOH+ 1.794e-013 1.765e-013 -12.746 -12.753 -0.007 (0)
Cl 1.128e-004
Cl- 1.128e-004 1.110e-004 -3.948 -3.955 -0.007 18.06
H(0) 1.160e-014
304
H2 5.800e-015 5.800e-015 -14.237 -14.237 0.000 28.61
K 2.557e-006
K+ 2.557e-006 2.517e-006 -5.592 -5.599 -0.007 8.99
KSO4- 1.046e-010 1.029e-010 -9.981 -9.988 -0.007 (0)
Mg 2.057e-005
Mg+2 2.053e-005 1.927e-005 -4.688 -4.715 -0.028 -21.89
MgSO4 2.629e-008 2.629e-008 -7.580 -7.580 0.000 5.84
MgHCO3+ 5.043e-009 4.963e-009 -8.297 -8.304 -0.007 5.46
MgOH+ 8.085e-012 7.958e-012 -11.092 -11.099 -0.007 (0)
MgCO3 2.159e-012 2.159e-012 -11.666 -11.666 0.000 -17.09
Na 9.135e-005
Na+ 9.134e-005 8.990e-005 -4.039 -4.046 -0.007 -1.41
NaSO4- 2.664e-009 2.622e-009 -8.574 -8.581 -0.007 18.40
NaHCO3 1.113e-009 1.113e-009 -8.954 -8.954 0.000 2.13
NaCO3- 1.997e-013 1.965e-013 -12.700 -12.707 -0.007 -0.63
NaOH 1.035e-023 1.035e-023 -22.985 -22.985 0.000 (0)
O(0) 0.000e+000
O2 0.000e+000 0.000e+000 -63.907 -63.907 0.000 30.40
S(-2) 8.020e-020
H2S 7.916e-020 7.916e-020 -19.102 -19.101 0.000 37.16
HS- 1.046e-021 1.030e-021 -20.980 -20.987 -0.007 20.59
S-2 1.508e-029 1.415e-029 -28.822 -28.849 -0.028 (0)
S(6) 6.246e-006
SO4-2 6.202e-006 5.819e-006 -5.207 -5.235 -0.028 13.95
MgSO4 2.629e-008 2.629e-008 -7.580 -7.580 0.000 5.84
CaSO4 9.678e-009 9.679e-009 -8.014 -8.014 0.000 7.50
HSO4- 5.055e-009 4.974e-009 -8.296 -8.303 -0.007 40.26
NaSO4- 2.664e-009 2.622e-009 -8.574 -8.581 -0.007 18.40
KSO4- 1.046e-010 1.029e-010 -9.981 -9.988 -0.007 (0)
CaHSO4+ 5.684e-013 5.594e-013 -12.245 -12.252 -0.007 (0)
Si 1.046e-004
H4SiO4 1.046e-004 1.046e-004 -3.980 -3.980 0.000 52.08
H3SiO4- 1.783e-009 1.754e-009 -8.749 -8.756 -0.007 27.96
H2SiO4-2 1.446e-017 1.357e-017 -16.840 -16.867 -0.028 (0)
Anhydrite -5.99 -10.26 -4.28 CaSO4

Aragonite -6.62 -14.96 -8.34 CaCO3
Calcite -6.48 -14.96 -8.48 CaCO3
CH4(g) -20.47 -23.32 -2.84 CH4
Chalcedony -0.43 -3.98 -3.55 SiO2
Dolomite -12.51 -29.60 -17.09 CaMg(CO3)2
Gypsum -5.68 -10.26 -4.58 CaSO4:2H2O
H2(g) -11.14 -14.24 -3.10 H2
H2O(g) -1.50 -0.00 1.50 H2O
H2S(g) -18.05 -26.04 -7.99 H2S
Halite -9.57 -8.00 1.57 NaCl
O2(g) -61.01 -63.91 -2.89 O2
Quartz 0.00 -3.98 -3.98 SiO2
SiO2(a) -1.27 -3.98 -2.71 SiO2
Sulfur -12.90 -8.01 4.88 S
Sylvite -10.45 -9.55 0.90 KCl
305
Talc -21.13 0.27 21.40 Mg3Si4O10(OH)2
------------------
End of simulation.
------------------
------------------------------------
------------------------------------
USE SOLUTION 0 60
EXCHANGE 1-60
equilibrate 1
X 250 #from Elrashidi et al., 2001
COPY cell 1 60
ADVECTION
cells 60
shifts 75 #15 shifts per year
time_step 0.06667 yr
initial_time 0
punch_cells 60
punch_frequency 1
print_cells 60
print_frequency 15
SELECTED_OUTPUT
file exRainX.sel
reset false
step
ph
totals K Ca Na
USER_PUNCH
heading Time
10 PUNCH (STEP_NO)/15
END
-------------------------------------------------------
Beginning of initial exchange-composition calculations.
-------------------------------------------------------
Exchange 1.
X 2.500e+002 mol
Equiv- Equivalent Log

Species Moles alents Fraction Gamma
MgX2 7.004e+001 1.401e+002 5.603e-001 -0.028

CaX2 5.388e+001 1.078e+002 4.311e-001 -0.028
NaX 1.891e+000 1.891e+000 7.562e-003 -0.007
KX 2.653e-001 2.653e-001 1.061e-003 -0.007
-----------------------------------------
-----------------------------------------
Reaction step 1.
Using solution 0. Solution after simulation 3.

Using exchange 1. Exchange assemblage after simulation 5.
306
-----------------------------Exchange composition------------------------------
X 2.500e+002 mol

MgX2 7.004e+001 1.401e+002 5.603e-001 -0.013

CaX2 5.388e+001 1.078e+002 4.311e-001 -0.013
NaX 1.891e+000 1.891e+000 7.562e-003 -0.003
KX 2.653e-001 2.653e-001 1.061e-003 -0.003
C 1.133e-005 1.133e-005
Ca 4.202e-007 4.202e-007
Cl 1.918e-005 1.918e-005
K 5.248e-007 5.248e-007
Mg 8.661e-007 8.661e-007
Na 1.874e-005 1.874e-005
S 7.287e-006 7.287e-006
Si 1.046e-004 1.046e-004

equilibrium
Volume (L) = 1.00297
Electrical balance (eq) = -1.306e-017
Percent error, 100*(Cat-|An|)/(Cat+|An|) = -0.00
Iterations = 71
Total H = 1.110124e+002
Total O = 5.550648e+001
Log Log Log mole V

H+ 1.231e-005 1.222e-005 -4.910 -4.913 -0.003 0.00

OH- 8.349e-010 8.286e-010 -9.078 -9.082 -0.003 -4.13
H2O 5.551e+001 1.000e+000 1.744 -0.000 0.000 18.07
C(-4) 3.549e-022
CH4 3.549e-022 3.549e-022 -21.450 -21.450 0.000 32.22
C(4) 1.133e-005
CO2 1.093e-005 1.093e-005 -4.961 -4.961 0.000 29.09
HCO3- 4.010e-007 3.980e-007 -6.397 -6.400 -0.003 24.69
307
NaHCO3 4.163e-012 4.163e-012 -11.381 -11.381 0.000 2.13
MgHCO3+ 3.936e-012 3.906e-012 -11.405 -11.408 -0.003 5.45
CaHCO3+ 2.083e-012 2.067e-012 -11.681 -11.685 -0.003 9.65
CO3-2 1.575e-012 1.528e-012 -11.803 -11.816 -0.013 -5.60
MgCO3 1.223e-015 1.223e-015 -14.913 -14.913 0.000 -17.09
CaCO3 1.045e-015 1.045e-015 -14.981 -14.981 0.000 -14.60
NaCO3- 5.333e-016 5.293e-016 -15.273 -15.276 -0.003 -0.64
Ca 4.202e-007
Ca+2 4.197e-007 4.072e-007 -6.377 -6.390 -0.013 -18.23
CaSO4 5.110e-010 5.110e-010 -9.292 -9.292 0.000 7.50
CaHCO3+ 2.083e-012 2.067e-012 -11.681 -11.685 -0.003 9.65
CaHSO4+ 4.135e-014 4.103e-014 -13.384 -13.387 -0.003 (0)
CaOH+ 5.574e-015 5.532e-015 -14.254 -14.257 -0.003 (0)
CaCO3 1.045e-015 1.045e-015 -14.981 -14.981 0.000 -14.60
Cl 1.918e-005
Cl- 1.918e-005 1.903e-005 -4.717 -4.720 -0.003 18.05
H(0) 8.538e-014
H2 4.269e-014 4.269e-014 -13.370 -13.370 0.000 28.61
K 5.248e-007
K+ 5.248e-007 5.208e-007 -6.280 -6.283 -0.003 8.99
KSO4- 2.602e-011 2.583e-011 -10.585 -10.588 -0.003 (0)
Mg 8.661e-007
Mg+2 8.647e-007 8.388e-007 -6.063 -6.076 -0.013 -21.91
MgSO4 1.388e-009 1.388e-009 -8.858 -8.858 0.000 5.84
MgHCO3+ 3.936e-012 3.906e-012 -11.405 -11.408 -0.003 5.45
MgOH+ 2.512e-013 2.493e-013 -12.600 -12.603 -0.003 (0)
MgCO3 1.223e-015 1.223e-015 -14.913 -14.913 0.000 -17.09
Na 1.874e-005
Na+ 1.874e-005 1.860e-005 -4.727 -4.730 -0.003 -1.41
NaSO4- 6.630e-010 6.580e-010 -9.178 -9.182 -0.003 18.39
NaHCO3 4.163e-012 4.163e-012 -11.381 -11.381 0.000 2.13
NaCO3- 5.333e-016 5.293e-016 -15.273 -15.276 -0.003 -0.64
NaOH 1.541e-024 1.541e-024 -23.812 -23.812 0.000 (0)
O(0) 0.000e+000
O2 0.000e+000 0.000e+000 -65.641 -65.641 0.000 30.40
S(-2) 5.489e-016
H2S 5.438e-016 5.438e-016 -15.265 -15.265 0.000 37.16
HS- 5.130e-018 5.091e-018 -17.290 -17.293 -0.003 20.58
S-2 5.190e-026 5.034e-026 -25.285 -25.298 -0.013 (0)
S(6) 7.287e-006
SO4-2 7.276e-006 7.058e-006 -5.138 -5.151 -0.013 13.92
HSO4- 8.446e-009 8.382e-009 -8.073 -8.077 -0.003 40.25
MgSO4 1.388e-009 1.388e-009 -8.858 -8.858 0.000 5.84
NaSO4- 6.630e-010 6.580e-010 -9.178 -9.182 -0.003 18.39
CaSO4 5.110e-010 5.110e-010 -9.292 -9.292 0.000 7.50
KSO4- 2.602e-011 2.583e-011 -10.585 -10.588 -0.003 (0)
CaHSO4+ 4.135e-014 4.103e-014 -13.384 -13.387 -0.003 (0)
Si 1.046e-004
H4SiO4 1.046e-004 1.046e-004 -3.980 -3.980 0.000 52.08
H3SiO4- 1.272e-009 1.263e-009 -8.895 -8.899 -0.003 27.95
H2SiO4-2 7.249e-018 7.032e-018 -17.140 -17.153 -0.013 (0)
Anhydrite -7.26 -11.54 -4.28 CaSO4

Aragonite -9.87 -18.21 -8.34 CaCO3
Calcite -9.73 -18.21 -8.48 CaCO3
308
CH4(g) -18.61 -21.45 -2.84 CH4
Chalcedony -0.43 -3.98 -3.55 SiO2
CO2(g) -3.50 -4.96 -1.46 CO2
Dolomite -19.01 -36.10 -17.09 CaMg(CO3)2
Gypsum -6.96 -11.54 -4.58 CaSO4:2H2O
H2(g) -10.27 -13.37 -3.10 H2
H2O(g) -1.50 -0.00 1.50 H2O
H2S(g) -14.21 -22.21 -7.99 H2S
Halite -11.02 -9.45 1.57 NaCl
O2(g) -62.75 -65.64 -2.89 O2
Quartz -0.00 -3.98 -3.98 SiO2
SiO2(a) -1.27 -3.98 -2.71 SiO2
Sulfur -9.93 -5.04 4.88 S
Sylvite -11.90 -11.00 0.90 KCl
Talc -26.07 -4.67 21.40 Mg3Si4O10(OH)2
------------------------------------
Beginning of advection calculations.
------------------------------------
Beginning of advection time step 1, cumulative pore volumes 0.016667.

Cell 60.
Using solution 60. Solution after simulation 5.

Using exchange 60. Exchange assemblage after simulation 5.
Using pure phase assemblage 60. Pure-phase assemblage after simulation 5.
Moles in assemblage
CO2(g) -1.90 -3.36 -1.46 1.000e+001 1.000e+001 0.000e+000

Quartz 0.00 -3.98 -3.98 1.000e+001 1.000e+001 0.000e+000
-----------------------------Exchange composition------------------------------
X 2.500e+002 mol
309
MgX2 7.004e+001 1.401e+002 5.603e-001 -0.028

CaX2 5.388e+001 1.078e+002 4.311e-001 -0.028
NaX 1.891e+000 1.891e+000 7.562e-003 -0.007
KX 2.653e-001 2.653e-001 1.061e-003 -0.007
C 4.575e-004 4.575e-004
Ca 9.980e-006 9.980e-006
Cl 1.128e-004 1.128e-004
K 2.557e-006 2.557e-006
Mg 2.057e-005 2.057e-005
Na 9.135e-005 9.135e-005
S 6.246e-006 6.246e-006
Si 1.046e-004 1.046e-004

pe = -0.213 Adjusted to redox
equilibrium
Volume (L) = 1.00301
Total alkalinity (eq/kg) = 1.344e-005
Iterations = 1
Total H = 1.110125e+002
Total O = 5.550739e+001
Log Log Log mole V

H+ 8.931e-006 8.792e-006 -5.049 -5.056 -0.007 0.00

OH- 1.170e-009 1.151e-009 -8.932 -8.939 -0.007 -4.13
H2O 5.551e+001 1.000e+000 1.744 -0.000 0.000 18.07
C(-4) 1.923e-018
CH4 1.923e-018 1.923e-018 -17.716 -17.716 0.000 32.22
C(4) 4.575e-004
CO2 4.351e-004 4.351e-004 -3.361 -3.361 0.000 29.09
HCO3- 2.236e-005 2.201e-005 -4.650 -4.657 -0.007 24.69
MgHCO3+ 5.043e-009 4.963e-009 -8.297 -8.304 -0.007 5.46
CaHCO3+ 2.668e-009 2.626e-009 -8.574 -8.581 -0.007 9.66
NaHCO3 1.113e-009 1.113e-009 -8.954 -8.954 0.000 2.13
CO3-2 1.251e-010 1.174e-010 -9.903 -9.930 -0.028 -5.58
MgCO3 2.159e-012 2.159e-012 -11.666 -11.666 0.000 -17.09
CaCO3 1.845e-012 1.845e-012 -11.734 -11.734 0.000 -14.60
310
NaCO3- 1.997e-013 1.965e-013 -12.700 -12.707 -0.007 -0.63
Ca 9.980e-006
Ca+2 9.968e-006 9.353e-006 -5.001 -5.029 -0.028 -18.20
CaSO4 9.678e-009 9.679e-009 -8.014 -8.014 0.000 7.50
CaHCO3+ 2.668e-009 2.626e-009 -8.574 -8.581 -0.007 9.66
CaCO3 1.845e-012 1.845e-012 -11.734 -11.734 0.000 -14.60
CaHSO4+ 5.684e-013 5.594e-013 -12.245 -12.252 -0.007 (0)
CaOH+ 1.794e-013 1.765e-013 -12.746 -12.753 -0.007 (0)
Cl 1.128e-004
Cl- 1.128e-004 1.110e-004 -3.948 -3.955 -0.007 18.06
H(0) 2.916e-013
H2 1.458e-013 1.458e-013 -12.836 -12.836 0.000 28.61
K 2.557e-006
K+ 2.557e-006 2.517e-006 -5.592 -5.599 -0.007 8.99
KSO4- 1.046e-010 1.029e-010 -9.981 -9.988 -0.007 (0)
Mg 2.057e-005
Mg+2 2.053e-005 1.927e-005 -4.688 -4.715 -0.028 -21.89
MgSO4 2.629e-008 2.629e-008 -7.580 -7.580 0.000 5.84
MgHCO3+ 5.043e-009 4.963e-009 -8.297 -8.304 -0.007 5.46
MgOH+ 8.085e-012 7.958e-012 -11.092 -11.099 -0.007 (0)
MgCO3 2.159e-012 2.159e-012 -11.666 -11.666 0.000 -17.09
Na 9.135e-005
Na+ 9.134e-005 8.990e-005 -4.039 -4.046 -0.007 -1.41
NaSO4- 2.664e-009 2.622e-009 -8.574 -8.581 -0.007 18.40
NaHCO3 1.113e-009 1.113e-009 -8.954 -8.954 0.000 2.13
NaCO3- 1.997e-013 1.965e-013 -12.700 -12.707 -0.007 -0.63
NaOH 1.035e-023 1.035e-023 -22.985 -22.985 0.000 (0)
O(0) 0.000e+000
O2 0.000e+000 0.000e+000 -66.708 -66.708 0.000 30.40
S(-2) 3.204e-014
H2S 3.162e-014 3.162e-014 -13.500 -13.500 0.000 37.16
HS- 4.180e-016 4.113e-016 -15.379 -15.386 -0.007 20.59
S-2 6.022e-024 5.650e-024 -23.220 -23.248 -0.028 (0)
S(6) 6.246e-006
SO4-2 6.202e-006 5.819e-006 -5.207 -5.235 -0.028 13.95
MgSO4 2.629e-008 2.629e-008 -7.580 -7.580 0.000 5.84
CaSO4 9.678e-009 9.679e-009 -8.014 -8.014 0.000 7.50
HSO4- 5.055e-009 4.974e-009 -8.296 -8.303 -0.007 40.26
NaSO4- 2.664e-009 2.622e-009 -8.574 -8.581 -0.007 18.40
KSO4- 1.046e-010 1.029e-010 -9.981 -9.988 -0.007 (0)
CaHSO4+ 5.684e-013 5.594e-013 -12.245 -12.252 -0.007 (0)
Si 1.046e-004
H4SiO4 1.046e-004 1.046e-004 -3.980 -3.980 0.000 52.08
H3SiO4- 1.783e-009 1.754e-009 -8.749 -8.756 -0.007 27.96
H2SiO4-2 1.446e-017 1.357e-017 -16.840 -16.867 -0.028 (0)
Anhydrite -5.99 -10.26 -4.28 CaSO4

Aragonite -6.62 -14.96 -8.34 CaCO3
Calcite -6.48 -14.96 -8.48 CaCO3
CH4(g) -14.87 -17.72 -2.84 CH4
Chalcedony -0.43 -3.98 -3.55 SiO2
Dolomite -12.51 -29.60 -17.09 CaMg(CO3)2
Gypsum -5.68 -10.26 -4.58 CaSO4:2H2O
311
H2(g) -9.74 -12.84 -3.10 H2
H2O(g) -1.50 -0.00 1.50 H2O
H2S(g) -12.45 -20.44 -7.99 H2S
Halite -9.57 -8.00 1.57 NaCl
O2(g) -63.82 -66.71 -2.89 O2
Quartz 0.00 -3.98 -3.98 SiO2
SiO2(a) -1.27 -3.98 -2.71 SiO2
Sulfur -8.70 -3.81 4.88 S
Sylvite -10.45 -9.55 0.90 KCl
Talc -21.13 0.27 21.40 Mg3Si4O10(OH)2
312
Appendix F Proposed Revised Florida Methods
F.1 Florida Method of Test for pH of Soil and Water, FM 5-550
Florida Method of Test for pH of Soil and Water

Designation: FM 5-550
1. SCOPE
1.1. This method covers the determination of pH in soil and water using a
pH/mV meter with automatic temperature compensation (ATC) and a
combination electrode that includes a silver/silver chloride reference
electrode, a glass bulb indicating electrode, and a thermocouple.
2. APPARATUS
2.1. pH Meter and Electrode System: Portable or benchtop pH/mV meter with
automatic temperature compensation (ATC) and accuracies for mV, the
greater of ± 0.2 mV or ± 0.05%; for pH, ± 0.002 pH units; for temperature,
± 0.3C, or better; a display of calibration data including %slope and
offset; and a refillable double-junction 3-in-1 combination electrode that
includes a silver/silver chloride (Ag/AgCl2) reference electrode, glass bulb
indicating electrode, and thermocouple with pH/ATC connectors that
match the meter inputs, capable of measuring 0 to 14 pH from -5C to
100C with a minimum resolution of 0.01 pH units (for example, Fisher
Scientific pH/mV meter kit catalog #13-636-AB150).
2.2. Analytical Balance: An analytical balance with a capacity of 2,000 g or

more and a resolution of 0.01 g or better.
2.3. 100 and 250 mL beakers, wash bottle, glass stirring rods, 100-mL scoop,
and delicate glass cleaning tissues.
3. REAGENTS
3.1. Standard buffered solutions for pH 4, 7, and 10.
3.2. Electrode filling solution and electrode storage solution (refer to electrode
manual).
313
3.3. Deionized or distilled water (dilution water) with a resistivity equal or
greater than 200,000 ohm-cm.
4. SAMPLES
4.1. Soil Sampling: Every effort should be made to obtain a soil sample that is
representative of the bulk material. Use clean tools for gathering samples.
Excessive moisture should be avoided by sampling from an area that has
been allowed to gravity drain for a short time. If the soil sample has
excess free moisture, place approximately 2.2 pounds (1 kg) of the soil on
top of a suitable sieve and cover with plastic. Allow the sample to drain for
a minimum of one hr. This step may be performed in the lab prior to
testing. If the soil sample is obtained from a heap that has been sitting for
a long time, take the sample from a depth below the weathered surface
where the moisture content appears to have stabilized. Avoid taking the
sample from near a weathered soil surface. If sampling from ground level,
remove top 12” (30 cm) to eliminate vegetation and debris before
sampling. The soil sample may be taken from underneath standing water,
but excess water should not be included with the sample. Soil samples
should be placed in plastic or plastic-lined bags. Squeeze the bag down
snugly around the sample and seal tightly to minimize contact with air.
4.2. Water Sampling: Water samples should be obtained from the main
channel of rivers and streams. Sampling from other bodies of water such
as lakes or ponds should be obtained from areas conducive to the capture
of representative samples. Care should be observed not to sample from
stagnant or pooled water, unless a structure will be placed in such an
area. Sample the water just below the surface to alleviate introduction of
floating debris, i.e. leaves, sticks, foam or trash. Fill the sample container
to the top to eliminate introducing air into the sample and tightly seal the
lid. The sample container shall be clean, at least 1 quart (1 liter) in size,
and be either glass or plastic with an airtight lid. When possible, submerge
the sample container below the surface of the water to completely fill and
secure lid underwater.
4.3. Field Measurement of Soil pH: Fill a 250-mL beaker to the 100-mL mark
with soil from sample container and add 100 mL of dilution water. Break
up any clumps of soil. Stir thoroughly the soil and water mixture for ~20
seconds every 10 minutes over a 30-minute period. Continue with test
procedure from Section 6.1.
4.4. Transporting the samples: Maintain test samples in a cool dark area after
sampling and during transport to the test facility.
314
4.5. Storing Samples: If samples need to be stored, store water and soil
samples at or below 39F (4C). Care should be taken to prevent freezing
of the samples. Analyze samples within seven (7) days.
5. SAMPLE PREPARATION
5.1. Preparation of Water: Allow test sample to reach room temperature.
5.2. Preparation of Soil: Allow the soil sample and dilution water to reach room
temperature. Place 100 g of sample in a 250 mL beaker and then add 100
mL of dilution water. Break up any clumps of soil. Stir thoroughly the soil
and water mixture for ~20 seconds every 10 minutes over a 30-minute
period.
6. TEST PROCEDURE
6.1. Calibration of pH Meter and Electrode System: Pour ~50 mL of pH 4, pH

7, and pH 10 standard buffer solutions into separate 100-mL beakers and
allow solutions to reach room temperature prior to calibration (Note 1).
Remove pH electrode from storage solution, rinse with dilution water and
blot or dab dry with tissue, uncover the fill hole, then place in pH 7 buffer
solution for ~10 minutes prior to calibration. Calibrate the pH/mV meter
and 3-in-1 combination electrode with three standard buffers according to
the directions in the manufacturer's instruction manual. When the
calibration is complete, record the calibration slope in %, offset in mV, and
buffer temperature in C (Note 2). The slope must be within 95% to 102%,
the offset within ± 10 mV, and the buffer temperatures within 25 ± 5C; if
otherwise, follow the instruction manual to clean, fill, equilibrate, re-
calibrate, re-activate, or replace the electrode so that the slope, offset, and
temperatures are within these values.
6.2. Care of Electrode: Check the electrode filling solution and keep filled to
the fill hole. Between any two pH measurements, rinse electrode with
dilution water and blot or dab the electrode tip dry with a delicate glass
tissue (Note 3).
6.3. Check of Calibration: Check the calibration by measuring the pH of each

of the standard buffers. The measured pH should be within 0.05 pH units
of the buffer pH at the test temperature. Buffer pH versus test temperature
data are typically tabulated on the buffer bottle. If the measured pH is
outside 0.05 pH units of the buffer pH at test temperature, take corrective
steps and re-calibrate the electrode (Note 4).
315
6.4. Determination of pH: Insert the pH electrode in the buffer, water or soil
sample such that the portion of the electrode that contains the glass bulb,
thermocouple, and liquid junction is submersed, typically about ½” (1.3
cm) of the electrode tip. Do not immerse the electrode to a depth greater
than the electrode’s internal reference solution, as the sample may
penetrate into and contaminate the reference solution. Very gently stir the
electrode in the sample (or alternately, gently move the sample beaker
beneath the electrode) to assure a pH representative of the bulk solution
(Note 5). Wait until the meter’s stable reading indicator stays on for 10
seconds or longer (Note 6). Remove the electrode from the sample, rinse
the tip with dilution water and blot or dab dry, and return the electrode to
the sample for a second pH measurement. Wait until the meter’s stable
reading indicator stays on for 10 seconds or longer. Record the pH to the
nearest hundredth pH units and sample temperature to the nearest tenth
C, for example, pH 7.00 @ 25.1C. After the second reading, return the
pH electrode to the pH 7 buffer and check that the electrode is still within
calibration.
6.5. Storage of Electrode (Note 7): Store the pH electrode overnight or longer
in its storage bottle either in the electrode storage solution with the fill hole
open in pH 4 or 7 buffer with the fill hole closed. Do not store in dilution
water as this will have an adverse effect on electrode performance. Tap
water or standard buffer solutions are appropriate for short-term electrode
storage.
Note 1: Always use fresh buffers for electrode calibration.
Note 2: For some meters, the slope and offset are presented separately for two buffers (for
example, pH 4 and pH 7, or pH 7 and pH 10). Record the slope and offset for the range
that includes the sample pH.
Note 3: With some 3-in-1 combination electrodes, the narrow space between the glass bulb and
the temperature sensor traps liquid, whether buffer, sample, or dilution water. This
trapped liquid can be transferred into the sample and may affect the pH reading.
Note 4: If after a sample pH measurement the calibration check fails, suspect a partially clogged
liquid junction on the electrode. Add electrode filling solution to flush the liquid junction,
and then re-check calibration.
Note 5: Ideally, once the pH reading has stabilized, the reading will not change whether gently
stirred or unstirred. For low ionic strength solutions, however, the pH may not stabilize if
stirred. If the pH reading does not stabilize after ~ 5 min with gentle stirring, obtain and
record a stable reading with no stirring.
Note 6: Disable the “hold” feature of the pH meter.
316
Note 7: Store the electrode in accordance with manufacturer’s instructions if different from this
section.
7. PRECISION AND BIAS
7.1. Bias: Single-operator, single laboratory bias for this method using an
Accumet meter/electrode AB150/13-620-631 was evaluated at 20C with
repeated measures of each six aqueous standards ranging from pH 5.00
to pH 9.00. The average bias was -0.01 pH units.
7.2. Reproducibility: For two test materials, multi-laboratory standard

deviations of a single test result were dependent on soil pH (Table 1). For
materials of similar pH, the results of two properly-conducted tests in
different laboratories on the same material are not expected to differ by
more than the pH units shown in column labeled “Acceptable Range of
Two Results.”
Table 1 Multi-Laboratory Precision for FM 5-550 pH

Average, Standard Deviation, Acceptable Range of
Material
pH units pH units Two Results
Sand (A-3) 7.76 0.32 0.91
Sand (A-3) 5.11 0.15 0.42
317
F.2 Florida Method of Test for Minimum Resistivity of Soil and Water, FM 5-551
Florida Method of Test for Minimum Resistivity of Soil and Water

1. SCOPE
1.1 This method covers the laboratory determination of resistivity of soil and
water using a soil resistance meter and soil box.
2. APPARATUS
2.1 Resistivity Meter: Any four-pin terminal, null-balancing ohmmeter or

multimeter capable of four-wire resistance measurements from one to
one million ohms, either analog or digital (as examples, resistivity meters
MC Miller Model 400A, Nilsson Model 400, and Tinker & Rasor Model SR-
2).
2.2 Soil Box: Designed such that the cross-sectional area (cm2) of the
sample, with the box filled level, divided by the distance (cm) between the
pins is equal to 1 cm (for example, MC Miller catalog #37008).
2.3 Analytical Balance: An analytical balance with a capacity of 2,000 g or

2.4 Other: Test leads, thermometer, flat spatula, mixing spoon, large non-
corrosive (glass, plastic, or stainless steel) bowl for mixing, 100-mL
graduated cylinder, squeeze bottle for cleaning, and disposable nitrile
gloves.
3. REAGENTS
3.1 Conductivity standard, sodium chloride, 250 µS/cm (for example, Fisher
Scientific catalog #22366032).
3.2 Distilled or deionized water (hereafter referred to as dilution water) with a

resistivity of 200,000 ohm-cm or greater (Note 1).
Note 1: Deionized or distilled water stored in containers that are not airtight will over time absorb
ions from acidic and basic gases in the atmosphere. Absorbed ions will lower the water’s
resistivity.
318
4. SAMPLES
4.1 Soil Sampling: Every effort should be made to obtain a soil sample that is
Excessive moisture should be avoided by sampling from an area that
has been allowed to gravity drain for a short time. If the soil sample has
excess free moisture, place approximately 2.2 pounds (1 kg) of the soil
on top of a suitable sieve and cover with plastic. Allow the sample to
drain for one hr. This step may be performed in the lab prior to testing. If
the soil sample is obtained from a heap that has been sitting for a
long time, take the sample from a depth below the weathered surface
sampling. The soil sample may be taken from underneath standing
water, but excess water should not be included with the sample. Soil
samples should be placed in plastic or plastic-lined bags. Squeeze the
bag down snugly around the sample and seal tightly to minimize contact
with air.
4.2 Water Sampling: Water samples should be obtained from the main
as lakes or ponds should be obtained from areas conducive to the
capture of representative samples. Care should be observed not to
sample from stagnant or pooled water unless a structure will be placed in
such an area. Sample the water just below the surface to alleviate
introduction of floating debris such as leaves, sticks, foam or trash. Fill
the sample container to the top to eliminate introducing air into the
sample and tightly seal the lid. The sample container shall be clean, at
least 1 quart (1 liter) in size, and be either glass or plastic with an airtight
lid. When possible, submerge the sample container below the surface of
the water to completely fill and secure lid underwater.
4.3 Transporting the Samples: Maintain test samples in a cool dark area after
4.4 Storing Samples: Store water samples at or below 39F (4C). Care
should be taken to prevent freezing of the samples. Analysis of a soil
sample “as received” is preferred. If, however, soil samples cannot be
analyzed within ~ 1 day of receipt, dry soil samples per Section 5.2 and
store dried soil at room temperature prior to analysis. Analyze samples
within seven (7) days.
319
5.1 Preparation of Water: Allow test sample to reach room temperature.
5.2 Preparation of Soil:
A. Loose Granular Soils: Spread the sample in a thin layer on a clean

tray and dry under ambient conditions until a constant mass is
achieved, or dry in an oven at no higher than 140F (60C) for
approximately four hr or until a constant mass is achieved. Sieve
through a No. 10 mesh (2 mm) sieve. Split the sample per ASTM C
702-98 to obtain approximately 1,200 g.
B. Muck and Soils with Clay: Spread the sample in a thin layer on a
clean tray and dry under ambient conditions until a constant mass
is achieved, or dry in an oven at no higher than 140F (60C) for
approximately four hr or until a constant mass is achieved. Using a
rawhide mallet or other suitable device pulverize the sample and
sieve through a No. 10 mesh (2 mm) sieve. Split the sample per
ASTM C 702-98 to obtain approximately 1,200 g.
6. TEST PROCEDURE
6.1 Equipment Set-up: Rinse the soil box, mixing bowl, and utensils with
dilution water before starting and after completion of each test. Follow
the manufacturer's instructions for properly connecting the test leads
between the meter and the soil box. A standard soil box will have plates
for application of current at both ends and a pair of electrode pins for
reading potential evenly spaced between the plates (Figure 1).
320
6.2 Determine Resistivity of “As-Received” Soil: If an “as received” soil
resistivity is requested, bring the soil sample and dilution water to room
temperature. Place 1,000 g of “as-received” soil sample into a large bowl.
Remove any debris (such as grass, roots, sticks or rocks, etc.). Fill the
soil box to the top with the soil sample, taking care to fill any voids, and
strike off any excess soil on the top of soil box. Connect the soil box to the
resistivity meter as instructed in the manufacturer’s instruction manual and
determine the soil resistivity. Record the “as-received” resistivity in ohm-
cm. Proceed to Section 6.3 to determine minimum resistivity.
6.3 Determine Minimum Resistivity of “As-Received” or of Dried, Sieved Soil:

Bring the soil sample and dilution water to room temperature. Place 1,000
g of soil sample into a large bowl.
6.4 Add 100 mL of dilution water to the soil sample and mix thoroughly (Note
2).
6.5 Fill the soil box to the top with the soil sample, taking care to fill any voids,
and strike off any excess soil on the top of soil box. Connect the soil box to
the resistivity meter as instructed in the manufacturer’s instruction manual
and determine the soil resistivity. Record the t o t a l volume of water
used in mL a n d the resistivity in ohm-cm. Place soil sample back into
mixing bowl (Note 3).
6.6 Repeat steps 6.4 and 6.5. The measured resistivity should decrease.
When the resistivity of the soil sample begins to increase, record the
t o t a l volume of water used in mL a nd the higher resistivity, and
measure and record the temperature in C of the sample in the soil box.
6.7 Record the lowest resistivity obtained as the minimum resistivity and the
temperature of the sample. Report the minimum resistivity in ohm-cm and
the temperature in C, for example, 3,000 ohm-cm @ 25C.
6.8 Determine Resistivity of a Small Soil Sample: If soil resistivity is requested

for a sample that is less than 1,000 g, complete steps 6.4 and 6.5 using
increments of 10% water instead of 100 mL (Note 4).
Note 2: In some soils, even a trace amount of salt will affect the resistivity reading. Avoid
transferring salt from hands to sample. For soils that are difficult to mix, wear disposable
gloves and mix soil with gloved hands.
Note 3: In many cases the minimum resistivity will occur after soil slurry is formed. As water is
added, mix thoroughly the soil slurry and pour the water from the slurry into the box until
321
the box is filled. If the water alone cannot fill the soil box, add enough of the remaining
soil to completely fill the box.
Note 4: A sample size of less than 1,000 g is not appropriate for select backfill (FDOT material
092L).
7. BEST PRACTICES
7.1 Determine Resistivity of Water: Refer to the manufacturer’s operation

manual for determining resistivity.
7.2 Determine Resistivity of Conductivity Standard: Test the conductivity

standard at least once per quarter and after the resistivity meter has
undergone repair or replacement. Bring the conductivity standard to room
temperature. Fill the soil box to the top with conductivity standard.
Connect the soil box to the resistivity meter as instructed in the
manufacturer’s instruction manual and verify that the conductivity standard
resistivity is 4,000 ohm-cm @ 25C. Refer to Table 1 for the conductivity
of the standard for testing at temperatures between 20C and 30C.
Record the conductivity standard resistivity in ohm-cm and the
measurement temperature in C. If the resistivity at measurement
temperature is outside 5%, troubleshoot and correct the problem, then
re-test. Rinse the soil box well with dilution water after testing the
conductivity standard.
Table 1 Temperature-Dependence of a 250 µS/cm

NaCl Conductivity Standard
Temperature, C Resistivity, ohm-cm
20 4,440
21 4,350
22 4,250
23 4,170
24 4,080
25 4,000
26 3,920
27 3,850
28 3,780
29 3,710
30 3,650
322
8.1 Bias: Single-operator, single laboratory bias for this method was evaluated
using a Nilsson Model 400 analog meter and a McMiller 280-mL soil box
and repeated measures of two aqueous standards with resistivities at
25C of 2,000 ohm-cm and 4,000 ohm-cm. After correcting for
measurement temperatures in the range of 21oC to 23oC, average biases
were -30 ohm-cm (-1.3%) for the 2,000 ohm-cm standard and 30 ohm-cm
(0.7%) for the 4,000 ohm-cm standard. With no correction for temperature,
average biases were 100 ohm-cm (5%) for the 2,000 ohm-cm standard
and 300 ohm-cm (7%) for the 4,000 ohm-cm standard.
8.2 Reproducibility: For two test materials, multi-laboratory standard deviations

of a single test results were dependent on soil minimum resistivity (Table
2). For materials with similar minimum resistivity, the results of two
properly-conducted tests in different laboratories on the same material are
not expected to differ by more than the ohm-cm shown in the column
labeled “Acceptable Range of Two Results.”
Table 2 Multi-Laboratory Precision for FM 5-551 Minimum Resistivity

Average, Standard Deviation, Acceptable Range of
Material
ohm-cm ohm-cm Two Results
Sand (A-3) 12,200 1,090 3,090
Sand (A-3) 2,310 430 1,230
323
F.3 Florida Method of Test for Chloride in Soil and Water, FM 5-552
Florida Method of Test for Chloride in Soil and Water

1. SCOPE
1.1 This method covers the determination of chlorides in soil and water using
either a screening approach based on a chloride reagent system or an
analytical approach as found in the Standard Methods for the Examination
of Water and Wastewater, Section 4500-Cl- B or Section 4110 B
(SMEWW).
1.2 Both the screening and analytical approaches use a silver nitrate titration
of a sample that is amended with a potassium chromate. Silver combines
first with chloride and then with chromate in a mixture of the two. Both
silver chloride and silver chromate are relatively insoluble in water but
silver chromate is a brown-red crystal and its appearance in the sample
marks the titration endpoint.
2. APPARATUS
2.1 Chloride Reagent System: Chloride Low Range Test Kit Model 8-P, 5-400
ppm (Hach catalog #144001) and Chloride High Range Test Kit Model
CD-51 (Hach catalog #208601), or equivalent, may be used for chloride
determinations. For the Hach Model 8-P test kit, additional glass mixing
bottles with a 23-mL mark are needed (for example, Hach catalog
#232706). Alternatively, the laboratory instruments referred to in the
SMEWW may be used for chloride determination.
2.2 Analytical Balance: An analytical balance with a capacity of 2,000 g or

2.3 Vacuum Filtration System: A vacuum filtration system is required. Such as

system includes a vacuum pump, 0.25” (6 mm) inner diameter flexible
vacuum hose, 300-mL filter funnel/holder for a 47-mm diameter filter, and
a 1-liter side arm vacuum filtration flask or similar.
2.4 Other: Whatman 41 filter paper (or equivalent), glass or plastic funnel, two
or more 500-mL Erlenmeyer flasks, 47-mm diameter 0.45-micron pore
size mixed cellulose ester (MCE) membrane filters, 0.10-g measuring
spoon, small ceramic dish, a transfer pipette (variable from 0.100 to 1.000
324
mL, capable of delivering increments of 0.01 mL) and tips, 100-mL
graduated cylinder, and disposable nitrile gloves.
3. REAGENTS
3.1 Potassium dichromate (K2Cr2O7), sodium bicarbonate and silver nitrate

(AgNO3) titrant are required for the chloride test kit (Note 1). The
potassium dichromate and sodium bicarbonate are supplied as Chloride 2
powder pillows (for example, Hach catalog #105766). For the SMEWW
method, reagent purchasing and preparation information is provided in the
method procedures.
3.2 Chloride standard solution, 1,000 ppm (mg/L) as Cl-, NIST-traceable.
3.3 Ferric sulfate, hydrated reagent grade (for example, Fisher Scientific
catalog #S25322 or #S25322A).
3.4 Deionized or distilled water (dilution water) with a resistivity equal or

greater than 200,000 ohm-cm.
Note 1: Potassium chromate and silver nitrate are hazardous materials. Refer to the chemical
safety data sheets (SDS) regarding the safe storage, handling, and disposal of these
hazardous materials. Dispose of hazardous material in accordance with federal, state,
and local mandates.
4. SAMPLES
been allowed to gravity drain for a short time. If the soil sample has
excess free moisture, place approximately 2.2 pounds (1 kg) of the soil on
top of a suitable sieve and cover with plastic. Allow the sample to drain for
a minimum of one hr. This step may be performed in the lab prior to
testing. If the soil sample is obtained from a heap that has been sitting for
a long time, take the sample from a depth below the weathered surface
sampling. The soil sample may be taken from underneath standing water,
but excess water should not be included with the sample. Soil samples
should be placed in plastic or plastic-lined bags. Squeeze the bag down
snugly around the sample and seal tightly to minimize contact with air.
325
4.2 Water Sampling: Water samples should be obtained from the main
as lakes or ponds should be obtained from areas conducive to the capture
of representative samples. Care should be observed not to sample from
stagnant or pooled water, unless a structure will be placed in such an
area. Sample the water just below the surface to alleviate introduction of
floating debris such as leaves, sticks, foam or trash. Fill the sample
container to the top to eliminate introducing air into the sample and tightly
seal the lid. The sample container shall be clean, at least 1 quart (1 liter) in
size, and be either glass or plastic with an airtight lid. When possible,
submerge the sample container below the surface of the water to
completely fill and secure lid underwater.
4.3 Transporting the samples: Maintain test samples in a cool dark area after
4.4 Storing Samples: Store water samples at or below 39F (4C). Care
should be taken to prevent freezing of the samples. Dry soil samples
without delay per Section 5.2 and store dried soil at room temperature
prior to analysis. Analyze samples within seven (7) days.
5.1 Preparation of Water: Allow test sample to reach room temperature. If

water sample contains suspended solids or color, gravity filter the water
through a Whatman 41 filter (or equivalent), and if necessary, vacuum
filter the water sample through a 0.45-micron pore size MCE membrane
filter.

achieved, or dry in an oven at 140F (60C) for approximately four
hr or until a constant mass is achieved. Sieve through a No. 10
mesh (2 mm) sieve. Split the sample per ASTM C 702-98 to obtain
approximately 400 g.
clean tray and dry under ambient conditions until a constant mass
is achieved, or dry in an oven at no higher than 140F (60C) for
326
ASTM C 702-98 to obtain approximately 400 g.
5.3 Preparation of Soil Extract: Place 100 grams of dried soil in a 500 mL
Erlenmeyer flask (Note 2). Add 300 mL of dilution water, stopper, and
shake vigorously for 20 seconds. Let stand for one hr and repeat agitation.
Let stand for a minimum of 12 hr.
A. Suspended particles (turbidity) or color may interfere with the

chloride determination. If the extract is cloudy or colored due to the
suspension of fine particles, add ferric sulfate. Dissolve in a small
ceramic dish ~0.10 g of ferric sulfate into 1 mL of water and pipette
the dissolved ferric sulfate into the extract. Allow the sample to
stand for three to four hr as the suspended particles settle.
B. Filter by gravity the soil extract through a Whatman 41 filter (or

equivalent) into a 500-mL Erlenmeyer flask. Slowly decant into the
filter the water layer followed by the soil slurry from the sample
extract and allow the liquid to drain until the liquid stops dripping.
C. Filter by vacuum the soil extract through a 0.45-micron pore size

membrane filter (Note 3) into a clean 1-liter vacuum filtration flask;
repeat if necessary to get a clear extract.
Note 2: Prepare separate soil extracts for sulfate and chloride analyses. If after 12 hr of settling
the both extracts are clear enough to filter, only one of two extracts needs to be filtered
for both sulfate and chloride analysis.
Note 3: The MCE membrane filter is white; spacers are blue.
6. TEST PROCEDURE
6.1 Low Range: For the SMEWW analytical approach or an equivalent test kit,
follow the kit instructions and include preparation and testing of a blank
and 30-ppm check standard. For the Hach Model 8-P chloride test kit (5-
400 ppm), follow these instructions:
A. Check the Reagents: Check that the silver chloride titrant and
Chloride 2 Indicator powder pillows have not reached their
expiration date.
B. Prepare a Reagent Blank: Fill one mixing bottle with dilution water
327
to the 23-mL mark. Prepare a blank at least once per day on days
in which samples are tested. Prepare a 30 mg/L (30 ppm) Check
Standard: Add to a second mixing bottle 0.690 g of 1,000 mg/L
chloride standard solution and fill with dilution water to the 23-mL
mark.
C. Prepare a check standard at least once per day on days in which

samples are tested.
D. Prepare a Water Sample or Soil Sample Extract: Fill a third mixing

bottle with the water sample or soil sample extract to the 23-mL
mark.
E. Determine the Blank Chloride Concentration: Add one Chloride 2

Indicator powder pillow (potassium dichromate/sodium bicarbonate)
to mixing bottle that contains the blank. Gently swirl the bottle to
achieve a uniform yellow color. Add one drop of 0.0493 N silver
nitrate to the blank and gently swirl the bottle to mix. Each drop is
equal to ~5 mg/L (5 ppm) of chloride (Note 4). With one drop of
silver nitrate the color of the blank should turn from yellow to red-
brown; this red-brown color will serve as a guide (Note 5). If two or
more drops of silver nitrate are necessary to titrate the dilution
water to a red-brown color, then either the dilution water is
contaminated with chloride or the titrant has gone bad (Notes 6 and
7).
Sample Calculation: 1 drop x 5 ppm/drop = 5 ppm per blank
F. Determine the Check Standard Chloride Concentration: Add one

Chloride 2 Indicator powder pillow (potassium dichromate/sodium
bicarbonate) to mixing bottle that contains the check standard. Add
one drop at a time of 0.0493 N silver nitrate and gently swirl the
bottle to mix. The check standard should turn from yellow to red-
brown after the addition of 6 to 8 drops. Record the total number of
drops and subtract from this number one drop for the blank.
Multiply the net number of drops by 5 ppm per drop. The check
standard concentration should fall between 25 ppm and 35 ppm. If
the concentration is less than 25 ppm or greater than 35 ppm the
volume of added chloride standard or the drop size of silver nitrate
may be off; otherwise the concentrations of either the chloride
check standard or the silver nitrate titrant may be off. In either case,
prepare a new check standard and re-test.
328
Sample Calculation: 7 drops x 5 ppm/drop = 35 ppm – 5 ppm blank = 30 ppm for
check standard
G. Determine the Water Sample or Soil Sample Extract Chloride

Concentration (0-60 ppm): Add one Chloride 2 Indicator powder
pillow (potassium dichromate/sodium bicarbonate) to mixing bottle
that contains the water sample or soil sample extract. Gently swirl
the bottle to achieve a uniform yellow color. Add one drop at a time
of 0.0493 N silver nitrate until the color of the extract changes from
yellow to red-brown. Record the total number of drops and multiply
each drop by 5 ppm of chloride per drop, then subtract from this
concentration the 5 ppm of chloride for the blank. If this result is
zero, record that the chloride concentration of the sample as below
detection. Otherwise, multiply the chloride concentration by the
dilution factor, which may be 1 for a water sample but 3 to account
for the initial dilution of a soil sample (100 grams of soil to 300 mL
water). Report this concentration in ppm.
Sample Calculation: 3 drops x 5 ppm/drop = 15 ppm – 5 ppm blank = 10 ppm x 3

dilution factor for soil extract = 30 ppm soil concentration
H. Determine the Water Sample or Soil Sample Extract Chloride

Concentration (60-400 ppm): If the water sample or soil sample
extract has not changed color after the addition of 13 drops (60
ppm), stop the titration. Prepare a new sample extract: fill a clean
mixing bottle to the 10-mL mark. Add one drop at a time of 0.0493
N silver nitrate until the color of the extract changes from yellow to
red-brown. Record the total number of drops and multiply each
drop by 20 ppm of chloride per drop, then subtract from this
concentration the 5 ppm of chloride for the blank. Otherwise,
multiply the chloride concentration by the dilution factor, which may
be 1 for a water sample but 3 to account for the initial dilution of a
soil sample (100 grams of soil to 300 mL water). Report this
concentration in ppm.

6.2 High Range: If water or soil has a high chloride concentration (for
example, seawater has a chloride concentration of ~19,000 ppm) use the
Hach Model CD-51 chloride test kit (500-100,000) (or equivalent) and
follow the kit instructions. For a check standard, fill the measuring tube
329
with a 1000-ppm chloride standard and continue with the test instructions.
The standard should be titrated from yellow to red-brown with 1-3 drops of
1.88-N silver nitrate titrant (500 ppm per drop). If more than 3 drops are
required to titrate the standard, the titrant may be off.
6.3 Sample Dilution: Dilution of the water sample or soil sample extract will be
necessary if the low range concentration is greater than 400 ppm. To
dilute by a factor of 10, mix in a 100-mL graduated cylinder 10 mL of water
sample or soil sample extract and 90 mL of dilution water. For the soil
extract, multiply the chloride concentration by 3 for the initial dilution, then
by 10 for the final dilution (factor of 30 for the total dilution).

Note 4: A dropper dispenses more accurately when held in a vertical position.
Note 5: White paper serves as a suitable background when observing solution colors.
Note 6: Silver nitrate degrades upon exposure to light and concentrates over time with
evaporation. Keep opaque container tightly sealed and away from heat when not in use.
Note 7: If per FM 5-551, the resistivity of dilution water is greater than 200,000 ohm-cm, the
concentration of chloride in dilution water is less than 5 ppm.
using a Hach Chloride Low Range Test Kit Model 8-P, 5-400 ppm, and
repeated measures of a 100-ppm aqueous standard, which was diluted by
a factor of three to be in the target range of the test kit. Six replicate
standards were tested, three with one lot number and three with a different
lot number of Chloride 2 indicator. The average bias was -2.5 ppm (-
2.5%).
7.2 Reproducibility: For a test material (sand, A-3) with an average chloride
concentration of 66 ppm, the multi-laboratory standard deviation of a
single test result has been found to be 12 ppm. Therefore, results of two
properly-conducted tests in different laboratories on the same material are
not expected to differ by more than 34 ppm.
330
F.4 Florida Method of Test for Sulfate in Soil and Water, FM 5-553
Florida Method of Test for Sulfate in Soil and Water

1. SCOPE
1.1 This method covers the determination of sulfate in soil and water using
either a screening approach based on a sulfate reagent system or an
analytical approach as found in the Standard Methods for the Examination of
Water and Wastewater, Section 4500-SO42- E or Section 4110 B(SMEWW).
1.2 Both the screening and analytical approaches use the reaction of water-
dissolved barium and sulfate to form an insoluble white precipitate of barium
sulfate. The concentration (turbidity) of the precipitate is proportional to the
sulfate concentration.
2. APPARATUS
2.1 Sulfate Reagent System: Sulfate, Pocket Colorimeter II Test Kit (Hach
catalog #5870029), 2 to 70 mg/L, or equivalent, may be used for sulfate
determinations. Additional glass sample cells with a 10-mL mark are needed
(for example, Hach catalog #2427606). Alternatively, the apparatus referred
to in the SMEWW may be used for sulfate determinations.
2.2 Analytical Balance: An analytical balance with a capacity of 2,000 g or more

and a resolution of 0.01 g or better.
2.3 Vacuum Filtration System: A vacuum filtration system is required. Such as

system includes a vacuum pump, 0.25” (6 mm) inner diameter flexible
vacuum hose, 300 mL filter funnel/holder for a 47-mm diameter filter, and a
1-liter side arm vacuum filtration flask or similar.
2.4 Other: Whatman 41 filter paper (or equivalent), glass or plastic funnel, two or
more 500-mL Erlenmeyer flasks, 47-mm diameter 0.45-micron pore size
mixed cellulose ester (MCE) membrane filters, transfer pipette (variable
pipette from 0.100 to 1.000 mL, capable of delivering increments of 0.01 mL)
and tips, timer (10 min, minimum), 100-mL graduated cylinder, and
disposable nitrile gloves.
331
3. REAGENTS
3.1 Barium chloride (BaCl2) (Note 1) and citric acid are required for the sulfate
test kit. Barium chloride and citric acid are supplied as SulfaVer 4 powder
pillows (for example, Hach catalog #21067-69). For the SMEWW method,
reagent purchasing and preparation information is provided in the method
procedures.
3.2 Sulfate standard solution, 1,000 ppm (mg/L) as SO42-, NIST-traceable.
3.3 Hydrochloric acid (HCl) (Note 1), ~37%, reagent grade.
3.4 Deionized or distilled water (dilution water) with a resistivity equal or greater
than 200,000 ohm-cm.
Note 1: Barium chloride and hydrochloric acid are hazardous materials. Refer to the chemical
material data safety (MSDS) sheets regarding the safe storage, handling, and disposal of
these hazardous materials. Dispose of hazardous material in accordance with federal,
state, and local mandates.
4. SAMPLES
been allowed to gravity drain for a short time. If the soil sample has excess
free moisture, place approximately 2.2 pounds (1 kg) of the soil on top of a
suitable sieve and cover with plastic. Allow the sample to drain for a
minimum of one hr. This step may be performed in the lab prior to testing. If
the soil sample is obtained from a heap that has been sitting for a long time,
take the sample from a depth below the weathered surface where the
moisture content appears to have stabilized. Avoid taking the sample from
near a weathered soil surface. If sampling from ground level, remove top 12”
(30 cm) to eliminate vegetation and debris before sampling. The soil sample
may be taken from underneath standing water, but excess water should not
be included with the sample. Soil samples should be placed in plastic or
plastic-lined bags. Squeeze the bag down snugly around the sample and
seal tightly to minimize contact with air.
4.2 Water Sampling: Water samples should be obtained from the main channel
of rivers and streams. Sampling from other bodies of water such as lakes or
ponds should be obtained from areas conducive to the capture of
representative samples. Care should be observed not to sample from
332
stagnant or pooled water, unless a structure will be placed in such an area.
Sample the water just below the surface to alleviate introduction of floating
debris, i.e. leaves, sticks, foam or trash. Fill the sample container to the top
to eliminate introducing air into the sample and tightly seal the lid. The
sample container shall be clean, at least 1 quart (1 liter) in size, and be
either glass or plastic with an airtight lid. When possible, submerge the
sample container below the surface of the water to completely fill and secure
lid underwater.
4.3 Transporting Samples: Maintain test samples in a cool dark area after
4.4 Storing Samples: Store water samples at or below 39C (4C). Care should
be taken to prevent freezing of the samples. Dry soil samples without delay
per section 5.2 and store dried soil at room temperature prior to analysis.
Analyze samples within seven (7) days.
5.1 Preparation of Water: Allow test sample to reach room temperature. If water
sample contains suspended solids or color, gravity filter the water sample
through a Whatman 41 filter (or equivalent), and if necessary, vacuum filter
the water sample through a 0.45-micron pore size membrane filter.

approximately four hr or until a constant mass is achieved. Sieve
through a No. 10 mesh (2 mm) sieve. Split the sample per ASTM C
702-98 to obtain approximately 400 g.
clean tray and dry under ambient conditions until a constant mass is
ASTM C 702-98 to obtain approximately 400 g.
333
C. Suspended particles (turbidity) or color will interfere with the sulfate
determination. If the extract is cloudy or colored due to the
suspension of fine particles, add a few (3-5) drops of hydrochloric
acid. Allow the acidified extract to stand for one to four hr as the
suspended particles settle (Note 3).
D. Filter by gravity the soil extract through a Whatman 41 filter into a

500-mL Erlenmeyer flask. Slowly decant into the filter the water layer
followed by the soil slurry from the sample extract and allow the liquid
to drain until the liquid stops dripping.
E. Filter by vacuum the soil extract through a 0.45-micron pore size

membrane filter (Note 4) into a filtration flask; repeat if necessary to
get a clear extract.
Note 2: Prepare separate soil extracts for sulfate and chloride analyses. If after 12 hr of settling
the both extracts are clear enough to filter, only one of two extracts needs to be filtered for
both sulfate and chloride analysis.
Note 3: More than a few drops of concentration hydrochloric acid will acidify the sample to less
than pH 2, upon which the sample becomes hazardous material.
6. TEST PROCEDURE
6.1 Low Range: For the SMEWW analytical approach, refer to the SMEWW
step-by-step instructions. For the screening approach using the sulfate test
kit (2-70 ppm), follow these instructions:
A. Check the Reagents and Glassware: Check that the SulfaVer 4

reagent powder pillows have not reached their expiration date. Check
and clean as necessary glass sample cells so that cells are free of
scratches, stains, deposits, or films that could affect light transmission
through the glass.
B. Prepare a Reagent Blank: Fill one sample cell with dilution water to
the 10-mL mark. Prepare a blank at least once per day on days in
which samples are tested.
C. Prepare a 30 ppm Check Standard: Add to a second sample cell 0.30

g of 1,000-ppm (mg/L) sulfate standard solution and fill the sample
cell to the 10-mL mark with dilution water. Prepare a check standard
at least once per day on days in which samples are tested.
334
D. Prepare Water Sample or Soil Sample Extract (Sample) Blank and
Sample: Fill the third and fourth cells with sample to the 10-mL mark.
Use one of these cells as a blank to exclude (zero) any remaining
turbidity or color.
E. React the Blanks, Standards, and Sample: Add one SulfaVer 4

powder pillow to each of these: the reagent blank, the 30 ppm check
standard, and only one of the two cells that contain the sample. Cap
all four sample cells and gently invert each cell 10 times to mix. Treat
each of the four sample cells the same, as the mixing action may
entrain air bubbles that will interfere with the measurement. Wipe the
glass cells clean of fingerprints. Wait for a 5-minute reaction period
with the sample cells undisturbed. Test the blanks, check standard,
and samples within 10 minutes of adding the SulfaVer 4 reagent.
F. Select a Range for the Measurements (see Note 5 and Figure 1):
Select a range on the photometer that gives the best results for the
reagent lot number. To switch from Range 1 to Range 2, press the
Menu key, then the Read key. Note the small arrow under the Range
label. Use the same range for blanks, check standard, and samples.
G. Measure the Check Standard: Place the reagent blank into the
sample cell holder and cap the holder. Press the photometer Zero
key. After a few seconds the digital display should read 0. Remove
the reagent blank and place the 30 ppm check standard into the
sample cell holder and cap the holder. Press the photometer Read
key. After a few seconds the digital display should read 30 (25 to 35).
Record the reading. If the measured concentration is less than 25
ppm or greater than 35 ppm, check first for scratched or dirty
glassware or a light leak through the cell holder; otherwise suspect
the potency of SulfaVer 4 reagent; purity or volume of sulfate
standard, or purity of dilution water. Troubleshoot and correct the
problem before testing the sample extract
H. Measure the Sample Extract: Place the extract blank into the sample
cell holder and cap the holder. Press the photometer Zero key. After a
few seconds the digital display should read 0. Remove the extract
blank and place the reacted extract into the sample cell holder and
cap the holder. Press the photometer Read key. After a few seconds
a reading will appear. If reading is 2 ppm or less, record reading as
below detection. Otherwise, record the photometer reading and
multiply the reading by 3 (dilution factor) to get the soil sulfate
concentration. Record the dilution factor and the soil sulfate
335
concentration in units of ppm.
6.2 Sample Dilution: Dilution of the test sample will be necessary if the sulfate
concentration is greater than the maximum limit of the photometer (digital
display will flash if reading is too high). To dilute by a factor of 10, mix in a
100-mL graduated cylinder 10 mL of sample and 90 mL of dilution water.
Repeat steps 6.1 E, F, and H for the sample. For the soil extract, multiply the
instrument reading by 3 for the initial dilution, then by 10 for the final dilution
(factor of 30 for the total dilution).
Note 5: The Hach photometer includes a pre-programmed calibration curve on each range, Range
1 and Range 2. To check the built-in calibration curves, use a series of check standards
from 0 to 70 ppm and plot the results. An example plot is shown below. Re-check the
curves if the photometer is repaired or replaced, for each new lot of SulfaVer 4 powder
pillows, and if a problem with the curve is suspected. Add in sequence to eight sample cells
0.00 g, 0.10 g, 0.20 g, 0.30 g, 0.40 g, 0.50 g, 0.60 g, and 0.70 g of 1,000 ppm (mg/L)
sulfate standard solution. Fill each sample cell to the 10-mL mark with dilution water to
obtain sulfate concentrations of 0, 10, 20, 30, 40, 50, 60, and 70 ppm. Test the calibration
curve standards as described in the above procedures (6.1 E, F, and G). The Hach
photometer also accepts a user-generated calibration curve as outlined in the instrument’s
instruction manual.
336
80
60
40
20
0
0 20 40 60 80
Standard Sulfate Concentration, ppm
Range 1 Lot 4120 Range 2 Lot 4120
Figure 1. Plot of a Hach photometer response to sulfate standard solutions. For this
photometer, the measured sulfate concentration was consistently higher on
Range 2 than on Range 1. For both Range 1 and Range 2, moreover, the
measured sulfate concentration was lower for SulfaVer 4 Lot 3303 than for
Lot 4120, even though both lots were within their expiration dates.
using a Hach Sulfate, Pocket Colorimeter II Test Kit and repeated measures
of a 100-ppm aqueous standard, which was diluted by a factor of three to be
in the target range of the test kit. Six replicate standards were tested, three
with one lot number and three with a different lot number of SulfaVer 4
reagent. The average bias was -0.5 ppm (-0.5%).
7.2 Reproducibility: For at test material (sand, A-3) with an average sulfate
concentration of 67 ppm, the multi-laboratory standard deviation of a single
test result has been found to be 16 ppm. Therefore, results of two properly-
conducted tests in different laboratories on the same material are not
337
Appendix G Audit Checklists
G.1 Checklist for Florida Method of Test for pH of Soil and Water
APPARATUS
 pH Meter and Electrode System: Portable or benchtop pH/mV meter with

automatic temperature compensation (ATC) and accuracies for mV: the greater
of ± 0.2 mV or ± 0.05%, for pH: ± 0.002 pH units, and for temperature: ±
0.3C, or better, a display of calibration data including %slope and offset; and
a refillable double-junction 3-in-1 combination electrode that includes a
silver/silver chloride (Ag/AgCl2) reference electrode, glass bulb indicating
electrode, and thermocouple with pH/ATC connectors that match the meter
inputs, and capable of measuring 0 to 14 pH from -5C to 100C with a
minimum resolution of 0.01 pH units (for example, Fisher Scientific pH/mV
meter kit catalog #13-636-AB150).
 Analytical Balance: An analytical balance with a capacity of 2,000 g or more

 100 and 250 mL beakers, wash bottle, glass stirring rods, 100-mL scoop, and
delicate glass cleaning tissues.
REAGENTS
 Standard buffered solutions for pH 4, 7, and 10.
 Electrode filling solution and electrode storage solution (refer to electrode

manual).
 Deionized or distilled water (dilution water) with a resistivity equal or greater

SAMPLES
 Storing Samples: If samples need to be stored, store water and soil samples at
or below 39F (4C). Care should be taken to prevent freezing of the samples.
Analyze samples within seven (7) days.
SAMPLE PREPARATION
 Preparation of Water: Allow test sample to reach room temperature.
338
 Preparation of Soil: Allow the soil sample and dilution water to reach
room temperature. Place 100 g of sample in a 250 mL beaker and then
add 100 mL of dilution water. Break up any clumps of soil. Stir
thoroughly the soil and water mixture for ~20 seconds every 10 minutes
over a 30-minute period.
TEST PROCEDURE
 Calibration of pH Meter and Electrode System: Pour ~50 mL of pH 4, pH 7,

and pH 10 standard buffer solutions into separate 100-mL beakers and allow
solutions to reach room temperature prior to calibration (Note 1). Remove pH
electrode from storage solution, rinse with dilution water and blot or dab dry
with tissue, uncover the fill hole, then place in pH 7 buffer solution for ~10
minutes prior to calibration. Calibrate the pH/mV meter and 3-in-1
combination electrode with three standard buffers according to the directions
in the manufacturer's instruction manual. When the calibration is complete,
record the calibration slope in %, offset in mV, and buffer temperature in C
(Note 2). The slope must be within 95% to 102%, the offset within ± 10 mV,
and the buffer temperatures within 25 ± 5C; if otherwise, follow the
instruction manual to clean, fill, equilibrate, re-calibrate, re-activate, or replace
the electrode so that the slope, offset, and temperatures are within these values.
 Care of Electrode: Check the electrode filling solution and keep filled to the fill
hole. Between any two pH measurements, rinse electrode with dilution water
and blot or dab the electrode tip dry with a delicate glass tissue (Note 3).
 Check of Calibration: Check the calibration by measuring the pH of each of the

standard buffers. The measured pH should be within 0.05 pH units of the
buffer pH at the test temperature. Buffer pH versus test temperature data are
typically tabulated on the buffer bottle. If the measured pH is outside 0.05 pH
units of the buffer pH at test temperature, take corrective steps and re-calibrate
the electrode (Note 4).
 Determination of pH: Insert the pH electrode in the buffer, water or soil sample
such that the portion of the electrode that contains the glass bulb,
thermocouple, and liquid junction is submersed, typically about ½” (1.3 cm) of
the electrode tip. Do not immerse the electrode to a depth greater than the
electrode’s internal reference solution, as the sample may penetrate into and
contaminate the reference solution. Very gently stir the electrode in the sample
(or alternately, gently move the sample beaker beneath the electrode) to assure
a pH representative of the bulk solution (Note 5). Wait until the meter’s stable
reading indicator stays on for 10 seconds or longer (Note 6). Remove the
electrode from the sample, rinse the tip with dilution water and blot or dab dry,
and return the electrode to the sample for a second pH measurement. Wait
339
until the meter’s stable reading indicator stays on for 10 seconds or longer.
Record the pH to the nearest hundredth pH units and sample temperature to the
nearest tenth C, for example, pH 7.00 @ 25.1C. After the second reading,
return the pH electrode to the pH 7 buffer and check that the electrode is still
within calibration.
 Storage of Electrode (Note 7): Store the pH electrode overnight or longer in its
storage bottle either in the electrode storage solution with the fill hole open in
pH 4 or 7 buffer with the fill hole closed. Do not store in dilution water as this
will have an adverse effect on electrode performance. Tap water or standard
buffer solutions are appropriate for short-term electrode storage.
Note 1: Always use fresh buffers for electrode calibration.
Note 2: For some meters, the slope and offset are presented separately for two buffers
(for example, pH 4 and pH 7, or pH 7 and pH 10). Record the slope and offset for
the range that includes the sample pH.
Note 3: With some 3-in-1 combination electrodes, the narrow space between the glass
bulb and the temperature sensor traps liquid, whether buffer, sample, or dilution
water. This trapped liquid can be transferred into the sample and may affect the
pH reading.
Note 4: If after a sample pH measurement the calibration check fails, suspect a partially
clogged liquid junction on the electrode. Add electrode filling solution to flush the
liquid junction, and then re-check calibration.
Note 5: Ideally, once the pH reading has stabilized, the reading will not change whether
gently stirred or unstirred. For low ionic strength solutions, however, the pH may
not stabilize if stirred. If the pH reading does not stabilize after ~ 5 min with
gentle stirring, obtain and record a stable reading with no stirring.
Note 6: Disable the “hold” feature of the pH meter.
Note 7: Store the electrode in accordance with manufacturer’s instructions if different

from this section.
340
G.2 Checklist for Florida Method of Test for Minimum Resistivity of Soil and Water
APPARATUS
 Resistivity Meter: Any four-pin terminal, null-balancing ohmmeter or

multimeter capable of four-wire resistance measurements from one to one
million ohms, either analog or digital (as examples, resistivity meters MC
Miller Model 400A, Nilsson Model 400, and Tinker & Rasor Model SR-2).
 Soil Box: Designed such that the cross-sectional area (cm2) of the sample,
with the box filled level, divided by the distance (cm) between the pins is
equal to 1 cm (for example, MC Miller catalog #37008).

 Other: Test leads, thermometer, flat spatula, mixing spoon, large non-corrosive
(glass, plastic, or stainless steel) bowl for mixing, 100-mL graduated cylinder,
squeeze bottle for cleaning, and disposable nitrile gloves.
REAGENTS
 Conductivity standard, sodium chloride, 250 µS/cm (for example, Fisher

Scientific catalog #22366032).
 Distilled or deionized water (hereafter referred to as dilution water) with a

resistivity of 200,000 ohm-cm or greater (Note 1).
Note 1: Deionized or distilled water stored in containers that are not airtight will over
time absorb ions from acidic and basic gases in the atmosphere. Absorbed ions
will lower the water’s resistivity.
SAMPLES
 Storing Samples: Store water samples at or below 39F (4C). Care should be
taken to prevent freezing of the samples. Analysis of a soil sample “as
received” is preferred. If, however, soil samples cannot be analyzed within ~ 1
day of receipt, dry soil samples per Section 5.2 and store dried soil at room
temperature prior to analysis. Analyze samples within seven (7) days.
SAMPLE PREPARATION
341
 Preparation of Water: Allow test sample to reach room temperature.
 Preparation of Soil:
Loose Granular Soils: Spread the sample in a thin layer on a clean tray
and dry under ambient conditions until a constant mass is achieved, or
dry in an oven at no higher than 140F (60C) for approximately four hr
or until a constant mass is achieved. Sieve through a No. 10 mesh (2
mm) sieve. Split the sample per ASTM C 702-98 to obtain
approximately 1,200 g.
Muck and Soils with Clay: Spread the sample in a thin layer on a clean
tray and dry under ambient conditions until a constant mass is achieved,
or dry in an oven at no higher than 140F (60C) for approximately
four hr or until a constant mass is achieved. Using a rawhide mallet or
other suitable device pulverize the sample and sieve through a No. 10
mesh (2 mm) sieve. Split the sample per ASTM C 702-98 to obtain
approximately 1,200 g.
TEST PROCEDURE
 Equipment Set-up: Rinse the soil box, mixing bowl, and utensils with dilution
water before starting and after completion of each test. Follow the
manufacturer's instructions for properly connecting the test leads between the
meter and the soil box. A standard soil box will have plates for application of
current at both ends and a pair of electrode pins for reading potential
evenly spaced between the plates (Figure 1).
 Determine Resistivity of “As-Received” Soil: If an “as received” soil

resistivity is requested, bring the soil sample and dilution water to room
temperature. Place 1,000 g of “as-received” soil sample into a large bowl.
342
Remove any debris (such as grass, roots, sticks or rocks, etc.). Fill the soil box
to the top with the soil sample, taking care to fill any voids, and strike off any
excess soil on the top of soil box. Connect the soil box to the resistivity meter
as instructed in the manufacturer’s instruction manual and determine the soil
resistivity. Record the “as-received” resistivity in ohm-cm. Proceed to Section
6.3 to determine minimum resistivity.
 Determine Minimum Resistivity of “As-Received” or of Dried, Sieved Soil:

Bring the soil sample and dilution water to room temperature. Place 1,000 g of
dried soil sample into a large bowl.
Add 100 mL of dilution water to the soil sample and mix thoroughly
(Note 2).
Fill the soil box to the top with the soil sample, taking care to fill any
voids, and strike off any excess soil on the top of soil box. Connect the
soil box to the resistivity meter as instructed in the manufacturer’s
instruction manual and determine the soil resistivity. Record the t ot al
volume of water used in mL and the resistivity in ohm-cm. Place soil
sample back into mixing bowl (Note 3).
Repeat steps 6.5 A and 6.5 B. The measured resistivity should decrease.
When the resistivity of the soil sample begins to increase, record the
t ot al volume of water used in mL and the higher resistivity, and
measure and record the temperature in C of the sample in the soil box.
Record the lowest resistivity obtained as the minimum resistivity and

the temperature of the sample. Report the minimum resistivity in ohm-cm
and the temperature in C, for example, 3,000 ohm-cm @ 25C.
 Determine Resistivity of a Small Soil Sample: If soil resistivity is requested for

a sample that is less than 1,000 g, complete steps 6.5 A and 6.5 B using
increments of 10% water instead of 100 mL (Note 4).
Note 2: In some soils, even a trace amount of salt will affect the resistivity reading.
Avoid transferring salt from hands to sample. For soils that are difficult to mix,
wear disposable gloves and mix soil with gloved hands.
Note 3: In many cases the minimum resistivity will occur after soil slurry is formed. As
water is added, mix thoroughly the soil slurry and pour the water from the slurry
into the box until the box is filled. If the water alone cannot fill the soil box, add
enough of the remaining soil to completely fill the box.
343
Note 4: A sample size of less than 1,000 g is not appropriate for select backfill (FDOT
material 092L).
BEST PRACTICES
 Determine Resistivity of Water: Refer to the manufacturer’s operation manual

for determining resistivity.
 Determine Resistivity of Conductivity Standard: Test the conductivity standard

at least once per quarter and after the resistivity meter has undergone repair or
replacement. Bring the conductivity standard to room temperature. Fill the soil
box to the top with conductivity standard. Connect the soil box to the resistivity
meter as instructed in the manufacturer’s instruction manual and verify that the
conductivity standard resistivity is 4,000 ohm-cm @ 25C. Refer to Table 1 for
the conductivity of the standard for testing at temperatures between 20C and
30C. Record the conductivity standard resistivity in ohm-cm and the
measurement temperature in C. If the resistivity at measurement temperature
is outside 5%, troubleshoot and correct the problem, then re-test. Rinse the
soil box well with dilution water after testing the conductivity standard.
Table 1 Temperature-Dependence of a 250 µS/cm NaCl Conductivity Standard

Temperature, C Resistivity, ohm-cm
20 4,440
21 4,350
22 4,250
23 4,170
24 4,080
25 4,000
26 3,920
27 3,850
28 3,780
29 3,710
30 3,650
344
G.3 Checklist for Florida Method of Test for Chloride in Soil and Water
APPARATUS
 Chloride Reagent System: Chloride Low Range Test Kit Model 8-P, 5-400 ppm
(Hach catalog #144001) and Chloride High Range Test Kit Model CD-51
(Hach catalog #208601), or equivalent, may be used for chloride
determinations. For the Hach Model 8-P test kit, additional glass mixing bottles
with a 23-mL mark are needed (for example, Hach catalog #232706).
Alternatively, the laboratory instruments referred to in the SMEWW may be
used for chloride determination.

 Vacuum Filtration System: A vacuum filtration system is required. Such as

system includes a vacuum pump, 0.25” (6 mm) inner diameter flexible vacuum
hose, 300-mL filter funnel/holder for a 47-mm diameter filter, and a 1-liter side
arm vacuum filtration flask or similar.
 Other: Whatman 41 filter paper (or equivalent), glass or plastic funnel, two or
more 500-mL Erlenmeyer flasks, 47-mm diameter 0.45-micron pore size mixed
cellulose ester (MCE) membrane filters, 0.10-g measuring spoon, small
ceramic dish, a transfer pipette (variable from 0.100 to 1.000 mL, capable of
delivering increments of 0.01 mL) and tips, 100-mL graduated cylinder, and
disposable nitrile gloves.
REAGENTS
 Potassium dichromate (K2Cr2O7), sodium bicarbonate and silver nitrate

(AgNO3) titrant are required for the chloride test kit (Note 1). The potassium
dichromate and sodium bicarbonate are supplied as Chloride 2 powder pillows
(for example, Hach catalog #105766). For the SMEWW method, reagent
purchasing and preparation information is provided in the method procedures.
 Chloride standard solution, 1,000 ppm (mg/L) as Cl-, NIST-traceable.
 Ferric sulfate, hydrated reagent grade (for example, Fisher Scientific catalog
#S25322 or #S25322A).
345
Note 1: Potassium chromate and silver nitrate are hazardous materials. Refer to the
chemical safety data sheets (SDS) regarding the safe storage, handling, and
disposal of these hazardous materials. Dispose of hazardous material in
accordance with federal, state, and local mandates.
SAMPLES
 Storing Samples: Store water samples at or below 39F (4C). Care should be
taken to prevent freezing of the samples. Dry soil samples without delay per
Section 5.2 and store dried soil at room temperature prior to analysis. Analyze
samples within seven (7) days.
SAMPLE PREPARATION
 Preparation of Water: Allow test sample to reach room temperature. If water

sample contains suspended solids or color, gravity filter the water through a
Whatman 41 filter (or equivalent), and if necessary, vacuum filter the water
sample through a 0.45-micron pore size MCE membrane filter.
dry in an oven at 140F (60C) for approximately four hr or until a
constant mass is achieved. Sieve through a No. 10 mesh (2 mm) sieve.
Split the sample per ASTM C 702-98 to obtain approximately 400 g.
or dry in an oven at no higher than 140F (60C) for approximately four
hr or until a constant mass is achieved. Using a rawhide mallet or other
suitable device pulverize the sample and sieve through a No. 10 mesh (2
mm) sieve. Split the sample per ASTM C 702-98 to obtain
approximately 400 g.
 Preparation of Soil Extract: Place 100 grams of dried soil in a 500 mL

Erlenmeyer flask (Note 2). Add 300 mL of dilution water, stopper, and shake
vigorously for 20 seconds. Let stand for one hr and repeat agitation. Let stand
for a minimum of 12 hr.
Suspended particles (turbidity) or color may interfere with the chloride
346
determination. If the extract is cloudy or colored due to the suspension of fine
particles, add ferric sulfate. Dissolve in a small ceramic dish ~0.10 g of ferric
sulfate into 1 mL of water and pipette the dissolved ferric sulfate into the
extract. Allow the sample to stand for one to four hr as the suspended particles
settle.
Filter by gravity the soil extract through a Whatman 41 filter (or equivalent)
into a 500-mL Erlenmeyer flask. Slowly decant into the filter the water layer
followed by the soil slurry from the sample extract and allow the liquid to drain
until the liquid stops dripping.
Filter by vacuum the soil extract through a 0.45-micron pore size membrane
filter (Note 3) into a clean 1-liter vacuum filtration flask; repeat if necessary to
get a clear extract.
Note 2: Prepare separate soil extracts for sulfate and chloride analyses. If after 12 hr of
settling the both extracts are clear enough to filter, only one of two extracts needs
to be filtered for both sulfate and chloride analysis.
TEST PROCEDURE
 Low Range: For the SMEWW analytical approach or an equivalent test kit,
follow the kit instructions and include preparation and testing of a blank and
30-ppm check standard. For the Hach Model 8-P chloride test kit (5-400 ppm),
follow these instructions:
Check the Reagents: Check that the silver chloride titrant and Chloride 2
Indicator powder pillows have not reached their expiration date.
Prepare a Reagent Blank: Fill one mixing bottle with dilution water to the
23-mL mark. Prepare a blank at least once per day on days in which
samples are tested.
Prepare a 30 mg/L (30 ppm) Check Standard: Add to a second mixing

bottle 0.690 g of 1,000 mg/L chloride standard solution and fill with
dilution water to the 23-mL mark. Prepare a check standard at least once
per day on days in which samples are tested.
Prepare a Water Sample or Soil Sample Extract: Fill a third mixing bottle
with the water sample or soil sample extract to the 23-mL mark.
347
Determine the Blank Chloride Concentration: Add one Chloride 2
Indicator powder pillow (potassium dichromate/sodium bicarbonate) to
mixing bottle that contains the blank. Gently swirl the bottle to achieve a
uniform yellow color. Add one drop of 0.0493 N silver nitrate to the blank
and gently swirl the bottle to mix. Each drop is equal to ~5 mg/L (5 ppm)
of chloride (Note 4). With one drop of silver nitrate the color of the blank
should turn from yellow to red-brown; this red-brown color will serve as a
guide (Note 5). If two or more drops of silver nitrate are necessary to
titrate the dilution water to a red-brown color, then either the dilution
water is contaminated with chloride or the titrant has gone bad (Notes 6
and 7).
Sample Calculation: 1 drop x 5 ppm/drop = 5 ppm per blank
Determine the Check Standard Chloride Concentration: Add one Chloride

2 Indicator powder pillow (potassium dichromate/sodium bicarbonate) to
mixing bottle that contains the check standard. Add one drop at a time of
0.0493 N silver nitrate and gently swirl the bottle to mix. The check
standard should turn from yellow to red-brown after the addition of 6 to 8
drops. Record the total number of drops and subtract from this number
one drop for the blank. Multiply the net number of drops by 5 ppm per
drop. The check standard concentration should fall between 25 ppm and
35 ppm. If the concentration is less than 25 ppm or greater than 35 ppm
the volume of added chloride standard or the drop size of silver nitrate
may be off; otherwise the concentrations of either the chloride check
standard or the silver nitrate titrant may be off. In either case, prepare a
new check standard and re-test.
Sample Calculation: 7 drops x 5 ppm/drop = 35 ppm – 5 ppm blank = 30

ppm for check standard
Determine the Water Sample or Soil Sample Extract Chloride

Concentration (0-60 ppm): Add one Chloride 2 Indicator powder pillow
(potassium dichromate/sodium bicarbonate) to mixing bottle that contains
the water sample or soil sample extract. Gently swirl the bottle to achieve
a uniform yellow color. Add one drop at a time of 0.0493 N silver nitrate
until the color of the extract changes from yellow to red-brown. Record
the total number of drops and multiply each drop by 5 ppm of chloride per
drop, then subtract from this concentration the 5 ppm of chloride for the
blank. If this result is zero, record that the chloride concentration of the
sample as below detection. Otherwise, multiply the chloride concentration
by the dilution factor, which may be 1 for a water sample but 3 to account
for the initial dilution of a soil sample (100 grams of soil to 300 mL
water). Report this concentration in ppm.
348
ppm x 3 dilution factor for soil extract = 30 ppm soil concentration
Determine the Water Sample or Soil Sample Extract Chloride

Concentration (60-400 ppm): If the water sample or soil sample extract
has not changed color after the addition of 13 drops (60 ppm), stop the
titration. Prepare a new sample extract: fill a clean mixing bottle to the
10-mL mark. Add one drop at a time of 0.0493 N silver nitrate until the
color of the extract changes from yellow to red-brown. Record the total
number of drops and multiply each drop by 20 ppm of chloride per drop,
then subtract from this concentration the 5 ppm of chloride for the blank.
Otherwise, multiply the chloride concentration by the dilution factor,
which may be 1 for a water sample but 3 to account for the initial dilution
of a soil sample (100 grams of soil to 300 mL water). Report this
concentration in ppm.

ppm x 3 dilution factor for soil extract = 165 ppm soil concentration
 High Range: If water or soil has a high chloride concentration (for example,
seawater has a chloride concentration of ~19,000 ppm) use the Hach Model
CD-51 chloride test kit (500-100,000) (or equivalent) and follow the kit
instructions. For a check standard, fill the measuring tube with a 1000-ppm
chloride standard and continue with the test instructions. The standard should
be titrated from yellow to red-brown with 1-3 drops of 1.88-N silver nitrate
titrant (500 ppm per drop). If more than 3 drops are required to titrate the
standard, the titrant may be off.
 Sample Dilution: Dilution of the water sample or soil sample extract will be
necessary if the low range concentration is greater than 400 ppm. To dilute by a
factor of 10, mix in a 100-mL graduated cylinder 10 mL of water sample or soil
sample extract and 90 mL of dilution water. For the soil extract, multiply the
chloride concentration by 3 for the initial dilution, then by 10 for the final
dilution (factor of 30 for the total dilution).
Sample Calculation: 3 drops x 20 ppm/drop = 60 ppm – 5 ppm blank =

55 ppm x 30 dilution factor for soil extract = 1650 ppm soil
concentration
Note 4: A dropper dispenses more accurately when held in a vertical position.
349
Note 5: White paper serves as a suitable background when observing solution colors.
Note 6: Silver nitrate degrades upon exposure to light and concentrates over time with
evaporation. Keep opaque container tightly sealed and away from heat when not
in use.
Note 7: If per FM 5-551, the resistivity of dilution water is greater than 200,000 ohm-cm,
the concentration of chloride in dilution water is less than 5 ppm.
350
G.4 Checklist for Florida Method of Test for Sulfate in Soil and Water
APPARATUS
 Sulfate Reagent System: Sulfate, Pocket Colorimeter II Test Kit (Hach catalog
#5870029), 2 to 70 mg/L, or equivalent, may be used for sulfate
determinations. Additional glass sample cells with a 10-mL mark are needed
(for example, Hach catalog #2427606). Alternatively, the apparatus referred to
in the SMEWW may be used for sulfate determinations.

 Vacuum Filtration System: A vacuum filtration system is required. Such as

system includes a vacuum pump, 0.25” (6 mm) inner diameter flexible vacuum
hose, 300 mL filter funnel/holder for a 47-mm diameter filter, and a 1-liter side
arm vacuum filtration flask or similar.
 Other: Whatman 41 filter paper (or equivalent), glass or plastic funnel, two or
more 500-mL Erlenmeyer flasks, 47-mm diameter 0.45-micron pore size mixed
cellulose ester (MCE) membrane filters, transfer pipette (variable pipette from
0.100 to 1.000 mL, capable of delivering increments of 0.01 mL) and tips,
timer (10 min, minimum), 100-mL graduated cylinder, and disposable nitrile
gloves.
REAGENTS
 Barium chloride (BaCl2) (Note 1) and citric acid are required for the sulfate
test kit. Barium chloride and citric acid are supplied as SulfaVer 4 powder
pillows (for example, Hach catalog #21067-69). For the SMEWW method,
reagent purchasing and preparation information is provided in the method
procedures.
 Sulfate standard solution, 1,000 ppm (mg/L) as SO42-, NIST-traceable.
 Hydrochloric acid (HCl) (Note 1), ~37%, reagent grade.

Note 1: Barium chloride and hydrochloric acid are hazardous materials. Refer to the
chemical material data safety (MSDS) sheets regarding the safe storage, handling,
and disposal of these hazardous materials. Dispose of hazardous material in
accordance with federal, state, and local mandates.
351
SAMPLES
 Storing Samples: Store water samples at or below 39C (4C). Care should be
taken to prevent freezing of the samples. Dry soil samples without delay per
section 5.2 and store dried soil at room temperature prior to analysis. Analyze
samples within seven (7) days.
SAMPLE PREPARATION
 Preparation of Water: Allow test sample to reach room temperature. If water

sample contains suspended solids or color, gravity filter the water sample
through a Whatman 41 filter (or equivalent), and if necessary, vacuum filter the
water sample through a 0.45-micron pore size membrane filter.
dry in an oven at no higher than 140F (60C) for approximately four hr
or until a constant mass is achieved. Sieve through a No. 10 mesh (2 mm)
sieve. Split the sample per ASTM C 702-98 to obtain approximately 400
g.
or dry in an oven at no higher than 140F (60C) for approximately four
hr or until a constant mass is achieved. Using a rawhide mallet or other
suitable device pulverize the sample and sieve through a No. 10 mesh (2
mm) sieve. Split the sample per ASTM C 702-98 to obtain approximately
400 g.
 Preparation of Soil Extract: Place 100 grams of dried soil in a 500-mL

Erlenmeyer flask (Note 2). Add 300 mL of dilution water, stopper, and shake
vigorously for 20 seconds. Let stand for one hr and repeat agitation. Let stand
for a minimum of 12 hr.
Suspended particles (turbidity) or color will interfere with the sulfate

determination. If the extract is cloudy or colored due to the suspension of fine
particles, add a few (3-5) drops of hydrochloric acid. Allow the acidified
extract to stand for one to four hr as the suspended particles settle (Note 3).
Filter by gravity the soil extract through a Whatman 41 filter into a 500-mL
Erlenmeyer flask. Slowly decant into the filter the water layer followed by the
soil slurry from the sample extract and allow the liquid to drain until the liquid
stops dripping.
352
Filter by vacuum the soil extract through a 0.45-micron pore size membrane
filter (Note 4) into a filtration flask; repeat if necessary to get a clear extract.
Note 2: Prepare separate soil extracts for sulfate and chloride analyses. If after 12 hr of
settling the both extracts are clear enough to filter, only one of two extracts needs
to be filtered for both sulfate and chloride analysis.
Note 3: More than a few drops of concentration hydrochloric acid will acidify the sample
to less than pH 2, upon which the sample becomes hazardous material.
TEST PROCEDURE
 Low Range: For the SMEWW analytical approach, refer to the SMEWW step-
by-step instructions. For the screening approach using the sulfate test kit (2-70
ppm), follow these instructions:
Check the Reagents and Glassware: Check that the SulfaVer 4 reagent
powder pillows have not reached their expiration date. Check and clean
as necessary glass sample cells so that cells are free of scratches, stains,
deposits, or films that could affect light transmission through the glass.
Prepare a Reagent Blank: Fill one sample cell with dilution water to the
10-mL mark. Prepare a blank at least once per day on days in which
samples are tested.
Prepare a 30 ppm Check Standard: Add to a second sample cell 0.30 g of

1,000-ppm (mg/L) sulfate standard solution and fill the sample cell to the
10-mL mark with dilution water. Prepare a check standard at least once
per day on days in which samples are tested.
Prepare Water Sample or Soil Sample Extract (Sample) Blank and

Sample: Fill the third and fourth cells with sample to the 10-mL mark.
Use one of these cells as a blank to exclude (zero) any remaining
turbidity or color.
React the Blanks, Standards, and Sample: Add one SulfaVer 4 powder
pillow to each of these: the reagent blank, the 30 ppm check standard,
and only one of the two cells that contain the sample. Cap all four sample
cells and gently invert each cell 10 times to mix. Treat each of the four
sample cells the same, as the mixing action may entrain air bubbles that
will interfere with the measurement. Wipe the glass cells clean of
fingerprints. Wait for a 5-minute reaction period with the sample cells
undisturbed. Test the blanks, check standard, and samples within 10
minutes of adding the SulfaVer 4 reagent.
353
Select a Range for the Measurements (see Note 5 and Figure 1): Select a
range on the photometer that gives the best results for the reagent lot
number. To switch from Range 1 to Range 2, press the Menu key, then
the Read key. Note the small arrow under the Range label. Use the same
range for blanks, check standard, and samples.
Measure the Check Standard: Place the reagent blank into the sample cell
holder and cap the holder. Press the photometer Zero key. After a few
seconds the digital display should read 0. Remove the reagent blank and
place the 30 ppm check standard into the sample cell holder and cap the
holder. Press the photometer Read key. After a few seconds the digital
display should read 30 (25 to 35). Record the reading. If the measured
concentration is less than 25 ppm or greater than 35 ppm, check first for
scratched or dirty glassware or a light leak through the cell holder;
otherwise suspect the potency of SulfaVer 4 reagent; purity or volume of
sulfate standard, or purity of dilution water. Troubleshoot and correct the
problem before testing the sample extract
Measure the Sample Extract: Place the extract blank into the sample cell
holder and cap the holder. Press the photometer Zero key. After a few
seconds the digital display should read 0. Remove the extract blank and
place the reacted extract into the sample cell holder and cap the holder.
Press the photometer Read key. After a few seconds a reading will
appear. If reading is 2 ppm or less, record reading as below detection.
Otherwise, record the photometer reading and multiply the reading by 3
(dilution factor) to get the soil sulfate concentration. Record the dilution
factor and the soil sulfate concentration in units of ppm.
 Sample Dilution: Dilution of the test sample will be necessary if the sulfate
concentration is greater than the maximum limit of the photometer (digital
display will flash if reading is too high). To dilute by a factor of 10, mix in a
100-mL graduated cylinder 10 mL of sample and 90 mL of dilution water.
Repeat steps 6.1 E, F, and H for the sample. For the soil extract, multiply the
instrument reading by 3 for the initial dilution, then by 10 for the final dilution
(factor of 30 for the total dilution).
Note 5: The Hach photometer includes a pre-programmed calibration curve on each

range, Range 1 and Range 2. To check the built-in calibration curves, use a series
of check standards from 0 to 70 ppm and plot the results. An example plot is
shown below. Re-check the curves if the photometer is repaired or replaced, for
each new lot of SulfaVer 4 powder pillows, and if a problem with the curve is
suspected. Add in sequence to eight sample cells 0.00 g, 0.10 g, 0.20 g, 0.30 g,
0.40 g, 0.50 g, 0.60 g, and 0.70 g of 1,000 ppm (mg/L) sulfate standard solution.
Fill each sample cell to the 10-mL mark with dilution water to obtain sulfate
concentrations of 0, 10, 20, 30, 40, 50, 60, and 70 ppm. Test the calibration curve
354
standards as described in the above procedures (6.1 E, F, and G). The Hach
photometer also accepts a user-generated calibration curve as outlined in the
instrument’s instruction manual.
80
60
40
20
0
0 20 40 60 80
Standard Sulfate Concentration, ppm
Figure 1. Plot of a Hach photometer response to sulfate standard solutions. For this photometer,
the measured sulfate concentration was consistently higher on Range 2 than on Range
1. For both Range 1 and Range 2, moreover, the measured sulfate concentration was
lower for SulfaVer 4 Lot 3303 than for Lot 4120, even though both lots were within
their expiration dates.
355

FDOT BDV25 977 03 RPT

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

FDOT BDV25 977 03 RPT

Uploaded by

Copyright:

Available Formats

Distribution of Chloride, pH, Resistivity, and

Sulfate Levels in Backfill for Mechanically-

Dr. Manjriker Gunaratne, Principal Investigator

SYMBOL WHEN YOU KNOW MULTIPLY BY TO FIND SYMBOL

SYMBOL WHEN YOU KNOW MULTIPLY BY TO FIND SYMBOL

SYMBOL WHEN YOU KNOW MULTIPLY BY TO FIND SYMBOL

SYMBOL WHEN YOU KNOW MULTIPLY BY TO FIND SYMBOL

SYMBOL WHEN YOU KNOW MULTIPLY BY TO FIND SYMBOL

4. Title and Subtitle 5. Report Date

7. Author(s) 8. Performing Organization Report No.

15. Supplementary Notes

 Perform a Florida-wide inter-laboratory study of the revised FMs. A Florida-wide inter-

Construction of mechanically-stabilized earth (MSE) walls in metropolitan areas within FDOT’s

1.2 Scope and Objectives

1.2.1 Literature review

1.2.2 Trends in backfill properties

1.2.3 Single-laboratory contributions to method reproducibility

Single-laboratory contributions to method reproducibility of factors that relate to method field

1.2.4 Replicate studies

1.2.5 Multi-laboratory contributions to method reproducibility

Multi-laboratory contributions to method reproducibility were estimated pre- and post-method

1.2.6 Geochemical modeling

1.2.7 Mine Information

Table 1-1 Mine Information

 Trends in backfill properties were placed in Chapter 4 Trends in Select Backfill;

 Single-laboratory contributions to method reproducibility were split into four chapters,

 Multi-laboratory contributions to method reproducibility were grouped into Chapter 6

 Geochemical modeling was organized as Chapter 12 Geotechnical Modeling of Ion

2.1.1 Features of an MSE wall and properties of select backfill

2.1.2 Corrosion of MSE wall reinforcement

2.2 Quality Assurance (QA) Plans for Construction Materials

 Assimilate data and information to determine relevant construction parameters;

Certify material Reject material

Transport material to MSE

Figure 2-4. Example operating characteristic (OC) for pH N = 2.

2.2.2.1 Process control for corrosion AQC

2.2.2.2 Product Acceptance for Corrosion AQC

2.2.3 Sampling of select backfill

 Obtain a representative sample;

 Design a sampling plan for specific case under consideration;

For pits or banks:

 Identify different stratum and select samples from each stratum;

 Obtain a representative sample;

2.2.4 Good laboratory practices

 A workspace that is clean, comfortable, and safe;

2.2.5 Other QA/QC activities

Table 3-1 Selected Test Methods for Soil Electrochemical Properties

3.2.1 Theory of measurement

3.2.1.1 Acidity, activity, and ionic strength

3.2.1.2 pH measurement system and electrode design

Indicating electrode Reference electrode

Glass bulb Liquid junction

Figure 3-1. Schematic of a pH measurement system.

3.2.1.3 Liquid junction

3.2.1.4 Temperature error

The temperature-dependence of pH measurements has at least three sources: (1) the

Table 3-2 Influence of Temperature on the pH of Pure Water

3.2.1.5 Sodium (alkaline) and acid errors

3.2.1.7 Stirring error

3.2.1.8 Ionic strength

An electrochemical pH measurement is a challenge in low ionic strength solutions. In pure water

3.2.1.9 Carbonate Equilibria

3.2.1.10 Suspension error