The Pennsylvania State University

The Graduate School

College of Engineering

THE EXPANSIVE EFFECTS OF CONCENTRATED PYRITIC ZONES

IN SHALES OF THE MARCELLUS FORMATION

A Dissertation in

Civil Engineering

by

Shad E. Hoover

© 2008 Shad E. Hoover

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

May 2008

The dissertation of Shad E. Hoover was reviewed and approved* by the following:

Mian C. Wang Professor of Civil and Environmental Engineering Dissertation Adviser Chair of Committee

Brian A. Dempsey Professor of Civil and Environmental Engineering

Angelica M. Palomino Assistant Professor of Civil and Environmental Engineering

Susan Trolier-McKinstry Professor of Ceramic Science and Engineering

David Lehmann Professional Geologist – Environmental Resources Management Special Member

Peggy A. Johnson Professor of Civil and Environmental Engineering Head of the Department of Civil and Environmental Engineering

*Signatures are on file in the Graduate School.

ii ABSTRACT

Expansive pyritic shales have caused untold amounts of damage to civil infrastructure throughout the world. The traditional characterization of potentially expansive pyritic shales only considers the presence of finely disseminated microscopic pyrite and references to case histories. The research presented in this paper shows that concentrated pyritic zones can have a significant impact on establishing microenvironments that are conducive to the production of heave inducing sulfates.

Oxidation tests are conducted in a laboratory environment to assess the potential to form heave inducing sulfates through swell measurements and geochemical markers.

Simple swell modeling is simulated with the PHREEQC geochemical computer program and a regression analysis of shale with significant gypsum infilling verifies calcite and pyrite concentrations. A detailed case history of micropile underpinning over expansive pyritic shales highlights the challenges associated with extensive structural remediation.

The O2 diffusion controlled oxidation process in calcareous shales with finely disseminated pyrite is impractically slow under intense conditions and does not adequately explain how gypsum infilling can occur over time periods of less than a decade. Geochemical laboratory testing and theoretical modeling suggest that an acidic or low pH environment is not possible with a significant presence of calcite with zones of availability. Gypsum is most likely to crystallize in a low pH environment when highly oxidative conditions are present. The research shows that microenvironments within highly concentrated pyritic zones more adequately explain the acidic conditions necessary that lead to accelerated oxidation. Acidic capillary pore water, which is not subject to flushing by a fluctuating or flowing water table through the vadose zone,

iii influences the calcareous microfractures and discontinuities resulting in the crystallization of sulfates.

iv TABLE OF CONTENTS

LIST OF TABLES ...... viii

LIST OF FIGURES ...... x

LIST OF MATHGRAMS ...... xvi

ACKNOWLEDGEMENTS ...... xvii

Chapter 1. INTRODUCTION ...... 1 1.1 General ...... 1 1.2 Microenvironmental Considerations ...... 3 1.3 Thesis Organization ...... 5

Chapter 2. LITERATURE REVIEW ...... 8 2.1 Introduction ...... 8 2.2 Pyritic Bedrock Geology...... 10 2.3 Pyrite Oxidation and Sulfate Precipitation Chemistry ...... 15 2.4 Weathering of Pyritic Shales ...... 20 2.5 Identification of Potentially Expansive Pyritic Shales ...... 21 2.6 Pyritic Sulfur Identification Techniques ...... 22 2.6.1 Munsell Color Guidelines ...... 23 2.6.2 Static Laboratory Testing Techniques ...... 24 2.6.3 Kinetic Laboratory Testing Techniques ...... 26 2.7 Laboratory Testing Methods – Swell Test ...... 27 2.8 Remnant Stresses and Horizontal Fracturing ...... 28

Chapter 3. CHEMICAL AND PHYSICAL EXPLANATIONS ...... 32 3.1 Introduction ...... 32 3.2 Materials and Methods ...... 33 3.2.1 Experiments Using 30% H2O2 ...... 33 3.2.2 Experiments Using 10% H2O2 ...... 34 3.3 Results ...... 36 3.3.1 Experiments Using 30% H2O2 ...... 36 3.3.2 Experiments Using 10% H2O2 ...... 39 3.4 Microscopic Observations ...... 41 3.5 Conclusions ...... 43

Chapter 4. PHREEQC HYDROGEOCHEMICAL TRANSPORT MODEL ...... 45 4.1 Introduction ...... 45 4.2 Irreversible Reaction Model ...... 46 4.2.1 Introduction ...... 46 4.3 Input Parameters ...... 48 4.3.1 First Run (0.1% S2 and 5% CaCO3 ...... 48 4.3.2 Second Run (0.5% S2 and 5% CaCO3 ...... 48

v 4.3.3 Third Run (Concentrated FeS2 and 5% CaCO3 ...... 48 4.4 Results ...... 49 4.4.1 Moles in Assemblage and Volume Change ...... 49 4.4.2 Swell Model ...... 51 4.5 Conclusions ...... 53

Chapter 5. EMPIRICAL REGRESSION MODEL ...... 59 5.1 Introduction ...... 59 5.2 Observational Data ...... 59 5.3 Experimental Data ...... 65 5.3.1 Image Analysis...... 65 5.3.2 Measurements ...... 68 5.4 Regression Analysis ...... 69 5.5 Conclusions ...... 70

Chapter 6. GEOTECHNICAL LABORATORY STUDIES ...... 75 6.1 Introduction ...... 75 6.2 Materials and Methods ...... 76 6.2.1 Controlled Experiments with Bacterial Oxidation ...... 76 6.2.1.1 Bacterial Preparation Procedure ...... 76 6.2.1.2 Swell Experiment Procedure ...... 78 6.2.2 Swell Experiments Using Kinetic Oxidation Techniques .....81 6.2.2.1 Swell Experiments Procedure ...... 81 6.3 Results ...... 87 6.3.1 Controlled Experiments with Bacterial Oxidation ...... 87 6.3.2 Swell Experiments Using Kinetic Oxidation Techniques .....89 6.4 Conclusions ...... 96 6.4.1 Controlled Experiments with Bacterial Oxidation ...... 96 6.4.2 Swell Experiments Using Kinetic Oxidation Techniques .....97

Chapter 7. CASE HISTORY: MICROPILE UNDERPINNING OVER EXPANSIVE PYRITIC SHALES ...... 100 7.1 Introduction ...... 100 7.2 Shale Expansion History ...... 105 7.3 Field and Laboratory Investigations ...... 106 7.3.1 Field Investigation ...... 106 7.3.2 Laboratory Testing ...... 109 7.4 Observations During Demolition and Overexcavation ...... 113 7.4.1 Demolition and Removal of Floor Slab ...... 113 7.4.2 Swelled Shale Observations ...... 114 7.5 Micropile Design ...... 115 7.5.1 Structural Loading Information ...... 115 7.5.2 Design Method 1 – Telephone Room ...... 117 7.5.3 Design Method 2 – Corridor ...... 121 7.5.4 Design Method 3 – Isolated Spread Footings ...... 124 7.6 Construction Considerations ...... 133

vi 7.6.1 Methods and Equipment ...... 133 7.6.2 As-Built Challenges ...... 134 7.7 Conclusions ...... 136

Chapter 8. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS ...... 137 8.1 Summary ...... 137 8.2 Conclusions ...... 138 8.3 Recommendations ...... 139 8.3.1 Changes to the State-of-Practice ...... 139 8.3.2 Future Research ...... 140

Appendix A: VISUAL IDENTIFICATION OF GEOMATERIALS ...... 141

Appendix B: CHEMICAL TEST PROCEDURES ...... 143

Appendix C: OXIDATION EXPERIMENTS DATA ...... 152

Appendix D: VISUAL MINTEQ EVALUATIONS FOR 30% H2O2 EXPERIMENTS ...... 156

Appendix E: X-RAY DIFFRACTION OF NODULE ...... 158

Appendix F: PHREEQC INPUT FILES FOR IRREVERSIBLE REACTIONS ...... 160

Appendix G: PHREEQC GRAPHICAL OUTPUT FOR IRREVERSIBLE REACTIONS ...... 164

Appendix H: VOLUME AND CELL HEIGHT CHANGE SPREADSHEETS FOR IRREVERSIBLE REACTIONS ...... 168

Appendix I: DIGITAL IMAGES OF IMAGE J ANALYSIS ...... 173

Appendix J: SPREADSHEETS OF ALL IMAGE J MEASUREMENTS ...... 180

Appendix K: ACIDITHIOBACILLUS FERROOXIDANS GROWTH MEDIUM ...... 183

Appendix L: ILLITE SHALE GRAIN SIZE ANALYSIS ...... 185

REFERENCES ...... 187

vii LIST OF TABLES

Table 2.1 Molar Volumes, densities and solubilities of key minerals ...... 19

Table 5.1 Image J Average Measurements for Sample S1 (PSSC) ...... 68

Table 5.2 Image J Average Measurements for Sample S2 (PSJ)...... 68

Table 6.1 Grain-size distribution of illite shale fragments ...... 78

Table 6.2 Modeled groundwater chemistry for controlled experiments with bacterial oxidation...... 81

Table 6.3 Particle size distribution requirement of pyritic shale using kinetic oxidation techniques ...... 83

Table 6.4 Sulfate anion and pH analytic results on effluent for homogeneous (B) and concentrated (A) pyrite mixtures ...... 88

Table 6.5 Forms of sulfur and calcium carbonate test results on shale fragments with (A2) and without (A1) obvious signs of macroscopic pyrite as determined with a hand lens ...... 90

Table 7.1 Comparison between forms of sulfur and chromium reducible sulfur testing methods on control samples provided to Mahaffey Laboratories, Ltd...... 112

Table C.1 Total Sulfur Analysis on shale samples from 10% H2O2 Oxidation Experiments ...... 153

Table C.2 Experimental Data from 30% H2O2 Oxidation Experiments ...... 154

Table C.3 Experimental Data for 10% H2O2 Oxidation Experiments ...... 155

Table D.1 Visual MINTEQ evaluations for H2O2 Oxidation Experiments ...... 157

Table H.1 Volume change calculations for 0.1% S2 and 5% CaCO3 ...... 169

Table H.2 Volume change calculations for 0.5% S2 and 5% CaCO3 ...... 170

Table H.3 Volume change calculations for concentrated FeS2 and 5% CaCO3 ...... 171

Table H.4 Cell height change calculations for varying concentrations of S2 and 5% CaCO3...... 172

viii

Table J.1 Measurements for cross-section S1 (PSSC) ...... 181

Table J.2 Measurements for cross-section S2 (PSJ) ...... 182

ix LIST OF FIGURES

Figure 1.1 Expansive pyritic shale model as a step-wise progression during a typical construction project. (a) Elevated water table with overburden materials prior to construction. (b) Water table is lowered and overburden removed as part of the excavation process resulting in increased fracturing and oxidation of available pyrite. (c) Capillary water evaporates and hydrous sulfates crystallize resulting in expansion of the shale matrix (H2 > H1) ...... 2

Figure 2.1 Schematic diagram illustrating the range of conditions that favor formation of various forms of early diagenetic pyrite (Brett and Baird, 1986) ...... 10

Figure 2.2 Devonian Shales throughout North American (Ettensohn, 1998) ...... 11

Figure 2.3 Geologic Units Containing Significant Acid-Producing Sulfide Minerals (Bureau of Topographic and Geologic Survey – DCNR, 2005) ...... 12

Figure 2.4 Geographic extent and hazards rating for sulfide-bearing materials in Virginia (Orndorff, 2001)...... 14

-37.1 Figure 2.5 Solubility of amorphous Fe(OH)3, Ksp=10 , as a function of pH at 25ºC. Lines indicate the solubility contributions of Fe3+ and individual Fe-hydroxy complexes (Langmuir, 1997) ...... 16

Figure 2.6 Eh-pH diagram for the system Fe-O2-S-H2O at 25ºC showing stability fields of goethite (α-FeOOH), pyrite (FeS2), and -2 monoclinic pyrrohotite (Fe7S8 = Fe0..877S) for ∑S(aq) = 10 mol/kg, and total carbonate 10-4 mol/kg (Barnes and Langmuir, 1979) ...... 17

Figure 2.7 Edale shale from Derbenshire, England showing pyrite (P) and gypsum (G) as a result of pyrite oxidation and reaction with calcium carbonate. Scale bar = 10 microns (Krinsley, et al., 1998) ...... 21

Figure 2.8 Experimental setup for expansion measurement with heating gradient and bacteria (Ballivy and Bellaloui, 1999) ...... 27

Figure 2.9 A Brown-Hoek stress profile to a depth of 3 km in sedimentary basins (Plumb, 1994) ...... 29

x Figure 3.1 Saw cut shale rock cores samples from Lewisburg, Pennsylvania ...... 35

Figure 3.2 Crushed shale rock core sample from Lewisburg, Pennsylvania ...... 35

Figure 3.3 pH versus time during oxidation of initial pyritic shale samples using 30% H2O2 ...... 36

Figure 3.4 Sulfate concentration versus time during oxidation of pyritic shale samples using 30% H2O2 ...... 37

Figure 3.5 Visual MINTEQ Evaluation: Saturation Index for sample T2B reacted with 30% H2O2. SI=log(Q/Keq) where Q is the product of the activities of the dissolved species ...... 39

Figure 3.6 pH values versus time of reaction in experiments using 10% H2O2 (1A through 6B) ...... 40

Figure 3.7 Optic image of sulfate crystal in sample T2B (30% H2O2 experiment) ...... 41

Figure 3.8 SEM image of sulphate crystal in sample 2B (10% H2O2 experiment...... 42

Figure 3.9 BSE image of the edge of a nodule in the pyritic shale from the 30% H2O2 experiment ...... 43

Figure 4.1 Cell model for irreversible reaction. a) Shale block measuring 0.34 m x 0.34 m x 0.34 m with 20 fractures or zones of availability at 1.619 cm spacing. b) Shale block after expansion and infilling of zones of availability with gypsum ...... 47

Figure 4.2 Total volume change associated with the dissolution of CaCO3, oxidation of FeS2, precipitation of gypsum and change in pH for a) 0.1% S2, b) 0.5% S2 and c) concentrated FeS2. See Tables H.1, H.2 and H.3 in Appendix H ...... 50

Figure 4.3 Swell model for 0.1% S2, 0.5% S2 and concentrated FeS2 cells with 20 fracture zones of availability. See Table H.4 in Appendix H ...... 52

Figure 5.1 Replacement pyrite fossil (fungia) encountered in the Marcellus Formation in Washingtonville, Pennsylvania. (a) Digital photograph of large spheroidal replacement fossil. (b) SEM Secondary Electron (SE) image of same pyrite fossil ...... 61

xi Figure 5.2 Digital photograph of fully oxidized zones of pyrite replacement along bedding plane of shale beneath floor slab at Evangelical Hospital in Lewisburg, Pennsylvania (Hoover and Pease, 2007) ...... 62

Figure 5.3 Digital photograph of cross section of joint in swelled pyritic shales beneath heaved section of floor slab at Evangelical Hospital in Lewisburg, Pennsylvania (Hoover and Pease, 2007) ...... 62

Figure 5.4 Concentrated Pyritic Shale Oxidation-Expansion Model. (a) Presence of concentrated pyrite within calcareous shale under overburden pressure. (b) Overburden materials are removed, which induces stress relief fracturing and initiation of pyrite oxidation. (c) Pyrite oxidation produces sulfuric acid migration through fractures resulting in the formation of gypsum. Swelling of shale matrix begins to heave overlying structure. (d) Pyrite is fully oxidized resulting in complete infilling of gypsum within the fractures and swelling is terminated ...... 64

Figure 5.5 Image S1L-7 thin section of saw cut perpendicular to joint and bedding plane (PSSC) ...... 66

Figure 5.6 Image S1M-6 thin section of saw cut perpendicular to joint and bedding plane (PSSC) ...... 66

Figure 5.7 Image S2M-5 thin section of saw cut parallel to joint and perpendicular to bedding plane (PSJ) ...... 67

Figure 5.8 Image SL3-20 thin section of saw cut parallel to bedding plane (PSP)...... 67

Figure 5.9 Simplified Model of Image J analysis of sample S2 (PSJ) with dimensions in millimeters (mm) ...... 69

Figure 6.1 Pyrite dissolution of 0.863 grams of FeS2 passing No. 60 sieve by inoculated (acidithiobacillus ferrooxidans) solution and uninoculated solution ...... 77

Figure 6.2 Swell testing apparatus for controlled experiments with bacterial oxidation ...... 79

Figure 6.3 Pyrite replacement fossils (fungia) from shale of the Marcellus Formation at a site near Washingtonville, Pennsylvania ...... 82

xii Figure 6.4 Possible hydrothermal pyrite deposit from shale of the Marcellus Formation at a site near Washingtonville, Pennsylvania ...... 82

Figure 6.5 a) Schematic diagram and b) photograph of actual setup for swell experiments using kinetic oxidation techniques ...... 86

Figure 6.6 Swell test results on controlled experiments with bacterial oxidation. Note that each of the samples actually experienced a measured decrease in total height over the course of the experiment ...... 88

Figure 6.7 Top view of concentrated pyritic source in sample B. Note the absence oxidation and sulfate crystallization features ...... 89

Figure 6.8 Sulfate concentration of effluent in kinetic oxidation column experiment...... 91

Figure 6.9 pH of effluent in kinetic oxidation column experiment ...... 91

Figure 6.10 Backscatter image of shale fragment from “A” column in the kinetic oxidation experiment...... 93

Figure 6.11 Backscatter images of framboidal clusters of pyrite microcrystals from column “B” in the kinetic oxidation experiment...... 93

Figure 6.12 Backscatter image of oxidation front into a shale fragment from column “A” in the kinetic oxidation experiment ...... 94

Figure 6.13 Backscatter image of oxidation front into a shale fragment from column “B” in the kinetic oxidation experiment...... 95

Figure 6.14 Backscatter image of macroscopic pyrite intentionally added to column “B” for the kinetic oxidation experiments ...... 96

Figure 7.1 USGS Topographic Map - Lewisburg Quadrangle (Berg and Dodge, 1981)...... 100

Figure 7.2 Pyrite Location Plan (Burt Hill, 2006) ...... 103

Figure 7.3 First floor transition between 1982 and 1996 additions ...... 104

Figure 7.4 Core Location and Floor Flatness Plan in the Information Sciences sector of the Evangelical Hospital ...... 107

xiii Figure 7.5 Percent by weight results for pyritic sulfur (forms of sulfur and chromium reducible) and sulfate sulfur (forms of sulfur) at core locations a) C-1, b) C-2 and c) C-3 in the IS area of the Evangelical Hospital ...... 110-111

Figure 7.6 Upward release of floor slab under pressure at the southwestern column in the IS area of the Evangelical Hospital ...... 114

Figure 7.7 Swelled shale sections beneath core location C-1 with in-situ photograph ...... 115

Figure 7.8 Underpinning plan for the Evangelical Hospital ...... 116

Figure 7.9 Design Method 1: Telephone room underpinning detail for main bearing wall at Evangelical Hospital ...... 118

Figure 7.10 Underpinning excavation around cantilever C-channel beams adjacent to the telephone room at Evangelical Hospital ...... 119

Figure 7.11 Underpinning excavation around micropiles and under adjacent footings along telephone room at Evangelical Hospital ...... 120

Figure 7.12 Telephone room underpinning close up of cantilever at Evangelical Hospital ...... 120

Figure 7.13 Broad view of telephone room underpinning at Evangelical Hospital ...... 121

Figure 7.14 Design Method 2: Wall and column/spread footing underpinning detail along 1982 addition at the Evangelical Hospital ...... 122

Figure 7.15 Saddle underpinning view from top along basement corridor adjacent to IS area of the Evangelical Hospital ...... 123

Figure 7.16 Saddle underpinning view from bottom along corridor adjacent to IS area of the Evangelical Hospital ...... 124

Figure 7.17 Design Method 3: Interior spread footing underpinning detail within IS area at Evangelical Hospital ...... 125

Figure 7.18 Design Method 3 shear ring a) detail and b) field casing for the IS area of the Evangelical Hospital ...... 127

xiv Figure 7.19 Design Method 3 footing to micropile bond detail for the IS area at the Evangelical Hospital ...... 128

Figure 7.20 Load test at the 489 kN footing in the IS area at Evangelical Hospital ...... 129

Figure 7.21 Load test at the 1,285 kN footing in the Computer Room a Evangelical Hospital ...... 129

Figure 7.22 489 kN column footing bond stress load test results in the IS area at the Evangelical Hospital ...... 131

Figure 7.23 1,285 kN column footing bond stress load test results in the Computer Room at the Evangelical Hospital ...... 132

Figure 7.24 Klemm KR 702 electric drill rig utilized to installed micropiles at the Evangelical Hospital ...... 134

Figure A.1 Reference Chart created by Bryant (2003) using Macromedia Freehand 10® Munsell® Book of Color Library. Color Guidelines from Sobek et al. (1978), USDA-NRCS (1998), and Hosterman and Whitlow (1980, 1983) ...... 142

Figure E.1 X-Ray Diffraction results from nodule in shale in oxidation experiments ...... 159

Figure G.1 PHREEQC graphical output of Assemblage (moles) versus O2 (moles) for 0.1% S2 and 5% CaCO3 ...... 165

Figure G.2 PHREEQC graphical output of Assemblage (moles) versus O2 (moles) for 0.5% S2 and 5% CaCO3 ...... 166

Figure G.3 PHREEQC graphical output of Assemblage (moles) versus O2 (moles) for Concentrated FeS2 and 5% CaCO3 ...... 167

xv LIST OF MATHGRAMS

Mathgram 2.1 Remnant Stress Example to Explain the Potential for Microcrack Formation in Pyritic Shales ...... 30

Mathgram 4.1 0.1% S2 and 5% CaCO3 Calculations ...... 55

Mathgram 4.2 0.5% S2 and 5% CaCO3 Calculations ...... 56

Mathgram 4.3 Concentrated FeS2 and 5% CaCO3 Calculations ...... 57

Mathgram 4.4 Change in Cell Height Calculations ...... 58

Mathgram 5.1 Regression Analysis Calculations ...... 72

xvi ACKNOWLEDGEMENTS

I owe my deepest debt of gratitude to my wife Jennifer. She has been extremely patient and supportive throughout my education and I can’t thank her enough. I also thank my daughters, Anne, Tessa Jane and Maizy, for putting up with their Daddy sequestering himself to the attic for hours on end. I want to thank my parents, Sheldon and Betty Ann Hoover, for always believing in me and giving me the confidence to tackle anything in life. None of this research or my education would be possible without Frank Welsh at CMT. Thanks to Frank for being such a progressive mentor and boss. I also want to thank all of the employees at CMT for supporting me and for always looking for that special pyritic rock. Thanks to my advisor Mian Wang for encouraging me to continue with my education. Thanks also to committee members Brian Dempsey, David Lehmann, Susan Trolier-McKinstry and Angelica Palomino for guiding me in my research. Thanks to Larry Mutti at Juniata College for helping me with the scanning electron microscope and X-ray diffraction equipment. Thanks also to Norm Siems for inspiring me early in my academic career and for giving me the opportunity to teach at my alma-mater. I would also like to recognize David Barrett, Fenglong Sun, John Senko and Jennifer Sloppy for their efforts in this research. Finally, thanks to Jesus Christ my savior for always watching over and guiding me.

xvii CHAPTER 1

INTRODUCTION

1.1 General

The study of expansive pyritic shales is becoming a rapidly relevant geotechnical

topic with the decrease in favorable building sites in highly populated areas. Our

understanding of shale expansion has been limited to historical studies of building sites

where there is documentation of structural damage. Other than case study reviews there has been relatively little study of expansive pyritic shales with the notable exception of

Ballivy and Bellaloui (1999) and Cormier (2000). There are many gaps in the science of

expansive pyritic shales where there needs to be better agreement between characterizing the potential for swelling and developing responsible geotechnical engineering techniques.

A description of the expansion process in a typical construction project is modeled in Figure 1.1. During initial excavation of a construction site, pyritic shale that typically had been in an intact and chemically reduced condition fractures due to unloading expansion and oxidizes as a result of lowering of the water table. As a result of these initial microenvironmental changes, there is an increase in fracturing, bedding- plane discontinuities, and permeability which allows for further availability of pyrite to oxidize. Capillary movement of groundwater within the vadose zone provides a transport mechanism for the migration of an acidic front. The pyrite oxidation leads to the formation of hydrous sulfates, and the entire shale cell heaves within the vadose zone.

1 The model shown in Figure 1.1 presents a visual description of the typical

conditions that lead to heave in pyritic shale environments (Hawkins and Pinches, 1987

and Hoover and Pease, 2007).

SURCHARGE LE UNLOADING AB (OVERBURDEN) T ER AT W H0 H0 OXIDIZED VADOSE ZONE H0 PH = 3 FOR BACTERIAL H0 WATER TABLE SURVIVAL HORIZONTAL PE POSITIVE DISCONTINUITY/ LE AB FRACTURE T ER T ED A ER W W VERTICAL LO JOINT OR FLOW H1>H0 H0 PATHWAY CAPILLARY MIGRATION (TRANSPORT) MIGRATION CAPILLARY INCREASE IN FRACTURING WITH WATER TABLE UNLOADING EVENT

LOWERED (TRANSPORT) MIGRATION CAPILLARY CAPILLARY MIGRATION (TRANSPORT) MIGRATION CAPILLARY CAPILLARY MIGRATION (TRANSPORT) MIGRATION CAPILLARY REDUCED ZONE ANAEROBIC BELOW WT REDUCED ZONE PH = 7 ANAEROBIC BELOW WT PE NEGATIVE PH = 7 PE NEGATIVE Z1 Z

a) b)

SWELL

H0 EVAPORATION OF PORE LEGEND: WATER AND H0 = UNCHANGED PLAN DIMENSIONS AND ORIGINAL HEIGHT OF H0 OXIDIZED VADOSE ZONE. PRECIPITATION OF H1 = NEW HEIGHT OF OXIDIZED VADOSE ZONE DUE TO STRESS HYDROUS SULFATES RELIEF FRACTURING. (GYPSUM) H2 = NEW HEIGHT OF OXIDIZED VADOSE ZONE DUE TO GYPSUM INFILLING OF MICRO-FRACTURES AND DISCONTINUITIES. LE Z = ORIGINAL HEIGHT OF REDUCED ZONE. AB T Z1 = NEW HEIGHT OF REDUCED ZONE DUE TO STRESS RELIEF ER T ED FRACTURING. A ER W W H2>H1>H0 LO

SWELLING AT DISCONTINUITIES WATER TABLE LOWERED REDUCED ZONE ANAEROBIC BELOW WT PH = 7 PE NEGATIVE

Z1

c)

Figure 1.1 Expansive pyritic shale model as a step-wise progression during a typical construction project. (a) Elevated water table with overburden materials prior to construction. (b) Water table is lowered and overburden removed as part of the excavation process resulting in increased fracturing and oxidation of available pyrite (H1>H0). (c) Capillary water evaporates and hydrous sulfates crystallize resulting in expansion of the shale matrix (H2>H1>H0).

2 1.2 Microenvironmental Considerations

The cause of most swelling in pyritic shales is from the oxidation of pyrite and subsequent formation of hydrous sulfates, with accompanying increase in molar volume

of the solid phases. The hydrous sulfate most commonly referenced in cases of heave

and subsequent structural distress is gypsum or calcium sulfate dihydrate (CaSO4•2H2O).

Although other metal hydrous sulfates result from oxidation of pyrite, these minerals are

rarely responsible for documented cases of swelling and damage to civil infrastructure

and are more likely to be responsible for the production of acid upon dissolution. The metal hydrous sulfates are often transitory, whereas the lower solubility of gypsum and

stability of gypsum relative to competing phases over a broad pH range results in a more stable condition.

Because gypsum is the mineral of primary importance, we must also be interested in the presence of calcite in pyritic shales. Calcite or some source of Ca(II) must be available in order for gypsum to form, which is typically through the interaction with sulfuric acid from the oxidation of pyrite (see Equation 2.5 and Chapter 4 on the

PHREEQC Geochemical Model). Pyrite may be present in a microscopic form within the shale matrix (framboidal or microcrystalline) or in a macroscopic or concentrated form (nodules, burrows and/or replacement fossils). Microscopic pyrite is generally defined as having a particle size less than 10 microns and macroscopic is anything greater than 10 microns but typically recognizable with a conventional hand lens (>10x).

This research for my dissertation provides empirical and theoretical evidence that macroscopic pyrite is a significant contributor to expansive pyritic shales that contain abundant calcite. The limited availability of microscopic forms of pyrite to oxidation is

3 the leading cause for skepticism about the primary role of framboidal or microcrystalline

pyrite to be the driving mechanism for the production of gypsum in pyritic shales with

significant amounts of calcite. Although microscopic pyrite will more readily oxidize

than larger forms due to increased surface area (Brady et al. 1998), the microscopic forms

of pyrite are only physically available at the surface of fractures and other discontinuities.

The availability to oxidize the microscopic pyrite within the shale matrix comes with

weathering processes and breakdown of the shale structure. The crystallization of gypsum typically occurs in original discontinuities or small stress relief fractures that result from unloading of the shale during construction (see Section 2.6 on remnant stresses).

The rate of pyrite oxidation increases significantly at low pH environments with abundant ferric iron (Rimstidt and Newcomb, 1993). The metabolism of Thiobacillus ferrooxidans is also high at low pH (Jaynes, 1984a and 1984b). Both the abiotic and biotic pathways are expected to be slow if the presence of calcite keeps the system at a near neutral pH. Groundwater monitoring data from the Hamilton Group of

Pennsylvania’s Ambient and Fixed Station Network (FSN) Monitoring program (Reese and Lee, 1998) reports that the pH of the groundwater to be at an average of 7.26.

This reasoning leads to the importance of macroscopic forms of pyrite within the

Devonian shales as a key indicator for swell potential. The release of overburden pressures associated with construction projects will likely produce an abundance of stress relief fractures at the interface of the concentrated macroscopic pyrite sources due to the discrepancy in hardness between pyrite and the surrounding shale mineralogy. These fractures allow for the oxidation process to begin around the concentrated pyrite sources.

4 The drop in pH from the production of acid around the microenvironments of the

concentrated pyrite sources lead to favorable conditions that allow for ferric iron and

bacteria to rapidly increase oxidation rate. These microenvironmental “hot spots” can

influence the surrounding materials through the migration of capillary pore water that

spreads from the concentrated source through fractures and discontinuities and into the

surrounding calcareous shale. This acidic migratory front then reacts with the calcite or

mobilizes dissolved calcium carbonate to form gypsum. The crystallization of gypsum

leads to heave of the shale matrix.

1.3 Thesis Organization

Chapter 2 is a literature review describing case studies of construction projects where expansive pyritic shales have caused structural distress to civil infrastructure. The geology of pyrite formation in carbonaceous shales is briefly covered with a particular emphasis on Devonian Shales and current mapping techniques that describe specific formations having sulfides that could potentially cause acid rock or mine drainage problems. Acid production and sulfate precipitation chemistry is presented to give an overview of the important processes that cause expansion. Laboratory testing methods for determining the percent by weight of finely disseminated or microscopic pyrite in rock and soil samples is presented. Finally, the remnant stress theory highlights the potential for stress relief fracturing.

Chapter 3 gives chemical and physical explanations of the expansive shale process through the oxidation of pyritic shales. Experimental oxidation of pyritic shales with varying amounts of pyrite nodules is accomplished through the introduction of hydrogen peroxide of varying concentration. The resulting temperature and pH and

5 sulfate ion, calcium, aluminum, potassium and iron concentrations are tracked and

compared with theoretical predictions utilizing the Visual MINTEQ (Gustafson, 2007)

hydrogeochemical program.

The PHREEQC (Parkhurst and Appelo, 1999) hydrogeochemical model is

presented in Chapter 4 to highlight the processes involved in the oxidation of pyrite and

subsequent production of gypsum. A generalized expansion model is offered in an

attempt to quantify the expected differences in heave that could result as calcite dissolves

in the reaction with sulfuric acid resulting in gypsum precipitation.

An empirical regression model technique is shown in Chapter 5 as a means of

uncovering clues to the origin of sulfate formation. Samples of shale that have

undergone expansion beneath the Evangelical Hospital in Lewisburg, Pennsylvania

(Hoover, 2004 and Hoover and Pease, 2007) are microscopically analyzed. Each thin

section of expanded rock is analyzed to estimate the total amount of gypsum produced

and then back calculated to estimate the amount of pyrite and calcite required in the

reaction.

Two separate geotechnical laboratory studies are presented in Chapter 6 in an

effort to produce an expansion test capable of determining the conditions responsible for

heave related failures. The first test consists of two molds with known quantities of illite,

calcite and pyrite. A bacteria rich solution is used to inundate the samples in order to establish oxidation of the pyrite that is present in various concentrations within the

samples. The second test combines a revised version of the Acid Drainage Technology

Initiative (ADTI-WP2) Leaching Column Method with the capability of measuring

expansion of the pyritic shale samples.

6 Chapter 7 is a case history involving micropile underpinning over expansive pyritic shales. The Evangelical Hospital in Lewisburg, Pennsylvania has undergone significant structural distress due to expansive pyritic shales. Comprehensive field and laboratory testing was accomplished and is detailed in this chapter. Also, extensive micropile underpinning techniques were utilized to support the structure and protect against future heave related distress.

Chapter 8 summarizes the findings of the laboratory and theoretical exercises and conclusions are presented concerning the characterization of the physical and geochemical processes of pyritic shales of the Marcellus Formation. Finally, recommendations are offered for changes to the current state-of-practice and future research challenges are highlighted.

The appendices provide laboratory testing data, established testing procedures and other miscellaneous information that support this research into expansive pyritic shales of the Marcellus Formation.

7 CHAPTER 2

LITERATURE REVIEW

2.1 Introduction

Of particular importance to the understanding of the expansion phenomenon is the geology of the shale bedrock, chemistry of pyrite oxidation and sulfate precipitation, empirical evidence of heave, laboratory techniques of identifying concentrations of pyrite and potential for heave and the physical explanations behind the formation of stress relief fracturing.

Pyrite is found in black, carbonaceous shales throughout the world. These potentially expansive pyritic shales have been documented in areas of the United States such as Tennessee (Long and Williams, 1990), Virginia, Ohio (ENR, 1960), West

Virginia (Dubbe et al., 1984), Pennsylvania (Hoover, 2004), Missouri and Kansas

(Coveney and Parizek, 1977). The expansive pyritic shale problem is also found in

England (Hawkins and Pinches, 1987), Norway (Moum and Rosenquist, 1959), Canada

(Tanguay et. al., 1999 and Penner and Eden, 1972) and Sweden (Jangdal, 1971).

Undocumented cases of structural damage due to expansive pyritic shales is certainly possible given that the problem is generally not recognized to those in the construction industry and to a significant proportion of geotechnical engineers and geologists.

Recently, the expansive shale problem has produced millions of dollars worth of damage to the Trinity South elementary school in Western Pennsylvania. The coal bearing shales beneath the school building, which was constructed in 1995, heaved shortly after construction. The school district has received over $1.1 million in damages

8 from contractors for negligence and breach of contract (Observer-Reporter, 2006 and

2007).

Research indicates that there is a sulfide to sulfate conversion process that results

from the oxidation of iron sulfide (FeS2) or pyrite (Dougherty and Barsotti, 1972, Belgeri and Siegel, 1998, Freeman, 2003, Hoover, 2004). The various sulfate minerals have molar volumes in large excess of the parent minerals resulting in expansion of the bedrock matrix and subsequent heave of overlying structures.

A study into the catastrophic failure of a shale embankment dam at Carsington,

Derbyshire, U.K., in 1984 revealed that displacive growth of secondary gypsum crystals played an important role in disrupting the shale fabric and provided strong observational evidence for the importance of pyrite oxidation and dissolution of primary carbonate minerals in the embankment fill (Pye and Miller, 1988). Oxidation of pyrite in the shale produced sulfuric acid, which leached much of the carbonate from the shale. The overall effect of these reactions was estimated to have increased the porosity of some parts of the shale fill by up to 10%, with possibly significant implications for its geotechnical behavior.

Crystallization of gypsum in small pores is known to produce pressures of up to

100 MPa (2,089 kips/ft2). Hoover (2002) has measured pressures of gypsum

crystallization in poorly-graded shale fragments, well-graded shale fragments and intact

rock cores of up to 17 kPa (360 lbs/ft2), 23 kPa (490 lbs/ft2) and 627 kPa (13,104 lbs/ft2), respectively.

9 2.2 Pyritic Bedrock Geology

The shale samples utilized in the laboratory studies outlined in subsequent

chapters are from the Marcellus Formation in Central Pennsylvania; however, other formations are discussed to provide a broad understanding of the abundance of pyrite in various rocks.

The reasons for the presence of pyrite, whether in macroscopic form (nodules, burrows, fossil replacement) or microscopic form (framboids or microcrystals), within shale bedrock is a relatively well understood phenomenon. As indicated in Figure 2.1, the presence of the various forms of pyrite is generally a function of the amount of oxygen in the overlying water and sedimentation rate (Brett and Baird, 1986).

Figure 2.1 Schematic diagram illustrating the range of conditions that favor formation of various forms of early diagenetic pyrite (Brett and Baird, 1986).

10 Syngenetic sulfides, such as pyrite, formed in bedded sediments during deposition

and diagenesis (changes to a sediment after deposition). The metallic components were

deposited in an anoxic sedimentary environment where they interacted with microbial

bacilli during diagenesis to extract sulfur from sea water (euxinic conditions) and grow

sulfide minerals such as pyrite (Gold and Doden, 2007).

The middle Devonian Marcellus Formation is commonly considered to have been

deposited in deep anoxic waters (Schultz, 1999). The substrate may have been nearly

anoxic and the bottom waters moderately well oxygenated most of the time. Typical

locations of Devonian Shales throughout North America are shown in Figure 2.2.

Figure 2.2 Devonian Shales throughout North American (Ettensohn, 1998).

The presence of epigenetic pyrite, or pyrite encountered in veins deposited from hydrothermal fluid near the surface of the earth after consolidation and lithification of the

11 host rock, in the Ordovician age sandstones (Juniata and Bald Eagle Formations) along the proposed I-99 corridor in Centre County, Pennsylvania has led to an increased interest in the presence of pyrite and its potential for acid rock drainage (ARD) and expansion

(Gold and Doden, 2007). Figure 2.3 shows the presence of potentially acid producing sulfide minerals in Pennsylvania. Note that the light green color designates the presence of Marcellus Formation within the Ridge and Valley Province in the central part of the state. The colors shown on the map represent geologic formations where potentially acid-producing sulfide minerals are present and do not designate a particular hazard rating.

Washingtonville

Lewisburg

Huntingdon

Altoona

Figure 2.3 Geologic Units Containing Significant Acid-Producing Sulfide Minerals (Bureau of Topographic and Geologic Survey – DCNR, 2005).

Note that Figure 2.3 does not currently rate the potential hazards associated with these formations, such as the potential to produce ARD, acid mine drainage (AMD) or expansion. The different colors only represent the different geologic formations that

12 have potentially significant acid-producing sulfide minerals. A sulfide hazard map is in

the progress of being created with the aid of researchers and students from the

Pennsylvania State University. Figure 2.4 shows a sulfide hazards rating map for

Virginia (Orndorff, 2001).

Note that Potential Peroxide Acidity (PPA) is a test utilized to determine the

amount of acidity produced by rapidly oxidizing the material with hydrogen peroxide

(H2O2). The results are presented in terms of the amount of calcium carbonate (CaCO3) required for neutralization (Barnhisel and Harrison, 1976).

13

Figure 2.4 Geographic extent and hazards rating for sulfide-bearing materials in Virginia (Orndorff, 2001).

14 2.3 Pyrite Oxidation and Sulfate Precipitation Chemistry

The most widely studied hydrous sulfate in the expansion process is gypsum or calcium sulfate dihydrate (CaSO4•2H2O). Pyrite, or iron sulfide (FeS2), oxidizes to form

sulfuric acid (H2SO4). The sulfuric acid then reacts with any calcium carbonate (CaCO3)

and precipitates to form gypsum in saturated conditions (Saturation Index >1). The

formation of acidity is described in the following reactions (Brady et al., 1998):

2+ 2- + FeS2(s) + 3.5 O2 + H2O = Fe + 2 SO4 + 2H 2.1

2+ + 3+ Fe + 0.25 O2 + H = Fe + 0.5 H2O 2.2

3+ 2+ 2- + FeS2(s) + 14 Fe + 8 H2O = 15 Fe + 2 SO4 + 16 H 2.3

3+ + Fe + 3 H2O = Fe(OH)3(s) + 3 H 2.4

Pyrite oxidation can occur due to advection of oxic groundwater into a pyritic

area, but solubility of O2 in water is relatively low and therefore the extent of reaction

may be limited. Reaction is usually more extensive when gaseous O2 diffuses into an

unsaturated zone. The low pH values cause a variety of minerals to precipitate including

gypsum and numerous ferric hydrosulfates (Appelo and Postma, 1993). Flux of gaseous

O2 can be quite large, resulting in rapid reaction and producing high sulfate and iron concentrations and pH of 2 or less where there is limited calcite availability. As pH is decreased the solubility of Fe(III) increases resulting a high flux of oxidizing Fe3+ into

adjacent saturate zones (Rimstidt and Newcomb, 1993). The solubility diagram in Figure

2.5 illustrates how the solubility of Fe3+ increases as the pH decreases.

15

-37.1 Figure 2.5 Solubility of amorphous Fe(OH)3, Ksp=10 , as a function of pH at 25ºC. Lines indicate the solubility contributions of Fe3+ and individual Fe-hydroxy complexes (Langmuir, 1997).

Microbial oxidation by Acidithiobacillus ferrooxidans can also increase the

reaction rate. Jaynes (1984a and 1984b) reported that prime bacterial activity is at a

maximum near pH ≈ 3.25. Relative humidity, temperature, surface area and morphology

also affect the rate of reaction. The stability of pyrite is also well illustrated in the Eh-pH diagram in Figure 2.6. This diagram shows that pyrite is typically stable in a reducing

16 environment (Eh(V) < 0), but can also be stable in a oxidizing environment (0 < Eh(V) <

0.24) with a pH less than 5.

Figure 2.6 Eh-pH diagram for the system Fe-O2-S-H2O at 25ºC showing stability fields of goethite (α-FeOOH), pyrite (FeS2), and monoclinic pyrrohotite (Fe7S8 = -2 -4 Fe0..877S) for ∑S(aq) = 10 mol/kg, and total carbonate 10 mol/kg (Barnes and Langmuir, 1979).

Although generally utilized to mitigate Acid Mine Drainage (AMD) and Acid

Rock Drainage (ARD), the use of calcium carbonate can lead to geotechnical problems associated with expansion and heave of the overlying structures. The sulfuric acid

produced during the oxidation of pyrite reacts with calcium carbonate to produce

gypsum, which involves an expansion in volume (Bell, 2000).

CaCO3 + H2SO4 + 2H2O → CaSO4•2H2O + CO2 + H2O 2.5

17 Note that as the carbon dioxide (CO2) is released during the reaction, pH will

increase in an open system and require more acid, not allowing for the complete

dissolution of the calcium carbonate. By contrast, a closed system, which prevents the

release of CO2 and keeps the pH acidic, maintains a lower pH and is more conducive to

the dissolution of calcium carbonate and conversion to gypsum. This concept is

illustrated in the AMD industry as a mean of determining the amount of calcium

carbonate or limestone rock required to react with and neutralize a material containing a

certain percentage of pyritic sulfur (S2). Equation 2.6 describes an open system in which

31.25 tons of CaCO3 are required to react with 1000 tons of material containing 1% S2 by weight and Equation 2.7 illustrates a closed system in which 62.50 tons of CaCO3 are required to react with 1000 tons of material containing 1% S2 by weight (Brady et al.,

1998).

2- 2+ FeS2 + 2 CaCO3 + 3.75 O2 + 1.5 H2O = Fe(OH)3 + 2 SO4 + 2 Ca + 2 CO2(g) 2.6

2- 2+ - FeS2 + 4 CaCO3 + 3.75 O2 + 3.5 H2O = Fe(OH)3 + 2 SO4 + 4 Ca + 4 HCO3 2.7

Equation 2.6 shows that 2 moles of CaCO3 are required to react with 1 mole of

FeS2 in an open system; whereas, Equation 2.7 shows that 4 moles of CaCO3 are required

to react with 1 mole of FeS2 in a closed system. The open system is generally the

accepted method of AMD or ARD remediation (Brady et al., 1998); however, the closed

system may be a more accurate representation of the conditions that are typically

encountered in expansion studies. The CaCO3 to FeS2 ratio will become very important

as the conditions required for precipitation of gypsum are clarified. As the CaCO3 to

18 FeS2 ratio drops, the conditions necessary for a drop in pH and crystallization of hydrous

sulfate become much more favorable.

Other hydrous sulfates that are produced from the oxidation of pyrite include

melanterite, halotrichite and jarosite. These transitional metal sulfates are formed during

a precipitation process described as follows (Hammarstrom et al., 2004):

2+ 2- Melanterite: Fe + SO4 + 7H2O → FeSO4·7H2O 2.8

2+ 2- 3+ Halotrichite: Fe + 4SO4 + 2Al + 22H2O → FeAl2(SO4)4·22H2O 2.9

3+ 2- + + K-Jarosite: 3Fe + 4SO4 + K + 6H2O → KFe3(SO4)2(OH)6 + 2H 2.10

These metal sulfates are capable of producing additional acid upon dissolution,

unlike gypsum (Brady et al., 1998). A summary of the density, molar volume and

solubility products for various parent minerals and hydrous sulfates are presented in

Table 2.1.

Table 2.1 Molar volumes, densities and solubilities of key minerals.

Molar Density Volume Solubility 3 †† 3 † Mineral (g/cm ) (cm /mol) Product (pKsp) Crystalline Pyrite 5.01 23.9 18.51** Framboidal Pyrite 4.12 29.1 18.51** Calcite 2.71 36.9 8.48* Gypsum 2.30 74.8 4.58* Melanterite 1.89 147.1 2.21* Halotrichite 1.84 483.9 -- K-Jarosite 3.09 162.1 11** *From Lanmuir, 1997 **K-Jarosite and Pyrite Ksp from V-Minteq database (Gustafson, 2007) †Estimated from density and molecular weight data ††From Roberts, 1990

19

The expansion that results from the conversion to the various hydrous sulfates cannot simply be determined by comparing the molar volume change of the principal constituents. The dissolution and transfer of minerals and elements within each reaction must be considered in any effective expansion model.

Research conducted by Hammarstrom, et al. (2004) into secondary sulfate minerals associated with acid drainage suggests that gypsum is the most stable of the sulfates produced as a result of the pyrite oxidation process. Essentially, modeling reveals that gypsum, which is less soluble than other hydrous sulfates, is most likely to precipitate in pyritic shale that contains abundant calcium carbonate. The focus of the research in the subsequent chapters will be on the formation of gypsum.

2.4 Weathering of Pyritic Shales

Pyrite is typically found in organic sedimentary rock because it typically forms as a result of the bacterial reduction of seawater sulfate using organic matter as an electron accepter (Berner, 1984). SEM, porosity, permeability and specific surface area studies of the pyritic New Albany Shales (Devonian) in Clay City, Kentucky show that inward migrating gaseous or dissolved O2 reacts first with modern soil organic matter and then with any remaining shallow pyrite. Once the pyrite is gone the O2 attacks the organic matter. A small amount of residual O2 migrates beyond the front to react with pyrite resulting in loss of some pyrite but little loss or gain of organic matter (Wildman, et al.,

2004). This oxidation front and subsequent dissolution of pyrite result in a porosity that decreases with depth. Thus, unless new fractures open, the weathering process is surficial and self-limiting.

20 Krinsley et al, (1998) utilized backscatter scanning electron microscopy to study weathered pyritic shale and noted that growth of the gypsum causes a volume expansion that ruptures the lamination, thereby accelerating the pyrite oxidation process.. Figure

2.7 shows weathered pyritic shale with growth of gypsum crystals (G) between partially opened shale laminae.

Figure 2.7 Edale shale from Derbenshire, England showing pyrite (P) and gypsum (G) as a result of pyrite oxidation and reaction with calcium carbonate. Scale bar = 10 microns (Krinsley, et al., 1998).

2.5 Identification of Potentially Expansive Pyritic Shales

The current state-of-practice for identifying expansive pyritic shales has been to compare expansive shale case studies to current pyritic conditions. Specifically, the amount of microscopic pyritic sulfur in the bedrock is compared to values presented where expansive shale has been documented to cause structural distress.

Empirical evidence has thus far been the greatest indicator in identifying potentially expansive pyritic shales. The Johnson City Public Library has undergone heave of up to 11 cm (4.33 inches) with laboratory tests indicating that the sulfide sulfur

21 content in the unweathered shale typically ranged from 0.59 to 0.71 percent (Long and

Williams, 1990). Freeman (2003) states in a report that Appalachian shales with sulfide

sulfur contents in the range of 0.1 to 0.5 percent or greater have caused heave in

overlying structures. Hawkins and Pinches (1987) report that heave of up to 6 cm (2.36

inches) has occurred at the Llandough Hospital in England with underlying total sulfur

contents of the shale ranging between 0.26 and 1.99 percent. Dougherty and Barsotti

(1972) suggest that data indicates that “sufficient” material for expansion is in the range

of 0.1 percent sulfide sulfur.

These empirical studies do not provide sufficient detail to differentiate if the samples were taken from within the oxidized or swell zone or from a deeper unoxidized region of the shale formation.

2.6 Pyritic Sulfur Identification Techniques

Munsell Color Guidelines and established static and kinetic laboratory testing

techniques are utilized to estimate or measure the amount of pyrite or pyritic sulfur in soil

or rock.

2.6.1 Munsell Color Guidelines

The ability of a material, such as shale bedrock, to produce an acidic environment

can be attributed to the Munsell color. The Munsell Color Charts are convenient to use

when identifying finely disseminated minerals, such as pyrite, in geologic materials. The

charts use hue, value (lightness), and chroma (color separation) to establish a standard

color system. The color notation has the form HAB ν/c that can be defined as follows:

22 H – numeral corresponding to AB color hue (0 – 100)

ν – value (0 – pure black to 10 – pure white)

c – chroma (0 – neutral colors to 30 – fluorescent)

Pyritic materials typically have a neutral designation with value but no color or hue (Bryant, 2003).

Research conducted by Sobek et al. (1978) indicates that higher chromas values

(≥ 3) are typical of materials in weathered, oxidized zones whereas pyritic materials with a high sulfur content (>1.0% sulfur) may have a low chroma (≤ 2) and a black streak or powder with a value ≤ 3. More significantly, chromas ≤ 2 often indicate sufficient pyrite to cause a pH < 4.0 upon oxidation, while chromas ≥ 3 typically correspond with negligible percentages of pyritic sulfur. The color chart is provided in Figure A.1 in

Appendix A.

Of more importance to the theories presented in this paper, the amount of calcium in pyritic shale is a key factor in the formation of gypsum. Calcite impacts the color value of calcareous shales by affecting carbon absorption during formation (Hosterman and Whitlow, 1980). Shales deposited in moderately high calcium environments tend to be darker (lower color value). In a study on Munsell color values of Devonian shales by

Hosterman and Whitlow (1980), shales deposited in high calcium environments range from 4.0 – 5.5 on the Munsell color value system and shales formed in low- or no- calcium environments typically range from 4.8 - 6.1.

Examples of recorded Munsell color designations for pyritic, calcareous shales include: 10 YR 4.4:0.5 for dark gray samples of the Sonyea Formation (averaging 5% pyrite and 5% calcite); N 6.0 (neutral) for samples of the Java Formation (averaging 5%

23 pyrite); and N 3.4 for samples of Millboro Shale (averaging 5% pyrite and 25% calcite)

(Hosterman and Whitlow 1983).

What is interesting about this Munsell color research into the correlation between

the potential to produce acidic environments and the amount of calcite within pyritic

shale is that a calcareous pyritic shale will likely not produce an acidic environment.

With Munsell values above 3.4, the potential ability to produce AMD is restricted. Also,

the potential of calcareous shale to produce an environment conducive to bacterial

oxidation (pH≈3) is severely limited by the presence of calcite within pyritic shales.

2.6.2 Static Laboratory Testing Techniques

The most common method of characterizing the potential for swell is to run either total sulfur or forms of sulfur tests and compare to historical cases of heave. The research conducted by Dougherty and Barsotti (1972), that suggests a lower limit of 0.1 percent sulfide or pyritic sulfur, is often referenced in case studies (Bryant, 2003).

The total sulfur test is conducted with a LECO Total Sulfur Analyzer and is

typically utilized in Acid Base Accounting (ABA) for determining the acid producing and

neutralizing capabilities of AMD or ARD. The total sulfur test is widely utilized due to

acceptable levels of agreement between laboratories.

The forms of sulfur test is utilized to determine the actual amount of total sulfur,

sulfide or pyritic sulfur and sulfate sulfur. Sulfate sulfur is rarely found in reducing

environments typical of the samples collected for these tests. Hence, the total sulfur in

these samples consists almost entirely of organic and pyritic sulfur. The forms of sulfur

test is often utilized in geotechnical studies due to the interest in pyritic sulfur content of

the shale, but it is not utilized with appreciable frequency for AMD or ARD analyses due

24 to problems with consistency between laboratories (Brady et al. 1998). Three methods for forms of sulfur testing are identified and accepted in the Pennsylvania Department of

Environmental Regulation (DER) Overburden Sampling and Testing Manual (1988):

A. ASTM Standard Method D2492: Appropriate for determining forms of sulfur

in coal, but not intended for any other matrix. Overestimation of pyritic sulfur

content is possible. Acceptable precision for pyritic sulfur is +/-0.05% for

total sulfur less than 2.0% and +/-0.10% for total sulfur greater than 2.0%.

B. Modified EPA Method: Appropriate for all coals and overburden; however, if

any unreacted pyrite is left in the sample, then underestimation of pyritic

sulfur content would result. Acceptable precision for pyritic sulfur is +/-

0.15% for total sulfur less than 2.0% and +/-0.25% for total sulfur greater than

2.0%.

C. ASTM/EPA Combination Method: Considered applicable to all coals and

overburden samples common to Pennsylvania. Acceptable precision for

pyritic sulfur is +/-0.10% for total sulfur less than 2.0% and +/-0.15% for total

sulfur greater than 2.0%.

The ASTM/EPA method is typically utilized to determine pyritic sulfur content in shale; however, it should be recognized that the acceptable precision of 0.10% is equal to previously defined thresholds for characterizing the potentially expansive nature of a bedrock formation (Freeman, 2003; Dougherty and Barsotti, 1972). Note also that this test method utilizes only 1 to 2 grams of sample to determine pyritic sulfur content for an unspecified volume of rock, which is typically far from homogenous. Unless large quantities of a particular bedrock formation are tested, there is the potential for significant

25 error in characterizing whether a material is expansive. No research has been conducted to determine the statistical variance of pyrite within materials that have experienced heave; therefore, a more detailed study of these shales is warranted.

The Laboratory Methods Guideline for Acid Sulfate Soils (Sullivan et al., 2000) states that the chromium reducible sulfur (SCR) (Method 22B) is a viable alternative to the

peroxide oxidizable sulfur (Method 21D) for measuring pyritic sulfur (Sullivan et al.,

1999). A comparison of this method to the forms of sulfur and total sulfur tests is

presented in Chapter 7.

The laboratory testing procedures for the Total Sulfur, Forms of Sulfur

(ASTM/EPA Combination Method) and Chromium Reducible Sulfur tests are presented

in Appendix B.

2.6.3 Kinetic Laboratory Testing Techniques

The ADTI-WP2 Leaching Column Method for Overburden Analysis and

Prediction of Weathering Rates method is being developed by Hornberger and Brady

(1998), Geidel et al. (2000) and Hornberger et al. (2000) in an attempt to provide an industry wide standard for a kinetic test method for predicting the quality of drainage from coal and metal mines. A number of humidity cell and leaching column methods have been developed; however, there is no currently accepted method (Brady et al. 1998).

The ADTI-WP2 method has shown great promise based on recent testing and has lead to an increased interest in this method for incorporation into swell experiments.

This method is referenced in Chapter 6 and a complete description of the procedure is provided in Appendix B.

26

2.7 Laboratory Testing Methods – Swell Test

Ballivy and Bellaloui (1999) have attempted to establish a reproducible swell test

in an effort to quantify the expansion potential of pyritic materials. This test essentially

involves compacting a material with known sulfur content into a conventional California

Bearing Ratio (CBR) mold and subjecting the material to a heating gradient and water

with a bacteria culture (Thioparus thiobacillus). The setup for this method is shown in

Figure 2.7. The material is measured for expansion over a period of time. This method has produced some success in terms of swell versus pyritic sulfur content and research is continuing.

Figure 2.8 Experimental setup for expansion measurement with heating gradient and bacteria (Ballivy and Bellaloui, 1999).

27 2.8 Remnant Stresses and Horizontal Fracturing

As is typical with conventional construction projects and a common theme to case

studies involving expansive pyritic shales (Hawkins and Pinches, 1987 and Hoover and

Pease, 2007), the removal of overburden is required in order to construct below grade

structures such as basements, elevator shafts and utilities. Removal of overburden

changes the stress state within a body of rock but near surface stress measurements show

that exhumation never completely relieves horizontal stress as long as the bedrock

remains intact (McGarr and Gay, 1978). This near surface horizontal compression results in buckling of the bedrock near the surface and is a common occurrence on quarry floors where rapid removal is necessary (Adams, 1982). These remnant horizontal stresses are known to produce horizontal microcracks in rocks in the upper 2 km of the earth’s crust.

Figure 2.9 shows the variation of ratio of average horizontal stress to vertical stress with depth below surface.

28

Figure 2.9 A Brown-Hoek stress profile to a depth of 3 km in sedimentary basins (Plumb, 1994).

Mathgram 2.1 gives a generalized description of how remnant stresses could be

utilized to partially explain the formation of microfractures as overburden materials are removed on a typical construction project. As the bedding angle of shale bedrock

becomes more parallel to the loading direction, a significantly reduced compressive

strength results (Hsu and Nelson, 2002). Horizontally bedded shale bedrock would thus be more susceptible to stress fractures as a result of remnant stresses. Note that this is a very simplified approach and does not buckling failure theory and crystallization pressures.

29 Mathgram 2.1 - Remnant Stress Example to Explain the Potential for Microcrack Formation in Pyritic Shales

Input Parameters z := 6 Unitless Depth of Stress Element (m) z':= 6 m Depth of Stress Element h:= 5.5 m Excavation Depth for Construction of Foundation kN γ':= 26.0 Assumed Density of Shale to be Removed 3 m

kN γ' := 26.0 3 m

30 Mathgram 2.1 - Remnant Stress Example to Explain the Potential for Microcrack Formation in Pyritic Shales (Cont'd)

Average Vertical to Horizontal Stress Ratio Calculations (Plumb, 1994) 100 k +:= 0.3 min z kmin = 17.0 Minimum Ratio of Average Horizontal Stress to Vertical Stress 1500 k +:= 0.5 max z kmax= 250.5 Maximum Ratio of Average Horizontal Stress to Vertical Stress Vertical and Horizontal Stress Calculations - Before Excavation sV1 γ'z⋅:= s 156⋅= kPa Original Vertical Stress on Element V1 s k ⋅:= s H1 min V1 s H1 2647⋅= kPa Original Horizontal Stress on Element (Considering Minimum Ratio) Vertical and Horizontal Stress Calculations - After Excavation s γ'z'h⋅:= ()− V2 s V2 13⋅= kPa New Vertical Stress on Element s := s H2 H1 s H2 2647⋅= kPa "Lock-In" Horizontal Stress on Element Change in Stress State and Resulting Buckling Failure Scenario

qu := 2000kPa Assumed Compressive Strength of Shale Parallel to Bedding Plane

BucklingFailure:= "POSSIBLE"if s > q H2 u "UNLIKELY" otherwise

BucklingFailure= "POSSIBLE" Neglecting Residual Vertical Stress, sV2

31 CHAPTER 3

CHEMICAL AND PHYSICAL EXPLANATIONS

3.1 Introduction

The objective of this research is to seek a greater understanding of the different forms of pyrite within the shale of the Devonian Marcellus Formation and how this might influence the potential for heaving through the production of hydrous sulfates. The hypothesis is that the physical structure and arrangement of pyrite and calcium carbonate affects the tendency of pyritic shale to produce heave. More specifically, it is hypothesized that oxidation of macroscopic forms of pyrite, such as nodules and fossil replacements, results in chemical environments that are primarily responsible for producing heave in shale bedrock and fill. If this is true, then total pyritic sulfur content might be an incomplete and inadequate indicator of the potential for heave of Devonian

Shales. This hypothesis also suggests that the oxidation availability of the microscopic forms of pyrite, such as depositional framboids or microcrystals, may not be sufficient to cause heave or the production of significant amounts of gypsum.

Shale rock core samples containing nodules with visible pyrite from the Devonian

Marcellus Formation were utilized for the oxidation experiments. The rock cores were separated based on the presence or absence of macroscopic nodules. These samples (plus a control sample) were physically and chemically manipulated, prior to the oxidation experiments, in order to test the hypothesis stated herein.

Note that the test methods do not consider the effects of CO2 induced pH reduction since of the solution since the experiments were conducted in an open environment; however, these effects are expected to be negligible.

32 3.2 Materials and Methods

3.2.1 Experiments Using 30% H2O2

Rock core samples (5.1 centimeters (2.0 inches) in diameter) were obtained from

a potential building site located in Lewisburg, Pennsylvania. The dark gray to black

shale samples are from the Marcellus Formation of the Hamilton Group (Devonian Age) with an abundance of pyritic and/or siderite nodules (Schultz, 1999). Samples were

obtained using NQ-II wireline rock coring equipment at depths ranging between 5.8 and

7.3 meters (1.76 and 2.22 feet) below the existing surface grades.

Initial testing was performed on two rock core samples to determine feasibility for the oxidation experiments. Each of the rock cores was saw cut in half, with one of the

samples left intact and the other crushed to pass through the No. 4 sieve (4.75 mm).

Percentages of calcite or pyrite were not determined, but the sample contained macroscopic pyritic nodules and reacted with 10% hydrochloric acid indicating that the parent rock was calcitic. Similar samples within the same formation and location had calcium carbonate amounts greater than 2% by weight. Framboidal or microcrystalline pyrite (microscopic) is not detectable with a conventional hand lens, but these forms of pyrite are known to be common in this formation as evidenced in the total sulfur testing in the 10% H2O2 experiment. Photos of the trial samples are presented in Figures 3.1 and

3.2.

The filtered (2-μm) and unfiltered water-phase samples from the initial tests were analyzed using a Leeman Labs PS3000UV Inductively Coupled Plasma (ICP) Emission

Spectrometer for aluminum, iron, calcium and potassium. Calculations of equilibrium or under-saturation of the filtrate samples with respect to gypsum, calcite, ferrihydrite, and

33 H-jarosite were based on measured concentrations of metals, sulfate and pH and were

evaluated using the chemical speciation model V-MINTEQ (Gustafsson, 2007).

3.2.2 Experiments Using 10% H2O2

Six new rock cores were taken from the same location at between 5.3 and 7.6 meters (1.62 and 2.32 feet) below the existing surface grades. The samples were taken from a zone beneath the groundwater table and thus were in a reducing environment prior to extraction. There was no evidence of sulfate precipitation on the samples at the time of collection. The cores were saw-cut with half of the core crushed to pass the No. 4 sieve and the other half left intact. The crushed samples were given the “A” designation and the intact core samples were given the “B” designation.

The concentration of hydrogen peroxide was decreased from 30% to 10% to prevent boiling and to allow easier collection of samples during the oxidation process.

Preparation of samples was similar to that shown in Figures 3.1 and 3.2. Sample

6 was considered a “control sample” since there was no nodules and additional pyrite was added.

Total sulfur was analyzed on representative samples of the parent rock in the cores. The nodules were removed from these samples. Total sulfur was measured instead of pyritic sulfur due to reproducibility concerns stated in Section 2.3.3. Sulfate sulfur is rarely found in reducing environments typical of the samples collected for these tests. Hence, the total sulfur in these samples consists almost entirely of organic and sulfide sulfur. The total sulfur was determined with the LECO Sulfur Analyzer. The total microscopic sulfur in the samples ranged between 0.74 and 1.28 percent by weight.

A complete list of the total sulfur tests is presented in Table C.1 in Appendix C.

34

Figure 3.1 Saw cut shale rock cores samples from Lewisburg, Pennsylvania.

Figure 3.2 Crushed shale rock core sample from Lewisburg, Pennsylvania.

35 3.3 Results

3.3.1 Experiments Using 30% H2O2

Sufficient hydrogen peroxide was used so as to exceed the amount needed to react

with all of the pyrite. Figures 3.3 and 3.4 summarize the changes in pH and sulfate

concentration, respectively, during the reaction of each of the three samples. Figure 3.5

shows the saturation condition of filtrate sample T2B (intact core with nodules).

pH Change - 30% H2O2 Experiment 8

7

6

5 pH T 1B (Rock Core - No Nodules) T2B (Rock Core - Nodular) 4 T2A (Crushed Core - Nodular) 3

2 0 20 40 60 80 100 120 140 Time (min)

Figure 3.3 pH versus time during oxidation of initial pyritic shale samples using 30% H2O2.

36 Sulfate Concentration - 30% H2O2 Experiment 0.03 T1B (Rock Core - No Nodules) 0.025 T2B (Rock Core - Nodular) T2A (Crushed Core - Nodular) 0.02

0.015

0.01

Sulfate Conc. (M) 0.005

0 0 20 40 60 80 100 120 140 Time (min)

Figure 3.4 Sulfate concentration versus time during oxidation of pyritic shale samples using 30% H2O2.

A total of 161 mL of 30% H2O2 was used to oxidize sample T1B (bulk sample

with no macroscopic nodules). The pH rose throughout the experiment and there was no

color change, no observed rise in temperature, and no apparent formation of new

minerals. There was no visual evidence with a conventional hand lens of precipitation of

gypsum in this sample. This observation was consistent with the V-MINTEQ evaluation of the chemical composition of the filtrate.

A total of 155 mL of 30% H2O2 was used to oxidize sample T2B (bulk sample

with macroscopic nodules). The oxidation of this sample produced a violent reaction,

dramatic decrease in pH and heat sufficient to result in boiling. The reaction was so

violent that measurement of pH and sample collection was inhibited due to safety

concerns. Gypsum crystallized on the surface of the rock core upon evaporation. The

37 filtrate was in equilibrium with precipitation of gypsum, H-jarosite and ferrihydrite,

based on V-MINTEQ evaluation and on visual observation. V-MINTEQ calculations of

equilibrium are illustrated in Figure 3.5, where SI=0 means equilibrium with the solid

phase, SI<0 means under-saturation with respect to the solid phase, and SI>0 means

supersaturation with respect to the solid phase.

A total of 114 mL of 30% H2O2 was used to oxidize sample T2A (crushed sample with macroscopic nodules). The pH was relatively unchanged during the first 48 minutes of the experiment, but a violent reaction ensued during an unsupervised period that damaged the pH electrode. There was significant iron release in sample T2A. A V-

MINTEQ evaluation, based on chemical composition of the filtrate, indicated that the sample was in equilibrium with both calcite and gypsum.

It is significant that only the crushed sample came into equilibrium with calcite.

The bulk samples remained under-saturated with respect to calcite, although they contained a stoichiometric excess of calcite compared to the acidity that could be produced by oxidation of all of the initial pyrite. Only the crushed sample (T2A) with much greater surface area and therefore greater access to calcite within the parent rock material came into equilibrium with calcite. Calcite remained under-saturated in the core samples containing microscopic (T1B) or macroscopic (T2B) nodules of pyrite.

38 30% H2O2 Experiment Sample T2B - Nodular Rock Core 2 7

0 6

-2 5

-4 4 SI pH

-6 3

Calcite -8 Gypsum 2 Ferrihydrite -10 H-Jarosite 1 Al(OH)3(am) 1 pH 10 100 Time (min)

Figure 3.5 Visual MINTEQ Evaluation: Saturation Index for sample T2B reacted with 30% H2O2. SI=log(Q/Keq) where Q is the product of the activities of the dissolved species.

The data obtained for the 30% H2O2 experiments are presented in Table C.2 in

Appendix C. Also, the results of the 30% H2O2 V-MINTEQ evaluations are presented in

Table D.1 in Appendix D.

3.3.2 Experiments Using 10% H2O2

The volume (mL) of 10% H2O2 used for the comprehensive oxidation

experiments corresponds to the weight of the sample (g). This provided large

stoichiometric excess of H2O2 compared to pyrite in the samples. Except for 2B (bulk

core sample with nodules) and 6B (control samples with added pyrite), the pH steadily increased during these tests, indicating relatively fast dissolution of calcite. The results of the changes in pH throughout the experiment are shown in Figure 3.6. This was surprising given the results of the 30% H2O2 experiments.

39

pH Change - 10% H2O2 Experiments 9

8 1A 1B 7 2A 2B 6 3A 3B

pH 4A 5 4B 5A 4 5B 6A 3 6B

2 1 10 100 1000 10000 Time (min)

Figure 3.6 pH values versus time of reaction in experiments using 10% H2O2 (1A through 6B).

Iron and aluminum concentrations were much lower for oxidation using 10% versus 30% H2O2, due to higher pH values, and this prevented accurate analysis of saturation with respect to hydroxides of those metals, however a brown precipitate was observed in sample 2B and an orange precipitate was observed in sample 6B. All A samples (crushed) were super-saturated with respect to gypsum. Among the whole core samples, only 2B and 6B (control) were equilibrated with gypsum. Rust staining upon evaporation was noted on all samples containing nodules. Gypsum crystallization occurred on the sides of rock cores 2B, 3B and 6B.

Overall, experiments using 10% H2O2 supported the conclusions from the 30%

H2O2 experiments regarding formation of hydrated sulfates.

40 The results of the experimental data obtained for the 10% H2O2 experiments are

presented on Table C.3 in Appendix C.

3.4 Microscopic Observations

Figure 3.7 is an optical image of a sulfate precipitate on the core of sample T2B

from the 30% H2O2 experiment. Figure 3.8 is a Secondary Electron (SE) Scanning

Electron Microscope (SEM) image of a sulfate precipitate on the core of sample 2B from

the 10% H2O2 experiment.

Scale: 100 µm

Figure 3.7 Optic image of sulfate crystal in sample 2B (30% H2O2 experiment).

41

Figure 3.8 SEM image of sulphate crystal in sample 2B (10% H2O2 experiment).

The crystal morphology suggests that the sulfate precipitate shown in the above

figures is gypsum or calcium sulfate dihydrate (CaSO4•2H2O), which is consistent with

sulfates that have been characterized with X-Ray Diffraction techniques within this same geologic formation (Hoover, 2002). Note that the characteristics of these crystals are the same as those that have formed on samples 3B and 6B in the 10% H2O2 oxidation

experiments.

The last unknown with respect to the minerals present within the shale samples is

the nodules or the location of the macroscopic pyrite. A backscatter electron (BSE)

image of the edge of one of the nodules within the shale is shown in Figure 3.9. X-Ray

Diffraction (XRD) revealed the presence of calcium carbonate, pyrite and quartz. The

conformation of pyrite within the nodules explains the chemical and physical

observations during and after oxidation with hydrogen peroxide. The results of the X-

42 Ray Diffraction are presented in Figure E.1 in Appendix E. The Marcellus Formation is

known for pyrite (FeS2) and siderite (FeCO3) nodules according to the Pennsylvania

Geologic Survey (Geyer and Wilshusen, 1982).

Figure 3.9 BSE image of the edge of a nodule in the pyritic shale from the 30% H2O2 experiment.

3.5 Conclusions

Gypsum crystallization occurred through oxidation with H2O2 in bulk core

samples of Marcellus shale when macroscopic pyrite nodules were present. Gypsum was

not produced in the absence of macroscopic pyrite nodules, either on the surface of

discontinuities or within the rock itself.

Oxidation of microscopic pyrite (framboidal or microcrystalline) did not result in

precipitation of gypsum. Field observations at project sites in Huntingdon, PA,

Lewisburg, PA and Washingtonville, PA were consistent with the laboratory findings.

43 The laboratory and field results indicate that testing for pyritic sulfur alone is inadequate for characterization of the heaving potential of Marcellus shale.

The experimental results are consistent with the hypothesis that low-pH “hot spots” are produced around macroscopic pyrite, due to the local excess acidity, and that this results in dissolution of calcium carbonate and subsequent precipitation of gypsum at discontinuities and microfractures in bedding planes.

Behavior of other pyritic shale materials could be consistent with that of the

Marcellus shale. High concentration of pyrite should be a particular concern when constructing civil infrastructure of these shales.

44 CHAPTER 4

PHREEQC HYDROGEOCHEMICAL TRANSPORT MODEL

4.1 Introduction

PHREEQC is a computer program written in the C programming language that is designed to perform a wide variety of low-temperature aqueous geochemical calculations. PHREEQC is based on an ion-association aqueous model and has capabilities for (1) speciation and saturation-index calculations; (2) batch-reaction and one-dimensional (1D) transport calculations involving reversible reactions, which include aqueous, mineral, gas, solid-solution, surface-complexation, and ion-exchange equilibria, and irreversible reactions, which include specified mole transfers of reactants, kinetically controlled reactions, mixing of solutions, and temperature changes; and (3) inverse modeling, which finds sets of mineral and gas mole transfers that account for differences in composition between waters, within specified compositional uncertainty limits

(Parkhurst and Appelo, 1999).

PHREEQC can be utilized to model some of the processes involved in expansive pyritic shales; however, a comprehensive expansion model written entirely in PHREEQC is presently out of reach. An irreversible reaction model can be formulated to simulate the oxidation of pyrite with oxygen at various concentrations of calcite and pyrite and the subsequent precipitation of gypsum.

45 4.2 Irreversible Reaction Model

4.2.1 Introduction

The first and second runs represent typical amounts of pyritic sulfur that are

thought to be present in potentially expansive shales (Bryant, 2003). Specifically, 0.1

and 0.5 percent by weight pyritic sulfur are modeled in runs one and two, respectively.

The third run, consisting of approximately 1% pyritic sulfur by weight, considers the effects of more concentrated sources of pyrite available for oxidation within shale

bedrock. All runs considered 5% CaCO3 by weight available for reaction.

The cell model shown in Figure 4.1 describes the physical constraints applied to the irreversible reaction model. Figure 4.1a depicts a cubic block of shale measuring 34 cm x 34 cm x 34 cm with 20 fractures or sites of availability for reaction. There is 1.619 cm spacing between fractures, which is reasonable based on field observations. Figure

4.1b shows the cell after oxidation of the available pyrite in its various forms and subsequent reaction with calcite and final precipitation of gypsum. The result is a change in volume of the cell and expansion or elongation in the vertical direction given that the model is constrained from moving in the x and y directions. This assumption is considered to be a conservative approach since the main concern for swell is in the vertical direction.

46

20 Fractures at 1.619cm Spacing

34cm

34cm Fixed 34cm a) Fixed

20 Fractures with Gypsum Infilling 34cm + Swell 34cm

34cm Fixed 34cm b) Fixed

Figure 4.1 Cell model for irreversible reaction. a) Shale block measuring 34 cm x 34 cm x 34 cm with 20 fractures or zones of availability at 1.619 cm spacing. b) Shale block after expansion and infilling of zones of availability with gypsum.

47 4.3 Input Parameters

Oxygen and halite are added at a 2:1 ratio and pyrite, calcite, goethite and illite are allowed to dissolve to equilibrium and gypsum is allowed to precipitate if it becomes supersaturated. The CO2(g) partial pressure has been increased to -2.5 (log of partial

pressure) to simulate a semi-closed environment at a temperature of 25ºC. The input files

for each of the following runs are presented in Appendix F.

4.3.1 First Run (0.1% S2 and 5% CaCO3)

The first run considers 0.1% S2 sulfur by weight and 5% CaCO3 by weight

available within a 34 centimeter cube block of shale with 20 fractures or zones of

availability. Specifically, there are 0.742 moles of FeS2 and 12.705 moles of CaCO3 available within a penetration distance of 2 mm on either side of the fracture zones. The mole calculations for the first run are presented in Mathgram 4.1 at the end of this chapter.

4.3.2 Second Run (0.5% S2 and 5 % CaCO3)

The second run considers 0.5% S2 sulfur by weight and 5% CaCO3 by weight

available within a 34 cm meter cube block of shale with 20 fractures or zones of

availability. Specifically, there are 3.709 moles of FeS2 and 12.705 moles of CaCO3 available within a penetration distance of 2 mm on either side of the fracture zones. The mole calculations for the second run are presented in Mathgram 4.2 at the end of this chapter.

4.3.3 Third Run (Concentrated FeS2 and 5% CaCO3)

The third run considers approximately 1% S2 by weight (concentrated

replacement pyrite) within the fractures available for oxidation and 5% CaCO3 by weight

48 available within a 34 centimeter cube block of shale with 20 fractures or zones of

availability. The model considers the FeS2 nodules to be 1.5 cm in diameter and 0.3 cm

in thickness with an average of 14.5 nodules per fracture. Specifically, there are 6.420

moles of FeS2 and 12.705 moles of CaCO3 available for reaction. The mole calculations

for the third run are presented in Mathgram 4.3 at the end of this chapter.

4.4 Results

4.4.1 Moles in Assemblage and Molar Volume Change

The moles in assemblage were calculated by PHREEQC in each of the

irreversible reactions described previously. The PHREEQC graphical output for the

0.1% S2, 0.5% S2 and concentrated FeS2 irreversible reactions are presented in Figures

G.1, G.2 and G.3, respectively, in Appendix G. For each of the irreversible reactions,

there is an obvious dissolution of CaCO3, oxidation of FeS2, precipitation of

CaSO4•2H2O and drop in pH. Figure 4.2 shows the total volume change associated with

the dissolution of CaCO3, oxidation of FeS2, precipitation of CaSO4•2H2O and the

change in pH for each of the three irreversible reactions described in the previous section.

Figure 4.2a shows that there is little change in total volume associated with the 0.1% S2 irreversible reaction and very little drop in pH. Figure 4.2b shows a more significant increase in total volume with the 0.5% S2 irreversible reaction and a more substantial

drop in pH; although, the pH remains above 6. The concentrated pyrite run shown in

Figure 4.2c depicts a similar trend with the irreversible reaction in 4.2b until the end of

the reaction where there is a significant drop in pH.

49 0.1% S2 - 5% CaCO3 500 10

450 9

400 8 350 7 ) 3 300 6 TOTAL 250 5

GYPSUM pH 200 CALCITE 4 pH Volume (cm Volume 150 3 100 2

50 1

0 0 0 5 10 15 20 25

O2 (moles added) a)

0.5% S2 - 5% CaCO3 600 10 TOTAL GYPSUM 500 CALCITE 9 pH

) 400 8 3

300 7 pH

Volume (cm Volume 200 6

100 5

0 4 0 5 10 15 20 25 30 35 O (moles added) b) 2

Concentrated FeS2 - 5% CaCO3 1000 11

900 TOTAL 10 800 GYPSUM 9 CALCITE 700 pH 8 ) 3 600 7

500 6 pH 400 5 Volume (cm Volume 300 4

200 3

100 2

0 1 0 5 10 15 20 25 30 35 40 45 50 55 c) O2 (moles added) Figure 4.2 Total volume change associated with the dissolution of CaCO3, oxidation of FeS2, precipitation of gypsum and change in pH for a) 0.1% S2, b) 0.5% S2 and c) concentrated FeS2. See Tables H.1, H.2 and H.3 in Appendix H.

50 The results indicate that gypsum precipitates and pH stays near 7 as long as

CaCO3 remains in the sample. The pH drops radically as the CaCO3 dissolves out of the

system and FeS2 continues to oxidize. It is interesting to note that there is a significant

increase in Fe3+ concentration at the very end of the irreversible reaction for the

concentrated FeS2.

4.4.2 Swell Model

The swell model considers the total volume change associated with the

dissolution of calcite and precipitation of gypsum. The molar volume difference in these

minerals results in a total volume change. The loss of pyrite was not considered in the

change in volume calculations for the 0.1% and 0.5% S2 scenarios because of the

potential for a “honeycombing effect” and given that precipitation of other minerals such

as jarosite would likely fill the micro voids within zones of availability. The

“honeycombing effect” suggests that there would not be a collapse of voids left behind from the dissolution of pyrite and that gypsum would not replace with void space upon crystallization. Given the high density of pyrite in comparison to the other minerals in

the reaction, the resulting change in volume would not be significant. Also, for the

concentrated FeS2 model, it was assumed that the source of the FeS2 was within nodules

and that an oxidation front would move out from these areas to produce volume change.

The swell estimates shown in Figure 4.3 consider that the x and y directions of the cell

are fixed and that the change in volume corresponds directly to a change in height.

51 Estimated Swell 0.450

0.400

0.350

0.300 0.1% S2 0.250 0.5% S2 0.200 Concentrated FeS2 Swell (cm) Swell 0.150

0.100

0.050

0.000 0 5 10 15 20 25 30 35 40 45 50 55

O2 (moles added)

Figure 4.3 Swell model for 0.1% S2, 0.5% S2 and concentrated FeS2 cells with 20 fracture zones of availability. See Table H.4 in Appendix H.

The results of this exercise correspond very well with field observations.

Specifically, if the fracture spacing obtained from the regression analysis in Chapter 5 is

applied to this model and a cell height of 61 cm is estimated from core location C-1 in

Chapter 7, where 9.6 cm of heave was measured, the estimated heave would be

approximately 8.2 cm.

The change in height for the 0.1% S2, 0.5% S2 and concentrated FeS2 irreversible

reactions are estimated as 0.134%, 0.710% and 1.223%, respectively. The spreadsheets showing the calculations of volume change and swell are presented in Appendix H and examples of the swell calculations are shown in Mathgram 4.4 at the end of this chapter.

52 4.5 Conclusions

The irreversible reactions run by PHREEQC broadly describe some of the geochemical processes that occur in expansive pyritic shales. Field observation indicates that the formation of sulfates, such as gypsum, is the main mechanism of heave in pyritic shales with a calcareous component. PHREEQC may be utilized as part of a comprehensive model given what this exercise has shown; however, this program will not be able to explain the mechanics involved in the evolution of sulfate precipitation and expansion. Also, PHREEQC cannot analyze low pH microenvironments and migration of acidic capillary pore water thought to be responsible for expansive pyritic shales of the

Marcellus Formation.

The variables described in this model do not do justice to the complexities involved in real world cases of expansive shales. The number of fractures or zones of availability play a significant role in the potential for heave to occur in pyritic shale zones in which the right geochemical conditions existing for oxidation. The model is able to show that the presence of CaCO3 above a certain threshold is an important consideration in maintaining a relatively high pH. Oxidation of the pyrite by bacterial sources or significant concentrations of ferric iron is not possible until the system is overwhelmed by the presence of pyrite and lack of calcite. This low pH environment was not evident until the final stages of oxidation in the concentrated FeS2 irreversible reaction and not at all in the first two runs in which the only source of pyrite was in small quantities. The increased change in height and presence of a low pH environment highlights the importance of higher concentrates of pyrite within calcareous shales.

53 It is important to note that the mole ratio of calcite to pyrite was calculated as

17.12, 3.42 and 1.98 in the first, second and third runs, respectively. As the mole ratio

came closer to 2, the production of gypsum increased and the potential for the pH to drop

also increased. As stated in Section 2.2, a calcite to pyrite mole ratio of 2 is required to

prevent AMD or ARD in an open environment; however, as this ratio drops below this value, the pH can also decrease significantly.

54 MATHGRAM 4.1 - 0.1% S2 and 5% CaCO3 Calculations

g ρ := 2.75 Shale 3 Density of the Shale Block cm

N Fracture := 20 Number of Fractures for Reaction Availability x:= 0.34m Dimension of Shale Cube P2m:= m Penetration Distance into Shale For Reaction 2 VShale x ⋅()2P⋅ ⋅:= NFracture 3 V 0.009⋅= m Shale Volume of Shale Available For Reaction

WShale VShale⋅:= ρShale

WShale 25432⋅= g Weight of Shale Available For Reaction

CaCO3%:= 5% Concentration of Calcium Carbonate S2%CaCO3%:= 0.1:=%5 % Concentration of Pyritic Sulfur g AM := 64.13 Atomic Mass of S S2 mol 2 g AM := 100.09 Atomic Mass of Calcium Carbonate CaCO3 mol g AMFeS2 := 119.97 Atomic Mass of Pyrite mol WCaCO3:= CaCO3%⋅ WShale

WCaCO3 1271.6⋅= g Mass of Calcium Carbonate

VShale⋅ρShale mCaCO3:= ⋅CaCO3% AMCaCO3 mCaCO3 = 12.705 mol Moles of CaCO3 Utilized in PHREEQC Analysis

WFeS2 S2%⋅:= WShale

WFeS2 25.432⋅= g Mass of Pyrite

WShale mS2 ⋅:= S2% AMS2 mS2 = 0.397mol Moles of S2

⎛ AMFeS2 ⎞ mFeS2 mS2⋅:= ⎜ ⎟ ⎝ AMS2 ⎠ m FeS2 0.742⋅= mol Moles of FeS2 Utilized in PHREEQC Analysis

55 MATHGRAM 4.2 - 0.5% S2 and 5% CaCO3 Calculations g ρ := 2.75 Shale 3 Density of the Shale Block cm

N Fracture := 20 Number of Fractures for Reaction Availability x:= 0.34 m Dimension of Shale Cube P2mm:= Penetration Distance into Shale For Reaction 2 VShale x ⋅()2P⋅ ⋅:= NFracture 3 V 0.009⋅= m Shale Volume of Shale Available For Reaction

WShale VShale⋅:= ρShale

WShale 25432⋅= g Weight of Shale Available For Reaction

CaCO3%:= 5% Concentration of Calcium Carbonate S2%:= 0.5% Concentration of Pyritic Sulfur g AM := 64.13 Atomic Mass of S S2 mol 2 g AM := 100.09 Atomic Mass of Calcium Carbonate CaCO3 mol g AM FeS2 := 119.97 Atomic Mass of Pyrite mol WCaCO3:= CaCO3%⋅ WShale

WCaCO3 1271.6⋅= g Mass of Calcium Carbonate

VShale ⋅ρShale mCaCO3 := ⋅CaCO3% AM CaCO3 mCaCO3 = 12.705mol Moles of CaCO3 Utilized in PHREEQC Analysis

WFeS2 S2%⋅:= WShale

WFeS2 127.16⋅= g Mass of Pyrite

WShale mS2 ⋅:= S2% AMS2 mS2 = 1.983mol Moles of S2

⎛ AMFeS2 ⎞ mFeS2 mS2⋅:= ⎜ ⎟ ⎝ AMS2 ⎠ mFeS2 3.709⋅= mol Moles of FeS2 Utilized in PHREEQC Analysis

56 MATHGRAM 4.3 - Concentrated FeS2 and 5% CaCO3 Calculations g ρ := 2.75 Shale 3 Density of the Shale Block cm g ρ := 5.01 Density of Crystalline Pyrite FeS2 3 cm

N Fracture := 20 Number of Fractures for Reaction Availability x:= 0.34m Dimension of Shale Cube P2m:= m Penetration Distance into Shale For Reaction 2 VShale x ⋅()2P⋅ ⋅:= NFracture 3 V 0.009⋅= m Shale Volume of Shale Available For Reaction WShale VShale⋅:= ρShale WShale 25432⋅= g Weight of Shale Available For Reaction CaCO3%:= 5% Concentration of Calcium Carbonate d FeS2 := 1.5 cm Diameter of Pyrite Fossil h FeS2 := 0.3 cm Height of Pyrite Fossil N FeS2 := 14.5 Number of Pyrite Fossils per Fracture

⎛ 2 ⎞ ⎜ π ⋅dFeS2 ⎟ V := ⎜ ⋅h ⎟⋅N ⋅N FeS2 ⎝ 4 FeS2⎠ FeS2 Fracture 3 VFeS2 153.742cm⋅= Total Volume of Pyrite Fossils Available for Reaction g AM := 100.09 Atomic Mass of Calcium Carbonate CaCO3 mol g AM := 119.97 Atomic Mass of Pyrite FeS2 mol

WCaCO3:= CaCO3%⋅ WShale WCaCO3 1271.6⋅= g Mass of Calcium Carbonate

VShale ⋅ρShale mCaCO3 := ⋅CaCO3% AM CaCO3 mCaCO3 = 12.705mol Moles of CaCO3 Utilized in PHREEQC Analysis WFeS2 := VFeS2⋅ρFeS2

WFeS2 770.246⋅= g Mass of Pyrite

WFeS2 mFeS2 := AMFeS2 mFeS2 = 6.420mol Moles of FeS2 Utilized in PHREEQC Analysis

57 MATHGRAM 4.4 - Change in Cell Height Calculations

Input Parameters x:= 0.34 m Width of Cell (Fixed) y:= 0.34 m Length of Cell (Fixed) z:= 0.34 m Height of Cell (Subject to Change from Gypsum Infilling

V original xy⋅ ⋅:= z 3 Voriginal 0.0393m⋅= Original Volume of Cell

Final Change in Cell Height Calculations 3 ΔV0.1%S2:= 52.49cm Change in Total Volume from PHREEQC Model (0.1%S2) 3 ΔV0.5%S2:= 279.09cm Change in Total Volume from PHREEQC Model (0.5%S2) 3 ΔVConcFeS2 := 480.53cm Change in Total Volume from PHREEQC Model (Concentrated FeS2 Source) Voriginal + ΔV0.1%S2 Δz := − z 0.1%S2 xy⋅

Δ z 0.045⋅= cm 0.1%S2 Swell Amount in 0.1%S2 Δz 0.1%S2 %Δz := 0.1%S2 z

%Δz 0.134⋅= % 0.1%S2 Swell Percentage in 0.1%S2 Voriginal + ΔV0.5%S2 Δz := − z 0.5%S2 xy⋅

Δ z 0.241⋅= cm 0.5%S2 Swell Amount in 0.5%S2 Δz 0.5%S2 %Δz := 0.5%S2 z

% Δ z 0.710⋅= % 0.5%S2 Swell Percentage in 0.5%S2 Voriginal + ΔVConcFeS2 Δz := − z ConcFeS2 xy⋅

Δz ConcFeS2 0.416⋅= cm Swell Amount in Concentrated FeS2

Δz ConcFeS2 %Δz := ConcFeS2 z

%Δz 1.223⋅= % ConcFeS2 Swell Percentage in Concentrated FeS2

58 CHAPTER 5

EMPIRICAL REGRESSION MODEL

5.1 Introduction

This study gives a broad outline of the chemical and physical nature of expansive

shales and discusses current methods of characterizing their potential for swelling. The

samples obtained for this research are from the Marcellus Formation of the Hamilton

Group (Devonian Age) and specifically from the Evangelical Hospital in Lewisburg,

Pennsylvania (see Chapter 7 for further information); however, shale with similar

concentrations of pyrite and calcium carbonate will likely behave similarly to the

Marcellus shale because the chemical and physical properties will essentially be the

same.

The experiments presented focus on regressive analyses of pyritic shale that has

undergone significant swelling and produced heave in overlying structures. These

experiments will aid investigators in identifying bedrock conditions leading to significant

ground heave and provide a direction for future research into this phenomenon.

5.2 Observational Data

The permeability in shales decreases very rapidly at shallow depths and is greatest

parallel to the orientation of the compacted clay and laminae or concentrations of silt

sized particles (Ingebritsen et al., 2006). Wildman, et al. (2004) showed that black shale

(Upper Devonian, New Albany Formation) has a water permeability of 6.5 +/- 1. x 10-8 m2 at an effective pressure of 1.4 bars. Availability of the pyrite in the shale to oxidize in

a short period of time (< 10 yrs) is relatively non-existent within unfractures shale given

such low permeability, but may increase significantly if the shale fractures.

59 The key question for any predictive model of shale expansion is whether the

pyrite within the shale, or the pyrite that is typically tested as part of the laboratory

analysis, is sufficiently available for oxidation, reaction with calcite, and supplying sulfur

for crystallization of gypsum. The heterogeneity of a particular unit of black shale and of exposed pyrite and calcite within the shale are parameters that need to be considered for theoretical and predictive models.

Devonian black shales contain a wide variety of diagenetic pyrite morphologies, including disseminated fine crystalline pyrite, framboids, concretions, and sand sized pyritic spheres. As outlined in detail in Section 1.2 entitled, “Microenvironmental

Considerations”, it is the higher forms of concentration or macroscopic pyrite within zones of availability (discontinuities) that requires significant consideration; specifically, in nodules and fossil replacements such as Tazmanites cysts (Schieber and Baird, 2001).

The sand sized pyrite spheres found in the Devonian shales, while undeniably early

diagenetic on textural grounds are atypical compared with other early diagenetic pyrite

occurrences in the shale. The size of these grains, as well as their in situ abundance

suggests that there was much more reactive iron (grain coatings of iron oxides and oxyhydroxides) (Carroll, 1958) in the system than observed in studies of pyrite formation

in modern sediments (Canfield and Raiswell, 1991; Canfield et al., 1992). An example of

one of these large spheres is presented in Figure 5.1a as a digital photograph and Figure

5.1b as a SEM image.

60 Scale: 1cm

a)

Edge of Sphere Replacement Pyrite Fossil

Scale: 200um

b)

Figure 5.1 Replacement pyrite encountered in the Marcellus Formation in Washingtonville, Pennsylvania. (a) Digital photograph of large spheroidal pyrite replacement. (b) SEM Secondary Electron (SE) image of same pyrite replacement.

The replacement pyrite fossil shown in Figures 5.1a and 5.1b was present within an unoxidized and unweathered zone beneath a floor slab that was damaged by swelling shales. This fossil measures approximately 1 cm in diameter. Evidence of fully oxidized

61 replacement pyrite fossils is abundant throughout zones of swelling. These fully oxidized zones contain obvious iron oxides, jarosite and gypsum minerals. An example of these fully oxidize portions of the shale are shown in Figures 5.2 and 5.3.

Fully Oxidized Pyrite Replacement

Scale: 1cm

Figure 5.2 Digital photograph of fully oxidized zones of pyrite replacement along bedding plane of shale beneath floor slab at Evangelical Hospital in Lewisburg, Pennsylvania (Hoover and Pease, 2007).

Figure 5.3 Digital photograph of cross section of joint in swelled pyritic shales beneath heaved section of floor slab at Evangelical Hospital in Lewisburg, Pennsylvania (Hoover and Pease, 2007).

62 The observed presence of highly concentrated pyrite zones (Figure 5.1a and

Figure 5.1b), evidence of full oxidation of these zones (Figure 5.2) and subsequent evidence of abundant gypsum crystal formation (Figure 5.3) suggests that it is the presence of the early diagenetic (change that happens to a sediment after deposition) pyrite that is the key ingredient to identifying potentially expansive shales of the

Marcellus Formation.

Figure 5.4 presents a Concentrated Pyritic Shale Oxidation-Expansion Model of the progression of the oxidation process that leads to expansion.

63 OVERBURDEN

a)

UNOXIDIZED JOINT PYRITE SPHERE

FRACTURES

UNAFFECTED BEDROCK ZONE

RELEASE OF OVERBURDEN

b) HEAVE

UNOXIDIZED JOINT EXPANSION PYRITE SPHERE OF FRACTURES

NEW FRACTURING

UNAFFECTED BEDROCK ZONE

c) HEAVE

PARTIALLY OXIDIZED JOINT GYPSUM FILLED PYRITE SPHERE FRACTURES

FRACTURES

UNAFFECTED BEDROCK ZONE

HEAVE d)

JOINT FULLY OXIDIZED GYPSUM FILLED PYRITE SPHERE FRACTURES

UNAFFECTED BEDROCK ZONE

Figure 5.4 Concentrated Pyritic Shale Oxidation-Expansion Model. (a) Presence of concentrated pyrite within calcareous shale under overburden pressure. (b) Overburden materials are removed, which induces stress relief fracturing and initiation of pyrite oxidation. (c) Pyrite oxidation produces sulfuric acid migration through fractures resulting in the formation of gypsum. Swelling of shale matrix begins to heave overlying structure. (d) Pyrite is fully oxidized resulting in complete infilling of gypsum within the fractures and swelling is terminated.

64 5.3 Experimental Data

5.3.1 Image Analysis

Image J (National Institutes of Health) analysis of thin sections of swelled pyritic

shales reveals the presence of distinct gypsum infilling in fractures within the shale.

Image J is a computer program that allows for detailed analysis of thin sections, such as

measuring area, mean, standard deviation, lengths and angles. Three thin sections were

taken from a single piece of shale beneath a floor slab area that has heaved approximately

7.6 centimeters (3.0 inches) (Hoover and Pease, 2007). The shale contains abundant

crystalline gypsum in very fine fractures. The thin sections consisted of the following

descriptions:

S1 (PDJ): Thin Section Perpendicular to Joint

S2 (PLJ): Thin Section Parallel to Joint

S3 (PLBP): Thin Section Parallel to Bedding Plane

Three sections (Left-Middle-Right) were measured for the S1 and S2 samples to

obtain the quantitative data. Digital images of each of the Image J thin sections are

presented in Appendix I. A total of 220 and 224 measurements were taken for samples

S1 and S2, respectively. The spreadsheet results of the S1 and S2 measurements are

presented in Tables J.1 and J.2, respectively, in Appendix J. Specifically, the distance

between gypsum infilling was measured along the entire depth of the sample at three

separate cross-sections. Figures 5.5, 5.6, 5.7 and 5.8 show examples of images utilized in

the Image J analysis.

65

Gypsum Infilling

Loss of Gypsum During Thin Sectioning Scale: 1 mm

Figure 5.5 Image S1L-7 thin section of saw cut perpendicular to joint and bedding plane (PDJ).

Loss of Gypsum During Thin Sectioning Gypsum Infilling

Scale: 1 mm

Figure 5.6 Image S1M-6 thin section of saw cut perpendicular to joint and bedding plane (PDJ).

66 Gypsum Infilling

Scale: 1 mm

Figure 5.7 Image S2M-5 thin section of saw cut parallel to joint and perpendicular to bedding plane (PLJ).

Gypsum Infilling

Scale: 100 um

Figure 5.8 Image SL3-20 thin section of saw cut parallel to bedding plane (PLBP).

67 5.3.2 Measurements

The Image J analysis was completed on an optic microscope at 5X power. The analysis consisted of measuring the distance across the gypsum filled fractures to determine the amount of gypsum present within the sample, spacing between fractures and percent swell. The following tables present the average values for each of the measurements:

Table 5.1 Image J Average Measurements for Sample S1 (PDJ).

Average Total Section Thickness (mm) = 16.412

Average Total Thickness of Gypsum Infilling (mm) = 4.630

Average Change in Thickness (%) = 39.742

Table 5.2 Image J Average Measurements for Sample S2 (PLJ).

Average Total Section Thickness (mm) = 19.503

Average Total Thickness of Gypsum Infilling (mm) = 4.821

Average Change in Thickness (%) = 33.115

68 5.4 Regression Analysis

This quantitative data was utilized to determine the amount of pyrite required to

produce the calculated amount of gypsum infilling present within the sample. Figure 5.9

is a simplified model that represents the data in Table 5.2 (Sample S2 (PLJ)).

19.5030

0.4821 GYPSUM INFILLING

SHALE "PARENT ROCK"

19.5030

19.5030

Figure 5.9 Simplified Model of Image J analysis of sample S2 (PLJ) with dimensions in millimeters (mm).

The model is a regressive view of the swelling process in expansive pyritic shales

that is intended to determine the possible geochemical and physical mechanisms. The

input parameters in the model are based on known physical data, which is entered into the

regressive stoichiometry of acid generation through oxidation of pyrite. The cell model

reveals a gypsum volume of 1.839 cm3 and a total gypsum mass of 4.229 g. Mathgram

69 5.1 presented at the end of this Chapter shows the S2 and CaCO3 calculations required to

produce the swell in the shale samples.

Stoichiometric calculations reveal a value of 5.63% S2 by weight required, which

is significantly greater than the threshold limits (0.1% S2) described earlier. Also, and

more significantly, this amount of S2 is significantly greater than any measured amount of

S2 from forms of sulfur testing of the shale bedrock at this site (see Chapter 7).

Specifically, the forms of sulfur test on the unoxidized shale show approximately 2.5% to

3.5% pyritic sulfur by weight. The pyritic sulfur in this test does not account for higher

concentrations of pyrite observed in the unoxidized zone (see Fig. 5.1). The regression

analysis shows that greater amounts of pyritic sulfur are required to produce the amount

of gypsum infilling in the microfractures.

5.5 Conclusions

The results of the regression analysis show that zones of high concentrations of pyrite must be available for oxidation in order to produce the amount of gypsum occurring in fractures below this area of structural heave. Although this model is a very simplified exercise for extremely complex phenomena, it does give some idea as to the likely dynamics of the pyritic shale expansion. If shale bedrock only contains between

0.1% to 0.5% pyritic sulfur, the simple stoichiometry doesn’t allow for the amount of expansion and hydrous sulfate precipitation that is seen in the damaged structure overlying this shale. However, our results suggest that localized higher concentrations of pyrite within the shale matrix are more likely responsible for the expansion and that the documented cases of lower values may be an affliction of the sampling process.

70 An interesting correlation can be made between the results of the regression

analysis and the PHREEQC analysis presented in Chapter 4. The third PHREEQC run

models a high concentration of pyrite that would be similar to the presence of pyritic

nodules or replacement pyrite fossils. Specifically, there are 6.420 moles of pyrite and

12.705 moles of calcite in the PHREEQC model producing the greatest heave, which

gives a calcite to pyrite mole ratio of 1.98. The regression model presented in this

Chapter calculates 0.012 moles of pyrite and 0.025 moles of calcite, which gives a calcite to pyrite mole ratio of 2.08. Note that this does not assume that all available pyrite and

calcite are completely consumed, which suggests additional precipitation of gypsum would be possible. The agreement in mole ratio between two very different models is

very compelling support for the microenvironmental theory and the importance of highly

concentrated sources of pyrite in the Devonian Shales of the Marcellus Formation.

71 MATHGRAM 5.1 - Regression Analysis Calculations

Input Parameters g ρ := 2.7 Shale 3 Assumed Density of the Shale Block cm g ρ := 2.3 Density of Gypsum (Roberts, 1990) Gypsum 3 cm

N Fracture := 10 Number of Fractures for Reaction Availability x2 := 1.953 cm Dimension of Shale Cube After Heave 3 VShale2 := x2 3 VShale2 7.449⋅= cm Total Volume of Shale Cube After Heave

HGypsum := 0.04821cm Average Height of Gypsum Infilling From Image J Analysis x1 x2 −:= HGypsum⋅NFracture x1 1.471⋅= cm Original Height of Shale Block 2 VShale1 x2 ⋅:= x1

3 VShale1 5.61⋅= cm Total Volume of Shale "Cube" Before Heave

g AM := 100.09 Atomic Mass of Calcium Carbonate CaCO3 mol g AM := 119.97 Atomic Mass of Pyrite FeS2 mol g AM := 66.128 Atomic Mass of Sulfide Sulfur S2 mol g AM := 172.14 Atomic Mass of Gypsum Gypsum mol g AM := 98.06 Atomic Mass of Sulfuric Acid H2SO4 mol Stoichiometry of Calcite and Sulfuric Acid Reaction to Produce Gypsum

CaCO3 + H2SO4 + 2H2O = CaSO42H2O + CO2 + H2O Reaction Equation 1 2 VGypsum := HGypsum⋅x2 ⋅NFracture 3 VGypsum 1.839cm⋅= Volume of Gypsum Produced

W Gypsum := ρGypsum ⋅VGypsum

WGypsum 4.229⋅= g Calculated Mass of Gypsum Produced

72 MATHGRAM 5.1 - Regression Analysis Calculations (Cont'd)

M1 CaCO3 := 1mol Mols of Calcite From Reaction Equation 1 M1 Gypsum := 1mol Mols of Gypsum From Reaction Equation 1

M1 H2SO4 := 1mol Mols of Sulfuric Acid From Reaction Equation 1

WGypsum mGypsum := AMGypsum mGypsum = 0.0246mol Moles of Gypsum Produced

⎛ M1CaCO3 ⎞ mCaCO3:= mGypsum⋅⎜ ⎟ ⎝ M1Gypsum ⎠ mCaCO3 = 0.025mol Moles of Calcite Required

W CaCO3 := mCaCO3 ⋅AM CaCO3

W 2.46⋅= g CaCO3 Mass of Calcite Required M1H2SO4 mH2SO4 mCaCO3⋅:= M1CaCO3 mH2SO4 = 0.025mol Moles of Sulfuric Acid Required

W H2SO4 := mH2SO4 ⋅AM H2SO4 Mass of Sulfuric Acid Required W H2SO4 2.409⋅= g

Stoichiometry of Sulfuric Acid Production From Oxidation with Pyrite

4FeS2 + 15O2 + 14H2O = 4Fe(OH)3 + 8H2SO4 Reaction Equation 2

M2FeS2 := 4mol Moles of Pyrite From Reaction Equation 2 M2H2SO4:= 8mol Moles of Sulfuric Acid From Reaction Equation 2

M2FeS2 mFeS2 mH2SO4⋅:= M2H2SO4 mFeS2 = 0.012mol Moles of Pyrite Required From Reaction Equation 2

W FeS2 mFeS2 ⋅:= AM FeS2

W FeS2 1.474⋅= g Mass of Pyrite Required From Reaction Equation 2

73 MATHGRAM 5.1 - Regression Analysis Calculations (Cont'd)

Percent By Weight Calculations for S2, FeS2 and CaCO3 WCaCO3 %CaCO3:= VShale1⋅ρShale %CaCO3 16.234⋅= % Percent by Mass of CaCO3 Required WFeS2 %FeS2 := VShale1⋅ρShale

%FeS2 9.729⋅= % Percent by Mass of FeS2 Required AMS2 %S2 %FeS2⋅:= AMFeS2 %S2 5.363⋅= % Percent by Mass of S2 Required

74 CHAPTER 6

GEOTECHNICAL LABORATORY STUDIES

6.1 Introduction

The objective of this study was to develop a laboratory method to characterize the swell potential of pyritic shales and be able to compare the potential differences associated with homogeneous versus heterogeneous distributions of pyrite. Current techniques for characterizing the potential for swell are limited to pyritic sulfur analyses and comparisons to previous studies of heave. Attempts have been made by Ballivy and

Bellaloui (1999) and Cormier (2000) to provide more direct laboratory testing techniques for measuring swell potential; however, none of these studies has focused on the effects of microscopic and macroscopic forms of pyrite on the potential for heave or the significance of calcium carbonate in the production of gypsum.

This research has progressed through trial and error and taken into consideration past attempts by Hoover (2002) to oxidize pyritic shales in various environments. The first part of the research endeavors to oxidize various concentrations of known amounts of pyrite with bacteria in a controlled matrix of calcium carbonate and illite shale. The second part of the research focuses on kinetic oxidation methods similar to the ADTI-

WP2 Leaching Column Method currently in the process of being developed by

Hornberger, et al. (2000).

75 6.2 Materials and Methods

6.2.1 Controlled Experiments with Bacterial Oxidation

6.2.1.1 Bacteria Preparation Procedure

Acidithiobacillus ferrooxidans were obtained from American Type Culture

Collection (ATCC) in Manassas, Virginia and were shipped in a vial. This strain was recommended by ATCC for “Determining the Rate of Bioleaching of Iron from Pyrite

(ASTM Standard Test Method E1257-90)”. The bacteria were kept alive with the following propagation procedures recommended by ATCC:

1. Incubate test tube cultures under a temperature of 26ºC in an aerobic

atmosphere upon receipt. Transfer the culture to acidithiobacillus

ferrooxidans medium (see Appendix K) within one week.

2. Gently vortex the test-tube to dislodge the iron that has oxidized onto the

glass. Aseptically withdraw approximately 1.0 ml of broth culture and

transfer into 5 ml of fresh broth.

3. Incubate the broth in a static and slanted position.

4. Growth is evident within one to two weeks or when yellow-orange iron oxide

deposits are observed.

5. Transfer culture every two to four weeks.

6. Culture will remain viable for a minimum of one month and for up to three

months when stored at room temperature and without shaking.

In order to determine a rate of pyrite dissolution by Acidithiobacillus ferrooxidans, an iron assay of the growth medium was determined in order to figure out the amount of pyrite required. Specifically, the molar concentration of iron in 0.1 L of

76 growth medium was calculated to be 0.0719 mol/L. Therefore, 0.863 g of pyrite was required to match the mass of iron in the growth medium. One sample was inoculated with Acidithiobacillus ferrooxidans and one sample was uninoculated as a control measure. The results of the dissolution rate experiment on 0.863 g of pyrite passing the

No. 60 sieve are provided in Figure 6.1.

FeS2 Dissolution by A. Ferrooxidans

4

3.5 Inoculated 3 Uninoculated

2.5

2

1.5

Dissolved Fe (mM) 1

0.5

0 1 10 100 Log Time (Days)

Figure 6.1 Pyrite dissolution of 0.863 grams of FeS2 passing No. 60 sieve by inoculated (acidithiobacillus ferrooxidans) solution and uninoculated solution.

Figure 6.1 shows that it took 38 days to oxidize about 5% of the FeS2, but that the rate is increasing with time. The rate of oxidation with the bacteria is fairly slow for the purpose of a practical laboratory testing procedures for expansive materials but still much faster than without the aid of bacteria. This bacterial oxidation rate is consistent with empirical data for heaved shales.

77 A volume of 500 ml of Acidithiobacillus ferrooxidans solution was prepared for

the geotechnical testing phase of this experiment. This amount of solution was prepared based on an estimated sample matrix void ratio of 0.388.

6.2.1.2 Swell Experiment Procedure

Two samples were prepared with mixtures of illite shale fragments, limestone

(calcium carbonate) fragments and pyrite fragments. The illite shale

((K,H3O)(Al,Mg,Fe)2(Si,Al)4O10[(OH)2,H2O]) was obtained from Ward’s Science

(Catalog Reference #46E0315) and was crushed in order to achieve a Unified Soils

Classification System (USCS) designation of a “well-graded sand” or SW (ASTM, 2000) gradation shown in Table 5.1. A well-graded material was specified to limit the amount of void spaces for crystal growth. The complete grain-size analysis is presented in

Appendix L.

Table 6.1 Grain-size distribution of illite shale fragments.

Sieve Size (mm) Percent Finer 4.750 99.5 2.000 85.0 0.800 59.5 0.595 49.9 0.425 42.0 0.250 32.2 0.150 18.1

The pyrite (FeS2) was also obtained from Ward’s Science (Catalog Reference

#46E6445) and crushed to pass the No. 60 sieve (0.250mm). The high calcium limestone

(<5% MgO) was obtained from the Graymont Limestone Mine in Pleasant Gap,

Pennsylvania. The limestone was also crushed to pass through the No. 60 sieve.

78 Two expansion molds were crafted specifically for this experiment. Each of the molds contained illite shale, pyrite and calcium carbonate meeting the gradation requirements set forth in the previous section. The first mold contained 0.2% pyritic sulfur (S2), 4% limestone and 95.6% illite in a homogeneous mixture. This resulted in 3 g of FeS2, 32 g of CaCO3 and 765 g of illite. The second mold contained the same amounts of each constituent; however, approximately 1 g of pyrite was placed in three distinct layers throughout the mold instead of mixing homogeneously. A schematic diagram of the test set-up is presented in Figure 6.2.

DIAL GAGE DIAL GAGE

FRAME

POROUS PLATE PYRITE LAYER

PYRITE LAYER GROUNDWATER GROUNDWATER RESERVOIR RESERVOIR

PYRITE LAYER CLEAR CYLINDER CLEAR CYLINDER

POROUS PLATE VALVE VALVE

PYRITE, CALCITE ILLITE AND CALCITE AND ILLITE HOMOGENEOUS HOMOGENEOUS MATRIX MATRIX WITH PYRITE LAYERS

Figure 6.2 Swell testing apparatus for controlled experiments with bacterial oxidation.

The pyrite for each of the tests was inundated with 50 ml of acidithiobacillus ferrooxidans solution prior to mixture or placement with illite shale fragments and limestone.

79 The first sample, consisting of pyrite, limestone and illite, was then

homogeneously mixed with 11% moisture (Acidithiobacillus ferrooxidans solution) and

compacted in 2.5 cm (1 in) loose layers and tamped 140 times per layer with a stainless steel tamper.

The second sample, consisting of illite shale fragments and limestone, was homogeneously mixed with 11% moisture (acidithiobacillus ferrooxidans solution) and

compacted in 2.5 cm (1 in) loose layers and tamped 140 times per layer with a stainless steel tamper. 1 g of the minus No. 60 mesh pyrite was placed in each of the 3rd, 5th and

7th layers.

Each of the samples was compacted to a moist density of 22.1 KN/m3 (141 lbs/ft3) with a void ratio of 0.4.

The molds were then placed in the swell measuring frame and initial readings taken. Each of the samples was then inundated with the bacterial solution for a period of

54 days. Vertical movement of each of the samples was then measured over the next 90 day period during cycles of inundation and drainage of groundwater solution.

The water utilized to inundate the samples during the swell measurements was prepared to mimic the groundwater chemistry of the Hamilton Group compiled by Reese and Lee (1998). A pH of 7.2 was achieved with the molarity requirements for each of the chemical constituents listed in Table 6.2.

80 Table 6.2 Modeled groundwater chemistry for controlled experiments with bacterial oxidation.

Component Molarity (μM)

KNO3 20

K2HPO4 2

CaSO4 428.5

MgSO4 189.5

Na2SO4 1287

(NH4)2SO4 9

Also, at the beginning of the experiment approximately 5 ml of bacterial solution

was added to 1 g of pyrite and allowed to evaporate. The purpose of this side experiment was to verify that oxidation did in fact progress with time using the bacterial solution.

6.2.2 Swell Experiments Using Kinetic Oxidation Techniques

6.2.2.1 Swell Experiment Procedure

Bulk shale samples from the Marcellus Formation were taken from an

undisclosed site located near Washingtonville, Pennsylvania. The black shale contained possible pyrite replacements and possible vein-filling pyrite deposits as shown in Figures

6.3 and 6.4, respectively. Also samples near the surface of the site contained evidence of swelling due to the presence of various hydrous sulfates, such as gypsum and jarosite.

81 Scale: 1cm

Figure 6.3 Pyrite replacement from shale of the Marcellus Formation at a site near Washingtonville, Pennsylvania.

Scale: 1 cm

Figure 6.4 Vein-filling pyrite deposit from shale of the Marcellus Formation at a site near Washingtonville, Pennsylvania.

82

The itemized procedure for the swell experiments using kinetic oxidation techniques is described as follows:

1. Prepare two (2) samples of carbonaceous pyritic shale. Sample A shall contain

carbonaceous pyritic shale from the Marcellus Formation with no obvious

macroscopic pyrite. Sample B shall contain carbonaceous shale from the same

location as A; however, this sample should have at a significant percentage (>2%

by weight) of macroscopic pyrite that has been identified as fossil replacement,

hydrothermal and/or in any concentrated form.

2. Prepare each sample by crushing or fracturing the shale through a 12.70 mm sieve

so that the particle size distribution shown in Table 6.3 is achieved.

Table 6.3 Particle size distribution requirement of pyritic shale using kinetic oxidation techniques.

Percent Sieve Size (mm) Retained 9.525 40 4.750 35 2.360 20 2.000 2 1.180 2 0.850 1

Prepare at least 1800 grams of the shale to ensure that enough will be available to

fill the 6.4 cm (2.5 in) diameter column with at least 17.8 cm (5.0 cm) of material.

Do not include the higher concentrated forms pyrite in the shale prepared during

this step.

83 3. Separate out enough of each sample to run two forms of sulfur (DER Combination

Method for Overburden Materials) and calcium carbonate tests of the shale. The

first sample should be analyzed with a hand lens (20x) to make sure there is no

macroscopic pyrite. The second sample should contain millimeter sized flecks of

pyrite that are obvious with a hand lens (>10x).

4. After preparing the columns to allow for introduction and release of humidified

air with 10% CO2 and measurement of swell, place porous disks in the bottom of

the sample. Weigh the molds prior to introduction of sample. Compact Sample A

crushed shale in a dry condition per the grain-size distribution in Item 2 in five

layers and tamp or compact. Prepare Sample B in a similar fashion; however,

place three distinct layers of a known quantity (>2% by weight total) of

macroscopic pyrite between layers 1-2, 3-4 and 4-5. Weigh each mold with

sample and then place porous disc on top of each sample.

5. Install lines for introduction and release (through modified top cap) of oxidizing

humidified air with 10% CO2.

6. Take initial measurement of dial gages on Samples A and B.

7. Introduce humidified air with 10% CO2 into each column for six days to allow for

oxidation of pyrite. A 9:1 ratio of air to CO2 shall be established to allow for a

maximum test period of approximately 90 days. A flow rate of 2.0 L/min shall be

established and maintained for each column.

8. Take dial gage measurement of each sample on the 3rd and 6th day of the oxidation

cycle.

84 9. On the seventh day, introduce dry air into sample for 24 hours in an attempt to

encourage evaporation of capillary water.

10. Take dial gage measurement of sample at the end of the 24 hour drying period

and note any obvious formation of crystallized sulfates.

11. Repeat items 7 through 10 for 2 to 3 weeks.

12. Inundate each sample with DI water while continuing to maintain flow of air/CO2

mixture for a period of 24 hours.

13. Drain each sample and continue with air/CO2 mixture for a period of 2 weeks.

Collect effluent for sulfate anion and pH testing. Observe the response of the

samples and note drying, swelling or the crystallization of sulfates.

14. If sample has not dried after 2 weeks, repeat item 9 and note drying, swelling or

crystallization.

15. Repeat items 12 through 14 until the end of the testing period.

16. Fill void spaces within each of the samples with EpoHeat (Buehler, Ltd.) epoxy to

allow for microanalysis with an SEM.

A schematic diagram and a photograph of the setup for the swell experiments using kinetic oxidation techniques are shown in Figures 6.5a and 6.5b, respectively.

85 DIAL GAGE DIAL GAGE

EXIT VALVE EXIT VALVE

POROUS PLATE PYRITIC SHALE PYRITIC SHALE FLOW REGULATOR (2.0 L/MIN) MACROSCOPIC PYRITE

CYLINDER CLEAR CYLINDER CLEAR POROUS PLATE VALVE VALVE

HUMIDIFIED AIR/CO2 LINE

AIR CO2

3.6 0.4 1CO2:9AIR a) HOUSE AIR CO2 LINE

b)

Figure 6.5 a) Schematic diagram and b) photograph of actual setup for swell experiments using kinetic oxidation techniques.

86 6.3 Results

6.3.1 Controlled Experiments with Bacterial Oxidation

Each of the molds containing the homogeneous mixture of pyrite (A) and concentrated mixture of pyrite (B) were placed in the swell measurement frame after inundation with the bacterial solution. Measurements of vertical swelling or shrinkage of the sample were then taken periodically over the next 152 days. The valves to the molds were both open at day 56; however, no effluent was released indicating that the pore water not held by capillary tension had evaporated. The groundwater mixture was then

added at day 90 and subsequently drained at day 97. Groundwater was again added at day 144 and the experiment was terminated at day 152 due to absence of swell and

obvious hydrous sulfate crystallization. The vertical swelling or shrinkage for each of the

mixtures is shown in Figure 6.6.

87 Swell Test with Bacterial Oxidation 0 A-Homogeneous Mix

-0.05 B-Concentrated Mix

) -0.1 Drain GW Soln -0.15

Swell (mm Swell -0.2 Open Valve Inundate w/ -0.25 (No Effluent) Inundate w/ GW S oln GW S oln -0.3 0 20 40 60 80 100 120 140 160 Time (Days)

Figure 6.6 Swell test results on controlled experiments with bacterial oxidation. Note that each of the samples actually experienced a measured decrease in total height over the course of the experiment.

Sulfate anion analyses on the effluent at 97 days and 151 days was taken on each of the samples. Also, the pH of the effluent was measured upon collection. These analytical results are presented in Table 6.4.

Table 6.4 Sulfate anion and pH analytic results on effluent for homogeneous (B) and concentrated (A) pyrite mixtures.

Sample A Sample B Time (Days) Sulfate (M) pH Sulfate (M) pH 97 0.00715 8.1 0.00583 8.1 152 0.00442 7.9 0.00543 7.9

Note that the groundwater has a calculated sulfate anion concentration of 0.00132 mol/L. Also, hydrous sulfates did form within the pyrite samples that were inundated

88 outside of the experiment. The fibrous nature of the sulfates indicated that halotrichite

possibly crystallized upon evaporation. Each of the molds were saw cut upon completion

of the “swell” experiments for the purpose of identifying the presence of hydrous sulfates. No hydrous sulfates were obvious with a conventional hand lens. A cross section of top placement of the 1 g concentrated pyritic source in sample B is shown in

Figure 6.7.

Concentrated Pyrite

Scale: 1cm

Figure 6.7 Top view of concentrated pyritic source in sample B. Note the absence oxidation and sulfate crystallization features.

6.3.2 Swell Experiments Using Kinetic Oxidation Techniques

The Marcellus shale samples were subjected to forms of sulfur and calcium

carbonate testing before placement in the molds and initiation of the kinetic oxidation

and after final completion of the testing. The results of these tests are presented in Table

6.5. Note that sample A1 consisted entirely of shale fragments that were analyzed with a hand lens and were separated because of no obvious sign of macroscopic pyrite and

89 sample A2 consisted of fragments of shale that did contain visual evidence of

macroscopic pyrite “flecks”.

Table 6.5 Forms of sulfur and calcium carbonate test results on shale fragments, prior to kinetic oxidation, with (A2) and without (A1) obvious signs of macroscopic pyrite as determined with a hand lens.

Forms of Sulfur (% by Weight) Calcium Carbonate Sample Pyritic Sulfate Organic Total (% by Weight) A1 1.57 0.03 0.05 1.65 5.7 A2 2.49 0.02 0.05 2.56 5.7

During the kinetic oxidation experiment there was no heave or vertical swelling of the samples that would suggest the formation of sulfates. The only appreciable swelling or shrinkage of the samples was immediately during the initiation of the inundation and drying/oxidation periods. This swelling or shrinkage had more to do with buoyancy and particle rearrangement effects than with alteration of minerals within the sample.

The sulfate anion concentration and pH of the effluent were taken after the various inundation periods and are presented in Figures 6.8 and 6.9, respectively. These results were used to analyze the progression of pyrite oxidation and the changes to the bulk chemistry of the solution.

90 Kinetic Oxidation Column Experiment 0.016

0.014

) 0.012

0.01

0.008

0.006 A - Pyritic Shale

B - Pyritic Shale

Sulfate Concentration (M Concentration Sulfate 0.004 (Concentrated)

0.002

0 0 102030405060708090 Time (Days)

Figure 6.8 Sulfate concentration of effluent in kinetic oxidation column experiment.

Kinetic Oxidation Column Experiment 7.3

7.2

7.1

7

6.9 pH

6.8 A - Pyritic Shale 6.7 B - Pyritic Shale (Concentrated) 6.6

6.5 0 102030405060708090 Time (Days)

Figure 6.9 pH of effluent in kinetic oxidation column experiment.

91 There is estimated 2% pyritic sulfur content in the shale particles utilized for both

columns and an additional 1.6% concentrated pyritic sulfur added to column “B”. Given

the amount of sulfur leached and the estimated pyritic sulfur contents in each of the columns, a total of approximately 1% of the pyritic sulfur has been oxidized during the

84 days of the experiment. There is no geochemical evidence to suggest that the added

macroscopic pyrite in column B has undergone any appreciable oxidation.

After completion of the 84 day kinetic oxidation experiment, each of the samples

was dried and inundated with epoxy and saw cut into sections. Each of these sections

was then taken to the SEM for microanalysis and elemental mapping. Figures 6.10 and

6.11 represent the different forms of pyrite present within the shale samples and included

in each of the oxidation columns. Note the cubic morphology of the macroscopic pyrite

and the framboidal clusters of microcrystals that are invisible to the naked eye. The

microcrystals shown in Figure 6.11 measure from hundredths to tenths of a micrometer in

maximum dimension, which highlight the potentially highly reactive nature of this pyrite

morphology.

92 Macroscopic Pyrite (Cubic) Framboidal Pyrite

Quartz Vein

Figure 6.10 Backscatter image of shale fragment from “A” column in the kinetic oxidation experiment.

Pyrite Framboids (Microcrystal Clusters)

Pyrite Microcrystals

Figure 6.11 Backscatter images of framboidal clusters of pyrite microcrystals from column “B” in the kinetic oxidation experiment.

93

The microanalysis of the shale fragments and elemental mapping of each of the

samples did not show evidence of sulfate formation; specifically, there was no

precipitation of gypsum either on the edge of the particles or within the fractured zones.

Figures 6.12 and 6.13 show that an obvious oxidation “front” was induced as a result of

the kinetic experiments, but that the only byproducts were potential iron oxides.

Framboidal Pyrite Epoxy

Oxidized Zone

Edge of Unoxidized Shale Zone Fragment

Figure 6.12 Backscatter image of oxidation front into a shale fragment from column “A” in the kinetic oxidation experiment.

94 Framboidal Pyrite

Epoxy

Unoxidized Zone Oxidized Zone Edge of Shale Fragment

Figure 6.13 Backscatter image of oxidation front into a shale fragment from column “B” in the kinetic oxidation experiment.

The oxidation front has seemingly only penetrated approximately 40 to 50 microns into the shale fragments. Also, the macroscopic forms of pyrite in column “B” did not show evidence of oxidation, such as pitting or weathering, as shown in Figure

6.14.

95 Macroscopic Pyrite (Cubic Morphology)

Calcite Grain

Figure 6.14 Backscatter image of macroscopic pyrite intentionally added to column “B” for the kinetic oxidation experiments.

6.4 Conclusions

6.4.1 Controlled Experiments with Bacterial Oxidation

The controlled experiment with specified concentrations of pyrite, calcite and

illite shale produced vertical shrinkage through the various inundation and drawdown

periods due to particle rearrangement. The acidithiobacillus ferrooxidans showed

promise as a potential accelerating agent for the oxidation of the pyrite during pre-trial experiments and even produced hydrous sulfates in the side experiment on the pyrite;

however, there was no evidence of pyrite oxidation within the controlled experiment

columns. Although attempts were made to attach the bacteria to the pyrite during the

mixing process for both columns, there was no sulfate crystallization that led to a positive

change in height or heave of the entire matrix. Although sulfates were in solution based

on the geochemical data, the saturation index (SI) of the solution was apparently below 1

96 for gypsum. SI is defined as ∆Gr/(2.3026 RT), where ∆Gr is the Gibb’s free energy of the

reaction, R is the gas constant and T is the temperature (Langmuir, 1997)

The intent of the experiment was to try and induce acidic microenvironments

either around the pyrite grains and/or concentrated sources of pyrite. These microenvironments would have potentially allowed the bacteria to flourish and accelerate the pyrite oxidation by maintaining a local pH low enough for survival. The presence of the calcite throughout the sample matrix most likely kept the bulk chemistry fairly alkaline and thus did not allow the bacteria to remain alive. If the pyrite was able to be isolated from the higher concentrations of calcite, the bacteria may have been able to

establish a favorable microenvironment and produce enough acid to react with the

surrounding calcite to produce gypsum. This separation between highly concentrated

sources of pyrite and the surrounding shale bedrock exists in situ, but is not easily

duplicated in the laboratory. This may be achieved on a much smaller scale in the

laboratory by creating a slurry of bacteria saturated pyrite grains surrounded by calcite

grains in a Petri dish. The boundary between the two materials could then be monitored

for the formation and migration of gypsum through the calcite. This would not produce

any real geotechnical data regarding heave potential as a function of pyrite concentration,

but would be valuable in understanding the processes required for bacteria survival

within these microenvironments.

6.4.2 Swell Experiments Using Kinetic Oxidation Techniques

As with the controlled experiments with bacterial oxidation, the swell

experiments using kinetic oxidation techniques did not produce heave or the formation of

hydrous sulfates. The ADTI-WP2 Leaching Column data showed terrific promise with

97 respect to reproducing close to ideal oxidizing conditions for various pyritic materials.

The techniques for the leaching column experiment were massaged in order to produce an environment more conducive to the crystallization of hydrous sulfate by extending the

oxidation periods and reducing the amount of inundation cycles. The data produced

compelling evidence that diffusive oxidation was taking place within the shale particles,

but that only about 1% of the sulfur was oxidized. Also, geochemical results and

microanalysis with the SEM showed that little or none of the macroscopic pyrite had

oxidized during the 84 day oxidation cycle.

The calcite to pyrite molar ratio in the shale samples was estimated to be around

1.8, which suggests that a low pH environment may have eventually been possible even

without the elevated concentrations of CO2. The potential for a low pH environment

conducive to the solubility of ferric iron (Fe3+) and/or the survival of acid loving bacteria,

such as acidithiobacillus ferrooxidans, is more likely to occur in the “B” sample where

higher concentrations of pyrite were artificially introduced. The time required for

complete oxidation under the established laboratory conditions is not practical for swell

prediction studies. Although this kinetic leaching test has proven to be a good predictor

for AMD and ARD studies, the cross over for geotechnical concerns may not be entirely

practical. The surface area available for reaction with the crushed samples does not

represent the field conditions well enough. Specifically, the surface area available for

reaction of intact bedrock is significantly less than crushed samples. This test may work

better on rock cores with various concentrations and forms of pyrite within calcareous

shales.

98 Given the diffusion controlled kinetics required for pyritic sulfur weathering demonstrated in this experiment, it may take decades to weather the entire samples under the oxidizing conditions established in the laboratory. The weathering rates in the field would be even greater since the conditions are typically less severe. This data suggests that oxidation of concentrated sources of pyrite may be necessary in the Devonian Shales in order to produce appreciable amounts of gypsum to cause swelling. The oxidation of the concentrated sources is not likely to proceed without pyrite oxidizing bacteria to increase reaction time.

99 CHAPTER 7

CASE HISTORY: MICROPILE UNDERPINNING OVER EXPANSIVE PYRITIC SHALES

7.1 Introduction

The Evangelical Hospital is located approximately 1.6 km (1.0 mile) north of

Lewisburg, Pennsylvania in Kelly Township, Union County. The hospital is located within the Marcellus and Mahantango Formations of the Hamilton Group (Devonian

Age) as shown in Figure 7.1.

Evangelical Hospital

Figure 7.1. USGS Topographic Map - Lewisburg Quadrangle (Berg and Dodge, 1981)

100 The hospital has undergone multiple expansions since it was built in 1953. The

original building was built on the brown shales of the Mahantango Formation and black

pyritic shales of the Marcellus Formation. A perched groundwater table required the use

of foundation drains and sump pumps to keep the basements and first floor relatively free

of moisture.

The foundations consisted of conventional spread and continuous wall footings,

which supported structural steel columns and masonry load bearing walls. No signs of

significant structural distress occurred during the next forty years. In 1982, a small

addition measuring approximately 16.8 m (55 ft) square in plan dimension was built onto

the western corner of the hospital. The 1982 addition was lowered approximately 1.5 m

(5 ft) with foundations and floor slabs bearing directly on the black shales of the

Marcellus Formation. Conventional underpinning was employed to allow for support of

the walls of the original 1953 structure. In 1996, a much larger addition was built onto

the western portion of the hospital and adjacent to the 1982 addition. The 1996 addition

was lowered another 0.91 m (3.0 ft) from the 1982 addition and required a moderately

extensive permanent dewatering system to control moisture in the basement areas.

In 1998, CMT Laboratories, Inc. (CMT) was employed by Evangelical Hospital

to investigate cracking and crowning of the floor slab in the corridor between the 1982 and 1996 addition (Hoover and Thornton, 1998). The heave in the floor slab was estimated at approximately 5.1 cm (2.0 in) at the center. Hand sampling techniques revealed the presence of black shale beneath the floor slab. The shale fragments contained visible crystals, which were identified as gypsum by X-Ray Diffraction techniques. It was determined that pyrite (FeS2), present within the shale, contributed to

101 the formation of gypsum and subsequent expansion (see discussion of expansive shales in next section). The shale fill materials were subsequently overexcavated in 2002 and a structural slab was built over a 30.5 cm (12 in) void to allow for future expansion. Also, during the original investigation of the corridor in 1998, it was noted that the floor slab of the 1982 addition had begun to heave in several locations.

In 2006, CMT was again notified by Evangelical Hospital of swelling related structural distress throughout the 1982 addition as well as the 1996 addition. The swelling had become more pronounced in the floor slab areas and had begun to progress to the superstructure through the load bearing wall between the two additions. The project architect, Burt Hill, identified areas throughout the hospital that had experienced heave related distress and produced the Pyrite Location Plan shown in Figure 7.2.

102

Figure 7.2 Pyrite Location Plan (Burt Hill, 2006)

103

The area identified as the “epicenter” is within the 1982 addition and experienced

the worst structural distress. The floor slab had heaved upwards of 7.6 cm (3.0 in) in several locations and the load bearing wall was thrust upwards at least 2.5 cm (1.0 in), which has made the transition between buildings challenging to pedestrian traffic (Figure

7.3).

Figure 7.3. First floor transition between 1982 and 1996 additions.

The project’s structural engineer, Highland Associates, expressed concern over

the amount of movement and the potential for a catastrophic collapse if the expansion

were allowed to continue. Also, the utility lines that traversed between buildings could

possibly be compromised with additional differential movement between the structures.

104 Also, the heave of the floor slab in the 1982 addition had made this area impractical for continued use.

The project team decided that major renovations to the structure were in order and that the focus of the efforts would be on the 1982 addition and the load bearing wall between the 1996 and 1982 additions. Major renovations to the 1996 addition were not pursued at this time because the majority of the distress was confined to the floor slab areas and the superstructure was not involved.

CMT and Hayward Baker Inc. (HBI) developed the concept of utilizing micropiles to underpin the load bearing wall between additions and to provide support for the column footings within the 1982 addition in an effort to prevent any future heave of the superstructure.

7.2 Shale Expansion History

Prior to the initiation of comprehensive field and laboratory studies, the project team had to come to a theory of how the expansion progressed throughout the hospital.

The following steps give a brief synopsis of the theory behind the structural distress caused by expansive pyritic shales at Evangelical Hospital (Hoover, 2004):

1. Overburden shale and residual soils were removed during construction of the

basement areas in the 1982 addition. This allowed for stress relief fractures to

form within the shale matrix thus increasing the surface area of the pyrite

available for oxidation.

2. The site was dewatered to an elevation just below the newly excavated subgrade.

This allowed the oxidation process to begin, although the majority of the shale

was still saturated and below the groundwater table.

105 3. The foundations and floor slabs were constructed directly on top of the saturated

pyritic shale bedrock.

4. Construction of the 1996 addition required that the perched groundwater table,

located below the 1982 addition, be lowered by more than 1.2 m (4 ft).

5. The exposed shale beneath the 1982 addition is now no longer inundated and is

subject to oxidation by oxygen and bacteria (acidithiobacillus ferrooxidans). The

bacteria increase the oxidation rate of the pyrite by orders of magnitude.

6. The oxidation of replacement pyrite, which was identified in high concentrations

throughout the heaved zone, produces sulfuric acid. The sulfuric acid is

mobilized throughout the fractures and discontinuities within the vadose zone and

reacts with any calcium carbonate located within the shale to produce gypsum.

7. Gypsum forms within the fractures, but also within the intact shale matrix and

produces swelling of the shale and displacement of the overlying structures.

8. The swelling produces markedly more heave over lightly loaded floor slabs than

beneath foundations; however, the main bearing walls are forced upwards over

2.5 cm (1.0 in).

7.3 Field and Laboratory Investigations

7.3.1 Field Investigation

The field investigation consisted of coring through the concrete slab at three

locations for the purpose of sampling the underlying subgrade materials, coring the slab

adjacent to the existing wall to verify the location of the adjacent continuous wall footing foundation and to measure the flatness of the floor throughout the heave areas. The three cores for the subgrade sampling are labeled C-1, C-2 and C-3 and the core adjacent to the

106 footing is labeled CC-1. The approximate locations of the cores are shown on the Floor

Flatness Survey in Figure 7.4.

0.00 (REF. ELEV.)

CC-1 +1.90 +2.95 +0.90 C-2 +3.15 +2.55 +0.90 +2.10

+1.70 +0.60 +1.70

+2.15 +0.55 +1.75 +2.20 +2.95 +2.7015 .55 .70 65 +3. 75 50 0 0 1. 3. 3. +0.85 +0.45 + + +1.10 + +3.30 + + +2.70 +0.90 .50 .80 C-1 75 2 +0.95 + +2.85 +3 +3.70 +1. +0.90 +0.90 +0.90 +0.90 +0.90 +0.80 +0.75 +0.75 +3.45 +0.75 +0.20 +3.75 +0.65 +2.70 +0.50 +0.25 +1.65 C-3 +0.10 -0.20 -0.10 +0.75 -0.25 -0.65 -0.25 -0.65 -0.50 -0.55 -0.80 -0.65 -0.60 -0.35

Figure 7.4 Core Location and Floor Flatness Plan in the Information Sciences sector of the Evangelical Hospital.

107 Core location C-1 was taken in the area experiencing the greatest amount of

heave. The concrete slab measured approximately 6.4 cm (2.5 in) in thickness and the

core was fractured vertically. The concrete slab was underlain by a plastic vapor barrier, which was overlain by about 10.2 cm (4.0 in) of AASHTO #57 limestone gravel. The underlying materials were sampled to a depth of about 61.0 cm (24.0 in) below the top of the slab with a Denison Tube. These materials consist of black shale, which appeared to have been placed as fill. A split-spoon sampler was then extended to a depth of about

88.4 cm (34.8 in) below the top of the slab. The materials encountered within the split- spoon sampler appeared to consist of natural, weathered black shale. These natural materials had moisture contents ranging between 3 and 6 percent by weight.

Core location C-2 was taken in the office adjacent to core location C-1. The concrete slab measured approximately 7.6 cm (3.0 in) in thickness and the core was fractured vertically. The concrete slab was underlain by a plastic vapor barrier, which was overlain by about 10.2 cm (4.0 in) of AASHTO #57 limestone gravel. The underlying materials were sampled to a depth of about 42.7 cm (16.8 in) below the top of the slab with a split-spoon sampler. A Denison Tube was then utilized to sample the materials to a depth of about 73.2 cm (28.8 in) below the top of the slab. The materials appeared to consist of natural, weathered black shale, which had a moisture content of about 3 percent by weight.

Core location C-3 was taken in the office adjacent to the existing storage room.

The concrete slab measured approximately 7.6 cm (3.0 in) in thickness. The concrete slab was underlain by a plastic vapor barrier, which was overlain by about 10.2 cm (4.0 in) inches of AASHTO #57 limestone gravel. The underlying materials were sampled to

108 a depth of about 42.7 cm (16.8 in) below the top of the slab with a split-spoon sampler.

A 10.2 cm (4.0 in) diameter core barrel was then utilized to sample the materials to a

depth of about 79.2 cm (31.2 in) below the top of the slab. The materials consisted of

natural, weathered black shale.

Core location CC-1 revealed the presence of a concrete footing directly beneath

the floor slab. The floor slab measured approximately 7.6 cm (3.0 in) in thickness and

the footing extended about 20.3 cm (8.0 in) away from the existing wall.

The results of the floor flatness survey are also shown on Figure 7.4. The values shown on the plan are in inches of variance from the reference in the existing corridor.

The red values indicate a floor level above the reference and the blue values indicate a floor level below the reference. The maximum upward change in floor elevation was about 9.65 cm (3.80 in) in the vicinity of core location C-1. An upward elevation difference of 8.00 cm (3.15 in) was measured in the area surrounding core location C-2.

The office area adjacent to the existing storage room was measured because of recent movement noticed by facility engineering personnel in the form of the door scraping on the floor. The maximum upward elevation difference in this office was about 2.41 cm

(0.95 in).

7.3.2 Laboratory Testing

Mahaffey Laboratories, Ltd. performed forms of sulfur analysis (sulfide (pyritic),

sulfate and organic), chromium reducible sulfur and calcium carbonate testing on the

samples taken from C-1, C-2 and C-3. The pyritic and sulfate sulfur percent by weight

results from the forms of sulfur test and the pyritic sulfur from the chromium reducible

109 sulfur test are presented in Figure 7.5a, 7.5b, and 7.5c for core locations C-1, C-2 and C-

3, respectively.

Core Location C-1

3 S2 (Forms of Sulfur)

2.5 S2 (Chromium Reducible)

SO4 (Forms of Sulfur) 2 t

1.5

% byWeigh 1

0.5

0 0 1020304050607080 Depth (cm) a)

Core Location C-2

1.4

1.2

1 t

0.8

0.6 % byWeigh

0.4 S2 (Forms of Sulfur)

0.2 S2 (Chromium Reducible) SO4 (Forms of Sulfur) 0 0 10203040506070 Depth (cm) b)

110 Core Location C-3 3.5

S2 (Forms of Sulfur) 3 S2 (Chromium Reducible)

SO4 (Forms of Sulfur) 2.5 t 2

` 1.5 % byWeigh

1

0.5

0 0 1020304050607080 Depth (cm) c)

Figure 7.5 Percent by weight results for pyritic sulfur (forms of sulfur and chromium reducible) and sulfate sulfur (forms of sulfur) at core locations a) C-1, b) C- 2 and c) C-3 in the IS area of the Evangelical Hospital.

Average calcium carbonate was measured to be 8.9% by weight. The pyritic

sulfur values increase with increasing depth and the sulfate sulfur values, with the

exception of C-1, decrease with increasing depth. These values indicate that the upper weathered zone of the shale bedrock has undergone pyrite oxidation and formation of sulfates and that the deeper zones are relatively unoxidized. The high pyritic sulfur values in the deeper unoxidized zone obtained from the forms of sulfur tests indicate either a high percentage of microscopic pyrite or the inclusion of macroscopic pyrite in the laboratory samples.

In order to compare the accuracy between the forms of sulfur and the chromium reducible sulfur tests, three samples were prepared with known quantities of pyrite

111 (sulfide sulfur), gypsum (sulfate sulfur) and albumin (organic sulfur). These samples

were sent to Mahaffey Laboratories, Ltd. and the results and corresponding control

sample percentages are shown on Table 7.1.

Table 7.1 Comparison between forms of sulfur and chromium reducible sulfur testing methods on control samples provided to Mahaffey Laboratories, Ltd. Chromium Forms of Sulfur Reducible Sulfide Control Sulfate Control Organic Control Sulfide Control Sample (%) (%) (%) (%) (%) (%) (%) (%) 0.46 0.50 0.25 0.20 0.07 0.10 0.48 0.50 A Δ = 8% Δ = 25% Δ = 30% Δ = 4% 0.94 1.00 0.41 0.40 0.18 0.20 1.06 1.00 B Δ = 6% Δ = 2.5% Δ = 10% Δ = 6% 1.83 2.00 0.81 0.80 0.36 0.40 2.14 2.00 C Δ = 8.5% Δ = 1.25% Δ = 10% Δ = 7%

The results of the control sample tests indicate relatively good accuracy between

the forms of sulfur and chromium reducible sulfur tests. The results for pyritic sulfur

indicate that either test would be a good method with the chromium reducible method

being slightly better. The results for sulfate and organic sulfur in the forms of sulfur test

indicate that greater accuracy comes with increased sample amounts. Given that the

forms of sulfur and chromium reducible sulfur tests require different sample sizes, it is difficult to determine the most accurate method of measuring sulfide sulfur unless a truly homogeneous sample can be provided.

112 7.4 Observations During Demolition and Overexcavation

7.4.1 Demolition and Removal of Floor Slab

Prior to implementation of the underpinning work to be accomplished by HBI, the

existing slab-on-grade in the Information Sciences portion of the 1982 addition and the

structural slab in the corridor area of the 1996 addition (renovated 2002) had to be

removed by the demolition contractor. Removal of the carpet and tiling within the

Information Sciences area revealed interesting distress to the floor slab in the form of

extensive cracking and heaving. The demolition contractor commenced with saw-cutting

of the slab-on-grade in the area surrounding southwestern most column of the IS area. As a square shape was cut out of the slab in a dimension similar to the anticipated footing size, the entire floor slab outside of the column surged upwards a distance of approximately 2.5 cm (1.0 in). This surprising turn of events destroyed the saw blade of

the cutting tool and sent a loud blasting noise throughout the basement of the hospital.

Figure 7.6 shows the location of the incident, which gave the project team a healthy

respect for expansive shales.

113

Figure 7.6. Upward release of floor slab under pressure at the southwestern column in the IS area of the Evangelical Hospital.

7.4.2 Swelled Shale Observations

The upper swelled zone beneath core location C-1 is shown in Figure 7.7a. This area was hand excavated along a joint in the shale bedrock. There was obvious iron stained and the presence of jarosite and gypsum was readily observed with a conventional hand lens.

114

Figure 7.7 Swelled shale sections beneath core location C-1 with in-situ photograph.

Evidence of “hot spots” within the shale indicates that the replacement pyrite may have undergone oxidation and radiated sulfuric acid out into the microenvironments within the surrounding shale matrix.

7.5 Micropile Design

7.5.1 Structural Loading Information

As discussed previously, the goal of retrofitting the 1982 addition was to provide

a reasonably cost effective method of creating distance between the structure and the

underlying expansive pyritic shales. Since there is no known method of treating the

shales to limit future movement and the heterogeneous nature of the shales precludes

estimation of when the shales might stop expanding, underpinning the conventional

foundations was the only viable option. An overview of the project area is presented on

the Underpinning Plan shown in Figure 7.8.

115

1996 Addition Main Bearing Wall Computer Room 1996 Addition

Telephone Room

Information Sciences 1982 Addition

Figure 7.8 Underpinning plan for the Evangelical Hospital.

116 The maximum unfactored service load on the main bearing wall was given to be

75.8 kN/m (5.2 kips/ft). The maximum unfactored service loads on the column footings

within the Information Sciences portion of the 1982 addition were given to range

between 489 kN (110 kips) and 890 kN (200 kips). The maximum unfactored service

loads on the column footings along the main bearing wall were given to range between

334 kN (75 kips) and 378 kN (85 kips). The maximum unfactored service load in the

Computer Room of the 1996 addition was given to be 1,285 kN (289 kips). Using a

factor of safety (FS) of 2.5, a maximum allowable bond zone frictional capacity of 172

kPa (25 psi) was utilized in the design of the micropiles (FHWA, 2000).

7.5.2 Design Method 1 – Telephone Room

Design Method 1 consisted of underpinning the continuous wall footing

foundation adjacent to the “telephone room” area, which stipulated that the underpinning

activities could only take place on one side of the wall. Removal of the extensive

telephone operations was not considered as an option by the hospital due to obvious

financial and logistical obstacles. The underpinning consisted of installing two (2)

micropiles on the one side of the wall and extending two C-channels beneath the wall and

attaching to the micropiles. Two (2) C12x25 Grade 36 C-channels were chosen based on

a maximum calculated bending moment of 73.5 kN-m (650 kip-in) per section.

Essentially, the far micropile acts in tension and the near micropile acts in compression.

The micropiles consisted of 17.8 cm (7.0 in) OD – 15.7 cm (6.184 in) ID 552 MPa (80

ksi) threaded steel pipe in 0.914 m (3.0 ft) sections. One No. 10 full length all-thread

reinforcement bar was utilized in the tension pile with the same reinforcement utilized in

117 the bond zone of the compression pile. This underpinning detail is presented in Figure

7.9.

Figure 7.9 Design Method 1: Telephone room underpinning detail for main bearing wall at Evangelical Hospital.

118 The construction sequence of Method 1 consisted of first installing the piles along

the corridor adjacent to the telephone room at 1.2 m (4.0 ft) centers. Next, excavation

around all odd micropile pairs and under the adjacent footing to allow for installation of

the C-channels was completed. After the non-shrink grout between the C-channels and

the footing had cured for 3 days, excavation around even micropile pairs and under the

adjacent footing commenced 1 pair at a time, with a 3-day waiting period between.

Excavation of the shale bedrock was very challenging and the carbonaceous nature of the

black shale made the working environment similar to a mining operation as depicted in

Figure 7.10 and 7.11. Figure 7.12 shows a close up of the cantilevered underpinning and

Figure 7.13 shows a broad view of the same underpinning for the telephone room area.

Figure 7.10 Underpinning excavation around cantilever C-channel beams adjacent to the telephone room at Evangelical Hospital.

119

Figure 7.11 Underpinning excavation around micropiles and under adjacent footings along telephone room at Evangelical Hospital.

Figure 7.12 Telephone room underpinning close up of cantilever at Evangelical Hospital.

120

Figure 7.13 Broad view of telephone room underpinning at Evangelical Hospital.

7.5.3 Design Method 2 – Corridor

Design Method 2 consisted of installing micropile underpinning on both sides of

the continuous wall footing foundation as well as the column/spread footing foundations

along the former Information Sciences portion of the 1982 addition. Access to both the

corridor and the interior of the 1982 addition permitted the design of a saddle-type underpinning scenario, which allowed the designers to utilize the piles in compression on both sides of the wall. The same micropile was utilized as outlined in Design Method 1; however, one No. 10 full length all-thread reinforcement bar was utilized in both piles.

This underpinning detail is presented in Figure 7.14.

121

Figure 7.14 Design Method 2: Wall and column/spread footing underpinning detail along 1982 addition at the Evangelical Hospital.

122 Sequencing of the underpinning commenced in a similar fashion to the telephone

room; however, special attention was paid to the 1.2 m (4.0 ft) wide spread/column footings, which required that two pairs of underpinning micropiles be installed approximately 0.76 m (2.5 ft) from one another. After each of the four micropiles were installed adjacent to the footing, one portion of the spread footing was overexcavated, the

C-channels installed, then the other portion of the footing was supported with temporary cribbing while the final set of C-channels were attached to the other pair of micropiles.

Figures 7.15 and 7.16 show close up views of the saddle type underpinning beneath the spread/column footings.

Figure 7.15 Saddle underpinning view from top along basement corridor adjacent to IS area of the Evangelical Hospital.

123

Figure 7.16 Saddle underpinning view from bottom along corridor adjacent to IS area of the Evangelical Hospital.

7.5.4 Design Method 3 – Isolated Spread Footings

Design Method 3 of underpinning originally consisted of bonding the micropiles

to the spread footing foundations through shear rings and high strength grout and

overexcavating 30.5 cm (12.0 in) of expansive shale beneath the footings. This would

essentially create stilts around the footings to allow for continued expansion of the shale

without creating upward lift on the footings and superstructure. A schematic of the

intended underpinning design for the spread footing foundations is presented in Figure

7.17.

124

Figure 7.17 Design Method 3: Interior spread footing underpinning detail within IS area at Evangelical Hospital.

There were seven spread footings identified by the project team that required underpinning with maximum service loads ranging between 489 kN (110 kips) and 1,285 kN (289 kips). Six of the spread footings were located within the Information Sciences portion of the 1982 addition and the remaining footing (1,285 kN footing) was located within the Computer Room of the 1996 addition. Based on the original structural

125 drawings provided by the hospital, the footings sizes ranged between 1.5 m x 1.5 m x

35.6 cm (5 ft x 5 ft x 14 in) for the 489 kN (110 kip) column and 2.4 m x 2.4 m x 55.9 cm

(8 ft x 8 ft x 22 in) for the 1,285 kN (289 kip) column.

It was determined that four 14.0 cm (5.50 in) OD – 11.9 cm (4.67 in) ID 552 MPa

(80 ksi) threaded steel casing were to be installed at each of the spread footings. The orientation of the micropiles was determined based on accessibility to each of the columns and the depths were determined based on the aforementioned geotechnical parameters.

The design was based on a 20.3 cm (8.0 in) diameter roughened core hole through the footing with a 14.0 cm (5.50 in) OD casing with 9.5 mm x 9.5 mm (3/8 in x 3/8 in) shear rings welded on at 7.62 cm (3.0 in) on center. Shear rings were placed within the bond zone to allow for increased load transfer (Sehn, 1998 and Sehn, 2000). The design allowable bond strength was 860 kPa (125 psi) with a factor-of-safety (FS) = 2.0. A detail of these shear rings is shown in Figure 7.18a and a photograph is presented in

Figure 7.18b.

126

a)

b)

Figure 7.18 Design Method 3 shear ring a) detail and b) field casing for the Information Sciences (IS) area of the Evangelical Hospital.

127 The biggest concern to the project team was the ability of the micropiles to bond

to the spread footings and provide a minimum factor-of-safety of 2.0 given the lack of redundancy for the new system. Therefore, two load tests were run by HBI to determine the allowable bond strength between the footing and the casing. Figure 7.19 shows a schematic of the load test set up and Figures 7.20 and 7.21 are photos of the actual load tests for the 489 kN (110 kip) and 1,285 kN (289 kip) footings, respectively.

HEX NUT 20.3cmX20.3cmX2.54cm STEEL PLATE

1.33 MN JACK O.D. PISTON

30.5cm O.D. STEEL PLATE WITH 20.3cm I.D. HOLE 1.33 MN JACK NO. 18 ALL-THREAD 20.3cm DIA. CORED HOLE 15.2cm O.D. PISTON NEAT CEMENT GROUT 14.0cm O.D. STEEL CASING 30.5cm O.D. STEEL PLATE CONCRETE WITH SHEAR RINGS WITH 20.3cm I.D. HOLE FOOTING NO. 18 ALL-THREAD GRADE 75 PLAN VIEW

15.2cm O.D. STEEL PLATE HEX NUT

ELEVATION VIEW

Figure 7.19 Design Method 3 footing to micropile bond detail for the IS area at the Evangelical Hospital.

128

Figure 7.20 Load test at the 489 kN footing in the IS area at Evangelical Hospital.

Figure 7.21 Load test at the 1,285 kN footing in the Computer Room at Evangelical Hospital.

129 Each of the 14.0 cm (5.5 in) OD load test casing sections were installed in a

roughened 20.3 cm (8 in) diameter cored hole with a 34.4 MPa (5 ksi) neat cement grout

mix. Actual laboratory compressive strength tests on grout samples from this mix resulted in 63.8 MPa (9,270 psi). The load test for the 489 kN (110 kip) column footing,

which actually measured 38.1 cm (15.0 in) in thickness, was taken to 583 kN (131.25

kips) and the load test for the 1,285 kN (289 kip) column footing, which measured 55.9

cm (22 in) in thickness, was taken to 667 kN (150 kips). This resulted in mobilized

average test bond strengths of 50.6 kPa (348 psi) and 39.4 kPa (271 psi) for the 489 kN

and 1,285 kN column footings, respectively. The recorded movements for the 489 kN

(110 kip) and 1,285 kN (289 kip) columns were 0.18 cm (0.071 in) and 0.33 cm (0.129

in), respectively with permanent sets of .086 cm (0.034 in) and 0.19 cm (0.076 in),

respectively. The results of the load tests for the 489 kN (110 kip) and 1,285 kN (289

kip) column footing bond are presented in Figures 7.22 and 7.24 respectively.

130 0.08

0.06

0.04 ELONGATION (IN)

0.02

MINIMUM ELONGATION PREDICTED = 80% PL/AE

0 0 20 40 60 80 100 120 140 160 LOAD (KIPS)

Figure 7.22 489 kN column footing bond stress load test results in the IS area at the Evangelical Hospital.

131 0.16

0.14

0.12

0.1

0.08 ELONGATION (IN)

0.06

0.04

0.02 MINIMUM ELONGATION PREDICTED = 80% PL/AE

0 0 20 40 60 80 100 120 140 160 LOAD (KIPS)

Figure 7.23 1,285 kN column footing bond stress load test results in the Computer Room at the Evangelical Hospital.

The results of these tests were considered satisfactory by the project team provided that each of the footings were symmetric and poured to dimensions close to the

1982 and 1996 drawings.

132 7.6 Construction Considerations

7.6.1 Materials and Equipment

Installation of these micropiles began with demolition of the existing floor slab to

expose the top of each footing. This provided access to each micropile location with the

drill rig. A 20.3 cm (8.0 in) core drill was penetrated through the footing at each

micropile location. The thickness of the footing was verified. The side walls of the core hole were roughened with a grinding wheel mounted in a die grinder to promote bond capacity of the micropile to the footing. The plain casing was extended to design depth, again in 0.914 m (3.0 ft) sections. The top piece of casing featured the shear rings as

described previously. The central reinforcing bar was installed and the micropile was

grouted with the tremie method, including the annulus between the existing footing and

the shear ring casing.

Figure 7.24 shows a Klemm KR 702 electric-hydraulic drill rig that was utilized by HBI to access the basement of the hospital. The drill rig and associated equipment

had to be lowered in sections through the service elevator. The compressor was set up

outside of the hospital to operate the down-the-hole hammer and flush the drill hole. A rotary percussive duplex method was utilized to advance the casing and construct the micropiles.

133

Figure 7.24 Klemm KR 702 electric drill rig utilized to installed micropiles at the Evangelical Hospital.

7.6.2 As-Built Challenges

To the surprise of each of the team members as well as the owner, overexcavation

activities after micropile installation within the Information Sciences room of the 1982

addition revealed some surprising spread footing configurations. It was determined that

the footings were not poured symmetrically around each of the columns and in fact were offset by as much as 30.5 cm (12.0 in) from the intended locations shown on the project

drawings. The spread footings were each poured utilizing the surrounding shale as

forms. One of the footings was basically nonexistent on the northern side of the

reinforced concrete column. Four of the seven column footings had one or more locations where there was not sufficient distance from the edge of the micropile to the edge of the footing, which was considered a distance equal to the footing depth.

134 Additionally, the thickness of these footings was significantly reduced at the edge of the footing compared to the location of the micropiles. This revelation forced the team to go back to the drawing board and determine a suitable path forward for the project, leading the project team to reconsider the use of micropiles as underpinning in the intended “stilt” configuration. The project team decided that the lack of redundancy required the utmost confidence in the work of the preceding contractors.

Since the goal of the underpinning was to limit or eliminate the potential for heave of the footings from swelling of the expansive shale, the project team decided to analyze the potential for utilizing the piles as tie-downs. The designers knew that the weight of the footings provided some source of limiting the expansion of the hydrous sulfate crystals within the pyritic shales, but the contribution of the micropiles, originally designed in compression, was in question. The bond of the casing to the footings was known and well tested and the threading and tensile strength of the casing could provide us with the capacity that was needed. Also, the development length of the reinforcing steel into the casing was sufficient as was the bond zone strength. Therefore, the pile itself could resist the expansive force anticipated from the shale. The last question that remained was the integrity of the spread footings. After a finite element analysis of each of the footings using PCA Mats (Portland Cement Association, 2003), it was decided to rebuild the footings with new concrete and steel doweling to produce footings that could provide the minimum required edge distance as originally expected in the “stilt” design.

The shale was left in place and the micropiles took on the role as tie-downs to limit upward movement of the footings.

135 7.7 Conclusions

The use of micropiles to underpin a structure that has been damaged by expansive pyritic shales is considered a viable means of remediation. Typically, the majority of the

damage induced by expansive pyritic shales is limited to lightly loaded floor slab areas

that require the use of a structural slab tied into the surrounding foundations; however,

when the foundation expansion reaches to the superstructure, then the use of micropiles

becomes a perfect fit for remediation. Micropile technology has the flexibility to reach

into tight spaces and install high capacity elements into existing conventional shallow

foundations. The limitations involved with the precision of micropile placement should

be well understood before design is complete and certainly before mobilization of the

drilling equipment. Finally, the design engineers cannot take for granted that the existing

structural elements have been built as shown on project drawings. A detailed

investigation of the existing structural elements should be implemented as part of the

design process in order to limit challenges to the project in the future.

136 CHAPTER 8

SUMMARY, CONCLUSIONS AND RECOMMENDATIONS

8.1 Summary

Case studies on potentially expansive shales have focused on microscopic framboidal and microcrystalline pyrite as the leading mechanism responsible for the formation of hydrous sulfates. This accepted hypothesis is an artifact of the identification process by which the shales are tested in the laboratory.

The H2O2 oxidation experiments showed that gypsum would crystallize on the surface of rock cores that contained a concentrated source of nodular pyrite. Gypsum did not crystallize on the surface of cores containing only microscopic sources of pyrite.

Crushing the samples showed that calcite availability would overwhelm the system during oxidation and keep the solution fairly alkaline.

The PHREEQC geochemical modeling and regression analysis showed that a calcite to pyrite molar ratio of approximately 2:1 is needed to produce significant heave in the form of gypsum precipitation. This finding is consistent with AMD science where there is CO2 release in an open environment. The PHREEQC modeling highlighted the importance of greater amounts of pyrite available for oxidation in the presence of a consistent source of calcite. The regression analysis showed that gypsum infilling in shales beneath an area of documented structural damage would require a higher percentage of pyrite than is available in microscopic form. The empirical evidence gathered from the Marcellus Formation indicates that this ratio is not likely to drop below

2:1 unless a concentrated pyritic source is available for oxidation.

137 The geotechnical laboratory studies indicate oxidation of pyrite and subsequent

precipitation of gypsum is not easily achievable. Bacterial oxidation of pyrite in a

calcareous environment is not possible unless the pyrite can be isolated. Creating

accelerated oxidative conditions is possible with conventional kinetic laboratory

techniques; however, oxidation of the microscopic pyrite is limited by slow oxygen

diffusion into the shale fragments. Also, oxidation of the macroscopic pyrite does not

progress without significant lowering of the bulk pH to allow for acidithiobacillus

ferrooxidans survival and dissolution of ferric iron.

8.2 Conclusions

The geochemical and geotechnical science of the potentially expansive pyritic

shales of the Marcellus Formation is highly complex. Total sulfur and/or forms of sulfur

laboratory tests on shale fragments, as a sole indicator of swell potential, has led to

oversimplifications. Oxidation of microscopic pyrite is possible through oxygen

diffusion into the shale, but this slow weathering process does not explain the rapid evolution of gypsum precipitation into microfractures and discontinuities that have been

documented in case studies. The macroscopic or higher concentrations of pyrite within

the calcareous shales of the Marcellus Formation aid in producing acidic

microenvironments. The acidic microenvironments can produce conditions that will

allow for survival of acidithiobacillus ferrooxidans and dissolution of ferric iron that is

necessary to explain the fast precipitation of gypsum within the microfractures and

discontinuities of the calcareous shale matrix. The absence of flushing or movement of

capillary water in these microenvironments allows for the build up of sulfates and

eventually supersaturation of gypsum. The precipitation of gypsum within the vadose

138 zone produces enough volume change and crystallization pressure to damage overlying

infrastructure.

8.3 Recommendations

8.3.1 Changes to the State-of-Practice

Remediation of potentially expansive pyritic shales should be conducted only

after a thorough field, laboratory and desktop study. If pyritic shales are anticipated

based on a geologic review of a particular site, the amount of test borings should be

increased by a minimum factor of 2. Test pits should also be considered whenever

possible so that the bedding planes and fracturing can be adequately assessed. Rock

coring should commence upon split-spoon refusal instead of auger or casing refusal so

that the character of the intact shale can be better analyzed. The shale should be

meticulously analyzed by an experienced geologist or geotechnical engineer for the

presence of concentrated forms of pyrite. The unoxidized shale samples that do not show

visual evidence of concentrated pyrite should be subjected to forms of sulfur testing to

determine the amount of microscopic pyrite available for reaction. The amount of

calcium carbonate in the sample should be critically assessed to determine the potential

for conversion to gypsum. Any sulfate crystals in the weathered or oxidized zones should be classified with X-Ray Diffraction and/or elemental mapping with a Scanning

Electron Microscope. The shale cores and intact samples should also be subjected to microanalysis to determine fracture spacing and potential zones of gypsum infilling.

This data should then be utilized to create a stoichiometric model of swell potential. A zone of potential swelling should be established and conditions identified which could lead to accelerated oxidation of macroscopic forms of pyrite. The depth of

139 the vadose zone beneath a structural member should be established where capillary pore

water can migrate within microfractures and discontinuities.

Each project should be treated on a case-by-case basis given the complexities

involved with expansive pyritic shales. Recognize that simple laboratory tests and

comparisons to case histories involving heave is an oversimplified approach and could

lead to vastly conservative or liberal judgments on the part of the geologist or engineer.

8.3.2 Future Research

The goal of this research was to create or simulate conditions that would lead to the formation of gypsum in an attempt to better understand the processes that lead to expansion of pyritic shales of the Marcellus Formation. This research was successful at highlighting the importance of higher concentrations of pyrite and the microenvironments in which these complex geochemical processes take place. The laboratory tests conducted for this research where conducted on a bulk basis with proven geochemical

testing and microanalytical techniques. The next step in research should be to observe the conditions that lead to the oxidation of the various forms of pyrite and the resulting production of gypsum from a much more focused perspective. An attempt should be made to show how acidic microenvironments can be established within pyritic shales and how accelerated oxidation can become established in the presence of bacteria and ferric iron. Once these conditions are established, the effective migration of the sulfuric acid into the surrounding microfractures can be studied. This testing would further explain

how expansion through the production of sulfates can occur in a relatively short period of time.

140

APPENDIX A

VISUAL IDENTIFICATION OF GEOMATERIALS

141 APPENDIX A – VISUAL IDENTIFICATION OF GEOMATERIALS

Figure A.1 Reference Chart created by Bryant (2003) using Macromedia Freehand 10® Munsell® Book of Color Library. Color guidelines from Sobek et al. (1978), USDA-NRCS (1998), and Hosterman and Whitlow (1980, 1983).

142

APPENDIX B

CHEMICAL TEST PROCEDURES

143 APPENDIX B – CHEMICAL TEST PROCEDURES

ACID SULFATE SOILS LABORATORY METHODS GUIDELINES 2004

Published by Department of Natural Resources, Mines and Energy, Indooroopilly, Queensland, Australia, May 2004

CHROMIUM REDUCIBLE SULFUR (SCR)—METHOD CODE 22B

Procedure: 1. Weigh 0.5 g of finely ground (eg. ring mill ground, see Section B1.3) sample (or other appropriate weight as described in the introduction) into a double-neck round- bottom digestion flask. Record sample weight (m) to the nearest 0.001 g. 2. Add 2.0 g of chromium powder and then 10 mL ethanol (95% concentration) to the digestion flask and swirl to wet the sample.

Caution: Chromium dust may be toxic if inhaled and may represent a combustion risk. Avoid the use of very fine chromium powder.

3. Place the digestion flask in the heating mantle and connect to the condenser. The digestion apparatus should be set up in a fume hood. 4. Attach the pressure equalising funnel making sure the gas flow arm is facing the condensers and that the solution tap is shut. Attach Pasteur pipette to the outlet tube at the top of the condenser and insert it into a 100 mL Erlenmeyer flask containing 40 mL zinc acetate solution. 5. Turn on the water flow around the condenser and make sure that all ground glass fittings are tight. Add 60 mL of 6 M HCl to the glass dispenser in the pressure equalising funnel. 6. Connect the N2 flow to the pressure equalising funnel and adjust the flow to obtain a bubble rate in the zinc acetate solution of about 3 bubbles per second. Allow the N2 gas to purge the system for about 3 min. 7. Slowly release the 6 M HCl from the dispenser.

Caution: The 6 M HCl should be added to the sediment and chromium powder very slowly in a fume hood.

8. Wait for 2 min before turning on the heating mantle and adjust the heat so that a gentle boil is achieved. Check for efficient reflux in the condenser. Allow to digest for 20 min.

Caution: H2S gas (a hazardous gas) can be evolved during this digest. Consequently, this part of the procedure should be undertaken in a fume hood.

9. Remove the Erlenmeyer flask and wash any ZnS on the Pasteur pipette into the Erlenmeyer flask with a wash bottle containing deionised water. Add 20 mL of 6 M HCl and 1 mL of the starch indicator solution to the zinc acetate solution and gently mix by swirling or by placing on a magnetic stirrer.

144 APPENDIX B – CHEMICAL TEST PROCEDURES

Note: If a large amount of ZnS has formed on the tip of the Pasteur pipette (and is not easily removed by washing with deionised water, the pipette can be left in the Erlenmeyer flask (and trapping solution), washed with a small amount of 6 M HCl and remain there during the titration.

10. Whilst stirring, titrate the zinc acetate trapping solution with the iodine solution to a permanent blue end-point. Record the volume of titrant (A) in mL. Perform the same titration on the blank sample and record the volume of titrant (B) in mL.

Warning: H2S gas (a hazardous gas) can be evolved after the acid is added to the zinc acetate trapping solution. Consequently, this part of the procedure should be: 1) carried out with a minimum of delay after the acid has been added, and 2) undertaken in a fume hood or with the aid of a fume extractor. It is recommended that laboratories be equipped suitable gas monitors to guard against accidental exposure to H2S.

Caution: The acidic chromium digest solution (in the round-bottomed flask) generated by this procedure must not be disposed of down the sink. Consult local or state regulatory authorities for its safe disposal.

Calculations

The concentration of chromium reducible sulfur (SCR) in %S is calculated as follows:

SCR (%) = (A – B) x C x 3.2066 m

Where: A = The volume of iodine (in mL) used to titrate the zinc acetate trapping solution following the soil digestion B = The volume of iodine (in mL) used to titrate the zinc acetate trapping solution following a blank digestion C = The molarity of the iodine solution (in M) as determined by titration of this

solution with the standard 0.025 M Na2S2O3 solution (see below)

C = 0.025 x D 2 x E

D = Titration volume of standard Na2S2O3 solution (in mL) E = Volume of iodine solution titrated (in mL) m = The mass of the soil weighed (in g)

145 APPENDIX B – CHEMICAL TEST PROCEDURES

OVERBURDEN SAMPLING AND TESTING MANUAL

Small Operator Assistance Program Bureau of Mining and Reclamation Pennsylvania Department of Environmental Resources April 29, 1988

ASTM/EPA COMBINATION METHOD

Procedure

1. Analyse the sample for total sulfur using one of the procedures that are explained in this manual. 2. Analyse the sample for sulfate sulfur.

a. Weight approximately 2 grams of sample and record weight to nearest 0.1 mg. b. Transfer the sample to a 150 ml beaker, wet with 25% methanol and add 50 ml of 40% HCl. c. Boil gently for 30 minutes and remove from the hotplate. d. Filter through #40 Whatman filter paper into a 400 ml beaker. Rinse the beaker twice and the residue six times with cold water. Discard the filter and residue. e. Add approximately 5 ml of bromine water to the filtrate and boil for 5 minutes. This step insures that all iron present in the filtrate will be oxidized to the ferric form for subsequent removal as ferric hydroxide. f. Stir in 25 ml of 1:1 NH4OH to precipitate any iron and manganese in the filtrate. g. Filter into a 400 ml beaker using #4 Whatman filter paper. Rinse the beaker twice and the residue six times with hot water. Discard the filter paper and residue.

NOTE: If a large amount of manganese was dissolved with the iron as evidenced by a fine textured brown precipitate instead of an orange colloidal precipitate, #42 Whatman filter paper will have to be used to effectively remove the metals. Some samples have such a high concentration of iron and manganese that steps (e) through (g) will have to be repeated in order that a clear filtrate is obtained.

h. Adjust the volume of filtrate to approximately 250 ml and make the solution slightly acidic with 40% HCl using the methyl orange solution as the indicator, then add an additional 1 ml of 40% HCl. i. Place the beakers on the hotplate and heat to boiling. j. Slowly add 10 ml of barium chloride solution and boil for 15 minutes. Let the samples remain standing for several hours, preferably overnight. This will facilitate quantitative filtration of the BaSO4 because of the formation of larger crystals.

146 APPENDIX B – CHEMICAL TEST PROCEDURES

k. Filter the precipitate using #42 Whatman filter paper. Rinse the beaker twice with water after triturating with a policeman. Rinse the filter paper and precipitate an additional 8 times. l. Transfer the filter paper, lightly folded, to a 15 ml crucible and place in a cool muffle furnace. m. Smoke off the filter paper by heating to 500 degrees Celsius. After ½ hour raise the temperature to 800 degrees Celsius and maintain for ½ hour. n. Weigh the residue after the crucible has cooled. 3. Analyse the sample for organic sulfur.

a. Weigh approximately 0.5 gram of sample, to the nearest 0.1 mg, into a clean 100 ml beaker. b. Add 50 ml of 12.5% nitric acid to the beaker and stir thoroughly. Cover the beaker with a watch glass and either boil gently for ½ hour (fume hood required) or let the sample stand overnight at room temperature. c. Filter the sample, washing the beaker twice and the residue and additional six times.

i. Use #40 Whatman if you use the Eschka Method for analysing the residue. Use glass fiber filters if you use any of the high temperature combustion methods to analyse the residue. ii. If the iodimetric or the acid base titration techniques are to be used then the filtrate will have to be tested for the presence of nitrates with the Nessler’s reagent. If nitrates are detected, wash the residue until the filtrate does not turn yellow upon the addition of the Nessler’s reagent. Allow one minute for the reaction. d. Dry the sample and determine the sulfur content using one of the total sulfur techniques. The bomb washing method would be cumbersome due to the mass of the filter paper.

Calculations

%Sulfate Sulfur = (A – B) * 13.738/C

A = weight of BaSO4 B = weight of blank residue C = weight of sample

%Organic Sulfur = Value from 3d.

%Pyritic Sulfur = %Total Sulfur – (%Sulfate Sulfur + %Organic Sulfur)

147 APPENDIX B – CHEMICAL TEST PROCEDURES

LECO SC132 SULFUR ANALYSER

LECO Corporation, 1980

TOTAL SULFUR ANALYSIS

Procedure

1. The sample is weighed into a ceramic crucible and iron chips and tin pellets are added. 2. The sample is then combusted in an induction furnace. 3. The SO2 evolved is then bubbled into a moderately dark blue colored starch solution containing HCl, KI, and a small amount of KIO3 in the following equilibrium:

KIO3 + 5KI + 6HCl = 6KCl +3I2 + 3H2O

When the SO2 is introduced, the reaction

SO2 + I2 + 2H2O = H2SO4 + 2HI

goes to the right and the starch loses its blue color due to the removal of I2.

4. To restore the blue endpoint, a known concentration KIO3 solution is added (titrated) to the solution. 5. The amount of KIO3 required to balance the reaction is proportional to the amount of SO2 evolved.

Calculations

FR = Final Reading IR = Initial Reading

-5 -3 Total Sulfur = [KIO3 (g/l) * (FR – IR) * 5x10 * 100]/[(2.247x10 * Weight (mg))]

148 APPENDIX B – CHEMICAL TEST PROCEDURES

ADTI-WP2 LEACHING COLUMN METHOD FOR OVERBURDEN ANALYSIS AND PREDICTION OF WEATHERING RATES

Procedure Re-typed from Roger J. Hornberger, Keith B.C. Brady, Barry E. Scheetz, William B. White and Stephen C. Parsons (Unpublished)

Procedure

1. Construct the leaching columns with 2-inch diameter clear polycarbonate plastic tubing in general accordance with the following figure:

2. The rock samples are crushed to a nominal maximum size of 3/8” diameter using a jaw crusher, and then mixed and homogenized using a riffle splitter and procedures described in ASTM C-702-98 and Noll, et al., (1988). The particle size distribution of the crushed sample is determined using a series of sieves (#4 (4.76 mm), #10 (2.00 mm), #16 (1.19 mm), #35 (0.50 mm), and #60 (0.250 mm) sieve sizes) to yield 7 particle size classes (i.e. including >3/8” (9.52 mm) and <60 (0.250 mm) fines). 3. The homogenized sample is then chemically analyzed for percent total sulfur and neutralization potential. 4. The reconstruction of the particle size distribution following the crushing of each rock sample is included in the method because variations in particle size distribution of the same lithologic unit can occur due to differences in crushing equipment, and the particle size distribution may vary significantly by rock type. The following table describes the intended particle size distribution:

149 APPENDIX B – CHEMICAL TEST PROCEDURES

5. Attempt to achieve a target 10% CO2 atmosphere in the weathering apparatus with one tank of CO2 with a regulator, and mixed it with filtered house air (i.e. the compressed air piped throughout the lab) in the reagent water reservoir, prior to entry in the leaching columns and humidity cells. Take precautions to trap any drops of oil from the air compressor in the air lines prior to gas mixing, because any oil residue coating the rock samples would cause serious interferences in the weathering test. 6. The ADTI-WP2 simulated weathering procedure consists of alternating cycles of saturation (i.e., water inundation) and unsaturation (drained) with circulated humidified, CO2-enriched air. The humidified gas mixture is introduced continuously through the gas inlet port of each leaching column and humidity cell during the periods of time between leaching episodes. These inter-leach periods of time are called “humidified air cycles” or “drying cycles”. The leaching episodes are called “wetting cycles” or “saturation cycles.” The gas mixture is also introduced into the leaching columns and humidity cells during periods of saturation (i.e. when the apparatus is filled with water). 7. Once the leaching column has been filled with the rock sample and sealed, reagent water (distilled, deionized) is introduced through the water inlet port at the bottom of the column until the column is full and all visible pore spaces are saturated. The first leaching episode is called the “initial flush”, in which the reagent water is drained from the column after a 1-hour contact time. During this initial flush the column is filled and drained again until the conductivity of the flush water stabilizes. This initial flush is intended to wash the rock samples of any oxidized materials that have accumulated during handling and storage. 8. The initial flush is followed by a one-week humidified air cycle. 9. Following this first and each successive humidified air cycle, reagent water is introduced through the water inlet port to just above the rock sample surface, and the saturation cycle begins. During this saturation cycle, the rock sample is in contact with the reagent water in the column for a 24-hour period. 10. Following this 24-hour saturated condition, the column is drained and the leachate (effluent water) is tested for analytes of concern (e.g. acidity, alkalinity, Fe, Mn, Al, sulfate). Then the next one-week humidified air cycle commences, followed by the next 24-hour saturation cycle, and this weekly pattern of alternating humidified air cycles and saturation cycles continues until method implementation is complete (e.g. 15 weeks).

150 APPENDIX B – CHEMICAL TEST PROCEDURES

11. The surface area of each fraction of sieved starting material was determined by BET (Brunauer et al., 1938; Yates, 1992) instrumentation using N2 gas bulk adsorption. This method measures the accessible surface of the rock to N2 gas molecules. 12. Surface areas were measured on the composite sieve fractions and at the completion of the testing, the rock was again sieved and remeasured. The bulk surface areas of rock were determined by taking the individual masses of the sieve fractions, multiplying each mass by the surface area (SA), and combining their fractional percent of the total as a weighted linear average.

151

APPENDIX C

OXIDATION EXPERIMENTS DATA

152 APPENDIX C - OXIDATION EXPERIMENTS DATA

Table C.1 Total Sulfur Analysis on shale samples from 10% H2O2 Oxidation Experiments.

Sample Weight Initial Final KIO3 Total Number (mg) Reading Reading Difference Conc. (g/L) Sulfur (%) 1A1 93.1 7 18 11 0.4 0.11 1A2 85.0 9 131 122 0.4 1.28 2A1 91.1 8 91 83 0.4 0.81 2A2 84.8 7 89 82 0.4 0.86 3A1 91.3 14 90 76 0.4 0.74 3A2 83.9 8 80 72 0.4 0.76 4A1 90.7 9 105 96 0.4 0.94 4A2 82.3 8 97 89 0.4 0.96 5A1 82.5 8 80 72 0.4 0.78 5A2 83.4 8 78 70 0.4 0.75 6A1 85.6 8 84 76 0.4 0.79 6A2 84.5 10 83 73 0.4 0.77 CS1 57.7 8 139 131 0.4 2.04 CS2 57.8 10 139 129 0.4 2.01 CS3 61.4 11 143 132 0.4 1.93

153 APPENDIX C - OXIDATION EXPERIMENTS DATA

Table C.2 Experimental Data from 30% H2O2 oxidation experiments.

Al Al Fe Fe Ca Ca K K SO42- Sample (ppm) (mol/L) (ppm) (mol/L) (ppm) (mol/L) (ppm) (mol/L) (mol/L) pH Filtered T1B #1 0.53 0.0020 <.02 0.4 0.0010 0.7 0.0018 0.0009425 4.836 Filtered T1B #2 0.41 0.0015 <.02 0.59 0.0015 0.7 0.0018 0.0007313 5.971 Filtered T1B #3 0.49 0.0018 <.02 1.6 0.0040 0.6 0.0015 0.0012089 6.854 Filtered T2B #1 0.54 0.0020 <.02 0.37 0.0009 0.7 0.0018 0.0008995 4.811 Filtered T2B #2 0.39 0.0014 0.02 0.0000 0.94 0.0023 0.7 0.0018 0.0039652 3.3 Filtered T2B #3 0.50 0.0019 0.29 0.0005 2.0 0.0050 0.7 0.0018 0.0149525 2.8 Filtered T2B #4 0.48 0.0018 0.41 0.0007 6.8 0.0170 0.6 0.0015 0.0280459 2.9 Filtered T2B #5 0.43 0.0016 0.05 0.0001 8.4 0.0210 0.6 0.0015 0.0168417 4.13 Filtered T2A #1 0.51 0.0019 <.02 4.8 0.0120 0.8 0.0020 0.0016987 6.389 Filtered T2A #2 <.05 <.02 48 0.1198 0.9 0.0023 0.0135212 7.78 Unfiltered T1B #1 0.47 0.0017 <.02 0.32 0.0008 0.8 0.0020 Unfiltered T1B #2 0.42 0.0016 <.02 0.54 0.0013 0.5 0.0013 Unfiltered T1B #3 0.50 0.0019 0.02 0.0000 1.6 0.0040 0.7 0.0018 Unfiltered T2B #1 0.52 0.0019 0.06 0.0001 0.33 0.0008 0.5 0.0013 Unfiltered T2B #2 0.54 0.0020 0.31 0.0006 0.9 0.0022 0.6 0.0015 Unfiltered T2B #3 0.56 0.0021 1.1 0.0020 2.0 0.0050 0.4 0.0010 Unfiltered T2B #4 0.45 0.0017 1.1 0.0020 6.4 0.0160 0.7 0.0018 Unfiltered T2B #5 0.49 0.0018 1.9 0.0034 7.8 0.0195 0.7 0.0018 Unfiltered T2A #1 0.50 0.0019 0.32 0.0006 4.7 0.0117 0.6 0.0015 Unfiltered T2A #2 0.83 0.0031 2.7 0.0048 67 0.1672 1.1 0.0028

154 APPENDIX C - OXIDATION EXPERIMENTS DATA

Table C.3 Experimental Data for 10% H2O oxidation experiments. All units are mol/L unless otherwise noted. TB-3, Depth = 25.0’, Black Shale with Calcite or Siderite in Vertical Fractures (Rough) A: < No. 4 Sieve = 230.3 grams B: Single Piece Core = 224.3 grams Sample Time Temp Sample Time Temp No. (min) pH (F) SO42- K Ca Al No. (min) pH (F) SO42- K Ca 1A 5 7.09 77 0.0014 0.0005 0.0025 0.0037 1B 5 5.95 72.8 -0.0027 0.0005 0.0002 10 7.06 77 0.0032 0.0005 0.0027 0.0041 10 6.42 73.2 -0.0005 0.0005 0.0003 20 7.07 85.6 0.0048 0.0005 0.0025 0.0037 20 6.61 73.2 -0.0006 0.0003 40 7.13 90.3 0.0074 0.0003 0.0030 0.0044 40 6.82 74.1 -0.0043 0.0005 80 7.17 93.6 0.0168 0.0005 0.0032 0.0048 80 7.04 75.4 -0.0046 0.0008 0.0010 160 7.3 94.2 0.0315 0.0008 0.0045 0.0067 160 7.19 79.7 -0.0049 0.0005 0.0012 270 7.4 89.7 0.0437 0.0060 0.0089 320 7.27 81.6 -0.0040 0.0003 0.0015 4545 7.6 78.8 0.1117 0.0005 0.0162 0.0241 4410 7.79 75.6 0.0127 0.0005 0.0030

TB-3, Depth = 23.5’, Black Shale with Trace Gypsum Crystals. Nodular Pyrite on Single Piece of Vertically Fractured Core A: < No. 4 Sieve with 3.63 gram Nodule = 247.2 grams B : 2 Piece Core with 2.4 gram Nodule = 245.6 grams Sample Time Temp Sample Time Temp No. (min) pH (F) SO42- K Ca Al No. (min) pH (F) SO42- K Ca 2A 5 6.98 77 0.0040 0.0008 0.0017 0.0026 2B 5 6.38 73 0.0035 0.0003 0.0002 10 6.98 77 0.0056 0.0003 0.0022 0.0033 10 6.37 74.3 -0.0050 0.0005 0.0002 20 7.01 91.8 0.0092 0.0005 0.0025 0.0037 20 6.46 74.3 0.0092 0.0003 0.0004 40 7.07 97.3 0.0166 0.0008 0.0032 0.0048 40 6.48 76.8 0.0126 0.0003 0.0009 80 7.19 103.9 0.0278 0.0003 0.0047 0.0070 80 6.52 81.1 0.0203 0.0017 160 7.39 95.4 0.0406 0.0005 0.0067 0.0100 160 5.5 92.6 0.0343 0.0003 0.0032 270 7.48 89.2 0.0491 0.0003 0.0075 0.0111 320 4.45 87.3 0.0619 0.0055 4547 7.65 78.8 0.1162 0.0005 0.0165 0.0244 4412 7.57 74.5 0.0708 0.0082

TB-4, Depth = 19.0’, Black Shale with Large Pyritic or Siderite Nodule (Nodule Weight = 108.2 grams) A: < No. 4 Sieve with 50.0 gram Nodule = 197.3 grams B : 3 Piece Core with 57.0 gram Nodule = 233.6 grams Sample Time Temp Sample Time Temp No. (min) pH (F) SO42- K Ca Al No. (min) pH (F) SO42- K Ca 3A 5 7.14 77 0.0037 0.0030 0.0044 3B 5 5.79 72.8 -0.0019 0.0005 0.0000 10 7.14 77 0.0023 0.0005 0.0025 0.0037 10 6.2 73.4 -0.0026 0.0001 20 7.17 85.4 0.0074 0.0003 0.0025 0.0037 20 6.64 73.4 -0.0023 0.0003 0.0004 40 7.14 88.1 0.0044 0.0025 0.0037 40 6.84 74.3 -0.0032 0.0003 0.0007 80 7.18 90.1 0.0090 0.0010 0.0030 0.0044 80 7.04 75.9 -0.0028 0.0003 0.0011 160 7.21 93.4 0.0266 0.0003 0.0042 0.0063 160 7.17 81.3 0.0021 0.0005 0.0020 270 7.35 91.7 0.0284 0.0008 0.0042 0.0063 320 7.14 84.3 0.0090 0.0003 0.0032 4549 7.79 78.8 0.1002 0.0003 0.0145 0.0215 4414 7.86 74.7 0.0383 0.0055

TB-2, Depth = 17.5’, Black Shale with Calcite or Siderite in Bedding Plane. Large Siderite or Pyrite Nodule. A: < No. 4 Sieve with 3.5 gram Nodule = 340.2 grams B: 3 Piece Core with 4.1 grams of 2 Nodule Pieces = 325.0 g Sample Time Temp Sample Time Temp No. (min) pH (F) SO42- K Ca Al No. (min) pH (F) SO42- K Ca 4A 5 7.12 77 0.0017 0.0005 0.0020 0.0030 4B 5 6.41 72.8 -0.0047 0.0003 0.0000 10 7.15 77 0.0027 0.0005 0.0017 0.0026 10 6.61 73.5 -0.0026 0.0003 0.0002 20 7.13 90 0.0028 0.0010 0.0017 0.0026 20 6.68 73.5 -0.0030 0.0005 0.0003 40 7.14 92.7 0.0038 0.0003 0.0017 0.0026 40 6.86 74.3 -0.0046 0.0003 0.0006 80 7.18 98.8 0.0094 0.0022 0.0033 80 7.07 75.9 -0.0065 0.0003 0.0008 160 7.33 107.3 0.0351 0.0005 0.0060 0.0089 160 7.21 80.3 -0.0056 0.0003 0.0013 270 7.48 96.8 0.0608 0.0005 0.0077 0.0115 320 7.27 82.5 -0.0042 0.0005 0.0016 4551 7.93 78.8 0.1163 0.0005 0.0185 0.0274 4416 8 75 0.0209 0.0008 0.0035 TB-2, Depth = 24.0’, Black Shale with 2 Pyritic or Siderite Nodules A: < No. 4 Sieve with 1.3 gram Nodule = 130.6 grams B: 3 Piece Core with 2 Small Nodules Intact = 172.1 grams Sample Time Temp Sample Time Temp No. (min) pH (F) SO42- K Ca Al No. (min) pH (F) SO42- K Ca 5A 5 7.15 77 0.0000 0.0000 0.0012 0.0019 5B 5 6.42 73 -0.0065 0.0001 10 7.19 77 0.0037 0.0003 0.0020 0.0030 10 6.49 73 -0.0045 0.0000 20 7.2 87.9 0.0026 0.0020 0.0030 20 6.82 74.2 -0.0054 0.0006 40 7.21 89.4 0.0050 0.0025 0.0037 40 7 75.2 -0.0048 0.0005 0.0006 80 7.32 90.3 0.0122 0.0003 0.0027 0.0041 80 7.02 77.3 -0.0031 0.0011 160 7.39 89.6 -0.0006 0.0008 0.0037 0.0056 160 7.13 82.2 0.0016 0.0020 270 7.49 86.5 0.0245 0.0042 0.0063 320 7.15 84.2 0.0117 0.0005 0.0027 4553 n/a n/a 0.0005 0.0015 0.0022 4418 8.01 74.5 0.0332 0.0010 0.0052

TB-2 (CONTROL), Depth = 23.0’, Black Shale A: < No. 4 Sieve = 200.0 grams B: < No. 4 Sieve with 4.0 grams of Pyrite = 200.0 grams Sample Time Temp Sample Time Temp No. (min) pH (F) SO42- K Ca Al No. (min) pH (F) SO42- K Ca 6A 5 7.11 77 0.0065 0.0025 0.0037 6B 5 4.81 74 0.0035 0.0006 10 7.12 77 0.0067 0.0025 0.0037 10 4.01 75 0.0352 0.0004 20 7.13 89.2 0.0063 0.0022 0.0033 20 5.56 76 0.0268 0.0009 40 7.16 92.9 0.0089 0.0003 0.0030 0.0044 40 3.21 84 0.0678 0.0017 80 7.24 98.7 0.0188 0.0005 0.0037 0.0056 80 2.97 117.5 0.1461 0.0003 0.0040 160 7.4 97.2 0.0448 0.0010 0.0057 0.0085 160 4.7 92.8 0.0829 0.0003 0.0072 270 7.47 90.3 0.0582 0.0075 0.0111 320 5.15 82.2 0.0673 0.0080 4555 7.74 78.8 0.1249 0.0003 0.0170 0.0252 4420 7.65 74.5 0.0657 0.0085

155

APPENDIX D

VISUAL MINTEQ EVALUATIONS FOR 30% H2O2 EXPERIMENTS

156 APPENDIX D - VISUAL MINTEQ EVALUATIONS FOR 30% H2O2 EXPERIMENTS

Table D.1 Visual MINTEQ Saturation Index (SI) evaluations for 30% H2O2 oxidation experiments.

Time (min) 4 30 137 4 21 33 49 54 7 111 Mineral Filt T1B #1 Filt T1B #2 Filt T1B #3 Filt T2B #1 Filt T2B #2 Filt T2B #3 Filt T2B #4 Filt T2B #5 Filt T2A #1 Filt T2A #2

Al(OH)3 (am) 0.131 1.634 1.634 0.09 -4.904 -6.885 -6.825 -2.89 2.019 0.244

Al(OH)3 (Soil) 2.641 4.144 4.144 2.6 -2.394 -4.375 -4.315 -0.38 4.529 2.754

Al2O3 2.21 5.217 5.217 2.128 -7.86 -11.822 -11.702 -3.831 5.985 2.438

Al4(OH)10SO4 7.758 11.738 11.738 7.607 -8.563 -14.94 -14.734 -1.712 12.514 3.028

AlOHSO4 0.895 0.365 0.365 0.866 -0.321 -0.754 -0.728 0.488 -0.012 -4.173 Alunite 9.707 11.29 11.29 9.583 0.695 -2.686 -2.57 5.042 11.365 2.645 Anhydrite -2.419 -2.008 -2.008 -2.49 -1.389 -0.727 -0.14 -0.205 -1.144 -0.014 Aragonite -5.782 -3.337 -3.337 -5.865 -8.57 -9.456 -8.836 -6.181 -1.712 1.805 Boehmite 2.353 3.857 3.857 2.312 -2.682 -4.663 -4.603 -0.668 4.241 2.467

CaCO3xH2O -6.973 -4.529 -4.529 -7.056 -9.762 -10.648 -10.028 -7.373 -2.904 0.612 Calcite -5.638 -3.193 -3.193 -5.721 -8.426 -9.312 -8.692 -6.037 -1.568 1.948 Diaspore 4.058 5.562 5.562 4.017 -0.977 -2.958 -2.898 1.037 5.946 4.172 Ettringite -35.865 -24.29 -24.29 -36.409 -51.211 -55.848 -52.112 -36.468 -16.056 -5.694

Fe2(SO4)3 -24.553 -28.293 -28.293 -24.573 -17.233 -13.643 -13.245 -17.285 -29.8 -34.351 Ferrihydrite 2.556 3.736 3.736 0.487 -0.04 0.21 2.269 4.124 5.427 Ferrihydrite (aged) 3.066 4.246 4.246 3.038 0.997 0.47 0.72 2.779 4.634 5.937 Gibbsite (C) 3.191 4.694 4.694 3.15 -1.844 -3.825 -3.765 0.17 5.079 3.304 Goethite 5.265 6.445 6.445 5.238 3.196 2.669 2.919 4.978 6.833 8.137 Gypsum -2.169 -1.758 -1.758 -2.24 -1.139 -0.477 0.109 0.045 -0.894 0.234 Hematite 12.93 15.29 15.29 12.875 8.793 7.738 8.238 12.357 16.066 18.675 H-Jarosite -3.875 -4.401 -4.401 -2.443 -0.929 -0.247 0.492 -4.76 -5.624 K-Alum -8.386 -9.81 -9.81 -8.428 -7.328 -6.748 -6.753 -7.012 -10.504 -15.685 K-Jarosite 3.781 4.394 4.394 3.697 3.668 4.648 5.336 7.319 4.481 4.994 Lepidocrocite 4.385 5.565 5.565 4.358 2.316 1.789 2.039 4.098 5.953 7.257 Lime -26.212 -23.767 -23.767 -26.295 -29.001 -29.887 -29.266 -26.611 -22.142 -18.626 Maghemite 5.126 7.486 7.486 5.071 0.989 -0.066 0.434 4.553 8.262 10.871 Portlandite -16.217 -13.772 -13.772 -16.3 -19.005 -19.892 -19.271 -16.616 -12.147 -8.632 Vaterite -6.204 -3.76 -3.76 -6.287 -8.993 -9.879 -9.258 -6.603 -2.134 1.382 pH 4.836 5.971 6.854 4.811 3.3 2.8 2.9 4.13 6.389 7.78

157

APPENDIX E

X-RAY DIFFRACTION OF NODULE

158 APPENDIX E – X-RAY DIFFRACTION OF NODULE

Figure E.1 X-Ray Diffraction results from nodule in shale in oxidation experiments. Second line defines actual X-Ray Diffraction patter of sample.

159

APPENDIX F

PHREEQC INPUT FILES FOR IRREVERSIBLE REACTIONS

160 APPENDIX F – PHREEQC INPUT FILES FOR IRREVERSIBLE REACTIONS

PHREEQC INPUT FILE: 0.1% S2 AND 5% CaCO2

Oxidation of Pyrite by Oxygen and Precipitation of Gypsum 0.1% Sulfide Sulfur 5% Calcium Carbonate in Pyritic Shales 0.32mx0.32mx0.32m Cube with 20 Fractures Assume 2 mm Penetration into Shale For Availability SOLUTION 1 PURE WATER ph 7.0 temp 25.0 EQUILIBRIUM_PHASES 1 Pyrite 0.0 0.740 # 88.78 grams of pyrite Goethite 0.0 Calcite 0.0 12.68 # 1,269.14 grams of calcium carbonate Illite 0.0 65.19 # Parent Rock (Inert) CO2(g) -2.5 Gypsum 0.0 0.0 REACTION 1 O2 1.0 NaCl 0.5 0.0 0.01 0.05 0.1 0.5 1 5 6 8 10 15 20 21 22 SELECTED_OUTPUT -file Pyrite.out -total Cl -si Gypsum -equilibrium_phases pyrite goethite melanterite calcite CO2(g) gypsum USER_GRAPH -headings "O2_Added" "Gypsum_(moles)" "Calcite_(moles)" "Pyrite_(moles)" "pH" -chart_title 0.1%_Pyritic_Sulfur_5%_Calcite -initial_solutions false -axis_titles "O2 (moles)" "Assemblage (moles)" "pH" -axis_scale x_axis 0 25 5 -axis_scale y_axis 0 15 1 -axis_scale sy_axis 0 14 1 -start 10 graph_x tot("Cl") * 2 20 graph_y equi("gypsum") equi("calcite") equi("pyrite") 30 graph_sy -la("H+") -end END

161 APPENDIX F – PHREEQC INPUT FILES FOR IRREVERSIBLE REACTIONS

PHREEQC INPUT FILE: 0.5% S2 AND 5% CaCO2

Oxidation of Pyrite by Oxygen and Precipitation of Gypsum 0.5% Sulfide Sulfur 5% Calcium Carbonate in Pyritic Shales 0.32mx0.32mx0.32m Cube with 20 Fractures Assume 2 mm Penetration into Shale For Availability SOLUTION 1 PURE WATER ph 7.0 temp 25.0 EQUILIBRIUM_PHASES 1 Pyrite 0.0 3.70 # 443.89 grams of pyrite Goethite 0.0 Calcite 0.0 12.68 # 1,269.14 grams of calcium carbonate Illite 0.0 65.19 # Parent Rock (Inert) CO2(g) -2.5 Gypsum 0.0 0.0 REACTION 1 O2 1.0 NaCl 0.5 0.0 0.01 0.05 0.1 0.5 1 5 6 8 10 15 20 21 22 SELECTED_OUTPUT -file Pyrite.out -total Cl -si Gypsum -equilibrium_phases pyrite goethite melanterite calcite CO2(g) gypsum USER_GRAPH -headings "O2_Added" "Gypsum_(moles)" "Calcite_(moles)" "Pyrite_(moles)" "pH" -chart_title 0.5%_Pyritic_Sulfur_5%_Calcite -initial_solutions false -axis_titles "O2 (moles)" "Assemblage (moles)" "pH" -axis_scale x_axis 0 25 5 -axis_scale y_axis 0 15 1 -axis_scale sy_axis 0 14 1 -start 10 graph_x tot("Cl") * 2 20 graph_y equi("gypsum") equi("calcite") equi("pyrite") 30 graph_sy -la("H+") -end END

162 APPENDIX F – PHREEQC INPUT FILES FOR IRREVERSIBLE REACTIONS

PHREEQC INPUT FILE: CONCENTRATED FeS2 AND 5% CaCO2

Oxidation of Pyrite by Oxygen and Precipitation of Gypsum Concentrated Pyrite Replacement Fossils(1.5cm dia. & 0.3cm height)at 4.35 per Fracture. Total of 20 Fractures in 0.34mx0.34mx0.34m cube. Reaction Available Within 2mm on Each Face of the Fracture. 5% Calcium Carbonate by Weight. SOLUTION 1 PURE WATER ph 7.0 temp 25.0 EQUILIBRIUM_PHASES 1 Pyrite 0.0 6.42 # 770.20 grams of pyrite Goethite 0.0 Calcite 0.0 12.68 # 1,269.14 grams of calcium carbonate Illite 0.0 65.19 CO2(g) -2.5 Gypsum 0.0 0.0 REACTION 1 O2 1.0 NaCl 0.5 0.0 0.01 0.05 0.1 0.5 1 5 10 15 20 22 23 24 24.5 SELECTED_OUTPUT -file Pyrite.out -total Cl -si Gypsum -equilibrium_phases pyrite goethite melanterite calcite CO2(g) gypsum USER_GRAPH -headings "O2_Added" "Gypsum_(moles)" "Calcite_(moles)" "Pyrite_(moles)" "pH" -chart_title Concentrated_Pyrite_5%_Calcite -initial_solutions false -axis_titles "O2 (moles)" "Assemblage (moles)" "pH" -axis_scale x_axis 0 55 5 -axis_scale y_axis 0 15 1 -axis_scale sy_axis 0 14 1 -start 10 graph_x tot("Cl") * 2 20 graph_y equi("gypsum") equi("calcite") equi("pyrite") 30 graph_sy -la("H+") -end END

163

APPENDIX G

PHREEQC GRAPHICAL OUTPUT FOR IRREVERSIBLE REACTIONS

164

APPENDIX G – PHREEQC GRAPHICAL OUTPUT FOR IRREVERSIBLE REACTIONS

Figure G.1 PHREEQC graphical output of Assemblage (moles) versus O2 (moles) for 0.1% S2 and 5% CaCO3.

165

APPENDIX G – PHREEQC GRAPHICAL OUTPUT FOR IRREVERSIBLE REACTIONS

Figure G.2 PHREEQC graphical output of Assemblage (moles) versus O2 (moles) for 0.5% S2 and 5% CaCO3.

166

APPENDIX G – PHREEQC GRAPHICAL OUTPUT FOR IRREVERSIBLE REACTIONS

Figure G.3 PHREEQC graphical output of Assemblage (moles) versus O2 (moles) for Concentrated FeS2 and 5% CaCO3.

167

APPENDIX H

VOLUME AND CELL HEIGHT CHANGE SPREADSHEETS FOR IRREVERSIBLE REACTIONS

168 APPENDIX H VOLUME AND CELL HEIGHT CHANGE SPREADSHEETS FOR IRREVERSIBLE REACTIONS

Table H.1 Volume change calculations for 0.1% S2 and 5% CaCO3.

Volume Volume

O2 Gypsum Calcite Gypsum Calcite Total Volume (moles) (moles) (moles) (cm3) (cm3) Change (cm3) pH 0.00 0.00 12.68 0.0 467.9 0.00 7.62 0.01 0.00 12.67 0.0 467.7 -0.18 7.37 0.05 0.01 12.65 0.7 466.9 -0.28 7.22 0.10 0.03 12.63 2.6 465.9 0.60 7.21 0.50 0.24 12.41 17.9 458.0 8.12 7.18 1.02 0.50 12.15 37.5 448.2 17.80 7.16 5.31 1.44 11.20 107.4 413.2 52.77 7.02 6.37 1.44 11.20 107.4 413.2 52.74 6.99 8.50 1.44 11.20 107.4 413.3 52.80 6.94 10.62 1.44 11.20 107.5 413.3 52.88 6.89 15.93 1.44 11.20 107.6 413.3 53.05 6.77 21.24 1.44 11.20 107.4 413.3 52.85 6.65 22.30 1.43 11.20 107.3 413.3 52.70 6.62 23.36 1.43 11.20 107.1 413.3 52.49 6.60

169 APPENDIX H VOLUME AND CELL HEIGHT CHANGE SPREADSHEETS FOR IRREVERSIBLE REACTIONS

Table H.2 Volume change calculations for 0.5% S2 and 5% CaCO3.

O2 Gypsum( Calcite Gypsum Calcite Total Volume (moles) moles) (moles) (cm3) (cm3) Change (cm3) pH 0.00 0.00 12.68 0.0 467.9 0.00 7.62 0.01 0.00 12.67 0.0 467.7 -0.18 7.37 0.05 0.01 12.65 0.7 466.9 -0.28 7.22 0.10 0.03 12.63 2.6 465.9 0.60 7.21 0.50 0.24 12.41 17.9 458.0 8.12 7.18 1.02 0.50 12.15 37.5 448.2 17.80 7.16 5.60 2.63 10.01 196.5 369.5 98.09 7.01 6.88 3.16 9.48 236.5 349.8 118.44 6.98 9.66 4.23 8.41 316.7 310.4 159.25 6.91 12.74 5.31 7.35 396.9 271.1 200.10 6.84 21.41 7.38 5.28 552.3 194.8 279.24 6.65 28.55 7.38 5.28 552.3 194.8 279.24 6.48 29.97 7.38 5.28 552.2 194.8 279.18 6.45 31.40 7.38 5.28 552.1 194.8 279.09 6.41

170 APPENDIX H VOLUME AND CELL HEIGHT CHANGE SPREADSHEETS FOR IRREVERSIBLE REACTIONS

Table H.3 Volume change calculations for concentrated FeS2 and 5% CaCO3.

O2 Gypsum( Calcite Gypsum Calcite Total Volume (moles) moles) (moles) (cm3) (cm3) Change (cm3) pH 0.00 0.00 12.68 0.0 467.9 0.00 7.62 0.01 0.00 12.67 0.0 467.7 -0.18 7.37 0.05 0.01 12.65 0.7 466.9 -0.28 7.22 0.10 0.03 12.63 2.6 465.9 0.60 7.21 0.50 0.24 12.41 17.9 458.0 8.12 7.18 1.02 0.50 12.15 37.5 448.2 17.80 7.16 5.60 2.63 10.01 196.5 369.5 98.09 7.01 12.74 5.31 7.35 396.9 271.1 200.10 6.84 22.18 7.99 4.68 597.3 172.7 302.12 6.63 35.22 10.66 2.01 797.3 74.3 403.73 6.32 41.94 11.73 0.95 877.2 34.9 444.26 6.16 45.73 12.26 0.41 917.0 15.3 464.46 6.08 49.32 12.68 0.00 948.5 0.0 480.61 4.87 50.57 12.68 0.00 948.4 0.0 480.53 1.49

171 APPENDIX H VOLUME AND CELL HEIGHT CHANGE SPREADSHEETS FOR IRREVERSIBLE REACTIONS

Table H.4 Cell height change calculations for varying concentrations of % S2 and 5%

CaCO3. 0.1% S2 0.5% S2 Concentrated FeS2

Total Height Total Height Total Height Change Change (cm) Change (cm) (cm) 0.00 0.0000 0.0000 -0.0002 -0.0002 -0.0002 -0.0002 -0.0002 -0.0002 0.0005 0.0005 0.0005 0.0070 0.0070 0.0070 0.0154 0.0154 0.0154 0.0456 0.0849 0.0849 0.0456 0.1025 0.1731 0.0457 0.1378 0.2613 0.0457 0.1731 0.3492 0.0459 0.2416 0.3843 0.0457 0.2416 0.4018 0.0456 0.2415 0.4158 0.0454 0.2414 0.4157

172

APPENDIX I

DIGITAL IMAGES OF IMAGE J ANALYSIS

173 APPENDIX I – DIGITAL IMAGES OF IMAGE J ANALYSIS (Scale bar = 1000 μm unless noted otherwise)

S1L-1 S1L-2 S1L-3

S1L-4 S1L-5 S1L-6

S1L-7 S1L-8 S1L-9

S1L-10X (Bar = 100 μm) S1L-20X (Bar = 50 μm) S1M-1

S1M-2 S1M-3 S1M-4

174 APPENDIX I – DIGITAL IMAGES OF IMAGE J ANALYSIS (Scale bar = 1000 μm unless noted otherwise)

S1M-5 S1M-6 S1M-7

S1M-8 S1M-9 S1M-10

S1M-10X (Bar = 100 μm) S1M-20X (Bar = μm) S1R-1

S1R-2 S1R-3 S1R-4

S1R-5 S1R-6 S1R-7

175 APPENDIX I – DIGITAL IMAGES OF IMAGE J ANALYSIS (Scale bar = 1000 μm unless noted otherwise)

S1R-8 S1R-9 S1R-10X (Bar = 100 μm)

S1R-20X (Bar = 50 μm) S2L-1 S2L-2

S2L-3 S2L-4 S2L-5

S2L-6 S2L-7 S2L-8

S2L-9 S2L-10 S2L-11

176 APPENDIX I – DIGITAL IMAGES OF IMAGE J ANALYSIS (Scale bar = 1000 μm unless noted otherwise)

S2L-10X (Bar = 100 μm) S2L-20X (Bar = 50μm) S2M-1

S2M-2 S2M-3 S2M-4

S2M-5 S2M-6 S2M-7

S2M-8 S2M-9 S2M-10

S2M-11 S2M-12 S2M-10X (Bar = 100 μm)

177 APPENDIX I – DIGITAL IMAGES OF IMAGE J ANALYSIS (Scale bar = 1000 μm unless noted otherwise)

S2M-20X (Bar = 50μm) S2R-1 S2R-2

S2R-3 S2R-4 S2R-5

S2R-6 S2R-7 S2R-8

S2R-9 S2R-10 S2R-11

S2R-10X (Bar = 100 μm) S2R-20X (Bar = 50 μm) S3L-5X

178 APPENDIX I – DIGITAL IMAGES OF IMAGE J ANALYSIS (Scale bar = 1000 μm unless noted otherwise)

S3L-10X (Bar = 100 μm) S3L-20X (Bar = 50 μm) S3M-5X

S3M-10X (Bar = 100 μm) S3M-20X (Bar = 50 μm) S3R-5X

S3R-10X (Bar = 100 μm) S3R-20X (Bar = 50μm)

179

APPENDIX J

SPREADSHEETS OF ALL IMAGE J MEASUREMENTS

180 APPENDIX J - SPREADSHEETS OF ALL IMAGE J MEASUREMENTS

Table J.1 Measurements for cross-section S1 (PDJ).

Vertical Distance Vertical Distance Vertical Distance Section Between Section Between Section Between S1L-1 Gypsum (mm) S1M-1 Gypsum (mm) S1R-1 Gypsum (mm) Total Thickness 1 103.623 87 147 -90 1.78 Total Thickness 1 94.264 79 189 -90 1.452 Total Thickness 1 97.573 81 147 -90 1.532 2 105.122 98 118 -90 0.192 2 107.304 98 136 -90 0.088 2 100.843 92 124 -90 0.48 3 105.343 97 127 -90 0.136 3 108.889 104 118 -90 0.032 3 98.391 89 116 -90 0.252 4 100.235 86 112 -90 0.132 4 98.426 92 107 -90 0.184 4 93.35 90 97 -90 0.076 5 103.25 99 110 -90 0.06 5 94.673 87 109 -90 0.192 5 88.357 81 108 -90 0.5 6 101.667 98 109 -90 0.08 6 92.216 87 102 -90 0.144 S1R-2 7 101.326 95 115 -90 0.168 7 88.167 82 109 -90 0.188 Total Thickness 1 104.025 87 140 -90 1.884 8 94.979 89 104 -90 0.184 8 84.15 79 90 -90 0.076 2 104.05 93 116 -90 0.236 9 93.483 86 102 -90 0.228 9 85 79 93 -90 0.108 3 105.841 96 118 -90 0.248 10 90.294 85 97 -90 0.064 10 83.591 80 93 -90 0.084 4 104.492 97 117 -90 0.232 S1L-2 S1M-2 5 105.207 99 112 -90 0.112 Total Thickness 1 103.555 88 125 -90 1.784 Total Thickness 1 103.72 70 142 -90 1.896 6 101 94 112 -90 0.108 2 105.727 100 115 -90 0.084 2 103.255 90 117 -90 0.2 7 98.732 91 108 -90 0.22 3 105.042 101 108 -90 0.092 3 109.307 99 125 -90 0.348 8 93.469 84 114 -90 0.252 4 108.143 97 125 -90 0.108 4 106.317 100 121 -90 0.236 S1R-3 5 107.574 99 121 -90 0.184 5 105.392 95 129 -90 0.2 Total Thickness 1 104.509 87 163 -90 1.828 6 105.4 95 114 -90 0.036 6 98.833 93 103 -90 0.068 2 106.947 99 119 -90 0.39 7 107.737 100 123 -90 0.072 7 99.4 94 111 -90 0.076 3 102.053 97 111 -90 0.072 8 102.151 95 113 -90 0.208 8 96.174 91 103 -90 0.088 4 104.255 94 120 -90 0.216 9 95.667 89 103 -90 0.116 9 95.982 87 113 -90 0.216 5 98.789 90 109 -90 0.376 10 97.462 88 110 -90 0.152 S1M-3 6 94.673 84 102 -90 0.216 11 95.563 91 106 -90 0.06 Total Thickness 1 104.049 61 159 -90 1.868 S1R-4 S1L-3 2 108.125 99 123 -90 0.092 Total Thickness 1 105.328 87 135 -90 1.788 Total Thickness 1 104.623 86 132 -90 1.808 3 100.302 63 121 -90 0.208 2 107.72 101 122 -90 0.296 2 109.347 99 129 -90 0.192 4 107.723 100 119 -90 0.256 3 109.762 105 118 -90 0.08 3 106.467 100 113 -90 0.056 5 107.421 101 115 -90 0.072 4 108.8 102 120 -90 0.076 4 107.552 96 118 -90 0.112 6 104.704 99 112 -90 0.104 5 105.4 98 114 -90 0.296 5 105.762 99 113 -90 0.08 7 101.707 95 112 -90 0.16 6 98.342 87 111 -90 0.44 6 104.263 99 111 -90 0.072 8 97.127 86 117 -90 0.404 7 98.211 88 105 -90 0.072 7 99.9 96 105 -90 0.036 S1M-4 S1R-5 8 100.684 96 107 -90 0.072 Total Thickness 1 103.103 64 126 -90 1.856 Total Thickness 1 103.786 90 127 -90 1.792 9 99.294 94 106 -90 0.064 2 106.8 101 120 -90 0.056 2 106.681 99 127 -90 0.272 10 99.364 90 123 -90 0.128 3 97.5 64 116 -90 0.068 3 105.194 97 114 -90 0.264 11 93.917 88 103 -90 0.14 4 107.317 101 112 -90 0.16 4 106.353 99 132 -90 0.132 S1L-4 5 106.217 97 117 -90 0.18 5 107.278 97 119 -90 0.14 Total Thickness 1 104.111 87 181 -90 1.868 6 105.73 99 121 -90 0.248 6 100.187 91 110 -90 0.296 2 105.038 96 124 -90 0.42 7 104 100 111 -90 0.064 7 96.238 90 102 -90 0.08 3 107.387 99 123 -90 0.368 8 98.292 89 113 -90 0.42 8 98.244 91 109 -90 0.176 4 104.08 96 120 -90 0.196 9 95.941 88 113 -90 0.132 S1R-6 5 98.304 94 104 -90 0.088 10 93.684 86 109 -90 0.072 Total Thickness 1 103.213 85 128 -90 1.876 6 103.789 95 121 -90 0.072 S1M-5 2 105.826 98 117 -90 0.34 7 95.492 87 107 -90 0.256 Total Thickness 1 104.528 63 239 -90 1.844 3 106.682 100 112 -90 0.084 S1L-5 2 107.938 102 113 -90 0.06 4 106.714 99 113 -90 0.052 Total Thickness 1 103.774 88 137 -90 1.908 3 83.389 63 118 -90 0.068 5 106.152 96 122 -90 0.392 2 108.233 100 132 -90 0.168 4 102.692 100 107 -90 0.048 6 99.652 96 108 -90 0.088 3 109.95 104 117 -90 0.076 5 105.681 99 116 -90 0.184 7 98.5 95 106 -90 0.036 4 108.921 99 124 -90 0.248 6 106.172 101 119 -90 0.112 8 93.219 83 110 -90 0.416 5 106.538 97 127 -90 0.256 7 104 98 114 -90 0.12 S1R-7 6 102.24 93 114 -90 0.196 8 104.72 94 131 -90 0.196 Total Thickness 1 104.009 87 137 -90 1.852 7 98.143 93 103 -90 0.08 9 96.794 89 104 -90 0.132 2 106.296 98 125 -90 0.388 8 95.073 88 105 -90 0.216 10 95.222 90 112 -90 0.068 3 108.529 101 124 -90 0.276 S1L-6 11 95.216 86 113 -90 0.144 4 100.31 89 120 -90 0.668 Total Thickness 1 105.259 87 132 -90 1.756 S1M-6 5 91.623 85 97 -90 0.208 2 112.083 101 123 -90 0.092 Total Thickness 1 102.664 59 183 -90 1.852 S1R-8 3 109.235 102 122 -90 0.132 2 105 97 132 -90 0.088 Total Thickness 1 103.053 86 128 -90 1.868 4 106.283 96 117 -90 0.392 3 83.957 61 109 -90 0.088 2 104.103 95 114 -90 0.268 5 98.719 88 123 -90 0.224 4 102.107 97 120 -90 0.108 3 106.234 100 118 -90 0.184 6 94.543 87 111 -90 0.32 5 99.538 93 106 -90 0.152 4 104.92 97 114 -90 0.096 S1L-7 6 103.061 95 116 -90 0.324 5 104.346 97 113 -90 0.204 Total Thickness 1 104.282 87 124 -90 1.796 7 98.763 91 111 -90 0.232 6 95.754 86 109 -90 0.468 2 104.621 97 119 -90 0.344 8 89.084 83 105 -90 0.376 7 91.957 86 98 -90 0.18 3 112.2 104 124 -90 0.076 S1M-7 S1R-8 4 104.947 92 120 -90 0.372 Total Thickness 1 104.5 63 166 -90 1.756 Total Thickness 1 103.491 85 161 -90 1.788 5 96.012 87 118 -90 0.324 2 92.143 64 106 -90 0.08 2 103.418 95 117 -90 0.216 6 94.579 89 120 -90 0.072 3 105.5 100 110 -90 0.06 3 104.556 95 125 -90 0.212 S1L-8 4 107.2 101 124 -90 0.256 4 105.048 91 133 -90 0.244 Total Thickness 1 104.063 83 133 -90 1.912 5 105.6 98 120 -90 0.076 5 98.731 90 110 -90 0.412 2 108.411 103 125 -92 0.14 6 114 101 148 -90 0.04 6 95.902 85 115 -90 0.324 3 110.814 102 119 -90 0.119 7 100.744 93 128 -90 0.308 Total Section Thickness (mm) = 16.208 4 113.99 106 125 -90 0.138 8 98 93 102 -90 0.06 Total Thickness of Gypsum Infilling (mm) = 3.846 5 107.863 103 115 -90 0.213 9 96.5 91 106 -90 0.124 Change in Height (%) = 31.111 6 108.492 102 126 -90 0.109 10 98.946 91 113 -90 0.144 Average Gypsum Thickness (mm) = 0.087 7 104.046 95 116 -90 0.125 11 93 90 97 -90 0.036 Average Spacing Between Gypsum(mm)= 0.281 8 96.526 91 103 -90 0.203 S1M-8 9 94.602 89 100 -90 0.063 Total Thickness 1 104.826 63 141 -90 1.86 Average Total Section Thickness (mm) = 16.412 10 92.154 85 98 -90 0.175 2 98.273 63 120 -90 0.26 Average Total Thickness of Gyp sum Infilling (mm) = 4.630 S1L-9 3 105.766 100 116 -90 0.304 Average Change in Height (%) = 39.742 Total Thickness 1 97.345 85 135 -90 1.468 4 106.577 100 113 -90 0.1 Total Average Gypsum Thickness (mm) = 0.084 2 98.868 90 135 -90 0.148 5 109.111 101 120 -90 0.068 Total Average Spacing Between Gypsum (mm) = 0.306 3 98.033 93 103 -90 0.116 6 102.333 94 114 -90 0.2 4 98.009 91 114 -90 0.444 7 99.962 91 115 -90 0.204 60 5 98 93 105 -90 0.096 8 94.722 86 105 -90 0.212 6 94.077 86 112 -90 0.36 S1M-9 Total Section Thickness (mm) = 16.080 Total Thickness 1 103.129 59 125 -90 1.856 Total Thickness of Gypsum Infilling (mm) = 5.335 2 99.351 59 115 -90 0.292 Change in Height (%) = 49.651 3 106.578 101 115 -90 0.176 Average Gypsum Thickness (mm) = 0.094 4 105.73 100 115 -90 0.292 Average Spacing Between Gypsum (mm) = 0.189 5 98.596 93 113 -90 0.204 6 97.615 91 107 -90 0.204 75 7 94.395 88 101 -90 0.168 S1M-10 Total Thickness 1 86.292 51 125 -90 0.708 2 80.577 51 95 -90 0.28 3 87.871 80 115 -90 0.368 Total Section Thickness (mm) = 16.948 Total Thickness of Gypsum Infilling (mm) = 4.708 Change in Height (%) = 38.464 Average Gypsum Thickness (mm) = 0.072 Average Spacing Between Gypsum (mm) = 0.449

181 APPENDIX J - SPREADSHEETS OF ALL IMAGE J MEASUREMENTS

Table J.2 Measurements for cross-section S2 (PLJ).

Vertical Distance Vertical Distance Vertical Distance Section Between Section Between Section Between S2L-1 Gypsum (mm) S2M-1 Gypsum (mm) S2R-1 Gypsum (mm) Total Thickness 1 101.08 79 134 -90 1.7 Total Thickness 1 96.776 80 139 -90 1.584 Total Thickness 1 91.078 75 147 -90 1.384 2 107.133 95 132 -90 0.056 2 113.8 105 139 -90 0.076 2 102.333 89 126 -90 0.248 3 99.947 93 111 -90 0.072 3 106 93 123 -90 0.052 3 88.226 79 108 -90 0.544 4 94.098 87 123 -90 0.444 4 101.308 96 110 -90 0.048 4 84.774 77 102 -90 0.12 5 88.649 82 101 -90 0.292 5 95.173 90 118 -90 0.32 5 83.185 77 92 -90 0.104 6 85.889 78 93 -90 0.176 6 91.581 85 101 -90 0.292 6 80.263 75 90 -90 0.072 S2L-2 7 84.987 80 93 -90 0.3 7 79.333 77 86 -90 0.044 Total Thickness 1 102.535 85 136 -90 1.708 8 89.875 83 100 -90 0.028 S2R-2 2 103.792 98 111 -90 0.092 S2M-2 Total Thickness 1 102.882 86 148 -90 1.832 3 104.169 97 110 -90 0.256 Total Thickness 1 103.372 86 163 -90 1.736 2 105.885 96 116 -90 0.1 4 102.95 98 107 -90 0.076 2 104.057 99 118 -90 0.208 3 104.952 98 113 -90 0.08 5 104.522 97 113 -90 0.088 3 105.2 101 113 -90 0.096 4 105.71 99 116 -90 0.12 6 100.608 93 121 -90 0.2 4 105.878 100 123 -90 0.16 5 105.9 97 128 -90 0.116 7 98.196 90 108 -90 0.2 5 108.143 99 124 -90 0.08 6 100.838 97 108 -90 0.144 8 95.614 86 108 -90 0.328 6 101.463 95 111 -90 0.212 7 102.76 95 115 -90 0.096 S2L-3 7 99.133 94 108 -90 0.056 8 101.389 94 109 -90 0.068 Total Thickness 1 103.294 87 129 -90 1.82 8 95.326 90 102 -90 0.18 9 95.083 90 105 -90 0.092 2 106.175 99 122 -90 0.248 9 94.526 89 103 -90 0.072 10 96 89 107 -90 0.288 3 103.81 96 120 -90 0.164 10 88.696 85 94 -90 0.088 11 90.892 86 99 -90 0.144 4 101.311 93 109 -90 0.176 11 90.696 85 95 -90 0.088 12 92 88 98 -90 0.036 5 100.147 94 116 -90 0.132 S2M-3 S2R-3 6 98.273 94 102 -90 0.04 Total Thickness 1 106.491 86 158 -90 1.804 Total Thickness 1 103.206 88 157 -90 1.824 7 95.543 86 105 -90 0.364 2 104.29 96 114 -90 0.244 2 104.286 99 115 -90 0.052 8 90.818 88 96 -90 0.04 3 109.25 101 121 -90 0.06 3 105.857 99 110 -90 0.08 9 92.182 87 105 -90 0.084 4 103.786 99 109 -90 0.052 4 103.405 97 115 -90 0.144 S2L-4 5 102.907 95 127 -90 0.34 5 110.828 99 157 -90 0.112 Total Thickness 1 104.022 87 183 -90 1.84 6 97.794 91 107 -90 0.132 6 102.757 98 110 -90 0.144 2 107.295 99 127 -90 0.172 7 96.943 86 123 -90 0.344 7 103.818 99 112 -90 0.084 3 105.194 99 119 -90 0.284 8 94 89 110 -90 0.052 8 101.524 95 111 -90 0.08 4 108.047 97 125 -90 0.168 S2M-4 9 98.932 91 108 -90 0.232 5 102.238 96 114 -90 0.316 Total Thickness 1 104.7 87 143 -90 1.876 10 96.467 93 102 -90 0.056 6 96.628 90 108 -90 0.308 2 106.117 98 126 -90 0.236 11 95.886 88 114 -90 0.312 7 102.816 88 255 -90 0.148 3 107.412 102 124 -90 0.132 12 95.706 89 108 -90 0.064 S2L-5 4 105 100 113 -90 0.188 S2R-4 Total Thickness 1 103.521 85 171 -90 1.832 5 106.212 100 117 -90 0.204 Total Thickness 1 101.784 83 145 -90 1.844 2 105.486 101 113 -90 0.144 6 101.553 93 126 -90 0.184 2 105.174 99 112 -90 0.088 3 103.018 95 130 -90 0.224 7 95.573 89 108 -90 0.296 3 105.688 96 122 -90 0.252 4 104.167 99 111 -90 0.164 8 91.32 87 95 -90 0.096 4 103.545 99 107 -90 0.04 5 105.196 98 112 -90 0.22 9 94.364 90 100 -90 0.04 5 102.875 94 120 -90 0.092 6 101.405 93 113 -90 0.144 S2M-5 6 101.079 95 115 -90 0.148 7 94.81 86 101 -90 0.08 Total Thickness 1 105.747 81.1 213 -90 1.892 7 99.414 95 105 -90 0.112 8 89.233 85 98 -90 0.116 2 97.956 93 105 -90 0.176 8 97.456 91 115 -90 0.356 9 91.474 88 96 -90 0.072 3 99.209 93 110 -90 0.264 9 101.294 93 110 -90 0.064 S2L-6 4 101.912 95 119 -90 0.224 10 91.545 84 111 -90 0.172 Total Thickness 1 104.06 86 125 -90 1.864 5 97.287 86 127 -90 0.456 11 92.286 89 97 -90 0.052 2 107.741 101 122 -90 0.336 6 90.721 81 103 -90 0.24 S2R-5 3 107.897 101 121 -90 0.112 7 85.333 81 91 -90 0.092 Total Thickness 1 104.721 84 195 -90 1.82 4 110.788 101 123 -90 0.128 S2M-6 2 101.529 96 109 -90 0.064 5 105.055 96 120 -90 0.36 Total Thickness 1 103.682 86 142 -90 1.832 3 102.98 98 112 -90 0.196 6 101.138 94 108 -90 0.112 2 107.65 99 137 -90 0.236 4 104.226 97 122 -90 0.208 7 95.375 86 125 -90 0.412 3 105.333 102 108 -90 0.02 5 107.2 98 121 -90 0.176 S2L-7 4 108.467 102 124 -90 0.056 6 103.094 97 116 -90 0.124 Total Thickness 1 104.676 89 239 -90 1.836 5 105.099 98 112 -90 0.36 7 99.229 93 109 -90 0.136 2 108.136 100 122 -90 0.524 6 109.086 99 142 -90 0.136 8 96.392 88 113 -90 0.2 3 102.33 89 239 -90 1.112 7 104.205 92 118 -90 0.172 9 93.308 89 102 -90 0.048 S2L-8 8 97.368 86 114 -90 0.464 10 93 87 102 -90 0.092 Total Thickness 1 102.693 86 138 -90 1.9 9 94.174 87 102 -90 0.088 S2R-6 2 106.744 98 138 -90 0.356 S2M-7 Total Thickness 1 103.011 86 133 -90 1.868 3 105.6 100 122 -90 0.196 Total Thickness 1 104.015 86 176 -90 1.848 2 108.161 100 120 -90 0.12 4 101.766 91 133 -90 0.372 2 106.34 99 124 -90 0.208 3 104.667 99 112 -90 0.14 5 105.778 98 115 -90 0.068 3 108.444 100 129 -90 0.104 4 106.654 102 112 -90 0.1 6 96.622 89 107 -90 0.388 4 106.981 98 138 -90 0.42 5 107.235 102 113 -90 0.064 7 93.32 86 104 -90 0.096 5 103.946 96 112 -90 0.22 6 103 95 113 -90 0.168 S2L-9 6 101.011 93 135 -90 0.36 7 102.259 98 111 -90 0.104 Total Thickness 1 103.514 85 158 -90 1.896 7 93.759 86 113 -90 0.328 8 106.886 96 133 -90 0.136 2 105.682 94 146 -90 0.512 S2M-8 9 95.354 86 115 -90 0.652 3 99.014 91 116 -90 0.276 Total Thickness 1 103.434 85 192 -90 1.828 S2R-7 4 98.245 89 112 -90 0.208 2 104.927 99 124 -90 0.216 Total Thickness 1 104.478 84 161 -90 1.828 5 95.038 85 134 -90 0.308 3 111.467 97 155 90 0.236 2 107.455 99 115 -90 0.04 S2L-10 4 104.451 96 116 -90 0.324 3 108.188 97 131 -90 0.188 Total Thickness 1 102.116 79 148 -90 1.784 5 98.114 89 113 -90 0.524 4 105.26 96 137 -90 0.304 2 96.786 89 106 -90 0.332 6 92.649 85 111 -90 0.292 5 109.313 99 124 -90 0.06 3 88.216 79 105 -90 0.664 S2M-9 6 100.76 92 136 -90 0.38 S2L-11 Total Thickness 1 104.454 87 167 -90 1.82 7 93.544 87 103 -90 0.452 Total Thickness 1 90.25 83 104 -90 0.684 2 102.85 98 113 -90 0.076 S2R-8 Total Section Thickness (mm) = 18.864 3 102.85 98 112 -90 0.076 Total Thickness 1 102.603 87 208 -90 1.82 Total Thickness of Gypsum Infilling (mm) = 5.220 4 102.889 97 108 -90 0.068 2 105.143 98 115 -90 0.304 Change in Height (%) = 38.259 5 104.556 100 111 -90 0.032 3 103.776 95 120 -90 0.3 Average Gypsum Thickness (mm) = 0.119 6 105.538 97 113 -90 0.204 4 99.927 92 115 -90 0.436 Average Spacing Between Gypsum (mm) = 0.310 7 107.722 100 124 -90 0.14 5 94.323 87 108 -90 0.256 8 102.664 89 153 -90 0.424 6 91.065 87 96 -90 0.12 Average Total Section Thickness (mm) = 19.503 9 98.596 86 165 -90 0.452 S2R-9 Average Total Thickness of Gypsum Infilling (mm) = 4.821 S2M-10 Total Thickness 1 99.829 85 127 -90 1.732 Average Change in Height (%) = 33.115 Total Thickness 1 96.901 68 219 -90 1.896 2 103.65 96 110 -90 0.156 Total Average Gypsum Thickness (mm) = 0.089 2 81.25 75 116 -90 0.892 3 100.744 93 116 -90 0.528 Total Average Spacing Between Gypsum (mm) = 0.268 3 74 67 112 -90 0.188 4 96.071 90 101 -90 0.108 S2M-11 5 93.512 88 125 -90 0.16 Total Thickness 1 102.837 85 137 -90 1.832 6 90.871 85 98 -90 0.276 65 2 105.576 96 129 -90 1.108 S2R-10 3 96.23 85 125 -90 0.536 Total Thickness 1 103.324 83 160 -90 1.812 4 93.133 88 104 -90 0.056 2 112.611 94 160 -90 0.14 S2M-12 3 94.768 80 123 -90 1.1 Total Thickness 1 83.727 77 100 -90 0.64 4 92.059 88 96 -90 0.064 Total Section Thickness (mm) = 20.588 S2R-11 Total Thickness of Gypsum Infilling (mm) = 4.524 Total Thickness 1 93.531 84 116 -90 1.292 Change in Height (%) = 28.162 2 91.943 87 100 -90 0.488 Average Gypsum Thickness (mm) = 0.077 3 92.42 84 109 -90 0.596 Average Spacing Between Gypsum (mm) = 0.272 Total Section Thickness (mm) = 19.056 Total Thickness of Gypsum Infilling (mm) = 4.720 82 Change in Height (%) = 32.924 Average Gypsum Thickness (mm) = 0.073 Average Spacing Between Gypsum (mm ) = 0.221

182

APPENDIX K

ACIDITHIOBACILLUS FERROOXIDANS GROWTH MEDIUM

183 APPENDIX K – ACIDITHIOBACILLUS FERROOXIDANS GROWTH MEDIUM

ATCC Medium: 2039 Acidithiobacillus ferrooxidans medium

Solution A:

(NH4)SO4: ...... 0.8 g MgSO4•7H2O ...... 2.0 g K2HPO4 ...... 0.4 g Wolfe’s Mineral Solution (see below) ...... 5.0 ml Distilled water ...... 800.0 ml

Adjust Solution A to pH 2.3 with H2SO4. Filter-sterilize.

Solution B: FeSO4•7H2O ...... 20.0 g Distilled water ...... 200.0 ml

Stir Solution B to dissolve and quickly filter-sterilize. Aseptically combine Solutions A and B. (A yellow precipitate is normal; it becomes darker as the iron oxidizes.)

Wolfe’s Mineral Solution: Available from ATCC as a sterile ready-to-use liquid (Trace Mineral Supplement, catalog no. MD-TMS).

Nitrilotriacetic acid ...... 1.5 g MgSO4•7H2O ...... 3.0 g MnSO4•H2O ...... 0.5 g NaCl ...... 1.0 g FeSO4•7H2O ...... 0.1 g CoCl2•6H2O ...... 0.1 g CaCl2 ...... 0.1 g ZnSO4•7H2O ...... 0.1 g CuSO4•5H2O ...... 0.01 g AlK(SO4)•12H2O ...... 0.01 g H3BO3 ...... 0.01 g Na2MoO4•2H2O ...... 0.01 g Distilled water ...... 1.0 L

Add nitrilotriacetic acid to approximately 500 ml of water and adjust to pH 6.5 with KOH to dissolve the compound. Bring volume to 1.0 L with remaining water and add remaining compounds one at a time.

184

APPENDIX L

ILLITE SHALE GRAIN SIZE ANALYSIS

185 APPENDIX L - ILLITE SHALE GRAIN SIZE ANALYSIS

MAIN OFFICE: 2380 Commercial Boulevard, State College, PA 16801, Phone (814) 231-8845, Fax (814) 231-8846

BRANCH OFFICES: 155 Phillips Park Drive, Williamsport, PA 17702, Phone (570) 567-1016, Fax (570) 567-1017 4500 Sixth Avenue, Suite 124, Altoona, PA 16602, Phone (814) 941-6904, Fax (814) 941-6941

Sieve Analysis - Dry Sieve (ASTM C136) Specification: None Gravel Sand Silt & Clay

100

90 80

70 60

50 40 Percent Finer 30 20

10 0 100 10 1 0.1 0.01 Grain Size (mm)

% Silt % -#200 Grain Size % +3" % Gravel % Sand & Clay (Wash) D60 5.40 Coefficients D 2.200 C 2.24 0.0 0.0 100.0 0.0 0.0 30 c D10 0.400 Cu 13.50 Sieve Sieve Sieve Sieve Size Size Percent Specification Number Size Percent Specification (inches) (mm) Finer Bottom Top Size (mm) Finer Bottom Top No. 4 4.75 99.5 0.00 0.0 No. 10 2.00 85.0 0.00 0.0 No. 20 0.80 59.5 0.00 0.0 No. 30 0.60 49.9 0.00 0.0 No. 40 0.43 42.0 0.00 0.0 No. 60 0.25 33.2 0.00 0.0 No. 100 0.15 18.1 0.00 0.0 Project: Doctoral Thesis - Shad E. Hoover File Number: -- Civil Engineering Sample Location: Ward's Science Date: 28-Nov-06 0 Material Description: Illite Shale (Crushed) CMT I.D. No.: -- Well-Graded Sand (SW) Client: Penn State University University Park, PA CMT Laboratories, Inc.

186 REFERENCES

Adams, J. (1982), “Stress-relief buckles in the McFarland quarry, Ottawa,” Canadian Journal of Earth Science, Vol. 19, pp. 1883-1887.

American Society for Testing and Materials (2000), Annual Book of ASTM Standards, Vol. 04.08, Conshohocken, Pennsylvania.

American Society for Testing and Materials (2001), Annual Book of ASTM Standards, Vol. 11.06, Conshohocken, Pennsylvania.

Anonymous (1960), “Structures Don’t Settle in this Shale, but Watch Out for Heave,” Engineering News Record, Vol. 164, No. 5, pp. 46-48.

Appelo, C.A.J., and Postma, D. (1993), Geochemistry, Groundwater and Pollution, Rotterdam, Netherlands, and Brookfield, Vermont, A.A. Balkema.

Ballivy, Gerard and Bellaloui, Achour (1999), “New Swelling Test to Characterize the Expansive Potential of Pyritic Rockfill,” Geotechnical News, Vol. 17, No. 4, pp. 53-55.

Barnes, H.L. and Langmuir, D. (1978), Geochemical Prospecting Handbook for Metals and Associated Elements, National Science Foundation Grant No. AER77- 06511A02, Annual Report.

Barnhisel, R. I. and J. Harrison (1976), “Estimating Lime Requirement by a Modified Hydrogen Peroxide Potential Acidity Method,” Lexington, KY. (Unpublished method for KY Agr. Ex. Sta., Soil Testing Laboratory).

Belgeri, J.J. and T.C. Siegel (1998), “Design and Performance of Foundations in Expansive Shale,” Ohio River Valley Soils Seminar XXIX, Louisville, Kentucky.

Bell, Fred G. (2000), Engineering Properties of Soils and Rock, Fourth Edition, Blackwell Science, pp. 277-281.

Berner, R.A. (1984), “Sedimentary pyrite formation: an update,” Geochemica Acta, Vol. 48, pp. 605-615.

Berg, T. M., and Dodge, C. M., eds. (1981), “Atlas of Preliminary Geologic Quadrangle Maps of Pennsylvania,” Pennsylvania Geological Survey.

Brady, Keith B.C., Smith, Michael W., and Schueck, Joseph ed. (1998), Coal Mine Drainage Prediction and Pollution Prevention in Pennsylvania. Pennsylvania Department of Environmental Protection. Harrisburg, Pennsylvania.

187 Brett, Carlton E. and Gordon C. Baird (1986), “Comparative Taphonomy: A Key to Paleoenvironmental Interpretation Based on Fossil Preservation,” Palaios, Vol. 1, No. 3, pp. 216 & 217.

Bryant, Lee Davis (2003), “Geotechnical Problems with Pyritic Rock and Soil,” Master of Science Thesis, Virginia Polytechnic Institute and State University, Blacksburg, Virginia.

Bureau of Topographic and Geologic Survey (2005), “Geologic Units Containing Potentially Significant Acid-Producing Sulfide Minerals,” Pennsylvania Department of Conservation and Natural Resources and the Pennsylvania Geological Society.

Burt Hill (2006), “Pyrite Remediation Assessment Report - Evangelical Community Hospital,” Project No. 06009.01.

Canfield, D.E., and Raiswell, R. (1991), Pyrite Formation and Fossil Preservation, in Allison, P.A. and Briggs, D.E.G., eds, Taphonomy: Releasing the data Locked in the Fossil Record, New York, Plenum Press, pp. 337-387.

Canfield, D.E., Raiswell, R., and Bottrell, S. (1992), “The Reactivity of Sedimentary Iron Minerals Towards Sulfide”, American Journal of Science, Vol. 292, pp. 659-683.

Carroll, D. (1958), “Role of Clay Minerals in the Transportation of Iron,” Geochimica et Cosmochinica Acta, Vol. 14, pp. 1-27.

Cormier, Marie-Claude (2000), “Analyse Statisque de Raports D’Expertise CTQ-M200 Relevant Les Dommages Structuraux Attribuables A La Presence de Pyrite Dans Les Remeblais Sous Dalles Debatiments Residentiel,” Québec Société d’habitation du Québec.

Coveney, R. M. and E. J. Parizek (1977), "Deformation of Mine Floor by Sulfide Alteration,” Bulletin of Association of Engineering Geologists XIV(1): pp. 131- 156.

Dougherty, Murray T. and Barsotti, Nino J. (1972), “Structural Damage and Potentially Expansive Sulfide Minerals,” Bulletin of the Associations of Engineering Geologists, Vol. IX, No. 2, pp. 105-125.

Dubbe, D.E., Usmen, M.A. and Moulten, L.K. (1984), “Expansive Pyritic Shales,” A Paper Offered for Publication by the Transportation Research Board, Washington, D.C. 36p.

Ettensohn, F. R. (1998). “Sedimentary Features indicating Erosion, Condensation, and Hiatuses in the Chattanooga Shale of Central Tennessee: Relevance for Sedimentary and Stratigraphic Evolution, Shales and Mudstones I: Basin Studies,

188 Sedimentology, and Paleontology,” J. Schieber, W. Zimmerle and P. S. Sethi. Stuttgart, E. Schweizerbart'sche, Verlagsbuchhandlung, Vol. I, pp. 109-128.

Freeman, Thomas E. (2003), “Final Report – Forensic Investigation of Pavement Distress: Old Airport Road in Bristol, Virginia,” Virginia Transportation Research Council. Charlottesville, Virginia.

Geidel, G., F.T. Caruccio, R.J. Hornberger and K.B.C. Brady, (2000), “Guidelines and recommendations for use of kinetic tests for coal mining (AMD) prediction in the Eastern U.S.” In: Prediction of Water Quality at Surface Coal Mines, Morgantown,: West Virginia University, National Mine Land Reclamation Center, pp. 99-139.

Geyer, Alan R. and Wilshusen, J. Peter (1982), Engineering Characteristics of the Rocks of Pennsylvania, Pennsylvania Geologic Survey, Fourth Series, Harrisburg, PA. 2nd Ed.

Gold, David P. and Doden, A.G. (1996), “The Age, Depth, and Nature of Acid- Generating Mineralization Along Segments of I-99 on Bald Eagle Ridge,” Pennsylvania Department of Transportation Bureau of Planning and Research, Harrisburg, Pennsylvania.

Gustafsson, Jon Peter, (2007), Visual MINTEQ, ver. 2.52 based on MINTEQA2 ver. 4.0, Department of Land and Water Resources Engineering, Stockholm, Sweden.

Hammarstrom, J.M., Seal II, R.R., Meir, A.L., Kornfield, J.M. (2004), “Secondary Sulfate Minerals Associated with Acid Drainage in the Eastern U.S.: Recycling of Metals and Acidity in Surficial Environments,” Chemical Geology Including Isotope Geoscience, Volume 215, pp. 407-431.

Hawkins, A.B. & Pinches, G.M. (1987), “Cause and significance of heave at Llandough Hospital, Cardiff – a case history of ground floor heave due to gypsum growth,” Quarterly Journal of Engineering Geology, London, Vol. 20, pp. 41-57.

Hoover, Shad E. (2002), “Swelling Behavior of Pyritic Shales in the Marcellus Formation,” Master of Science Thesis, Pennsylvania State University, Department of Civil and Environmental Engineering, University Park, PA.

Hoover, Shad E. (2004), “Structural Damage Induced by Pyritic Shale,” Fifth International Conference on Case Histories in Geotechnical Engineering, New York, NY (USA).

Hoover, Shad E. and Pease, James B. (2007), “Micropile Underpinning Over Expansive Pyritic Shales,” 8th International Workshop on Micropiles, Toronto, Canada.

189 Hoover, Shad E. and Thornton, Jr., James P. (1998), “Geotechnical Report - Evangelical Hospital”, CMT Laboratories, Inc. CMT File No. 92886X.

Hornberger, R.J. and K.B.C. Brady (1998), “Kinetic (leaching) tests for the prediction of mine drainage quality”, In: K.B.C. Brady, et al., eds., Coal Mine Drainage Prediction and Pollution Prevention in Pennsylvania. Pennsylvania Department of Environmental Protection, 5600-BK-DEP2256, pp. 7-1 to 7-54.

Hornberger, R.J., K.A. Lapakko, G. E. Krueger, C.H. Bucknam, P.F. Ziemkiewicz, D.J.A. vanZyl. and H.H. Posey (2000), “The Acid Drainage Technology Initiative (ADTI),” In: Proceedings from the Fifth International Conference on Acid Rock Drainage. Littleton CO: Society for Mining Metallurgy and Exploration, pp. 41- 50.

Hosterman, J. W. and S. I. Whitlow (1980), “Munsell Color Value as Related to Organic Carbon in Devonian Shale of Appalachian Basin,” Reston, VA. U.S. Geological Survey: U.S. Geological Survey Open-file Report 80-660.

Hosterman, J. W. and S. I. Whitlow (1981), “Clay Mineralogy of Devonian Shales in the Appalachian Basin,” USGS Open-File Report 81-585, US Geological Survey, 170p.

Hsu, Sung-Chi and Nelson, Pricilla P. (2002), “Characterization of Eagle Ford Shale,” Engineering Geology, Vol. 67, pp. 169-183.

Ingebritsen, Steve, Sanford, Ward, and Neuzil, Chris (2006), Groundwater in Geologic Processes – 2nd Edition. Cambridge University Press. Cambridge, New York.

Jangal, C. E. (1971), “Swelling Shale in the Ostersund Area,” National Swedish Building Research, R35, p. 116.

Jaynes, D.B, Pionke, H.B., and Rogowski, A.S. (1984a), “Acid Mine Drainage from Reclaimed Coal Strip Mines – 2 Simulation Results of Model,” Water Resources Research, Vol. 20, No. 2, pp. 243-250.

Jaynes, D.B., Rogowski, A.S., and Pionke, H.B. (1984b), “Acid Mine Drainage from Reclaimed Coal Strip Mines – 1 Model Description,” Water Resources Research, Vol. 20, No. 2, pp. 233-242.

Krinsley, David H., Pye, Kenneth, Boggs, Sam and Tovey, N. Keith (1998), Backscattered Scanning Electron Microscopy and Image Analysis of Sediments and Sedimentary Rock, Cambridge University Press.

Langmuir, Donald (1997), Aqueous Environmental Geochemistry, Prentice-Hall, Inc., Upper Saddle River, New Jersey.

190 LECO Corporation, (1980), Instruction manual SC-132 sulfur system: St. Louis, Missouri.

Long, J.D. and Williams, P.A. (1990), “Report of Structural Distress Investigation – Phase I, Johnson City Public Library,” S&ME, Blountville, Tennessee, 31p.

McGarr, A. and Gay, N.C., (1978), “State of stress in the earth’s crust,” Annual Reviews of Earth and Planetary Sciences, Vol. 6, pp. 405-436.

Moum, J. and I. Rosenqvist (1959), "Sulfate Attack on Concrete in the Oslo Region,” Journal of the American Concrete Institute, pp. 257-264.

Observer-Reporter (2006), “Damages Piling up at Trinity South,” Price, Heidi, February 4, 2006, p. 1B.

Observer-Reporter (2006), “Trinity South Repair, Replacement Tops Substantial List of Capital Needs,” Richards, John, June 6, 2006, p. 3B.

Observer-Report (2007), “PDE Could Back a New Trinity South,” Mansfield, Karen, September 21, 2007, p. 2B.

Orndorff, Zenah (2001), “Evaluation of Sulfidic Materials in Virginia Highway Corridors”, Doctoral Thesis, Crop and Soil Environmental Sciences - Virginia Polytechnic Institute and State University, Blacksburg, VA, 171p.

Parkhurst, D.L., and Appelo, C.A.J. (1999), User's guide to PHREEQC (Version 2)--a computer program for speciation, batch-reaction, one-dimensional transport, and inverse geochemical calculations: U.S. Geological Survey Water-Resources Investigations Report 99-4259, 312 p.

Penner, E, and Eden, W.J. (1972), “Expansion of Pyritic Shales,” Canadian Building Digest. CBD-152. National Research Council Canada. www.nrc.ca/irc/cbd/cbd152e.html

Pennsylvania Department of Environmental Protection (1988), “Overburden Sampling and Testing Manual,” Small Operator Assistance Program, Bureau of Mining and Reclamation, Harrisburg, Pennsylvania.

Plumb, R.A. (1994), Variations of the least horizontal stress magnitude in sedimentary rocks: in Nelson, P.P. and Laubach, S.E., Rock Mechanics Models and Measurements: Challenges from Industry, A.A. Balkema, Rotterdam, Netherlands, pp. 71-78.

Portland Cement Association (2003), “PCA Structure Point – PCA Mats Finite Element Analysis Program.”

191 Pye, K. and Miller, J.A. (1988), “Chemical and biochemical weathering of pyritic mudrocks in a shale embankment,” Quarterly Journal of Engineering Geology, Vol. 23, pp. 365 – 381.

Reese, Stuart O. and Lee, Joseph J. (1998), Summary of Groundwater Monitoring Data 1985 – 1997) from Pennsylvania’s Ambient and Fixed Station Network (FSN) Monitoring Program - Selected Groundwater Basins in Southwestern, Southcentral and Southeastern Pennsylvania, Commonwealth of Pennsylvania - Department of Environmental Protection, Bureau of Water Supply Management.

Rimstidt, J.D. and Newcomb W.D. (1993), “Measurement and analysis of rate data - the rate of reaction of ferric iron with pyrite,” Geochimica et Cosmochimica Acta, Vol. 57, pp. 1919-1934.

Roberts, Willard L. (1990), Encyclopedia of Minerals, 2nd Edition, Van Nostrand Reinhold, New York, New York, 979p.

Schieber, Jurgen and Baird, Gordon (2001), “On the Origin and Significance of Pyrite Spheres in Devonian Black Shales North America,” Journal of Sedimentary Research, Vol. 71, No. 1, pp. 155-166.

Shultz, Charles H. Editor (1999), The Geology of Pennsylvania, Special Publication I. Pennsylvania Geological Survey & Pittsburgh Geological Society, pp. 117, 125, 428, 429, and 702.

Sehn, Allen L. (1998), “Evaluation of Minipile Connection Capacity for the Westport Library Project,” Special Study Report. Hayward Baker, Inc.

Sehn, Allen L. (2000), “Evaluation of Minipile Grouts and Minipile Connection Capacity at the Hotel Roanoke Project,” Special Study Report. Hayward Baker, Inc.

Sobek, A.A., Schuller, W.A., Freeman, J.R. and Smith, R.M. (1978), Field and Laboratory Methods Applicable to Overburdens and Minespoils, U.S. Environmental Protection Agency Environmental Protection Technology, EPA- 600/2-78-054.

Sullivan, L.A., Bush, R.T., McConchie, D., Lancaster, G., Haskins, P.G., and Clark, M.W. (1999), “Comparison of Peroxide-Oxidisable Sulfur and Chromium Reducible Sulfur Methods for Determination of Reduced Inorganic Sulfur in Soil,” Journal of Soil Research, Vol. 37, pp. 255-265.

Sullivan, L.A., Bush, R.T. and McConchie, D. (2000), “A Modified Chromium- Reducible Sulfur Method for Reduced Inorganic Sulfur: Optimum Reaction Time for Acid Sulfate Soil,” Australian Journal of Soil Research, Vol. 38, pp. 729-734.

192 Tanguay, Charles et al. (1999), Pyrite and Your House, Association Des Consommateurs Pour La Qualite Dans La Construction, Montreal, Quebec, 52p.

T.O.P. Drilling Supply Co., Inc. (2006), “Casing, 5 ½” O.D. x .415” wall x 3’ S-S Male x Female Thread,” Drawing 0002-53.

US Department of Transportation – Federal Highway Administration (2000), “Micropile Design and Construction Guidelines – Implementation Manual”, Publication No. FHWA-SA-97-070.

Wildman, R.A., Berner, R.A., Petch, S.T., Bolton, E.W., Eckert, J.O., Mok, U. and Evans, J.B. (2004), “The Weathering of Sedimentary Organic Matter as a Control on Atmospheric O2: I. Analysis of a Black Shale,” American Journal of Science, Vol. 304, pp. 234-249.

193 CURRICULUM VITA - SHAD E. HOOVER

EXPERIENCE: Technician - Pennsylvania Department of Transportation: Harrisburg, PA: May 1994 to August 1994 Geotechnical Engineer - Professional Service Industries, Inc.: Pittsburgh, PA: June 1995 to February 1997 Cleveland, OH: February 1997 to April 1998 Managing Engineer - CMT Laboratories, Inc.: State College, PA: April 1998 to Present

EDUCATION: Bachelor of Science Degree (B.S.) in Natural Sciences - Juniata College (1995) Bachelor of Science Degree (B.S.) in Civil Engineering - Pennsylvania State University (1995) Master of Science Degree (M.S.) in Civil Engineering - Pennsylvania State University (2002) Doctor of Philosophy Degree (Ph.D.) in Civil Engineering - Pennsylvania State University (2008)

CERTIFICATIONS: Engineer-in-Training (EIT), PA (1995) Professional Engineer (PE): PA (2000), NY (2002), MD (2003), WV (2005)

PUBLICATIONS: Hoover, Shad E. (2002), “Swelling Behavior of Pyritic Shales in the Marcellus Formation,” Master of Science Thesis, Pennsylvania State University, Department of Civil and Environmental Engineering, University Park, PA. Hoover, Shad E. (2004), “Structural Damage Induced by Pyritic Shale,” Fifth International Conference on Case Histories in Geotechnical Engineering, New York, NY. Doden, Arnold G., Gold, David P., Hoover, Shad E., Scheetz, Barry, Ellsworth, Chad (2008), “Road Bed and Building Heave from Alteration of Sulfide and Sulfate Minerals,” Northeastern Geological Society of America Conference, Buffalo, NY.

MEMBERSHIPS: American Society of Civil Engineers (1995) International Society for Soil Mechanics (2000) Deep Foundations Institute (2005) International Society of Micropiles – Founding Member (2006)

TEACHING EXPERIENCE: Juniata College, Spring 2004. Engineering Mechanics – Dynamics (Visiting Instructor) Pennsylvania State University, Fall 2004. Foundation Design (Visiting Instructor)