UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Factors affecting the permeability of gas shales Pathi, Venkat Suryanarayana Murthy 2008

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2008_fall_pathi_venkat_suryanarayana_murthy.pdf [ 4.04MB ]
Metadata
JSON: 24-1.0052381.json
JSON-LD: 24-1.0052381-ld.json
RDF/XML (Pretty): 24-1.0052381-rdf.xml
RDF/JSON: 24-1.0052381-rdf.json
Turtle: 24-1.0052381-turtle.txt
N-Triples: 24-1.0052381-rdf-ntriples.txt
Original Record: 24-1.0052381-source.json
Full Text
24-1.0052381-fulltext.txt
Citation
24-1.0052381.ris

Full Text

FACTORS AFFECTING THE PERMEABILITY OF GAS SHALES by VENKAT SURYANARAYANA MURTHY PATHI B.Sc., Andhra University, 1995 MUSe., Andhra University, 1997 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Geological Science) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) October, 2008 ©Venkat Suryanarayana Murthy Pathi, 2008 Abstract The mechanical properties and matrix permeability of gas shales are the most important properties in determining their production capacity. In this research, I have investigated the matrix permeability and rock mechanical properties of Western Canadian and Woodford shales. The matrix permeability was measured using pulse decay experiment. The pulse decay experiment was employed with triaxial experiments combined with mercury porosimetry, helium pydnometry, Rock-Eval pyrolysis, SEM and X-ray diffraction analysis to measure rock strength, pore size, porosity, total organic content, fabric and composition of samples. The permeability results were correlated with effective stress, anisotropy, fabric, rock strength, porosity, pore size and total organic content. Mineralogy plays an important role in determining the permeability of Canadian and Woodford shales. Higher permeability was observed in samples with high clay content, and low permeability was observed in samples with high quartz and carbonate content. Among the clay-, silica-, and calcite-rich Canadian shales, the calcite-rich shales had a very low permeability (1O md) compared to other shales. The permeability of all shales decline exponentially with increasing effective stress. Samples that were tested parallel to bedding had higher permeabilities than samples were tested normal to bedding. Among shales, the quartz-rich shales showed differences of 11 three to four orders of magnitude for the samples tested parallel to bedding, compared to those tested normal to the bedding. The largest anisotropy was found in the clay-rich samples. Clay-rich shales also have a well developed fabric with a strongly preferred orientation, while the quartz-rich shales had random orientation of the fabric. The porosimetry results suggest that fluid flow is mostly in the meso (2-50 nm) and macro pores (>50 nm) of the Woodford shales. Samples with higher clay content (>30%) showed a higher intrusion volume in macro pores, while samples with higher quartz content showed intrusion volume in micro pores. Porosity is correlated to permeability in the Western Canadian shales and showed a linear correlation within the Woodford shales. Even though calcite-rich Canadian shales and quartz-rich Woodford shales have high TOC content, TOC was not seen to effect permeability. Triaxial compression rock testing was conducted on the Woodford shales to measure the elastic properties and strength behaviour of shale. Lithologic composition plays an important role in the strength and pore compressibility of shale. Quartz-rich or carbonate rich shales have a brittle behaviour and clay-rich shales have a ductile behaviour. Pore compressibility is greater in the clay-rich shales, and less in the quartz-rich shales. 111 TABLE OF CONTENTS Abstract ii Table of Contents iv List of Tables ix List of Figures x Acknowledgements xv CHAPTER-i INTRODUCTION i 1.1 INTRODUCTORY STATEMENTS 1 1.2 STRUCTURE OF THESIS 3 1.3 REFERENCES 4 CHAPTER-2 FACTORS AFFECTING THE PERMEABILITY OF SHALES IN THE WESTERN CANADIAN BASIN 6 2.1 ABSTRACT 6 2.2 INTRODUCTION 8 2.3 STUDY AREA AND DATA 10 2.4. EXPERIMENTAL PROCEDURES 12 iv 2.4.1 Sample Preparation.12. 2.4.2 Permeability through Pulse Decay Experiment 12 2.4.3 Procedure 14 2.4.4 Theory 16 2.4.5 Porosity Measurements 20 2.4.6 Mineralogy 21 2.4.7 Total Organic Content (TOC) 22 2.4.8 Scanning Electron Microscopy (SEM) 23 2.5 EXPERIMENTAL RESULTS 24 2.5.1 Permeability-Effective Stress Measurements 24 2.5.2 Permeability Anisotropy 35 2.5.3 SEM Observations 40 2.5.4 Mineralogy 47 2.5.5 TOC 51 2.5.6 Permeability vs. Porosity 53 2.6 DISCUSSION 54 2.7 CONCLUSIONS 55 2.8 REFERENCES 56 V CHAPTER-3 TRIAXIAL TESTING AND MECHANICAL BEHAVIOUR OF WOODFORD SHALES 63 3.1 ABSTRACT 63 3.2 INTRODUCTION 65 3.3 GEOLOGY 67 3.4 EXPERIMENTAL PROCEDURES 70 3.4.1 Sample description and preparation 70 3.4.2 Apparatus 72 3.4.3 Testing procedure 76 3.5 RESULTS 77 3.5.1 Stress-strain curves 77 3.5.2 Influence of Lithology 79 3.5.3 Shale composition and pore compressibility 79 3.5.4 Confining pressure 88 3.5.5 Anisotropy 96 3.5.6 Failure envelope 103 3.6 DISCUSSION 108 3.7 CONCLUSION 109 3.8 REFERENCES 110 vi CHAPTER-4 TRANSIENT PRESSURE PULSE DECAY PERMEABILITY OF WOODFORD SHALE 117 4.1 ABSTRACT 117 4.2 INTRODUCTION 119 4.3 GEOLOGY 121 4.4 EXPERIMENTAL TECHNIQUES 124 4.4.1 Sample Preparation 124 4.4.2 Permeability through the Pulse Decay Experiment 124 4.4.3 Procedure 126 4.4.4 Porosity Measurements 128 4.4.5 Mineralogy 129 4.5.6 Total Organic Content (TOC) 130 4.5.7 Scanning Electron Microscopy (SEM) 131 4.5.8 Mercury Porosimetry 131 4.6 RESULTS 133 4.6.1 Permeability vs. Effective Stress 133 4.6.2 Anisotropy 141 4.6.3 Pore Size Distribution 146 4.6.4 Scanning Electron Microscopy 150 4.6.5 Porosity 158 4.6.6 Total Organic Content 159 vii 4.7 DISCUSSION 160 4.8 CONCLUSIONS 161 4.9 REFERENCES 162 CHAPTER-5 CONCLUSIONS 169 5.1 CONCLUDING REMARKS 170 5.2 FUTURE RECOMMENDATIONS 172 5.3 REFERENCES 174 viii LIST OF TABLES Table 2-1. Permeability of WC 1 and WC2 (argillaceous shale) at different effective stresses 30 Table 2-2. Permeability of WC3 and WC4 (siliceous shale) at different effective stresses. 31 Table 2-3. Permeability of WC5 (calcareous shale) at different effective stresses 32 Table 2-4. Permeability of WC6 (calcareous shale) at different effective stresses 33 Table 2-5. Lithologic composition of core samples 34 Table 2-6. Permeabilities measured parallel and normal to the bedding for WC3.2 and WC3.3 samples 36 Table 2-7. Permeability measured parallel and normal to the bedding for the WC6.1, WC6.3, and WC6.4 samples 38 Table 3-1. Test specimen description 71 Table 3-2. Specifications of Triaxial Apparatus 73 Table 3-3. Elastic properties of the triaxial samples tested 87 Table 3-4. Mineralogy of the tested triaxial samples (WS-1, WS-2 and WS-3) 87 Table 3-5. Correlation of elastic constant of Woodford shales at different confining pressures. CP 27* and 35* have microfractures and show a non-linear increase of Young’s modulus with confining pressure 90 Table 3-6. Anisotropic elastic properties of WS-A1 and WS-A2 samples 97 Table 4-1. Permeability of Woodford shale samples from WS- 1 to WS-6 at different effective stresses 139 Table 4-2. Permeability of Woodford shale samples from WS-7 to WS-13 at different effective stresses 140 Table 4-3. Lithologic composition, porosity and TOC of core samples 141 Table 4-4. Permeabilities measured parallel and normal to bedding for samples WS-A1, WS-A2 and WS-A3 142 Table 4-5. Lithologic composition of samples with permeabilities measured parallel and normal to bedding 142 ix LIST OF FIGURES Figure 1-1. Graph shows the North American current gas supply and production supply by 2020 (Elsie, 2008) 1 Figure 2-1. Map showing the Western Canadian Sedimentary Basin (WCSB) from where samples were collected 10 Figure 2-2. Schematic diagram of the modified pulse decay experiment 13 Figure 2-3. Pressure profiles across the sample reservoirs according to the time decay. Hu upstream pressure; Hd= downstream pressure 15 Figure 2-4. Changes in the log time with the increase of confining pressures in WC-5.4. Changes in the confining pressures from 6.89 MPa to 20.68 MPa will result in longer pressure decay times 15 Figure 2-5. Permeability of well WC1, as a function of effective stress 25 Figure 2-6. Permeability of well WC2, as a function of effective stress 26 Figure 2-7. Permeability of well WC3 as a function of effective stress 26 Figure 2-8. Permeability of well WC4, as a function of effective stress 27 Figure 2-9. Permeability of well WC5, as a function of effective stress. Permeability loss even at low effective stresses is clearly visible 27 Figure 2-10. Permeability of well WC6, as a function of effective stress 28 Figure 2-11. Exponential regression of all WCSB samples tested 34 Figure 2-12. Permeability vs. effective stress for the WC3.2 sample; flow measured parallel and normal to bedding 37 Figure 2-13. Permeability vs. effective stress of the WC3.3 sample; flow measured parallel and normal to bedding 37 Figure 2-14. Permeability vs. effective stress of the WC6.1 sample; flow measured parallel and normal to bedding 39 Figure 2-15. Permeability vs. effective stress of the WC6.3 sample; flow measured parallel and normal to bedding 39 Figure 2-16. Permeability vs. effective stress of the WC-6.4 sample; flow measured parallel and normal to bedding 40 Figure 2-17. Clay particle orientation in the argillaceous shale 42 Figure 2-18. Micro fractures are visible in the argillaceous matrix which increase the fluid flow rate 42 Figure 2-19. Fractures filled with authigenic illite and silica 43 x Figure 2-20. Random orientation of clay particles with the increase of silt material in siliceous shale 43 Figure 2-21. Sample depicting edge-to-edge and face-to-face particle contacts with randomly shaped voids. The random orientation of fabric is also seen 44 Figure 2-22. Isolated quartz crystals in the matrix 44 Figure 2-23. Randomly distributed micro pores present in the clay matrix of siliceous shale 45 Figure 2-24. Fabric is enriched with dolomite 45 Figure 2-25. Calcite crystal present in the clay matrix 46 Figure 2-26. Biogenic silica sourced from diatoms present in the siliceous shale 46 Figure 2-27. Micro pores in the matrix of calcareous shale 47 Figure 2-28. High permeability is visible in the argillaceous samples (WC-2), which has high clay content and low quartz content 48 Figure 2-29. Permeability change with the influence of mineralogy in siliceous shale (WC-3). High permeability is visible in the samples, which has average clay content (>30%) and low quartz and carbonate content 49 Figure 2-3 0. Permeability change with the influence of mineralogy in siliceous shale (WC-4). High permeability is visible in the samples, which has average clay content(>30%) and low quartz content 49 Figure 2-31. Permeability change with the influence of mineralogy in calcareous shale (WC-5). High permeability is visible in the samples, which has high clay content and low carbonate content 50 Figure 2-32. Permeability change with the influence of mineralogy in calcareous shale (WC-6). High permeability is visible in the samples, which has high clay content and low carbonate content 50 Figure 2-33. Correlation between TOC and permeability of argillaceous shale (WC-l and WC-2). No significant correlation is apparent 51 Figure 2-34. Correlation between TOC and permeability of siliceous shale (WC-3 and WC-4). A weak correlation is seen 52 Figure 2-35. Weak correlation is visible between TOC and permeability of carbonaceous shale 52 Figure 2-36. Correlation between permeability and porosity of argillaceous (WC-1 and WC-2) and siliceous shale (WC-4). WC-1 and WC-2 show weak correlation between porosity and permeability. WC-4 shows a negative correlation 53 Figure 3-1. Physical and stratigraphic map of the Woodford shale (Johnson and Cardott, 1992) 69 xi Figure 3-2. Schematic diagram of the triaxial machine 74 Figure 3-3. Triaxial arrangement of the sample 75 Figure 3-4. Deviator stress-axial strain plot of all the three triaxial samples tested 81 Figure 3-5. The full stress-strain response of Woodford shale (WS-2) in triaxial compression 82 Figure 3-6. Change of volumetric strain with axial strain of triaxial samples 83 Figure 3-7. Volumetric response with loading. WS-1 has higher clay content and shows more compaction than does WS-2 and WS-3 84 Figure 3-8. Change of porosity and pore size of WS-1(clay rich) sample before and after the triaxial test 85 Figure 3-9. Change of porosity and pore size of WS-2 (quartz-rich) sample before and after the triaxial test 86 Figure 3-10. Shear fractures at 5, 10, 15, 27, and 35 confining pressures for the WS-S1 sample 91 Figure 3-11. Stress-strain plot at different confining pressures 92 Figure 3-12. Volumetric-axial strain at different confining pressures 93 Figure 3-13. Stress-volumetric strain of Woodford shales at different confining pressures 94 Figure 3-14. The onset of Dilatancy C at different confining pressures 95 Figure 3-15. Deviatoric stress vs. axial strain of WS-A1 samples. The strength is greater in the samples that were deformed normal to the bedding, rather than those deformed parallel to the bedding 98 Figure 3-16. Deviatoric stress vs. volumetric strain of the WS-A1 sample. Larger dilatancy is observed in samples that were deformed parallel to the bedding, compared to those deformed normal to the bedding 99 Figure 3-17. Deviatoric stress vs. axial strain of WS-A2 samples 100 Figure 3-18. Deviatoric stress vs. volumetric strain of WS-A2 samples. The bedding plane is more compressed and allows no space for increase in the lateral strain of samples deformed normal to the bedding plane; whereas, in the samples deformed parallel to the bedding, the bedding plane is less compressed and more dilatancy occur 101 Figure 3-19. Samples were tested for anisotropy strength for parallel to bedding and normal to the bedding 102 xii Figure 3-20. Strength envelope of shear and normal stresses of WS-S 1 sample at 5, 10, and 15 MPa confining pressures. Cohesion is 9.2 MPa and friction angel is 55 degrees 105 Figure 3-21. Strength envelope of shear and normal stresses at 27 and 35 MPa confining pressures. Cohesion is 22 MPa and friction angle is 34 degrees 106 Figure 3-22. Strenght envelope of shear and normal stresses of WS-1 sample at 5, 10, 15, 27 and 35 MPa confining pressures. The change in trend line shows a bilinear envelope with a transition at higher confining pressures 107 Figure 4-1. Physical and stratigraphic map of the Woodford shale (Johnson and Cardott, 1992) 123 Figure 4-2. Schematic diagram of the modified pulse decay experiment 126 Figure 4-3. Pressure profiles across the sample reservoirs according to time decay 127 Hu = upstream pressure, Hd= downstream pressure 127 Figure 4-4. Permeability of samples WS- 1 to WS-4 as a function of effective stress. Note the decrease in permeability in relation to the increase in effective stress. ... 135 Figure 4-5. Permeability of samples WS-5 to WS-13 as a function of effective stress.. 137 Figure 4-6. Changes in permeability with quartz content. Note the lower permeability in quartz- rich samples (60-80%) and higher permeability in low quartz (<60%) samples 138 Figure 4-7. Permeability vs. effective stress for the WS-A1 sample; flow measured parallel and normal to fabric. Change in permeability is less at low effective stress and more in high effective stress 143 Figure 4-8. Permeability vs. effective stress for the WS-A2 sample; flow measured parallel and normal to fabric. Anisotropy is observed at low and high effective stresses 144 Figure 4-9. Permeability vs. effective stress for the WS-A3 sample; flow measured parallel and normal to fabric 145 Figure 4-10. Pore size distribution for tested high permeability samples (WS-2, WS 3,WS-9,WS-10 and WS-13). The macro pores of clay rich samples (WS-2, WS-3 and WS-10) and meso pores of quartz samples (WS-9 and WS-13) show higher pore volumes 147 Figure 4-11. Pore size distribution for tested low permeability samples (WS-1, WS-4, WS-5, WS-6, WS-8, WS-12). Meso pores of quartz (WS-1, WS-4, WS-6, WS 8 and WS-12) and macro pores of clay rich sample (WS-5) show higher pore volumes 148 Figure 4-12. Pore size distribution of very low permeability samples (WS-7 and WS-11). Meso pores of quartz rich (WS-7 and WS-11) show higher pore volumes. WS xiii 7 has k 10 md to 106 md at 0.69 MPa to 3.45 MPa, WS-1 1 has k i0 md to i0 md at 10.35 MPa to 12.42 MPa 149 Figure 4-13. Uncompressed Tasmanites in the clay matrix 151 Figure 4-14. Compressed Tasmanite in the clay matrix 152 Figure 4-15. Highly compressed (around six fold) Tasmanite 153 Figure 4-16. SEM micrograph showing the preferred clay orientation of clay rich samples 154 Figure 4-17. SEM micrograph showing the random orientation of clay particles in quartz rich samples 155 Figure 4-18. Microfractures in the Woodford shales 156 Figure 4-19. Development of fractures in the quartz grain of Woodford shales 157 Figure 4-20. Correlation between porosity and permeability of clay-rich Woodford shales 158 Figure 4-21. X-plot of TOC and permeability 159 xiv ACKNOWLEDGEMENTS I would like to thank all people who have guided and inspired me during my Masters thesis.First of all I would like thank my supervisor Dr. R. Marc Bustin for accepting me into the Masters program even though I had a long gap between geology studies. I would not have finished my thesis without his encouragement, inputs, proper guidance, and valuable time and especially to his patience. I am also thankful to my supervisory committee Dr. Erik Eberhardt and Dr. Roger Beckie for their research ideas and guidance. My sincere thanks to lab-mate Dr. Laxmi Chikatamarla, for his assistance in operating the lab instruments and also for the moral support through thick and thin. I am thankful to other lab mates Dr. Daniel Ross, Dr. Gareth Chalmers, and Dr. Amanda M. M. Bustin for their valuable suggestions in writing my thesis. Finally and most importantly, I would like to thank my parents and Santhoshi (my wife) for their unconditional love and support in finishing my thesis. xv Chapter 1 Introduction 1.1 INTRODUCTORY STATEMENTS Natural gas is a clean-burning fuel, found in abundance in U.S and Canada as a mixture of gases in porous rock formations. According to EIA (2007), Natural gas consumption in North America is projected to increase at an average annual rate of 1.0 percent from 2004 to 2030. Most of the additional supply for domestic natural gas production will come from low-permeability reservoirs, such as tight sands, coal-bed methane and shale gas reservoirs. These unconventional gas supplies will play an important role in North American natural gas demand, making up one half (48%) of its production rate by 2020 (Figure 1) (Elsie, 2008) Figure 1-1 has been removed due to copyright restrictions. This figure shows the increase of Unconventional gas demand from 2007 to 2020 (Elsie, 2008). Figure 1-1. Graph shows the North American current gas supply and production supply by 2020 (Elsie, 2008) 1 Shales can be a source, reservoir and trap for gas. Gas shales are unconventional resources with continuous type gas accumulations (Schmoker, 1996). Total gas shale resources in the United States have been estimated in the range of 500-600 tcf, whereas in Canada it is around 1000 tcf (Bustin, 2005). Gas in shales is stored in the adsorbed state in the micropores of organic matter and as free gas in the pores of the rock matrix and in solution. Permeability is the most important rock property affecting gas shale production. Most of the gas is stored in the matrix of the shale, which have permeabilities on the order of 1 0 md (Dewhurst et. al., 1998; Kwon et al., 2004). In this investigation, matrix permeability was measured using pressure pulse decay experiment on Western Canadian and Woodford shales of Oklahoma. The pulse decay experiment was employed with triaxial experiments combined with mercury porosimetry, helium pycnometry, Rock-Eval pyrolysis, SEM and X-ray diffraction analysis to measure rock strength, pore size, porosity, total organic content, fabric and composition of samples. The permeability results were correlated with effective stress, anisotropy, fabric, rock strength, porosity, pore size and total organic content. 2 1.2 STRUCTURE OF THESIS This thesis is written in a manuscript format. Unpublished individual papers concentrated on the shales of The Western Canadian Sedimentary Basin of Canada and Woodford shale of Oklahoma, USA. These are in 3 main chapters: Chapter- 2 discusses the factors that affect the permeability of shales of the Western Canadian sedimentary basin. All factors were studied on samples of contrasting lithologies. Chapter-3 emphasises the importance of rock mechanical properties which are useful in reservoir production. The functionality of triaxial undrained experiment is explained in detail. Chapter-4 investigates the factors affecting the permeability of Woodford shales of Oklahoma, US. Conclusions and recommendations for further research are then presented. 3 1.3 REFERENCES Elsie, Ross. 2008. Study sees Unconventional gas driving North American supply growth, Oil Daily Bulletin, September, 2008. ETA (Energy Information Administration). 2007. Annual energy review 2007: U.S. Department of Energy, Energy Information Administration report DOE/ETA 0484(2007), May-2007, p. 230: http://www.eia.doe/gov/oiaf/ieo/index.html Schmoker, J. W., 1996, Method for assessing continous-type (unconventional) hydrocarbon accumulations, in D. L. Gautier, G. L. Dolton, K. I. Takahashi, and K. L. Varnes, eds., 1995 National assessment of United States oil and gas resources — Results, methodology, and supporting data: U.S. Geological Survey Digital Data Series DDS-30, Release 2, [CD-ROM]. Bustin, R.M. 2005. Gas Shales Trapped for Big Play: AAPG Explorer, February. Dewhurst, D.N., Aplin, A.C., Sarda, J.-P. and Yang, Y. 1998. Compaction-driven evolution of porosity and permeability in natural mudstones: An experimental study. Journal of Geophysical Research, v. 103, p. 651—661. Kwon, 0., Kronenberg, A.K., Gangi, A.F., Johnson, B., and Herbert, B.E. 2004. Permeability of Illite-bearing shale: 1. Anisotropy and effects of clay content and loading, J.Geophys. Res., B 10205, doi: 10.1 029/2004JB003052. 4 CHAPTER 2 Factors Affecting the Permeability of Shales in the Western Canadian Sedimentary Basin (WCSB) 5 Chapter 2 Factors Affecting the Permeability of Shales in the Western Canadian Sedimentary Basin (WCSB) 2.1 ABSTRACT The matrix permeability (k) of major shale intervals in the Western Canadian Sedimentary Basin (WCSB) was investigated to assess their gas shale potential. Potential gas shales in the WCSB range from comparable shallow and soft water-sensitive shales in the east to relatively deep, brittle and silica-rich shales in the deeper parts of the basin to the south west. We measured the permeability and stress sensitivity of the permeability using the transient pulse decay method to resolve the correlation between effective pressure, mineralogy, porosity, total organic content (TOC), fabric and permeability anisotropy. The permeability of all shales declines exponentially with increasing effective stress, though the amount of stress-sensitivity depends on mineralogy and fabric. In clay-rich shales (WC-2.3), the permeability (k) decreases from 3 .29E-03 md to 1.1 7E-04 md as the effective stress increases from 3.45 MPa to 17.24 MPa. In biogenic quartz-rich shales (WC 3.3), k decreases from 8.85E-02 md to 1 .92E-02 md with the effective stress increasing from 6.9 MPa to 27.59 IviPa; and in calcareous shales (WC-6.3), k decreases from 3.03E- 05 to 9.16E-07 with effective stress increasing from 3.45 MPa to 27.59 MPa. Clay mineralogy plays an important role in determining permeability, which is dependent on the 6 porosity, fabric, TOC, and mechanical properties of the strata. In argillaceous shales (WC 2), high quartz content (> 26%) samples show low permeability. Biogenic quartz-rich shales(WC-3) have a comparatively low permeability in high quartz (>46%) and carbonate rich (>3%) samples and high permeability in average clay content (>30%) samples. Calcite-rich calcareous shales (WC-6) have a very low permeability in high carbonate (>45%) samples and high permeability in high clay (>12%) samples. Among the shales, porosity is moderately correlated with permeability in argillaceous shale, but there is no correlation between porosity and permeability in calcareous shale. Surprisingly, siliceous shale shows a negative correlation between porosity and permeability. TOC is correlated with the calcareous shales, though not with argillaceous or siliceous shales. Quartz-rich shales show three to four orders of magnitude difference whereas calcite-rich shales have two to three orders of magnitude difference in permeability between the samples tested parallel to bedding and normal to bedding. Overall permeability anisotropy is more marked for the quartz-rich shales. 7 2.2 INTRODUCTION Fluid transport properties of rocks at both laboratory and field scales are critical for evaluating the production potential of gas shale reservoirs. Permeability is the most important rock fluid property governing and affecting the fluid flow in the subsurface, and thus, production (Marek, 1979; Simon et al., 1982). Generally, shales have storage mainly in the matrix which has permeability in the order of 1023 m2 (1O md) (Bredehoeft et al., 1988; Brace, 1980; Morrow et a!., 1984; Mudford and Best, 1989; Dewhurst et al., 1998; Kwon et a!., 2004) and the matrix permeability may be production rate limiting. Apart from the study of reservoir seals, there is little data available on permeability of gas through shale. This lack of data leads to significant uncertainties as to the productive capacity of prospecting shales. In shales, the production rate is generally considered to be governed by the Darcy flow rate in fractures and by diffusion and Darcy flow through the matrix (Bustin et al., 2007). The flux (diffusion and advection) through the matrix of gas shales can be measured in the laboratory using pulse decay experiments. The gas flows through the matrix are stress dependent and in many shale reservoirs that have wide fracture spacing, the production may be limited by matrix flow rates (Bustin, 2007). The importance of permeability in hydrocarbon accumulation has led geologists to investigate the controlling factors. Permeability depends on the effective stress (Zoback and Byerlee, 1975; Bustin, 1997; Kwon et al., 2001; Bell, 2006; Yang and Aplin, 2007), 8 anisotropy (Bolton et al., 2000; Meyer and Krause, 2006; Kwon et al., 2004; Scholes et a!., 2007; Clennell et al., 1999; Horsrund et al., 1998; Meyer, 2002), porosity (Katsube et a!., 1991; Bloch, 1991; Davies et al., 1991; Revil and Cathles, 1999; Kwon et al., 2004; Yang and Aplin, 2007), mineralogical composition (Marion et a!., 1989; Ahmed et a!., 1991; Bustin, 1997; Howard, 1992) and microfabric (O’Brien, 1970; O’Brien and Slatt, 1990). By understanding the relation between effective stress, shale fabric, composition, and porosity in production, more effective exploration and exploitation of potential gas shales may be possible (Bustin et a!., 2007). The primary objective of this study is to investigate the matrix permeability of shales of the Western Canadian Sedimentary Basin (WCSB) using pulse decay experiments. The interrelationship of effective pressure, fabric, anisotropy, porosity, TOC, composition with permeability of shale is also investigated. 9 2.3 STUDY AREA AND DATA Core samples were investigated from shales of contrasting lithologies taken from different locations in the WCSB (Figure 2-1). Figure 2-1. Map showing the Western Canadian Sedimentary Basin (WCSB) from where samples were collected. Samples that were collected from the wells of the WCSB are here labelled WC1 to WC6 to preserve location confidentiality. WC1 and WC2 are rich in clay (>45% clay content), WC3 and WC4 are rich in silica (>45% Quartz), and WC5 and WC6 are rich in calcite (>45% carbonate content). Samples were analysed to determine: 1) permeability vs. 10 effective stress; 2) anisotropy or directional permeability; 3) mineralogy; 4) porosity; 5) TOC; and 6) fabric (examined by scanning electron microscope, SEM). 11 2.4. EXPERIMENTAL PROCEDURES 2.4.1 Sample Preparation Shale samples 3 cm diameter were re-cored from full diameter preserved cores with original diameters of approximately 8.9 cm. Samples were cored using a diamond drill bit and air as a coolant to avoid reactions between the samples and water. 2.4.2 Permeability through Pulse Decay Experiment The pulse decay apparatus was constructed based on the design of Jones (1997) (Figure 2). The core holder is a RocTest ® Hoek cell, which consists of a hollow steel cylinder with threaded removable caps. A urethane rubber membrane, which is present inside the cell, acts as a pressurization chamber for hydraulic fluid. The Hoek cell can maintain confining pressures up to 69 MPa with hydraulic oil. The core holder (Figure 2-2) is connected to an upstream reservoir of volume, Vi, and the downstream reservoir of volume, V2. At the onset of experiment the sample is confined and the volumes of the upstream and downstream vessels are taken slightly higher than the sample pore volume, Vp (Jones, 1997). A differential pressure transducer measures the difference between the two reservoirs with two absolute transducers present at the upstream and the downstream reservoirs to 12 measure the pressures. Helium gas is used to supply the pore pressure through the sample end plates. A computer-based data acquisition system is used to record all physical changes that are monitored and recorded at specified time intervals. Figure 2-2. Schematic diagram of the modified pulse decay experiment. 13 2.4.3 Procedure The sample is assembled in the Hoek cell between the two pistons. The confining pressure, which is calculated from the reservoir effective stress, is applied to the Hoek cell using a hydraulic pump. With valves at 1, 2, 3 and 4 open (see Figure 2-2), helium gas flows through the reservoirs and the sample at a pressure of between 6.9 MPa and 10.3 MPa. The higher mean pore pressures reduce the gas slippage effect (Dicker, et al., 1988; Jones, 1997). After gas fills the system, the fill valve is closed and the system allowed to attain equilibrium pressure. Once all pressures are at equilibrium, the differential transducer is set to zero. Valve 2 is then closed, and the shutoff valve is slowly opened (creating a pressure increase in the upstream reservoir) until the differential transducer reaches less than 0.1 MPa. The shutoff valve and valve 4 are closed, and the following pressures are monitored: tP (differential pressure), P1 (upstream pressure), and P2 (downstream pressure). The pressure at the upstream end decreases slowly while the downstream pressure increases with time (Figure 2-3). The pore pressure decay (or increase) depends on fluid viscosity, sample pore volume, and sample and fluid compressibility (Brace et al., 1968; Em, 1977; Hsieh et al., 1981; Trimmer 1982; Jones, 1997). Measurements are nonnally terminated after a 50% decay of the upstream pressure. Permeability is measured at different confining pressures. As the confining pressure increases, the time of decay increases and the permeability of the sample decreases (Figure 2-4). 14 Figure 2-3. Pressure profiles across the sample reservoirs according to the time decay. Hu = upstream pressure; Hd= downstream pressure. 100 101 102 Log (time) Figure 2-4. Changes in the log time with the increase of confining pressures in WC-5.4. Changes in the confining pressures from 6.89 MPa to 20.68 MPa will result in longer pressure decay times. 15 Upper Reservoir time Sample Down Reservoir Hd time 440 438 C 436 E 434 432 430 428 426 424 CP6.89 MPa CP=1O.34 MPa CP=13.79 MPa CPI 7.24 1Pa CP=20.68 MPa 1 1 2.4.4 Theory 2.4.4.1 Boundary Conditions Brace (1968) presented the pulse technique and Trimmer et al. (1980) developed the theory by showing the importance of compressive storage in porous rocks. The necessity of compressive storage and sample boundary conditions were considered by Hsieh et al. (1981). Subsequently, Dicker and Smith (1988) modified the boundary equations of Heish et al. (1981) for use in the oil industry. The boundary conditions of the pulse decay experiment are as follows (Dicker and Smith, 1988): Fluid pressure, before the increase of incremental pressure, AP, is uniform throughout the sample and equal to the downstream pressure. P(x,0)P2(0 for0<x<L (1) The initial boundary conditions of the upstream and downstream reservoirs are as follows: P(0,t)P1(t fort0 (2) P (L, t) = P2 (t) for t 0 (3) 16 The differential equation across the sample, P (x, t) that obeys Darcy’s Law and the conservation of mass is: 82P(xt) cpçbãP(x,t) for 0< x <Land t> 0 (4) 8x2 k ôt With the increase of pressure in the upstream reservoir, the boundary conditions in both the reservoirs change to (5) and (6) and express the conservation of mass at the sample faces (fluid flow in the sample is equal to the pressure increase at the downstream reservoir). dPi k V8ÔP = ____— fort>0 (5) dt cpq5Li 8X x=0 dP2 = — k for t> 0 (6) di’ c1uq5Lv2 ax x=l Where P = pressure, x = distance from the upstream end of the sample, P1 = upstream reservoir pressure, P2 = downstream reservoir pressure, 17 t = the time from the increment of upstream pressure, k = sample permeability, L = length of the sample, c = fluid (gas) compressibility, = fluid (helium) viscosity, and 0 = sample porosity. Solutions for the pressure in the upstream and downstream reservoirs contain a dimensionless variable: tD (dimensionless time): tD = kt cpL a = (ratios of compressive storage of sample’s pore volume to the upstream reservoir) b = (ratios of compressive storage of sample’s pore volume to the downstream reservoir) Several researchers have commented about the importance of the size of the downstream and upstream reservoirs. Dicker and Smith (1988) emphasized using identical and larger values of a and b such that the time decay is rapid. The values of a and b should be equal to the pore volume. Jones (1997), however, suggested using reservoirs 2 to 10 times 18 larger than the pore volume of the sample. The present pulse decay experiments were conducted based on the recommendation of Jones (1997). 2.4.4.2 Calculation of Permeability Gas permeability is calculated using the following equation (Jones, 1997): kg= —l4696mipgLfz (7) 1 1 fiApm v12 Where ml = slope of the linear equation, Pm mean absolute pressure, tg = viscosity of the gas [viscosity of the helium gas was taken from (API RP4O)], L = length of the core plug, fi mass flow correction factor, A = cross-sectional area of cylindrical core plug, Vi = volume in small upstream reservoir, V2 = volume in small downstream reservoir, and = Gas compressibility correction factor [used for the deviation from ideal gas behaviour for helium at various pressures]. 19 2.4.5 Porosity Measurements Porosity is the storage capacity of the rock, and the storage capacity of petroleum reservoir fluids depends on the porosity of the rock. Reservoir porosity (i) is expressed by the following relationship: — pore volume — bulk volume (8) The porosity of the rocks can be determined by routine core analysis (lab measurements) and well logging techniques. Between these two methods, routine core analysis is probably the most commonly used for determining porosity. Pore volume and the bulk volume of the rock are measured to determine the porosity of the rock, based on Equation (8). The bulk volume of the sample is measured with a mercury-based Archimedes principle that uses observations of the volume of the fluid displaced by a sample. Pore volume can be measured using mercury porosimetry and helium pycnometry. All well samples are tested by helium pycnometry, with 20 to 30 g of powdered sample required for the test. The method uses Boyle’s Law. The apparatus consists of two equal chambers (or cells), a reference chamber and a sample chamber. The reference chamber has a volume of Vi at initial pressure of P1 and the sample chamber has an unknown volume V2 and an initial pressure P2. The unknown sample volume is calculated after opening 20 the sample volume when it is at equilibrium. By knowing both the bulk volume and the pore volume, the porosity of the rock can be determined. 2.4.6 Mineralogy X-ray powder diffraction is the best available method for identifying minerals in shale samples. The Rietveld method is the most acceptable quantitative phase analysis (Rietveld, 1967, 1969). However, due to platy nature of clay crystallites and disorder that create a preferred orientation, the quantitative analysis of clays is not accurate (Srodon, 2001). Samples were prepared using procedures described by Raudsepp and Pani (2003). Samples were crushed in a mortar to pass through a 0.4 mm sieve. Samples were grounded in ethanol using a vibratory McCrone ® Micronising Mill with corundum elements, for around 7 mm. The grinding with ethanol, instead of water, was preferred to avoid swelling of the clays. After grinding, the sample was x-rayed using a Siemens D5000® diffractometer, equipped with a theta-theta goniometer. The initial phase recognition was done with EVA® software, using conventional search-match procedures applied to the database. The data was entered into the Rietveld program and the mineralogy was quantified using TOPAS® 3.0 software. 21 2.4.7 Total Organic Content (TOC) The TOC of the shale samples was measured using Rock-Eval pyrolysis. Samples were ground to a 100 mg powder before the pyrolysis was run, and samples were then weighed accurately into metal crucibles before being subjected to programmed heating in an inert atmosphere (helium) to determine the free hydrocarbons contained in the samples. Hydrocarbon- and oxygen-containing compounds were volatilized during the cracking of the organic matter in the samples (Tissot and Welte, 1984). The total organic carbon content is calculated with the following equation: K(S1 + S2) S4 TOC= 10 10 (9) Where K Elemental composition of natural substances, S1+S2 = Total potential (Amount of free hydrocarbons and the amount of hydrocarbons generated through thermal cracking of organic matter, and S4 = Quantity of CO2 produced during oxidation of the residual organic carbon. 22 2.4.8 Scanning Electron Microscopy (SEM) Samples were prepared with procedures described by Bohor and Hughes (1971). Samples were fractured to reveal freshly exposed fracture surfaces. The fractured rock chips were then glued to the aluminium stub and gold-coated to a thickness of about 5-10 nm by evaporation of gold under high vacuum. SEM and BSEM observations were carried out using a Philips XL-30 SEM with a Princeton® Gamma-Tech PRISMIG energy-dispersive spectrometer at 15 kV and images were taken with a 10 mm working distance. 23 2.5 EXPERIMENTAL RESULTS 2.5.1 Permeability-Effective Stress Measurements The experimental results for permeabilities of WCSB shale samples, from wells WC1 to WC6 (argillaceous shale, siliceous shale, and calcareous shale), are listed in Table 2-1 to 2-4. Shales from wells WC1 and WC2 are clay-rich (>45% clay content); WC3 and WC4 are quartz-rich (>45% quartz content); and WC5 and WC6 are calcite-rich (>45% carbonate content). All permeabilities measured during the experiment are relative to gas present in the shale. In Figures 2-5 to 2- 10, log permeabilities are plotted against effective stress for the analysed samples. In WC1, WC2, and WC4 permeability loss mostly occurred in the low effective stress range. Permeability decreases with increasing effective stress in the initial stage as a result of the closure of the aspect ratio pores at the lower effective stresses. The permeability loss that occurred among all shales was at effective stresses within the range of 1 .OE-00 to 9.6E-07 md. In clay-rich shales (WC-2.3), permeability (k) decreases from 3.29E-03 md to 1.17E-04 md, as the effective stress increases from 3.45 MPa to 17.24 MPa. In quartz-rich shales (WC-3.3), k decreased from 8.85E-02 md to 1.92E-02 md with the effective stress increasing from 6.9 MPa to 27.59 MPa and in calcareous shales (WC-6.3), k decreases from 3.03E-05 to 9.16E-07 md with the effective stress increasing from 3.45 MPa to 27.59 MPa. A drastic change in permeability with the effective stress is not prominent in the argillaceous and siliceous shale, though a big difference (two order 24 decrease) in permeability loss is seen in calcareous shale. The permeability loss with effective stress is higher in calcareous shales than in other shales (Figure 2-10). Samples from well WC4 (Figure 8) had natural fractures and, hence prior to running the permeability test, samples were kept under a high confining pressure for 10 hrs to close the fractures. Higher confining pressures were applied if the permeability values could not be generated because of the uncompressible fractures. 100 -110 I i02. I WC-1.1 —regression • WC-1.2 regresslon • WC-1.3 regression 205 10 1ffective stresc MPi 15 Figure 2-5. Permeability of well WC1, as a function of effective stress. 25 - t CD C CD C) cn C) -4 C C T j GQ C -t CD t; J N C CD C) L’J (ID C) -4 C C I lo g pe rm ea bi lit y, m d 0 (I’ -L 01 t) 0 IJ ON • WC-4.1 regresslon * WC-4.2 regresslon * WC-4.3 —regression V WC-4.4 —regression 10 10 — 10 i02 10 15 20 25 30 35 40 45 Effective stress, MPa Figure 2-8. Permeability of well WC4, as a function of effective stress. WC-5.1 regresslon 10_i • WC-5.Z —regression 4 WC-&3 • regres5ionE -2 • WC-5.410 regression — • WC-5.5 l -3 regresslon 10 I 1o4 10-s 5 10 15 20 25 Effective stress, MPa Figure 2-9. Permeability of well WC5, as a function of effective stress. Permeability loss even at low effective stresses is clearly visible. 27 r j p I Table 2-1. Permeability of WC 1 and WC2 (argillaceous shale) at different effective stresses. Core Effective Stress Permeability sample MPa md 3.77E-01 2.18E-0 1 1.78E-01 1.24E-01 9.20E-02 5 .74E-02 2.74E-0 1 1.65E-01 1.30E-01 1.03E-01 6.79E-02 4.5 1E-02 2.33E-0 1 1.81 E-0 1 1.52E-01 1.19E-01 8.13E-02 4.90E-02 2.61E-02 8 .24E-03 3 .04E-03 1 .39E-03 6.06E-0 1 3.40E-01 2.90E-0 1 1.83E-01 1.23E-01 8.81E-02 3.29E-03 1 .99E-03 7.OOE-04 3 .60E-04 1.17E-04 WC1 wc1.1 WC1.2 WC1.3 WC2 WC2.1 WC2.2 WC2.3 5.24 7.31 9.38 11.45 13.52 15.59 3.71 5.78 7.85 9.92 11.99 14.06 3.88 5.95 8.02 10.09 12.16 14.23 3.45 6.90 10.34 13.79 3.45 6.90 10.34 13.79 17.24 20.69 3.45 6.90 10.34 13.79 17.24 29 Table 2-2. Permeability of WC3 and WC4 (siliceous shale) at different effective stresses. Core Effective Stress Permeability sample MPa md WC3 WC3.1 WC3.2 WC3.3 WC4 WC4.1 WC4.2 WC4.3 WC4.4 17.93 21.37 24.82 28.27 14.48 17.93 21.37 24.82 28.27 6.90 10.34 13.79 17.24 20.69 24.14 27.59 6.89 13.79 20.69 27.58 34.48 41.38 20.69 27.58 34.48 41.38 13.79 20.69 27.58 34.48 41.38 6.89 13.79 20.69 27.58 34.48 41.38 4.79E-03 4.47 F -03 4.OOE-03 3.28E-03 7.81E-04 5.32E-04 3.62E-04 2.1 8E-04 1 .20E-04 8.85E-02 6.74E-02 5.62E-02 4.42E-02 3.24E-02 2.87E-02 1 .92E-02 1 .90E+00 1 .45E+00 1 .08E+00 8.52E-01 6.76E-01 5.68E-01 1.14E+00 1.O1E+00 7.96E-01 6.12E-01 1.16E-01 4.82E-02 2.64E-02 1 .48E-02 8.56E-03 7.66E-01 5.35E-01 3.89F-01 2.38E-01 1.51E-01 1.13E-01 30 Table 2-3. Permeability of WC5 (calcareous shale) at different effective stresses. Core Effective Stress Permeability sample 1’1Pa md WC5 WC5 .1 WC52 WC5.3 WC5.4 WC5.5 6.9 10.34 13.79 17.24 20.69 6.9 10.34 13.79 1724 20.69 6.9 1034 13.79 17.24 20.69 6.9 1034 13.79 1724 20.69 6.9 1034 13.79 1724 20.69 3 .20E-05 2.02E-05 1 .28E-05 8.09E-06 5. 12E-06 4.02E-0 1 1.36E-01 6.60E-02 3 .70E-02 1 .91E-02 2.68E-04 1 .26E-04 5.97E-05 2.82E-05 1 .33E-05 2.15E-03 1 .47E-03 1 .O1E-03 6.88E-04 4.70E-04 2.64E-04 2.47E-04 2.3 1E-04 2. 17E-04 2.03E-04 31 Table 2-4. Permeability of WC6 (calcareous shale) at different effective stresses. Core Effective Stress Permeability sample MPa md WC6 WC6.1 WC6.2 WC6.3 WC6.4 3.45 6.90 10.34 13.79 17.24 20.69 24.14 27.59 3.45 6.90 10.34 13.79 17.24 20.69 24.14 27.59 3.45 6.90 10.34 13.79 1724 20.69 24.14 27.59 3.45 6.90 10.34 13.79 1724 20.69 24.14 27.59 2.22E-04 1 .65E-04 I .22E-04 9.04E-05 6.69E-05 4.96E-05 3 .67E-05 2.72E-05 1.60E-01 9.22E-02 5.32E-02 3 .07E-02 1.77E-02 1.02E-02 5 .89E-03 3.40E-03 3 .03E-05 1 .84E-05 1.12E-05 6.77E-06 4.1OE-06 2.49E-06 1.51E-06 9. 16E-07 7.68E-02 5.14E-02 3 .45E-02 2.3 1E-02 1.55E-02 1 .04E-02 6.96E-03 4.67E-03 32 Ta bl e 2- 5. Li th ol og ic co m po sit io n o fc o re sa m pl es k TO C 0 Qu ar tz Py rit e C al ci te D ol om ite A lb ite Ill ite K ao lin ite c hl or ite G yp su m To ta l-C la y T ot al -C ar b sa m pl es (w t% )( wt %) (% ) (% ) (% ) (% ) (% ) (% ) (% ) (% ) (% ) (% ) (% ) W C1 .1 1.2 17 .8 34 .8 2.1 6.5 1.9 2. 2 40 .1 8.3 2. 8 1.5 51 .1 8. 4 W C1 .2 2.1 17 ,2 35 .4 2.1 6.5 1.0 1.1 43 .5 6. 2 2.1 2.1 51 .8 7. 5 W C1 .3 3.1 14 .6 37 .7 2. 7 8.9 1.5 2. 0 36 .3 7. 8 1.1 2.1 45 .2 10 .3 W C2 .1 1.1 2. 0 46 .0 1.0 0. 2 0.1 2. 5 31 .6 9. 8 6. 7 2. 0 48 .2 0.3 W C 2. 2 11 .4 3.5 26 .2 5. 7 0, 4 1.7 1.5 34 .8 13 .7 11 .5 4. 7 59 .9 2. 0 W C2 .3 1.4 1.5 31 .2 1.5 0.1 1.3 2. 7 38 .1 11 .1 11 .4 2. 7 60 .5 1.4 W C3 .1 2. 0 2. 4 58 .4 2. 5 1.7 1.7 1.4 31 .5 0. 2 0.5 2.1 32 .3 3. 4 W C3 .2 1.9 3. 4 50 .5 3.1 8. 7 3. 6 3. 9 22 .8 4. 9 1.4 1.1 29 .1 12 .3 W C3 .3 3.1 3.8 46 .5 4. 0 2. 8 2. 0 6. 2 27 .1 9. 2 0. 9 1.3 37 .2 4. 8 W C4 .1 2.3 2. 4 54 .5 2. 5 0. 0 2.1 2. 2 34 .3 1.2 2. 5 0. 7 38 .0 2.1 W C4 .2 3.3 1.9 49 .0 4. 6 0. 0 1.1 3.4 39 .9 0. 8 0. 9 0.3 41 .6 1.1 W C4 .3 2. 4 5.8 81 .8 0.8 0. 0 0. 7 2. 2 13 .1 0.1 0. 7 0. 6 13 .9 0. 7 W C4 .4 1.1 3.1 39 .9 2. 0 0.3 12 .8 2. 9 37 .7 1.1 0. 9 0.5 39 .7 13 .1 W C5 .1 3.3 5.5 8. 2 0. 9 85 .0 3. 6 0. 0 1.7 0. 2 0. 0 0.5 1.9 88 .6 W C5 .2 14 .1 4.1 12 .8 3.1 63 .1 8.8 0. 0 3. 4 8.6 0. 0 0. 2 12 .0 71 .9 W C5 .3 11 .5 4.1 48 .2 3. 0 18 .7 19 .4 0. 7 5. 6 0. 6 1.0 2. 8 7.2 38 .1 W C 5. 4 12 .1 16 .3 9. 7 1.6 76 .6 6.1 0.5 2. 2 1.0 0. 7 1.6 3.9 82 .8 W C5 .5 1.4 19 .6 7.1 0. 2 62 .1 27 .1 0.5 1.3 0. 9 0. 0 0. 9 2.2 89 .2 W C6 .1 0.6 12 .7 9. 4 3. 6 62 .7 2. 4 0. 0 8.8 6.3 2. 7 4. 3 17 .8 65 .0 W C6 .2 5.4 12 .6 42 .4 1.8 36 .6 1.3 0. 0 1.3 6. 7 9. 9 0. 0 17 .9 38 .0 W C6 .3 1.5 6. 4 17 .4 1.1 54 .8 12 .7 0. 7 1.0 7. 6 4. 0 0. 0 12 .6 67 .5 W C6 .4 5.0 7. 7 17 .9 1.4 39 .6 5. 9 1.9 19 .3 11 .1 2. 9 0. 0 33 .4 45 .5 33 The relationship between permeability and effective stress has been investigated by various researchers. An exponential relationship (Wyble, 1958; Pedrosa, 1986; Best and Katsube, 1995), power-law (Jones and Owens, 1986; Ostensen, 1986), and logarithmic relationships are most widely used to fit the permeability data. In the shale samples from wells WC1 to WC6, the permeability reduction can be approximated by an exponential function: K = k0 exp( —aPe) (10) Where k0 is the permeability at atmospheric pressure; Pe is the effective stress; and a is the slope of the curve. The normalized permeability values (klko) of all the samples with the effective stress, had trends that follow the exponential form (shown in Figure 2-1 1). * we-i WC-2 I. * WC-3 WC-4 A IWC-5 AWC6 a * I * A I.* 0.4 ‘ •l * 4 A * 4.. 02 ‘ I *• 4 V I• q o • 4., , . 0 5 10 15 20 25 30 35 40 45 Effective stress, MPa Figure 2-11. Exponential regression of all WCSB samples tested. 34 2.5.2 Permeability Anisotropy Permeabilities were measured parallel and normal to the bedding at different effective stresses on the siliceous (WC3.2 and WC3.3) (Table 2-6) and the calcareous (WC6.1, WC6.3, and WC6.4) (Table 2-7) shale samples. The results of permeability testing are shown in Figures 2-12 to 2-16. In the siliceous shales, permeabilities measured parallel to the bedding are 3 to 4 orders of magnitude greater than those measured normal to the bedding. The anisotropy occurs at the lowest effective stress and increases with increasing effective stresses. Within the calcareous shales, the permeability measured parallel to the bedding are about 2 to 3 orders of magnitude higher, compared with those measured normal to the bedding. In calcareous shale, at an effective stress of 3.45 MPa, the permeability is isotropic, but with increasing stress, the anisotropy increases. 35 Table 2-6. Permeabilities measured parallel and normal to the bedding for WC3.2 and WC3.3 samples. Parallel to Normal to Effective bedding bedding Sample pressure (k) (k) MPa md md 3.45 3.71E-03 2.19E-05 WC-3.2 5.52 2.81E-03 1.30E-05 6.90 2.33E-03 8.79E-06 8.97 1.76E-03 5.22E-06 3.45 1.13E-01 1.66E-05 WC-3.3 5.52 9.76E-02 1.55E-05 6.90 8.85E-02 1.46E-05 8.97 7.65E-02 1.37E-05 WC3.3 (Figure 2-13) and WC6.4 (Figure 2-16) have more clay preferred orientation and show a greater anisotropy than do the other samples. 36 • WC-3.2(normal to bedding) —regression WC-3.2(parallel to bedding) — regression I I I I I I i0 3 456789 10 Effective stress, MPa Figure 2-12. Permeability vs. effective stress for the WC3.2 sample; flow measured parallel and normal to bedding. 100 A WC-33(norrnal to bedding) —regression .WC-3.3(peraIlel to bedding) —regression i10.1 ,. - . 10 -4 —10 A 3 4 5 6 7 8 9 Effective stress, MPa Figure 2-13. Permeability vs. effective stress of the WC3.3 sample; flow measured parallel and normal to bedding. 37 Table 2-7. Permeability measured parallel and normal to the bedding for the WC6.1, WC6.3, and WC6.4 samples. Parallel to Normal to Effective bedding bedding Sample pressure (k) (k) MPa md md WC 6.1 WC 6.3 WC 6.4 3.45 6.90 10.34 13.79 17.24 20.69 24.14 27.59 3.45 6.90 10.34 13.79 17.24 20.69 24.14 27.59 3.45 6.90 10.34 13.79 17.24 20.69 24.14 27.59 2.22E-04 1 .65E-04 1 .22E-04 9.04E-05 6.69E-05 4.96E-05 3.67E-05 2.72E-05 2.22E-04 1.65E-04 1.22E-04 9.04E-05 6.69E-05 4.96E-05 3.67E-05 2.72E-05 7.68E-02 5.14E-02 3.45E-02 2.31E-02 1.55E-02 1.04E-02 6.96E-03 4.67E-03 1.42E-04 6.69E-05 3.16E-05 1 .49E-05 7.06E-06 3.33E-06 1.57E-06 7.44E-07 3.03E-05 1 .84E-05 1.12E-05 6.77E-06 4.1OE-06 2.49E-06 1.51E-06 9.16E-07 5.46E-04 2.71E-04 1 .35E-04 6.69E-05 3.32E-05 1.65E-05 8.19E-06 4.07E-06 38 • WC-6J(parallel to bedding) —regression * WC-.1(normaI to bedding) —regression] i0-5 106 0 5 10 15 20 25 30 Effective stress, MPa Figure 2-14. Permeability vs. effective stress of the WC6.1 sample; flow measured parallel and normal to bedding. • WC-6.3(parallel to bedding) —regression * WC-6.3(normal to bedding) —regression] 10 5 10 15 20 25 30 Effective stress, MPa Figure 2-15. Permeability vs. effective stress of the WC6.3 sample; flow measured parallel and normal to bedding. 39 • WC-6.4(parallel to bedding) — regression * WC-6.4(normal to bedding) — regressionj I 15 20 2 30 Effective stress,MPa Figure 2-16. Permeability vs. effective stress of the WC-6.4 sample; flow measured parallel and normal to bedding. 2.5.3 SEM Observations The fabric of the shale samples was examined using a scanning electron microscope (SEM). The mineralogy in SEM was identified using EDS (energy dispersive spectrometer). In the clay-rich shales (WC1 and WC2) (Figure 2-17), clay particles had a strong preferred orientation. Fabric was delineated by illite, kaolinite, and chlorite minerals. Silt-sized grains were absent in these samples. Permeability is controlled by micro fractures and fissures in the high permeability shales. Fractures were visible in the WC1 samples (Figure 2-18), which also had higher 40 permeabilities when compared with another clay-rich well (WC2). Fracture filled clay minerals, such as illite were visible in these samples (Figure 2-19). In the quartz-rich samples, the clays have a near random particle orientation (Figure 2- 20). The clay has edge-to-edge and face-to-face particle contacts in the quartz rich samples (Figure 2-2 1). Quartz minerals are present in the fabric isolated from the clay matrix (Figure 2-22) and are generally difficult to identify with SEM as they are masked by argillaceous material. The fluid flow mostly occurs through the micro pores that are present in the clay matrix of the quartz-rich siliceous shale (Ross, 2007) (Figure 2-23). The calcareous shales have less fabric due to the presence of authigenic calcite and dolomite minerals that are visible in the illite matrix (Figures 2-24 and 2-25). The biogenic silica is recrystallized diatoms as remnants are recognized in the matrix (Figure 2-26). The pores of the calcareous shale (WC6.3) that have a very low permeability can be seen in Figure 2-27). 41 Figure 2-18. Micro fractures are visible in the argillaceous matrix which increase the fluid flow rate. Figure 2-17. Clay particle orientation in the argillaceous shale. 42 Figure 2-20. Random orientation of clay particles with the increase of silt material in siliceous shale. Figure 2-19. Fractures filled with authigenic illite and silica. 43 Figure 2-21. Sample depicting edge-to-edge and face-to-face particle contacts with randomly shaped voids. The random orientation of fabric is also seen. Figure 2-22. Isolated quartz crystals in the matrix. 44 Figure 2-23. Randomly distributed micro pores present in the clay matrix of siliceous shale. Figure 2-24. Fabric is enriched with dolomite. 45 Figure 2-25. Calcite crystal present in the clay matrix. Figure 2-26. Biogenic silica sourced from diatoms present in the siliceous shale. 46 Figure 2-27. Micro pores in the matrix of calcareous shale. 2.5.4 Mineralogy In shales, permeability changes with the effective stress are strongly dependent on the mineralogy (Table 2-5). Within the argillaceous shale sub population (WC-2), an increase of quartz content (> 26%) show low permeability (Figure 2-28). In siliceous shale sub population (WC-3), average clay content samples (>30%) show increase in permeability and high quartz (>46%) and carbonate content (>3%) samples show decrease in permeability (Figure 2-29). In the other siliceous shale (WC-4) the same correlations were observed but carbonate content did not appear to affect the permeability (Figure 2- 30). The kaolinite content is clearly correlated with permeability in siliceous shale. In calcareous shale sub population (WC-6), an increase of clay content in samples (>12%) show an increase in permeability and increase of carbonate content (>45%) in samples 47 show decrease in permeability (Figures 2-3 1, 2-32). Increases in illite and kaolinite lead to increased permeability in calcareous shale. Figure 2-28. High permeability is visible in the argillaceous samples (WC-2), which has high clay content and low quartz content. A WC-Zi regression V WC-2.2 regressIon * WC-2.3 regression 101 Clay:48%; Quartz:46% 10 10_a CIay:60%; Quartz:31% * 2 4 6 8 10 12 14 Effective stress, MPa 16 18 20 48 * WC-8.1 —regression • WC-3.2 —regression • WC-3.3 —regression QZ6Clay:37%;CarbonatI% • 102 Quartz:58%; Clay:32%; Carbonate:3% 16 51oEt::2% 30 Effective stress. MPa Figure 2-29. Permeability change with the influence of mineralogy in siliceous shale (WC-3). High permeability is visible in the samples, which has average clay content (>30%) and low quartz and carbonate content. 101 • WC-41 —regression A WC-4.2 regresslon • WC-4.3 regresslon V WC-4.4 regsession I Clay:41%; Carbonate:1% ‘—QO%;c1ay:4O%;carbonae:13°, 105 Effective stress, MPa Figure 2-30. Permeability change with the influence of mineralogy in siliceous shale (WC-4). High permeability is visible in the samples, which has average clay content(>30%) and low quartz content. 49 • WC5.L rsgracdm • WC52 r.gretsitn d WC53 .çetd.n ——WC5.4 flr.sflot • WCS.5 r.gnni.nJ Carbonateay2%71% Clay 4%: Carbonate 82% 4? B 4? ______________________________________ -4 Z1%;carbte:890 10 15 20 25 Effective stress, MPa Figure 2-31. Permeability change with the influence of mineralogy in calcareous shale (WC-5). High permeability is visible in the samples, which has high clay content and low carbonate content. V WC6.1 —rgrsien * WC 6.2 regression WC 6.3 regressIon • WC 6.4 regressIon1 33%;carbonate:45°o 4? Clay 1 7%;Carbonate65% .410 .2 106 0 5 10 15 20 25 30 35 Effective stress, MPa Figure 2-32. Permeability change with the influence of mineralogy in calcareous shale (WC-6). High permeability is visible in the samples, which has high clay content and low carbonate content. 50 2.5.5 TOC Weak correlation is visible between permeability and the organic content of all the WCSB shales (Figures 2-33, 2-34, 2- 35). Permeability vs TOC 10 * . 10_i E * 10 - * 2 4 6 10 12 TOC ( %) Figure 2-33. Correlation between TOC and permeability of argillaceous shale (WC-1 and WC-2). No significant correlation is apparent. 51 Permeability vs TOC 10 _ -110 -2 . 10 io 1 1.5 2 2.5 3 3.5 TOC (wt %) Figure 2-34. Correlation between TOC and permeability of siliceous shale (WC-3 and WC-4). A weak correlation is seen. Permeability vs TOC 10 i02 2 110: 10 0 106 10 15 TOC (wt%) Figure 2-35. Weak correlation is visible between TOC and permeability of carbonaceous shale. 52 2.5.6 Permeability vs. Porosity Figure 2-36 shows the correlation between shale permeability and porosity, of argillaceous (WC-1, WC-2) and siliceous shales (WC-3, WC-4). Among all three litho type shales, the argillaceous shales (WC- 1, WC-2) have higher porosities and show a weak correlation between permeability and porosity. Porosities increase with increasing illite content in argillaceous shale. In the siliceous shales (WC-4) a negative correlation is shown: thus, as porosity increases, permeability decreases. 100 (±4) Iwc-1 * WC-3 4 -2 * A IWC-5 10 AWC-6 * 10 A 2 25 Gas filled porosity (%) Figure 2-36. Correlation between permeability and porosity of argillaceous (WC-1 and WC-2) and siliceous shale (WC-4). WC-1 and WC-2 show weak correlation between porosity and permeability. WC-4 shows a negative correlation. 53 2.6 DISCUSSION The results in section 4.0 show the effect of clay mineralogy on permeability, effective stress, anisotropy, shale fabric and porosity. In all three sample populations (argillaceous, siliceous and calcareous) mineralogical composition plays an important role. Low quartz, carbonate and high clay samples show higher permeability whereas high quartz, carbonate and low clay samples show lower permeability. The importance of fabric is implied but difficult to quantify. The permeability of Western Canadian Sedimentary basin samples were compared to shale permeabilities reported elsewhere (Dewhurst et al., 1998, 1999; Kwon et al., 2001; Yang and Aplin, 2007). The permeability of argillaceous shales is within the range (10 md to 1 0 md) of other reported samples and the permeability decreases effectively with the increase of confining pressures. Like others (Yang and Aplin, 2007; Kwon et al., 2001) permeability was influenced by the clay content. The results of this study provide the permeability ranges of argillaceous, siliceous and calcareous shale samples of Western Canadian Sedimentary Basin. In the present work, matrix permeability was measured using helium gas. Future research should also be carried out performing experiments with other gases (Ar, CH4N2) that adsorb during the permeability testing. 54 2.7 CONCLUSIONS Matrix permeability of Western Canadian shales were measured using transient pulse decay experiments. Permeability declines exponentially with increase of effective stress and stress-sensitivity depends on the mineralogy and fabric. Changes in the clay mineralogy influence the permeability. In clay, silica and calcite rich shales, permeability increases with the increase in clay content. High permeability clay rich shales had a well developed fabric due to clay orientation whereas quartz rich shales had random clay orientation. An increase of quartz content increased the randomness of fabric. The permeability loss with the effective stress is more observed in calcareous shale than argillaceous and siliceous shale. Some of the fractured siliceous shales (WC-4.1,4.2) had shown high permeabilities The fractures were not compressed much under confining pressures due to asperities. Directional permeability and TOC show positive correlation with permeability. Porosity shows a negative correlation within silica rich shales. Pore size distribution and mechanical properties were not mentioned in this paper are also to be considered as factors that affects the permeability of shales. This paper demonstrates some of the interrelations that exist in the shales with contrasting lithologies. 55 2.8 REFERENCES Ahmed, U., Crary, S. and Coates, G. 1991. Permeability estimation: the various sources and their interrelationships: Journal of Petroleum Technology, v.43,p.578-87. Aplin, A.C., Fleet, A.J., Macquaker, J.H.S. 1999. Muds and mudstones; physical and fluid-flow properties (in Muds and mudstone; physical and fluid-flow properties) Geological Society Special Publications, 158. Bell, J.S. 2006. In-situ and coal bed methane potential in Western Canada, Bulletin of Canadian Petroleum Geology, Vol. 54, p.1 97-220. Best, M.E., and Katsube, T.J. 1995. Shale permeability and its significance in hydrocarbon exploration, The Leading Edge, Vol. 14, Issue 3, pp. 165-170. Bjørlykke, K. 1999. Principal aspects of compaction and fluid flow in mudstones. Geological Society, London, Special Publications; v.158; p. 73-78 Bloch, 5. 1991. Empirical prediction of porosity and permeability in sandstones: AAPG Bulletin, V.90, P.11524-11532. Bohor, B.P., Hughes, R.E. 1971. Scanning electron microscopy of clays and clay minerals, Clays and Clay Minerals, V.19, pp. 49-54. Bolton, A.J., Maltman, A.J., Fisher, Q. 2000. Anisotropic permeability and bimodal pore- size distributions of fine-grained marine sediments, Mar. Pet. Geol., 17, 657-672. Brace, W.F. 1980. Permeability of crystalline and argillaceous rocks: Internat. J. Rock Mech. Mi Scie & Geomech., Abstr. 17, 241-25 1. Brace, W.F., Walsh, J.B., and Frangos, W.T. 1968. Permeability of granite under high pressure, J. Geophysical Research, 73, 2225. Bredehoeft, J.D., Neuzil, C.E., and Milly, P.C.D. 1983. Regional flow in the Dakota aquifer: A study of the role of confining layers, U.S. Geol. Surv. Water Supply Pap. 2237, 45 pp. 56 Bustin, A.M.M., Cui, X., Pathi, V.K. and Bustin, R.M. 2007. Importance of fabric on production rates of gas shales: experimental and numerical analysis (abs.). The American Association of Petroleum Geologists Annual Conference, April 3, 2007, Long Beach, CA, USA. Bustin, R.M. 1997. Importance of fabric and composition on the stress sensitivity of permeability in some coals, northern Sydney Basin, Australia: Relevance to coalbed methane exploration. The American Association of Petroleum Geologists Bulletin, v. 81, P. 1894—1908. Clennell, M.B., Dewhurst, D.N. Brown, K.M., Westbrook, G.K. 1999. Permeability anisotropy of consolidated clays. Geological Society, London, Special Publications; v. 158; p. 79-96. Coyner, K., Katsube, T.J., Best, M.E., and Williamsone, M. 1993. Gas and water permeability of tight shales from the venture gas field, offshore Nova Scotia; in current research, Part D; Geological survey of Canada, Paper 93-iD, p.129-136. Davies, K.D., Bryant, W.R., Vessell, R.K., Burkett, P.J. 1991. Porosity, permeabilities and microfabrics of Devonian shales. In: R.H. Bennett, W.R. Bryant and M.H. Hulbert, Editors, Microstructure of fine-grained sediments, Springer Verlag, New York,pp. 109—119. Dewhurst, D.N., Aplin, A.C. and Sarda, J.-P. 1999. Influence of clay fraction on pore- scale properties and hydraulic conductivity of experimentally compacted mudstones. Journal of Geophysical Research, v. 104, p. 29, 261—29,274. Dewhurst, D.N., Aplin, A.C., Sarda, J.-P. and Yang, Y. 1998. Compaction-driven evolution of porosity and permeability in natural mudstones: An experimental study. Journal of Geophysical Research, v. 103, p. 651—661. Dicker, A.!. and Smits, R.M. 1998. A practical approach for determining permeability from laboratory pressure-pulse decay measurements,” paper SPE 17578 presented at the 1998 SPE international meeting on petroleum engineering, Tianj in, China, 1-4 November. Faraj, Basim; Williams, Harold; Addison, Gary; McKinstry, Brian; Donaleshen, Roland; Sloan, Garth; Lee, James; Anderson, Tom; Leal, Rayleen; Anderson, Cohn; Lafleur, Carolle; and Ahistrom, Janice. 2004. Gas Shale Potential of Selected Upper Cretaceous, Jurassic, Triassic and Devonian Shale Formations, in the WCSB of Western Canada: Implications for Shale Gas production, GTI E&P Services Canada. 57 Horsrud, P., Sonstebo, E.F., Boe, R. 1998. Mechanical and petrophysical properties of North Sea Shales, Int.J.Rock. Mech. Min.Sci, Vol.35, 8, pp-1009-1020. Howard, J.J. 1992 Influence of authigenic-clay minerals on permeability (in Origin, Diagenesis, and Petrophysics of clay minerals in sandstones), SEPM special publication No.47. Hsieh, P.A et a!. 1981. A transient Laboratory Method for Determining the Hydraulic Properties of Tight Rocks- I theory.” Intl. J. Rock Mech. Mi Sci. and Geomech. Abstr, 18, 245. Jones, S.C. 1997. A technique for faster pulse-decay permeability measurements in tight rocks. Society of Petroleum Engineers Formation Evaluation, SPE 28450, p. 19—25. Jones, F.O. and Owens, W.W. 1980. A Laboratory Study of Gas Sands, JPT, pp. 1631- 40. Katsube, T.J. 2000. Shale permeability and pore-structure evolution characteristics; Geological survey of Canada, Current research 2000-E-15; 9p. Katsube, T.J., Best, M.E., and Mudford, B.S. 1991. Petrophysical characteristics of shales from the Scotian shelf; Geophysics, v.56, 10,p.1681-8 Kwon, 0., Kronenberg, A.K., Gangi, A.F., Johnson, B., and Herbert, B.E. 2004. Permeability of Illite-bearing shale: 1. Anisotropy and effects of clay content and loading, J.Geophys. Res., B 10205, doi: 10.1 029/2004JB003052. Kwon, 0., Kronenberg, A.K.,Gani, A.F., Johnson, B. 2001. Permeability of Wilcox shale and its effective pressure law, J. Geophys. Res., 106, 19,339- 19,353. Lin, W. 1977. Compressible fluid flow through rocks of variable permeability. Lawrence Livermore Laboratory, Livermore, California, Rept. UCRL-52304. Magara, K. 1978. Compaction and fluid migration: Practical Petroleum Geology, 319 pp., Elsevier Sci., New York. Marek, B.F. 1979. Permeability loss in depletion reservoirs. 5th Annu. Conf. Soc. Pet. eng., Lasvegas, NV, SPE pap.8433. 58 Marion, D., Nur, A. and Albert, F. 1989. Modelling the relationships between sonic velocity, porosity, permeability , and shaliness in sand, shale, and shaly sand: SPWLA Thirtieth Annual Logging Symposium Proceedings, paper G, p.1 -22. Meyer, R. 2002. Anisotropy of sandstone permeability, CREWES Research Report, Vol.14. Meyer, R., Krause, F.F. 2006. Permeability anisotropy and heterogeneity of a sandstone reservoir analogue. Bulletin of Canadian Petroleum Geology, Vol.54, 4, pp-30l- 318. Morrow, C., Shi L., and Byerlee, J. 1984. Permeability of fault gauge under confining pressure and shear stress: J. Geophys. Res., 89, 3 193-3200. Mudford, B.S., and Best, M.E. 1989. The Venture gas field, offshore Nova Scotia: A case history of overpressuring in a region of low sedimentation rate: AAPG Bull., 73, 1383-1396. Neuzil, C.E. 1994. How permeable are clays and shales? Water Resour. Res., 30, 145- 150. Nur, A., and Byerlee, J.D. 1971. An exact effective stress law for elastic deformation of rock with fluids, J. Geophys. Res., 76, 6414-6419. O’Brien, N. 1970. The fabric of shale- An electron-microscopic study. Sedimentalogy, vol.15, pp.229-46 O’Brien, N.R. and Slatt, R.M. 1990. Argillaceous Rock Atlas. New York, Spinger Verlag. Osborne, M.J, Swarbrick, R.E. 1997. Mechanisms for generating overpressure in sedimentary basins : A revaluation; AAPG Bulletin, Vol.81, no.6,pp.1023-41 Ostensen, R.W. 1986. The Effect of Stress-Dependent Permeability on Gas Production and Well Testing, SPEFE, pp.227-35. Pedrosa, O.A. 1986. Pressure Transient Response in Stress-Sensitive Formations, Paper No. SPE 15115, SPE California Regional Meeting, Oakland, California. 59 Raudsepp, M. and Pani, E. 2003. Application of rietveld analysis to environmental mineralogy, Mineralogical Association of Canada, Short Course 31. Recommended practices for core analysis (API RP 40). 1998. American Petroleum Institute, 2nd edition. Revil, A. Cathies, L.M., 1999. Permeability of shaly sands, Water Resour. Res., 35, 651— 662 Rieke, H.H. and Chilingarian, G.V. 1974. Compaction of Argillaceous sediments. Development of Sedimentology 16. Elsevier, Amsterdam. Rietveld, H.M. 1967. Line profiles of neuron powder-diffraction peaks for structure refinement. Acta Crystallogr. 22, pp.151-2. Rietveld, H.M. 1969. A profile refinement method for nuclear and magnetic structures. J. Appl. Crystallogr., 2, pp.65-71. Robin, P.Y.F. 1971. Note on effective pressure, J. Geophys. Res., 78, 2434-2437. Ross, D.J.K. 2007. Investigation into the importance of geochemical and pore structure heterogeneities for shale gas reservoir evaluation. Unpublished PhD thesis, University of British Columbia, Canada. Scholes, O.N., Clayton, S.A., Hoadley, A.F.A., Tiu, C. 2007. Permeability anisotropy due to consolidation of compressible porous media, Transp Porous Med, 68:365-387. Simon, D.E., Coulter, G.R., King, G. And Holmes, G. 1982. North sea completions- A laboratory study. J. Pet. Technology, pp. 253 1-2536. Skempton, A. W. 1961. Effective stress in soils, concrete and rocks, in Conference on pore pressure and suction in soils, pp.4-16, Butterworths, London. Srodon, J. 2001. Quantitative X-ray diffraction analysis of clay-bearing rocks from random preparations, Clay and Clay Minerals, vol. 49, No.6, pp.51428 Tissot, B.P., and Welte, D.H. 1984. Petroleum Formation and Occurrence (2’’ ed.): Heidelberg (Springer-Verlag). 60 Trimmer, D. 1982. Laboratory measurements of ultralow permeability of geologic materials, Rev. Sc Instrum, 53(8). Wardy, W.A., Zimmerman, R.W. 2004. Effective stress law for the permeability of clay- rich sandstones. Journal of Geophysical Research, Vol.109, B04203. Wyble, D.O. 1958. Effect of Applied Pressure on the Conductive Porosity and Permeability of Sandstones, Transactions AIME, Vol.2 13, pp. 430-32. Yang, Y. and Aplin, A.C. 2007. Permeability and petrophysical properties of 30 natural mudstones. Journal of Geophysical Research, v. 112, p. 1—14. Zoback, D.M. and Byerlee, D.J. 1975. The effect of microcrack dilatancy on permeability of Westerly granite, J. Geophys. Res., 80, 752-755. 61 CHAPTER 3 TRIAXIAL TESTING AND MECHANICAL BEHAVIOUR OF WOODFORD SHALES 62 Chapter-3 Triaxial Testing and Mechanical Behaviour of Woodford Shales 3.1 ABSTRACT Consolidated undrained triaxial compression tests were performed on samples of Woodford shales from Oklahoma, a quartz-rich shale with permeability of about i04 md. Woodford shales show a brittle-ductile transition at peak strength and transition changes with increasing confining pressure. The brittle-ductile transition correlates with the mineralogy; samples that are quartz-rich or carbonate-rich have a brittle post-peak behaviour and clay-rich samples have a ductile post peak behaviour. The stress-strain behaviour of samples was also studied with respect to the crack initiation and propagation thresholds. Young’s modulus shows a correlation with the rock compressive strength, whereas, the Poisson ratio shows no correlation. Shear dilation, which is effective in fracture permeability, mostly occurs at lower confining pressures whereas compaction occurs at high confining pressures. Samples that were deformed normal to bedding show greater stiffness and strength in comparison to samples deformed parallel to the bedding. A Mohr-Coulomb failure envelope was used to define the strength behaviour of the shales. Depending on the characteristics of the test samples, 63 cohesion was seen to vary from 9.2 MPa to 22.3 MPa and friction angle from 55 degrees to 34 degrees. 64 3.2 INTRODUCTION Shales are fine grained, low penneable sedimentary rocks with moderate to high clay content. Due to small pore size and low permeability, shales act as both source and reservoir rocks for hydrocarbons. Shales are responsible for 75% of drilling problems which includes weilbore stability, hole enlargement, stuck pipe, high torque and drag and side tracking in the petroleum industry (Dzialowski et al., 1993). It is estimated that wellbore stability problems costs the oil industry one billion dollars a year (Chen et al., 2002). To minimize weilbore stability problems, a proper understanding and knowledge of the mechanical properties of shale is essential. Knowledge of mechanical properties is also essential in the hydraulic fracturing of low permeable shales which is vital for commercial gas production. There is very little experimental data available on the mechanical behaviour of shales due to technical difficulties involved in sample retrieval, preparation and testing (Cook et al., 1990; Chenevert and Amanullah, 1997). Previous studies reporting rock mechanic results include those on the North Sea shales, Norway (Kvilhaug and Roaldset, 1998; Muniz and da Forntoura, 2005; Horsrud et al., 1998), Tournemire shale, France (Vales et al., 2004; Niandou et al., 1997); Wilcox shale, Louisiana, U.S. (William and Andreas, 1993) and Woodford shale, Oklahoma, U.S. (Younane et al., 2007). This paper presents the analysis and results of laboratory testing focussed on determining the mechanical properties of Woodford shales. A series of triaxial experiments were 65 performed on samples at different confining pressures to understand changes in the failure modes of the samples and to determine their intact shear strength properties. Also, we investigate the effect of dilation and lithology on failure. Changes in compressive strength and elastic behaviour of the samples, which were tested parallel and normal to the bedding plane, were also studied. 66 3.3 GEOLOGY The Woodford shale is a lateral continuous shale unit that occurs throughout the Oklahoma portion of the Anadarko Basin (Cardott and Lambert, 1985). This formation is of Late Devonian and Early Mississippian age, and is contained in a series of folds and fault blocks in the Arbuckle Mountains and Lawrence uplift (Figure 3-1) but becomes less deformed in southern Kansas through Oklahoma and into western Texas. The Woodford Formation is part of a black shale interval that covered much of North America in the Late Devonian. Similar aged formations includes the Antrim shale of Michigan, the New Albany shale of the Illinois Basin, the lower and upper members of the Bakken Formation of the Williston Basin, the Exshaw Formation of the Alberta Basin, and the Devonian shales of the Appalachian Basin (Hester et al., 1990). The Woodford Formation consists of dark-coloured to black siliceous and fissile shale, with rare, fossiliferous chert, bedded limestone, and dolomite. The shale ranges from 80 to 100 m in thickness, but locally, thicknesses of up to 210 m can occur (Amsden, 1975). The term Woodford is taken from the town of Woodford in Carter County and was first used by Taff (1903, 1904). The Woodford Formation lies above the Hunton Group carbonates (Sullivan, 1985) and below the Sycamore Formation. In the northern part of the Lawrence uplift, the Woodford Formation lies under the “pre-Welden shale” (Cooper, 1939) and the Welden Limestone. The Hunton Group consists of shallow marine carbonates of the Chimney Hill subgroup, argillaceous and silty carbonates of the middle Henryhouse, and Haragan-Bois 67 d’Arc formation (Jolly, 1988). The Mississippian Sycamore Formation consists of fine grained, silty limestone with interbedded thin dark shales (Jolly, 1988). The Woodford is typically quartz-rich (5O%80%), with illite (21% to 40%), and lesser concentrations of dolomite, kaolinite, pyrite, chlorite and albite. The organic matter consists of type II and type III kerogens (Heckel, 1972; Cluff, 1981). The porosities range from 3 to 9% and the permeability ranges from i0 to i0 md (Vulgamore, 2008). Thermal maturities vary from Ro 0.5%, which is the start of the oil window, to Ro 3.0%, which is the start of the dry gas window. 68 W oo df or d s ha le s tu dy a re a W E • A R E U C K L E S L A W R E N C E U PL IF T z C A N E Y C P N E Y II 0- — W E L D E N S Y C A M O R E W O O D F O R D W O O D F O R D z — — - ;- -- ,- -— -- — - - - - - - - . . - - - - - - Z H A RG A N U A R G N — % -- -- -- -- -- -- H E N R Y H O U SE H E N R Y H O U SE x A I. R N C H IM N EY H IL L C H IM N EY H IL L 1— - I SY LV A N SY LV A N Fi gu re 3- 1. Ph ys ic al an d st ra tig ra ph ic m ap o ft he W oo df or d sh al e (Jo hn so na n d Ca rd ot t, 19 92 ). z 0 0 V IO LA V IO LA 69 3.4 EXPERIMENTAL PROCEDURES 3.4.1 Sample description and preparation The shale samples tested in this project come from core recovered from the Woodford Formation, Oklahoma between depths of 2640 m to 2670 m. The Woodford is typically quartz- rich (50%—80%), with illite (21% to 40%), and lesser concentrations of dolomite, kaolinite, pyrite, chlorite and albite. The borehole was single-tube drilled with 90.76mm core diameter. The selected samples for triaxial testing were wrapped in aluminium foil and dipped into wax to preserve their water content. One set of samples was recored to 25.4 mm diameter using a diamond drill bit and air as a coolant to avoid reactions between the samples and water. These were then tested under various confining pressures. The smaller diameter allowed for samples to be drilled parallel and perpendicular to the core axis to test for anisotropy. The samples tested during triaxial analysis are listed in Table 3-1. After cutting the samples, they were jacketed with heat shrink to isolate the sample from the confining fluid. The axial and radial strains were measured using linear variable differential transducers (LVDT5). The axial strain was calculated dividing the change in axial length by the initial specimen length, and the volumetric strain was calculated using the equation: 6v8a+28r (11) 70 Where = volumetric strain La = axial strain Er = radial strain Table 3-1. Test specimen description Core Sample Sample Confining Sample Length Diameter Pressure ID Depth (m) Litho-Type (cm) (cm) (MPa) Test WS-1 2642.10-2647.8 shale 18.2 90.76 26.9 Triaxial WS-2 2657.4-2658.6 silty shale 18.3 90.76 27.8 Triaxial WS-3 2662.1-2662.9 silty shale 18.4 90.76 28 Triaxial WS-S1 2651.10-2653.0 siltyshale 50.8 25.4 5 Triaxial 51.45 25.4 10 Triaxial 52.26 25.4 15 Triaxial 51.06 25.4 27 Triaxial 51.95 25.4 35 Triaxial parallel to WS-A1 2654.10-2655.2 silty shale 50.9 25.4 27 bedding(triaxial) normal to WS-A1 2654. 10-2655.2 silty shale 51.06 25.4 27 bedding (triaxial) parallel to WS-A2 2655 .7-2658.1 silty shale 51.95 25.4 27 bedding (triaxial) normal to WS-A2 2655 .7-2658.1 silty shale 50.9 25.4 27 bedding (triaxial) 71 3.4.2 Apparatus 3.4.2.2 Load application system Figure 3-2 shows a schematic illustration of the triaxial cell which was used to measure the rock mechanical properties under consolidated undrained conditions without pore pressure measurements (ASTM D 2664-95 a). All tests were carried out at the UBC Earth and Oceans Sciences rock testing laboratory. The laboratory has an automatic, hydraulic servo-controlled load application system (Geotechnical Consulting and Testing Systems, GCTS). The load application system consists of a rigid 1000 KN load frame, and triaxial cell, in which the sample is mounted, to allow for applications of axial load and confining pressure during each test. The axial actuator, confining pressure, and pore pressure are independently controlled by an electro-hydraulic servo—controlled feedback system. The axial load is monitored using the load cell mounted between the load actuator and the loading piston. Loading can be controlled by stress rate or by strain rate. The triaxial apparatus specifications are given in Table 3-2. The well preserved sample is mounted between the top and bottom platens. It is jacketed in an impermeable Teflon sleeve (shrink wrap) that separates the confining fluid from the specimen. Two linear-variable differential transducers (LVDT) are clamped to the top and bottom platens to measure axial deformation of the specimen (Figure 3-3). Circumferential deformation of the sample is measured by a chain that is mounted as a belt around the sample, and the chain deformation is recorded during the test using an 72 LVDT. The confining pressures were monitored using a pressure transducer. Pore pressures were not measured in these experiments as these rocks are essentially impervious under these conditions and the strength values were determined in terms of total stress without correcting for pore pressures. Table 3-2. Specifications of Triaxial Apparatus Max.load 1000 kN (static load) 800 kN (dynamic load) Stiffness 750 kN/mm Max.confining pressure 70 MPa Inside diameter 150 mm, base diameter 450 Pressure cell mm, height 550 mm Sample dimensions Diameter 100 mm Length 200 mm Piston travel 50.8 mm Triaxial stages: 1.Consolidation 2.Dynamic loading 3.Universal 4. Static loading 73 Loading piston Internal LVDT’s Sample Membrane Load frame Triaxial cell — Top platen Confining ressure Inlet Fill/drain port Figure 3-2. Schematic diagram of the triaxial machine. Bottom platen L......Sample top pore pressure Axial load Sample bottom pore pressure 74 Figure 3-3. Triaxial arrangement of the sample. Axial deformation LVDT’s ircumferential LVDT “ Lower platen 75 3.4.3 Testing procedure Three series of triaxial tests were performed. The first involved triaxial tests performed on three specimens (WS-1, WS-2, and WS-3) of 90.76 mm diameter using confining pressures corresponding to reservoir pressures (Table 3-3). These were tested to study the effects of sample mineralogy on the strength behaviour of the shales. The second series of tests involved five samples of 25.4 mm diameter cored from samples of 90.76 mm diameter core (WS-S1) and were tested at different confining pressures (Table 3-5) to derive a Mohr-Coulomb strength envelope. The last series of tests involved two sets of samples (WS-A 1 and WS-A2) of 25.4 mm diameter (cored normal and parallel to bedding), which were tested to analyse the anisotropy. The preserved sample is placed between the lower and upper platens. The pressure vessel is then lowered on to the sample assembly while ensuring a proper seal with the base and connections to the confining pressure lines. A small axial load that is about 1% of the ultimate strength is applied to the sample and the chamber is filled with the confining fluid so that the confining pressure increases according to that specified for the test. Hydrostatic conditions are initially maintained. The axial load is then programmed to increase with a controlled axial strain rate of 0.05% per minute until the specimen fails. Once the sample fails the axial stress decreases to the initial hydrostatic condition and the confining pressure is reduced to zero. 76 3.5 RESULTS 3.5.1 Stress-strain curves A stress-axial strain plot of WS-1, WS-2 and WS-3 samples is shown in Figure 3-4. The stress strain response of laboratory tested rock has been investigated by various authors who divide the failure process into several key stages (Eberhardt et al., 1998; Corkum and Martin, 2007), as shown in Figure 3-5. These stages include: crack closure, crack initiation, crack damage, and peak strength. Upon the beginning of loading, fissures and some pores begin to close producing an initially inelastic, concave-upward stress-strain curve. Eventually, a transition occurs where a significant percentage of these fissures have closed and the rock response becomes more elastic. Crack closure marks the point where the axial strain curve changes from non-linear to linear behaviour. As loading increases, new cracks begin to initiate and propagate. This is recognized as a departure from linear radial strain as the sample begins to dilate (Martin, 1993; Eberhardt et al., 1998). The crack initiation threshold coincides with increased radial strain, and Poisson’s ratio increases as new cracks begin to form inside the sample. At the reversal of the volumetric strain curve, crack damage occurs (Eberhardt et al., 1998). This is believed to signify a point where the population of propagating cracks become critical, and is used as a measure of the long term in situ strength around a tunnel or borehole (Martin, 1993). The peak strength is the final stage when the sample fails and the maximum compressive strength is reached. 77 Most rocks exhibit stress-dependent elastic properties before the onset of significant cracking (Corkum and Martin, 2007). Elastic properties (listed in Table 3-3), such as Young’s modulus, were seen to increase with increases in confining pressure, whereas the Poisson’s ratio changed with respect to the lithology. The Young’s modulus was determined by the initial linear-least- square slope of the differential stress versus axial strain curve at 50% of the ultimate strength, and Poisson’s ratio was determined by the linear least square slope of the radial strain versus axial strain curve over the same interval where Young’s modulus was determined. WS-1 showed a semi-brittle to ductile failure mode, whereas WS-2 and WS-3 showed a brittle response upon failure (Figure 3-4). However, because of the limited number of samples tested, specific conclusions cannot be drawn except to say that sample WS-1 showed a more ductile post-peak behaviour than the other two samples. A volumetric strain- axial strain plot is shown in Figure 3-6 and the change in volumetric stain with load is shown in Figure 3-7. In both plots, the WS-1 sample shows more volumetric change than do the WS-2 and WS-3 samples. The increase in volume is stable in the initial stage due to the crack closure but shows a rapid increase prior to the failure. 78 3.5.2 Influence of Lithology The mineralogy of the Woodford shale was seen to have a strong effect on the triaxial compressibility of the samples. The mineralogy of the three samples is listed in Table 3-4. The two silica-rich (quartz > 60%) samples show higher compressive strengths and more brittle behaviour, whereas the medium clay rich sample (clay >35%) shows a lower compressive strength and ductile behaviour. WS-2 and WS-3 have less clay content than WS- 1 and showed a brittle regime with strain-softening after reaching peak stress. However, WS- 1 has a higher clay content (>35%) than that of WS-1 or WS-2 and showed a ductile regime with plastic yielding. Figure 3-7 shows dominant shear-enhanced compaction in the clay-rich shale (WS- 1), compared to the quartz-rich shales (WS-2 and WS-3). 3.5.3 Shale composition and pore compressibility The results show the effect of lithologic composition (Table 3-4) on the brittle-ductile transition and dilation of shale samples. During dilation, various changes occur in the porosity and pore size of the samples. Previous studies have shown that before the onset of dilitancy, and during crack closure, the porosity and permeability of the shale decreases; and after the onset of dilitancy, the porosity and permeability increases under applied stress (Zoback & Byerlee, 1975). 79 The changes in the porosity and pore size, under applied stress, depend on the lithologic composition of the shale samples. To investigate the changes in porosity and pore size under stress for the WS-1 and WS-3 samples, porosimetry experiments were performed. WS-1 had a high clay content (40%), whereas WS-3 had a high quartz content (72%). Figures 3-8 and 3-9 show changes in porosity and pore sizes of the WS- 1 and WS-3 samples, before and after triaxial testing. In the clay-rich sample (WS-1), the change in porosity (3.8% to 1.9%) is greater, whereas, in the quartz-rich shale (WS-3) a small difference (3.9% to 3.1%) is observed. A change in fluid volume under stress is seen in the macro and meso pores of the clay-rich sample, and the quartz-rich shale shows a change in fluid volume only in the macro pore region. 80 Figure 3-4. Deviator stress-axial strain plot of all the three triaxial samples tested. 160 140 120 80 .— I. 20 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Axial strain (% 81 Figure 3-5. The full stress-strain response of Woodford shale (WS-2) in triaxial compression. z I Strain (%) 82 0.5 Figure 3-6. Change of volumetric strain with axial strain of triaxial samples. 0.4 0.3 0.2 0.1 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Axial strain (%) 83 160 c 140 120 ri) 80 40 — ws-1 20 —WS-2 WS-3 0 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Volumetric strain (%) Figure 3-7. Volumetric response with loading. WS-1 has higher clay content and shows more compaction than does WS-2 and WS-3. 84 Figure 3-8. Change of porosity and pore size of WS-1(clay rich) sample before and after the triaxial test. 0.018 0.015 , 0.012 . — .— -.. 0.006 !r 23 5 1020 50100 1000 Pore diameter (nm) 10000 10000050000( 85 —.— -.S I Pore diameter (nm) Figure 3-9. Change of porosity and pore size of WS-2 (quartz-rich) sample before and after the triaxial test. 0.014 0.012 0.01 0.008 : ** Ilefore test (D 3.9%) • : •.. 4* After test ((I) 3.1%) ** *. —- It I :1. .* *0...I S w*I 0* a. .**•*‘ * * 0.006 0.004 0.002 0 23 5 1020 50100 1000 10000 10000050000( 86 Table 3-3. Elastic properties of the triaxial samples tested. Specimen No CP (MPa) a1-a3 (MPa) E (GPa) V WS-1 26.9 94.80 21.28 0.10 WS-2 27.8 156.90 23.96 0.22 WS-3 28.0 152.60 24.45 0.16 Table 3-4. Mineralogy (%) of the tested triaxial samples (WS-1, WS-2 and WS-3). Quarlz Pyiite Calcite Dolomite Albite lute Kaolinite Chlorite Gypsum Total-Clay Total-Carb Sample (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (°/ ‘0’S-i 56.0 0.8 0.0 0.4 2.7 36.1 2.1 1.6 0.2 39.7 0.4 4S-2 63.9 3.5 0.0 3.4 2.5 22.9 2.2 1.1 0.2 26.2 3.4 4S-3 71.6 2.0 0.0 0.7 1.9 21.7 0.7 1.0 0.5 23.4 0.7 87 3.5.4 Confining pressure A set of triaxial tests (Figure 3-10) were carried out on the 25.4 mm (1 in) core samples at different confining pressures. The 25.4 mm samples were cored, perpendicular to bedding, and a set of confining pressures were applied to each of the cores to investigate the brittle-ductile behaviour and dilatancy at various confining pressures. The stress-strain curves of the WS-S1 samples are shown in Figure 3-11. The peak stress shows a positive correlation with the confining pressures up to 15 MPa and after that, the correlation is not maintained. This might be due to the presence of pre-existing macro fractures in the two samples tested at the highest confining pressures, which were noted before the test (Figure 3-10), or due to the transition of sample failure from a brittle to ductile mode. The samples deformed at confining pressures of 5 MPa, 10 MPa, and 15 MPa followed a typical brittle fracturing regime whereas those at confining pressures of 27 MPa and 35 MPa appear to have deformed in a semi-brittle to ductile manner. The brittle-ductile transition is related to the strength of the rock (Mogi, 1965, 1 966a). In previous studies the brittle-ductile transition was found to occur when the confining pressures become equal to one-third of the stress difference at failure; in carbonate rocks, the difference is one-quarter (Paterson and Wong, 2005). Figure 3-12 shows the volumetric strain response with axial strain. At all confining pressures, excluding 35 MPa, the inelastic volumetric dilation occurs significantly below the peak strength (Figure 3-13). At confining pressures of 35 MPa, the shear compaction of the sample occurs 88 rather than dilatancy (Popp and Saizer, 2005). An increase in confining pressure reduces the dilitancy of the sample and increases the compactancy. The onset of dilation (C*) of the Woodford samples are shown in Figure 3-14. The onset of dilation with crack initiation is evident at the lower confining pressures of 5, 10, 15, and 27 MPa, and compaction or negative dilatancy is visible at the higher confining pressure of 35 MPa. At this confining pressure, the sample volume decreases; a phenomenon referred to as shear-enhanced compaction (Curran and Carroll, 1979). The triaxial compressive strength shows a correlation with the Young’s modulus and with Poisson’s ratio of the sample, at different confining pressures. As the confining pressures increase from 5 MPa to 15 MPa, stiffness increases from 11 GPa to 14 GPa, and Poisson ratio decreases from 0.23 to 0.18 (Table 3-5). The non-linear increase of Young’s modulus and the decrease of Poisson’s ratio at confining pressures of 27 MPa and 35 MPa are likely due to the presence of pre-test macro fractures. However, the horizontal macro fractures in the samples did not show any effect in the formation of shear fracture, which is clearly seen in Figure 3-10. 89 Table 3-5. Correlation of elastic constant of Woodford shales at different confining pressures. CP 27* and 35* have microfractures and show a non-linear increase of Young’s modulus with confining pressure. Specimen No CP (MPa) (MPa) E (GPa) V 5 103.1 11.00 0.23 10 168.3 11.00 0.21 WS-S1 15 192.5 14.06 0.18 27* 152.5 11.43 0.19 35* 172.8 13.90 0.14 90 Figure 3-10. Shear fractures at 5, 10, 15, 27, and 35 confining pressures for the WS-S1 sample. 91 Horizontal macro fractures Figure 3-11. Stress-strain plot at different confining pressures. 200 160 120 40 0 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 Axial strain (%) 92 Figure 3-12. Volumetric-axial strain at different confining pressures. 1.2 0.8 -0.4 -0.8 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 Axial strain (%) 93 Figure 3-13. Stress-volumetric strain of Woodford shales at different confining pressures. 180 150 120 . 90 c 60 30 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 Volumetric strain (%) 94 Figure 3-14. The onset of Dilatancy C at different confining pressures. c* o o.i 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Volumetric strain (% 95 3.5.5 Anisotropy Two sets (WS-A1 and WS-A2) of two samples each were cut parallel and normal to the bedding plane and tested for the strength under confinement. The anisotropy in shales is strongly related to the microstructure (bedding / lamination planes), foliation, pre-existing micro fractures and orientation of clays. Generally, anisotropy can be evaluated jn laboratory testing of rocks drilled in different directions. Because of the difficulty in coring samples in different directions, due to the fissility of shale, anisotropy is evaluated here parallel and normal to the bedding directions. In this experiment, the orientation 0 = 00 is parallel to bedding and 8 = 90° is normal to bedding. The reservoir confining pressure is used to test the peak strength at these two angles. Figures 3- 15 and 3-17 show that the WS-A1 and WS-A2 samples when loaded normal to the bedding are stiffer and stronger than those loaded parallel to bedding. Volumetric strain vs. deviatoric stress is plotted in Figures 3-17 and 3-19. In previous studies volumetric strain of shale was found to be strongly dependent on the loading orientation (Niandou et al., 1997). A non linearity in volumetric strain before reversal is evident when the deviatoric stress reaches the peak strength (Figures 3-17 and 3-19), in samples deformed parallel to the bedding plane compared to those deformed normal to the bedding. This is due to the bedding planes which are not compressed in 8 00; whereas, in 8 = 900, the bedding planes are compressed during shortening, which allows for more limited lateral strain and ultimately no further dilatancy is observed. The failed samples are shown in Figure 3-19. 96 The elastic properties are listed in Table 3-6. The anisotropy was also seen in the compressive strength values. It has been shown that stress-induced cracks favour propagating parallel to the loading direction. Thus when the planes of weakness represented in the fabric and foliation is also aligned parallel to the loading direction, the compressive strengths are significantly weaker. Table 3-6. Anisotropic elastic properties of WS-A1 and WS-A2 samples. Bedding plane Specimen No orientation CP (MPa) 1-G3 (MPa) E (GPa) V WS-A1 Parallel 27 94.1 10.8 0.32 Normal 27 124.9 11 0.2 WS-A2 Parallel 27 125 7.07 0.22 Normal 27 244.5 19.3 0.27 97 :, Figure 3-15. Deviatoric stress vs. axial strain of WS-A1 samples. The strength is greater in the samples that were deformed normal to the bedding, rather than those deformed parallel to the bedding. 140 120 100 80I. -.‘ .— -. 60 40 20 0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 Axial strain (%‘) 2 98 Figure 3-16. Deviatoric stress vs. volumetric strain of the WS-A1 sample. Larger dilatancy is observed in samples that were deformed parallel to the bedding, compared to those deformed normal to the bedding. I 140 120 100 80 60 40 20 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 Volumetric strain (%) 99 Normal Parallel Figure 3-17. Deviatoric stress vs. axial strain of WS-A2 samples. 100 250 200 150 100 50 __ 0 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5 Axial strain (%) Figure 3-18. Deviatoric stress vs. volumetric strain of WS-A2 samples. The bedding plane is more compressed and allows no space for increase in the lateral strain of samples defonned normal to the bedding plane; whereas, in the samples deformed parallel to the bedding, the bedding plane is less compressed and more dilatancy occur. I 250 200 150 100 50 0 0 0.1 0.2 0.3 Volumetric strain (%) 0.4 0.5 0.6 0.7 0.8 0.9 101 WS-A1 WS-A2 Figure 3-19. Samples were tested for anisotropy strength for parallel to bedding and normal to the bedding. Parallel to bedding Normal to bedding Parallel to bedding Normal to bedding 102 3.5.6 Failure envelope 3.5.6.1 Mohr-Coulomb failure The equation for the linear Mohr-Coulomb shear failure envelope is expressed as: r=c+cr’tanq5 (12) Where = shear stress c = cohesion intercept 0 = friction angle = effective normal stress on the failure plane Mohr-Coulomb failure envelopes for the Woodford shales (WS-S1) are presented in Figures 3- 20 and 3-21. Individual Mohr circles are constructed for different confining pressures, and a best-fit tangent is constructed along the various circles to give the Coulomb failure envelope. Based on these results, the values for confining pressures of 5, 10, and 15 MPa appear to show a 103 different failure envelope trend than those fractured samples at confining pressures of 27 and 35 MPa. Again it should be noted, that these latter two samples contained visible closed fractures, which may have led to their premature failure. Under low confinement, a cohesion of 9.2 MPa and internal friction angle of 56° was observed. At higher confinements, the apparent cohesion increases to 22 MPa and the friction angle decreases to 34°. A failure envelope (Figure 3-22) was also constructed for all five confining pressures and the change in the trend line shows a bilinear envelope with a transition at higher confining pressures. Such changes in triaxial behaviour (non-linear shear strength) is expected for weak rocks, and therefore this change may be confirming that the behaviour of the shale ranges from a more brittle material to a more plastic material with increasing confinement. 104 %%, Figure 3-20. Strength envelope of shear and normal stresses of WS-S1 sample at 5, 10, and 15 MPa confining pressures. Cohesion is 9.2 MPa and friction angel is 55 degrees. 105 c=9.29659, =55.6251 >7 / I ‘4, 300 270 240 210 180 150 - 120 90 60 30 0 / 0 30 60 90 120 150 180 210 Normal stress. a (MPa c=22.3273, 434.OO25 165 150 135 120 105 90 75 60 45 30 15 0 /7 0 15 30 45 60 75 90 105 120 135 150 165 180 195 210 Normal stress. (MPa Figure 3-21. Strength envelope of shear and normal stresses at 27 and 35 MPa confining pressures. Cohesion is 22 MPa and friction angle is 34 degrees. I 106 195 180 165 150 135 90 I) 60 45 30 15 0 0 30 60 Normal stress, Y (Mpa) Figure 3-22. Strenght envelope of shear and normal stresses of WS-1 sample at 5, 10, 15, 27 and 35 MPa confining pressures. The change in trend line shows a bilinear envelope with a transition at higher confining pressures. 90 120 150 180 21( 107 3.6 DISCUSSION Woodford shale exhibits a brittle to brittle-ductile transition at high confining pressures. At lower confining pressures shear fractures and brittleness dominates the behaviour and at high confining pressures, its ductility. This is consistent with reports of other shales tested (William and Andreas, 1993; Kvilhaug and Roaldset, 1998; Wong et al., 1997; Besuelle, 2000; Nygard et al., 2006 and Younane et al., 2007). Lithologic composition plays an important role in the shale strength and deformation properties. The Woodford shales show brittle behaviour, which enhances fracture dilation. The fracture dilation increases permeability and increases the reservoir production (Zoback and Byerlee, 1975). Clay rich shales are more ductile, which responds poorly to the hydrofracing stimulation. Porosity reduction and pore compressibility is more extensive in the clay rich rocks whereas it is less in the quartz rich shales. It shows the compressibility of deep Woodford shales under high confining conditions. Woodford shales were also seen to be anisotropic with respect to their deformation and strength characteristics measured. As the degree of anisotropy increases, the stability of the borehole decreases (Gupta and Zaman, 1999). 108 3.7 CONCLUSION Consolidated undrained triaxial experiments were conducted on Devonian-Mississippian Woodford shales. Woodford shale is one of the most silica rich gas shales with 50-80% quartz, and a permeability of around 1 o md. Samples were tested at different confining pressures to understand their brittle-ductile behaviour. Mineralogy was seen to correlate with brittle-ductile behaviour. Increases in quartz and carbonate content increase the brittleness, and increases in clay content increase the ductility of the samples. At low confining pressures, dilatancy was observed, due to brittle fracture, while at high confining pressures ductile behaviour was observed. Variations in stiffness and strength were seen in samples that had been cored perpendicular and parallel to the bedding. Higher strengths were seen in samples that had been loaded perpendicular to the bedding. 109 3.8 REFERENCES Abousleiman, Y., Tran, M., Hoang, S., Bobko, C., Ortega, A and Ulm, F. 2007. Geomechanics field and laboratory characterization of Woodford shale, SPE Annual Technical Conference and Exhibition, Anaheim, US.A. Amsden, T.W. 1975. Hunton Group (Late Ordovician, Silurian, and Early Devonian) in the Anadarko basin of Oklahoma: Oklahoma Geological Survey Bulletin 121, 213 p. ASTM. 1995. Standard test method for Triaxial compressive strength of undrained rock core specimens without pore pressure measurements (D 2664-95A). American Society for Testing and Materials (ASTM), Philadelphia. ASTM. 2008. Standard practices for preparing rock core as cylindrical test specimens and verifying conformance to dimensional and shale tolerances (D 4543). American Society for Testing and Materials (ASTM), Philadelphia, Pa. Attewell, P.B. and Sandford, M.R. 1974. Intrinsic shear strength of a brittle anisotropic rock I, experimental and mechanical interpretation. Int.J.Rock Mech. Mm. Sci & Geomech. Abstr.1, 423-430. Besuelle, P. 2001. Evolution of Strain Localisation with Stress in a Sandstone: Brittle and Semi Brittle Regimes, Phys. Chem. Earth (A), Vol.26, No.1-2, pp. 101-106. Birch, F. 1960. The velocity of compressional waves in rocks to 10 kilobars, 1: J. Geophys. Res., 65, 1083-1102. Bjorlykke, K and Hoeg, K. 1997. Effect of burial diagenesis on stresses, compaction and fluid flow in sedimentary basins, Marine and Petroleum Geology, Vol.14; No.3, pp.267- 276. 110 Brown, E.T., Richards, L.R. and Barr M.V. 1977. Shear strength characteristics of Delabole slate. In Proc. Conf. Rock Eng., Newcastle-Upon-Tyne, pp. 31-51. Cardott, B.J. and Lambart, M.W. 1985. Thermal maturation by vitrinite reflectance of Woodford shale, Anadarko basin, Oklahoma: American Association of Petroleum Geologists Bulletin, v.69, no.11,p.1982-98. Chen, X., Tan, C.P and Detournay, C. 2002. The impact of mud infiltration on wellbore stability in fractured rock masses, SPE/ISRM 78241 presented at the SPE/ISRM rock mechanics conference held in Irving, Texas. Chenevert, M.E. and Gatlin, C. 1965. Mechanical anisotropies of laminated sedimentary rocks. Soc. Pet. Eng. 15, 67-77. Chenevert, M.E. and Amanullah, M.D. 1997. Shale preservation and testing techniques for borehole stability studies. SPE 37672 presented at the 1997 SPEIIADC drilling conference held in Amsterdam, The Netherlands. Cluff, R.M. 1981. Mudrock fabrics and their significance- reply: Journal of Sedimentary Petrology, v.51, no.3,pp.1029-31 Cook, J.M., Sheppard, M.C. and Houwen, 0. H. 1990. Effects of strain rate and confining pressure on the deformation and failure of shale, SPE 19944 presented at the 1990 IADC/SPE drilling conference held in Houston, Texas. Cooper, C.L. 1939. Conodonts from a Bushberg-Hannibal horizon in Oklahoma, Journal of Palaeontology, 13. 379-422. Corkum, A.G. and Martin, C.D. 2007. The mechanical behaviour of week mudstone (Opalinus Clay) at low stresses, International Journal of Rock Mechanics & Mining Sciences, 44,196-209. Curran, J.H., Carroll, M.M. 1979. Shear stress enhancement of void compaction. J Geophys, Res, 84:1105-1112. 111 Donath, F.A. 1964. Strength variation and deformational behaviour of anisotropic rocks. In State of Stress in the Earth’s Crust (Edited by W. R. Judd), pp.281 -318. Elsevier, New York. Dzialowski, A., Hale, A. and Mahajan, 5. 1993. Lubricity and wear of shale: Effects of drilling fluids and mechanical parameters, SPE/IADC paper 25730, presented at the SPE/ IADC drilling conference in Amsterdam. Eberhardt, E., Stead, D., Stimpson, B. and Read, R.S. 1998. Identifying crack initiation and propagation thresholds in brittle rock. Canadian Geotechnical Journal: 35 (2), 222- 233. Gupta, D., and Zaman, M. 1999. Stability of boreholes in a geologic medium including the effect of anisotropy, Applied Mathematics and Mechanics, vol.20, no.8. Heckel, P.H. 1972. Recognition of ancient shallow marine environments, in Rigby, J., and Hamblin, W.K., eds., Recognition of ancient sedimentary environments: Society of Economic Palaeontologists and Mineralogists Special Publication 1,p.226-8. Hester, T.C.; Schmoker, J.W. and SahI, H.L. 1990. Log-Derived regional source-rock characteristics of the Woodford Shale, Anadarko Basin, Oklahoma, U.S. Geological Survey, 38pp. Hoek, E. 1964. Fracture of anisotropic rock. J.S. Afr. Inst. Mi Metall. 64 (10), 510-518. Horsrud, P., Eyvind, F.S. and Reider, B. 1998. Mechanical and petrophysical properties of North Sea Shales, mt. J. Rock Mech. & Mi Sci, pp.1009-1020. Ibanez, W., and Kronenberg, A.K. 1993. Experimental and deformation of shale : Mechanical properties and microstructural indicator of mechanisms, Int. J. Rock Mech. Mi Sci. & Geomech Abstr, vol.30, no.7, pp. 723-734. 112 Johnson, K.S., and Cardott, B.J. 1992. Geologic framework and hydrocarbon source rocks of Oklahoma, in K.S. Johnson and B.J. Cardott, eds., Source rocks in the Southern mid continent, 1990 symposium: OGS circular 93,p.21-37. Jolly, G.D. 1988. Correlation of the Woodford Formation in South-Central Oklahoma using Gamma-ray scintillation measurements of the natural background radiation. Unpublished Masters thesis, Stephen F.Austin State University, pp. 154. Kvilhaug, T. and Roaldset, E. 1998. Rock mechanical characterization of an offshore mudrock from Haltenbanken, Mid-Norway, Rock Mech. Rock Engng, 31(2), 95-115. Martin, C.D. 1993. The strength of massive Lac du Bonnet granite around underground openings. PhD. Thesis, Department of Civil and Geological Engineering, University of Manitoba, Winnipeg, Man. McLamore, R. and Gray, K.E. 1967. The Mechanical behaviour of anisotropic rocks. J. Eng. Ind. 89, 62-76. Mogi, K. 1965. Deformation and fracture of rocks under confining pressure (2). Elasticity and plasticity of some rocks. B Earthq Res I Tokyo 43: 349-379. Mogi, K. 1 966a. Pressure dependence of rock strength and transition from brittle fracture to ductile flow. B Earthq Res ITokyo 44:215-232. Muniz, E.S., da Fontoura, S.A.B. and Soares, A.C. 2005. CU Triaxial tests on North Sea shale, the 40th U.S symposium on Rock Mechanics (USRMS); Rock mechanics for energy, mineral and infrastructure development in the Northern Regions, Alaska. Niandou, H., Shao, J.F., Henry, J.P. and Fourmaintraux, D. 1997. Laboratory investigation of the mechanical behaviour of Tournemire shale, mt. J. Rock. Mech. Mi Sci, 34, pp. 3-16. Niandou, H.; Shao,J.F.; Henry, J.P.; Fourmaintraux, D. 1997. Laboratory Investigation of the mechanical behaviour of Tournemire Shale, Tnt. J. Rock Mech. Mi Sci. 34; 3-16. 113 Nygard, Runar., Gutierrez, Marte., Bratli, R.K. and Hoeg, K. 2006. Brittle-ductile transition, shear failure and leakage in shales and mudrocks, Marine and petroleum geology, 23, pp.201-212. Peterson, M.S. and Wong, T.F. 2005. Experimental Rock Deformation- The Brittle Field, 2nd ed., Springer-Verlag, Berlin and New York. Popp, T and Salzar, K. 2005. Triaxial deformation tests in a multi-anvil apparatus with ultrasonic monitoring. Technical report, Nagra. Sullivan, K. 1985. Organic facies variation of the Woodford shale in Western Oklahoma: Shale Shaker, v.35,p.76-89. Taff, J.A. 1903. Description of the Tishomingo quadrangle (Indian Territory): U.S. Geological Survey Geological Atlas. Folio 98, Scale 1:125,000, 8p. Taff, J.A. 1904. Preliminary report on the geology of the Arbuckle and Wichita mountains in Indian Territory and Oklahoma: U.S. Geological Survey professional paper 31, hip. Tissot, B.P. and Welte, D.H. 1984. Petroleum formation and occurrence (2 ed.): New York, Springer-Verlag, 699 p. Vales, F., Minh, D.N., Gharbi, H. and Rejeb, A. 2004. Experimental study of the influence of the degree of saturation on physical and mechanical properties in Tournemire shale, Applied Clay Science, 26, pp. 197-207. Vulgamore, T., Wolhart, S., Mayerhofer, M., Clawson, T. and Pope, C. 2008. Hydraulic Fracture Diagnostics Help Optimize Stimulations Of Woodford Shale Horizontals, The American Oil & Gas Reporter. Wong, T.F., David, C. and Zhu, W. 1997. The transition from brittle faulting to cataclastic flow in porous sandstone: mechanical deformation, J.Geo Phys. Res., 102, 3009-3025. 114 Zhao, J. 2000. Applicability of Mohr-Coulomb and Hoek-Brown strength criteria to the dynamic strength of brittle rock. International Journal of Rock Mechanics & Mining Sciences, 37, 1115-1121. Zoback, M.D., and Byerlee, J.D. 1975. The effect of microcrack dilatancy on the permeability of Westerly granite, J. Geophys. Res., 80, 752-755, 1975. 115 CHAPTER 4 Transient Pressure Pulse Decay Permeability of Woodford Shale 116 Chapter-4 Transient Pressure Pulse Decay Permeability of Woodford Shale 4.1 ABSTRACT The matrix permeability of Woodford shale in Oklahoma was investigated to assess gas shale production potential. This low permeability shale is rich in quartz (60-80%) and organic carbon (5-15%). We measured both the permeability and the permeability’s stress sensitivity using the transient pulse decay method. The aim was to resolve the correlation between effective stress, mineralogy, pore size, porosity, total organic content (TOC), depositional fabric, and permeability anisotropy. The permeability of Woodford shale declines exponentially with increasing effective stress, though the degree of stress-sensitivity depends on mineralogy and fabric. The tested samples are divisible into three populations: 1) high permeability (102 md to i0 md at 6.90 MPa to 20.69 MPa), 2) low permeability at high effective stresses (10k md to i0 md at 6.90 MPa to 20.69 MPa), and 3) very low permeability at low effective stresses (10 md to 106 md at 0.69 MPa to 3.45 MPa and i0 md to i0 md at 10.35 MPa to 12.42 MPa). In high permeability samples (quartz < 60%), permeability (k) decreases from 1 .22E-02 md to 8.43E-04 md as the effective stress increases from 6.90 MPa to 20.69 MPa. In low permeability samples (quartz > 60%), k 117 decreases from 1 .65E-04 to 2.75E-07 as the effective stress increases from 6.90 MPa to 20.69 MPa. In very low permeability samples (quartz> 60%), permeability decreases at low effective stress and decreases from 1 .39E-04 to 6.37E-06 as the effective stress increases from 0.69 MPa to 3.45 MPa. Mineralogical composition plays an important role in determining the permeability, which depends on pore size, porosity, fabric and TOC of the strata. In all three sample populations, low quartz content samples show higher permeability, and high quartz content samples show lower permeability. Porosity and permeability are positively correlated in the high permeability clay-rich samples but there is no correlation in the quartz- rich low permeability samples. Higher porosity is observed in the samples with high clay content. TOC is not correlated with permeability, even though the samples have TOC ranging between 5 to 15%. Porosimetry results suggest that fluid flow is mostly in the meso (2-5 0 nm) and macro pores (>50 nm) of the Woodford shales. Samples with higher clay content (>30%) show a higher intrusion in macro pores, whereas, samples with higher quartz content show intrusion in micro pores. Permeability was measured horizontal and normal to bedding at different effective stresses. Permeability is higher in all samples parallel to bedding, the amount of which varies with the mineralogy and the fabric. 118 4.2 INTRODUCTION Permeability plays an important role in governing and affecting the fluid flow and production potential of reservoirs (Marek, 1979; Simon et al., 1982). Fluid flow through low permeable (10 md) shales is a major consideration in recovering hydrocarbons in unconventional gas exploration. The Devonian Woodford shale is a major petroleum source rock and gas shale in the Anadarko Basin in northeast and southwest of Oklahoma (Cardott and Lambert, 1985). Because of the formation’s low matrix permeability (1 to 1 0 md) (Kareem, 1992), open natural fractures or hydraulic fracture stimulations play important roles in gas productivity (Schmoker, 1995; Surdam, 1997a; Popov et al., 2001). In shales, the production rate is governed by the Darcy flow rate in fractures and by diffusion and Darcy flow through the matrix (Bustin et al., 2007). The diffusion and flow through the matrix of the gas shales can be measured in the laboratory using the pulse decay experiments. The gas flows through the matrix are stress dependent and in many shale reservoirs that have wide fracture spacing, the production may be limited by matrix flow (Bustin et al., 2007). The importance of permeability in hydrocarbon accumulation has led geologists to investigate the controlling factors. Permeability depends on the effective stress (Zoback and Byerlee, 1975; Bustin, 1997; Kwon et al., 2001; Bell, 2006; Yang and Aplin, 2007), anisotropy (Bolton et al., 2000; Meyer and Krause, 2006; Kwon et al., 2004; Scholes et al., 2007; Clennell et al., 1999; 119 Horsrund et a!., 1998; Meyer, 2002), porosity (Katsube et a!., 1991; Bloch 1991; Davies et a!., 1991; Revil and Cathles, 1999; Kwon et a!., 2004; Yang and Aplin, 2007), mineralogical composition (Marion et a!., 1989; Ahmed et a!., 1991; Bustin, 1997; Howard, 1992), pore size distribution (Bolton et al., 2000; Yang and Aplin, 2007; Katsube et al., 1991) and microfabric (O’Brien, 1970; O’Brien and Sian, 1990). By understanding the relation between effective stress, shale fabric, composition, pore size, and porosity in production, more effective exploration and exploitation of potential gas shales may be possible (Bustin et al., 2007). The primary objective of this study is to investigate the matrix permeability of Woodford shale using the pulse decay experiment. We further investigate the interrelationship of effective stress, fabric, anisotropy, pore size distribution, porosity, TOC, and composition with shale permeability. 120 4.3 GEOLOGY The Woodford shale is a lateral continuous shale unit that occurs throughout the Oklahoma portion of the Anadarko Basin (Cardott and Lambert, 1985). This formation is of Late Devonian and Early Mississippian age, and is folded and faulted in the Arbuckle Mountains (Figure 4-1) but becomes less deformed in southern Kansas through Oklahoma and into western Texas. The Woodford Formation is part of a black shale interval that covered much of North America in the Late Devonian. Similar aged black shale formations includes the Antrim shale of Michigan, the New Albany shale of the Illinois Basin, the lower and upper members of the Bakken Formation of the Williston Basin, the Exshaw Formation of the Alberta Basin, and the Devonian shales of the Appalachian Basin (Hester et al., 1990). The Woodford consists of dark-coloured to black siliceous and fissile shale, with rare, fossiliferous chert, bedded limestone, and dolomite. The shale ranges from 80 to 100 m in thickness, but locally, thicknesses of up to 210 m can occur (Amsden, 1975). The term Woodford is taken from the town of Woodford in Carter County and was first used by Taff (1903, 1904). The Woodford Formation lies above Devonian Hunton Group carbonates (Sullivan, 1985) and below the Sycamore Formation. In the northern part of the Lawrence uplift, the Woodford Formation lies under the “pre-Welden shale” (Cooper, 1939) and the Welden Limestone. The Hunton Group consists of shallow marine carbonates of the Chimney Hill subgroup, argillaceous and silty carbonates of the middle Henryhouse, and Haragan-Bois d’Arc formation (Jolly, 1988). 121 The Mississippian Sycamore Formation consists of fine-grained, silty limestone with interbedded thin dark shales (Jolly, 1988). Samples are rich in quartz (50%—80%), with illite (21% to 40%) and lesser concentrations of dolomite, kaolinite, pyrite, chlorite and albite. The organic matter consists of type II and type III kerogen (Heckel, 1972; Cluff, 1981). Porosities range from 3 to 9% and the permeability ranges from 10.6 to i0 md (Vulgamore, 2008). Thermal maturities vary from Ro 0.5%, which is the start of the oil window, to Ro 3.0% and a dry gas window. 122 W o o d fo rd s h al e s tu d y a r e a \ J d f o r f 7 W F Fi gu re 4- 1. Ph ys ic al an d st ra tig ra ph ic m ap o ft he W oo df or d sh al e (Jo hn so na n d Ca rd ot t, 19 92 ). • A R B U C K L E S L A W R eN C e U P L IF T C A N E Y C A N E Y z < a_ — U , Z S Y C A M O R E W E L O E N W O O D F O R D W 0 0 0 F O R D a H A R G A N u J • . _ . _ ; : : — H A R G A N • A R C H E N R Y H O U S E z H E N R Y H O U S E A LA ID N C H IM N E Y H IL L C H IM N E Y H IL L SY LV A N SY LV A N - S z - J a 0 V IO LA V IO LA 12 3 4.4 EXPERIMENTAL TECHNIQUES 4.4.1 Sample Preparation Shale samples 3 cm in diameter were re-cored from full diameter preserved cores originally measuring approximately 8.9 cm. Samples were cored using a diamond drill bit with air as a coolant to avoid reactions between the samples and water. The samples will be tested at their insitu conditions. 4.4.2 Permeability through the Pulse Decay Experiment The pulse decay apparatus was constructed based on the design of Jones (1997) (Figure 4-2). The core holder is a RocTest® Hoek cell consisting of a hollow steel cylinder with threaded removable caps. A urethane rubber membrane inside the cell acts as a pressurization chamber for hydraulic fluid. The Hoek cell can maintain confining pressures up to 69 MPa with hydraulic oil. The core holder (Figure 2) is connected by the upstream reservoir of volume, Vi, and the downstream reservoir of volume, V2. At experiment onset, the sample is confined and the upstream and downstream vessel volumes are slightly higher than the sample pore volume, Vp (Jones, 1997). A differential pressure transducer measures the difference between the two reservoirs, using two absolute transducers at the upstream and downstream reservoirs. Helium gas is used to supply 124 pore pressure through the sample end plates. A computer-based data acquisition system is used to record all physical changes monitored and recorded at specified time intervals. Figure 4-2. Schematic diagram of the modified pulse decay experiment. 125 4.4.3 Procedure The sample is assembled in the Hoek cell between the two pistons. The confining pressure calculated from the reservoir effective stress is applied to the Hoek cell using a hydraulic pump. When valves at 1, 2, 3 and 4 are open, helium gas flows through the reservoirs and sample at a pressure between 6.9 MPa and 10.3 MPa. The higher mean pore pressures reduce gas slippage effect (Dicker, et al., 1988; Jones, 1997). After gas fills the system, the fill valve is closed and the system allowed to attain equilibrium pressure. Once all pressures are at equilibrium, the differential transducer is set to zero. Valve 2 is then closed, and the shutoff valve is slowly opened (creating a pressure increase in the upstream reservoir) until the differential transducer reaches less than 0.1 MPa. The shutoff valve and valve 4 are then closed, monitoring the following pressures: AP (differential pressure), P1 (upstream pressure) and P2 (downstream pressure). The upstream end pressure slowly decreases while the downstream pressure increases over time (Figure 4-3). The pore pressure decay (or increase) depends on fluid viscosity, sample pore volume, and sample and fluid compressibility (Brace et al., 1968; Lin, 1977; Hsieh et al., 1981; Trimmer 1982; Jones, 1997). Measurements normally stop after a 50% decay of the upstream pressure. Permeability is measured at different confining pressures. As the confining pressure increases, the decay time increases and the sample permeability decreases. 126 Figure 4-3. Pressure profiles across the sample reservoirs according to time decay. Hu = upstream pressure, Hd= downstream pressure 4.4.3.1 Calculation of Permeability Gas permeability is calculated using the following equation (Jones, 1997): kg= —14696 mipgLfz 1 fiApm _+_ v1 v2J Where (13) ml = slope of the linear equation, Pm = mean absolute pressure, JIg = viscosity of the gas [viscosity of the helium gas was taken from (API RP4O)], 127 L = length of the core plug, f1 = mass flow correction factor, A = cross-sectional area of cylindrical core plug, Vi = volume in small upstream reservoir, V2 = volume in small downstream reservoir, and gas compressibility correction factor [used for the deviation from ideal gas behaviour for helium at various pressures]. 4.4.4 Porosity Measurements The storage capacity of petroleum reservoir fluids depends on the porosity of the rock. Porosity is denoted by 1 and expressed by the following relationship: — pore volume — bulk volume (14) The porosity of the rocks is commonly determined by routine core analysis (lab measurements), as well as logging techniques. To determine the rock’s porosity based on equation (14), pore and bulk volumes are measured. The sample’s bulk volume is measured with a mercury-based Archimedes principle that observes the fluid volume displaced by a sample. Pore volume can be measured using helium pycnometry. All well samples are tested by helium pycnometry, with 20 128 to 30 g of powdered sample required for the test using Boyle’s Law. The apparatus consists of two equal chambers (or cells), a reference chamber and a sample chamber. The reference chamber has a volume of Vi at initial pressure of P1, and the sample chamber has an unknown volume V2 at initial pressure of P2. The unknown sample volume is calculated at equilibrium. The porosity of the rock can be determined when both the bulk and pore volume are known. 4.4.5 Mineralogy X-ray powder diffraction is the best available method for identifying minerals in shale samples. The Rietveld method is the most acceptable quantitative phase analysis (Rietveld, 1967, 1969). The quantitative analysis of clays is not accurate due to the platy nature and disorder of clay crystallites creating a preferred orientation (Srodon, 2001). In this study, samples were prepared using procedures described by Raudsepp and Pani (2003). Samples were crushed in a mortar then passed through a 0.4 mm sieve. They were ground in ethanol for around seven minutes using a vibratory McCrone® Micronising Mill with corundum elements. Grinding with ethanol instead of water was used to avoid clay swelling. After grinding, the sample was x-rayed using a Siemens D5000® diffractometer equipped with a theta-theta goniometer. The initial phase recognition was done with EVA software applying conventional search-match procedures to the database. The data was entered into the Rietveld program and the mineralogy was quantified using TOPAS® 3.0 software. 129 4.5.6 Total Organic Content (TOC) The TOC was measured using Rock-Eval pyrolysis. Prior to running the pyrolysis, samples were ground to a 100 mg powder and accurately weighed into metal crucibles. Samples were then subjected to programmed heating in an inert atmosphere (helium) to determine the amount of free hydrocarbons. Hydrocarbon and oxygen-containing compounds are volatilized when sample organic matter cracks (Tissot and Welte, 1984). The total organic carbon content is calculated with the following equation: K(S1 + S2) S4 TOC= +— (15) 10 10 Where K = elemental composition of natural substances, S 1+S2 = total potential (amount of free hydrocarbons and hydrocarbons generated through thermal cracking of organic matter), and S4 = quantity of CO2 produced during residual organic carbon oxidation. 4.5.7 Scanning Electron Microscopy (SEM) Following the procedures described by Bohor and Hughes (1971), samples were cracked to reveal the freshly exposed fracture. They were then glued to an aluminium stub and gold-coated 130 through evaporation under a high vacuum to a thickness of about 5-10 nm. SEM and BSEM observations were carried out using a Philips XL-30 SEM with a Princeton® Gamma-Tech PRISMIG energy-dispersive spectrometer at 15 kV. Images were taken with a 10 mm working distance. 4.5.8 Mercury Porosimetry The pore size and pore volume distributions of the Woodford shale samples were measured with a Micromeritics Autopore IV 9500 Series mercury porosimeter. The samples were prepared and run as per ASTM standards (ASTM D-4404-84). Sample pore sizes were measured by injecting mercury into the clean, dried samples in step-like increments up to pressures of 413 MPa. The required pressure depends on the contact angle, pore shape, and surface tension of the liquid. The relationship between intrusion pressure and pore diameter for cylindrical pores was given by Washburn (1921) as: — 4ycos6 P= (16) d 131 Where P = capillary pressure, F = interfacial tension of the mercury, o = contact angle, d = pore diameter. 132 4.6 RESULTS 4.6.1 Permeability vs. Effective Stress Woodford shale samples are labelled WS-1 to WS-13 to preserve location confidentiality. In this study, permeability of Woodford shales was measured at different effective stresses. Permeability vs. effective stress of the samples is plotted in Figures 4-4 and 4-5. The experimental results are listed in Tables 4-1 and 4-2 and composition in Table 4-3. The sample permeability markedly varies with the change in effective stresses. Generally, all samples show one order of difference in permeability from low to high effective stress; however, in samples WS-7 and WS-1 1 the permeability difference is two orders. For example, sample WS-7 shows a permeability of 1.39E-04 and 6.37E-06 at 0.6 MPa and 3.45MPa, and WS-l 1 shows a permeability of 4.67E-03 and 2.66E-05 at 10.35 MPa and 12.42 MPa. In these samples, the greater change in permeability occurs at the lower effective stresses. The influence of lithology on permeability was studied by dividing Woodford shale samples into three populations. The first population (WS-2, WS-3, WS-5, WS-6, WS-8, and WS-12) includes samples with high permeabilities (10.2 md to i0 md at 6.90 MPa to 20.69 MPa). The second population (WS-1, WS-4, WS-5, WS-6, WS-8, and WS-12), includes lower permeabilities at higher effective stresses (10 md to i0 md at 6.90 MPa to 20.69 MPa), and the third population (WS-7 and WS-11) shows very low permeabilities at lower effective stresses (10 md to 106 md 133 at 0.69 MPa to 3.45 MPa and i0 md to i0 md at 10.35 MPa to 12.42 MPa). In all three populations, samples with increased quartz content (>60-80%) show low permeability (Figure 4- 6). The first sample population (high permeability) shows higher permeability among samples with high clay content (30-40%) and average quartz content (<60%). Very low permeability samples have high quartz content and WS-1 1 has 80% quartz content. Two samples (WS-9 and WS-13) have high permeability despite lower clay content and high quartz content (Table 4-3). Natural or drilling induced fractures possibly explain the increased flux in these rocks. 134 is-s WS-2 Figure 4-4. Permeability of samples WS- 1 to WS-4 as a function of effective stress. Note the decrease in permeability in relation to the increase in effective stress. S S Effective stress, Ml’s WS-3 E I: .5 .5 S E 8 10 12 14 16 18 20 22 Effective stress, MPa IA 4. .5. WS-4 6 8 18 12 14 16 18 20 22 Effective stress, MPa 8 10 12 14 16 18 20 22 Effective stress, MPa 135 WS-5 • ws-6 1 8 10 12 84 16 IS 20 22 0 8 10 12 14 16 18 20 22Effective stress, MPa Effective stress, Ml’s — WS-7 io WS-8 Effective stress, MPa 6 8 10 12 14 16 18 20 22 Effective stress, M1’a WS-9 10’ ws-1o0.011 • 3 • 0.013 A 22 0 i1416182 22Effective stress, Effective stress, MPa 136 WS-12 0.022 0.02 0.018 0.016 0.014 0.012 0.01 0.008 WS-J3 8 10 12 14 1 18 20 22 1ffective stress, M1’a 2 8 10 12 14 16 18 20 22 Effective stress, MPa Figure 4-5. Permeability of samples WS-5 to WS-13 as a function of effective stress. 137 6 -- A Low-permeability A A very low-permeability 80 A High-permeability A ‘75 A e 70 o A A A A A -S 6 -4 010 10 10 10 1W 10 log permeability (md) Figure 4-6. Changes in permeability with quartz content. Note the lower permeability in quartz- rich samples (60-80%) and higher permeability in low quartz (<60%) samples. 38 Table 4-1. Permeability of Woodford shale samples from WS- 1 to WS-6 at different effective stresses Confining Pore Core pressure pressure Effective stress Permeability sample MPa MPa MPa md WS -1 WS -2 WS -3 WS -4 WS -5 WS -6 17.24 20.69 24.14 27.59 31.03 17.24 20,69 24.14 27.59 31.03 17,24 20.69 24.14 27.59 31.03 17.24 20.69 24.14 27,59 31.03 17.24 20.69 24.14 27.59 31.03 17.24 20.69 24.14 27.59 31.03 10.34 10.34 10.34 10,34 10.34 10.34 10,34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10,34 10.34 10.34 10.34 10.34 10,34 10.34 10.34 10.34 6.90 10.35 13,80 17.25 20.69 6.90 10.35 13.80 17.25 20.69 6.90 10.35 13.80 17.25 20.69 6.90 10.35 13.80 17.25 20,69 6.90 10.35 13.80 17,2 5 20.69 6.90 10.35 13.80 17.25 20.69 9.43E-05 9.27E-05 7.02E-05 6,14E-05 5,29E-05 8.33E-03 6.64E-03 5 .65E-03 4.58E-03 3 .77E-03 3.95E-03 3 .25E-03 2.3 1E-03 1.70E-03 1 .43E-03 2.3 8E-04 1.12E-04 6.68E-05 3.1OE-05 1.62E-05 2.68E-06 1 .49E-06 8.71E-07 4.82E-07 2.75E-07 1 .65E-04 1.14E-04 7.8 1E-05 5.37E-05 3.69E-05 139 Table 4-2. Permeability of Woodford shale samples from WS-7 to WS- 13 at different effective stresses Confining Pore Core pressure pressure Effective stress Permeability sample MPa MPa MPa md WS-7 WS-8 WS-9 ws-10 WS-1 1 WS-12 WS-13 11.03 11.72 12.41 13.10 13.79 17.24 20.69 24.14 27.59 31.03 17.24 20.69 24.14 27.59 31.03 17.24 20.69 24.14 27.59 31.03 20.69 21.38 22.07 22.76 17.24 20.69 24.14 27.59 31.03 17.24 20.69 24.14 27.59 31.03 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 10.34 0.69 1.38 2.07 2.76 3.45 6.90 10.35 13.80 17.25 20.69 6.90 10.35 13.80 17.25 20.69 6.90 10.35 13.80 17.25 20.69 10.35 11.04 11.73 12.42 6.90 10.35 13.80 17.25 20.69 6.90 10.35 13.80 17.25 20.69 1 .39E-04 6.43E-05 2.9 SE-OS 1 .54E-OS 6.37E-06 6.SSE-O6 6.33E-06 6.1 1E-06 5.91E-06 5.7 1E-06 1 .22E-02 1 .05E-02 9.53E-03 8.22E-03 7.24E-03 2.83E-03 1 .99E-03 1 .54E-03 1.14E-03 8.43E-04 4.67E-03 8.54E-04 1 .69E-04 2.66E-05 2.46E-04 2.22E-04 2.O1E-04 1 .82E-04 1 .65E-04 2.16E-02 I .83E-02 1 .46E-02 I .08E-02 7.78E-03 140 Table 4-3. Lithologic composition, porosity and TOC of core samples. Sample-ID Porosity TOC Quartz Pyrite Calcite Dolomite Albite Illite Kaolinite Chlorite Gypsum Total-Clay Total-Carb (wt%) (wt%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) WS-1 12.5 7.9 64.6 23 0.5 7.0 1.6 19.8 1.1 14 1.7 22.3 7.5 WS-2 2 I 5.3 51.8 2.6 7.1 2.4 1.4 29.9 1.4 28 0.7 34.1 9.5 WS-3 16 136 53.9 2.0 0.7 7.1 3.6 31.0 0.6 02 1.0 31.7 7.8 WS-4 2.4 14.9 67.8 3.1 0.5 4.5 25 19.9 1.0 0.6 0.3 21.5 5.0 WS-5 5.7 10.5 63.5 2.4 0.6 0.6 2.4 28.6 1.3 0.4 0.2 30.3 1.2 WS-6 5.2 6.4 72.3 1.7 0.8 0.6 1.7 16.8 2.2 1.3 2.7 20.3 1.3 WS-7 5.1 8.6 67.1 1.8 0.7 1.2 2.3 24.7 1.1 0.9 0.3 26.6 2.0 WS-8 2.5 10.8 64.9 2.6 0.1 1.5 1.9 25.3 2.8 0.6 0.3 28.7 1.6 WS-9 1.9 10.3 68.2 3.1 0.4 4.4 2.2 20.2 1.1 0.3 0.1 21.5 4.8 WS-10 1.5 11.2 59.8 3.9 0.2 1.0 2.7 28.4 2.8 1.0 0.3 32.1 1.2 WS-11 5.2 108 80.4 4.1 0.3 0.5 2.0 11.1 1.0 0.4 0.3 12.5 0.8 WS-12 4.7 11.7 67.3 3.9 0.5 2.3 2.1 22.0 0.8 0.8 0.4 23.6 2.8 WS-13 2.8 11.7 76.2 3.1 0.2 0.5 2.4 15.6 1.6 0.3 0.1 17.6 0.7 4.6.2 Anisotropy Directional permeability (parallel and normal to bedding) was measured on samples WS Al, WS-A2, and WS-A3 (Table 4-4), results are shown in Figures 4-7 to 4-9. Permeabilities were measured parallel and normal to bedding at different effective stresses. In WS-A1 and WS-A3, the permeability measured parallel to the bedding is one order of magnitude greater than the permeability measured normal to bedding. Whereas in WS-A2, the permeability measured parallel to bedding is two orders of magnitude greater than the permeability measured normal to the bedding. At lower effective stress, samples show isotropic behaviour. As the stress increases, anisotropy increases. Nevertheless, in WS-A2, anisotropy is visible at lower and higher effective stresses. The high clay content sample (WS-A2) has higher horizontal permeability. 141 Table 4-4. Permeabilities measured parallel and normal to bedding for samples WS-A1, WS-A2 and WS-A3. Confining Pore Parallel to Normal to Core pressure pressure Effective stress bedding (k) bedding (k) sample MPa MPa MPa md md WS-A1 17.24 10.34 6.90 1.65E-04 5.86E-05 20.69 10.34 10.35 1.14E-04 1.85E-05 24.14 10.34 13.80 7.81E-05 5.81E-06 27.59 10.34 17.25 5.37E-05 1.83E-06 31.03 10.34 20.69 3.69E-05 5.77E-07 WS-A2 17.24 10.34 6.90 2.26E-02 2.61E-04 20.69 10.34 10.35 1.43E-02 2.49E-04 24.14 10.34 13.80 1.06E-02 2.40E-04 27.59 10.34 17.25 7.61E-03 2.35E-04 31.03 10.34 20.69 5.44E-03 2.27E-04 WS-A3 15.86 10.34 5.52 1.93E-05 9.82E-06 16.55 10.34 6.21 1.83E-05 8.03E-06 17.24 10.34 6.9 1.73E-05 7.29E-06 17.93 10.34 7.59 1.64E-05 5.16E-06 Table 4-5. Lithologic composition of samples with permeabilities measured parallel and normal to bedding. Quartz Pyrite Calcite Dolomite Albite Illite Kaolinite Chlorite Gypsum Total-Clay Total-Carb Sample-ID (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) WS-A1 72.3 1.7 0.8 0.6 1.7 16.8 2.2 1.3 2.7 20.3 1.3 WS-A2 64.8 2.3 2.8 1.9 1.5 22.9 2.1 0.9 0.8 26.0 4.7 WS-A3 69.2 3.0 1.6 1.5 0.8 12.0 3.3 8.6 0.0 23.8 3.1 142 Figure 4-7. Permeability vs. effective stress for the WS-A1 sample; flow measured parallel and normal to fabric. Change in permeability is less at low effective stress and more in high effective stress. WS-A1 1o- A normal to bedding — regression V parallel to bedding — regression E . - 10 55 1o- 8 10 12 14 16 Effective stress, MPa 18 20 22 143 WS-A2 101 V psrallel to bedding —regression A normal to bedding — regression :141182 22 Effective stress, MPa Figure 4-8. Permeability vs. effective stress for the WS-A2 sample; flow measured parallel and normal to fabric. Anisotropy is observed at low and high effective stresses. 144 WS-A3 V parallel to bedding — regression A normal to bedding — regression iø6 5 5.5 6 6.5 7 7.5 8 Effective stress, MPa Figure 4-9. Permeability vs. effective stress for the WS-A3 sample; flow measured parallel and normal to fabric. 145 4.6.3 Pore Size Distribution Pore size distribution of Woodford shales were studied in three sample populations: 1) high permeability (102 md to iO md at 6.90 MPa to 20.69 MPa), 2) low permeability (10 md to i0 md at 6.90 MPa to 20.69 MPa), and 3) very low permeability (10 md to 106 md at 0.69 MPa to 3.45 MPa and i0 md to i0 md at 10.35 MPa to 12.42 MPa). Figures 4-11 and 4-12 shows the log differential intrusion vs. pore diameter of all three sample populations. A broad pore size distribution is observed in all three populations with pore volume mostly below 100 nm and above 10000 nm. Higher clay content samples (>30%) show a higher intrusion in the macro pores, whereas higher quartz content samples show an intrusion in meso pores. For example, WS-2, WS-3, WS-5, and WS-10 have clay content above 30% and show higher intrusions at macro pores of 100000 nm. WS- 11 has high quartz content (80%) and shows an intrusion of 0.025 mL/g (higher than clay samples) at meso pores of 3 nm. 146 —‘-—WS-2 —‘—-WS-3 —*-—WS-9 ——WS-1O —l—WS-13 0.01( 2.13E-03 md 1.22E-02 md ‘0.014 — , 0.012 8.33E-93 md 0.01 0.008 3.95E-03 md 0.006 4 0.004 t ‘\ 2.16E-02 md / .—, 0.002 V - -- --I1 ØO 10 102103 io 1o Pore size diameter (nm) Figure 4-10. Pore size distribution for tested high permeability samples (WS-2, WS-3,WS-9,WS-10 and WS-13). The macro pores of clay rich samples (WS-2, WS-3 and WS-10) and meso pores of quartz samples (WS-9 and WS-13) show higher pore volumes. 147 —.--WS-1 -4—WS-4 —.—WS-5 —--WS-6 —--WS-8 ---WS-12 0.016 2.3E—O4 md ‘0.014 2.46E-04 md 0.012 6.55E—06 md Z.68E-06 md 0.01 0.008 1.65E-04 md / - 0.006 ‘I’ 0 004 9.43E-O5 md ooo; - / i0 106 Pore size diameter (nm) Figure 4-11. Pore size distribution for tested low permeability samples (WS- 1, WS-4, WS-5, WS-6, WS-8, WS-12). Meso pores of quartz (WS-1, WS-4, WS 6, WS-8 and WS-12) and macro pores of clay rich sample (WS-5) show higher pore volumes. 148 ——WS-7 0.025 —.--ws-11. 4002 A flEE __ ‘—. . .* . *—I I, a I 100 101 102 1o5 106 Pore size diameter (nm) Figure 4-12. Pore size distribution of very low permeability samples (WS-7 and WS-11). Meso pores of quartz rich (WS-7 and WS-11) show higher pore volumes. WS-7 has k i04 md to 106 md at 0.69 MPa to 3.45 MPa, WS-1 1 has k i0 md to i0 md at 10.35 MPa to 12.42 MPa 149 4.6.4 Scanning Electron Microscopy The Woodford shale is characterized by a fabric mostly consisting of “Tasmanites,” marine algae. Tasmanites normally have a circular shape (Figure 4-13), which after compression forms into flattened spores (Figure 4-14 and 4-15) (O’Brien, 1990). High permeability samples with >30% clay content show preferred clay orientation (Figure 4-16), whereas, low permeability samples with high quartz content (60-80%) show random orientation of clay particles (Figure 4-17). Micro-fracturing is visible in the quartz dominated samples. The development of fractures across the quartz grain is shown in Figure 4-18 and 4-19. 150 151 Figure 4-13. Uncompressed Tasmanites in the clay matrix. 152 Figure 4-14. Compressed Tasmanite in the clay matrix. Figure 4-15. Highly compressed (around six fold) Tasmanite. 153 154 Figure 4-16. SEM micrograph showing the preferred clay orientation of clay rich samples. Figure 4-17. SEM micrograph showing the random orientation of clay particles in quartz rich samples. 155 I. 156 Figure 4-18. Mierofractures in the Woodford shales. 157 Figure 4-19. Development of fractures in the quartz grain of Woodford shales. 4.6.5 Porosity Porosity was measured using a helium pycnometer. Moderate correlation exists between porosity and permeability (Figure 4-20) in the high permeable clay rich samples but no correlation is visible in the quartz rich samples. 3 2.8 * 2.2 I * 2* 1 log permeability (md) Figure 4-20. Correlation between porosity and permeability of clay-rich Woodford shales 158 4.6.6 Total Organic Content Measured Total Organic Content (TOC) is between 5 to 15%. Permeability and TOC are not correlated (Figure 4-2 1). 1! • * 14 * 13 12 111 * ** Cs * cs 10 9 * * 0 *6 1 I *1 io8 io2 100 log permeability (md) Figure 4-21. X-plot of TOC and permeability. 159 4.7 DISCUSSION Quartz-rich Woodford shales have lower permeabilities (1 0 md) compared to shale permeabilities reported elsewhere which are mainly in the order of 1 0 md (Dewhurst et al., 1998, 1999; Kwon et a!., 2001; Yang and Aplin, 2007). In this study, samples WS-7 and WS-1 1 showed low permeabilities at lower effective stress levels (1O md to 106 md at 0.69 MPa to 3.45 MPa, and i0 md to i0 md at 10.35 MPa to 12.42 MPa) at lower effective stress levels. The lower permeabilities are likely due to biogenic recrystallized silica. These results show that effective pressure, anisotropy, pore size, and porosity affect the matrix permeability of Woodford shale. Lithology is also clearly important in affecting the matrix permeability. Clay-rich samples have higher porosity and permeability, when compared to quartz-rich samples. The porosimetry show, clay-rich samples have high pore volume in macro pores, while quartz-rich samples have segregated pore volume in meso pores. The results also show that porosity is correlated with clay content. Quartz-rich shales show a brittle behaviour, which corresponds with increased fracture dilation, and thus increased permeability and reservoir production (Zoback and Byerlee, 1975). Clay-rich shales show ductility, which more poorly responds to fracture stimulation. Porosity reduction and increased pore compressibility is more often seen in 160 the clay-rich rocks, but is less common in the quartz-rich shales. These results indicate the compressibility of deep Woodford shales under confining conditions. 4.8 CONCLUSIONS Our experimental data indicate that lithology plays an important role in determining permeability. Quartz controls the permeability in all three sample populations. The sample’s pore volume mostly correlates with clay and quartz content. Clay rich samples have high pore volume in macro pores, while quartz rich samples have segregated pore volume in mesopores. Correlation is visible between porosity and permeability in the high permeability clay rich samples and no correlation visible in the quartz rich low permeability samples. No visible correlation exists between TOC and permeability. 161 4.9 REFERENCES Ahmed, U., Crary, S., and Coates, G. 1991. Permeability estimation: the various sources and their interrelationships: Journal of Petroleum Technology, v.43,p.578-87. Al-Shaieb, X., Puckette, J.O., Abdalla, A.A., Tigert, V., and Ortoleva, P.J. 1994b. The banded character of pressure seals, in P.J. Ortoleva, ed., Basin compartments and seals: AAPG Memoir 61,p.351-67. Al-Shaieb, Z., Puckette, J.O., Abdalla, A.A., and Ely, P. 1994a. Megacompartment complex in the Anadarko basin: a completely sealed overpressured phenomenon, in P.J. Ortoleva, ed., Basin compartments and seals: AAPG Memoir 61, p.55-68. Al-Shaieb, Z., Puckette, P., Ely, P., and Tiger, V. 1992. Pressure compartments and seals in the Anadarko basin, in K.S. Johnson and B.J. Cardott, eds., Source rocks in the southern mid-continent, 1990 symposium: Oklahoma Geological Survey Circular, 93,p.210-28. Amsden, T.W. 1975. Hunton Group (Late Ordovician, Silurian, and Early Devonian) in the Anadarko basin of Oklahoma: Oklahoma Geological Survey Bulletin, p.121, 213. ASTM. 2004. Standard Test Method for Determination of Pore Volume and Pore Volume Distribution of Soil and Rock by Mercury Intrusion Porosimetry (D 4404-84). American Society for Testing and Materials (ASTM), Philadelphia, Pa. Bell, J.S. 2006. In-situ and coal bed methane potential in Western Canada, Bulletin of Canadian Petroleum Geology, v. 54, p.197-220. Bloch, S. 1991. Empirical prediction of porosity and permeability in sandstones: AAPG Bulletin, v.90, p.1 1524-11532. Bohor, B.P. and Hughes, R.E. 1971. Scanning electron microscopy of clays and clay minerals, Clays and Clay Minerals, v.19,p.49-54. 162 Bolton, A.J., Maltman, A.J., and Fisher, Q. 2000. Anisotropic permeability and bimodal pore-size distributions of fine-grained marine sediments, Mar. Pet. Geol., 17, p.657-672. Brace, W.F., Walsh, J.B., and Frangos, W.T. 1968. Permeability of granite under high pressure, Journal of Geophysical Research, v. 73, p.2225. Bustin, A.M.M., Cui, X., Pathi, V.K. and Bustin, R.M. 2007. Importance of fabric on production rates of gas shales: experimental and numerical analysis (abs.). The American Association of Petroleum Geologists Annual Conference, April 3, 2007, Long Beach, CA, USA. Bustin, R.M. 1997. Importance of fabric and composition on the stress sensitivity of permeability in some coals, northern Sydney Basin, Australia: Relevance to coalbed methane exploration. The American Association of Petroleum Geologists Bulletin, v. 81,p.1894—908. Cardott, B .J. and Lambart, M.W. 1985. Thermal maturation by vitrinite reflectance of Woodford shale, Anadarko basin, Oklahoma: American Association of Petroleum Geologists Bulletin, v.69, no.11,p.1982-98. Cardott, B.J. and Lambart, M.W. 1985. Thermal maturation by vitrinite reflectance of Woodford shale, Anadarko basin, Oklahoma: American Association of Petroleum Geologists Bulletin, v.69, no.11,p.1982-98. Clennell, M.B., Dewhurst, D.N., Brown, K.M., and Westbrook, G.K. 1999. Permeability anisotropy of consolidated clays. Geological Society, London, Special Publications; v. 158; p. 79-96. Cluff, R.M. 1981. Mudrock fabrics and their significance- reply: Journal of Sedimentary Petrology, v.51, no.3,p.1029-31 Cooper, C.L. 1939. Conodonts from a Bushberg-Hannibal horizon in Oklahoma, Journal of Palaeontology, 13, p. 379-422. 163 Davies, K.D., Bryant, W.R., Vessell, R.K., and Burkett, P.J. 1991. Porosity, permeabilities and microfabrics of Devonian shales. In: R.H. Bennett, W.R. Bryant and M.H. Hulbert, Editors, Microstructure of fine-grained sediments, Springer Verlag, New York, p.1 09—119. Dewhurst, D.N., Aplin, A.C. and Sarda, J.-P. 1999. Influence of clay fraction on pore- scale properties and hydraulic conductivity of experimentally compacted mudstones. Journal of Geophysical Research, v. 104, p. 29, 261—29,274. Dewhurst, D.N., Aplin, A.C., Sarda, J.-P. and Yang, Y. 1998. Compaction- driven evolution of porosity and permeability in natural mudstones: An experimental study. Journal of Geophysical Research, v. 103, p. 651—661. Dicker, A.I. and Smits, R.M. 1998. A practical approach for determining permeability from laboratory pressure-pulse decay measurements,” paper SPE 17578 presented at the 1998 SPE international meeting on petroleum engineering, Tianjin, China, 1-4 November. Haines, L. 2006. Activity builds in Woodford shale, Oil and Gas Investor, p.17. Heckel, P.11. 1972. Recognition of ancient shallow marine environments, in Rigby, J., and Hamblin, W.K., eds., Recognition of ancient sedimentary environments: Society of Economic Palaeontologists and Mineralogists Special Publication 1, p.226-286. Hester, T.C., Schmoker, J.W., and Sahi, H.L. 1990. Log-Derived regional source-rock characteristics of the Woodford Shale, Anadarko Basin, Oklahoma, U.S. Geological Survey, p.38. Horsrud, P., Sonstebo, E.F., and Boe, R. 1998. Mechanical and petrophysical properties of North Sea Shales, Int.J.Rock. Mech. Min.Sci, v.35, 8,p.1009-20. Howard, J.J. 1992. Influence of authigenic-clay minerals on permeability (in Origin, Diagenesis, and Petrophysics of clay minerals in sandstones), SEPM special publication No.47, PAGE Hsieh, P.A., Tracy, J.V., Neuzil, C.E., Bredehoeft and Silliman, S.E. 1981. A transient laboratory method for determining the hydraulic properties of tight rocks - I theory. Intl. J. Rock Mech. Mi Sci. and Geomech. Abstr, 18, p.245. 164 Hunt, J.M. 1990. Generation and migration of petroleum from abnormally pressured fluid compartments: AAPG Bulletin, v.72, p.1-2. Johnson, K.S., and Cardott, B.J. 1992. Geologic framework and hydrocarbon source rocks of Oklahoma, in K.S. Johnson and B.J. Cardott, eds., Source rocks in the Southern mid continent, 1990 symposium: OGS circular 93,p.21-37. Jolly, G.D. 1988. Correlation of the Woodford Formation in South-Central Oklahoma using Gamma-ray scintillation measurements of the natural background radiation. Unpublished Master’s thesis, Stephen F.Austin State University, p. 154. Jones, S.C. 1997. A technique for faster pulse-decay permeability measurements in tight rocks. Society of Petroleum Engineers Formation Evaluation, SPE 28450,p.19-25. Kareem, M.R. 1992. Geological constrained modeling of the temporal and spatial evolution of hydrocarbon generation in the Anadarko basin: Master’s thesis, University of Oklahoma, Norman, Oklahoma, p.191. Katsube, T.J., Best, M.E., and Mudford, B.S. 1991. Petrophysical characteristics of shales from the Scotian shelf; Geophysics, v.56, 10,p.1681-8 Kwon, 0., Kronenberg, A.K., Gangi A.F., and Johnson, B. 2001. Permeability of Wilcox shale and its effective pressure law, Journal of Geophysical Research., v. 106, 19, p.339-353. Kwon, 0., Kronenberg, A.K., Gangi, A.F., Johnson, B., and Herbert, B.E. 2004. Permeability of Illite-bearing shale: 1. Anisotropy and effects of clay content and loading, Journal of Geophysical Research, B 10205, doi:10.1 029/2004JB003052. Lin, W. 1977. Compressible fluid flow through rocks of variable permeability. Lawrence Livermore Laboratory, Livermore, California, Rept. UCRL-52304. Marek, B.F. 1979. Permeability loss in depletion reservoirs. 5th Annual Conference Soc. Pet. eng., Las Vegas, NV, SPE pap.8433, PAGE. 165 Marion, D., Nur, A., and Albert, F. 1989. Modelling the relationships between sonic velocity, porosity, permeability , and shaliness in sand, shale, and shaly sand: SPWLA Thirteenth Annual Logging Symposium Proceedings, paper G, p.1-22. Meyer, R. 2002. Anisotropy of sandstone permeability, CREWES Research Report, v.14, PAGE. Meyer, R. and Krause, F.F. 2006. Permeability anisotropy and heterogeneity of a sandstone reservoir analogue. Bulletin of Canadian Petroleum Geology, v.54, 4, p. 301-3 18. O’Brien, N. 1970. The fabric of shale- An electron-microscopic study. Sedimentalogy, v.15, p. 229-246. O’Brien, N.R. and Slatt, R.M. 1990. Argillaceous Rock Atlas. New York, Springer- Verlag. Popov, M.A., Nuccio, V.F., Dyman, T.S., Gognat, T.A., Johnson, R.C., Schmoker, J.W., Wilson, M.S. and Bartberger, C. 2001. Basin-centered gas systems of the U.S.: U.S. Geological Survey Open-File Report 01-135, version 1.0 (CD-ROM). Raudsepp, M. and Pani, E. 2003. Application of rietveld analysis to environmental mineralogy, Mineralogical Association of Canada, Short Course 31. Recommended practices for core analysis (API RP 40). 1998. American Petroleum Institute, 2nd edition, PAGE. Revil, A. and Cathles, L.M., 1999. Permeability of shaly sands, Water Resour. Res., 35, p. 651—662. Rietveld, H.M. 1967. Line profiles of neuron powder-diffraction peaks for structure refinement. Acta Crystallogr. 22,p.151-2. Rietveld, H.M. 1969. A profile refinement method for nuclear and magnetic structures. J. Appl. Crystallogr., 2,p.65-71. 166 Schmoker, J.W. 1995. Method for assessing continuous-type (unconventional) hydrocarbon accumulations, in D.L. Gautier, G.L. Dolton, K.I. Takahashi, and K.L. Varnes, eds., 1995 National Assessment of United States Oil and Gas Resources- Results, methodology, and supporting data: U.S. Geological Survey Digital Data Series DDS-30 (CD-ROM). Scholes, O.N., Clayton, S.A., Hoadley, A.F.A., and Tiu, C. 2007. Permeability anisotropy due to consolidation of compressible porous media, Transp Porous Med, 68, p.365-387. Simon, D.E., Coulter, G.R., King, G., and Holmes, G. 1982. North sea completions- A laboratory study. J. Pet. Technology, p. 253 1-2536. Srodon, J. 2001. Quantitative x-ray diffraction analysis of clay-bearing rocks from random preparations, Clay and Clay Minerals, v. 49, No.6,p.51428 Sullivan, K. 1985. Organic facies variation of the Woodford shale in Western Oklahoma: Shale Shaker, v.35,p.76-89. Surdam, R.C. 1997a. A new paradigm for gas exploration in anomalously pressured “tight gas sands” in the Rocky Mountain Laramide basins, in R.C.Surdam, ed., Seals, traps, and the petroleum system: AAPG Memoir 67, p. 283-298. Taff, J.A. 1903. Description of the Tishomingo quadrangle (Indian Territory): U.S. Geological Survey Geological Atlas. Folio 98, Scale 1:125,000, p. 8. Taff, J.A. 1904. Preliminary report on the geology of the Arbuckle and Wichita mountains in Indian Territory and Oklahoma: U.S. Geological Survey professional paper3l,p. 111. Tissot B.P. and Welte, D.H. 1984. Petroleum formation and occurrence (2w’ ed.): New York, Springer-Verlag, p. 699. Trimmer, D. 1982. Laboratory measurements of ultralow permeability of geologic materials, Rev. Sci. Instrum, 53(8), p. 1246-1254. 167 Vulgamore, T., Woihart, S., Mayerhofer, M., Clawson, T, and Pope, C. 2008. Hydraulic Fracture Diagnostics Help Optimize Stimulations Of Woodford Shale Horizontals, The American Oil & Gas Reporter, PAGE. Washburn, E.W. 1921. The dynamics of capillary flow. Physical Review, 17, p. 274-307. Yang, Y. and Aplin, A.C. 2007. Permeability and petrophysical properties of 30 natural mudstones. Journal of Geophysical Research, v. 112, p. 1-14. Zoback, D.M. and Byerlee, D.J. 1975. The effect of microcrack dilatancy on permeability of Westerly granite, Journal of Geophysical Research, v. 80, p. 752-755. 168 CHAPTER-5 CONCLUSIONS 169 Chapter-5 Conclusions 5.1 CONCLUDING REMARKS Matrix permeability of Western Canadian and Woodford shales was determined using pressure pulse decay experiment. The study included understanding and evaluating the controlling factors that affect permeability. Based on the results and analysis of the experiments, the following conclusions can be drawn: 1) The pulse decay method was applicable to measure low permeability coefficients range from 1W’ md to iO md. 2) The permeability of all shales decline exponential with increasing effective stress (Wybe, 1958; Pedrosa, 1986; David et al., 1995; Best and Katsube, 1995). 3) In clay rich shales of Western Canada, the permeability decreases from 3.29E-03 md to 1.17E-04 md with the effective stress increasing from 3.45 MPa to 17.24 MPa; and in siliceous shales the permeability decreases from 8.85E-02 md to 1 .92E-02 md with the effective stress increasing from 6.9 MPa to 27.59 MPa. In calcareous shales, the permeability decreases from 3.03E-05 to 9.16E-07 md with 170 the effective stress increasing from 3.45 MPa to 27.59 MPa. The permeability loss with effective stress is more observed in calcareous shales. 4) Quartz rich Woodford shales had the permeability coefficients range from 1 02 md to iO md. 5) Mineralogy plays an important role in determining the permeability of Canadian and Woodford shales. Higher permeability was observed in samples with high clay content, and low permeability was observed with high quartz and carbonate content. 6) Samples that were tested parallel to bedding had higher permeabilities than samples tested normal to bedding. Among Western Canadian shales, the quartz rich shales showed differences of three to four orders of magnitude for the samples tested parallel to bedding, compared to those tested normal to bedding. 7) Porosimetry results on Woodford shales suggest that fluid flow is mostly observed in the meso and macro pores. Samples with higher clay content (>30%) showed a higher intrusion volume in macro pores, while samples with higher quartz content (>60%) showed intrusion volume in micro pores (Daniel, 2007). 8) Porosity was weakly correlated to permeability in the Western Canadian shales and showed a linear correlation within the Woodford shales. 9) TOC was not seen any proper correlation with the permeability of Western Canadian and Woodford shales. 171 10) Triaxial results on Woodford shales showed a brittle-ductile behavior at different confining pressures. Lithologic composition plays an important role in the strength and pore compressibility of shale. Quartz rich or carbonate rich shales had brittle behaviour and clay rich shales had ductile behaviour. Pore compressibility is greater in clay-rich shales, and less in the quartz rich shales. 5.2 FUTURE RECOMMENDATIONS The results of this study explain the factors affecting the permeability of gas shales. Other factors such as grain size, moisture content, and surface tension, that may affect permeability, remain to be investigated. Fracture permeability should be measured at different confining pressures to investigate the influence of fracture properties (fracture spacing, length, aperture, and connectivity) in the hydrocarbon production of low permeability shales. Change in fracture permeability between mineralogically distinct shales also needs to be studied. In the present work, matrix permeability was measured with helium gas. The impact of gases other than helium that adsorb during the permeability testing requires further investigation. Rock mechanical tests at different pore pressures are to be studied. Shale anisotropy at different orientations under various confining pressures is also to be studied. 172 Mercury porosimetry can measure the macro and some of the meso pores of the matrix. Other techniques such as nitrogen or carbon dioxide adsorption can also be used to measure the micro pores of the shale. These kinds of measurements provide knowledge about the entire pore size distribution of the sample. 173 5.3 REFERENCES Best, M.E., and Katsube, T.J. 1995. Shale permeability and its significance in hydrocarbon exploration, The Leading Edge, Vol. 14, Issue 3, pp. 165-170. Pedrosa, O.A. 1986. Pressure Transient Response in Stress-Sensitive Formations, Paper No. SPE 15115, SPE California Regional Meeting, Oakland, California. Ross, D.J.K. 2007. Investigation into the importance of geochemical and pore structure heterogeneities for shale gas reservoir evaluation. PhD thesis, unpublished. Wyble, D.O. 1958. Effect of Applied Pressure on the Conductive Porosity and Permeability of Sandstones, Transactions AIME, Vol.2 13, pp. 430-32. 174

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0052381/manifest

Comment

Related Items