INFORMATION TO USERS This manuscript has been reproduced from the microfihn master. UME films the text directly fi'om the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter &ce, while others may be from ai^ type of computer printer. The quality^ of this reproduction is dependent upon tiie quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect rqnoduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, b^irming at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book. Photographs included in the original manuscript have been reproduced xerographically in this copy. Kgher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. UMI A Bell & Howell Information Company 300 North Zed> Road, Ann Arbor MI 48106-1346 USA 313/761-4700 800/521-0600 SINGLE- AND CROSS-HOLE PNEUMATIC INJECTION TESTS IN UNSATURATED FRACTURED TUFFS AT THE APACHE LEAP RESEARCH SITE NEAR SUPERIOR, ARIZONA by Walter Arthur Illman (Sho Yamamoto) Copyright © Walter Arthur Illman 1999 A Dissertation Submitted to the Faculty of the DEPARTMENT OF HYDROLOGY AND WATER RESOURCES In Partial Fulfillment of the Requirements For the Degree of DOCTOR OF PHILOSOPHY WITH A MAJOR IN HYDROLOGY In the Graduate College THE UNIVERSITY OF ARIZONA 1999 DMX Niunber: 9927499 Copyright 1999 by Illmsm, Walter Arthur All rights resorved. UMI Microform 9927499 Copyright 1999, by UMI Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. UMI 300 North Zeeb Road Ann Ari>or, MI 48103 2 THE UNIVERSITY OF ARIZONA « GRADUATE COLLEGE As members of the Final Examination Committee, we certify that we have read the dissertation prepared by entitled Walter Arthur Illman Single- and Cross-hole Pneumatic Injection Tests in Unsaturated Fractured Tuffs at the Apache Leap Research Site near Superior, AZ and recommend that it be accepted as fulfilling the dissertation requirement for the Degree of Shlomo Doctor of Philosophy Da^ Neuman ' D a t e ^ ^ Peter J. Wierenga 2 Date G. Zreda Dona.m E. Mye Date Arthur W. Warrick Date Final approved, and acceptance of this dissertation is contingent upon the candidate's submission of the final copy of the dissertation to the Graduate College. I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation rei Dissertation Director Shlomo P. Neuman 'ate 3 STATEMENT BY AUTHOR This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at the University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library. 3rief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgement of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the copyright holder. SIGNED: [ 4 ACKNOWLEDGEMENTS I am indebted to Regents' Professor Shiomo Peter Neuman, my dissertation director and finend for iiis guidance, support, patience and time taken from his very busy schedule. I will never forget his lectures (HWR 503, 504, and 603), conference presentations, numerous discussions on various topics pertaining to hydrogeology, and his commitment to teaching and research. He always took time to listen to my problems, whether it pertained to academics, research, or personal matters. His skills in challenging students to improve their understanding of hydrogeology/mathematics are urunatched and commendable. I would like to thank Professors Peter J. Wierenga, Donald E. Myers, Arthur W. Warrick, and Marek G. Zreda for serving on my dissertation committee. I would also like to thank Professor James E. Smith for his personal advice in teaching and research. I wish to acknowledge with gratitude the unwavering and effective support, advice and encouragement provided to me throughout this study by our NRC Project Manager, Mr. Thomas J. Nicholson, without whom this work would never have materialized. I would also like to thank Mr. Dick L. Thompson, for his commitment and excellence as a field technician during my field project. Our success in conducting crosshole pneumatic tests and its substantial cost-savings were accomplished through his ingenious inventions and persistence in the field. I must acknowledge and thank Dr. Daniel M. Tartakovsky for being my good friend, improving my understanding of mathematics, and teaching me the simplicity and beauty of analytical solutions. My parents (William and Hatsue Illman) must be acknowledged for their support throughout my life and studies. Difficult times encountered throughout my research and studies at the University of Arizona were always resolved with telephone calls to Japan. Finally, I would like to thank Ms. Carla Marie Frascari for providing me emotional support during my last year at the University of Arizona. It has been very difficult this last year, but her love, positive attitude, kindness, forgiveness and patience is leading me to become a better person. The majority of this research was supported by the United States Nuclear Regulatory Commission under contracts NRC-04-95-038 and NRC-04-97-056. It was also supported in part by a National Science Foundation Graduate Traineeship during 1994-1995, a University of Arizona Graduate College Fellowship during 1997-1998, the Horton Doctoral Research Grant from the American Geophysical Union during 19971998, and the John and Margaret Harshbarger Doctoral Fellowship from the Department of Hydrology and Water Resources at The University of Arizona during 1998-1999. 5 Those who understand classical music have a good grasp of mathematics. (translated from Japanese) Hatsue Yamamoto Illman Dreaming of success leads to success. One must also remember that success is a result of diligence, patience, honesty, and persistence. William Richard Illman 6 To My Carla 7 TABLE OF CONTENTS STATEMENT BY AUTHOR 3 ACKNOWLEDGEMENTS 4 TABLE OF CONTENTS 7 LIST OF ILLUSTRATIONS 10 LIST OF TABLES 19 LIST OF PLATES 21 ABSTRACT 22 1. INTRODUCTION 24 I. I ROLE OF PNEUMATIC TESTS IN SITE CHARACTERIZATION 1.2 THE ALRS AND PREVIOUS SITE INVESTIGATIONS 1.2.1 Site Description 1.2.2 Previous Work in Study Area of ALRS 1.2.3 Single-Hole Pneumatic Injection Tests 1.2.4 Kriged Estimates of Air Permeability at the ALRS 1.2.5 Geophysical Surveys Conducted at Site 1.3 TYPE-CURVE ANALYSIS OF SINGLE-HOLE PNEUMATIC INJECTION TESTS 1.4 CROSS-HOLE PNEUMATIC INJECTION TESTS 1.5 SCOPE OF THE DISSERTATION 2. GAS FLOW IN POROUS AND FRACTURED MEDIA 2.1 THEORY OF GAS FLOW IN POROUS MEDIA 2.1.1 Literature Review 2.1.2 Two-Phase and Single-Phase Representations 2.1.3 Linearization of the Gas Flow Equation 3 TYPE-CURVE MODELS FOR THE INTERPRETATION OF SINGLEHOLE TEST DATA 3.0 CONCEPTUALIZATION OF FLOW THROUGH FRACTURED ROCKS 3.1 SPHERICAL GAS FLOW MODEL 3.2 RADIAL GAS FLOW MODEL 3.3 UNIFORM FLUX HORIZONTAL AND VERTICAL FRACTURES 4 TYPE-CURVE INTERPRETATION OF SINGLE-HOLE PNEUMATIC INJECTION TEST DATA 24 26 26 28 44 49 53 61 62 64 65 65 67 71 78 80 80 85 88 94 101 8 TABLE OF CONTENTS - Continued 4.1 SINGLE-HOLE TEST METHODOLOGY 4.2 PHENOMENOLOGY OF SINGLE-HOLE TESTS 4.3 TYPE-CURVE INTERPRETATION OF SINGLE-HOLE TESTS 5 TYPE-CURVE MODELS FOR THE INTERPRETATION OF CROSSHOLE TEST DATA 5.1 5.2 5.3 5.4 5.5 5.6 6 129 INTRODUCTION 129 POINT-INJECTION/POINT-OBSERVATION 133 LINE-INJECTION/POINT-OBSERVATION 137 POINT-INJECTION/LINE-OBSERVATION 142 LINE-INJECTION/LINE-OBSERVATION 145 MODIFIED FORM OF HSIEH AND NEUMAN (1985a) SOLUTION TO ACCOUNT FOR OBSERVATION WELLBORE STORAGE AND SKIN EFFECTS 148 TYPE-CURVE INTERPRETATION OF CROSS-HOLE PNEUMATIC TEST DATA 6.1 CROSS-HOLE TEST METHODOLOGY 6.1.1 Instrumentation Used In Cross-Hole Tests 6.2 CROSS-HOLE TESTING PROCEDURE WITH EMPHASIS ON TEST PP4 6.3 TYPE-CURVE INTERPRETATION OF CROSS-HOLE TEST PP4 7 101 107 115 FINDINGS AND CONCLUSIONS APPENDIX A; ASSESSMENT OF ELECTRICAL RESISTIVITY TOMOGRAPHY (ERT) IN OBTAINING THE SUBSURFACE DISTRIBUTION OF WATER CONTENT IN THE VADOSE ZONE 157 157 158 171 189 211 218 A.1 INTRODUCTION-APPLICATIONS TO ENVIRONMENTAL MONITORING 218 A.2 THEORY AND CONCEPTS 220 A.3 INVERSION OF TRANSFER RESISTANCES - VARIOUS RECONSTRUCTION ALGORITHMS 221 A.4 APPLICATION OF ERT TO MONITOR VADOSE ZONE WATER MOVEMENT 225 A.5 DISCUSSION OF RESULTS: RELATIVE MERITS OF APPLYING ERT TO MONITOR VADOSE ZONE WATER MOVEMENT 229 A.6 REFERENCES PERTAINING TO APPENDIX A 234 APPENDIX B: NEUTRON PROBE METHODOLOGY B.l INTRODUCTION B.2 THEORY BEHIND THE NEUTRON ATTENUATION METHOD 237 237 238 9 TABLE OF CONTENTS - Continued B.3 THE STANDARD CALIBRATION METHOD OF NEUTRON PROBE IN GRANULAR MATERIALS 241 B.4 FACTORS THAT AFFECT NEUTRON COUNTS 243 B.5 REFERENCES PERTAINING TO APPENDIX B 248 APPENDIX C; SLIP FLOW 250 APPENDIX D: AIR COMPRESSIBILITY AND VISCOSITY IN RELATION TO PRESSURE AND TEMPERATURE AT THE ALRS 256 DERIVATION OF SINGLE-PHASE GAS FLOW EQUATION 259 APPENDIX F: TYPE-CURVE SOLUTION OF SPHERICAL GAS FLOW 262 APPENDIX G: CODE FOR NUMERICAL INVERSION OF LAPLACE TRANSFORM 269 CODE FOR NUMERICAL DIFFERENTIATION OF PNEUMATIC TEST DATA 278 MODIFICATION OF HSEEH AND NEUMAN (1985a) SOLUTION TO ACCOUNT FOR STORAGE AND SKIN IN MONITORING INTERVALS 282 RECORD OF BOREHOLE TELEVISION (BHTV) SURVEY CONDUCTED AT THE ALRS 285 APPENDIX E: APPENDIX H: APPENDIX I: APPENDIX J: REFERENCES 298 10 LIST OF ILLUSTRATIONS Figure 1.1; Location map of the covered site (indicated as red star) 30 Figure 1.2: Plan view of boreholes, plastic cover, and field laboratory at ALRS 33 Figure 1.3; Three-dimensional perspective of the boreholes at the site 34 Figure 1.4: Lower hemisphere Schmidt equal-area projection of fractures identified by Rasmussen et al. (1990). Contours indicate number of fi^actures per unit area of projection circle 39 Geometric parameters associated with a single-hole hydraulic injection test (adapted from Tidtwell, 1988) 42 Air permeability versus fi:acture density (data fi^om Rasmussen elal. 1990) 43 Perspective toward the Northeast showing center locations of 1m single-hole pneumatic test intervals; overlapping circles indicate re-tested locations (adapted fi-om Guzman et al. 1996) 48 Kriged log k estimates obtained using \-m scale data from boreholes X2, Y2, Y3, Z2, V2, and W2A (left) and same together with 3-m scale d a t a firom boreholes X I , X3, YI, Z I , and Z3 (right) (adapted from Illman et al. 1998) 51 Three-dimensional representation of kriged log k. (adapted from Illman et al. 1998) 52 Image of cross-hole, electrical resistivity tomography (ERT) obtained between boreholes VI and V3 56 Variation of volumetric water content with depth in W-series boreholes 57 Variation of volumetric water content with depth in X-series boreholes 57 Variation of volumetric water content with depth in Y-series boreholes 58 Variation of water content with depth in Z-series boreholes 58 Figure 1.5: Figure 1.6: Figure 1.7: Figure 1.8: Figure 1.9: Figure 1.10: Figure 1.11: Figure 1.12: Figure 1.13: Figure 1.14: 11 LIST OF ILLUSTRATIONS - Continued Figure 1.15: Histogram of volumetric water content values from the W-, X-, Y-, and Z-series boreholes 59 Figure 1.16; EM survey conducted along borehole Y2 60 Figure 2.1; Variation of compressibility factor Z with pressure and temperature 74 Figure 2.2: Variation of air viscosity fj. *vith pressure and temperature 75 Figure 2.3: Variation of fiZ with pressure and temperature 75 Figure 3.1: Type-curves of dimensionless pseudopressure in injection interval versus normalized dimensionless time for various Co and 5 = 0 under spherical flow 91 Type-curves of dimensionless pseudopressure and its derivative in injection interval versus dimensionless time for various Co and s under spherical flow 92 Type-curves of dimensionless pseudopressure and its derivative in injection interval versus dimensionless time for various Co under radial flow 93 Geometry of idealized horizontal fracture in an infinite flow domain 98 Type-curves of dimensionless pseudopressure in injection interval, normalized by dimensionless height, versus dimensionless time for various ho in uniform flux horizontal fracture model 99 Figure 3.2: Figure 3.3: Figure 3.4: Figure 3.5: Figure 3.6: Figure 4.1: Figure 4.2: Figure 4.3: Type-curves of dimensionless pseudopressure and its derivative in injection interval versus dimensionless time in uniform flux vertical fracture model at center of fracture (x^ = =0) 100 Schematic diagram of the air injection system (adapted from Guzman et al. 1996) 105 Example of multi-rate single-hole test (VCClOOl in borehole V2 at 10.37 m from LL marker) 111 Arithmetic plot of pressure data from CGAl 120 112 12 LIST OF ILLUSTRATIONS - Continued Figure 4.4: Logarithmic plot of pressure data CGAI120 112 Figure 4.5: Logarithmic plot of pressure data from test JGA0616 113 Figure 4.6: Logarithmic plot of pressure data from test JHB0612 113 Figure 4.7: BHTV image taken in borehole Y2 (high permeability zone) 114 Figure 4.8: Logarithmic plot of pressure data from test ZDC0826 114 Figure 4.9a: Type-curve match of data from single-hole pneumatic test CAC0813 with liquid, spherical flow model 120 Type-curve match of data from single-hole pneumatic test CAC0813 with gas, spherical flow model 120 Figure 4. lOa: Type-curve match of data from single-hole pneumatic test CHB0617 with liquid, spherical flow model 121 Figure 4. lOb: Type-curve match of data from single-hole pneumatic test CHB0617 with gas, spherical flow model 121 Figure 4.9b: Figure 4.11: Figure 4.12: Figure 4.13: Scatter diagram of results of permeability obtained from steadystate and /7-based spherical models 122 Scatter dia^am of results of permeability obtained from steadystate and -based spherical models 122 Scatter diagram of results of permeability obtained from steadystate and p-znd p'-based spherical models 123 Figure 4.14a: Type-curve match of data from single-hole pneumatic test JGA0605 with liquid, radial flow model 124 Figure 4.14b: Type-curve match of data from single-hole pneumatic test JGA0605 with gas, radial flow model 124 Figure 4.15a: Type-curve match of data from single-hole pneumatic test JJA0616 with gas, spherical flow model 125 Figure 4.15b: Type-curve match of data from single-hole pneumatic test JJA0616 with liquid, radial flow model 125 13 LIST OF ILLUSTRATIONS - Continued Figure 4.16: Figure 4.17; Figure 4.18: Figure 4.19: Figure 4.20: Figure 4.21: Figure 5.1: Figure 5.2: Figure 5.3: Figure 5.4: Figure 5.5: Type-curve match of data from single-hole pneumatic test J T O 0 6 I 2 t o the horizontal fracture model 126 Type-curve match of data from single-hole pneumatic test JHB0612 to the vertical fracture model 126 Type-curve match of data from single-hole pneumatic test CDB1007 to the gas, radial flow model 127 Type-curve match of data from single-hole pneumatic test JNA0713 to the gas, radial flow model 127 Type-curve match of data from single-hole pneumatic test JKC0625 to the gas, spherical flow model 128 Type-curve match of data from single-hole pneumatic test YFB0621 to the gas, spherical flow model 128 The point source solution (solid curve) and its time derivative (dashed curve) 136 Logarithmic plot of (solid curves) and d (dashed curves) versus for a range of or, values. The Theis (1935) solution (open circles) and its derivative (open triangles) are included for purposes of comparison 140 Plot of normalized pressure head for the line-injection/pointobservation case (solid curves) and its derivatives (dashed curves) plotted against . The TT/c/j (1935) solution (open circles) and its derivative (open triangles) are included for purposes of comparison 141 Logarithmic plot of(solid curves) and (dashed curves) versus where the point injection/pointmonitoring solution (open circles) and its derivative (open triangles) have also been plotted 144 Logarithmic plot for a, = yS, = 0 for various values of or, = ySJ. The Theis (1935) solution (open circles) and its derivative (open triangles) are included for purposes of comparison 147 14 LIST OF ILLUSTRATIONS - Continued Figure 5.6: Type-curves of dimensionless pressure in observation interval (solid curves) and its derivative (dashed curves) versus dimensionless time for various Q and fit = 5.0, ySo = 0.01. Open circles represent dimensionless pressure in the solution of Hsieh and Neuman (1985a) and open triangles represent their derivatives 152 Type-curves of dimensionless pressure in observation interval (solid curves) and its derivative (dashed curves) versus dimensionless time for various Q and Pi = OA, 0.01. Open circles represent dimensionless pressure in the solution of Hsieh and Neuman (1985a) and open triangles represent their derivatives 153 Type-curves of dimensionless pressure in observation interval (solid curves) and its derivative (dashed curves) versus dimensionless time for various and Pi = 0.01, ^2= 0.01. Open circles represent dimensionless pressure in the solution of Hsieh and Neuman (1985a) and open triangles represent their derivatives 154 Type-curves of dimensionless pressure in observation interval (solid curves) and its derivative (dashed curves) versus dimensionless time for various Q and or/ = 0.01, orj = 0.01. The red solid line represents dimensionless pressure in the solution of Hsieh and Neuman (1985a) and red dashed line their derivatives. The open circles represent the Theis (1935) solution while the open triangles represent their derivatives 155 Type-curves of dimensionless pressure in observation interval (solid) and its derivative (dashed) versus dimensionless time for various H and ai = 0.01, a2= 0.01. The blue solid line represents dimensionless pressure in the solution of Theis (1935) and the blue dashed line their derivatives. The open circles represent the Black andKipp (1977) solution while the open triangles represent their derivatives 156 Figure 6.1: Injection and monitoring systems used for cross-hole tests 163 Figure 6.2: Air injection system installed in the field laboratory 169 Figure 5.7: Figure 5.8: Figure 5.9: Figure 5.10: 15 LIST OF ILLUSTRATIONS - Continued Figure 6.3: Locations of centers of injection and monitoring intervals. Large solid circle represents injection interval, small solid circles represent short monitoring intervals, and open circles represent long monitoring intervals 174 Figure 6.4: Barometric pressure during cross-hole test PP4 178 Figure 6.5: Flow rate during cross-hole test PP4 178 Figure 6.6: Packer pressure during cross-hole test PP4 179 Figure 6.7: Battery voltage during cross-hole test PP4 179 Figure 6.8: Fluctuations in relative humidity during cross-hole test PP4 180 Figure 6.9: Surface injection pressure and pressure in injection (Y2M) and monitoring (Y2U, Y2B) intervals within borehole Y2 during cross-hole test PP4 180 Figure 6.10: Pressure in monitoring interval VI during cross-hole test PP4 181 Figure 6.11: Pressure in monitoring intervals V2U, V2M, and V2B during cross-hole test PP4 181 Pressure in monitoring intervals V3U, V3M, and V3B during cross-hole test PP4 182 Figure 6.13: Pressure in monitoring interval W1 during cross-hole test PP4 182 Figure 6.14: Pressure in monitoring interval W2 during cross-hole test PP4 183 Figure 6.15: Pressure in monitoring intervals W2AU, W2AM, W2AL, and W2AB during cross-hole test PP4 183 Pressure in monitoring intervals W3U, W3M, and W3B during cross-hole test PP4 184 Figure 6.17: Pressure in monitoring interval XI during cross-hole test PP4 184 Figure 6.18: Pressure in monitoring intervals X2U, X2M, and X2B during cross-hole test PP4 185 Pressure in monitoring interval X3 during cross-hole test PP4 185 Figure 6.12: Figure 6.16: Figure 6.19: 16 LIST OF ILLUSTRATIONS - Continued Figure 6.20; Pressure in monitoring intervals YIU, YIM, and YIB during cross-hole test PP4 186 Pressure in monitoring intervals Y3U, Y3M, and Y3B during cross-hole test PP4 186 Figure 6.22: Pressure in monitoring interval Zl during cross-hole test PP4 187 Figure 6.23: Pressure in monitoring intervals Z2U, Z2M, Z2L, and Z2B during cross-hole test PP4 187 Pressure in monitoring intervals Z3L, Z3M, and Z3B during cross-hole test PP4 188 Figure 6.25: Type-curve match of pressure data from monitoring interval VI 195 Figure 6.26: Type-curve match of pressure data from monitoring interval V2M 195 Type-curve match of pressure data from monitoring interval V3U 196 Type-curve match of pressure data from monitoring interval V3M 196 Type-curve match of pressure data from monitoring interval V3B 197 Figure 6.30: Type-curve match of pressure data from monitoring interval Wl 197 Figure 6.31: Type-curve match of pressure data from monitoring interval W2AU 198 Type-curve match of pressure data from monitoring interval W2AM 198 Type-curve match of pressure data from monitoring interval W2AL 199 Type-curve match of pressure data from monitoring interval W2AB 199 Type-curve match of pressure data from monitoring interval W3U 200 Figure 6.21: Figure 6.24: Figure 6.27: Figure 6.28: Figure 6.29: Figure 6.32: Figure 6.33: Figure 6.34: Figure 6.35: 17 LIST OF ILLUSTRATIONS - Continued Figure 6.36: Type-curve match of pressure data from monitoring interval W3M 200 Figure 6.37; Type-curve match of pressure data from monitoring interval XI 201 Figure 6.38: Type-curve match of pressure data from monitoring interval X2U 201 Type-curve match of pressure data from monitoring interval X2M 202 Type-curve match of pressure data from monitoring interval X2B 202 Type-curve match of pressure data from monitoring interval YIU 203 Type-curve match of pressure data from monitoring interval YIM 203 Type-curve match of pressure data from monitoring interval Y2U 204 Type-curve match of pressure data from monitoring interval Y2B 204 Type-curve match of pressure data from monitoring interval Y3U 205 Type-curve match of pressure data from monitoring interval Y3M 205 Type-curve match of pressure data from monitoring interval Y3B 206 Figure 6.48: Type-curve match of pressure data from monitoring interval Z1 206 Figure 6.49: Type-curve match of pressure data from monitoring interval Z2U 207 Type-curve match of pressure data from monitoring interval Z2M 207 Figure 6.39: Figure 6.40: Figure 6.41: Figure 6.42: Figure 6.43: Figure 6.44: Figure 6.45: Figure 6.46: Figure 6.47: Figure 6.50: 18 LIST OF ILLUSTRATIONS - Continued Figure 6.51; Type-curve match of pressure data from monitoring interval Z2L 208 Type-curve match of pressure data from monitoring interval Z2B 208 Type-curve match of pressure data from monitoring interval Z3U 209 Type-curve match of pressure data from monitoring interval Z3M 209 Type-curve match of pressure data from monitoring interval Z3B 210 Relative frequency of equivalent mean pore diameter plotted on semi-logarithmic scale, showing bi modal distribution of pores in tuff matrix at the ALRS (data from Rasmussen et al. 1990) 255 Figure F.1: Mass conservation in injection system 267 Figure F.2: Impact of positive skin on pseudopressure around the injection interval 268 Figure 6.52: Figure 6.53; Figure 6.54: Figure 6.55: Figure C.1: 19 LIST OF TABLES Table 1.1: Statistics of effective porosity measurements on ALRS core. Medium size sample measurements were obtained using mercury intrusion method which underestimates effective porosity (after Rasmussen et al. 1990) 36 Statistics of hydraulic conductivity values were obtained from 105 large core samples at the ALRS (after Rasmussen et al. 1990) 36 Table 1.3: Statistics for van Genuchten's a (after Rasmussen et al. 1990) 36 Table 1.4; Statistics of laboratory determined air permeabilities for various values of suction and at oven-dried (OD) conditions (after Rasmussen et al. 1990) 37 Statistics of laboratory determined Klinkenberg slip flow coefficient (after Rasmussen et al. 1990) 37 Statistics of fracture density and orientation obtained from ALRS cores (after Rasmussen et al. 1990) 37 Statistics of field determined outflow rates and of saturated hydraulic conductivity values at the ALRS (after Rasmussen et al., 1990) 38 Statistics of 3-m scale air permeability data. One interval exceeded the capacity of the measurement device, while a second interval was less than the measurement threshold (after Rasmussen et al. 1990) 38 Field water content on various days at the ALRS (after Rasmussen et al. 1990) 38 Table 1.2; Table 1.5; Table 1.6; Table 1.7; Table 1.8; Table 1.9; Table 4.1; Table 4.2; Table 4.3; Table 6.1; Coordinates of boreholes subjected to air permeability testing (adapted from Guzman et al. 1996) 103 Nominal Scale and Number of Single-Hole Pneumatic Injection Tests at the ALRS.(adapted from Guzman et al. 1996) 103 Air permeabilities obtained usingp- and p"-based spherical flow models 119 Cross-hole tests completed at the ALRS 160 20 LIST OF TABLES - Continued Table 6.2; Table 6.3: Table 6.4: Table 6.5: Table C.1: Table D. 1: Table D.2: Table D.3: Table D.4: Information on injection and monitoring intervals with pressure transducer types during phase 3 cross-hole tests 164 Coordinates of centers of monitoring intervals relative to origin at center of injection interval, interval lengths, radial distances between centers of injection and monitoring intervals, geometric parameters Pi and ^2 and maximum recorded pressure change 177 Pneumatic parameters obtained from type-curve analysis of pressure buildup data collected during cross-hole test PP4 193 Sample statistics of directional air permeabilities and air-filled porosities obtained from type-curve interpretation of cross-hole test PP4, and of air permeabilities from steady state interpretations of l-m scale single-hole tests. Numbers in parentheses represent corresponding actual values 194 Cumulative mercury intrusion volume as a function of equivalent pore diameter (adapted from Rasmussen et al. 1990) 254 Gaseous composition of U. S. standard atmosphere {CRC Handbook, 1992-1993) 257 Data concerning stable pressures encountered during \-m scale single-hole pneumatic injection tests at the ALRS (summarized from Guzman et al. 1996) 257 Data concerning stable temperatures encountered during \-m scale single-hole pneumatic injection tests at the ALRS (summarized from Guzman et al. 1996) 257 Variation of air dynamic viscosity with temperature and pressure (data from Vasserman et al. 1966) 258 21 LIST OF PLATES Plate 1.1; Photograph of the Apache Leap 31 Plate 1.2: Photograph of the covered borehole site (CBS) at ALRS where pneumatic testing program took place 32 Photograph of the single-hole packer system employed during the study of Guzman et al. (1996) 106 Plate 6.1; Photograph of the single packer monitoring system 161 Plate 6.2: Photograph of the modular packer system 162 Plate 6.3: Photograph of the air manifold installed at the site 167 Plate 6.4: Photograph of the injection interval with neutron probe installed 168 Photograph of the air injection system and electronic equipment installed in the field laboratory 170 Plate 4.1: Plate 6.5: 22 ABSTRACT This dissertation documents research results from a series of field experiments and analyses used to test interpretive models for investigating the role of fractures in fluid flow through unsaturated, fractured tuffs. It summarizes the experimental design of single- and cross-hole pneumatic injection tests, including borehole configuration and testing schedules, data collection system, interpretive models developed and tested, data, and conclusions. Single-hole tests were interpreted by Guzman et al. (1996) by means of steady-state analysis to obtain permeability values based solely on late pressure data. This dissertation and Illman et al. (1998) employ pressure and pressure-derivative typecurves to analyze transient data. Air permeabilities determined from transient analyses agree well with those derived from steady-state analyses. Cross-hole pneumatic tests were analyzed by means of a graphical matching procedure using newly-developed pressure and pressure-derivative type-curves. Analyses of pressure data from individual monitoring intervals using these new type-curves, under the assumption that the rock acts as a uniform and isotropic fractured porous continuum, yield results that are comparable with parameters obtained from a numerical inverse procedure described in Illman et al. (1998). The results include information about pneumatic connections between the injection and monitoring intervals, corresponding directional air permeabilities, and airfilled porosities. Together with the results of earlier site investigations, single- and crosshole test analyses reveal that at the Apache Leap Research Site in central Arizona: (I) the pneumatic pressure behavior of fractured tuff is amenable to analysis by methods that treat the rock as a continuum on scales ranging from meters to tens of meters; (2) this 23 continuum is representative primarily, but not exclusively, of interconnected fractures; (3) its pneumatic properties vary strongly with location, direction and scale, in particular, the mean of pneumatic permeabilities increases, and their variance decreases with scale; (4) this scale effect is most probably due to the presence in the rock of various size fractures that are interconnected on a variety of scales; and (5) given a sufficiently large sample of spatially varying pneumatic rock properties on a given scale of measurement, these properties are amenable to analysis by geostatistical methods, which treat them as correlated random fields defined over a continuum. 24 1. INTRODUCTION 1.1 ROLE OF PNEUMATIC TESTS IN SITE CHARACTERIZATION Issues associated with the site characterization of fractured rock terrains, the analysis of fluid flow and contaminant transport in such terrains and the efficient handling of contaminated sites are typically very difficult to resolve. A major source of this difficulty is the complex nature of the subsurface "plumbing systems" of pores and fractures through which flow and transport in rocks take place. There is at present no well-established field methodology to characterize the fluid flow and contaminant transport properties of unsaturated fractured rocks. In order to characterize the ability of such rocks to conduct water, and to transport dissolved or suspended contaminants, one would ideally want to observe these phenomena directly by conducting controlled field hydraulic injection and tracer experiments within the rock. In order to characterize the ability of unsaturated fractured rocks to conduct non-aqueous phase liquids such as chlorinated solvents, one would ideally want to observe the movement of such liquids under controlled conditions in the field. In practice, there are severe logistical obstacles to the injection of water into unsaturated geologic media, and logistical as well as regulatory obstacles to the injection of non-aqueous phase liquids. There also are important technical reasons why the injection of liquids, and dissolved or suspended tracers, into fractured rocks may not be the most feasible approach to site characterization when the rock is partially saturated with water. Injecting liquids and dissolved or suspended tracers into an unsaturated rock would cause them to move predominantly 25 downward under the influence of gravity, and would therefore yield at best limited information about the ability of the rock to conduct liquids and chemical constituents in directions other than the vertical. It would further make it difficult to conduct more than a single test at any location because the injection of liquid modifies the ambient saturation of the rock, and the time required to recover ambient conditions may be exceedingly long. Many of these limitations can be overcome by conducting field tests with gases rather than with liquids, and with gaseous tracers instead of chemicals dissolved in water. Experience with pneumatic injection and gaseous tracer experiments in fractured rocks is limited; much of this experience has been accumulated in recent years by The University of Arizona at the Apache Leap Research Site (ALRS) near Superior, Arizona and by the U.S. Geological Survey (USGS) near the ALRS {LeCain, 1995), and at Yucca Mountain in Nevada {LeCain, 1996a,b; LeCain, 1998). Two research groups fi-om the Lawrence Berkeley National Laboratory (LBNL) also conducted numerous air injection tests at Yucca Mountain {Wang et aL 1998; Cook and Wang 1998) and in unsaturated fractured basalts at Box Canyon in Idaho {Benito et al. 1998). To our knowledge, the earliest pneumatic injection tests were conducted by Boardman and Skrove (1966) to determine fracture permeability following a contained nuclear explosion. Their analysis was based on the assumption of steady-state, isothermal, radial flow of ideal gases. Other early testing of fractured rocks with air is described by Montazer (1982), Trauz (1984), Mishra et aL (1987), and Guzman et aL (1996). 26 This dissertation focuses on single-hole and cross-hole pneumatic injection tests conducted by our group at the ALRS under the auspices of the U.S. Nuclear Regulatory Commission (NRC). These tests were part of confirmatory research in support of NRC's role as the licensing agency for a potential high-level nuclear waste repository in unsaturated fractured tuffs at Yucca Mountain. However, unsaturated fi-actured porous rocks similar to tuffs are found at many locations, including some low-level radioactive waste disposal sites, nuclear decommissioning facilities and sites contaminated with radioactive as well as other hazardous materials. The test methodologies we have developed, and the understanding we have gained concerning the pneumatic properties of tuffs at the ALRS, are directly relevant to such facilities and sites. It should be stated that test methodologies developed by us for unsaturated fi-actured rocks are equally applicable to unsaturated porous media (sand, gravel, fractured clays and alluvium). 1.2 THE ALRS AND PREVIOUS SITE INVESTIGATIONS 1.2.1 Site Description The Apache Leap Research Site is situated near Superior in central Arizona, approximately 160 km north of Tucson at an elevation of 1,200 m above sea level (Figure 1.1-covered site indicated with a red star). The site is similar in many respects to Yucca Mountain in southern Nevada where the candidate high-level nuclear waste (HLW) repository site is being characterized. ALRS is located near the extreme western edge of the Pinal Mountains. Lying immediately east of Superior, Arizona, is the Apache Leap, which forms a 600-/n west-facing escarpment that exposes a volcanic zoned ash-flow tuff 27 sheet and an underlying carbonate stratum (see Plate 1.1). The dacite ash-flow sheet {Peterson, 1961) covers an area of 1,000 km^ and varies considerably in thickness with an average of about 300 m. The tuff is a consolidated deposit of volcanic ash with particle diameters of less than 0.4 mm, resulting from a turbulent mixture of gas and pyroclastic materials at high temperature about 19 million years ago. The climate is temperate and dry, with a mean annual precipitation of less than 50 cm. Most of the precipitation occurs during two periods, from mid-July to late-September and from mid-November to lateMarch. During periods of high temperature and evapotranspiration demand in the summer, rain is characterized by high intensity, short duration thunderstorms. During cooler periods with much lower evapotranspiration demand in the winter, storms are of longer duration and lower intensity. The regional water table lies at a variable depth of more than 600 m. Except for a relatively thin perched zone of saturation at a depth of approximately 150 m, the rock above the water table is unsaturated. The covered site (Plate 1.2) under investigation consists of a cluster of 22 vertical and inclined (at 45") boreholes that have been completed to a maximum vertical depth of 30 m within a layer of slightly welded unsaturated tuff. A plan view of the boreholes is shown in Figure 1.2 and a three-dimensional perspective in Figure 1.3. Recently surveyed wellhead locations in Figure 1.2, and borehole geometries in Figure 1.3, are given with references to Cartesian coordinates (x, y, z) with origin at the lower lip of the casing in borehole Z3 and a vertical r-axis pointing downward. Boreholes having the designations X, Y and Z were drilled during the initial stages of the project, prior to those designated V, W and G. Shortly after the completion of drilling, a surface area of 1500 28 m' that includes all boreholes was covered with a thick plastic sheet to minimize infiltration and evaporation. The V, W, X, Y and Z boreholes span a surface area of approximately 55 OT by 35 m and a volume of rock on the order of 60,000 m^. The vertical G boreholes were drilled with conventional rotary equipment using water as cooling fluid; are about 20 m deep; and lie to the west of the plastic cover. A total of 270 m of oriented core was retrieved fi'om the boreholes and stored at the University of Arizona Core Storage Facility. The upper 1.8 m of each borehole was cased. Borehole television images are available for boreholes VI, V2, V3, Wl, W2, W2A, W3, XI, X2, X3, Yl, Y2, Y3, ZI, Z2 and Z3. 1.2.2 Previous Work in Study Area of ALRS Earlier work related to our area of study at the ALRS is described by Evans (1983), Schrauf and Evans (1984), Huang and Evans (1984), Green and Evans (1987), Rasmussen and Evans (1987, 1989, 1992), Yeh et al. (1988), Weber and Evans (1988), Chuang et al. (1990), Rasmussen et al. (1990, 1993, 1996), Evans and Rasmussen {\99\), and Bassett et al. (1994). The earlier work included drilling 16 boreholes (VI, V2, V3, Wl, W2, W2A, W3, XI, X2, X3, Yl, Y2, Y3, Zl, Z2, and Z3) and conducting numerous field and laboratory investigations. Laboratory measurements of matrix properties were conducted on core segments of various sizes taken from 3-m borehole intervals at 105 locations (indicated in Figure 5 of Evans and Rasmussen, 1991) within nine of the boreholes (XI, X2, X3, Yl, Y2, Y3, ZI, Z2, Z3). The measurements include interstitial properties such as bulk density (Table 1 in Rasmussen et al. 1990), effective porosity 29 (Table 1.1), skeletal density (Table 3 in Rasmussen et al. 1990), pore surface area (Table 4 in Rasmussen et al. 1990), and pore size distribution (Table 5 in Rasmussen et al. 1990); hydraulic properties such as saturated and unsaturated hydraulic conductivity (Table 1.2) and moisture retention characteristics (Table 8 in Rasmussen et al. 1990); a value of the van Genuchten moisture characteristic function (Table 1.3); pneumatic properties such as oven-dry and unsaturated air-phase permeability (Table 1.4); and Klinkenberg slip-flow coefficient (Table 1.5). Moisture retention properties of the matrix were characterized by the van Genuchten (1980) parzuneter a while holding other parameters of the van Genuchten retention model constant: the dimensionless parameter n at 1.6, and the residual water content Or at zero {Rasmussen et al. 1990). 30 Figure l.l: Locatioii map of the covered sfte (indicated as red star) 31 Pbte 1.1: Photogci^hofthe Apache Leap 32 Plate 1.2: Pfaotogci^ of the coveted boidioie sfte (CBS) at ALRS where (meumatic testing program took place. 33 J W3 1\« Field N Labontoty W2 W2A Black PlHtk Cover »G4 • G3 06 • G2 • *05 « Wl • • « VI V2 V3 Gl « -20 -10 • 0 • • X2 X3 • • Y1 Y2 • • 10 20 Z2 Z3 •30 • XI Y3 Z1 30 X f/ml F^;iire 1.2: Plan view of botdules, plastic cover, and field laboratory at ALRS. 40 34 * Iml Figure 1.3; Three-dimensional perspective of the boreholes at the site. 35 Information about the location and geometry of fractures in the study area has been obtained from surface observations, the examination of oriented cores, and borehole televiewer records. A summary of data concerning the orientation, dip and density of fractures in boreholes can be found in Rasmussen et al. (1990). According to Rasmussen et al. (1990), a total of 79 fractures have been identified in boreholes at the site. The fractures appear to be exponentially distributed in a manner consistent with a Poisson process of fracture locations. Fracture density, defined by Rasmussen et al. (1990) as number of fractures per meter in a 3-/n borehole interval, ranges from zero to a maximum of 4.3 per meter (Table 1.6). A Schmidt equal-area projection of fracture orientations, with contours indicating number of fractures per unit area of the projection circle, is shown in Figure 1.4. Though the fractures exhibit a wide range of inclinations and trends, most of them are near vertical, strike north-south and dip steeply to the east. The stereonet in Figure 1.4 is based on the data of Rasmussen et al. (1990); an earlier stereonet presented by Yeh et al. (1988) has the strikes of all fractures rotated by 180° degrees. 36 Table l.l: Statistics of effective porosity measurements on ALRS core. Medium size sample measurements were obtained using mercury intrusion method which underestimates effective porosity jdiAex Rasmussen et al. 1990). Effective Porosity [%1 Large Small Medium 17.54 Mean 17.15 14.62 Coef. Var. 13% 16% 26% 14.30 Minimum 11.02 9.18 Median 16.52 17.21 14.31 Maximum 24.73 47.58 27.51 Table 1.2: Statistics of hydraulic conductivity values were obtained from 105 large core Mean Coef. Var. Minimum Median Maximum Hydraulic Conductivity [*10*^ /n/s] Suction [kPd\ 0 25 10 3.346 1.475 21.31 301% 105% 156% 0.69 0.126 0.110 2.610 0.556 4.24 25.750 14.588 438.28 50 0.908 115% 0.002 0.498 5.041 100 0.364 112% 0.005 0.235 2.541 Table 1.3: Statistics for van Genuchten's a (after Rasmussen et al. 1990). Van Genuchten's Moisture Characteristic Function a-value [kPa'] 0.0224 Mean 31.T/0 Coef Var. 0.0102 Minimum 0.0203 Median 0.0643 Maximum 37 Table 1.4; Statistics of laboratory determined air permeabilities for various values of suction and at oven-dried (QD) conditions (after Rasmussen et al. 1990). Air Permeability^rni^*io-"'i Suction kPd\ 300 10 25 50 100 500 OD 1.54 26.67 35.11 Mean 11.20 16.88 38.23 57.12 309% Coef. Var. 434% 436% 344% 326% 295% 272% <0.01 <0.01 Minimum 0.02 0.25 1.29 1.91 3.81 0.05 5.09 Median 0.10 0.39 2.10 6.04 12.08 41.90 780.50 333.10 678.70 780.50 1012.60 Maximum 389.80 Table 1.5: Statistics of laboratory determined Klinkenberg slip flow coefficient Rasmussen et al. 1990). Klinkenberg Coefficient {kPa\ Mean 322.1 Coef Var. 82% 35 Minimum 217 Median 1277 Maximum Table 1.6: Statistics of fracture density and orientation obtained from ALRS cores {Z&&V Rasmussen et al. 1990). Fracture Fracture Orientation Strike [deg] Density [m"'] Dip \deg\ Mean 0.77 214.4 64.5 108% 56% Coef Var. 37% 0.00 Minimum 3 I Median 0.67 109 55 Maximum 359 4.33 89 38 o Table 1.7: Statistics of field determined outflow rates and of saturated hydraulic conductivity values at the ALRS (after/towmssew et al. 1990). Outflow Hydraulic Conductivity [• r*10-^#nVjl Philip Glover Dachler Mean 26.349 59.42 29.10 30.20 Coef Var. 612% 729% 729% 662% Minimum 0.016 0.27 0.48 0.41 Median 0.633 11.63 5.64 10.33 Maximum 1232 39224 19126 17900 Table 1.8; Statistics of 3-/n scale air permeability data. One interval exceeded the capacity of the measurement device, while a second interval was less than the measurement threshold (after Rasmussen et al. 1990). Field Air Permeability [•10*"'m^l 178.1 Mean 667% Coef Var. <0.420 Minimum 4.02 Median >13366 Maximum Table 1.9: Field water content on various days at the ALRS (after Rasmussen et al. 1990). Borehole Water Content (m /m ) Julian Day 279 372 41 236 255 406 448 505 After 1/1/87 Mean 13.75 14.59 14.50 14.36 14.26 14.12 14.32 14.32 Coef Var. 11% 12% 11% 12% 12% 12% 12% 12% Minimum 10.77 9.99 9.85 10.19 10.85 10.89 9.53 10.14 Median 13.71 14.04 14.46 14.48 14.33 14.35 14.18 14.23 Maximum 17.63 17.29 18.75 18.42 17.74 17.77 17.57 17.51 39 Figure 1.4; Lower hemisphere Schmidt equal-area projection of fractures identified by Rasmussen et al. (1990). Contours indicate number of fractures per unit area of projection circle. 40 Surface fracture traces include an additional, steeply dipping east-west set. Laboratory experiments have also been conducted on large blocks of fractured tuff, including a study of aperture distribution in a large natural fracture (Vickers et al. 1992). Single-hole hydraulic injection tests were performed by Tidwell (1988) in 87 out of 105 3-m intervals in boreholes XI, X2, X3, Yl, Y2, Y3, Zl, Z2, and Z3 from which core samples had been taken. The hydraulic tests were conducted by maintaining a constant water level near the top of a borehole until a constant injection rate was established. The injection rate was converted into equivalent hydraulic conductivity using three different formulae for steady state flow. One formula, due to Glover (1953), was modified by Tidwell (1988, p. 65, eq. 4.26) to account for borehole inclination according to O sinh"'(Z,/r)- Ljh (1.1) where K is hydraulic conductivity {LT'\ Q is flow rate [L^T'], L is length of injection interval [Z,], r is borehole radius [Z,], h is borehole length [£.], 2, is distance from bottom of borehole to lower edge of test interval [Z,], and fi is angle of borehole inclination relative to ground surface (Figure 1.5), and the dimensions of each variable are specified in terms of mass \M\, length [Z,] and time [7] throughout this dissertation. Another formula, due to Philip (1985), was modified in a likewise manner by Tidwell (1988, p. 79, eq. 4.61) to read 41 where Q is a geometric factor related to the eccentricity of an assumed prolate spheroid representing the borehole, and z, is distance along the borehole from its bottom to the upper edge of the test interval [Z,]. The third formula, due to Dachler (1936), is written by Rasmussen et al. (1990; p.27, eq. 19) as (1.3) where is pressure head in the injection interval [Z,]. Summary statistics are listed in Table 1.7 and show that calculated hydraulic conductivities range over five orders of magnitude. According to Rasmussen et al. (1990), the corresponding hydraulic conductivities are log-normally distributed and strongly skewed toward high values. Single-hole pneumatic injection tests were conducted in 87 intervals of length 3 m in 9 boreholes (XI, X2, X3, YI, Y2, Y3, Zl, Z2 and Z3) by Rasmussen et al. (1990, 1991, 1993). According to Rasmussen et al. (1993), the tests were conducted by injecting air at a constant mass rate between two inflated packers while monitoring pressure within the injection interval. Pressure was said to have reached stable values within minutes in most test intervals. Air permeability was calculated using Dachler's (1936) steady state formula, adapted by Rasmussen et al. (1990, p.35, eq. 24 and 1993) to isothermal airflow, . where Q.M.P. rlrf\ p 2 - P a2\) ,,4\ • / is intrinsic permeability to air [L^], Qa is volumetric airflow rate at standard temperature and pressure [L^T^ water \ML'^T\ is air viscosity y is the specific weight of is atmospheric pressure \ML''T\ and p is absolute air pressure 42 [ML''T^'\ in the test interval. Summary statistics are listed in Table 1.8. Figure 5b of Rasmussen et al. (1993) suggests a good correlation (r = 0.876) between pneumatic and hydraulic permeabilities at the ALRS. Ground Surfoce H sin B Figure 1.5; Geometric parameters associated with a single-hole hydraulic injection test (adapted from Tidwell, 1988). 43 Figure 1.6 shows a scatter plot of pneumatic permeability versus fracture density for 3-m borehole intervals based on the data of Rasnnissen et al. (1990). It suggests a lack of correlation between fracture density and air permeability. 1E-11 1E-12 -= 1E-13 ^ £ § 1E-14 • • : g. ^ 1E-15 -4 :! • • I. 1E-16 ^ : i 5 • • 1E-17 0.0 2.0 4.0 Fracture density [counts/m] 6.0 Figure 1.6: Air permeability versus fracture density (data from Rasmussen et al. 1990). Rasmussen et al. (1990) conducted in situ determinations of volumetric water content by means of neutron probes at 105 locations within the nine boreholes listed earlier. They took measurements on eight separate occasions 41, 236, 255, 279, 372, 406, 448, and 505 Julian days following 1/1/87 (Table 1.9). Temporal variations in 44 neutron readings were slow during this period. As matrix porosities are much larger than the porosity of fractures, and pore sizes within the matrix are generally much smaller than fracture openings, most of the water resides in the matrix and its water content is much higher than that of the fractures. We therefore attribute the neutron probe measurements, and corresponding water contents, primarily to the matrix. A gaseous tracer experiment was conducted at the ALRS (Yeh et al. 1988) by injecting helium into one borehole and monitoring its arrival in neighboring boreholes by means of a thermal conductivity meter (utilizing the low thermal conductivity of helium relative to that of standard atmosphere). A test performed by injecting helium into borehole X2, below a packer set at a distance of about 20 m from the surface, showed breakthrough of helium into borehole XI at a distance of 9.5 - 10 m from the surface. The breakthrough was attributed by the authors to a fracture that had been encountered in the injection and the detection intervals. No helium was detected in borehole X3 during the test, but the data collected in borehole XI are summarized in Figure 10 of Yeh et al. (1988). 1.2.3 Single-Hole Pneumatic Injection Tests Single-hole pneumatic injection tests conducted at the ALRS by Rasmussen et al. (1990, 1991, 1993) were of relatively short duration and involved relatively long test intervals. Guzman et al. (1994, 1996) and Guzman and Neuman (1996) conducted a much larger number of single-hole pneumatic injection tests of considerably longer duration over shorter borehole intervals. Their tests were conducted under highly 45 controlled conditions, subject to strict quality assurance, within six boreholes (V2, W2A, X2, Y2, Y3, and Z2) that extend over a horizontal area of 32 m by 20 m. Five of the boreholes are 30-m long (V2, W2A, X2, Y2, Z2) and one has a length of 45 m (Y3); five are inclined at 45° (W2A, X2, Y2, Y3, Z2) and one is vertical (V2). A total of 184 borehole segments were tested by setting the packers 1 m apart as shown in Figure 1.7. Additional tests were conducted in segments of lengths 0.5, 2.0 and 3.0 m in borehole Y2, and 2.0 m in borehole X2, bringing the total number of tests to over 270. The tests were conducted by maintaining a constant injection rate until air pressure became relatively stable and remained so for some time. The injection rate was then incremented by a constant value and the procedure repeated. Three or more such incremental steps were conducted in each borehole segment while recording the air injection rate, pressure, temperature and relative humidity. For each relatively stable period of injection rate and pressure, air permeability was estimated by treating the rock around each test interval as a uniform, isotropic continuum within which air flows as a single phase under steady state, in a pressure field exhibiting prolate spheroidal symmetry. The results of these steady state interpretations of single-hole air injection tests are listed in Guzman el al. (1996). The authors found, and noted, that {Guzman et al. 1994, \996-, Guzman andNeuman, 1996) I. Air permeabilities determined in situ fi-om steady state single-hole test data are much higher than those determined on core samples of rock matrix in the laboratory, suggesting that the in situ permeabilities represent the properties of fractures at the site. 46 2. It is generally not possible to distinguish between the permeabilities of individual fractures, and the bulk permeability of the fractured rock in the immediate vicinity of a test interval, by means of steady state single-hole test data. 3. The time required for pressure to stabilize in the injection interval typically ranges from 30 to 60 min, increases with flow rate, and may at times exceed 24 hrs, suggesting that steady state permeability values published in the literature for this and other sites, based on much shorter air injection tests, may not be entirely valid. 4. Steady state interpretation of single-hole injection tests, based on the assumption of radial flow, is acceptable for intervals of length equal to or greater than 0.5 m in boreholes having a radius of 5 cm, as is the case at the ALRS. 5. Pressure in the injection interval typically rises to a peak prior to stabilizing at a constant value, due to a two-phase flow effect whereby water in the rock is displaced by air during injection. 6. In most test intervals, pneumatic permeabilities show a systematic increase with applied pressure as air displaces water under two-phase flow. 7. In a few test intervals, intersected by widely open fractures, air permeabilities decrease with applied pressure due to inertia! effects. 8. Air permeabilities exhibit a hysteretic variation with applied pressure. 9. The pressure-dependence of air permeability suggests that it is advisable to conduct single-hole air injection tests at several applied flow rates and/or pressures. 10. Enhanced permeability due to slip flow (the Klinkenberg effect) appears to be of little relevance to the interpretation of single-hole air injection tests at the ALRS. 47 11. Local-scaie air permeabilities vary by orders of magnitude between test intervals across the site. 12. Spatial variability in permeability is much greater than that due to applied pressure and lends itself to meaningful statistical and geostatistical analysis. 13. Air permeabilities are poorly correlated with fracture densities, as is known to be the case for hydraulic conductivities at many water-saturated fractured rock sites worldwide {Neuman, 1987), providing further support for Neuman's conclusion that the permeability of fractured rocks cannot be reliably predicted from information about fracture geometry (density, trace lengths, orientations, apertures and their roughness) but must be determined directly by means of hydraulic and/or pneumatic tests. 14. Air permeabilities vary systematically with the scale of measurement as represented nominally by the distance between packers in an injection interval. The work of Guzman el al. (1994, 1996) and Guzman and Neuman (1996) strongly suggests that air injection tests yield properties of the fracture system, which are relevant to both unsaturated and saturated conditions. In particular, numerical simulations by these authors show that, whereas the intrinsic permeability one determines from such tests is generally lower than the intrinsic permeability to water of fractures which surround the test interval, it nevertheless approaches the latter as the applied pressure goes up. This is so because capillary forces tend to draw water from fractures into the porous (matrix) blocks of rock between the fractures, thereby leaving the latter saturated primarily with air. Water saturation in the matrix blocks is therefore typically 48 much higher than that within the fractures, making it relatively difHcuIt for air to flow through such blocks. It follows that, during a pneumatic injection test, the air moves primarily through fractures (most of which contain relatively little water) and the test therefore yields flow and transport parameters which reflect the intrinsic properties of these largely air-fllled fractures. Locatians of 1.0 m Air MmmmwiimiU (raria* do net indicf «phf» of influ»ne*) Figure 1.7; Perspective toward the Northeast showing center locations of l-m single-hole pneumatic test intervals; overlapping circles indicate re-tested locations (adapted from Guzman etal. 1996). 49 1.2.4 Kriged Estimates of Air Permeability at the ALRS Core and single-hole measurements, conducted over short segments of a borehole, provide information only about a small volume of rock in the immediate vicinity of each measurement interval. Available data from the ALRS indicate that rock properties, measured on such small scales, vary erratically in space in a manner, which renders the rock randomly heterogeneous and pneumatically anisotropic. A major question is how to describe this spatial and directional dependence of medium properties in untested portions of the rock. Analyses to date suggest (Bassett et al. 1994, 1997; Guzman et al. 1996, Illman et al. 1998) that it is possible to interpolate some of the core and single-hole measurements at the ALRS between boreholes by means of geostatistical methods, which view the corresponding variables as correlated random fields. This is especially true about air permeability, porosity, fracture density, water content, and the van Genuchten water retention parameter a, for each of which we possess enough measurements to constitute a workable geostatistical sample. Standard geostatistical analysis provides best (minimum variance) linear unbiased estimates of how each such quantity varies in three-dimensional space, together with information about the quality of these estimates. A geostatistical analysis of the above site variables has been conducted by Chen et al. (1997) and by Illman et al. (1998) in which the details are provided. Here, kriged maps of single-hole air permeability values are provided for the reader to give a subsurface picture of air permeability distribution at the ALRS. 50 Figure 1.8 compares kriged images of log air permeability generated by lllman et al. (1998) along four vertical sections at = 0, 5, 7, and 10 m using \-m data (left column) and the combined set o f \ - m and 3-m data, the latter from boreholes X I , X3, YI, Z1 and Z3 (right column). The figure shows boreholes intersected by, or located very close to, each cross-section. The two sets of kriged images are considerably different from each other. This is most pronounced at = 0 m, which passes through the Z-series of boreholes; here the inclusion of data from boreholes Z1 and Z3 has caused estimated permeability in the upper right comer of the section to be much higher than it is v/ithout these data. The effect extends to all four cross-sections, which exhibit elevated permeabilities near the upper-right comer. Along sections ^X.y = S m and 7 m, which pass close to the Y and V series, respectively, the addition of data from YI affects the shape and size of a prominent high-permeability zone which extends through Y2. The addition of data from X3 to the set reveals corresponding high- and low-permeability zones in section= 10 /n which correlate well with similar zones, intersected by Y3, in section j/ = 5 m. A three-dimensional representation of kriged log permeability based on the combined set of data is shown in Figure 1.9. Illmcm et al. (1998) show that the above data are amenable to continuum geostatistical analysis and exhibit distinct spatial correlation stmctures. This suggests that the data can be viewed as samples from a random field, or stochastic continuum, as proposed over a decade ago by Meuman (1987) and affirmed more recently by Tsang et al. (1996). This is so despite the fact that the rock is fractured and therefore mechanically discontinuous. Our finding may be contrary to geologic intuition, but it is supported 51 strongfy similar findings in many other ftactured tock terrains including crystalline rocks at Oracle, Arizona; Strqia, Sweden; and Fanay-Ai^res, France. It strongly siqiports the triplication ofcontinuum flow and tramqwrt theories and models to fiactured porous tuffi on scales of a metn or more. Figure 1.8: Kriged log k estimates obtained using l-m scale data from boreholes X2, Y2, Y3, Z2, V2, and W2A (left) and same together with 3-m scale data fiom boreholes XI, X3, Yl, Zl, and Z3 (pg^) {adapted Smmlllnumetal. 1998). Log ifc where it [m'l Figure 1.9: Three-dimeiisioiial rqnesentation of kriged tog k (ad^ed finm Ulnum et al. 1998). 53 1.2.5 Geophysical Surveys Conducted at Site Modeling of gas/liquid transport through the vadose zone by analytical and numerical techniques requires inter alia, the quantification of spatial and temporal variations in water content. Water content can be obtained in situ by means of neutron probes and possibly through the employment of cross-hole electrical resistivity tomography (ERT). Neutron probe measurements were taken using Boart-Longyear CPN Model 503 at 1-m intervals in boreholes Wl, W2A, W3, XI, X2, X3, Yl, Y2, Y3. Zl, Z2, and Z3 during July 1996. A cross-hole ERT survey was also conducted by student researchers from the University of Arizona, during January 1996. Here, results from both techniques are briefly discussed and summarized. Appendix A describes the use of ERT to estimate in situ values of water content. A review of the neutron probe methodology is given in Appendix B. Cross-hole electrical resistivity tomography (ERT) was employed (measurements tak^n by G. Morelli from Mining Engineering and his coworker) during January 1996 to obtain estimates of electrical resistivity (which can then be possibly related to water content) at the covered site. Figures 1.10 is an image of cross-hole ERT obtained through a numerical inversion technique described in Morelli and LaBrecque (1996). Data were obtained along the V-series boreholes (VI, V2, and V3). Contours of electrical resistivity are given in [Qni\. A relatively large area of high resistivity measurements seen in figure 54 1.10 suggests a possible (dry) angled fracture plane. Spatial variability in resistivity can result from a number of causes including textural variability of subsurface materials (e.g., sand versus clay; large voids such as fractures), distribution of capillary water, variability in the ionic content of water, etc. Neutron probe measurements were taken in boreholes W, X, Y, and Z to quantify the amount of moisture present in the unsaturated fractured tuffs, prior to the instrumentation of boreholes for cross-hole pneumatic testing. Due to time and financial constraints, a detailed calibration of neutron probes in unsaturated fractured rocks proposed by Illman (1996) could not be completed. For purposes of obtaining crude estimates of water content from thermalized neutron counts, a field calibration curve obtained by Rasmussen el al. (1990; p.3l, eq. 21) has been employed to convert the neutron counts to estimates of volumetric v.'ater content. Figures 1.11 through 1.14 show the spatial distribution of volumetric water content down the boreholes. The figures clearly show that downhole variability of water content in 13 boreholes is quite small (between 0.08 - 0.12), suggesting that capillary water is quite uniformly distributed throughout the fractured rock mass. The uniformity in the spatial distribution of capillary water may perhaps be due to the black, plastic cover shown on Plate 1.2. A study by McTigue et al. (1993) also showed relatively uniform moisture in unsaturated alluvium under a concrete slab at a field site near Albuquerque, NM. Figure 1.15 is a histogram of the 336 values of water content. One unusually high value obtained from the bottom of borehole Y3 (probably filled with water) was excluded from the statistical sample. The histogram clearly shows a bimodal distribution of 55 volumetric water content which is related to the pore size distribution of fractured tuffs shown in Figure C.l, supporting the notion that neutron probes measure primarily the properties of the matrix. This may further imply qualitatively that capillary water resides in the porous matrix and not in fractures, which are thus relatively dry. An EM survey was conducted along borehole Y2 by researchers from the Geosciences Department at the University of Arizona (personal conmiunication, A. Guzman, 1999). Figure 16 shows the results from the survey in which both small and large-scale oscillations of resistivity are visible. Approximately 54 feet from the lower lip of the casing an anomaly in resistivity is evident which corresponds to a large fracture seen in BHTV (or cavity - see Figure 4.7). ^ 56 Figure 1.10: Image of cross-hole, electrical resistivity tomogrtq>hy (ERT) obtained between bordioles VI andVB. 57 a 14 at2 ai ao8 ao6 ao4 • wt I ' W2A| W3 i ao2 0 S I O U 2 0 2 S X 3 S 4 O 4 S S O Figure 1.11: Variation of vohunetric water content with depth in W-series boreholes. 0.14 ai2 ai aos ao6 aoi • XI • X2 X3 0.02 10 IS 20 25 10 35 40 4S SO F^:ure 1.12: Variation of vohunetric water content widi depth in X-series boreholes. 58 0.14 0.12 I K >• 0.1 0.0S 0.06 0.04 • Yl 0.02 • Y2 Y3 10 iS 25 20 30 35 40 45 SO Dtmyrni Figure 1.13: Variatioii of vohimetric water conteoEt wfth depth in Y-series boreholes. 0.14 0.12 • • 0.1 I • a • 0.08 0.06 0.04 • Zl 0.02 m Z2 Z3 10 15 20 25 30 35 40 45 50 Otptkyml Figure 1.14: Variation of vohmietric water content wfth dqidi in Z-series boreholes. 59 8 f 20 ' 0, Figine 1.15: Histogcam of vohimetric water oontCTt vahies from the W-, X-, Y-, and Zseries boreholes. Y2 high k zone »V CCMSfJCWTL ?« ttpis tv «« «K?ir? Figure 1,16: EM survey conducted along borehole Y2. 61 1.3 TYPE-CURVE ANALYSIS OF SINGLE-HOLE PNEUMATIC INJECTION TESTS In this dissertation, interpretations of transient pressure data from the single-hole air injection tests previously conducted at the site by Guzman el al. (1994, 1996) and Guzman and Neuman (1996) are presented. These interpretations are based on typecurves derived analytically for single-phase gas flow by linearizing the otherwise nonlinear partial differential equations, which govern such flow in uniform, isotropic porous continua. The type-curves correspond to three different flow geometries: threedimensional flow with spherical symmetry, two-dimensional flow with radial symmetry, and flow in a continuum with an embedded high-permeability planar feature (a major fracture). Included in the type-curves are effects of gas storage in the injection interval (known as borehole storage effect) and reduced or enhanced permeability in the immediate vicinity of this interval (known as positive or negative skin effects). Our test interpretations rely not only on standard type-curves but also on the type-curves of pressure derivative versus the logarithm of time. Such pressure derivative curves have become popular in recent years because they accentuate phenomena which might otherwise be missed, help diagnose the prevailing flow regime, and aid in constraining the calculation of corresponding flow parameters. Transient analyses of single-hole pneumatic tests yield information about air permeability, air-filled porosity, skin factor and dimensionality of the flow regime on a nominal scale of I m in the immediate vicinity of each test interval. We show that transient air permeabilities agree well with previously determined steady state values. 62 1.4 CROSS-HOLE PNEUMATIC INJECTION TESTS Cross-hole pneumatic injection tests were conducted at the ALRS by the author with the help of a technician, Mr. Dick L. Thompson. These tests span the entire volume of fractured rock previously subjected to single-hole testing, and beyond. Their purpose was to a) directly characterize the pneumatic properties of the rock on a site-wide scale; b) determine the spatial extent and connectivity of fractures and/or high-permeability flow chaimels across the site; and c) compare the results with corresponding information that one might deduce from smaller-scale (laboratory and single-hole) tests. A total of forty-four cross-hole tests of diverse types (constant injection rate, step injection rates, instantaneous injection rate) have been conducted between various boreholes and borehole intervals at the site as part of this dissertation work. Several tests were conducted without inflating the packers in monitoring holes in order to ascertain the effect that such open holes may have on pressure response in the injection test interval, i.e., on single-hole pneumatic injection tests. Additional single- and cross-hole tests were conducted to investigate the effects of: 1) drierite (calcium sulfate which acts as a drying agent) on humidity in the injection air stream and two-phase flow effects seen in Guzman and Neuman (1994 and 1996); 2) air injection into a zero-permeability PVC pipe to quantify air leakage from the equipment; 3) barometric pressure fluctuations on air pressure in packed-off monitoring intervals under ambient conditions; 4) air injection while measuring changes in neutron counts in the rock surrounding the injection interval; and 5) variations in injection rate on pressure in the injection interval. To design our 63 cross-hole tests, we relied heavily on information about fiacture locations and medium properties derived from core, borehole televiewer, and single-hole air injection test data. A complete list of all cross-hole tests conducted by us at the ALRS is included in Table 6.1. In most of these tests, air was injected at a constant mass flow rate into a relatively short borehole interval of length I - 2 m while monitoring a) air pressure and temperature in the injection interval; b) barometric pressure, air temperature and relative humidity at the surface; and c) air pressure and temperature in 13 short (0.5-2 m) and 24 longer (4 - 20 m) intervals within the injection and surrounding boreholes. Only one of these tests, labeled PP4, was fiilly analyzed to date. During this test, which we describe in detail, pressure responses were detected in 12 of the 13 short monitoring intervals and 20 of the 24 longer intervals. We use newly developed, analytically-derived type-curves to interpret the PP4 cross-hole test data. An automatic parameter estimation method based on a three- dimensional finite volume code (FEHM), coupled with an inverse code (PEST) has been employed to analyze the same data set in Illman et al. (1998). The type-curve approach treats short and longer borehole intervals as either points or lines, depending on distance between injection and monitoring intervals, while accounting indirectly for storage effects in monitoring intervals due to the compressibility of air. The finite volume code allows representing borehole geometry and storage more realistically, and directly, by treating each borehole as a high-permeability cylinder having finite length and radius. Analyses of pressure data from individual monitoring intervals by the two methods, under the assumption that the rock acts as a uniform and isotropic fractured porous 64 continuum, yield comparable results. These results include information about pneumatic connections between the injection and monitoring intervals, corresponding directional air permeabilities, and air-filled porosities, all of which are found to vary considerably from one monitoring interval to another on a scale of 10 - 20 /n. Together with the results of earlier site investigations, our single- and cross-hole test analyses provide important new insight into the phenomenology of air flow through unsaturated fractured rocks at the ALRS, and about the spatial variability and scale-dependence of corresponding pneumatic parameters such as permeability, connectivity and air-filled porosity. 1.5 SCOPE OF THE DISSERTATION Chapter 2 of this dissertation presents the theory of gas flow in porous media with an emphasis on the case where water is relatively immobile. This theory forms the basis for all the analytical and numerical methods employed here to interpret single-hole and cross-hole pneumatic injection tests at the ALRS. Analytical formulae and corresponding type-curves for the interpretation of single-hole tests are presented in Chapter 3. Chapter 4 describes the type-curve interpretation of transient single-hole test results and their comparison with corresponding steady state test interpretations. Analytical formulae and corresponding type-curves for the interpretation of cross-hole tests are presented in Chapter 5. In Chapter 6, the design and conduct of cross-hole pneumatic interference tests at the site and interpretation of one of them by means of newly developed typecurves are described. The dissertation ends with a comprehensive list of flndings and conclusions in Chapter 7. 65 2. GAS FLOW IN POROUS AND FRACTURED MEDIA 2.1 THEORY OF GAS FLOW IN POROUS MEDIA The theory of gas (air) flow in porous and fi^ctured media has gained great importance in recent years. There are several advantages to using gases instead of liquids to conduct laboratory and field determinations of flow and transport parameters in unsaturated media. Injecting liquids and dissolved or suspended tracers into unsaturated media would cause them to move predominantly downward under the influence of gravity, and would therefore yield at best limited information about the ability of geologic media to conduct liquids and chemical constituents in directions other than the vertical. It would further make it difficult to conduct more than a single test at any location because the injection of liquid modifies the ambient saturation of the geologic medium, and the time required to recover ambient conditions may be exceedingly long. In the laboratory, the flow of liquids can cause plugging of a porous medium upon continuous flow of water {Mttskat, 1946). Many of these limitations can be overcome by conducting field and laboratory tests with gases rather than with liquids, and with gaseous tracers instead of chemicals dissolved in water. Gases also have much lower densities in comparison to liquids which makes it possible for one to employ lighter (or less durable yet more flexible) materials (PVC as opposed to steel) to conduct experiments both in the field and in the laboratory. In unsaturated flow systems, gases fill all void space not occupied by liquid. This may be one reason why air injection experiments at the ALRS (single- and cross-hole 66 tests - to be discussed later) almost always reach steady-state or three-dimensional flow. If one were able to disregard nonlinear effects such as slip flow, inertia, and multi-phase flow, one would expect to see similar pressure responses in saturated fi-actured media if water was employed as the extraction/injection fluid. This is because intrinsic permeability (not relative permeability in multi-phase flow) is independent of the physico-mechanical properties of homogenous fluids passing through a stable, nonreactive material. Intrinsic permeability is by definition a constant fully determined by the structure of the medium. The apparent permeability, on the other hand is a pressure dependent quantity and is usually associated with gas flow at low velocities in geologic formations of low permeability or when other fluids coexist within the medium. The equations that describe gas flow in partially saturated porous media are nonlinear due to the compressible nature of gases, its capillary interaction with water, and non-Darcian behavior at high Reynolds numbers. A complete description of gas-water (in our experiments air-water) interaction requires two systems of coupled partial differential equations, one for each phase. The resulting two-phase flow equations can then be solved by numerical methods of the kind developed by Mendoza and Frind (1990), Pruess (1991) or Zyvoloski et al. (1988, 1996, 1997). In special cases it may be possible to solve simplified versions of these equations analytically. We are particularly interested in the use of analytically derived type-curve methods for the interpretation of single-hole and cross-hole pneumatic test results. The development of corresponding analytical formulae requires that two-phase flow is approximated as single-phase gas flow and that water is treated as immobile. The resulting single-phase gas flow equations 67 must additionally be linearized to allow solving them either in terms of pressure, as is customary for liquids, or in terms of pressure-squared or pseudopressure, as is more common for gases. We provide a literature survey and describe this process of simplification and linearization in the subsequent sections. 2.1.1 Literature Review Early attempts to solve/study gas-flow problems were limited to solutions involving a single gas phase. Research concerning gas flow through porous media was conducted primarily in the petroleum/natural gas industry (also in chemical engineering/industry; ceramics and materials) that involved approximate analytical, graphical, analogue and numerical solutions. Muskat (1934) suggested a method, which employs a succession of steady-state solutions. Others {Heatherington et al. 1942; MacRoberts 1949, Cornell and Katz 1953, and others) have obtained approximate solutions by assuming constant gas properties throughout the duration of a well test and small pressure gradients. Variations of gas properties such as dynamic viscosity {fx) and compressibility factor (Z) with pressure have been neglected due to analytical difficulties in solving the nonlinear partial differential equation. Katz et al. (1959) list linearized versions of the gas flow equation for an ideal gas. Aronofsky ( \ 9 S A r ) considered the effect of gas slip on transient flow of gas through one and two-dimensional porous media. The gas slip permeability (or apparent permeability) was assumed to vary as a linear function with the reciprocal pressure. As discussed by Aronofsky (1954), Guzman and Neuman (1996), and in Appendix C of this 68 dissertation, gas slip most probably does not impact pressures during well tests; it is more important in the conduct and analysis of laboratory gas flow experiments. Other nonlinear effects such as two-phase flow and inertia have been found by Guzman et al. (1994, 1996) and Guzman and Neuman (1996), as well as lllman et al. (1998), to manifest themselves primarily in the immediate neighborhood of injection intervals. During the early period (petroleum industry in the 1930's through the 1950's) in which analytical solutions were sought for the gas flow equation, some authors such as Aronofsky and Jenkins (1952) Jenkins and Aronofsky (1953), Bruce et al. (1953), Aronofsky and Jenkins (1954), Aronofsky and Ferris (1954), and Aronofsky and Porter (1956) attempted to solve the nonlinear equation using numerical techniques. In particular, Douglas, Peaceman, andRachford, Jr. (1955) numerically solved the transient gas flow equation through porous media assuming a square reservoir with a single well in the center. The alternating direction implicit (ADI) method was employed for the calculation in which the numerical solutions compared favorably with published solutions for radial reservoirs. A numerical study by Aronofsky and Jenkins (1954) yielded one of the most important conclusions in gas flow through porous media during this period. The important conclusion was that the numerical solution of radial flow of ideal gas in porous media for various rates can be approximated quite well by the analytical solution for transient liquid flow in a radial system of van Everdingen and //wrs/ (1949). A major advance in theory was made when Al-Hussainy et al. (1966) published their application of the real gas pseudopressure concept to rigorously linearize the nonlinear flow equation of real gases through porous media. Pseudopressure is a term 69 employed in the petroleum industry, but is more widely known as the Kirchhoff transformation (Kirchhoff, 1894). According to Raghavan (1993), the concept of pseudopressure in the context of gas flow has been used by Carter (1966) prior to Al- Hussainy et al. (1966) in a slightly different form. It is Al-Hussainy et al. (1966) who first applied this concept to understand pressure behavior and well responses. We show later in this chapter how the Kirchhoff transformation is applied to our nonlinear gas diffusion equation to linearize it in a more tractable form. Before the introduction of KirchhofF transformation, data collected fi-om gas well tests were plotted in terms of pressure-squared (^*) in the petroleum industry. Recently, Massmcom (1989) and KicWhorter (1990) derived analytical solutions to unsteady radial flow of gas in the vadose zone. Massmann (1989), not concerned with the compressible nature of gases, instead addressed the issue of the validity of Darcy's law in describing the flow of gases through porous media and gave some conditions under which Darcy's law is applicable to gas flow. One of the conclusions reached by this author was that Darcy's law is valid and groundwater equations can be used to model gas flow in the vadose zone as long as slip flow can be neglected. In the literature, one finds data from gas injection or withdrawal tests plotted and analyzed in terms of pressure (p\ pressure-squared or pseudopressure (vf). Tracey (1956) and others (e.g., Matthews, 1961) suggested that when pressure in the system is sufficiently low, such data can be analyzed in terms of p as if the gas behaved like a slightly compressible liquid. This approach is quite common in the field of environmental engineering and hydrology. For example, Massmann (1989), Johnson et 70 al. (1990), Edwards (1994), Massmann and Madden, (1994), Beckett and Huntley (1994) and Edwards (1996) analyzed transient radial flow of gas in the vadose zone by means of standard groundwater flow equations written in terms of pressure without a gravity term. Massmann (1989) concluded that this is appropriate as long as gas pressure between two points is less than 0.5 atm (1 a/m=101,325 Pa). Above 0.5 atm, he recommends to write groundwater equations in terms of pressure-squared as suggested by the analytical investigation by Kidder (1957). In the same study, Massmann (1989) also gave condition under which the Klinkenberg effect may become important. A more rigorous analysis under steady-state conditions has been carried out by Baehr and Hult (1991) who had developed two analytical solutions for steady-state, twodimensional, axisymmetric airflow to a single well partially screened in the unsaturated zone. Other steady-state solutions have been developed by Shan et al. (1992), Baehr and Joss (1995), Kaluarachchi (1995), DePaoli et al. (1996), Ge and Liao (1996), and Ge (1998). Warrick and Rojano (1999) examined the effects of source cavity shape on steady, three-dimensional flow of soil gases. Transient solutions have been developed by McWhorter (1990) for a two-dimensional flow system and Falta (1995) for injection and extraction wells from horizontal wells. Guo (1997) obtained a quasi-analytical solution to the transient, nonlinear radial, gas flow equation using the Boltzman transformation in which the pressure is calculated using the Newton-Raphson method. 71 2.1.2 Two-Phase and Single-Phase Representations Consider the flow of two homogenous immiscible fluids, a liquid and air, in a rigid porous medium under isothermal conditions. In the absence of mass transfer between the fluid phases, the liquid and air mass conservation equations take the forms = (2.1-1) where the subscripts / and a refer to liquid and air, respectively; p is mass density q is flux density [LT'\, ^ is porosity [Z,®]; S is fluid saturation [L\ t is time [7]; V- is the divergence operator where the dimensions of each variable are specified in terms of mass [M\, length [Z,] and time [7]. In practice, it is common to assume that Darcy's law applies to both liquids and air. In the case of air, Darcy's law may not apply when Knudsen diffusion or slip flow (the so-called Klinkenberg effect) are important. Both of these phenomena are discussed briefly in Appendix C. Laboratory experiments by Alzaydi et al. (1978) have shown that air flux through a column of Ottawa sand is linear in the applied pressure gradient {i.e., Darcian) except at early time following the start of each experiment. They found the same to be true for kaolinite clay, except that here the time required to establish a Darcy regime was longer. As mentioned in Section 1.2.3 of the Introduction, the steady state interpretation of multistep single-hole pneumatic injection tests at the ALRS have 72 revealed inertial (nonlinear, non-Darcian) behavior of air flow in a few test intervals intersected by highly permeable fractures. In most other test intervals, the flow appeared to be multiphase Darcian. We therefore adopt the following standard forms of Darcy's law for liquid and air, respectively, ,,=-A£i,Vh, Ml q. = -—vp, f. where g is the gravitational constant [LT'\, is dynamic viscosity of air (2.1-2) k is intrinsic permeability [Z,*] for both air and water; V is the gradient operator [Z,"']; hi is hydraulic head [L\ defined as h,=-^ + z P,g p is pressure Z (2.1-3) is elevation about an arbitrary datum [Z,]; and the effect of gravity on air flow is neglected. Substituting (2.1-2) into (2.1-1) yields the liquid flow and air flow equations, V •[p, ^ (2.1-4) respectively. The two equations are coupled via the relationships 5,+^„ = l (2.1-5) Pc = Pa-Pi (2.1-6) 73 where is capillary pressure. To solve them requires the additionzd specification of appropriate equations of state (functional relationships between fluid properties and pressure, and between permeability, capillary pressure and saturation, for each phase) and forcing functions (source terms, initial and boundary conditions). As the equations of state are generally nonlinear, so are the above flow equations. For purposes of analyzing airflow around wells, it is common to treat the liquid phase as being immobile. Then Sa = I and the airflow equation simplifies to where the subscript a has been omitted as all equations now refer to a single gas phase. The corresponding equation of state is commonly written as {Burcik, 1957) pV = Z{pj)nRT (2.1-8) where V is volume [Z,^]; Z(/7, t) is a dimensionless compressibility factor; n is mass in moles [mo/]; and R is the universal gas constant [Jottles "FC'mor']. Upon rearranging eq. (2.1-8) to read where = V j n is specific volume the dependence of Z o n p and T is made explicit. One can thus readily calculate Z for any pressure, temperature and specific volume. A sample calculation of Z for conditions typically found at the ALRS (p = 200 kPa, 7 = 200° k) is included in Appendix D. Figure 2.1 shows corresponding variations of Z with pressure and temperature. The value of Z is seen to be 1 for a large 74 range of pressures ^10^ - \Qi*kPc^ and temperatures ^270-400 'K^ Dik^ sin^e-hole tests at the ALRS, pressures widim the iiyection mterval ranged between 80 kPa and 360 itPa (Table D.2) and tenpecatuies between 288.10" ATand 302.74' K (Table D.3). Thk allows us to treat ak, for purposes of test analysis at the ALRS, as an ideal gas with Z= 1. 15 ZM fldtai Ih"a fcbljr hign«y6ffiiwn L5 N 111 • ••• 05 UBHB Figure 2.1: Variation of con^nessibility fiictor Z widi pressure and temperature. P^29(K T=30(K LIB45 Pim] F^;ure 2J2: Variotioa of air viscosity // with pressure and tenqierature. -O— T-270K - MSOK —T-290K —T-300K 2MM 1M«M P [*^«l F^;ure 2.3: Variation of with i»essure and tempoature. 76 Figure 2.2 shows the variation of air viscosity // with pressure and temperature. Air viscosity is seen to increase slowly with pressure, and to increase with temperature at low pressures but decrease at higher pressures. Figure 2.3 has been included to show how the corresponding product /zZ varies with pressure for a range of temperatures. The product is seen to remain constant (equal to 1.81 x lO'^fa .v) for the range of pressures and temperatures encountered under field-testing conditions at the ALRS. Since Z is virtually constant under these conditions, so is jj, (equal to 1.81 x 10"'/'a 5). Upon considering isothermal conditions, and treating Z and n as constants, one finds from eq. (2.1-8) that mass density is proportional to pressure. For purposes of solving the gas flow equation analytically, it is additionally common to treat ^ as a constant. This allows rewriting (2.1-7) as (2.1-10) which can be rewritten as ^ k dt (2.1-11) or, alternatively, as ^2 2 ^^^£2 ^ k dt (2.1-12) where c is gas compressibility, defmed as c = -L^ = l ~ pdp~ p (2.1-13) 77 In the more general case where /iZ is not a constant, the equation in terms of includes an additional logarithmic term {Al-Hussairy et al. 1966; Raghavan^ 1993; see derivation in Appendix E), V2 2 _ dp ^ <l>fjc dp k dt (2.1-14) Yet another form of the gas flow equation can be obtained upon introducing the pseudopressure defined as {Al-Hussainy et al., 1966; Raghavan, 1993) wi where vv(p) has dimensions [ML''T% This is akin to the well-known Kirchhoff transformation {Kirchhqff, 1894) and has, according to Raghavan, been used earlier by Carter (1966) in a slightly different form. For our purposes, the lower limit of zero on p represents barometric pressure. In the case where /iZ is constant, w{p) = fjZ^Po (2.1-16) iXL where Pj, is the barometric pressure. Figure 2.3 shows that, at high pressures, /xZ varies more-or-less linearly with p. It then follows that {Gradshteyn and Ryzhik, 1994, p. 79, eq. 2.152), M p )« ' f ———dp = — h^a-^bp b In general. + bpJ (2.1-17) 78 dt _ dw{p) d p _ I p d p dp dt piZ dt Vw(p) = ^ ^ V p = ^ V p (2.1-18) and so the gas flow equation in terms of pseudopressure takes on a quasilinear form, k dt 2.1.3 Linearization of the Gas flow Equation Under steady state, equations (2.1-10) and (2.1-11) are linear in p', and equation (2.1-19) is linear in w. Linearization of the transient p-based equation (2.1-10) requires that the leftmost p be set equal to a constant, p (typically some average pressure). The equation can then be rewritten as k dt (2.-20) where, by virtue of (2.1-13), c = \l p. This equation is similar to that typically used for liquids in the absence of gravity (linear dif^usion equation). Setting c equal to a constant i n (2.1-10) o r (2.1-19), s a y c =\ l p , renders these equations linear i n p ' o r w , respectively, a at a dt where a = kjpS^. is pneumatic diffusivity in terms of a gas storage factor (2.1-21) (2.1-22) 79 (2-1-23) P The latter differs from the specific storage commonly used in hydrology which, in our case, would be defined asSs = pgSaIn Chapter 3 we modify a number of analytical type-curve expressions, derived by various authors for liquid flow under conditions similar in principle to those we encounter during single-hole and cross-hole pneumatic injection tests at the ALRS, in a manner which renders them applicable to air-flow. Among others, we recast these expressions in terms of pseudopressure, which by virtue of (2.1-16) is analogous to rewriting them in terms ofp' when /iZ is constant, as is the case at the ALRS. Our typecurve analyses of single-hole pneumatic air injection tests at the ALRS, described in Chapter 4, are conducted using both the /7-based equation (2.1-20), and the '-based equation (2.1-22) with pseudopressure defined according to (2.1-16). As the two sets of results are consistent, we conduct our type-curve analysis of cross-hole tests, described in Chapter 5, using only the simpler p-based form. 80 3. TYPE-CURVE MODELS FOR THE INTERPRETATION OF SINGLE-HOLE TEST DATA 3.0 CONCEPTUALIZATION OF FLOW THROUGH FRACTURED ROCKS For purposes of analyzing fluid flow and solute transport through fractured rocks on the fleld scale, it has become common to think of matrix blocks as forming a continuum, and of fractures as forming another, overlapping continuum {Barenblatt et al. 1960; Warren and Root, 1963). If one of these overlapping entities dominates all relevant aspects of flow and transport, one treats the rock as a single matrix- or fracturedominated continuum. Otherwise, one must allow for the possibility that fluids and solutes could migrate from one entity to the other under a pressure and/or concentration differential between the two. If such fluid and mass transfer between matrix and fractures is fast in comparison to flow and transport through the rock, one considers the two entities to be at equilibrium and treats the rock as an equivalent or effective (single) matrix-fracture continuum {Dykhuizen, 1990; Peters and Klavetter, 1988; Pruess et al. 1990). In the absence of such equilibrium, it is common to adopt the dual porosity model (Bibby, 1981; Moench, 1984; Zimmerman et al. 1993) in which the matrix acts as a nonconducting storage reservoir; the fractures form a conducting medium with negligible storage capacity; and the transfer of fluids (or solutes) between these two overlapping continua is proportional to the pressure (or concentration) differential between them at each point in space-time. A more general version of the latter is the dual permeability model {Duguid and Lee, 1977; Gerke and van Germchten, 1993a,b) in which both the matrix and fracture continua conduct fluids and solutes; still another version is one that 81 accounts for internal gradients of pressure (or concentration) within matrix blocks, referred to as interacting multiple continua by Pruess and Narasimhan (1985; see also Kazemi, 1969; Berkowitz et al. 1988; and Birkholzer et al. 1993). Carrera et al. (1990) were able to successfully reproduce pumping tests in a fractured block of monzonitic gneiss by treating discrete fracture zones, and the rest of the fractured rock mass, as juxtaposed (nonoverlapping) fracture-dominated (single) continua. The extent to which continuum (single, equivalent, dual, multiple, juxtaposed, deterministic or stochastic) concepts may or may not apply to fractured rocks has been the subject of intense research and debate for over two decades. In such rocks, flow and transport often take place preferentially through discrete fractures and channels. Some of these discontinuities can usually be identified and mapped in surface outcrops, boreholes, and subsurface openings. This has led to the belief on the part of some that it should be possible to delineate the geometry of the subsurface "plumbing system" through which most flow and advective transport take place. Several hydrologists consider it especially feasible to construct realistic models of fracture networks deterministically or stochastically. Typically, such networks consist of discrete polygonal or oval-shaped planes of finite size, embedded in an impermeable, or at times permeable, rock matrix. Each plane is assigned effective flow and transport properties, usually at random; in some single fracture studies, these properties are further treated as random fields defined at each point in the fracture plane. Fracture network models containing thousands of planes have been used to simulate flow and tracer migration at several experimental sites, most notably in crystalline rocks of the Site Characterization and Validation (SCV) complex at 82 the Stripa mine in Sweden {Dershowitz et al. 1991; Dverstorp et al. 1992; National Research Council, 1996) and the Fanay-Augeres mine in France {Cacas et al. 1990a,b). The conceptual-theoretical framework behind the discrete fracture modeling approach was seen by Neuman (1987, 1988) as lacking firm experimental support. Neuman also questioned the practicality of the approach on the grounds that existing field techniques make it extremely difficult, if not impossible, to reconstruct with any reasonable degree of fidelity either the geometry of the subsurface plumbing system (which consists of porous blocks, fractures, and channels that are known to evolve dynamically along fracture planes and fracture intersections) or the flow and transport properties of its individual components. Indeed, Tsang and Neuman (1995) recently reached a similar conclusion based on extensive experience gained during the six-year international ENTRAVAL project. The authors pointed out that several ENTRAVAL field hydraulic and tracer experiments have proven equally amenable to analysis by discrete and continuum models, rendering the validation of either approach difficult. The best models appeared to be those that were neither too simplistic nor too complex. A recent summary of the international Stripa project {SKB, 1993) has concluded that while it has been possible to construct working fracture network models with thousands of discrete planes for the SCV site by calibrating them against observed hydraulic and tracer data, these models have not necessarily performed better than much simpler and more parsimonious continuum models (only very elementary continuum models were considered in this comparison; more sophisticated stochzistic continuum models were not considered and could potentially perform much better). The idea of representing 83 fi^ctured rocks as stochastic continua, originally proposed by Neuman (1987), has been adopted by Tsang et al. (1996) in their recent analysis of flow and transport at the Aspo Island SKB Hard Rock laboratory in Sweden. Most single-hole and cross-hole pneumatic test data at the ALRS have proven amenable to analysis by means of a single fracture-dominated continuum representation of the fractured-porous tuff at the site. This is in line with the more general conclusion by Neuman (1987, 1988) that flow and transport in many fractured rock environments is amenable to analysis by continuum models which account in sufficient detail for medium heterogeneity and anisotropy. Only when one can distinguish clearly between distinct hydrogeologic units or features such as layers, faults, fracture zones, or major individual fractures of low or high permeability, on scales not much smaller than the domain of interest, should one in our view attempt to model them discretely (delineate their geometry deterministically); one should then still consider treating the internal properties of each such discrete unit as random fields. Recent evidence that the latter idea often works can be found in the University of Arizona theses of Kostner (1993) and Ando (1995), and in the work of Guzman and Neuman (1996). The two theses demonstrate that hydraulic and tracer tests in saturated fractured granites at the Fanay-Augeres mine in France can be reproduced by means of continuum indicator geostatistics {Joumel, 1983, 1989), and fracture-dominated single-continuum stochastic flow and transport models, with better accuracy than has been achieved previously with discrete fracture network models. Illman et al. (1998) demonstrate that small-scale, single-hole, steady state pneumatic permeability data from unsaturated fractured tuffs at the ALRS are likewise 84 amenable to continuum geostatistical analysis, exhibiting both anisotropic and random fractal behaviors. Most type-curve models currently available for the interpretation of single-hole and cross-hole fluid injection (or withdrawal) tests in fractured rocks fall into three broad categories; I) those that treat the rock as a single porous continuum representing the fracture network; 2) those that treat the rock as two overlapping continua of the dual porosity type; and 3) those that allow an additional major fracture to intersect the injection (or withdrawal) test interval at various angles. The prevailing interpretation of dual continua is that one represents the fracture network and the other embedded blocks of rock matrix. We take the broader view that multiple (including dual) continua may represent fractures on a multiplicity of scales, not necessarily fractures and matrix. When a dominant fracture is present in a type-curve model, it is usually pictured as a highpermeability slab of finite or infinitesimal thickness. To allow developing analytical solutions in support of type-curve models, the continua are taken to be uniform and either isotropic or anisotropic. The test interval is taken to intersect a dominant fracture at its center. Either flow across the walls of such a fracture, or incremental pressure within the fracture, are taken to be uniform in most models. Flow is usually taken to be transient with radial or spherical symmetry, which may transition into near-uniform flow as one approaches a major fracture that intersects the test interval. Some models account for borehole storage and skin effects in the injection (or withdrawal) interval. In this dissertation, interpretation of transient data from single-hole pneumatic injection tests at the ALRS are conducted by means of modified single-continuum type- 85 curve models developed for spherical flow by Joseph and Koederitz (1985); for radial flow by Agarwal et al. (1970); for a single horizontal fracture by Gringarten and Ramey (1974); and for a single vertical fracture by Gringarten et al. (1974). We interpret crosshole injection tests by means of a type-curve model developed for spherical flow in an anisotropic continuum by Hsieh artd Neuman (1985a). Our modifications consist of recasting the first three models in terms of pseudopressure; developing corresponding expressions and type-curves in terms of (pseudo)pressure-derivatives; and in Chapter 5 adding compressible storage in monitoring wells to the model of Hsieh and Neuman (1985a). A brief description of these modified models is given in the following sections and in Chapter 5. 3.1 SPHERICAL GAS FLOW MODEL Single-hole pneumatic pressure data tend to stabilize with time in most injection intervals at the ALRS. As recharge boundaries are not likely to be a major cause of such stabilization at the site, we consider instead a model in which air is injected at a constant volumetric flow rate 0 [L^T'] fi-om a spherical source of (equivalent or pseudo) radius [Z,] into a uniform, isotropic continuum of infinite extent. This results in a spherically symmetric flow regime governed by (2.1-22) and subject to the following initial and boundary conditions (Appendix F) vi/(r,0) = 0; r>r^ (3.1-1) lim M/(r,/) = 0; />0 (3.1-2) 86 2Q„rp^ dt " QJTp^\ dr 2;zArsw I T ' ^ (3.1-3) (3.1-4) JC where most variables have been defined in Chapter 2; p,„ is pressure in the injection line; Cw is storage coefficient of the injection interval w; is pseudopressure in the rock just outside the skin; ^ is a dimensionless skin factor; the subscript sc denotes standard conditions; and (Guzman and Neuman, 1994, eq. (C.7), p. C-2; see also Energy Resources Conservation Board, p.2-34) (3.1-5) ~ T^P Eq. (3.1-3) represents mass balance in the injection interval under the assumption that gas density within it is the same as in the injection system and the rock. Eq. (3.1-4) relates pressure in the injection interval to that in the rock by means of an additive skin factor. The latter represents an infinitesimal skin or membrane that resists flow but does not store fluid so that flow across it takes place at a steady rate. Reduced permeability around the injection interval is represented by a positive skin factor, enhanced permeability by a negative skin. The above formulation is analogous to that of Joseph andKoederitz (1985) who additionally write 87 where b is the length of the actual cylindrical injection interval; this follows from equating the spherical source volume to that of a prolate-spheroid that fits snugly into the cylindrical interval. Upon defining dimensionless pseudopressure, time, radius and borehole storage coefficient as \yv , __ 'd ~ r , <^^ir • 4 » =1 -— r _ C^p C.= A7r<f>rl (3.1-7) r>r_ (3.1-8) r >r„ (3.1-9) \ 3^ (3.1-10) respectively, it is possible to recast the problem in dimensionless form. Here, we have 4 _ defined the storage coefficient of the injection interval to be Q = -^LP The corresponding solution has the same form as that given by Joseph andKoederitz m'.d = - ! (3.1-11) /1^2|VXCd [(i - + jVI(i + + (l + where W ^D is dimensionless pseudopressure in the well, L' is inverse Laplace transform and X is the corresponding transform parameter. Inversion of the Laplace transform solution was accomplished numerically using an algorithm due to DeHoog et al. (1982). A computer code for numerical inversion of Laplace transformation based on this algorithm is provided in Appendix G. 88 Figure 3.1 is a logarithmic plot of versus for various values of C^, when s = 0. The early unit slope of the type-curve is a diagnostic feature of borehole storage. The stabilization of case spherical) flow. at late time is diagnostic of three-dimensional (in our Type-curves corresponding to Co> 2.5x10^ are difficult to distinguish from each other. Under test conditions at the ALRS, this means that one cannot use type-curves based on the spherical model to determine porosity from singlehole test data. The dimensionless time derivative of pseudo-pressure is obtained directly from (3.1-11) in the form /1^|VICd[(I Figure 3.2 shows type-curves of + (l + (solid) and (dashed) versus ^ for various values of Q and s. The derivative is seen to climb to a peak and then to diminish asymptotically toward zero at a rapid rate. 3.2 RADIAL GAS FLOW MODEL Some single-hole pneumatic test data from the ALRS appear to fit a radial flow model during part or the entire test. The model we use has been modified after Agarwal et al. (1970; see also Raghavan, 1993, p. 68, eq. 4.105) in the manner mentioned earlier. Their model is in turn a modification of an earlier solution due to Papadopulos and 89 Cooper (1967) by adding to it the effect of a skin. Airflow is governed by eq. (2.1-22) subject to \ P C '2-QJ'Ps. <Py)f J w(r,0) = 0 (3.2-1) Hm w{r,t) = 0 (3.2-2) f dt -Tdcb 7- — f r dw\ =1 Q^Tp^V dr)^ a IeAS Tdcbr^ r. J (3.2-3) (3.2-4) The dimensionless pseudo-pressure and its derivative are given by w "wD K„{A.) + sy[lK,[-jA) = / '^ ^ (3.2-5) A|VI/:.(>/I) + /iCo[^o(VI)+iVI/:,(VI)]|^ d w wD d \x\ t r _ T-i K^{X) + SyflK,[4J) (3.2-6) + ACo[^o(VI)+iVI^,(VI)]}^ which, in this case, can be inverted analytically to read Vf WD 4 f« r^ ^ 1 - exp(- v-/o) ^ ° ^'{[^^D"^o(v')-(i-Co^')y,(v)]"+[vCoK„(v)-(i-Cojrv-)?;(v)]"| (3.2-7) 4 f =— ^In/r w- Jc ° ,(l-exp(-v'fo)) dv +[vC„r„(v)-(i-Co^=)?;{v)]'} (3.2-8) 90 where now nkb( w Q^yTpJ (3.2-9) (3.2-10) r r>r r. C ^ 27d)r^<f> where ^ 2<l> = jArlp . The solution is plotted on logarithmic paper versus in Figure 3.3. At early time, Asymptotically, (3.2-11) (3.2-12) -^ = 0 (solid) exhibits a unit slope due to borehole storage. becomes proportional to In/^, and its derivative (dashed) with respect to In/^ becomes a constant. This makes radial flow easy to distinguish from spherical flow. sensitive to Both the pseudo-pressure and derivative type-curves are sufficiently to allow extracting porosity values from time-pressure data that match the radial type-curves. 91 l.C&Ol l.CEKX) I.G&Q2 I.C&04 l.QE+OO l.O&Ol 1.0&02 l.CGKB Figure 3.1; Type-curves of dimensionless pseudopressure in injection interval versus normalized dimensionless time for various Co and ^ = 0 under spherical flow. 92 I.0EH)5 1.0&-10 Figure 3.2: Type-curves of dimensionless pseudopressure and its derivative in injection interval versus dimensionless time for various Co and s under spherical flow. approxiinatB b^inning of pseudo-steacty state unit slope behavior l.O&Ol 1.0EK)I 1.0EK)3 tc/Oo Figure 3.3: Type-curves of dimensionless pseudopressure and its derivative in injection interval versus dimensionless time for various Co and j = 0 under radial flow. 94 3.3 UNIFORM FLUX HORIZONTAL AND VERTICAL FRACTURES Some of the early singie-hole pneumatic time-pressure data from the ALRS delineate a half-slope on logarithmic paper. Such behavior is typically observed in boreholes that are intersected by a high-permeability planar feature like a major fracture. Gringarten and Ramey (1974) developed a uniform flux model for a horizontal fracture in an infinite, uniform confined aquifer. This model was developed in the petroleum industry to model pressure responses in wells that were completed with large fractures induced by means of hydraulic fracturing to enhance oil recovery. This particular model can be classified as a hybrid model, which differs from the continuum models in a sense that a discrete feature is built explicitly into a porous continuum. In this model, flow rate across the fracture is uniform and equal to the rate of injection or withdrawal. The fracture has finite thickness is disk-shaped with radius /y, centered about the well, and lies at an elevation Zj. above the bottom of the aquifer (Figure 3.4). The presence of the fracture renders the aquifer anisotropic with horizontal permeability greater than the vertical permeability k,. Our modified version of this solution is r I 4 A • '^ fD / >I Jexpl n > ^ ( n - T T T n ^ — ^ I'd /o 4rJ I 2^-0 J ( ^ if P J •H — J •H — J •' ' " I J where the dimensionless variables are now defined as 95 (3.3-2) r (3.3-3) (3.3-4) = _ (3.3-5) The derivative with respect to In/^ is (Gradshteyn and Ryzhik, 1965, p. 23, eq. 0.410) Figure 3.5 shows type-curves of (solid) versus for various values of . Another model by Gringarten et al. (1974) considers a pumping well intersected by a vertical fracture of zero thickness that completely penetrates a confined, uniform and isotropic aquifer. Under the assumption that flux is distributed uniformly across the fracture, whose edges are at equal distances from the well, their solution at the center of the fracture where it intersects the active well reads, in modified form. (3.3-7) where 96 Tikbf T„ w Qsc^Tp^ (3.3-8) kfp erf is error function, Ei is exponential (3.3-9) integral, and is volumetric withdrawal/injection rate written in terms of standard conditions, from/into the fracture wliich, in turn, is equal to that from/into the well. The time derivative is given by Mathematica™ as (3.3-10) Figure 3.6 shows type-curves of (solid) and d-w^^ldXnt^ ^ (dashed) versus a pumping well at the center of the fracture. for Half-slope behavior is evident during early time when uniform flow into the fracture dominates. At late time, the flow pattern is predominantly radial. The derivative of pseudopressure (dashed line) reaches a constant value of V-i as in the radial flow case without a fracture. Solutions have also been developed for partially penetrating (Raghavan et al., 1978) and finite conductivity (Cinco-Ley et al., 1978) vertical as well as inclined (Cinco-Ley etal., 1975) fractures. These models are more complex than the ones already presented and include a larger number of parameters. They do not necessarily provide a more realistic representation of pneumatic test conditions at the ALRS than do the simpler models we just described, and many of their parameters are difficult to either define or measure at the site. We are therefore guided by the principle of parsimony in 97 working with the simplest models that nevertheless help us interpret the available pneumatic test data in a satisfactory manner. Figure 3.4; Geometry of idealized horizontal fracture in an infinite flow domain. 99 l.OEKE ho^^.OS hD=0.2 I.OEH)I hD^-3 hD=0.4 hD=0.5 hD=K).7 Q Q hD=l.O hD=2-I0 I.OEKX) l.OE-01 1.0&Q2 l.OE-OI I.OEKX) l.OE+OI 1.0EH12 to Figure 3.5; Type-curves of dimensionless pseudopressure in injection interval, normalized by dimensionless height, versus dimensionless time for various /loin uniform flux horizontal fracture model. 100 i.(e«i2 I.(eK)l : I.CKHX) : •s l.OEOl • I (^02 I.(®<12 — ^ l.OBOl l.OBKX) 1.0B401 l.OEKG l.OEKB 1.0&04 iorf Figure 3.6: Type-curves of dimensionless pseudopressure and its derivative in injection interval versus dimensionless time in uniform flux vertical fracture model at center of fracture (Xo = y ^ = 0 ) . lOl 4. TYPE-CURVE INTERPRETATION OF SINGLE-HOLE PNEUMATIC INJECTION TEST DATA 4.1 SINGLE-HOLE TEST METHODOLOGY Single-hole pneumatic tests were conducted to help characterize unsaturated fractured rocks at the ALRS on nominal scales (borehole interval lengths) of 0.5, 1.0, 2.0 and 3.0 m. The tests and their analyses by means of steady state formulae, were conducted by Guzman et al. (1994, 1996) and Guzman and Neuman (1996). Each test involved the injection of air into a section of a borehole isolated by means of inflatable packers. The corresponding air injection system is depicted schematically in Figure 4.1. The system included a straddle packer assembly consisting of two inflatable rubber bladders (each 3 m in length-see Plate 4.1), a set of flow meters and flow controllers, pressure valves and regulators, and an electronic monitoring system to automatically record field data. The distance between bladders was adjustable. Air pressure, temperature and relative humidity were recorded in the test interval. Mass flow meters and controllers were used to cover two ranges o f equivalent volumetric rate, 0 - 0 . 1 SLPM (standard liters per minute) and 0-20 SLPM. Flow rotameters served as visual backup to help maintain a constant volume rate between 0.01 SLPM and 20 SLPM. Measurements were recorded using a data logger connected to an optically isolated interface, which allowed periodic downloading onto an on-site personal computer. Recording was done at small intervals during early times when changes in pressures were expected to be largest and at successively longer intervals thereafter. 102 Table 4.1 lists the coordinates of six boreholes subjected to single-hole air permeability testing at the ALRS. The coordinate system is shown in Figure 1.2 and its origin coincides with the lower lip of the near-surface casing in borehole Z3. The table also lists the approximate length and dip of each borehole. A three-dimensional perspective of all l-m scale air permeability test locations, viewed from the Southwest toward the Northeast, is shown in Figure 1.7. Each point in the perspective represents the center of a borehole test interval between packers. Each test interval is identified by borehole and the distance along the borehole from its center to the L.L. point listed in Table 4.1. There were neither gaps nor significant overlaps between test intervals in a borehole. Prior to each air injection test, the packers were inflated to isolate the test interval, and the resulting pressure was allowed to dissipate. The test commenced by injecting air into the packed off interval at a constant mass flow rate. This continued until the pressure stabilized so as to vary by not more than the equivalent of 1 mm of mercury in 30 min. The test continued by incrementing the mass flow rate and monitoring pressure until it attained a new stable value. Most tests included three or more such incremental steps of mass flow rate. Injection was then discontinued and the pressure allowed to recover back to atmospheric. 103 Table 4.1: Coordinates of boreholes subjected to air permeability testing (adapted from Guzman etal. 1996) Bore L.L. X Borehole B B . r BB.y L.L. Z General L.L.y B.B.Z -hole [m] [m] length [m] dip [m] [m] [m] [m] direction V2 30 6.8 4.24 6.84 0.01 4.2 -30.0 Vertical W2a 5.42 5.4 -0.03 30 0.2 21.46 SSE -21.2 X2 20.44 30 -0.8 10.0 WSW 10.03 -0.02 -21.2 Y2 20.04 30 5.2 WSW 5.20 -0.31 -1.2 -21.5 Y3 30.07 5.3 WSW 45 -1.8 5.35 -0.27 -32.0 9.80 Z2 31.0 0.0 ENE 0.03 30 -0.20 -21.4 Coordinates are shown in Figure 1.2 and their origin is at lower lip of Z3 casing. Borehole length is approximate. L.L. marks lower lip of near-surface casing in each borehole; B.B. marks borehole bottom (approximate). Table 4.2; Nominal Scale and Number of Single-Hole Pneumatic Injection Tests at the ALRS adapted from Guzman et al. 1996). Nominal Scale Number of Intervals j Borehole (Length of Test Tested Interval) 1.0 m 21 V2 37 1.0 m W2A 30 1.0 m X2 2.0 m 10 X2 54 0.5 m Y2 28 l.O m Y2 2.0 m 14 Y2 9 3.0 m Y2 39 1.0 m Y3 1.0 m 28 Z2 270 Total The packers were deflated, the instrument was repositioned in the borehole, and testing resumed until the entire uncased length of a borehole has been tested. The method has proved reliable in that repeated tests of selected intervals, over several years, have 104 given highly reproducible permeability estimates. This was due in part to a strict quality assurance and quality control protocol at each stage of testing. Tests were conducted along borehole intervals of lengths 0.5, I.O, 2.0 and 3.0 m in borehole Y2; 1.0 and 2.0 m in borehole X2; and 1.0 m in boreholes V2, W2A, Y3 and Z2. A total of more than 270 single-hole injection tests have been completed, of which 180 were conducted along l-m sections in six boreholes (Figure 1.7 and Table 4.2) that span 20,000 of unsaturated porous fractured tuff. Guzman et al. (1996) relied on a steady state analysis of late data during each step to obtain a corresponding pneumatic permeability value. They used a modified formula originally developed by Hvorslev (1951; see also Hsieh et al. 1983) which assumes that, during each relatively stable period of injection rate and pressure, air is the only mobile phase within the rock near the test interval and its steady-state pressure field has prolate-spheroidal symmetry. This implies that the rock forms a uniform, isotropic porous continuum. The formula reads (after substituting eq. 3.1-5 into eq. 1, p. 10, of Guzman et al. (1996)) ^ _ Qpft\n{blrJ) (4-1) where k is permeability [Z,"], Q is volumetric flow rate at standard conditions [L^T'\ H is dynamic viscosity of air at standard conditions (l.81 x 10 ^ Ns/m'^, b is length of the test interval [Z,], interval, is borehole radius (0.05 m ) , p \ s steady-state pressure in the injection o is ambient air pressure. 105 During single-hole tests, air injection was conducted at a number of incremental volumetric flow rates. The type-curve analyses of transient single-hole test data that we present below concern only the first of any such multistep sequence. Date Logger CR-IO Flow CoatroUer Lme Dner, Futer, Pieanre Tmuducer. Thermictor Pressure Reservoir Pieaiure Tnmaducer (S) Valve Thennitrnr ^ Relative Hbunidity CeU KMolator Figure 4.1: Schematic diagram of the air injection system (adapted from Guzman et al. 1996). 106 Plate 4.1: Photogn^of single-hole packer system en^toyed during the study of Guzman «/a/.(1996). 107 4.2 PHENOMENOLOGY OF SINGLE-HOLE TESTS Figure 4.2 shows how air pressure and temperature varied during a typical multistep single-hole pneumatic injection test, labeled VCClOOl, conducted within a \-m interval whose center was located 10.37 m below the LL marker (Table 4.1) in borehole V2 at the ALRS (our designation of single-hole tests follows that established by Guzman et al. 1994, 1996). Whereas temperature was nearly constant throughout the test, pressure first rose to a peak and then declined toward a stable value. The same pressure phenomenon is seen more clearly in a plot of pressure versus time during the first step of test CGA1120 (Figure 4.3), conducted within a X-m interval whose center was located 20.15-/W below the LL marker (Table 4.1) in borehole X2. Simulations of two-phase flow in a porous medium by Guzman et al. (1996) have confirmed that this phenomenon is due to displacement of water by air in the immediate vicinity of the injection interval. When air is injected into a rock that contains water at partial saturation, the latter acts to block its movement. Hence the permeability one computes for air is lower than what one would compute in the absence of a water phase. It follows that the computed air permeability is less than the intrinsic permeability of the rock. Indeed, Guzman et aL (1996) were able to demonstrate computationally that the higher is the applied pressure during a test, the closer the computed air permeability is to its intrinsic value. They also found that, in most test intervals, pneumatic permeabilities show a systematic increase with applied pressure as air displaces water under two-phase flow. Only in a few test intervals, which were intersected by high-permeability fractures, did air permeability 108 decrease with applied pressure due to inertial efTects. In many cases, air permeability exhibited a hysteretic variation with applied pressure. The observed stabilization of pressure at late time during each step recurs in more than 90 percent of single-hole tests at the ALRS. This could, in principle, be due either to the establishment of three-dimensional flow around the test interval or the presence of an atmospheric pressure boundary in its vicinity. As pressure and pressure derivative data from most test intervals fit type-curves based on a spherical flow model in an infinite domain, we attribute the observed pressure stabilization to three-dimensional flow (Jllman et al. 1997; Illman et al. 1998). Figure 4.4 shows the same data as those in Figure 4.3 re-plotted on logarithmic paper. It reveals that, at early time, the pressure increases linearly with a unit slope that is diagnostic of borehole storage due to gas compressibility (van Everdingen and Hurst, 1949; Papadopulos and Cooper, 1967, Agarwal et al. 1970, Joseph and Koederitz, 1985). Only when this straight line starts curving does air start penetrating into the rock in a measurable way. The ensuing period of transitional flow, prior to pressure stabilization, is seen to be extremely short and dominated by two-phase flow effects. As such, it cannot be analyzed to yield reliable values of fracture porosity or borehole skin factor. A similar difficulty is encountered in the majority of single-hole tests at the ALRS in which the pressure stabilizes in a manner typical to three-dimensional flow. The two-phase flow efifect is more pronounced on an arithmetic plot than on a logarithmic plot of pneumatic pressure versus time. 109 Figure 4.5 shows a relatively rare example of a single-hole test (JJA06I6 in a \-m interval with center located 17.77 m from the LL mark along borehole) at the ALRS in which the pressure does not stabilize but continues to increase in a manner that is characteristic of radial flow (on semi-logarithmic paper, the late pressure data would delineate a straight line). Here the transient period, following that dominated by borehole storage, is sufficiently long to allow extracting from the corresponding data information about air-filled porosity and skin factor. Our attempts to do so, for this and other data that appear to fit radial type-curves, led to air-filled porosity values that are much smaller than those we obtain later by a numerical inverse analysis of similar data and from crosshole tests (Illman et ai 1998); we therefore suspect the reliability of porosity values obtained from radial type-curve fits at the ALRS and do not quote them in this dissertation. In a still smaller number of tests, such as that illustrated in Figure 4.6 (JHB0612 in a \-m interval with center located 15.81 m from the LL mark along borehole), pneumatic pressure data delineate a straight line with a half-slope at early time on logarithmic paper. Such behavior is diagnostic of a highly conductive planar feature such as a wide fracture. Indeed, a televiewer image of the test interval in Figure 4.7 reveals the presence of a wide-open fracture or cavity within it. At late time, the pressure in Figure 4.6 continues to climb as is typical of radial flow. We analyze this test in this chapter with the aid of various analytical continuum (radial, spherical) and fracture (horizontal, vertical) flow type-curve models. In contrast. Figure 4.8 shows an example of a test (ZDC0826 in a \-m interval with center located 13.56-m from the LL mark along borehole Z2 in which an early half-slope indicates the presence of a conductive planar 110 feature, and the stable late pressure data are indicative of a three-dimensional flow regime or possibly a recharge boundary. Some single-hole tests, during which the pressure eventually stabilizes, contain a sufficiently long transition period that is amenable to analysis by type-curves based on a radial flow model. This is true about numerous \-m test data from boreholes W2A and Y2. It suggests to us that flow around such test intervals evolves from radial to threedimensional with time. Though it seems to allow extracting from these tests information about air-filled porosity and skin factor, we consider the results of such analyses unreliable, for reasons mentioned earlier, and do not quote the corresponding parameter values in this dissertation. The fact that transient pressure behavior is not entirely consistent across the site, but varies from one test interval to another, provides a qualitative indication that the site is pneumatically nonuniform and the local rock is heterogeneous. soo § 40U) oc Ui 01 -aoap 2 o UJ o m Ui QC oA -4oaj) "3^ Q. Ui o 200 3 -— ffitchgwH — Protur* -OO TIME (MIN) Figure 4.2; Example of multi-rate single-hole test (VCCIOOI in borehole V2 at 10.37 /n from LL marker) (adapted from Guzman et al, 1996). 112 70000 60000 \ SOOOO 40000 / pcMan TMdiM Maaify vriiw £ a. 30000 20000 10000 5000 15000 10000 20000 25000 «^1 Figure 4.3: Aritfamedc plot ofpfessuie data fituii test CGAl 120 IBHJS Hernfy^itme 1EH)4 2piiaMflaw lEHO i. lE+02 lE+01 lE+00 1&02 lE-Ol lE+00 1EH)1 IEHJ2 /fuel lEHB Hgure4.4: LogahtfamicplotofptessuredatafiantestCGAl 120 lEW 1EH)5 16+06 113 1EHJ5 lE+Ot ibhb g. Q. lE+02 lE+OI 1E«» IBOl IfiHX) 1EH}1 1BH)2 1E«<)3 IE+04 1EH>S 1EH)6 Figure 4.5: Logarithmic plot ofpiessure data fiom test JGA0616 lE+OS lEHM paendo-iteKfy 1E«B 1EH)2 IB«)1 lE+00 lE-01 lEHW lEtOl lE+02 lB+03 /(MCI Figure 4.6: Logaridmnc plot of pressure data fiom test XHB0612 lEt04 lEHJS lEt06 Figure 4.7: BETTV image taken in b<KdioleY2(hi^penneabiIily zone) 1.<E«)5 T l.CEHM l.OE+iB I 1.0Et02 l.(eH)I l.(BH)0 l.Q&Ol ' • • l.flE+00 • • i • • i ....i l.(E+02 1.QEH13 . • • ...ii l.QBHM t\suc\ Figute4.8; Logaiithmic plot of pressure data from sinde-hcrie test ZDC0826 -• l.(E+05 l.QE+06 115 4.3 TYPE-CURVE INTERPRETATION OF SINGLE-HOLE TESTS Over 40 sets of l-m scale single-hole pneumatic injection test data were interpreted by means of the spherical, radial, vertical and horizontal fracture flow models described in Chapter 3. Some of the data were interpreted using both p-based and p'based formulae in order to check the extent to which the method of linearizing the gas flow equations affects test result. As will be illustrated below, the single-hole test data examined showed little sensitivity to the method of linearization. As was already mentioned, the majority of our data conform to the spherical flow model regardless of number or orientation of fractures in a test interval. We interpret this to mean that flow around most test intervals is controlled by a single continuum, representative of a threedimensional network of interconnected fractures, rather than by discrete planar features. Only in a small number of test intervals, known to be intersected by widely open fractures, have the latter dominated flow as evidenced by the development of an early half-slope on logarithmic plots of pressure versus time; unfortunately, the corresponding data do not fully conform to available type-curve models of fracture flow. Some pressure records conform to the radial flow model during early and intermediate times but none do so fully at late time. Figure 4.9a shows visual fits between pressure (circles) and pressure derivative (triangles) data from test CAC08I3 (in a l-m interval with center located 4.47 m from the LL mark along borehole X2), and /;-based type-curves corresponding to the spherical flow model. Pressure derivatives have been calculated using an algorithm described in Appendix H. Figure 4.9b shows similar fits between incremental squared pressure 116 (circles) and derivative (triangles) data, and type-curves corresponding to a spherical flow model expressed in derivative ^^w^/^ln/o » terms of dimensionless pseudopressure and its recall from (2.1-16) that pseudopressure, w, is proportional to incremental squared pressure, Ap', when /zZ is constant as we take it to be at the ALRS. Both the />-based and /7*-based sets of data exhibit a good match with type-curves that correspond to zero skin (s = 0); indeed, most test data from the ALRS show little evidence of a skin effect. The two sets of data yield similar values of air permeability, 1.29 X 10"'' based on p and 1.56 x 10"^' based on p^. Each set of early-time data falls on a straight line with unit slope, indicative of compressible air storage within the test interval. A pressure peak due to two-phase flow is not discernible on the logarithmic scale of Figures 4.9a,b. Figures 4.10a,b show similar type-curve matches for test CHB0617 (in a l-m interval with center located 24. l-m from the LL mark along borehole X2). Both the p- and p^-based sets of data exhibit a fair match with type-curves that correspond to zero skin for early to intermediate data. The two matches yield comparable permeabilities, 6.11 X 10"'^ m' based on p and 6.01 x 10"'^ m' based on p'. Late data do not match the type-curves due to apparent displacement of water by air, which manifests itself as a gradually decreasing skin effect; this two-phase flow effect is most clearly discernible when derivative data are plotted in terms of pressure squared. Each set of early-time data is strongly affected by compressible air storage within the test interval. A pressure peak due to two-phase flow is now discernible on the logarithmic scale of Figures 4. IOa,b. 117 A complete list of air permeability values and skin obtained by means of /7-based and />^-based spherical type-curve analyses is given in Table 4.3. Skin factors that appear to evolve with time due to two-phase flow are not included in the table. Figures 4.11 compares permeabilities obtained by steady-state and p-based transient analyses; Figure 4.12 compares those obtained by steady-state and p'-based transient analyses; and Figure 4.13 does so for permeabilities obtained by p'-based and /?-based analyses. Overall, all three data sets agree reasonably well with each other. The good agreement between p^based and ^-based results suggests that the method of linearization has little influence on our interpretation of \-m single-hole pneumatic air injection tests at the ALRS. This suggests that nonlinear effects due to gas compressibility have little effect on pneumatic pressure propagation through the rock, though our analysis shows that they have a major impact on gas storage within borehole injection and monitoring intervals. Figures 4.14 a,b show p- and p'-based type-curve matches for single-hole test JGA0605 in a \-m interval with center located 11.89 /n fi*om the LL mark along borehole Y2. In this case, the early and intermediate data appear to fit the radial flow model but the late pressure data stabilize, and the late pressure derivative data drop, in manners characteristic of three-dimensional flow. The same is seen to happen when we consider in Figures 4.15a and b data from single-hole test JJA0616 in a \-m interval with center located 17.77-m from the LL mark along borehole Y2. We take this to indicate that the flow regime evolves from radial to spherical with time. The \-m test interval JHB0612, with center located 15.81-/n from the LL mark along borehole Y2, intersects a fracture which on the televiewer (Figure 4.7) appears to 118 be widely open. Figure 4.16 depicts an attempt on our part to match the corresponding incremental squared pressure data to type-curves of pseudopressure based on the horizontal fracture flow model described in Chapter 3. Only the early time data appear to match one of these curves. We suspect that deviation of the late data from the typecurves is due to the fact that whereas in reality the flow evolves with time to become three-dimensional, in the model it evolves to become radial. Upon ignoring the late data and considering only the early match, we obtain an air permeability of 1.32 x 10"'^ m^. This is about four times the value of 4.8 x 10"''* m' obtained by Guzman et ai (1996) on the basis of a steady state analysis of the late data. An unsuccessful attempt to match the same data with a type-curve corresponding to the vertical fracture flow model, described in Chapter 3, is depicted in Figure 4.17. Pressure and, to a much greater extent, pressure derivative data from several single-hole pneumatic injection tests, some of which are illustrated in Figures 4.18 through 4.21, exhibit inflections that are suggestive of dual or multiple continuum behaviors. If so, we ascribe such behavior not to fractures and rock matrix as is common in the literature {Warren and Root, 1963; Odeh, 1965; Gringarten, 1979, 1982), but to fractures associated with two or more distinct length scales. However, these inflections show some correlation with barometric pressure fluctuations, implying that they might have been caused by the latter rather than by dual or multiple continuum phenomena. 119 Table 4.3; Air permeabilities obtained using p- and ^'-based spherical flow models. Test CAC0813 CBA0902 CDB1007 CGA1204 CHA0121 CHB0617 CIA0621 JFC0604 JGA0605 JGB0608 JGC0609 JHA0611 JHB0612 JJA0616 JJB0618 JKB0623 JKC062S JNA0713 2JG092 VCB0924 VCClOOl VFB0318 VHA0422 VHB0429 WFClOl YAA0301 YDA0426 YFB0621 YGB070S YGC0709 YIC1002 YLB1108 ZCB0819 ZDB0825 ZDC0826 ZFC0918 ZHA1006 ZIB1109 borehole location (m 4.5 X2 5.5 X2 12.3 X2 20.2 X2 X2 23.1 X2 24.1 26.0 X2 10.9 Y2 11.9 Y2 12.9 Y2 13.9 Y2 14.8 Y2 15.8 Y2 17.8 Y2 18.8 Y2 21.7 Y2 22.7 Y2 Y2 26.6 16.1 Y2 V2 9.4 V2 10.4 V2 18.4 V2 23.4 V2 24.4 W2A 19.6 Y3 2.6 Y3 11.6 Y3 18.6 21.6 Y3 Y3 22.6 Y3 28.6 Y3 36.6 9.6 Z2 12.6 Z2 13.6 Z2 19.5 Z2 23.5 Z2 27.5 Z2 p-based p^-based k [m'\ k [m'\ 1.3E-15 7.7E-16 l.lE-14 3.1E-16 1.4E-15 6.1E-17 5.8E-17 1.8E-15 4.3E-15 9.2E-15 2.2E-15 3.6E-15 5.1E-14 6.5E-15 3.9E-15 1.6E-15 8.2E-16 1.3E-14 3.4E-16 1.5E-15 6.0E-17 6.4E-17 1.9E-15 4.6E-15 1.2E-14 2.9E-15 4.0E-15 6.5E-14 8.0E-15 4.3E-15 9.1E-16 4.3E-15 1.7E-16 2.6E-14 3.2E-16 1.4E-15 8.3E-17 1.6E-17 2.9E-15 1.2E-16 3.1E-16 1.3E-15 1.5E-16 3.1E-17 2.5E-17 9.0E-16 4.9E-14 5.0E-16 1.2E-17 1.3E-17 4.4E-17 2.1E-17 l.lE-14 1.2E-14 4.6E-14 7.2E-17 7.4E-17 l.lE-15 9.1E-16 5.0E-16 l.lE-17 1.3E-17 3.5E-17 1.2E-14 5.9E-14 1.6E-16 1.2E-15 skin 1.4E-15 l.lE-15 1.8E-15 3.6E-16 1.6E-15 7.8E-17 5.9E-17 1.6E-15 3.0E-15 2.1E-15 2.0E-15 2.7E-15 4.8E-14 5.6E-15 3.4E-15 8.7E-16 1.2E-15 9.1E-17 2.8E-14 4.5E-16 1.3E-15 l.lE-16 2.7E-17 2.0E-17 8.0E-16 5.2E-14 5.6E-16 l.OE-17 l.OE-17 7.5E-18 3.9E-17 1.6E-17 1.3E-14 1.3E-14 5.3E-14 1.4E-16 8.2E-17 1.3E-15 0 negative 3 and incieasine negative 0 negative 0 0 0 and increasing 3 and increasing 0 and increasing 0 and increasing 0 and increasing 0 and increasing 0 and increasing 0 1 2 and increasing 0 and increasing 0 neagtive 0 and decreasing 0 and increasing ? 7 0 0 and decreasing ? 0 ? 7 0 0 0 negative 0 120 \\ l.QE-02 TGAC 0 Figure 4.9a: T^ipe-curve matdi of data fiom siii{^e-hde pneumatic test CAC0813 to the 1.0&02 lo/c/t Figure 4.9b: Type-curve match. oCdata ftom-siii^e-faQle pneumatic test CAC0813 to the gas, qihencal flow modd. 121 10 IE-01 lE-02 lE-02 lE-01 Figiire4.10a: Type<urve match ofdatafiomsmg^e-hdepneamatic test CHB0617 Kqoid, sphencal flow modd. o-s, 10 %• i L€&02 l.(E-01 l.(e400 LOE«01 l.(£^ l.<e^ 1.(^404 ta/cjt Figure 4.1Ob: Type<urve matdi of data fiom sin^e-hcle pneumatic test CHB0617 >with gas, qihencal flow modd. L<eH>S 122 l.OE-13 1.0E-I4 "jl l.OE-15 sv 2 l.OE-16 «5 1.0E-I7 1.0E-I8 I.OE-18 I.OE-17 l.OE-16 l.OE-15 l.OE-14 l.OE-13 l.OE-12 p-bastdk [jr 'I Figure 4.11;Scatter diagram of results of permeability obtained from steady-state and p -based spherical models. l.OE-13 l.OE-14 l.OE-15 s 3 l.OE-16 m l.OE-17 l.OE-18 l.OE-18 l.OE-17 l.OE-16 l.OE-15 l.OE-14 l.OE-13 l.OE-12 p'-bascdik [m'l Figure 4.12:Scatter diagram of results of permeability obtained from steady-state and p ^ -based spherical models t.OE-14 I.OE-15 9 l.OE-16 I.0E-I7 l.OE-18 1.0E-I8 I.OE-17 l.OE-16 I.0E-1S 1.0E-I4 l.OE-13 p-based ft [m'] Figure 4.13;Scatter diagram of results of permeability obtained firom p - andp -based spherical models. I.OE- 124 «« a a. lE+04 TO/CO FigDie4.14a: Type^nrve match of data from singte-hdeiaieuiiialic test JGA060S with Ucpud. racfi^ flow model l.(EM)2 4 «« 1.QB02 IX&Ol IXeHX) lUlEH)! IJOBKa lXSE*<a t/f/co Figure 4.14b: Tjpe-carve match of data frmn single-hole pneumatic test JGA0605 with gas, radial flow modeL IXBHM 125 l.OEtOl l.OB+Ol l.OE+02 1.0E«)4 Figiue 4. ISa: Type-curve match of data fixnn single-hcde pneumatic test JJA0616 to the gas, spbeiical flow model i^ppnBOBMte begimiog cf lE+01 Qy-10' lE+04 1&02 o Figure 4.1Sb; Type-curve match of data from single-hole pneumatic test JJA0616 to the gas, radial flow model. 126 VMcii fiadnra aohnioa >.05 LOE-tOl LOE-HU 'o Figiire4.16; Type-curve matdi of data fiom singje-hole pneumatic test JHB0612 to the horizontal fiacdue model. LOE-KU LOE-HM LdE-fOl 1.0E-HM Figuie4.17; Type-curve match of data from single-hole pneumatic test JiiB0612 to the vertical fracture mode!. 127 ^ LOE-HW to/co Figure 4.18: Type<urve matdi of data fiom sii^e-hde pneumatic test CDB1007 to the gas, radiai flow modd. LOE-HW LOE-HU LOE-HU Figute4.19: Type-curvematchofdatafiomsin^e-holepneumatictestJNA0713tottte gas, radial flow modd. LOE-HM 128 O-S.10 tf/c^ Fi0ire4.2O: lype-cmve match ofcfala from single-hole pnemnatK test JKC0625 to die 9B, spherical flow model O-S. 10 jkoeli^ tyc, Fi9iie4.21: TVpe-corvsmatchofpiessiiiedatafiamtestYFB0621 todie g^s, spheiical flowmodeL 129 CHAPTER 5: TYPE-CURVE MODELS FOR THE INTERPRETATION OF CROSS-HOLE TESTS 5.1: INTRODUCnON Subsurface materials including fractured media often exhibit anisotropy with respect to permeability. In the field, such anisotropy is due to fracturing or stratification in sedimentary materials. The permeability of an anisotropic material is described by a second-rank, symmetric, positive-definite tensor. Because anisotropy has an important effect on fluid flow and contaminant transport, the determination of this tensor is important for subsurface hydrology. Existing methods to determine the hydraulic conductivity or permeability tensor by means of pumping/injection tests are limited to horizontal aquifers in which one of the principal directions is vertical. Papadopulos (1965), Hantush (I966a,b), Hantush and Thomas (1966), Neuman et aL (1984), and Loo et al. (1984) developed methods to determine the two-dimensional anisotropic transmissivity tensor of an aquifer under conditions of horizontal flow. Weeks (1969) and Wc^ and McKee (1982) proposed methods to determine the vertical and horizontal hydraulic conductivities of confined aquifers. Hantush (1961), using partially penetrating wells presented a method to determine vertical anisotropy. A method that enables one to do the same for unconfined aquifers exhibiting delayed gravity response (or delayed yield) was developed by Neuman (1975). Interests in situations where the principal directions of the hydraulic conductivity tensor are not necessarily horizontal and vertical have risen primarily among geotechnical 130 engineers concerned with flow through fractured rocks. Virtually all the methods that have been proposed to date require that the principal directions be known prior to the test. In the absence of more reliable alternatives, these directions are usually predicted by the basis of fracture geometry information obtained from boreholes and surface or subsurface rock exposures. In the past, individual boreholes were drilled parallel to the conjectured principal directions in order to determine the corresponding principal hydraulic conductivities by conventional packer tests {Snow, 1966). Because the principal directions are orthogonal, the boreholes must be perpendicular to each other, which is a difiHcult drilling requirement. Furthermore, single-hole packer tests only give results that are valid in the close vicinity of the borehole. For example, Butler and Healey (1998) discuss the reliability of parameters obtained from single-hole tests. According to Butler and Healey (1998), skin effects may adversely affect single-hole tests such that parameter estimates from single-hole tests may be biased. They state that interference/cross- hole/pumping tests give more accurate estimates of medium properties. Illman et al. (1998) clearly show through their type-curve analysis that their single-hole data from the ALRS are not largely affected by skin effects (see section 4.3). To enable the measurement of principal hydraulic conductivities on a larger scale (between boreholes), Louis (1974) proposed a method that requires drilling one injection hole and two monitoring holes parallel to any one of the three (known) principal directions. The two monitoring holes are drilled so that the shortest lines between them and the injection hole are parallel to the remaining two principal directions. A "hydraulic triple probe", consisting of four packers, is inserted into the injection hole to create three 131 isolated intervals (see Figure 2 in Hsieh and Neumcm, 1985a). Fluid is injected into all three intervals at a known rate and pressure. The role of the two "guard intervals" is to minimize deviations from radial flow around the central interval. The measurement of pressure inside the short intervals between long packers in the monitoring holes makes it possible to compute all the three unknown principal hydraulic conductivities. In this chapter, a field method originally proposed by Hsieh and Neuman (1985a) is applied to the determination of the three-dimensional permeability tensor of an unsaturated anisotropic medium. Their method differs from previous approaches in two important ways: (1) it does not require that the principal directions be known prior to the tests and (2) the boreholes may be drilled in any directions that are technically feasible (e.g., they may be vertical, slanted or horizontal). Here, we present an extension of solutions given in Hsieh and Neuman (1985a), which account for gas flow conditions, observation wellbore storage and skin effects. In the absence of observation wellbore storage and skin effects, the solution approaches that given by Hsieh and Neuman (1985a). While the method is applicable to both porous and fi-actured media, we concentrate on the latter because it is fi-acturing that most often prevents one from determining a priori the principal directions. Because the permeability of fractured rocks is often quite low, we consider fluid injection rather than fluid withdrawal. However, for more permeable media in which pumping is possible, the same test could be performed by reversing the direction of flow without any changes in the method of interpretation. The proposed method is referred to as cross-hole testing {Hsieh and Neuman, 1985a), to distinguish it from the more conventional single-hole packer. It consists of 132 injecting fluids into packed-off intervals in a number of boreholes and monitoring the transient pressure response in monitoring intervals in neighboring boreholes. An important aspect of the method is that it provides direct field information on whether or not the rock behaves as a uniform, anisotropic continuum on the scale of the test. We first describe the mathematical model, then systematically present theoretical expressions for transient type-curves given in Hsieh omd Neuman (1985a) modified to account for gas flow conditions. Expressions of pressure derivatives not given in Hsieh and Neuman (1985a) are also given. We then describe a method to modify the Hsieh and Neuman (1985a) solutions to account for observation wellbore storage and present its pressure derivatives. The method to account for observation wellbore storage is based on the theory by Black and Kipp (1977). After we discuss the characteristics of type-curves, we employ those new curves to analyze a set of pressure data obtained fi"om a cross-hole pneumatic interference test conducted in unsaturated fractured rocks at the ALRS. For purposes of cross-hole test analysis, we represent the fractured rock by an infinite three-dimensional uniform, but anisotropic continuum, as was done by Hsieh and Neuman (1985a). In terms of pressure, single-phase airflow is then governed approximately (due to linearization) Ot where the permeability is written in terms of a tensor and the variables were defined previously. We choose a convenient set of "working" cartesian coordinate system, x,, / = 1,2,3, so that a point in the flow domain is located by the vector x = (x,}, and the 133 permeability tensor is represented by the matrix k = definite and symmetric. j. We assume that k is positive- The initial pressure distribution in the medium (and hence pressure) is assumed to be uniform: p{x,t) = 0- t=0 (5-2) We also begin by assuming that the flow domain is infinite, so that p ( x j )= 0 ; 1 * 1 ( 5 - 3 ) The injection and observation intervals are idealized, depending on their lengths, as points or lines. In the following analysis, we deal only with fluid injection into the test interval. The analysis is identical for fluid withdrawal if the increase in pressure is replaced by decrease in pressure. The following four cases are considered; (1) injection at a point, observation at a point; (2) injection along a line, observation at a point; (3) injection at point, observation along a line; and (4) injection along a line, observation along a line. 5.2: POINT INJECTION/POINT OBSERVATION Consider a point source injecting at a constant volumetric rate, Q, at the origin of the working coordinate system. The increase in pressure at a point x and at time / is given in dimensionless form by =e^c{l/(4/o)"'} (5-4) where etfc[ ) is the complementary error function and is the dimensionless pressure increase. 134 (5-5) The subscript denotes the case number and is the dimensionless time ta=Dt/(S^^J while D (5-6) is the determinant of k, i.e., D — ^11^22^33 ^11^23 ^33^12 (5-7) and G„ is the quadratic form, defined as (5-8) = x ^ \ x = x , x ^ A,J where the superscript T indicates transpose. Unless stated otherwise, summation of repeated indices is implied. The matrix A in (5-8) is known as the adjoint of k, its components are given by (no summation implied below) (5-9) ^•j = ^j' = (5-10) where /,y,and k are cyclic arrangements of 1,2, and 3 (i.e., /,7,^ = 1,2,3, or 2,3,1, or 3,1,2). It can be shown that A, like k, is symmetric and positive-definite. The time derivative of the point injection/point observation solution is c/ln/o = - M e x p ( - u) 2V;r where u = 1/4/^ . For isotropic media, ^12 = the (scalar) permeability. Consequently, Thus d = k ^ , k = = R^k', where R ' = ^13 = 0. (5-11) i = ^22 = ^33 = ^ ^ where k is , where I is the identity matrix. is the radial distance from the point source, and the 135 dimensionless expressions (5-5) and (5-6) take the simple form = AjipkR/Q/j and = k f p j . The solution for the isotropic case is seen to possess spherical symmetry about the point source. A logarithmic plot of (solid curve) and dpa^fdXvit^ (dashed curve) versus is shown in Figure 5.1. It is interesting to note that a comparison of the point source and spherical flow solutions with wellbore storage and skin (Figure 3.2) shows that the derivative decays much sooner for the latter model. This is because when borehole storage in the injection borehole is dominant (in comparison to borehole storage in observation intervals - shown later), the transient fluid behavior becomes very short in duration and approaches steady-state at a much faster rate (which is clearly shown through the derivatives). 136 l.OE+01 l.OE+00 • I.0E^2 1.0Er02 I.OE-Ol l.OE+00 l.OE+Ol l.OE+02 I.OE+03 l.OE+04 to Figure 5.1; The point source solution (solid curve) and its time derivative (dashed curve). 137 5.3: LINE-INJECnON/POINT-OBSERVATION The solution for a finite line source injecting at a constant volumetric rate, Q, can be obtained from the solution for a point source by integration. Let a line source of length L be centered at the origin of the x, coordinate system and be represented by the line BOA (see Figure 4 in Hsieh and Neuman, 1985a). Let OA be represented by the vector (5-12) so that OB is -1. In addition, let e and e, be unit vectors in the direction x and I respectively, so that x = Re and I = (Ll2)e,. If we assume that flux along the line source is uniform, then the solution at point x is given in the following dimensionless form Pd,=]: J -expf- (l - a2')y\-+ l/a.)]" *:rf[y"'{a2 - l/a,)]}«^ (5-13) where erf{^ ) is the error function, and the two geometric parameters a, and a,, are defined as (5-14) (5-15) and Gj = x'"AI = x,/^4, (5-16) 138 G„=rAl = /./,4 (5-17) The dimensionless pressure increase is now = inpQ^'-lQix the dimensionless time, (5-18) is the same as in (5-6). parameters a, and a, become evident when The meanings of geometric , and G„ are expressed in terms of R^L, e, and e,: =/?-(e'"Ae) (5-19) G^={RL/2)le''\e,) (5-20) G^=(^V4Xe/Ae,) (5-21) Then a, = (2 ^//:)[e'"Ae/e/ Ae, (5-22) a, = (e'^Ae,)^e'^Ae/e/Ae,j' (5-23) For an isotropic medium, (5-22) reduces to a, =(2/?//,), and (5-23) becomes OTj = e^e, = cos 0, where 0 is the angle between x and I. Thus a, is related to the ratio between R and L, and a, is related to the angle between x and I. For anisotropic cases, the effect of anisotropy is incorporated into the definitions. Since A is positive-definite, we have or, > 0 and 0 < ^ 1. 139 The partial derivative of eq. (5-13) can be calculated by means of Leibnitz's rule of differentiation of a definite integral with respect to a parameter {Gradshteyn and Ryzhik, p.23, eq. 0.410) 0~ Figure 5.2 shows a logarithmic plot of + I/a,)]- erf[u"-{a,_ - I/a,)]} and ^Po^l^XntQ versus (5-24) (^solid and dashed curves) for a range of or, values and the Theis (1935) solution and its derivative (open circles and triangles) are included for purposes of comparison. The line injection/pointobservation case approaches the solution of Theis (1935) when the observation point is close to the injection line. This can be seen from the plot of d/^In /j, for a, = 0.01. At late times, the derivative approaches a constant value of unity when flow reaches pseudo-steady state in which the pressure gradient becomes constant with time. As the distance between the injection line and the observation point increases, we observe that the derivative decreases with time indicating that flow becomes three-dimensional at late time. It is shown in Hsieh and Neuman (1985a - see figure 9 in their paper) that as ex, -^00, the quantityck,h' /2 approaches the point-injection/point-observation solution. o Figure 5.3 is a plot of normalized pressure head for the line-injection/point-observation case and its derivatives plotted against . We have also plotted the point-injection/point observation solution and its derivative (solid circle and triangles), the Theis (1935) model and its derivatives (open circles and triangles) to show the variability of the lineinjection/point observation solution with a,. Examination of the derivative curves shows 140 that this decrease becomes constant becomes constant) at late times owing to the three-dimensional nature of flow. a,-0.0 J & W(u) l.OE+00 a,=2.0 a,=5.0 3 Q. a, =2.0 Figure 5.2: Logarithmic plot of (solid curves) and d(dashed curves) versus for a range of a, values. The Theis (1935) solution (open circles) and its derivatives (open triangles) are included for purposes of comparison. 141 a,=0.2 l.OE-01 a,=0.2 o' l.OE-02 •''A a,=0.01. W{u) a,~O.OK dW(uydlnt| l.OE^OI Figure 5.3; Plot of normalized pressure head for the line-injection/point-observation case (solid curves) and its derivatives (dashed curves) plotted against The Theis (1935) solution (open circles) and its derivatives (open triangles) are included for purposes of comparison. 142 5.4: POINT-INJECnON/LINE-OBSERVATION The solution for the point injection/line observation case can be obtained by averaging the solution for the point injection/point observation case along the observation line. Consider an observation line of length B, such as B'CA' (see Figure 7 in Hsieh and Neuman, 1985a), and let x be the radius vertor of point C (midpoint of the observation line). Let CA' be represented by the vector b" =(A„A,>3) so that CB' is - b. In addition, let (5-25) be a unit vector in the direction of b so that b = (5/2)e, (5-26) The average pressure increase, Ap, over B'CA' can be obtained by integrating the dimensional form of the point-injection/point-observation solution along this line and dividing the result by B. However, due to our assumption of uniform flux along the line source, there is a symmetric relationship between the point-injection/line-observation case and the line-injection/point-observation case. In other words, the solution for the former case can be obtained directly from (5-13) by merely redefining some of the variables; pd, = ^ f - e x p f - ( i - A ' j j ' ] • ii=i/4fo y + y P x ) ]- )^dy (5-27) where 143 (5-28) (5-29) Pz=Gj{GJG^f- (5-30) = x^'Ab = (/25/2)(e'"Ae,) (5-31) and G,, = b'" Ab = (5V4)e/Ae, (5-32) The only difference between (5-13) and (5-27) is the replacement of I by b. Note, however, that the dimensionless pressure increases, and , are defined in different manners, which accounts for the difference in the appearances of the respective dimensionless solutions. The meanings of and are similar to those of a, and or, A=(2^S)[e''Ae/e/Ae,]'" (5-33) A=(e''Ae.)[e''Ae/e/Ae.]'" (5-34) In other words, of >9, is related to the ratio between the magnitudes of x and b, while is related to the angle between the two vectors. For an isotropic medium the dimensionless expressions (5-5) and (5-6) take the simple form = k^l[<(>fjR^^. = A7q}kR[Q jj. and The solution for the isotropic case is seen to possess spherical symmetry about the point source. The time derivative of (5-27) is ^^ = ^exp[-(l-A')"] M"'"(A -l/A)]} (5-35) 144 Figure 5.4 is a logarithmic plot of and dversus for various values of pi and p2 was icept at 0.01 (solid and dashed lines). The point injection/point-monitoring solution and its derivative has also been plotted (open circles and triangles). P5=0.01 p,=0.01.0.1.0.Z OJ. 0.4. 0.5. 1.0. 5.0 Pi-S.O and point source ^ —''"'^Pi=5.0 and derivative source solution P. =0.0 1.0E-Q2 P.=0.0 1.0E-Q2 l.OE+00 to Figure 5.4: Logarithmic plot of p,^ (solid curves) and (dashed curves) versus where the point injection/point-monitoring solution (open circles) and its derivative (open triangles) have also been plotted. 145 5.5: LINE-INJECnON/LINE-OBSERVATION The solution for the line injection/line observation case is obtained by averaging the solution for the line-injection/point-observation case along the observation line. The averaging procedure is given in Hsieh and Neuman (1985a). The final result in dimensionless form is pd , = \ f -exp[-(l-a,/);/1- |exp{-[(l-c-)/lY^,'+2(A-«2'^)A/>0;M ^ ii=i/4/o y i=-i (5-36) where ^ (i'-it) ai, a2. Pi and P2 are defined in (5-14), (5-15), (5-29), and (5-30), respectively, and c = GJ(G,IGJ'- (5-38) where = I'Ab = (iB/4)(e,''Ae.) (5-39) Note that c has a similar meaning as a? and P2, it is related to the angle between 1 and b. The time derivative for the line-injection/line-monitoring case is given by 146 ^ ^ = ^exp[-(l-a/)tt]- Jexp|-[(l-c-)ylY)ff,' +2(^2 -a2c)A/^,]wJ • + V«i + '^/A)](5-40) For the general case, the inner integral in (5-36) and (5-40) cannot be evaluated analytically. However, when b is parallel to I (the line source and observation line are in parallel boreholes), simplification is possible. For this case, a, = and c = 1. The exponent in the inner integral of (5-36) and (5-40) vanishes and the remaining error function can now be integrated analytically to yield (see Gradshteyn and Ryzhik, 1965, p.633, eq. 5.41) ^ ii=i/4fp y [exp(- yC^')- exp(- >C,') -exp(- yCi) + exp(- y C { n y ^ ' ^ d y (5-41) Similarly the pressure derivative simplifies as - C^erf[u''-C^) + C,erf[u"^'C,)+ Jexp(- mC,^ )- exp(- «C^) -exp^- uCj*) + exp(- uC,^ |du (5-42) where 147 C, = a, + 1/a, + \/fi^ C=a,+l/a,-I/^ C3=a,-l/a,+iM ^ ^ Q =«2-V«l -1/A Due to the large number of parameters in (5-41) and (5-42), numerous solution plots may be constructed. Here we consider only the case where 1 = b, so that a, = >55 Figure 5.5 shows a logarithmic plot (5-41) and (5-42) (solid and dashed lines) Oj = ^ =0 and various values of or, =^. Again, the Theis (1935) solution has been included for purposes of comparison. a, =P ,= 0.01, 02, 0.5, 1.0. Za 5.0 Oti= P2=0.01 1.0E-Q2 Figure 5.5: Logarithmic plot for l.OE+00 l.OE+02 for various values of l.OE+04 cc^=P\- 148 5.6 MODIFIED FORM OF HSIEH AND NEUMAN (1985A) SOLUTION TO ACCOUNT FOR OBSERVATION STORAGE AND SKIN EFFECTS The solutions for various injection and monitoring interval configurations derived by Hsieh and Neuman (198Sa) do not account for observation well storage and skin effects. We have seen from the analysis of single-hole data that borehole storage effects in the injection interval can be important and in some cases can cause significant delays for pressure transients to enter the formation. Here, we expect similar effects in the observation intervals arising from compounding effects of large interval volume and gas compressibility. Tongpenyai and Raghavan (1981) present a solution for radial flow, which considers storage and skin effects in both the pumping and the observation wells. We follow a much simpler approach due to Black and Kipp (1977) that is based on a concept introduced by Hvorslev (1951). According to their approach (see also Neuman and Gardner, 1989), the pressure p in the rock is related to observed pressure p^^ (5-44) dt where tg = C^^fFk, is the characteristic response time of the instrument, known as basic time lag; of volume F; - Vjp^ is storage coefficient associated with an interval is permeability (which we attribute to skin); and F is a geometric shape factor [Z,]. The basic time lag can be determined by means of a pressurized slug test so that there is no need to know either or F. Once tg has been determined, one can correct p^ for the effects of storage and skin by means of (5-44). The general solution of (5-44), in dimensionless form, is (Appendix I) 149 where 0 = 4^^/^/^^///?* is a dimensionless well response time, cr is a constant related to the surface area of the observation well intake (for example, o" = l/2;zr^6 for radial flow in which b is the length of the injection interval), and Sa is the gas storage factor [M''LT^'\ defined in Chapter 2 (eq. 2.1-23). Substituting (5-27) into (5-45) yields for dimensionless pressure, and its derivative, ^ »=l/4fo ^ • {l - exp[- I/a(l/w - l/^)]}i/^ (5-46) D TF=|/4FP ^ • {exp[- I/fi(l/« (5-47) Figures 5.6 through 5.8 depict corresponding type-curves for various values of Q, >^2 =0.01, and Pi = 5.0, 0.1 and 0.01, respectively where is represented by solid curves and^Pj^f^Xnti^ by dashed curves. There clearly is a delay in response as Q increases. The original solution (open circles, eq. 5-27) of Hsieh and Neuman (1985a) and its derivative (open triangles, eq. 5-35) are included for comparison. 150 In the case of Figure 5.6 where Pt = 5.0, radial distance between the centroids of the injection and observation intervals is large compared to the length of the observation interval. Here the pressure derivative decays towards zero as is typical of three- dimensional flow, which develops around the point injection interval. As Pi diminishes, so does the response, which is additionally delayed in time. In Figure 5.8 where y?/=O.OI, so that the monitoring interval is long relative to its distance from the injection interval, the pressure derivative corresponding to n= 1.0 and 10.0 is constant during intermediate time as is typical of radial flow (see Figure 3.3). We use these type-curves to interpret cross-hole test data from the ALRS in Chapter 6. The solution for the line-injection/point-observation with observation wellbore storage can be likewise obtained by substituting eqn. (5-13) into (5-45). P d 2 = \ J 7 ® * P [ ~ +1 / a , ) ] - - I / a , ) ] } ^ u=\/4to S - {1 - exp[- l/Q(l/tt -1/^)]}«/^ (5-48) D U=L/4F0 ^ • {exp[- l/n (l/M (5-49) 151 Figure 5.9 shows a family of type-curves for the line-injection/point-observation case with at = 0.01, aj = 0.01. The red solid line represents the dimensionless pressure in the solution of Hsieh and Neuman (1985a) while the red dashed line represents their derivatives. We notice that the type-curve with Q = 0.1 and the Hsieh and Neuman (1985a) solution coincide at late time but is slightly delayed at early time. This delay in pressure increases with the magnitude of Q as it did with the point-injection/lineobservation solution. The Theis (1935) solution and its derivatives (as open circles and triangles) are also included for purposes of comparison in this figure. We notice from this plot that the Hsieh andNettman solution matches exactly with the solution of Theis (1935) from the early to late time (except at to greater than 2000). The derivative of the Theis (1935) solution remains constant at late time as the pressure response reaches a pseudosteady state. The derivatives of the Hsieh and Neitman (1935) and our modified solution (5-51) begins to decrease as the pressure response "seen" in the point observation interval in response to line injection evolves from a radial to a spherical flow regime. This is true for all values of Q in our modified solution (5-50). The Black and Kipp (1977) solution (integral independently evaluated numerically) is included in Figure 5.10 to show the correspondence with the Hsieh and Neuman (1985a) at ai = 0.01, a2 = 0.01. 152 l.OEH)! 0=1000 0=10 I.OEtOO o=i Q 2 l.OBOI £3 O. 1.0E02 l.OBOI l.OE«)l I.OEKC l.OEKM 'o Figure 5.6: Type-curves of dimensionless pressure in observation interval (solid curves) and its derivative (dashed curves) versus dimensionless time for various Q and Pi = 5.0, >?2= 0.01. Open circles represent dimensionless pressure in the solution of Hsieh and Neuman (1985a) and open triangles their derivatives. 153 Q=IO Q=1 l.OE+OO l.OEtOI 1.0EK)2 I.OEKW to Figure 5.7: Type-curves of dimensionless pressure in observation interval (solid curves) and its derivative (dashed curves) versus dimensionless time for various O and ^/ = 0.1, ^2= 0.01. Open circles represent dimensionless pressure in the solution of Hsieh and Neuman (1985a) and open triangles their derivatives. 154 l.CEKX) : 0=100 1.0EK)l 1.0E-+03 to Figure 5.8; Type-curves of dimensionless pressure in observation interval (solid curves) and its derivative (dashed curves) versus dimensionless time for various £2 and >3/ = 0.01, >?2= 0.01. Open circles represent dimensionless pressure in the solution of Hsieh and Neuman (1985a) and open triangles their derivatives. 155 UEKU « L(E01 L(EC2 L(EOI L(EOI L(E(B L(EHB Figure 5.9: Type-cucves of dimmaonless ptessure in obsovstiim intovai (solid cuives) and its doivadve (dashed carves) versus dimensionless time for various and ai = 0.01, 0.01. The red solid line represents dimensionless pressure in die solution of Hsieh oite/JVisuman (1985a) and die red dadied line dieir derivatives. The opw circles represoit die Theis (1935) sohtion while the open triangles their daivatives. 156 i a I = 0.01 a2=qsi\ • £2=10 Q = 0.01;0.1 •• l.GB-04 1.0E-Q3 l.OE-Ol l.OE-HW to F^nue 5.10: Type-caives of dimoiafHiless piessme in obsovation interval (solid curves) and its dqivative (dashed corves) vctsus dimqisionless time for various and or/= 0.01, 02=0.01. The bine solid line r^resents dimaisionless pressure in the sohition of Theis (1935) and die blue dashed line their derivatives. The open circles represent die and Kipp (1977) sohition^lile die open tiiai^es dieir derivatives. 157 6. TYPE-CURVE INTERPRETATION OF CROSS-HOLE PNEUMATIC INJECTION TEST DATA 6.1 CROSS-HOLE TEST METHODOLOGY Single-hole air injection tests provide information only about a small volume of rock in the close vicinity of the injection interval. Our data indicate (see sections 1.2.4 and 4.2) that rock properties, measured on such small scales, vary rapidly and erratically in space so as to render the rock strongly and randomly heterogeneous. To determine the properties of the rock on a larger scale, we conducted cross-hole interference tests by injecting air into an isolated interval within one borehole, while monitoring pressure responses in isolated intervals within this and other boreholes. Of the 16 boreholes we used for cross-hole testing, 6 were previously subjected to single-hole testing. The results of the single-hole tests (primarily spatial distribution of air permeabilities and local flow geometry) together with other site information (primarily borehole televiewer images-see Appendix J) served as a guide in our design of the cross-hole tests. As there is little prior experience with pneumatic cross-hole tests in unsaturated fractured rocks, we conducted our tests at the ALRS in three phases. Phase 1 included line-injection/line-monitoring (LL) tests in which injection and monitoring took place along the entire length of a borehole that had been isolated from the atmosphere by means of shallow packers. Phase 2 consisted of point-injection/line-monitoring (PL) tests in which air was injected into a 2-m section in one borehole while pressure was recorded along the entire length of each monitoring borehole. During Phase 3, we conducted 158 point-injection/point-monitoring (PP) tests in which both the injection and the monitoring intervals were short enough to be regarded, for purposes of type-curve analysis, as points. A total of 44 cross-hole pneumatic interference tests of various types (constant injection rate, multiple step injection rates, instantaneous injection) have been conducted during the years 1995 - 1997 using various configurations of injection and monitoring intervals (LL, PL and PP). The type of cross-hole test; injection borehole, interval and rate; monitoring borehole and intervals; as well as brief comments on each test are listed in Table 6.1. A test was considered to be successful when all equipment functioned reliably throughout the entire period; injection rate was adequately controlled; and all data were recorded properly. 6.1.1 Instrumentation Used in Cross-Hole Tests Cross-hole tests were conducted using modular straddle packer systems that were easily adapted to various test configurations and allowed rapid replacement of failed components, modification of the number of packers, and adjustment of distances between them in both the injection and monitoring boreholes. For their construction we relied on relatively inexpensive PVC pipes, which can be worked on in the field using hand tools and are sufficiently flexible to slide with relative ease up and down an uneven borehole. Figure 6.1 is a schematic diagram of the main cross-hole injection string of packers. The air-filled volume of the injection interval was made smaller than it had been during single-hole testing so as to minimize borehole storage effects. The main injection string installed in borehole Y2 consisted of three packers, one near the soil surface to isolate the borehole from the atmosphere, and two to enclose the injection interval. Pressure 159 transducers were used to monitor absolute pressure and temperature in each of the three borehole sections that had been isolated from each other in this manner. In general, the most sensitive transducers (GEOKON-4500H-00I0) were placed close to the injection interval where we expected to see relatively pronounced pressure responses. Sensitive transducers were also placed furthest away fi-om the injection interval, and in monitoring intervals YIM, Z2M, Z2L, Z3M and Z3B, where we expected to see pronounced barometric pressure effects. Two types of borehole monitoring systems were employed, one with a single packer near the soil surface to monitor pressure along the entire length of a borehole (6 units - see Plate 6.1) and another, modular system with three or four packers to monitor pressure in several isolated segments of a borehole (9 units-see Plate 6.2). Monitoring intervals with a single packer near the soil surface are identified by borehole designation; for example VI, XI and Wl. Where a modular system separates a borehole into three isolated intervals, we append to the borehole designation a suffix U, M or B to identify the upper, middle or bottom interval, respectively; for example V3U, V3M and V3B. Where a modular system separates a borehole into four isolated intervals, we append to the borehole designation a suffix U, M, L or B to identify the upper, middle, lower or bottom interval, respectively; for example Z2U, Z2M, Z2L, and Z2B. Of the 9 modular monitoring systems, 7 were equipped with two tracer-sampling ports each, which we had use for air injection. 160 Table 6.1: Cross-hole tests completed at ALlRS Q Uipm\ Y2 Y2 Int. Int. [ml 100-30 0 10.0-30.0 I5.0-I70 I5.0-I70 I5.0-I70 21.0-230 IS.O-200 IS.O-200 IS.0-200 IS.O-200 26.0-28.0 23.0-25.0 21 0-23.0 CR Barometric CR CR CR Slug SlUR CR CR SiuK Bamnetnc Barometric CR CR Y2 Y2 Y2 Y2 Y2 Y2 Y2 Y2 Y2 Y2 Y2 Y2 Y2 Y2 21.0-23.0 21.0-23.0 21.0-23.0 21.0-23.0 21.0-23.0 21.0-23.0 21 0-23.0 21 0-23 0 21.0-23 0 21 0-23 0 21 0-23 0 21 0-23 0 21 0-23 0 21 0-23 0 10 00 10 10 15 10 2.0 10 10 00 0.0 00 PL26 PL27 PL28 PPI PP2 Step SH-neulron SH-aeutron CR CR Y2 W3 W3 Y2 Y2 21 0-23 0 14.7-169 16 9-19 I 15 0-170 15 0-170 05.10.15.10 PP3 PP4 PPS PP6 PP7 PPS PP9 PPIO CR CR Step Step Step CR Step CR Y2 Y2 X2 Z3 W3 Y2 Y2 Y2 15 0-170 150-170 18.5-207 15 9-179 19.2-204 15 0-170 15 0-170 15 0-170 150 500 5.0 and 10 0 5 0 and 10 0 5 0 and 10.0 500 75.0 and 50.0 250 PPI I Slug Y2 15 0-170 1000 PPI 2 PPI3 PPM CR CR SH-neutron; Step Y2 Y2 W3 15.0-17.0 15.0-17.0 25.7-269 250 05 0.5, 10. 13. 19 Test LL2 PLI PL2 PL3 PL4 PL5 PL6 PL7 PL8 PL9 PLIO PLII Flow CR CR CR CR CR CR CR CR CR CR CR CR CR Ini. Hole Y2 Y2 Y2 PLI2 PLI3 PLI4 PLIS PL 16 PL17 PLIS PLI9 PL20 PL2I PL22 PL23 PL24 PL25 LLI 1 ] 1 rz Y2 Y2 Y2 rz T2 Y2 n 50-0 100.0 8.S.I00 lao 2ao 1.0 1.0 1.0 l.o 10 10 1.0 10 ? 10 OS OS IS 18.0 Comments Pressure increase appears to be proportional to Q Could not maintain constant O RupciinnR of packer lines First successful test Responses in W2A and V2 Generator problems Generator problems Generator problems Response m VI [nadvertent lack of proper data recordmg No air iniection First point*to>point connection Same as PL14 but without driente Same with larRer O First cross>hole slufc test Repeat of PLI 7 with more data points Packer tnnation SIUK test Data lost? Added 3-packer monitoring system m V2 Isolated V2-W2 intersection No change in neutron counts No change m neutron counts Unknown flow rale Y3 and X2 surface guard packers kept uninflated All packers inflated Ail packers inflated All packers inflated All packers inflated All packers inflated Flow rate changes towards end of test Flow rale changes towards end of test All packers deflated except Y2 mjection stnng Monitored Y2M. YIM. W2AM. V3M and X2M at fastest possible rate All packers inflated Slight decrease ui neutron counts CR = constant rate SH-neutron = single-hole test with neutron probe installed in the injection interval Slug = slug injection of air Barometric = barometric test in which all packers were inflated, pressure in the intervals were monitored while there was no air injection at the injection interval Plate 6.1: Photogtiqrii of a single packer monitorii^ system. 162 Plate 6.2: Pfaotogcqrfi of tfie modular packer ss^em. 163 Upper Guard Packer Upper Monitoring Interval (with pressure transducer) A Middle Guard Packer^ Injection / Monitoring Interval (with pressure transducer and CPN neutron probe for W3) Lower Guard Packer t Lower Monitoring Interval Down-hole direction (with pressure transducer) Note; [>awtng Is not to scale Figure 6.1: Injection and monitoring systems used for cross-hole tests. 164 Table 6.2; Information on injection and monitoring intervals with pressure transducer types during phase 3 cross-hole tests Pressure transducer installed in injection and monitoring intervals Interval Interval type FT manufactioer Model PT Range PT Resolution VI V2U V2M V2B V3U V3M V3B W1 W2 W2AU W2AM W2AL W2AB W3U W3M W3B XI X2U X2M X2B X3 YIU YIM YIB Y2U Y2M Y2B Y3U Y3M Y3B Z1 Z2U Z2M Z2L Z2B Line Line Point Line Line Point Line Line Line Line Point Point Line Line Point Line Line Line Point Line Line Point Point Line Line Point Line Line Point Line Line Line Point Point Line Line Point Line Geokon Honeywell Geokon Honeywell Druck Geokon Druck Geokon Honeywell Honeywell Geokon Geokon Honeywell Honeywell Geokon Honeywell Geokon Geokon Geokon Geokon Geokon Druck Geokon Geokon Honeywell Geokon Honeywell Geokon Geokon Geokon Geokon Geokon Geokon Geokon Geokon Geokon Geokon Geokon 4500AL-25 Microswitch 4500AL-25 Microswitch PDCR 4500AL-25 PDCR 4500AL-25 Microswitch Microswitch 4500AL-25 4500AL-25 Microswitch Microswitch 4500AL-25 Microswitch 4500AL-25 4500AL-25 4500AL-25 4500AL-25 4500AL-25 PDCR 4500H-0010 4500AL-25 Microswitch 4500AL-25 Microswitch 4500AL-25 4500H-(K)10 4500H-0010 4500AL-25 4500AL-25 4500H-0010 4500H-OOIO 4500AL-25 4500AL-25 4500H-0010 4500H-0010 0-25 0-15 0-25 0-15 0-15 0-25 0-15 0-25 0-15 0-15 0-25 0-25 0-15 0-15 0-25 0-15 0-25 0-25 0-25 0-25 0-25 0-15 O-IO 0-25 0-15 0-25 0-15 0-25 0-10 0-10 0-25 0-25 0-10 O-IO 0-25 0-25 0-10 0-10 HijOi Medium HiRh Medium Medium High Medium HiRh Medium Medium High High Medium Medium High Medium High High High High High Medium Very High High Medium High Medium High Very High Very High High High Very High Very High High High Very High Very High Z3U Z3M Z3B Pressure transducers installed within injection system in field laboratory Injection system Packer line Barometer Geokon 4500AL-25 0-25 High Honeywell Geokon MS 4580-1-2.5 0-60 0-2.5 Medium Very High Tracer sampling ports (injection port) No No Yes No No No No No No No No No No No No No No No Yes No No No Yes No No Yes No No Yes No No No No Yes No No Yes No 165 In total, thirty-eight packers were used in each cross-hole test. The packers were approximately \-m long in all PP tests performed during phase 3 of the project except in monitoring system V3 where shorter packers, 50-cm in length, had been employed. An air manifold was constructed to distribute air pressure evenly and efficiently through the system (see Plate 6.3 - pressure gauge was added later). A pressure gauge was placed in the manifold to allow visual check of packer inflation. The system allowed inflating packers individually and, in this way, monitoring and recording their pressure as well as varying the lengths and configurations of monitoring intervals without necessarily moving the down-hole equipment. Packer inflation pressure was maintained at 60 Psi throughout each cross-hole test. Temperature readings were taken by downhole and surface pressure transducers that are temperature compensated as noted in Table 6.2. The equipment included twenty-seven Geokon 4500 pressure transducers which compensate pressure for fluctuations in temperature, one Geokon 4580 barometric transducer which provides temperature compensation, three Druck PDCR™ pressure transducers without such compensation, and ten Honeywell-MICROSWITCH™ pressure transducers without temperature compensation. The type of pressure transducers, their locations, range and relative resolutions are listed in Table 6.2. To monitor relative changes in volumetric water content around the injection interval, we installed a BOART LONGYEAR CPN* model 503 neutron probe in the injection interval toward the end of our project. As the probe was sensitive to PVC, the latter was replaced by galvanized steel (see Plate 6.4). 166 The air injection system installed in the field laboratory is illustrated schematically in Figure 6.2 (see Plate 6.5 also). It included a mass flow meter, several pressure valves and regulators, an oil and a water filter, a 1-ijm particle filter, a pressure transducer, a relative humidity sensor, drierite (anhydrous calcium sulfate) and electronic equipment to automatically collect field data. Pressure and temperature in the injection air stream were measured using a Geokon 4500 pressure transducer compensated for temperature. Relative humidity (RH) in the air stream was measured using a VAISALA™ SOY probe. Various Sierra Instruments Sidetrack™ 830/840 mass flow controllers and meters operated over ranges of 0-0.1 SLPM (standard liters per minute), 0-20 SLPM, and 0-100 SLPM, supplying a constant injection (volumetric) flow rate that could be adjusted to between O.I to 100 SLPM. For simplicity of test interpretation, we tried to maintain a constant injection (volumetric) flow rate during each test. As the mass flow controller was sensitive to variations in moisture content in the air stream, we used dry air for injection to help maintain the rate constant. This, we believe, had little if any effect on pressure variations in the monitoring intervals. Not shown in Figure 5.2 are a 02 Psi barometric transducer, two \0-kW generators, four compressors, and solar panels that made up the rest of the system. Data were recorded at 1- to 10-5ec intervals throughout the duration of each crosshole test using three Campbell Scientific™ CRIO dataloggers connected to a Campbell Scientific™ SC32a optically isolated interface, which allowed periodic downloading of data onto an on-site personal computer with a removable, large capacity disk drive. Three Campbell Scientific AM416 multiplexers were concatenated to allow 167 simuftaiieous recofdmg of downhole pressure and temperature data, packer pressure, battery voltage, mass flow rate, air temperature, air relative humidity and barometric pressure throu^mut the duration of each test. The electronic system was tested thorou^ily for defects. The computer program, which controls each datalogger was optimized for maximum efSciency. Plate 63: Photogrq)h of tiie air manifold installed at the site (does not show pressure gauge). Plate 6.4: Photogn^of the mjectioii interval with neutron probe installed. 169 1. pressure gauge 3. pressure gauge 1 r 4. oil & water filter 2. pressure regulator 5. Drierite: CaS04 _ 11H2O from air source 10. valve 7. mass flowmeter 8. pressure transducer flow to the injection interval n Va R o 6. 7 (un particle filter 9. relative humidity sensor Figure 6.2; Air injection system installed in the field laboratory. 170 Musflow contfdkr Plate 6.5: Photogr^di of the air injection system and electronic equ^ment installed in the &ld hbocatory. 171 6.2 CROSS-HOLE TESTING PROCEDURE WITH EMPHASIS ON TEST PP4 A typical cross-hole test consisted of packer inflation, a period of pressure recovery (due to packer inflation), air injection and another period of pressure recovery. Our system allowed rapid release of packer inflation pressure when the corresponding recovery was slow, but this feature was never activated even though recovery had sometimes taken several hours. Once packer inflation pressure had dissipated in all (monitoring and injection) intervals, air injection at a constant mass flow rate began. It generally continued for several days, until pressure in most monitoring intervals appeared to have stabilized. In some tests, injection pressure was allowed to dissipate until ambient conditions have been recovered. In other tests, air injection continued at incremental flow rates, each lasting until the corresponding pressure had stabilized, before the system was allowed to recover. In this dissertation we focus on the analysis of test PP4 conducted during the third phase of our program. This test was selected for type-curve analysis because it involved 1) injection into a high-permeability zone in borehole Y2 (see Figures 1.8, 1.9, and 1.16), which helped pressure to propagate rapidly across much of the site; 2) injection at a relatively high flow rate which led to unambiguous pressure responses in a relatively large number of monitoring intervals; 3) the largest number of pressure and temperature monitoring intervals among all tests; 4) a complete record of relative humidity, battery voltage, atmospheric pressure, packer pressure, and injection pressure; 5) the least number of equipment failures among all tests; 6) flow conditions (such as injection rate, fluctuations in barometric pressure, battery voltage, and relative humidity) that were 172 better controlled, and more stable, than in all other tests; 7) minimum boundary effects due to injection into the central part of the tested rock mass; 8) a relatively long injection period; 9) rapid recovery; and 10) a test configuration that allowed direct comparison of test results with those obtained from two line-injection/line-monitoring tests (LLI and LL2), and a point-injection/line-monitoring test (PL3), at the same location (see Table 6.1 for a list of cross-hole tests). Stable flow rate and barometric pressure made type-curve analysis of test PP4 results relatively straightforward. Test PP4 was conducted by injecting air at a rate of 50 SLPM into a 1-m interval located 15 - Mm below the LL marker (Table 4.1) in borehole Y2 as indicated by a large solid circle in Figure 6.3. The figure also shows a system of Cartesian coordinates x, y, z with origin at the center of the injection interval which we use to identify the placement of monitoring intervals relative to this center. Responses were monitored in 13 relatively short intervals (0.5 - 2 m) whose centers are indicated in the figure by small white circles, and 24 relatively long intervals (4 - 42.6 iri) whose centers are indicated by small solid circles, located in 16 boreholes. Several of the short monitoring intervals (V2M, V3M, W2AM, W2AL, W3M, X2M, Z2M, Z2L and Z3M) were designed to intersect a high permeability region (Figures 1.8, 1.9, and 1.16) that extends across much of the site at a depth comparable to that of the injection interval. Table 6.3 lists coordinates of the centers of all monitoring intervals, their lengths 5, radial distances R from the center of the injection interval, geometric parameters /?/ and P2 defined in (5-33) and (5-34), and maximum pressure during the test. 173 Data recorded at the surface during the test included barometric pressure (Figure 6.4), mass flow rate (Figure 6.5), packer pressure (Figure 6.6), battery voltage (Figure 6.7) and relative humidity in the injection air stream (Figure 6.8). Fluctuations in barometric pressure are seen to have been quite regular with an amplitude of about 0.25 kPa during the first 250,000 sec of the test. It later dropped by about I kPa and stayed nearly constant until the end of the test. Mass flow rate remained constant except for a slight drop of I to 2 SLPM zi about 175,000 sec. Packer pressure remained constant at 60 Psi throughout the first part of the test but dropped approximately 275,000 sec into the test, apparently in response to a concurrent drop in barometric pressure. Battery voltage, supplied by solar panels, increased during the day and decreased at night to form a square wave. Relative humidity in the injection stream varied diumally. Arithmetic plots of pressure responses in all monitoring intervals as well as in the injection interval are presented in Figures 6.9 - 6.24. Pressure responses in the 2.0-m injection interval Y2M, 7.1-m monitoring interval Y2U above it, and \2.9-m monitoring interval Y2B below it are depicted in Figure 6.9. The figure also shows back-pressure at the surface, recorded behind the mass flow controller (Figure 6.2), which is seen to be about twice as high as pressure in the injection interval. Fluctuations in back pressure could be due to variations in relative humidity and air temperature within the injection line at the surface. This is one primary reason why we placed our pressure transducers within the injection and monitoring intervals and not at the soil surface. As injection mass flow rate is constant, these fluctuations have no adverse effect on pressure in the injection interval. 174 Figure 6.3; Locations of centers of injection and monitoring intervals. Large solid circle represents injection interval, small solid circles represent short monitoring intervals, and open circles represent long monitoring intervals. Pressure in the injection interval (Figure 6.9) is seen to reach a stable value almost immediately after the start of injection and then to decline slowly with time. A similar pressure behavior was observed in many single-hole tests by Guzman et al. (1994, 1996) and attributed by them to a two-phase flow effect, as discussed in section 4.2. No such effect is seen in any of the monitoring intervals, not even in those situated inrunediately above and below the injection interval in borehole Y2. On the other hand, the Y2M 175 injection interval shows an inflection (175,000-200,000 sec) that we consider characteristic of dual continuum behavior. Barometric pressure fluctuations are included when they are deemed to have had a potential impact on downhole pressure. There was no measurable pressure response in monitoring interval YIB, and pressure transducers in monitoring intervals V2U, V2B, W3B and YIU appeared to have malfunctioned during cross-hole test PP4. In general, pressure responses tended to be largest in monitoring intervals with lengths ranging fi-om 0.5-m to 2-m and to diminish with distance fi-om the injection interval. Pressure in monitoring interval VI (Figure 6.10) shows an inflection at about 150,000 sec which may be an indication of dual continuum behavior; a similar inflection occurs in monitoring intervals V2M (Figure 6.11), V3M (Figure 6.12), W1 (Figure 6.13), W2AM (Figure 6.15), W3M (Figure 6.16), XI (Figure 6.17), X2U (Figure 6.18), X2M (Figure 6.18), X2B (Figure 6.18), Y2M (Figure 6.9), Y3B (Figure 6.21), and Z3M (Figure 6.24). At late time, the pressure in VI declines in apparent response to a concurrent decline in barometric pressure; this too is seen in several intervals including V2M (Figure 6.11), W1 (Figure 6.13), W2AM (Figure 6.15), W2AL (Figure 6.15), W3M (Figure 6.16), XI (Figure 6.17), X2U (Figure 6.18), X2M (Figure 6.18), X2B (Figure 6.18), X3 (Figure 6.19), YIM (Figure 6.20), YIB (Figure 6.20), Y3U (Figure 6.21), Y3M (Figure 6.21), Y3B (Figure 6.21), Z1 (Figure 6.22), Z2U (Figure 6.23), Z2M (Figure 6.23), Z2L (Figure 6.23), Z2B (Figure 6.23), Z3U (Figure 6.24), Z3M (Figure 6.24) and Z3B (Figure 6.24). Otherwise, barometric pressure fluctuations seem to have only a small effect on pressure in VI as well as in V2M (Figure 6.11), V3U (Figure 6.12), V3M 176 (Figure 6.12), V3B (Figure 6.12), W1 (Figure 6.13), W2AM (Figure 6.15), W2AL (Figure 6.15), X2U (Figure 6.18), X2M (Figure 6.18), X2B (Figure 6.18), Y2U (Figure 6.9), Y2M (Figure 6.9), Y2B (Figure 6.9), Z3U (Figure 6.24) and Z3M (Figure 6.24). On the other hand, such fluctuations seem to have a measurable impact on pressure in W3M (Figure 6.15), YIM (Figure 6.19), Y3B (Figure 6.21), Z3U (Figure 6.24) and Z3M (Figure 6.24), and to dominate pressure variations in X3 (Figure 6.19), Y3U (Figure 6.21), Y3M (Figure 6.21), Z1 (Figure 6.22), Z2U (Figure 6.23), Z2M (Figure 6.23), Z2L (Figure 6.23), Z2B (Figure 6.23) and Z3B (Figure 6.24). These intervals must have excellent pneumatic communication with the atmosphere through high-permeability fractures. We do not have a complete record of pressure recovery for cross-hole test PP4. We did observe, however, that the rate of recovery was much faster in intervals V2M, V3M, Wl, W2AM, W2AL, W3M, XI, Y2M, Y3M, Y3B and Z3M than in VI, V2B, V3U, V3B, W2AU, W2AB, W3U, YIU, Y2U, Y2B, Y3U, Z2U, Z2M and Z3U; was nearly identical in intervals X2U, X2M and X2B; and was imperceptible in interval YIU. 177 Table 6.3: Coordinates of centers of monitoring intervals relative to origin at center of injection interval, interval lengths, radial distances between centers of injection and monitoring intervals, geometric parameters Pi and P2 and maximum recorded pressure Interval VI V2U V2M V2B V3U V3M V3B W1 W2AU W2AM W2AL W2AB W3U W3M W3B XI X2U X2M X2B X3 YIU YIM YIB Y2U Y2M Y2B Y3U Y3M Y3B Z1 Z2U Z2M Z2L Z2B Z3U 23M Z3B r [OT| -7.2 -4.2 -4.2 -4.2 -1.2 -1.2 -1.2 -3.6 -3.2 -3.2 -3.2 -3.2 -3.4 -3.4 -3.4 -6.0 2.6 -1.8 -7.9 4.2 -0.7 -2.8 -7.3 4.0 0.0 -6.0 14.8 10.7 -0.5 18.9 10.5 14.8 16.9 21.1 -0.3 3.5 14.4 y[m\ 1.6 1.6 1.6 1.6 1.6 1.6 1.6 0.3 4.8 1.2 -0.8 -3.7 16.5 12.9 1.8 4.8 4.8 4.8 4.8 4.8 -0.1 -0.1 -0.1 0.0 0.0 0.0 0.1 0.1 0.1 -5.2 -5.2 -5.2 -5.2 -5.2 -5.2 -5.2 -5.2 PT = Pressure Transducer z{m\ -6.8 8.4 3.9 -7.2 6.1 -0.3 -9.5 4.9 0.4 -3.2 -5.2 -8.0 1.5 -2.1 -13.3 4.0 2.4 -1.9 -8.0 -5.9 9.4 7.3 2.8 4.0 0.0 -6.0 4.8 0.7 -10.5 3.8 2.6 -1.7 -3.8 -8.0 3.8 0.0 -10.9 B[m\ 29.5 5.0 2.0 18.2 9.6 0.5 18.1 11.9 6.7 1.5 2.1 4.0 6.8 1.2 28.6 14.4 8.2 2.2 12.9 42.6 2.1 1.8 8.8 7.1 2.0 12.9 7.5 2.0 27.8 13.0 8.1 2.0 2.0 7.7 6.8 2.0 26.0 10.0 9.6 6.0 8.5 6.4 2.0 9.7 6.1 5.7 4.7 6.1 9.4 16.9 13.5 13.8 8.7 6.0 5.5 12.2 8.7 9.4 7.8 7.8 5.6 0.0 8.5 15.6 10.8 10.5 19.9 12.0 15.8 18.1 23.2 6.4 6.3 18.8 P, 07 0.7 3.9 5.9 0.9 1.3 8.2 l.l 1.0 1.7 6.1 5.9 4.7 5.0 22.0 1.0 1.2 1.5 5.0 1.9 0.4 8.9 8.5 1.8 1.6 0.7 0.9 0.7 0.8 0.9 0.2 1.0 0.6 0.6 0.3 0.7 0.9 0.8 0.6 0.6 0.2 0.6 0.5 0.9 O.l 0.7 0.4 0.4 0.0 1.3 4.2 10.6 0.8 3.1 3.0 15.8 18.0 6.0 1.9 6.3 1.4 0.0 0.9 0.8 0.7 0.5 0.5 0.7 0.8 0.9 0.4 0.4 1.0 max Ap \kPa\ 8.3 Broken PT 9.8 Broken PT 6.3 49.2 23.5 9.2 2.9 13.8 7.5 6.9 4.5 4.8 Broken PT 7.0 9.2 17.9 7.3 0.2 broken PT 7.2 0.0 16.9 116.6 21.5 0.6 0.7 3.0 0.2 0.7 0.3 O.l O.l 6.5 6.9 0.2 i 100000 u 3 I 200000 150000 230000 300000 330000 -05 time [sec] F^nie 6.4: Barametxic piessore dnxii^cross-hole test PP4. 60 so 45 40 30 20 10 0 50000 100000 150000 200000 dme [MC] Figure 6.5: Flow rate during cross-hole test PP4. 230000 300000 350000 400000 179 70 60 SO I : 40 - 30 20 10 - 0 — 0 SOOOO 100000 ISOOOO 200000 250000 300000 3SOOOO 400000 230000 300000 330000 400000 tfane {ser I Figuce 6.6: Padcer pressure dnring ctoss^iQle test PP4. 0 30000 ISOOOO 200000 tteetMc] Figure 6.7: Battery vottage during oossp^Kiie test PP4. \ A ji V 50000 100000 IJOOOO ^ ! <f \ 200000 / ' 230000 % ^ 300000 330000 400000 >(s*c| Figure 6.8; Fluctuations in relative humidity dnting cross-hole test PP4. 24&0 / 19&0 14&0 98L0 Y2B 4&0 Y2U -ZO 0 50000 100000 150000 200000 250000 300000 350000 400000 / |Me| Figure 6.9: Surface injection pressure and pressure in injection (Y2M) and monitoring (Y2U, Y2B) intervals within borehole Y2 during cross-hcde test FP4. 181 10.0 8.0 6.0 VI £ 4.0 2Xt 0.0 SOOOO 130000 30000Q ^30000 400300 330000 400000 Figure 6.10: Pressure in monitonQg interval VI during cross-hole test PP4. V2M V2B OJ in bnometnc presswe 0 SOOOO 100000 ISOOOO 200000 2S0000 300000 ffmr] Figure 6.11:Pressure inmooitonag intervals V2U, V2M, and V2B duni^ cross^e test PP4. 182 58^ 483 383 V3M V3B 283 A 183 V3U \ • • -13 0 30000 100000 150000 200000 250000 300000 350000 400000 '6wcl Figure 6.12; Pressure in mocitorii^ intervals V3U, V3M, and V3B during cross4ude test PP4. 10.0 / CMW 3OOOO0 Figure 6.13; Pressure in moiiitorii^interval W1 durii^ cross-hole test PP4. ^50000 4mMW W2 OS ao aaqj^V i 300000 *^QOOOO 4OC|DOO 320000 400000 iabioomriciiiiMiiii -1.0 Figoie 6.14: Pressure in monitanng interval W2 <biring cross-hole test H>4. 16 14 12 W2AB 10 W2AL W3AM 8 6 W2AU 4 2 0 -2 -4 0 soooo 100000 150000 200000 230000 300000 Figure 6. IS; Pressure in numitning intervals W2AU, V^AM, W2AL, and W2AB during cross-hole testl?4. 184 9 7 I W3M ' 3 W3U W3B 1 -2 0 50000 100000 130000 200000 250000 300000 350000 400000 Figure 6.16; Pressure in monitoring intervals W3U, W3M, and W3B during cross^ide test PP4. XI £ % -1 soooo ISOOOO 300000 Figure 6.17: Pressure in monitoring interval XI during cros&4ide test PP4. 350000 40«D00 185 23J l&S 13.5 -IJ SOOOO 100000 150000 200000 250000 300000 350000 400000 Figure 6.18; Piessuie in monitonng intervals X2U, X2M, and X2B during cross4icde test PP4. 300000 •03 -1.0 -IJ t^] Figure 6.19; Pressure in mooitonng interval X3 dunog cros&lurie test PP4. 350000 400DOO 186 30000 100000 130000 200000 230000 300000 330000 400000 <Ehc| Figure 6.20 Pressure in nuwiiftwing intervals YIU, YIM, and YIB during cross-hole test FP4. 30000 100000 130000 200000 230000 300000 330000 400000 Figure 6.21 Pressure in monitonng intervals Y3U, Y3M, and Y3B daring cross-hole test PP4. 03 Zl 330000 I -IJ Figme 6.22; Pressure in manitariiig interval Zl during cioss-hole test PP4. 1.0 Z2U 0.0 i -1.0 Z2B 0 50000 100000 130000 200000 vc 230000 300000 330000 400000 /^I Figure 6.23; Pressure in monitoring intervals Z2U, Z2M, 22L, and Z2B during cross-iKrie test PP4. .2 0 30000 100000 130000 200000 230000 300000 330000 400000 tpecl Figure 6.24;Pressure in monitoring intervals Z3U, Z3M, andZ3B durii^ cross-bcde test PP4. 189 6.3 TYPE-CURVE INTERPRETATION OF CROSS-HOLE TEST PP4 Type-curve interpretation of pressure data from cross-hole test PP4 included intervals VI, V2M, V3U, V3M, V3B, WI, W2AU, W2AM, W2AL, W2AB, W3U, W3M, XI, X2U, X2M, X2B, YIU, YIM, Y2U, Y2B, Y3U, Y3M, Y3B, Zl, Z2U, Z2M, Z2L, Z2B, Z3U, Z3M and Z3B. Pressure data from intervals V2U, V2B, W2, W3B and YIB were not amenable for type-curve interpretation and have therefore been excluded. A special set of type-curves was developed for each pressure monitoring interval based on the point-injection/line-monitoring solution, modified for storage and skin in the monitoring interval, given by equations (5-46) and (5-47) in Chapter 5. Evaluation of the corresponding integrals was done by Rhomberg integration. The geometric parameters and yS, were calculated according to P^=1RIB (6.6-1) P = cos^ (6.6-2) where R is radial distance between the centroids of the injection and monitoring intervals, B is the length of the monitoring interval, and 6 is the angle between the corresponding radius vectors. Equations (6.6-1) and (6.6-2) are obtained from (5-33) and (5-34), respectively, upon treating the medium as if it was pneumatically isotropic. Indeed, our type-curve analysis additionally treats the rock as if it was pneumatically uniform. However, since the analysis of pressure data from different monitoring intervals yield different values of pneumatic parameters, our analysis ultimately yields information about the spatial and directional dependence of these parameters. 190 Figures 6.25 - 6.55 show how we matched each record of pressure buildup from cross-hole injection test PP4 to corresponding type-curves on logarithmic paper. Though the figures include type-curves of pressure derivatives, many of the pressure derivative data are noisy compared to the single-hole counterparts (in Chapter 4) and some that are not excessively so. Pressure derivatives have been computed using the algorithm by Bourdet et al. (1989), described in Appendix H. Those derivatives were calculated using variable intervals of pressure data (10, 20, 50, 100, and 150 data points) depending on the quality of data. Pressure derivatives were calculated using fewer pressure data points for good quality data from monitoring intervals with the strongest pressure responses. As the pressure response became noisier, a larger number of data points were employed for the calculation. Pressure derivatives allowed us to guide our type-curve match and to obtain better fits. The match between pressure data and its derivatives with corresponding typecurves in Figures 6.25, 6.27, 6.28, 6.32, 6.36, 6.38, 6.40, 6.42, 6.43, 6.44, 6.46, 6.47, 6.49 and 6.53 are excellent over the entire length of the buildup record; those in Figures 6.30, 6.31, 6.37, 6.43, and 6.44 are good except for intermediate time where the data exhibit an inflection, which suggests dual continuum behavior; in Figures 6.26, 6.33, 6.45, 6.50, and 6.54 intermediate time and late pressure data match the type-curves well, but early data lie above the type-curves; in Figures 6.39, intermediate time and late pressure data match the type-curves well but early data lie below the type-curves; in Figures 6.36, 6.40, and 6.47, early and intermediate time data fit the type-curves well, but late data fall below the type-curves; and in Figures 6.29 and 6.41 the matches are poor. 191 Fluctuations in barometric pressure (Figure 6.4) affect almost all of the late pressure buildup data. It is accentuated in the pressure derivatives for late data. The timing and magnitude of this barometric effect varies between intervals; it tends to be most pronounced in monitoring intervals that show a weak response to air injection into Y2M, in intervals close to the soil surface, and within the Z holes (close to the cliff face). The rapid and haphazard fluctuation in pressure derivatives also may be due to noise in pressure data. Our finding that most pressure buildup data match the type-curves well is a clear indication that the majority of cross-hole test PP4 results are amenable to interpretation by means of a continuum model, which treats the rock as being pneumatically uniform and isotropic while describing airflow by means of linearized, pressure-based equations. The fact that some of our data do not fit this model shows that the latter does not provide a complete description of pneumatic pressure behavior at the site. That the site is not pneumatically uniform or isotropic on the scale of cross-hole test PP4 is made evident by pneumatic parameters derived fi-om our type-curve matches. Table 6.4 lists values of the dimensionless well response time defined in section 5.6, pneumatic permeability, and air-filled porosity derived from these matches. The latter two parameters represent bulk properties of the rock between the corresponding monitoring interval and the injection interval. The permeabilities additionally represent directional values along lines that connect the centers of these intervals. Corresponding statistics are listed in Table 6.5, which compares them with similar statistics of 1-m scale permeabilities, obtained fi^om steady sate interpretations of single-hole test data. The directional permeabilities range 192 from l.lxlO""^m" to 4.6 X with a mean of-13.5 for logio-based k while a corresponding (anti-log) value is 2.9 x . The corresponding variance and coefHcient of variation (CV) are 53 x 10'' and - 5.4 x 10 ", respectively. Corresponding air-filled porosities range from 1.7 x 10'^ to 2.2 x 10"' with a geometric mean of 3.5 x 10"' while the variance and coefficient of variation are 8.0 x 10"' and - 3.6 x 10"', respectively. Permeabilities derived from cross-hole tests are seen to have a much higher mean, and lower variance, than those from the smaller-scale single-hole tests. 193 Table 6.4: Pneumatic parameters obtained from type-curve analysis of pressure buildup data collected during cross-hole test PP4 k[m'] Interval n 1.1E-I4 4.5E-03 VI 1.0 2.0E-14 5.7E-03 V2M 1.0 3.4E-14 8.2E-03 V3U 1.0 1.2E-14 5.4E-03 V3M 0.0 6.2E-15 5.7E-04 V3B 1.0 2. IE-14 1.4E-03 WI 10.0 1.2E-02 7.6E-14 W2AU 1.0 1.5E-14 2.0E-02 W2AM 1.0 2.4E-14 2.1E-02 W2AL l.O 1.9E-14 5.4E-03 10.0 W2AB W3U 1.6E-14 4.6E-03 1.0 1.5E-14 4.1E-03 W3M 1.0 l.lE-03 1.8E-14 10.0 XI 3.4E-03 2.0E-14 lO.O X2U I.lE-16 2.4E-05 1.0 X2M 9.8E-15 2.6E-03 1.0 X2B 3.2E-14 2.2E-01 YIU 1.0 7.7E-15 2.5E-03 lO.O YIM l.lE-14 7.4E-03 Y2U 1.0 1.5E-03 5.4E-I5 Y2B 1.0 7.6E-04 100.0 1.3E-13 Y3U 4.2E-02 1.4E-13 Y3M 1.0 1.5E-02 2.6E-14 Y3B 1.0 1.8E-02 4.2E-13 10.0 Z1 1.7E-05 1000.0 1.4E-13 Z2U 2.3E-13 2.9E-04 10000.0 Z2M 1.3E-03 4.6E-I3 1000.0 Z2L 3.7E-03 3.6E-I3 100.0 Z2B 2.4E-14 3.7E-02 Z3U 1.0 2.5E-14 2.3E-02 Z3M 1.0 1.3E-02 10.0 4.0E-13 Z3B 194 Table 6.S; Sample statistics of directional air permeabilities and air-filled porosities obtained from type-curve interpretation of cross-hole test PP4, and of air permeabilities from steady state interpretations of \-m scale single-hole tests. Numbers in parentheses Statistic Minimum Maximum Mean (Logio) Mean Variance CV Cross-hole values Logio k [m-] Logio 4> -16.0 (5.4 X 10-'^) -3.1 (7.6 X lO"') -12.3(4.6 X 10*'^) -13.5 2.9 X lO-'" -0.7 (2.2 X 10 ') -2.5 Single-hole values Logio k -17.1 -13.1 -15.3 lO-'" 3.5 X 10*^ 5.6 X 5.3 X 10-' 1.5 X 10-^ 7.6 X 10"' -5.4 X 10-^ -3.6 X 10*' -5.8 X 10"^ 195 i.(eH>i Q = lo QflOO Q^IOOO l.(EH)Q Q I a i.a&(n •S* a «. i.a&a2 i.a&«3 xxem I.(£-OI l.(»00 i.ce+oi I.(£H12 I.(£+<B l.(£-H>4 Figiiie6.25: Type-curve match afpiessure and its derivative fipominonitoniig interval VI. Q =10 1.0&03 1.0EH12 Figure6.26: Type-curve match afpressure and its derivative from monitoring interval V2M. 196 l.(£K)l /r, = IJ4 /»,=OiW 1.0EH)0 a I ft 1.C&01 ijobm i.a&03 1.0&<I3 l-OErOZ I.(£-OI UCCHX) l.OEHB l.OEHM Figure 6.27;Type-curve matdiafpiessuie and its derivativefianmamtoni^ interval V3U. /?;=S.l9 ^, = 0.16 0=10 0=100 0=1000 i.(eHX) h I ft I.(Kr01 '5 l.(£pQ2 1.0E-a3 l.OErOI I.O&OI l.(SH)0 \SS&<a. l.0E*C3 l.OEHM F^;uie 6.28;Type-curve match ofpressure and its derivative fircnnmanitonng interval V3M. 197 /I, = 1.07 ^,=058 Q=100 a 1 '5 I.0EH>4 'o Figure 6.29:Type-curve matcfaofpressnre and its derivative ftewnmonitofiiig interval V3B. Q=i 1 Figure 6.30: Type-curve matdi of pressure and its denvative fiom monitorii^ intenral W1. 198 0=100 O-l l-OBOI a« l.OEHW 'o Figare6Jl;Type-ciirve match ofpressuieazxlits derivative firaninaaitoringinterval W2AU. £>?10 Q=100 1.0&03 1.0EH)1 Figuie 6.32:Type-curve match ofpiessuie and its derivative fiommamtonng interval W2AM. 199 1.0EK>1 ^,=592 p2=0j» Q =10 l-OEHX) Q=IOO 0=^1000 Q I ^ l.(£«l * i.a&a2 i.a&a3 l.(£-Q3 O Oq I.(£<C i.(£«l l.(£H)l I.^HX2 l.(£H)3 1.0EHM Figure 6.33;Type-curve matdi of pressure audits derivative fixim mouitorii^ interval W2AL. 1.0EH)1 ft I =4.75 fij =0m Q = 100 Q = lOOO Q»10 1.(£4<W I ^ 1.0&01 '5 i.a&az l.(£^ i.(£-a2 I.(£ 01 p l.OEHM l.ae-kxz F^uie 6.34:Type-curve matdi of pressure andits derivative 6om monitoring interval W2AB. 200 p , =4J»7 p,=o.n Q=10 Q=IOO QjlOOO 1.CEHI0 I a i.«KOi • 1.0B<X2 l.a&413 i-oeroz i.(ep(» uoepOI l.CSHC t.0EH)3 l.QEHM 'o Figure 6.35;Type-curve matdi of pressure audits derivative from monitOTiiig interval W3U. i.aEHn P, =21.99 P,=0SJ Q=i Q=io Q^lOO Q"1000 I.aEH)0 I •fr '5 i.a&02 •• l.(£03 I.OE-02 'll ' : / // l.(EtOO I.0EH>3 l.OEHM Figure 6.36;Type-curve matdi of pressure and its derivative from monitotiiig interval W3M. 201 I.0EH)1 Bj =0.16 l.OE-01 i.a&c2 I.0E<I3 l.OBrm l.(£-01 I.OEtOO i.cbkj3 1.0Et04 Figure 6.37;Type<uive match ofpressure andits denvative from nusiitoniig interval XI. /?, = 1.47 Q=1 Q=10 Q = 100 Q=1000 I-OEHX) l.(Er01 l-OErCa l.(£-02 l.OE-01 l.oeha l.OEHB l.(£t04 Figure 6.38;Type-curve match of pressure and its derivative frcm manitoring interval X2U. 202 1.0EH)1 /»,=0.48 Q=tO Q-IOO Q = 1000 i.aEH» I ^ I-OEPOI • 1.0642 I.OEr^B 1.0&a2 l.(E-01 l.ffi+OO 1.(^1 l.OEHC 1.0EHB l.(SH)4 Figure 6.39;Type-curve match of pressure andits derivative from monitoiiug interval X2M. i.(e«<)i fij'usa Q°10 Q=100 Q=1000 LOEHX) 1.0Er01 L-OEROZ 1.0E43 I.(£P02 1.(£P01 l.OEHX) I.OE+01 1.^+02 l.(£K>3 l.(£HM Figure 6.40; Type-curve match of pressure and its derivative from monitonng intOTval X2B. 203 1.0EH)1 /?,=8.90 fij =0.65 Q=l l.aEH)0 QclO Q=100 Q=1000 ^ 1.0&01 Q «i ijo&m I.(£^ I.(£pa2 I.aE41 l.(£H)0 l.OEH)! I.(£+<B l.OEHB Figure 6.41;Type-curve matdi cf pressure audits derivative fixjmmanitonnginterval l.OEHM YIU. I.0EH)1 ^2=0.40 Q=10 Q=100 Q=1000 I-C&OZ I.0EM12 l.CeHX) 1.(^*03 l.OEHM Figure 6.42;Type-curve matdi of pressure andits derivative firom monitonng interval YIM. 204 p, =\st pj=om Q=1 Q=100 Q=10 Q=1000 Q l.OErOl ix&m. l.OE-02 I.OE-01 1.0EH)0 l.OEM)I LOEHG l.OEHB l.OEHM Figure 6.43;Type-curve matdi of pressure andits derivative firom monitoring interval Y2U. I.OE+OI P, = 131 pi =0.01 Q=l 1.0EH)0 Qfio ,= 1000 QtIOO «> l.OE-01 t "5 1.0Er02 1.0Ep03 l.OE-02 l.OE-OI I.OE+00 l.OEHG l.OEHB l.OEHM Figure 6.44:Type-curve xnatdiafptessure and its derivative fixminuxiitoiing interval Y2B. 205 /r,=4.i8 Q=l l-OEHW Q=10 Q^lOO Q=IOOO Q I 6 I.(]&01 A l-OErOS 1.0E^ l.(lBr(a l.OErOl 1.0EHX) 1.0EH)1 l.OE+02 l.OEtOS l.OEHM Figure 6.45;Type-curve match of pressure audits derivative firom monitoringinterval Y3U. fi, = 11163 Q=i a=io Q=100 Q=1000 l-CEHX) A I "% i.a&oi • *5 I-OEr^B ijo&m I-OEFOZ i.(e4>i l.(eH)0 uoeka Figure 6.46;Type<urve matdi of pressure audits derivative finon monitoring interval Y3M. 206 p, =0.76 /I, =0.74 = 10 en 'b Figure 6.47;Type-curve match of pressure and its derivative fi:om monitonag interval Y3B. Q=i 1.0&03 1.0EHX) Figure 6.48;Type-curve match ofpressure and its draivativefixnnmcnitoring interval Zl. 207 Q=10 Q=1 // l.OEHB Figure 6.49:Type-curve matdiof pressure andits derivative fixim manitonnginterval Z2U. tf,=0.74 Q=°10 Q=1000 Q=100 Q= 10000 I.CEHX) I.OE-02 1.0EP03 l.(K«K) l.(E+01 LOEHB I.QEHQ l.«®+04 l.OE+05 *0 F^pre 6.50: Type-curve matdiafpressure andits derivative fixwnmoaitoung interval Z2M. 208 l.<£t01 p, =isjn Q = .o Q=»«> 0=^1000 KKHX) 1.0E01 l.oe-oz xsmrol l.OEHX) 1.0EH)I 1.0ehz2 I.OEHB l.OEHVt r i^e 6.51 Type-curve match ofpiessuie and its derivative from monitonng interval Z2L. Q=1 Q=10 piooo I ft 1.0&01 •» // 1.C&01 Figure 6.S2; Type-curve matchof pressure and its derivative from monitoring interval Z2B. 209 Q=1 Q=10 Q=100 Figure 6.53:Type-curve match of pressure and its derivative firom monitom^ interval Z3U. P,=6.29 Q=10 Q=100 1.QE-Q3 Figure 6.54;Type-curve match of pressure and its derivative fiom manttoiing interval Z3M. 210 P, = 1.14 B,=03S Q = 10 l.OE-02 l.OE-01 l.OEHX) Q=100 I.OE+OI Q = 1000 LOEHB I.OEMB I.OEHM Figure 6.55; Type-curve match of pressure and its derivative fiom mamtoni^ interval 23 B. 211 7. FINDINGS AND CONCLUSIONS Our work, together with earlier studies concerning the pneumatic behavior of unsaturated fractured tuffs at the ALRS, most notably those by Guzman et al. (1994, 1996), Guzman and Neuman (1996), and Illman et al. (1998) lead to the following major findings and conclusions; 1. Issues associated with the site characterization of fractured rock terrains, the analysis of fluid flow and contaminant transport in such terrains, and the efficient handling of contaminated sites are typically very difficult to resolve. A major source of this difficulty is the complex nature of the subsurface "plumbing systems" of pores and fractures through which flow and transport in rocks take place. There is at present no well-established field methodology to characterize the fluid flow and contaminant transport properties of unsaturated fractured rocks. 2. In order to characterize the ability of unsaturated fractured rocks to conduct water, and to transport dissolved or suspended contaminants, one would ideally want to observe these phenomena directly by conducting controlled field hydraulic injection and tracer experiments within the rock. In order to characterize the ability of unsaturated fractured rocks to conduct non-aqueous phase liquids such as chlorinated solvents, one would ideally want to observe the movement of such liquids under controlled conditions in the field. In practice, there are severe logistical obstacles to the injection of water into unsaturated geologic media, and logistical as well as regulatory obstacles to the injection of non-aqueous liquids. There also are important technical reasons why the injection of liquids, and dissolved or suspended tracers. 212 into fractured rocks may not be the most feasible approach to site characterization when the rock is partially saturated with water. Many of these limitations can be overcome by conducting field tests with gases rather than with liquids, and with gaseous tracers instead of chemicals dissolved in water. 3. The University of Arizona has conducted successfully numerous single-hole and cross-hole pneumatic injection tests in unsaturated fi'actured tuffs at the Apache Leap Research Site (ALRS) near Superior, Arizona, under the auspices of the U.S. Nuclear Regulatory Commission (NRC). These tests were part of confirmatory research in support of NRC s role as the licensing agency for a potential high-level nuclear waste repository in unsaturated fractured tuffs at Yucca Mountain. However, unsaturated fractured porous rocks similar to tuffs (and alluvium) are found at many locations, including some low-level radioactive waste disposal sites, nuclear decommissioning facilities and sites contaminated with radioactive as well as other hazardous materials. The test methodologies we have developed, and the understanding we have gained concerning the pneumatic behavior and properties of tuffs at the ALRS, are directly relevant to such facilities and sites. 4. We found it possible to interpret both single-hole and cross-hole pneumatic injection tests at the ALRS by means of analytically derived type-curves, which account only for single-phase airflow through the rock while treating water as if it was immobile. Our type-curves are additionally based on linearized versions of the nonlinear partial differential equations that govern single-phase airflow in uniform, isotropic porous continua under three regimes; three-dimensional flow with spherical symmetry, two- 213 dimensional flow with radial symmetry, and flow in a continuum with an embedded high-permeability planar feature (a major fracture). The particular method of linearization appears to have only a minor impact on the results of our type-curve analyses. Included in our type-curves are effects of compressible air storage and skin in the injection interval during single-hole tests, and in monitoring intervals during cross-hole tests. Our analytical tools include type-curves of pressure derivative versus the logarithm of time, which accentuate phenomena that might otherwise be missed, help diagnose the prevailing flow regime, and aid in constraining the calculation of corresponding flow parameters. 5. Steady state type-curve interpretations of single-hole pneumatic tests yield air permeability values for the rock in the immediate vicinity of the test interval. Transient type-curve analyses of such tests provide additional information about the dimensionality of the corresponding flow regime, skin factors and compressible air storage effects. Under radial flow, or in the absence of a significant borehole storage effect, transient type-curve analyses may also yield values of air-filled porosity. At the ALRS, air permeabilities obtained from steady state and transient type-curve interpretations of single-hole pneumatic injection tests, conducted in borehole intervals of 1-m, agree closely with each other but correlate poorly with fracture density data. Airflow around the vast majority of these relatively short test intervals appears to be three-dimensional; borehole storage due to air compressibility is pronounced; and skin effects are minimal. The combined effects of three- 214 dimensional flow and borehole storage make it difficult to obtain reliable air-fiiled porosity values from these tests by means of type-curves. 6. During a pneumatic injection test, air moves primarily through fractures most of which contain relatively little water, and the test therefore yields permeabilities and porosities which reflect closely the intrinsic properties of the surrounding fractures. This is so because capillary forces tend to draw water from fractures into porous (matrix) blocks of rock, leaving the fractures saturated primarily with air, and making it difficult for air to flow through matrix blocks. Since the fractures contain some residual water, the corresponding pneumatic permeabilities and air-filled porosities tend to be somewhat lower than their intrinsic counterparts. The former nevertheless approach the latter as the rate of injection goes up. This is due to displacement of water by air, which under a constant rate of injection manifests itself in a rapid increase in pressure within the injection interval, followed by a gradual decrease. Two-phase flow of water and air additionally causes air permeabilities from single- hole pneumatic injection tests to exhibit a hysteretic variation with applied pressure. 7. In most single-hole pneumatic injection tests at the ALRS, pneumatic permeabilities increase systematically with applied pressure as air displaces water under two-phase flow. In a few single-hole tests, where the injection intervals are intersected by widely open fractures, air permeabilities decrease with applied pressure due to inertial effects. This pressure-dependence of air permeability suggests that it is advisable to conduct single-hole air injection tests at several applied flow rates and/or pressures. Pneumatic parameters derived from pressure data recorded in monitoring intervals 215 during cross-hole tests appear to be much less sensitive to the rate of injection, suggesting that two-phase flow and inertial phenomena decay rapidly with distance from the injection interval. Enhanced permeability due to slip flow (the Klinkenberg effect) appears to be of little relevance to the interpretation of single-hole or crosshole air injection tests at the ALRS. 8. Flow in the vicinity of most \-m single-hole pneumatic test intervals at the ALRS appears to be three-dimensional regardless of the number or orientation of fractures in the surrounding rock. We interpret this to mean that such flow is controlled by a single continuum, representative of a three-dimensional network of interconnected fractures, rather than by discrete planar features. Indeed, most single-hole and crosshole pneumatic test data at the ALRS have proven amenable to analysis by means of a single fracture-dominated continuum representation of the fractured-porous tuff at the site. Only in a small number of single-hole test intervals, known to be intersected by widely open fractures, have the latter dominated flow as evidenced by the development of an early half-slope on logarithmic plots of pressure versus time; unfortunately, the corresponding data do not fiilly conform to available type-curve models of fracture flow. Some pressure records conform to the radial flow model during early and intermediate times, but none do so fiilly at late time. 9. It is generally not possible to distinguish between the permeabilities of individual fractures, and the bulk permeability of the fractured rock in the immediate vicinity of a test interval, by means of pneumatic injection tests. Hence there is little justification for attempting to model flow through individual fractures at the site. The 216 explicit modeling of discrete features appears to be justified only when one can distinguish clearly between layers, faults, fracture zones, or major individual fractures on scales not much smaller than the domain of interest. 10. Air permeabilities obtained from single-hole tests are poorly correlated with fracture densities, as is known to be the case for hydraulic conductivities at many watersaturated fractured rock sites worldwide {Neuman, 1987). This provides further support for Neuman's conclusion that the permeability of fractured rocks cannot be reliably predicted from information about fracture geometry (density, trace lengths, orientations, apertures and their roughness) but must be determined directly by means of hydraulic and/or pneumatic tests. 1 i. Cross-hole pneumatic injection test data from individual monitoring intervals at the ALRS have proven amenable to analysis by type-curves, which treat the rock as a uniform and isotropic fractured porous continuum. Analyses of pressure data yielded information concerning pneumatic connections between injection and monitoring intervals, corresponding directional air permeabilities, and air-filled porosities. All of these quantities were found to vary considerably from one monitoring interval to another in a given cross-hole test on scales ranging from a few to over 20 meters. Thus, even though our type-curve analysis treats the rock as if it was pneumatically uniform and isotropic, it ultimately yields information about the spatial and directional dependence of pneumatic permeability and connectivity across the site. 12. Some single- and cross-hole pressure records reveal an inflection that is characteristic of dual continuum behavior. The prevailing interpretation of dual continua is that one 217 represents the fracture network and the other embedded blocks of rock matrix. We take the broader view that multiple (including dual) continua may represent fractures on a multiplicity of scales, not necessarily fractures and matrix. However, these inflections show some correlation with barometric pressure fluctuations, implying that they might have been caused by the latter rather than by dual or multiple continuum behavior. 13. The pneumatic permeabilities of unsaturated fractured strongly with location, direction and scale. tuffs at the ALRS vary In particular, the mean of pneumatic permeabilities increases, and their variance decreases, with distance between packers in a single-hole injection test, and with distance between injection and monitoring intervals in cross-hole injection tests. This scale effect is most probably due to the presence in the rock of various size fractures that are interconnected on a variety of scales. 218 APPENDIX A: ASSESSMENT OF ELECTRICAL RESISTIVITY TOMOGRAPHY (ERT) IN OBTAINING THE SUBSURFACE DISTRIBUTION OF WATER CONTENT IN THE VADOSE ZONE A.1: INTRODUCTION-APPLICATIONS TO ENVIRONMENTAL MONITORING Electrical resistivity tomography (ERT) has been recently applied to a number of field studies in subsurface hydrology. Interests in employing ERT in conjunction with hydraulic tests and remediation activities have increased due to its potential in providing high-resolution, two- and three-dimensional images of the subsurface that cannot be usually obtained by means of hydraulic and/or tracer tests. Needless to say, hydraulic/tracer tests can provide high-resolution images of the subsurface on the condition that data samples are large enough for interpolation and stochastic/geostatistical analysis. In practice, intensive field data collection is expensive and time consuming; hence, geophysical techniques such as ERT is becoming an attractive alternative. The underlying principle behind ERT is straightforward. Electrical currents are passed through electrodes (transmitters) placed on the soil surface and/or in the subsurface and voltage differences are measured in another set of electrodes (receivers). The known voltage difference and input current is related to the resistivity of subsurface materials. Voltage differences obtained from a large number of cross-hole measurements are subsequently inverted to reproduce the subsurface distribution of electrical resistivity. This variability in electrical resistivity is then related to flow and transport parameters through empirical laws. 219 Spatial variability in resistivity can result from textural variability of subsurface materials (e.g., sand versus clay; large voids such as fractures at the ALRS - see 1.2.5), distribution of capillary water, variability in the ionic content of water, etc. Emphasis is placed on monitoring changes in resistivity over time (it requires a succession of measurements to see changes in subsurface conditions). In the saturated zone, Rcanirez et al. (1993) employed ERT to map the subsurface distribution of a steam flood as a function time. Injected steam induced changes in soil resistivity because the application of steam displaces some of the capillary water, increasing the pore water and soil temperature. According to Ramirez et al. (1993), changes in resistivity are also due to alterations in the ionic content of capillary water. In another field study, Lundegard and LaBrecque (1995) employed ERT to monitor air sparging. This study showed that the radius of influence of air sparging can be delineated from reconstructed images of resistivity. In a more recent study, Stubben et al. (1998) applied ERT to monitor changes in groundwater conductivity in relation to in situ leaching of copper. Applications of ERT in the unsaturated zone include monitoring of infiltration experiments by Daily et al. (1992) and Stubben et al. (1998). Daily et al. (1992) conducted two types of infiltration experiments (point and line infiltration) in which they employed ERT to monitor the movement of water. More recently, Stubben et al. (1998) applied the technique to monitor an infiltration experiment conducted on a large scale in unsaturated alluvium at a site in Southern Arizona. In this appendix, a journal article by Daily et al. (1992) will be reviewed for purposes of evaluating ERT to obtain adequate 220 estimates of vadose zone water movement through monitoring infiltration experiments. Prior to the evaluation, an introduction to the theory and concepts behind ERT is provided, followed by a description of the reconstruction algorthim (numerical inverse procedure). The conduct of the infiltration experiments will not be discussed but a discussion on the relative merits of ERT will be provided. A.2: THEORY AND CONCEPTS ERT is a method of determining the distribution of electrical resistivity within some volume of subsurface materials fi^om discrete measurements of current and voltage. Electrodes are placed in contact with the formation on the soil surface and/or in boreholes. Two adjacent electrodes are driven by a known electrical current in which the resulting voltage differential is measured between all other adjacent pairs of electrodes (in both boreholes). Once the voltage differential is measured between two points in the transmitting and receiving boreholes, two other electrodes are picked sequentially and the process is repeated until electrical current is applied to all pairs of adjacent electrodes. For each pair of measurement, a transfer resistance is computed. Transfer resistance is the ratio of voltage at one pair of electrodes to the induced current {Daily and Owen, 1991). For n electrodes, there are n(n-3)/2 independent transfer resistances. The current imposed on electrodes in ERT follows the path of least resistance unlike X-rays employed primarily in medicine. This implies that current can flow in curved paths rendering the inversion process a non-unique problem. 221 Following Daily and Owen (1991), we define the mathematical problem, which describes the steady distribution of electrical potential in three-dimensional space. The distribution of electrical potential in a conductive material is governed by the Possion equation (A-1) subject to the following Neumann boundary condition (A-2) p \s)r] V where V - is a divergence operator, V is the gradient operator, p is the resistivity of the region s, y/ is the electric potential, tj is the outward normal unit vector, F is the current source, and g is the current density. There are slight differences in the statement of the mathematical problem when the works of various researchers (Yorkey, 1987, Daily and Owen, 1991, Daily et al. 1992, LaBrecque et al. 1996) are compared but the underlying principle is the same. The inversion problem is to estimate the spatial variability of resistivity p, given multiple sets of measurements of n/ and g. A.3: INVERSION OF TRANSFER RESISTANCES - VARIOUS RECONSTRUCTION ALGORITHMS As described briefly in the previous section, data collected through resistivity surveys must be inverted (analytically or numerically) to obtain a spatial distribution of resistivities. Here, we outline different inversion techniques and discuss some key characteristics. Early analyses included tabulation of measured apparent resistivity, which is difficult to interpret for all but the simplest subsurface structures and the trial 222 and error method {Daily and Owen, 1991). In the trial and error method, a conceptual model is heuristically proposed and the resistivity distribution resulting from this model is computed. The calculated values of resistivity are then compared to data and the model is calibrated by adjusting the parameters so that the model fits the data in a trial and error fashion. According to Daily and Owen (1991), a more sophisticated method was proposed by Tripp et al. (1988). They formulated a nonlinear inversion technique to adjust resistivity for a set of blocks, each of fixed size, shape, and position to fit a surface sampling of voltage and current. One key disadvantage to this technique is that the geometry of the model must be known a priori and cannot be determined from data collected in the field. LaBrecque and Ward (1988) studied borehole to borehole inversion data but their work was also limited to a specific geometry. Daily and Owen (1991) employed a finite element Newton-Raphson algorithm proposed by Yorkey (1986) in which there was no need for a priori knowledge of subsurface resistivities. The main advantage of their method was the flexibility of the inversion process, which can be applied to a variety of measurement geometries (cross-borehole, surface, and borehole to surface). Other researchers such as Beasley (1989), LaBrecque (1990), Sasaki (1990), Tang and Yang (1990) applied different inversion techniques (Marquardt's method, smoothest inversion, and other related methods) with varying degrees of success. More recently, LaBrecque et al. (1996) employed an inversion algorithm in which the geological formation is discretized into parameter blocks, each containing one or more elements. A finite element technique solves the forward problem and Occam's 223 inversion is employed to search for the smoothest two-dimensional model for which the Chi-squared statistic equals an a priori value. This data-weighting scheme effectively removes outliers (perhaps not a good thing). The inverse problem solved is an ill-posed problem yielding non-unique solutions. According to LaBrecque et al. (1996), solutions of least-squares inversions are often highly erratic and geologically unreasonable. Occam's inversion, on the other hand, controls the form of the final solution by finding an optimal, yet conservative solution that fits field data. Further details to the procedure are given in LaBrecque et al. (1996) and Morelli and LaBrecque (1996). LaBrecque et al. (1996, 1998) discuss the effects of data and procedural noise on the inverse modeling using Occam's method. Here, they refer to noise as variance or distribution of random and systematic errors in data collected in the field. Random errors, inter alia, arise fi'om fluctuations in applied current, and inherent variability in voltage differential in the subsurface. Systematic or field procedural errors are more severe and include interference between electrical cables within the borehole and misplacement of electrodes. Another systematic error commonly neglected includes the application of a crude two-dimensional inversion model to data collected in a threedimensional field environment. One intriguing aspect of the inversion procedure described is how the process handles outliers in transfer resistances. As stated earlier, Occam's inversion efficiently removes outliers in data noise in an effort to obtain smoother images. The removal of outliers fi'om data may achieve greater efficiency in the numerical inversion procedure. However, there is no guarantee that those outliers represent real data points collected in 224 highly heterogeneous formations. In many environmental applications, those outliers are data that most interest hydrogeologists (esp. in heterogeneous formations), thus the removal of outliers may contribute to over-smoothing of inverted images. Determination of noise in data appears to be the focus in increasing accuracy and of inverted images. According to LaBrecque et al. (1996), noise estimates can be obtained by comparing reciprocal data points (those with the receiver and transmitter interchanged). However, this may not be the optimal method to determine the noise level in the data due to the nonlinear nature of current path. Another (possible) way to test the validity of collected data is to vary current systematically to see if the transfer resistance changes. If the data collection system is flinctioning correctly and assuming a linear relationship of resistance with current and voltage differential, resistance should remain constant with changes in applied current. Increasing the current should also decrease random noise, as clearer signals should be obtained at the receiving end. In any case, the estimation of noise level is crucial in achieving greater accuracy. LaBrecque et al. (1996) state that overestimates of the noise will decrease the resolution of the image, particularly away from the boreholes. The method appears to be highly dependent on the correct estimation of noise. Even modest overestimates of the noise can cause dramatic increases in the resolution of the images. Morelli and LaBrecque (1996) discuss the effect of noise on reconstructed images through synthetic data with both random and systematic errors while applying their robust inversion technique to a field site in California. The inversion algorithms discussed above are deterministic in a sense that statistical information about the correlation among model parameters and observed data 225 are not incorporated. One promising alternative to previous approaches is the stochastic inversion of three-dimensional ERT data proposed by Xianjin and LaBrecque (1998). According to those authors, one significant advantage of this new algorithm is that any prior knowledge about the model parameters can be incorporated into the inversion process in the form of covariance model type and correlation length. Formulation of the inversion process in the context of an iterative geostatistical inversion procedure (KcA et al. 1996) should allow one to conduct joint hydrological-geophysical inversion with larger confidence. A.4: APPLICATION OF ERT TO MONITOR VADOSE ZONE WATER movemem-daily et al. (1992) STUDY Infiltration experiments were conducted near Lawrence Livermore National Laboratory (LBNL) by Daily et al. (1992) to demonstrate the capabilities and limitations of cross-hole ERT in monitoring the movement of vadose zone water. The plume of wetted soil can be monitored using ERT because coarser, well-drained soils (sands and gravels) appear as more resistive zones, whereas finer grained soils (silts and clays) that hold more water are imaged as less resistive zones. The secondary goal of this study was to leam about specifics of unsaturated flow in alluvial materials. Two types of infiltration experiments were conducted at the site. One was a point infiltration experiment in which a constant head of tap water was maintained in a shallow borehole for 2.5 hrs. The point infiltration event was monitored to a depth of about 17.3 m during a 22-hr period. Two boreholes adjacent to the injection borehole were instrumented with electrodes to conduct a cross-borehole ERT survey to monitor the 226 movement of water in the unsaturated zone. Another experiment, conducted at a nearby site, consisted of a 25-m long trench filled with tap water again maintained at constant head for 74 dc^s. This second infiltration experiment was monitored in two image planes; one perpendicular to the trench and the other image plane parallel to the trench. On both image planes, ERT data were collected to about 27-m below the soil surface. The authors refer to this second experiment as a line or one-dimensional infiltration experiment. In both experiments, images of resistivity distribution were obtained prior to and during infiltration. Cross-hole ERT data taken prior to infiltration in both experiments were considered as the initial distribution of resistivity. This background image was then employed to compare/separate the static resistivity distribution from the changes caused by water infiltration. Daily et al. (1992) evaluated consistency in their results by comparing inverted images to well log data, considered repeatability of their experiments, showed agreement of adjacent image planes along their common edges, considered the similarity of field scale images and the reasonableness of the time sequence of images. In addition, the authors claimed that site-wide average hydrogeologic properties were similar and that previous results {Toney, 1990) can be used to evaluate results from cross-hole ERT surveys. The authors stated that induction logs were employed to validate the reconstructed images. Daily et al. (1992) provided an adequate introduction to the concept and theory behind ERT including the mathematical procedure employed to conduct the numerical inversion of field data. The experimental procedure was explained in sufficient detail. 227 Information on site-geology, data acquisition and experimental results were systematically illustrated. The mathematical procedure employed in Daily et al. (1992)'s study is outlined below. A forward solution to eqn. (A-1) is obtained in Fourier transform space in which the potentials are calculated for a discrete number of transform variables at the nodes of a mesh consisting of quadrilateral elements. Potentials calculated in the transformed domain are then inverted to Cartesian domain using a method described by LaBrecque (1989). A numerical inversion procedure similar to that described in LaBrecque et al. (1996) was employed to obtain images of changes in resistivity. Experimental results from the point infiltration experiment show that cross- borehole ERT appears to capture the resistive and less resistive regions quite effectively. It is interesting to see in the background image (Plate la - see Daily et al. 1990), both resistive zones (one near the soil surface at about 3 m, while the other one is at about 13 m in depth) thinning out equidistant from the transmitting and receiving boreholes. The authors do not comment on the thinning out of the resistive zone but it is suspected that the sensitivity of ERT appears to decrease with distance between the transmitting and receiving points. Subsequent images (Plate lb through li) show snapshots of changes in resistivity, which the authors relate heuristically to changes in water content due to the infiltration event. Images of interest are Plates lb through Id, which show the effect of a capillary barrier, temporarily arresting the downward migration of water and Plates le through li which show preferential flow of water to the right side of the image caused by subsurface heterogeneity. The authors acknowledge that cross-hole ERT data was 228 collected on only this one two-dimensional plane. Anomalous "islands" of decreasing resistivity seen in Plates le through li may be due to: I) ERT not capturing the narrow zone of infiltrating water; and/or 2) water may be flowing out of the image plane (channeled flow). This effectively illustrates one limitation of cross-hole ERT as it was conducted by Daily et al. (1992). In heterogeneous materials (such as alluvium and fiactured rocks), this technique will most likely not be able to capture preferential flow in sufficient detail due to the lack of resolution and over-smoothing of inverted images. Cross-hole ERT, however, may be useful as a monitoring tool in obtaining images (to capture sufficient detail) in relatively homogeneous materials. One interesting finding from this study is the much faster wetting front speed inferred from the ERT images as compared to a rate reported by Taney {^1990). The line source infiltration experiment was conducted at another site using 3 image planes. The background data (Plate 4a) shows several regions of resistive and less resistive regions. The authors claim that there is "^good continuity at the image plane edges along BHl. This correlation is one important consistency check for the tomographs, and this is evidence that the reconstructions represent the in situ resistivity". However, there appears to be a discontinuity in resistivity where tomograph I was taken suggesting some problems with the inversion of data to obtain those images. Subsequent images (Plates 4b through 4d) show irregularities and discontinuities in the images rendering them not too reliable. In addition, there is an apparent increase in the lower part of plane 1-4 in Plate 4d in which the authors claim that there is undoubtedly a reconstruction artifact. They explain that '^''artifacts occur -when the algorithm generates 229 laarge contrasts in the image; a change in one part of the image is accompanied by a change of the opposite sense, in an adjacent region". The fact that there are discontinuities and artifacts in the images renders (on my part) further analysis and review of the line infiltration experiment fiitile. The authors also compare resistivity obtained from cross-hole ERT and induction logs showing a lack of consistency in the results. To obtain values of water content, Archie's law modified for the vadose zone was employed to relate resistivity and water content. The authors found a weak negative correlation between imaged resistivity and water content. This is another indication inter alia (such as discontinuities, reconstruction artifacts) that water content values cannot be accurately inferred from cross-hole ERT surveys. A.5: DISCUSSION OF RESULTS: RELATIVE MERITS OF APPLYING ERT TO MONITOR VADOSE ZONE WATER MOVEMENT The paper by Daily et al. (1992) discussed one application of cross-hole ERT to monitor point and line infiltration experiments conducted in the vadose zone. Results from this study showed that the technique was able capture the downward migration of water in a general sense. Major conclusions reached by the author include; 1. Images of resistivity before infiltration were consistent with what is known of the geology at the site. 2. From images of resistivity changes during infiltration, they inferred the size and shape of the plume in the subsurface over a 2A-hr period. 230 3. The development of the plume was consistent with what they knew of lithology from the driller's log and what could be concluded from these about capilary suction and permeability distribution. 4. The time sequence of images was internally consistent, showing a reasonable development of plume with time. 5. Numerical simulations further confirmed the accuracy of tomographs and revealed that tomography and numerical models can be complementary tools for understanding infiltration in heterogeneous formations. 6. The results showed that water moved through the unsaturated sediments more rapidly than previous investigations had estimated. 7. Results from the line infiltration experiment were reasonably consistent; tomographs of background conditions as well as infiltration yielded similar results for both experiments. For the most part, adjacent image planes matched along their common edges. There was, however, some disagreement in the line source experiment once the wetted regions produced fairly complex resistivity distributions and also in the 1-4 plane where the resistivity constrast was very large. 8. Tomographs matched qualitatively with the borehole induction logs. 9. On the basis of the internal consistency and agreement with other data, they believe that ERT has been shown to have a useful capability to delineate an infiltration plume in a hydrologically complex environment. This subsurface imaging technique will not only be a useful tool for studying contaminant transport but may find use in other fields of geophysics. 231 Based on Daily et al. (1992)'s conclusions, cross-hole ERT surveys appears to be useful in delineating the downward migration of the wetted front. However, there are several inconsistencies and questions arising from their results. The approach of validating the utility of ERT in monitoring vadose zone water movement using site-wide average hydrogeologic parameters appears to be acceptable if one were to employ ERT to obtain a general picture of the subsurface and changes in subsurface conditions induced by water infiltration. However, the authors appear to contradict themselves by trying to obtain a realistic image of the vadose zone using site-wide (or homogeneous) hydrogeologic parameters. In addition, when testing a new field technique in a complex geologic setting, it appears that independent validation of calculated parameters by means of neutron probes and tensiometers may be advisable for added confidence in results. The authors of the journal article criticize traditional approaches (one based on Darcy's law and the other based on tracers) to determine flow velocity and associated moisture distribution on the following grounds: 1) an approach based on Darcy's law requires information of hydraulic conductivity, hydraulic gradient, and porosity which is difficult to obtain; and 2) an approach based on tracer tests is simple and accurate but requires correction for dispersion of solutes, requires multiple measurement points, and inadequate sampling can lead to misleading flow fields. Their criticisms on traditional approaches are valid, yet troublesome, because ERT does not provide us with values of water content, capillary pressure and other important flow and transport parameters. Only those traditional approaches criticized by Daily et al. (1992) give us those estimates. 232 A more recent study by Stubben et al, (1998) conducted at the Maricopa Site in Southern Arizona appears to overcome this shortcoming. In this study, a variety of other hydrogeological and geophysical methods were employed in conjuction with cross-hole ERT which effectively evaluated the performance of ERT to monitor vadose zone water movement. Daily et al, (1992) should have instrumented the tomographed region with tensiometers, neutron probe access tubes and perhaps obtain estimates of air permeability and air-filled porosity through air injection tests to independently validate results from cross-hole ERT. The two-dimensional inversion of data can be troublesome because the collected data reflect water flowing in three-dimensional space. An alternative sampling plan and inversion technique covering three-dimensional space is probably needed to adequately characterize vadose zone water movement and to comment about capabilities and limitations of ERT. Regardless of Daily et al. (1992)'s questionable results, there are several advantages in employing ERT to monitor vadose zone water movement. Some of those include; 1) the technique is not an invasive procedure requiring less boreholes; 2) high resolution can be achieved with large number of electrodes; 3) rapid estimates of plume location in real-time can be obtained through automation {LaBrecque et al. 1998), and 4) if background data is available, one can effectively monitor changes in subsurface water movement. There are, however, numerous problems that must be addressed and corrected. Disadvantages in employing ERT to monitor vadose zone water movement include; 1) 233 complexity of instrumentation can have a high failure rate that can cause systematic errors; 2) collection of reliable data in the field requires good electrical contact of electrodes with the formation; 3) many subsurface variables such as porosity, water content, water chemistry, temperature fluctuations, etc., affect resistance; 4) unknown relationship(s) among variables such as spacing of electrodes, data noise and spatial variability of subsurface materials can cause artifacts in reconstructed images; S) interference of cable signals can cause excessive noise in data; 6) inversion of data can be difficult and faces the problem of non-uniqueness; 7) arbitrariness in weighting scheme for Occam's inverse procedure-particular concern is in the removal of outliers because those outliers may be of most interest to hydrogeologists; and 8) sensitivity of ERT decreases with distance which may cause artifacts in the images (needs to be put in the context of geostatistics so that one can quantify the variance in data points allowing the optimal placement of sensors in boreholes); and 9) mapping resistivity does not necessarily give the subsurface distribution of vadose zone water-one has to still rely on traditional approaches to obtain reliable estimates of water content. The largest problem in employing cross-hole ERT to monitor vadose zone water movement is in its data inversion. Inversion of transfer resistances yield stable, and unique solutions in medical applications because of our knowledge of anatomy of humans, geometry and vast experience in conditions that can cause anomalies in resistivity measurements. In environmental applications, the knowledge of subsurface distribution of flow and transport parameters is limited or unknown, hence rendering every inversion ill-posed, which will then yield solutions that are non-unique. 234 Despite (albeit) the complexity of equipment, logistics, and problems in inverting data to obtain reliable images, ERT appears to be a promising environmental monitoring tool once the problems discussed above are addressed and better quantified. According to LaBrecque et al. (1998), near-real time monitoring environmental processes is being undertaken including the monitoring of water movement in the vadose zone. Large data sets will be collected in the field, processed and imaged in less than 24 hours and still produce reliable, high-quality images. This particular paper describes the automation of data collection and inversion of collected data. We must remember that even if one were able to obtain accurate images of the subsurface, ERT does not give us direct measures of flow and transport parameters. Only traditional hydraulic/pneumatic and tracer tests provide us with estimates of flow and transport parameters that will allow us to model fluid and contaminant transport in the subsurface. A.6: REFERENCES PERTAINING TO APPENDIX A Beasley, C.W., Cross-borehole resistivity inversion: Theory and applications to monitoring enhanced oil recovery, Ph.D. Dissertation, Univ. of Utah, 1989. Daily, W. and E. Owen, Cross-borehole resistivity tomography. Geophysics, 56(8), 12281235, 1991. Daily, W., A. Ramirez, D. J. LaBrecque, and J. Nitao, Electrical Resistivity Tomography ofVadose Water Movement, Water Resour. Res., 28(5), 1429-1442, 1992. LaBrecque, D., and Ward, S., Two-dimensional inversion of cross-borehole resistivity data using multiple boundaries. Presented at the 58"' Ann. Intemat. Mtg., Soc. Expl. Geophys., Expanded Abstracts, 194-197, 1988. LaBrecque, D. J., Cross-borehole resistivity modeling and model fitting: Ph.D Thesis, Univ. of Utah, 1989. 235 LaBrecque, D. J., M. Miletto, W. Daily, A. Ramirez, and E. Owen, The effects of noise on Occam's inversion of resistivity tomography data. Geophysics, <5/(2), 538-548, 1996. LaBrecque, D., J. Bennett, G. Heath, S. Schima, and H. Sowers, Electrical Resistivity Tomography For Processs Control In Environmental Remediation, SAGEEP, Meeting of the Environmental and Engineering Geophysical Society - March 22-26, 1998, Palmer House EBlton Hotel, Chicago, Illinois, Conference Proceedings, 613-622, 1998. Morelli G. and D. J. LaBrecque, Advances in ERT Inverse modeling, European Journal of Environmental and Engineering Geophysics, 1, 171-186, 1996. Ramirez, A. and W. Daily, Monitoring an Underground Steam Injection Process Using Electrical Resistance Tomography, Water Resoitr. Res., 29(1), 73-87, 1993. Sasaki, Y., Model studies of resistivity tomography using boreholes: Presented at the Internal. Symp. On borehole geophysics: Petroleum, hydrogeology, mining, and engineering applications, Soc. Expl. Geophys., 1990. Stubben, M.A. and D. J. LaBrecque, 3-D ERT Inversion Used To Monitor An Infiltration Experiment, SAGEEP, Meeting of the Environmental and Engineering Geophysical Society - March 22-26, 1998, Palmer House Hilton Hotel, Chicago, Illinois, Conference Proceedings, 593-601, 1998. Stubben, M. A. and D. J. LaBrecque, 3-D ERT Inversion Used To Monitor An Injection Experiment, SAGEEP, Meeting of the Environmental and Engineering Geophysical Society - March 22-26, 1998, Palmer House Hilton Hotel, Chicago, Illinois, Conference Proceedings, 603-612 1998. Sullivan, E. and D. J. LaBrecque, Optimization of ERT Surveys, SAGEEP, Meeting of the Environmental and Engineering Geophysical Society - March 22-26, 1998, Palmer House Hilton Hotel, Chicago, Illinois, Conference Proceedings, 571-581, 1998. Toney, K.C., A hydrogeological investigation of groundwater recharge near the drainage retention basin, Lawrence Livermore National Laboratory, Livermore, California, thesis. Dept. ofGeol., San Jose State University, San Hose, Calif, 1990. Tong, L., and Yang, C., Incorporation of topography in two-dimensional resistivity inversion: Geophysics, 55, 354-361, 1990. Tripp, A.C., Hohmann, G.W., and Swift, C.M. Jr., Two-dimensional resistivity inversion: Geophysics, A9, 1708-1717, 1984. 236 Xianjin, Y. and DJ. LaBrecque, Stochastic Inversion of 3D ERT Data, SAGEEP, Meeting of the Environmental and Engineering Geophysical Society - March 22-26, 1998, Palmer House Hilton Hotel, Chicago, Dlinois, Conference Proceedings, 221-228, 1998. Yeh, T-C. J., M. Jin, and S. Hanna, An iterative stochastic inverse method: conditional effective transmissivity and hydraulic head fields. Water Resour., Res., 32, I, 85-92, 1996. Yorkey, T. J., J. G Webster, and W. J. Tompkins, Comparing Reconstruction Algorithms for Electrical Impedance Tomography, IEEE Transactions on Biomedical Engineering, BME-34, II, 843-852, 1987. 237 APPENDIX B: NEUTRON PROBE METHODOLOGY B.1: INTRODUCTION: One of the recommended indirect approaches to determine water content, both in the laboratory and field is the neutron attenuation method. The neutron attenuation method is a standardized method, recommended by agencies such as the American Standard of Testing Materials (ASTM D30I7-88) and the U.S. Dept. of Interior (1977). Previous studies and a comprehensive review of the methodology can be found in Visvalingam and Tandy (1972) and {Gardner, in Klute 1986). Employment of neutron probe to measure water content yields a value of volumetric water content {9J) instead of the gravimetric water content obtained by weighing the sample. Field determination of water content is necessary because 1) the collection of sample for laboratory analysis is a destructive process, 2) handling numerous samples for laboratory analysis may be cumbersome and inefficient and 3) in sitit measurements of water content may be more meaningful. A neutron probe may be operated with relative ease with a number of point measurements taken over a large volume of the geologic medium. It gives the analyst the capability to obtain a large data set of water content in a nondestructive manner. One of the traditional applications of a neutron probe is to lower inside an access tube at successive intervals to quantify the vertical distribution of moisture down the hole. Measurements taken over long periods (months) can be used to delineate the spatial and temporal variations of water content at a field site. For example, a neutron probe was 238 employed at the Apache Leap Research Site (ALRS) to measure water content in fractured tuffs (Rasmussen et al. 1990; see also 1.2.5 of this dissertation). Collected data was helpful in simulating the proposed infiltration experiment using a computer model {Chen et al. 1995 in Bassett et al. 1995). In another example, a neutron probe has been used extensively to determine the spatial and temporal variations of soil moisture under an impermeable cap at a field site in New Mexico for purposes of waste isolation {McTigue et al. 1994). Measurements of water content and temperature in seven boreholes over three years have led the workers to conclude that the spatial variability in water content decreases as depth increases and moves downward by diffusive vapor distillation. In each example, a calibration curve was fiimished to relate neutron counts to water content. Calibration of neutron probes for use in alluvial materials is well documented, practiced, and has become a standardized procedure. Calibration of neutron probes in consolidated rocks, however, has not been attempted previously, since the device is used primarily by soil physicists and agronomists, whom are usually not interested in vadose zone investigations in consolidated rocks. We begin with a discussion of the theory of neutron attenuation and a brief description of the standard calibration procedure using alluvial materials. Factors that affect neutron counts are then discussed in attempt to understand the capabilities and limitations of the neutron probe. B.2 TEDEORY BEHIND THE NEUTRON ATTENUATION METHOD The method for measuring volumetric water content in partially saturated soils and rocks is commonly referred to as the neutron attenuation method. The neutron 239 attenuation method was first proposed by Belcher et al. (1950). Gardner et al. (1952), however, first stated the theory and described the physical and operational aspects of the methodology. The device used for the investigation is a cylindrical probe, which consists of a small radiation source and a detector. The radiation source, which is usually a Ra-Be or Am-Be, both emit high energy neutron in the range of few million electron volts (MeV)Emitted neutrons collide with the nuclei of atoms that surround the neutron source. The collision of fast or "high-energy" neutrons with heavier nuclei such as C'*', and Mg^^ is virtually inelastic and the reduction of momentum of traveling neutrons is minimal. The neutrons become "thermalized" or slowed down, when the neutrons collide with smaller atoms such as hydrogen (H"). The number of thermalized neutrons Is proportional to the number of hydrogen atoms found in the surrounding material. A detecting device, which consists of boron trifluoride (BF3), registers neutrons moving at thermal velocities. Hydrogen atoms in soils and rocks are mostly in the form of water and neutron thermalization is dominated by neutron-hydrogen collisions, hence a relation between neutron counts and volumetric water content can be established for the surrounding material whether the water is in solid, liquid or gaseous phases. We must note, however, that some hydrogen is in the form of organic material, plastics (casing material and current injection interval for cross-hole tests) and hygroscopic water that is attached to the mineral grains. Consequently, measurements of thermalized neutrons may not be a true reflection of the amount of moisture surrounding the detector. Neutron counts are amenable to statistical analysis due to the inherent variance associated with 240 neutron-hydrogen collisions and the number of thermalized neutrons that are counted by the detector. For a homogenous soil with a uniform moisture distribution, the cloud of thermal neutrons surrounding the source will resemble a sphere. The phrase "sphere of influence" (Van Bavel et al, 1956; Glasstone and Edlund, 1957) or the "sphere of importance" (Olgaard, 1965) denotes the zone of neutron flux that is 95% of the total flux obtained in an infinite medium. An empirical expression suggested by [Olgaard, 1965) relates the surrounding water content to the radius of the "sphere of importance". 1.4 + 10^. where 9^ is the volumetric water content and R is the radius of the sphere [L] of importance in centimeters. The radius calculated in this manner will provide the analyst a rough estimate on the size of the container to be used for calibration purposes. For wet soils, the sphere of influence is smaller compared to dry soils. For example, a completely dry soil will have a R of 70cm, while a wet soil with a volumetric moisture content of = 0.35) will decrease the R to 22cm. The shape of the thermal neutron cloud will not be spherical for a heterogeneous soil. In a soil profile consisting of strata with large differences in lithology, the moisture content profile is discontinuous across those layers. The vertical resolution of the neutron probe is the ability of the neutron probe to distinguish changes in moisture content with depth and has been examined by McHenry (1963). The study was conducted by placing a soil type with higher moisture content between two dry soils. One of the most important conclusions reached by this author was 241 that a neutron probe can distinguish up to 1 inch of wet soil that intervenes two very dry soil layers. B.3: THE STANDARD CALIBRATION METHOD OF NEUTRON PROBE IN GRANULAR MATERIALS The neutron-scattering method as stated previously, is an indirect method of obtaining water content. Thermal neutron counts are related to volumetric water content through a site-specific calibration curve. The accuracy of the calculated water content values hence depends on the accuracy of the calibration curve. Calibration, although important, is often neglected in some applications when neutron probes are employed to measure moisture content. A factory calibration curve is usually shipped with the device and in some applications, this calibration curve is sufficient. Calibration of the instrument, if possible, should be taken seriously if one is to obtain credible results fi'om field measurements. Relative changes in water content from one sampled region to the other may be obtained, but an absolute water content cannot be obtained unless a detailed calibration study is conducted. The standard calibration procedure suggested by the American Society for Testing and Materials will be described below. Calibration should be performed following ASTM-D3017, so that results can be compared and replicated. The first step involved in the calibration of the instrument is to standardize the equipment to a homogeneous sample with a known value of water content. In order to standardize the equipment, the ratio of the measurement count rate to the count rate made on a reference standard is recorded. Standardization is required each time the instrument is employed to ensure that 242 the equipment is functioning correctly. Radioactive decay, leakage, failure of electrical components and other factors that may affect the performance of the instrument should be monitored continuously by standardizing the probe before the investigation commences. A calibration curve/line can be obtained from at least two samples of different known water content values. A large container of well-mixed, compacted material with a known water content and density should be prepared for the calibration study. Dimensions of the container that exceeds 610 mm * 460 mm. * 360 mm. has been suggested to be satisfactory. The use of a 55 Gal drum is widely accepted and an aluminum access tube is installed at the center of the drum. Measurements are taken by carefully lowering the probe down the access tube. The probe is lowered to the middle of the access tube, where neutron counts are taken. It is suggested that the measurements are taken at the central portion of the container to avoid boundary effects. Boundary effects at the top and bottom of the container can cause deviance from the true value of water content. Repetitive measurements should be taken to minimize instrumental errors. The water content should be determined by the evaporative method (ASTM D2216) and the wet density calculated fi-om the mass of the wet soil and the volume of the container. The water content of the material employed for calibration purposes should be varied through a range of water content values encountered in the field and the density of the material tested should be in the range of 1,600 to 22^Qkg/m^. A calibration line can be established with merely two points. However, multiple measurements of widely varying 243 water content values should be taken for a more reliable calibration curve. A best-fit line is obtained through linear regression analysis. The precision of the instrument should also be calculated periodically. Precision can be determined from the ratio of the standard deviation of thermal neutron counts to the slope of the calibration curve. B.4: FACTORS THAT AFFECT NEUTRON COUNTS The neutron attenuation method is valid based on the premise that I) the hydrogen atoms surrounding the probe is the only source of moderation of fast neutrons and that 2) the majority (95% or more) of thermalized neutrons is registered by the detector. In natural soils, however, thermal neutrons can be produced by other means and/or can be deflected or absorbed by other constituents of the surrounding material. Organic material in shallow soils, is one source of hydrogen and has been shown to affect neutron counts. Elements such as lithium (Li^), cadmium (Cd^*), and chlorine (CI") tend to absorb neutrons that are emitted fi-om the source. All of the aforementioned factors affect the reliability of the measurement. Therefore, an absolute water content, in some cases, may not be obtained with confidence. Early researchers such as Gardner et al. (1952) and Van Bavel et al. (1955) have claimed that a universal calibration curve can be employed in determining soil water content from thermal neutron counts. In contrast, research conducted by Holmes (1966), Lai (1974), Greacen and Hign&XX (1979) and others have shown that the bulk density (pj) of the tested material influences the thermal neutron counts. The bulk density of the 244 soil depends on factors such as the mineral constituent, texture, presence of fractures and clay content. Holmes (1966) compared calibration curves obtained from four different soil types and concluded that the slope of the calibration curve flattened with increasing clay content. The slope becomes flatter because more thermalized neutrons are absorbed for soils with higher clay content which have a larger "macroscopic cross section for absorption of slow neutrons." This author reports that the accuracy of the calibration can be improved by taking into account the absorption of thermalized neutrons. Lai (1974) studied the effect of bulk density and soil texture on the calibration curve. The study was conducted using five different ferrallitic soils obtained in a tropical region. This author reported that both bulk density and texture have a very large effect on the calibration curve and questioned the validity of a universal calibration curve for all soil types. The author clearly showed that coarse textured soils have greater slopes compared to fme-textured soils. The intercept of the calibration curve was found to be higher for higher clay and iron content. Greacen and Hignett (1979) studied the heterogeneous nature of bulk density in fractured clays. They reported that the intercepts of calibration curve can vary from site to site and that the bias resulting fi'om the heterogeneous nature of fractured clays can lead to misleading values of soil water content. The bias can be reduced significantly, if the calibration curve is corrected for variations in bulk density. Organic material in soil has also been found to affect neutron counts. It is known that the average hydrogen content of humus is approximately 5% of its weight, while the 245 hydrogen content in pure water is 11% of its weight. Calibration curves of soils with high organic content tend to have a best-fit line with a larger intercept. Therefore, hydrogen atoms found in humus can play a large role in the production of thermal neutrons and must be taken into account. Elements such as boron (B^^, cadmium (Cd^^, chlorine (CI ), lithium (Li"^), magnesium (Mg^^) and potassium (K^ are known to absorb neutrons, thus a lower thermal neutron count is observed in soils with high concentrations of those elements. Access tube characteristics on neutron counts are often neglected or taken very lightly in most hydrologic and agronomic applications. Research has shown that the characteristics of access tubes such as the diameter, construction material, compaction of surrounding soil and backfill (skin) may have a significant contribution to either an increase or decrease in the soil moisture content. Access tubes for shallow vadose zone investigations are made of aluminum and are usually two inches in diameter. In some cases, neutron probes are sometimes lowered into monitoring boreholes that have a much larger diameter (4 to 6 inches). Air space between the probe and the borehole wall seems to have a quite significant effect - a 35% reduction in neutron counts (personal communication, technician from Boart-Longyear CPN corp.). Tyler (1988) studied the effects of large air gaps between the detector and the surrounding material and came to a conclusion that a calibration of the instrument is possible with a nonlinear calibration curve. An in-house study performed by Boart-Longyear CPN corp. (1988), the manufacturer of a neutron (and density) probe, demonstrated that the calibration curve for 246 an over-sized tubing had a flatter slope compared to the factory calibration standard. Researchers compared the factory calibration to calibration curves obtained for the case when 1) the probe was placed against the wall and 2) when the probe was centered in the borehole. The calibration curves for the two cases are nearly identical with the slope slightly steeper for the case when the probe is placed against the wall. One of the conclusions reached by the workers was with proper calibration, the probe can be employed in access tubing larger than the suggested nominal 2 inch diameter. The material of access tubes and compaction of surrounding soil during the installation of access tubes have also shown to affect moisture content data. Aluminum is the obvious choice for the material because it is virtually invisible to the fast and thermalized neutrons. In some cases, PVC pipes have been used at field sites and found to be satisfactory, when the calibration is performed correctly. Compaction of surrounding soil and presence of water on access tube wall (from condensation) may increase the apparent moisture content. The sphere of influence or importance discussed previously is spherical for a homogenous soil with uniform moisture distribution. A value of water content (for this idealized case) obtained from neutron investigation in homogenous soils represents a unique value due to the uniform distribution of moisture. The sphere of influence, however, loses its symmetry when neutron probes are employed in layered or laterally heterogeneous soils/rocks. Water content determined in heterogeneous soils, is therefore, an averaged value of some volume depending on the sphere of influence. 247 Textural differences among layers sometimes cause discontinuities in water content across strata. McHenry (1963) studied the effect of layering of soils on the vertical moisture distribution using cadmium tape. He reported that simulating vertical heterogeneity by placing intervening wet layers between very dry layers was technically difficult. Therefore, sharp moisture boundaries were simulated by using cadmium tape in the access tube. Cadmium has the property of absorbing most neutrons emitted fi^om the source. He reasoned that portions of access tube taped with cadmium can be made to appear as a dry layer, while the untaped portion can be used to simulate a wet layer. He admitted that the loss of thermal neutrons by absorption through cadmium tape was not identical to the situation in which thermal neutrons were lost in very dry soils nevertheless, it was a good approximation. He considered three cases: I) the influence of a dry soil layer between two wet layers, 2) the influence of a wet layer between two dry layers and 3) multiple wet layers placed between dry layers. The thickness of intervening layers were varied from one to twelve inches. From his study, he found that a minimum thickness of one inch wet or dry layer can be distinguished, but an absolute water content cannot be determined because of the dominant influence of the over and underlying strata. He also determined that a minimum thickness (12.5 inches) of a soil layer to yield a measurement of neutron density unaffected by adjacent layers. 248 B.5: REFERENCES PERTAINING TO APPENDIX B ASTM D2216-7I Standard Method of Laboratory Determination of Moisture Content of Soil, 290-291 ASTM D3017-88 Standard Test Method for Water Content of Soil and Rock in Place by Nuclear Methods (Shallow Depth), 411-415 Belcher, D. J., T. R. Cuykendall, and H. S. Sack, The measurements of soil moisture and density by neutron and gamma-ray scattering. Civil Aero. Adm. Tech. Dev. Rept. v. 127, Washington D.C., 1950. Cotecchia V., A. Inzaghi, E. Pirastru, and R. Ricchena, Influence of the physical and chemical properties of soil on measurements of water content using neutron probes. Water Resour. Research, A, 1025-1028, 1968. CPN, An in-house study on the effect of air gaps on the 501 DR probe, 1988. Elder, A. N., 1988, Neutron Gauge Calibration Model for Water Content of Geologic Media, Unpublished M.S. Thesis, University of Arizona, Tucson, AZ, 181 pp., 1988. Gardner, Wilford, and D. Kirkham, Determination of soil moisture by neutron scattering. 73;391-401, 1952. Soil Sci. Glasstone, E., and M. C. Edlund, 1957, The elements of nuclear theory. Van Nostrand Co. Inc., New Jersey. Greacen, E. L., and C. T. Hignett, Sources and bias in the filed calibration of a neutron meter. Anst. J. Soil Res. 17:405-415, 1979. Greacen E. L. and G. Schrale, The Effect of Bulk Density on Neutron Meter Calibration, Vol. 14, pp. 159-169, 1976. Aust. J. Soil Res., Holmes, J. W., Influence of bulk density of the soil on neutron moisture calibration. Soil Sci. 102:355-360, 1966. Klute, A., (editor). Methods of Soil Analysis Part I-Physical and Mineralogical Methods, SSSABook Series: 5, 1986. Lai, R. The effect of soil texture and density on the neutron and density probe calibration for some tropical soils. Soil Sci. 117:183-190, 1974. 249 McHenry, J. R. 1963, Theory and application of neutron scattering in the measurement of soil moisture. Soil Sci. 95:294-307, 1963. McTigue D. F., C. L. Stein, W. A. Illman, M. J. Martinez, and T. D. Burford, Heat and Vapor Transport Beneath and Impermeable Barrier, Abstract presented at the 1993 Fall Meeting, San Francisco, CA, 1993. Olgaard, P. L., On the theory of the neutronic method for measuring the water content of soil. Danish Atomic Energy Comm., Riso Rep. 97, 1965. Rawitz, E., Installation and Field Calibration of Neutron-Scatter Equipment for Hydrological Research in Heterogeneous and Stony Soils, Water Resour. Res., 5(2), 519523, 1969. Sinclair, D. F., and J. Williams, Components of variance involved in estimating soil water content and water content change using a neutron moisture meter. Aust. J. Soil Res. 17:237-47, 1979. Tyler, S. W., Neutron Moisture Meter Calibration in Large Diameter Boreholes, Soil Sci. Soc. Am. J., V.52, pp.890-893, 1988. U.S. Dept. of Interior, National Handbook of Recommended Methods for Water Data - A United States Contribution to the International Hydrological Program, Chapter 6 - Soil Water, 6-1 to 6-3, 1977. Acquisition Van Bavel, C. H. M., N. Underwood, and R. W. Swanson. 1955, Soil moisture measurement by neutron moderation. Soil Sci., 82:29-41, 1955. Visvalingam, M., and J. D. Tandy, The neutron method for measuring soil moisture content-a review. Soil Sci., 13 A99-S\\, 1972. 250 APPENDIX C: SLIP FLOW Consider single-phase flow of a Newtonian fluid in a porous or fractured medium. Knudsen (1934) deflned a dimensionless number K„ = X/d (called Knudsen number) where A is the mean free path of fluid molecules (average distance between molecules) and ^ is a characteristic length. In porous or fractured media, d is typically a measure of pore diameter or firacture aperture {Present, 1958; Bear, 1972). When K„ <0.01, the fluid behaves (on the pore or fi-acture aperture scale) as a viscous continuum and satisfies the Navier-Stokes equation. At sufficiently small Reynolds numbers, a linear Stokes regime develops within pores or fractures in which viscous forces are large in comparison to nonlinear inertia so that the latter can be disregarded. The fluid continuum satisfies a noslip, or zero velocity, condition at its contact with the solid walls of pores or fi-actures. On a macroscopic scale that includes many pores or fracture apertures the Stokes regime, under the no-slip condition, manifests itself in the form of Darcy's law. As the Knudsen number increases toward 1, viscosity and drag along solid walls diminish sufficiently to allow slippage of fluid past solid walls to occur. This allows flow rate to be higher than predicted by Darcy's law, a phenomenon known as slip flow or the Klinkenberg (1941) effect. As K„ increases further, the fluid behaves less and less as a viscous continuum a n d m o r e a n d m o r e a s a diffusing collection o f fi-ee-flowing molecules. W h e n K„ > 1 , Knudsen diffusion becomes dominant. As the mean free path of gas molecules is much larger than that of liquid molecules, the former is associated with much larger Knudsen numbers and one must therefore consider the possibility that gas may exhibit the Klinkenberg effect or, less commonly, be subject to Knudsen diffusion. 251 A decrease in gas pressure brings about an increase in the mean free path of its molecules and a concomitant increase in the Knudsen number. If one uses Darcy's law to describe the associated Klinkenberg effect, one finds that permeability appears to increase as pressure decreases. Indeed, based on an analogy to gas flow in a capillary tube, Klinkenberg {\9AV) proposed that the apparent permeability of gas varies according to (C-1) where k is intrinsic permeability at high pressures (small Knudsen numbers), c ^ 1 is a proportionality factor, r is the radius of the capillary tube, and m is an empirical coefficient valid for porous media that depends on k and pore size. As pore size and permeability are related, m can be considered a function of k. Though such inverse relationship between apparent permeability and pressure is often observed in the laboratory, it was pointed out by Aronofsky (1954) that the Klinkenberg effect seldom affects the interpretation of gas well tests in the field. It was mentioned in the Introduction (section 1.2.3) that, according to Guzman et al. (1994), Knudsen diffusion and the Klinkenberg effect play virtually no role in the observed pressure dependence of air permeabilities obtained from single-hole pneumatic injection tests at the ALRS; the observed decrease in apparent permeability with pressure is due to inertia effects, and the observed increase is due to two-phase flow. Most pneumatic injection tests at the ALRS "see" fractures, which have mean apertures that are considerably larger than the mean pore diameter of intact matrix at the 252 site (see Table C. I for pore size distribution and Figure C. I for mean pore sizes of intact matrix cores; ydgt, 1988; Rasmussen et al. 1990). We suspect that this is a major reason why these tests appear to be unaffected by the Klinkenberg effect. To understand how this effect is related to pore size, we consider the work of Scott and Dullien (1962a) who proposed the following quasi-Darcian relationship for gasflow. F=- ^ rp _2 ArRT 8// ' Vp (C-2) RT where F is molar flux of gas [molL'^T'], r is mean pore radius [Z-], M is molecular weight and is mean molecular velocity [LT'^. Volume flux is related to molar flux via {Massmann, 1989) q=— (c-3) P This, coupled with the Boyle-Mariotte law for real gases {Burcik, 1957) P=— (C-4) ZRT transform (C-2) into q= (C-5) 8// In the absence of slip flow, (C-5) reduces to Darcy's law in the form 8/i (C-6) The first term on the right hand sides of (C-2) and (C-5) accounts for Darcian flow and the second for slip flow. These equations imply that (a) as r becomes large, the 253 importance of slip flow relative to Darcian flow diminishes; (b) the opposite happens as r becomes small; and (c) gas-flow involving light molecules that travel at low velocities through large pores can exhibit a large slip component. According to Harleman et al. (1963), intrinsic permeability is proportional to the square of mean pore diameter d, k = cd^ (C-8) where c is a constant. Therefore, eq. (C-6) can be rewritten as q = --Vp (C-9) which is Darcy's law expressed in terms of pressure when the flowing gas is assumed to b e ideal ( Z = l ) . 254 Table C.1: Cumulative mercury intrusion volume as a function of equivalent pore diameter (adapted from Rasmussen et al. 1990) Pore Size [fm\ Cumulative Volumetric Porosity Volumetric Porosity [m^/m^] Relative Frequency 13.8 9.19 7.64 6.3 5.08 4.06 3.27 2.61 2.07 1.65 1.3 1.03 0.802 0.627 0.494 0.385 0.302 0.237 0.187 0.146 0.115 0.0896 0.07 0.0548 0.0429 0.0336 0.0262 0.0205 0.016 0.0125 0.0098 0.0077 0.0003 0.0005 0.0011 0.0024 0.006 0.0125 0.0213 0.0308 0.041 0.0487 0.0544 0.059 0.0631 0.0666 0.0696 0.0726 0.0756 0.0787 0.0819 0.0859 0.0907 0.0969 0.1042 0.1116 0.1182 0.1237 0.1284 0.1327 0.1369 0.1397 0.1417 0.1428 Sum 0.0003 0.0002 0.0006 0.0013 0.0036 0.0065 0.0088 0.0095 0.0102 0.0077 0.0057 0.0046 0.0041 0.0035 0.003 0.003 0.003 0.0031 0.0032 0.004 0.0048 0.0062 0.0073 0.0074 0.0066 0.0055 0.0047 0.0043 0.0042 0.0028 0.002 0.0011 0.1428 0.002101 0.001401 0.004202 0.009104 0.02521 0.045518 0.061625 0.066527 0.071429 0.053922 0.039916 0.032213 0.028711 0.02451 0.021008 0.021008 0.021008 0.021709 0.022409 0.028011 0.033613 0.043417 0.05112 0.051821 0.046218 0.038515 0.032913 0.030112 0.029412 0.019608 0.014006 0.007703 1 255 0.2 0.18 0.16 0.14 i 0.12 er £ u, 0.1 u I 0.08 OS *• slip flow is negligible slip flow needs to be accounted for 0.06 • 0.04 • • • •• 0.02 0 0.001 A • 0.01 0.1 1 • 10 Equivalent Mean Pore Diameter (^m) Figure C.1: Relative frequency of equivalent mean pore diameter plotted on semilogarithmic scale, showing bimodal distribution of pores in tuff matrix at the ALRS (data Rasmussen et al. 1990). 100 256 APPENDIX D: AIR COMPRESSIBILITY AND VISCOSITY IN RELATION TO PRESSURE AND TEMPERATURE AT THE ALRS In this appendix we include a sample calculation of the compressibility factor Z for air under standard conditions, tables of pressures and temperatures encountered during single-hole pneumatic injection tests over \-m intervals at the ALRS, and a table of air viscosity values under a range of temperatures. The compressibility factor for gas is defined in (2.1-9) as where V, = V/n. To assess Z for air, we note from Table D. 1 (CRC Handbook, 19921993) that the molecular weight of standard atmospheric air in the United States is 28.96443 g/mol. = 1.246629 • Its specific volume is therefore {Vasserman et Imol. As al. 1966) the universal gas constant is /? = 8.31441 Nm /" AT mol we find that, under standard conditions of /? = 200 kPa and T = 300°A!', (200,000/'a)(l246629 x lO'"'"' (8314510 A/m/°A:/no/)(300°^) Otherwise Z varies with p and T as shown in Figure 2.1. Table D.2 lists the minimum, maximum, average and range of stable injection pressures attained in each borehole within which single-hole pneumatic injection tests took place over \-m intervals at the ALRS. Table D.3 shows the same for temperatures. 257 Table D.4 lists values of dynamic viscosity for air over a range of pressures and temperatures comparable to those encountered at the ALRS. The same data are shown graphically in Figure 2.2. Table D.l: Gaseous composition of U.S. standard atmosphere Molecular Weight \kg/kmo[\ 28.0134 31.9988 39.948 44.00995 20.183 4.0026 83.8 131.3 16.04303 2.01594 Gas Species N2 O2 AT CO2 Ne He Kr Xe CH4 H, Total Fractional Volume Idimensionlessl 0.78084 0.209476 0.00934 0.000314 0.00001818 0.00000524 0.00000114 0.000000087 0.000002 0.0000005 1" Fractional Molecular Weight \g/mo[\ 21.87398 6.702981 0.373114 0.013819 0.000367 2.1e-5 9.55e-5 1.14e-5 3.21e-5 I.Ole-6 28.96443 •1 Table D.2: Data concerning stable pressures encountered during \ - m scale single-hole Borehole V2 W2A X2 Y2 Y3 Z2 Minimum Pressure \Pa\ 88886.04 88419.42 89805.97 89632.65 88379.42 90072.61 Maximum Pressure \Pa\ 248592.9 240073.6 265058.3 233380.9 361610.3 240273.6 Mean Pressure [Pa\ 159460.9 147661.9 144005.1 146031.9 162972.3 152740.5 Range ofPressxues [ \Pa\ 159706.9 151654.2 175252.3 143748.2 273230.9 150201.0 Table D.3; Data concerning stable temperatures encountered during l-nt scale single-hole pneumatic in ection tests at the ALRS (summarizec from Guzman et al. 1996) Borehole V2 V/2A X2 Y2 Y3 Z2 Minimum Temp r'AH 293.09 292.57 293.04 288.50 288.10 291.32 Maximum Temp ["K] 300.90 299.07 302.60 293.93 298.93 302.74 Mean Temp m 294.40 293.40 294.82 293.14 293.02 294.22 Range of Temps fATI 7.81 6.50 9.56 5.43 10.83 11.42 258 Table D.4: Variation of air dynamic viscosity with temperature and pressure (data from ?{kPa\ l.OE+02 l.OE+03 2.0E-H)3 3.0E-H)3 4.0E-K)3 5.0E+03 6.0E-K)3 7.0E-K)3 8.0E-H)3 9.0E-H)3 1.0E-H)4 TfAl T = 270TC 1.6960E-0S 1.7110E-05 1.7280E-05 1.7480E-05 1.7690E-05 1.7930E-05 1.8190E-05 1.8460E-05 1.8750E-05 1.9070E-05 1.9410E-05 T[''iq T = 280°K I.7460E-05 1.7600E-05 1.7770E-05 1.7950E-05 1.8160E-05 1.8380E-05 1.8620E-05 1.8880E-05 1.9140E-05 1.9440E-05 1.9760E-05 TC/n T = 290'TC I.7960E-05 1.8100E-05 1.8260E-05 1.8430E-05 1.8630E-05 1.8830E-05 1.9060E-05 I.9300E-05 1.9550E-05 1.9830E-05 2.0120E-05 TL'An T = 300°K I.8460E-05 I.8590E-05 1.8740E-05 1.8910E-05 I.9100E-05 1.9290E-05 1.9510E-05 1.9730E-05 1.9970E-05 2.0230E-05 2.0500E-05 TI'Al T = 310'X 1.8960E-05 1.9100E-05 1.9230E-05 1.9390E-05 I.9570E-05 1.9760E-05 1.9960E-05 2.0I70E-05 2.0400E-05 2.0640E-05 2.0890E-05 TCAH T = 320TC 1.9450E-05 1.9580E-05 1.9710E-05 1.9860E-05 2.0030E-05 2.0210E-05 2.0400E-05 2.0610E-05 2.0820E-05 2.1050E-05 2.1290E-05 259 APPENDIX E: DERIVATION OF SINGLE-PHASE GAS FLOW EQUATION The continuity equation for single-phase gas flow in a rigid porous medium is -V.{pq) = ,S^ (E-1) Substituting the Boyle-Mariotte relationship {Burcik, 1957) pM P= (E-2) TRT and Darcy's law (2.1-2) for a gas into (E-I) yields M p M p k^ ^ V- vp As M and R are (E-3) dt RTZ n constant for a homogeneous gas, and T is constant under isothermal conditions, this reduces to V- (E-4) Z // The latter form of the single-phase gas flow equation is found, for example, in Hussainy et al. (1966) and Bear Al- (1972, p. 200, eq. 6.2.26). In a medium with uniform permeability. = £Af £ kaivz (E-5) which can be written as (E-6) V. 2pZ J k dt\ZJ 260 Expanding the left-hand-side of (E-6) yields \.2fjZ) ^ ^ =-—f-1 (E-7) kdt\z) Multiplying both sides by 2/zZ gives •vp' ^ ^ 2 = ^ m £ a ( p ] ^il ^ k (E-8) dAz) Remembering that /j Z is pressure dependent, we note that d \ <?xL/izJ d ( \ ^dp- _ d ^ 1 ^ d{^iZ) dp- dp'Ki^j dx 1 dp- d{^)dp- (pZ)' dp dx (E-9) <ip' (E-IO) dp' and therefore d? '' 1 ^ _ ^xv/izj 1 d\n{jij^dp' (/iZ) dp' dx (E-11) It follows from (E-8) and (E-11) that dp' k dt (I) (E-12) Upon taking p « constant and Z « 1, vv- 2it>p dp k dt (E-I3) 261 which is the same as (2.1-11) and a form found in Scheidegger (1974). Defining the compressibility of air as (E-I4) P dp we obtain from (E-2) and (E-14), for isothermal conditions. _Z{p)RT d ^ pM ^ _Z(p) d ^ P 'W = pM dp Z{p)RT, P dp zip)} (E-IS) This simplifies to / \ I c\p)= p I dZ{p) \ V Z{p) dp (E-16) If Z is a weak function of pressure, air behaves almost ideally and compressibility can be approximated by c{p)~P (E-I7) 1_( P \- P^{p)^P Z{p) dt (E-18) Otherwise, it follows from (E-15) that Substitution of (E-18) into (E-12) yields (Raghavari, 1993, p. 27, eq. 3.21) dp^ vp- k dt (E-19) 262 APPENDIX F: TYPE-CURVE SOLUTION OF SPHERICAL GAS FLOW In a manner analogous to Joseph andKoederitz (1985) who dealt with a liquid, we consider the injection of gas at a constant mass rate Q from a spherical source of radius into a uniform, isotropic continuum of infinite extent. The flow is spherically symmetric so that surfaces of equal pressure (isobars) form concentric spheres about the source. Pseudopressure is then governed by — + 2 Sw = I dw dr' r 3r _ ,. (F-I) a dt where r is radial distance from the center of the source. In our case the source is closer to a cylinder of length L and radius r„. Under steady state, the isobars around such a cylinder can be closely approximated by prolatespheroids or confocal ellipsoids {Moron and Finklea, 1962, Culham, 1974, and Joseph and Koederitz (1985). It is convenient to view the source as the innermost ellipsoid with semi axes a and b The radius = c. Joseph andKoedertiz r^^ of a sphere having similar volume is then given by (1985) as r L (F-2) Equation (F-1) is solved subject to the initial and outer boundary conditions w(r,0) = 0; r>r^ (F-3) lim w(ir,f) = 0; />0 (F-4) 263 Mass balance within the injection interval (Figure F.1) is written in terms of pressure as 0mA„ - QwPw = ^ (F-5) where the subscripts in and w refer to the injection line and injection interval, respectively, and is a borehole storage coefficient Mass flow rate from a spherical source into the surrounding rock is given in terms of pressure by Darcy's law, in combination with (E-2), as = (F.6) 'n. For a low-pressure system such as ours, eq. (2.1-16) implies that dr 2p dp /iZ dr dM/ dt 2p dp fjiZ. 3t so that (F-6) can be written in terms of pseudopressure as = -2^^^ dw dr. (F-7) By virtue of (E-2), the rate at which mass within the interval increases with time can be re written in terms of pseudopressure as ^ r ^ " dt ZRT "2/7^ dt ^ pM dw^ 2RT dt Substituting (F-7) and (F-8) into (F-5) yields the inner boundary condition in terms of pseudopressure. 264 ^2RT dt RT\ dr (F-9) QmPm If air density in the injection system and in the formation are the same {pi„ pj), (F-9) simplifies to ( 2 dw =a (F-IO) 2/?^ dt Equation (3.1-5), ^ (F-11) t„p allows rewriting (F-10) as '^qj'p^ p^ di 2 dw qjpk \ ~d? =1 (F-12) '•n. An infinitesimal skin may cause pressure and pseudopressure differences (F-13) (F-14) to develop across it, as shown schematically in Figure F.2, where the subscripts s and wf represent skin and the rock just outside the skin, respectively. As the skin is infinitesimally thin, it cannot store gas and therefore the pressure drop across it is constant, as required by the Laplace equation. This allows defining a constant dimensionless skin factor through s 265 r t- ^ Aw, = —qsc te^ 27dcr_ V t y (F-15) where 5 is a dimensionless skin factor. Substituting (F-15) into (F-14) yields the auxiliary inner boundary condition = "'w + a. (tp.\ (F-16) 27dcr„ KT„) The complete mathematical problem for spherical flow with storage and skin effects consists of the governing diffusion equation (F-1), initial condition (F-3), outer boundary condition (F-4), inner boundary condition (F-12) and auxiliary inner boundary condition (F-16). The following dimensionless variables reduce the dimensional boundary value problem into a dimensionless form: D~ r 'd , = 1 -— ' r>r_ 4 • r r >r (F-17) The dimensionless form of the complete mathematical problem is dlr =o) = 0; 0<rD < 1 (F-18) (F-I9) 266 lim{H'o(ro,/o)} = 0; t ^ > 0 r ° du =1 drr. »^«o('D)=»^D('b=0>'D)- where (F-20) (F-21) ro=0 (^) (F-22) =C^/'/4;r^^ is a dimensionless wellbore storage coefficient. Note that the dimensionless form of the boundary value problem is virtually identical to that of liquid flow given in Joseph and Koederitz (1985). The definition of the radial flow boundary value problem follows in a similar manner to the above development. 267 Q,rP„ i QwPw ^ • pC ^ * dt Figure F.1: Mass conservation in injection system. qmpm 268 w wwf /N ^w(pj = V w W. Infinitesimal skin zone •> r r^+5r Figure F.2; Impact of positive skin on pseudopressure around the injection interval. 269 APPENDIX G: CODE FOR NUMERICAL INVERSION OF LAPLACE TRANSFORM To invert analytical expressions from Laplace to real time domain we employ the algorithm of DeHoog et al. (1982). This algorithm relies on accelerated convergence of Fourier series obtained from the Laplace inversion integral by means of the trapezoidal rule. It is similar in principle to that of Crump (1976) but is more accurate, converges faster, and is thus computationally more efficient. The Fortran code included below was modified slightly fi-om Press et al. (1992) for double complex arithmetic to allow accurate computation of Bessel functions. The program computes one value of for each value of tp listed as a data statement in the main program. After the end of the computation, a data file is written in the same directory in which the executable file resides. Numerically inverted results are then imported into a standard spreadsheet program for the purposes of generating type curves. The sample program given below is the one that inverts the Laplace transform for the radial flow problem. The program consists of three main parts: 1) a main program which assigns a value of dimensionless time, evaluates the Laplace transform in a subroutine, and writes the calculated value into an output file, 2) subroutines used to evaluate Bessel functions, and 3) a function statement which does the actual inversion of the Laplace transform to real-time space. A few comments are added before key operations in the main programs for clarity. This computer program can be applied to essentially any problem, which requires the inversion of the Laplace transform into real time space by rewriting the Laplace transform listed in the subroutine (RADIAL). 270 PROGRAM WELLBORE IMPLICIT REAL*8 (A-H, 0-Z) REAL niax_time, time DOUBLE PRECISION inversion DOUBLE COMPLEX conc_lap( 15), p i DIMENSION TDW(8) DATATDW/ LODOO. 1.25DOO, 2.50D00, 3.750D00, 5.0D00, 6.250D00, " 7.50D00,8.250D00/ WRITE (*,*) 'max_time=?' READ (*,*) max time WRITC (*,*) Enter a rD value;' READ (*,*) rD WRITE (*,*) "Enter the dimensionless storage value (CD);' READ (•,•) CD C C Opening output data file OPEN (6,FILE= '2ider-l.dat', STATUS='UNXNOWN') C C C Writing file headings into file WRITE (6,*) 'rD=', rD, •CD=', CD WRITE (6,*) 'tDw, tDCD, C C C C C C C mD' Operation which takes the value of data statement given at the beginning of the main program and assigns a value on the logarithmic scale. Eight values of dimensionless time are input for each log cycle. This number can be changed in the data statement and in the DO statement given below for better resolution of the type curves. DO 3000 j= 1,30 DO 3000 k = 1, 8 tD= tDw(K) * (10.0D00**G-5)) max_time = sngl(tD) p0=8.634694/max_time 271 DOi=l, 15 p i = dcniplx( pO, (i-l)*3.141592654/(0.8*max_time) ) CALL RADIAL (p_i, CD, RD, conc_lap(i)) END DO time = maxtime wow = inversion(conc_lap, time, max time) mD = wow tDCD = tD / CD C C C Writing calculated results into the output data file WRITE (6,5000) tD, tDCD, mD WRITE (*,5000) tD, tDCD, mD 3000 CONTINUE 5000 FORMAT (6(E15.7, IX)) STOP END C C C C C Definition of the Laplace transform for the radial flow problem in Laplace space. Note that all input variables are converted to double -precision, complex arithmetic including numbers for improved accuracy. SUBROUTINE RADIAL (p_i, CD, RD, concjap) IMPLICIT DOUBLE COMPLEX (A-H, 0-Z) REAL*8 CD, RD DCCD=DCMPLX(CD) DIMR=DCMPLX(RD) RMD_i= CDSQRT(pJ) CALL IBESSKO (RMDJ*DIMR, SOLI) CALL IBESSKO (RMDJ, S0L2) CALL IBESSKl (RMDJ, SOLS) S0L4 = (p_i*SOL2*DCCD+RMD_i»SOL3»DCMPLX(0.5)) conclap = SOL1/SOL4 RETURN END 272 C C C C C A series of subroutines borrowed from Numerical Recipes (1986) that evaluate various Bessel function for a particular argument. Note that all functions and numbers are rewritten in double-precision, complex arithmetic. SUBROUTINE IBESSKO(X,ZZ) IMPLICIT DOUBLE COMPLEX (a-h,o-z) REAL*8 RP,RQ DIMENSION RP(7),RQ(7),P(7),Q(7) DATARP/-O.57721566DO,0.4227842ODO,O.23O69756DO, * 0.3488590D-l,0.262698D-2,0.10750D-3,0.74D-5/ DATA RQ/1.25331414D0,-0.7832358D-1,0.2189568D-1, * -0.1062446D-l,0.587872D-2,-0.251540D-2,0.53208D-3/ C DO 10 1=1,7 P(I) = DCMPLX(RP(I)) 10 QG) = DCMPLX(RQ(I)) IF (CDABS(X) LE. 2.0D00) THEN Y=X*X/DCMPLX(4.0) CALL IBESSIO(X,BESSIO) X2 = X/DCMPLX(2.0) X3 = -CDLOG(X2)*BESSIO ZZ=X3+(P( 1)+Y*(P(2)+Y*(P(3)+ * Y*(P(4)+Y*(P(5)+Y*(P(6)+Y*P(7))))))) ELSE Y=(DCMPLX(2.0)/X) ZZ=(CDEXP(-X)/CDSQRT(X))»(Q( 1 )+Y*(Q(2)+Y»(Q(3)+ Y*(Q(4)+Y*(Q(5)+Y*(Q(6)+Y*Q(7))))))) ENDIF RETURN END SUBROUTINE IBESSK1(X,ZZ) IMPLICIT DOUBLE COMPLEX (A-H,0-Z) REAL»8 RP,RQ DIMENSION RP(7),RQ(7),P(7),Q(7) DATA RP/1.0D0,0.15443144D0,-0.67278579D0, • -0.18156897D0,-0.1919402D-1,-0.1 l0404D-2,-0.4686D-4/ DATA RQ/1.25331414D0,0.23498619D0,-0.3655620D-1, » 0.1504268D-l,-0.780353D-2,0.325614D-2,-0.68245D-3/ C DO 10 1=1,7 273 10 P(I) = DCMPLX(RPa)) QO) = DCMPLX(RQ(I)) IF ( CABSCX) LE. 2.0 ) THEN Y=X»X/DCMPLX(4.0) CALL IBESSI1(X,BESSI1) ZZ=(CLOG(X/DCMPLX(2.0))*BESSI 1)+(DCMPLX( 1.0)/X)*(P( 1)+Y*(P(2)+ • Y*(P(3)+Y*(P(4)+Y*(P(5)+Y*(P(6)+Y*P(7))))))) ELSE Y=DCMPLX(2.0)/X ZZ=(CEXP(-X)/CSQRT(X))*(Q( I )+Y*(Q(2)+Y*(Q(3)+ * Y*(Q(4)+Y*(Q(5)+Y*(Q(6)+Y*Q(7))))))) ENDIF RETURN END SUBROUTINE IBESSIO(X,ZZ) IMPLICIT DOUBLE COMPLEX (A-H,0-Z) REAL»8 RP,RQ,AX DIMENSION RP(7),RQ(9),P(7),Q(9) DATA RP/1.0D0,3.5156229D0,3.0899424D0,1.2067492D0, • 0.2659732D0,0.360768D-l,0.45813D-2/ DATA RQ/0.39894228D0,0.1328592D-1, • 0.225319D-2,-0.157565D-2,0.916281D-2,-0.2057706D-1, » 0.26355370-1,-0.1647633D-l,0.392377D-2/ C DO 10 1=1,7 10 P(I) = DCMPLX(RP(I)) DO 20 1=1,9 20 Q(I) = DCMPLX(RQ(I)) C IF (CABS(X).LT.3.75) THEN Y=(X/DCMPLX(3.75))**2 ZZ=P(1)+Y*(P(2)+Y*(P(3)+Y*(P(4)+Y*(P(5)+Y*CP(6)+Y*P(7)))))) ELSE AX=CABS(X) Y=DCMPLX( 3.75/AX ) ZZ=DCMPLX( EXP(AX)/SQRT(AX) )*(Q(I)+Y*(Q(2)+Y*(Q(3)+Y*(Q(4) • +Y*(Q(5)+Y*(Q(6)+YnQ(7)+Y»(Q(8)+Y*Q(9))))))))) ENDIF RETURN END 274 SUBROUTINE IBESSIl(X,ZZ) IMPLICIT DOUBLE COMPLEX (A-H,0-Z) REAL*8 RP,RQ,AX DIMENSION RP(7),RQ(9),P(7),Q(9) DATA RP/0.5D0,0.87890594D0,0.51498869D0, • 0.I5084934D0,0.2658733D-l,0.30I532D-2,0.3241 ID-3/ DATA RQ/ • 0.39894228D0, -0.3988024D-1, » -0.362018D-2,0.163801D-2,-0.1031555D-1,0.2282967D-1, • -0.2895312D-l,0.1787654D-1,-0.4200590-2/ C DO 10 1=1,7 10 P(I) = DCMPLX(RP(I)) DO 20 1=1,9 20 QO) = DCMPLX(RQ(I)) C IF (CABS(X).LT.3.75) THEN Y=( X / DCMPLX(3.75) )**2 ZZ=X*(P( 1 )+Y*(P(2)+Y*(P(3)+Y*(P(4)+Y*(P(5)+Y*CP(6)+Y*P(7))))))) ELSE AX=CABS(X) Y=DCMPLX( 3.75/AX ) ZZ=DCMPLX(DEXP(AX)/DSQRT(AX))*(Q( 1)+Y*(Q(2)+Y»(Q(3)+Y*(Q(4)+ • Y*(Q(5)+Y*(Q(6)+Y*(Q(7)+Y*(Q(8)+Y*Q(9))))))))) ENDIF RETURN END C C Beginning of the DeHoog et al. (1982) inversion function C double precision function inversion(conc_lap,time,max_time) implicit none integer max_lap,i parameter (maxjap=15) real pi, max_time, pO parameter (pi=3.141592654) real time double complex record(max_lap), conc(max_lap),conc_lap(max_lap) common /transform/conc double complex temp_time, AM, BM lis p0=8.634694/max_time do i=l, max_lap conc(i)=conc_lap(i) end do call coeff_QD(record) temp_time=dcmplx(0.0, pi*time/0.8/max_time) teinp_time=exp(temp_time) call ab(am,bm,record,temp_time) if ( (am.eq.O).and.(bm.eq.O) ) then inversion=0 else if (bm.eq.O) then write(*,«)'ERRORS EXIST IN THE INVERSION CALCULATIONS' stop INVERSION ERROR* else inversion= exp(p0*time)/0.8/max_time*Dreal(AM/BM) end if return end subroutine coefF_QD(record) implicit none integer maxlap parameter (maxJap=15) double complex record(max_lap) integer count integer i, j, trafic common /trafic/trafic double complex epsilon(max_lap, 2), Term count=l doj=l,2 do i=l,maxjap epsilon(ij)=(0.0,0.0) end do end do record(count)=Term(O) do while (count.lt.max_lap) call add(count, epsilon) count=count+l record(count)=-epsilon(1,1) end do return end subroutine ab(am,bm,record,z) implicit none integer max_lap, i parameter (maxjap=15) double complex record(max_lap) double complex z, am, bm double complex temp_l(maxjap), tenip_2(max_lap), h temp_l (1)=record(1) temp_1(2)=record(1) temp_2(l HI-0,0.0) temp_2(2)= I +record(2)*z do i=3, max_lap if (i.lt.maxjap) then temp_l (i)=temp_ 1(i-1)+record(i)*temp_ 1 (i-2)*z temp_2(i)=temp_2(i-1 )+record(i)*temp_2(i-2)*z else if (i.eq.max lap) then h=0.5-K).5*z*( record(i-l) - record(i) ) h=-h*( 1 - sqrt(l+z*record(i)/h**2)) temp_1(i)=temp_1(i-1)+h*temp_ 1(i-2)*z temp_2(i)=temp_2(i-1)+h*temp_2(i-2)*z end if end do am=temp_1(maxjap) bm=temp_2(maxjap) return end subroutine add(count, epsilon) implicit none integer maxlap parameter (max_lap=l 5) double complex epsilon(maxJap, 2) integer count integer trafic, i, switch common /trafic/trafic double complex Term if (trafic.eq.1) then epsilon(count,2)=Term(count)/Term(count-1) switch=l do i=count-1,1,-1 if (switch.gt.O) then epsilon(i,2)=epsilon(i+1,1)+epsilon(i+1,2)-epsilon(i, I) Ill else if (switch.It.O) then epsiIon(i,2)=epsilon(i+1,1)*epsiIon(i+1,2)/epsilon(i, 1) end if switch=-switch end do do i=l, count epsilon(i, l)=epsilon(i,2) end do c end if return end double complex function Term(l) implicit none integer maxlap parameter (maxjap= 15) double complex conc(max_lap) common /transform/conc integer 1, trafic conunon /trafic/trafic if (l.eq.O) then Term= 0.5*Conc(l) else Term=Conc(l+l) end if if ( abs(Term).lt.exp(-50.0)) then trafic=0 else trafic=l end if return end C C C End of the inversion program 278 APPENDIX H: CODE FOR NUMERICAL DIFFERENTIATION OF PNEUMATIC TEST DATA Numerical differentiation of data can be accomplished either directly or indirectly. Direct numerical differentiation is applied to data relatively free of noise that do not require extensive pre-processing. Numerical differentiation, when performed indirectly, involves pre-processing of data b'm(3Uits application. Data that are very noisy may be first smoothed using spline fitting. A number of researchers in the petroleum industry have suggested various differentiation algorithms. Smoothing of data, however, needs to be conducted with extreme caution, because the procedure tends to mask the slight changes in pressure data that may be of interest to the analyst. Researchers such as Blasingame et al. (1989) and Onur et al. (1989) have recently proposed integrating pressure data prior to the approximation of derivatives. The main advantage of employing this technique is that the integration procedure tends to smooth data and makes type-curve matching easier. This technique, as in the case of spline fitting schemes, may cause the loss of character of the original data. The loss of the character in original data will make the subsequently performed pressure derivative analysis futile. Bourdet et al. (1989) suggested a method based on a weighted central difference approximation to calculate the derivative at a reference point i. This algorithm, over the last decade, has become the standard algorithm among petroleum engineers because of its simplicity to implement and has been used to calculate numerical derivatives of pneumatic pressure data colleaed at the ALRS. The algorithm takes the form: 279 dp _ axj dx, where AP, = P, - P,_^ •, APj = P,,i- P. \ - v^xj (H-1) (A^,+A^2) AA', = X, - X,_, ; AAf, = A",,, - A",. Data before (i-1) and after (i+1) the reference data point (i) is employed to calculate the numerical derivative. The weighted mean of the two calculated derivatives (one between the point before and the reference point; the other between the point after and the reference point) are calculated. One disadvantage in using the algorithm suggested by Bourdet et al. (1989) is that when consecutive points are put to use for the calculation, the derivative can be very noisy and may be unsuitable for type-curve analysis. This dilemma can be usually alleviated by choosing a larger interval length (L = AX1+AX2) between successive data points. The interval length cannot be increased blindly, when applied to selected pneumatic data. Manual selection of interval length becomes a very important component of the calculation and it is when the judgment of the analyst counts the most when employing this algorithm. Noisy data will require a larger interval length if this algorithm is employed. Increasing the interval length blindly will increase truncation errors as pointed out by Lane et al. (1991). The selection of the interval length is perhaps one of the few uncertain aspects of this algorithm. 280 CSdebug PROGRAM BOURDET C C C C C C A program which reads values from a given file with two columns (time and pressure). The program calculates the difference of every N pressure readings. If the difference is greater than zero, then the difference is written on an output file. The natural logarithm of the time is also calculated simultaneously and written in an output file. by Walter Illman IMPLICIT REAL*8 (A-H, O-Z) CHARACTER*12 INFILE, OUTFILE DIMENSION T(1500), deIP(1500) C write( *,*) "Enter input filename(.txt):' READ(*,'(A12y) INFILE write( *,*) "Enter output filename(.out):" READ(*,'(A12)") OUTFILE write( *,*) "Enter interval length;" READ •, NLEN write( *,*) "Enter the number of rows;" READ •, NROW C C C C C Number of rows must be specified accurately. If the number of rows is not specified accurately""subscript out of range"" run-time error will be displayed. NUM = NROW OPEN(UNIT= I ,FILE=INFILE,STATUS="OLD") 0PEN(UNIT=2,FILE=0UTFILE,STATUS="UNKNO\VN") C creates input and output files WRITE (2,2000) 2000 FORMAT (lOX, "t15X, "Int", 15X, "delP", I5X, "dP/dlnt") numm=num DO 10I0i=l, NUMm READ (l.*,end=l 111) T(i), delP(i) write(*,*) i, T(i), delP(i) numi=i 1010 continue 1111 continue 281 num=numi write(*,*)'— num = num if(NUM-NLEN .LT. NLEN+1) stop DO 1000 i=NLEN+l, NUM-NLEN il=i-NLEN i2=i i3=i+NLEN C C C C Do calculations in this loop; use an algorithm suggested by Bourdet et al (1989) "Use of Pressure Derivative in Well-Test Interpretation" til =0.0 ti2 = 0.0 ti3 = 0.0 Dint I =0.0 Dlnt2 = 0.0 Dlnt3 = 0.0 delPl =0.0 delP2 = 0.0 DERTV = 0.0 til =DLOG(T(il)) ti2 = DL0G(T(i2)) ti3 = DL0G(T(i3)) Dlntl = ti2 - til Dlnt2 = ti3 - ti2 DInt3 = ti3 - ti 1 delPl = delP(i2) - delP(il) delP2 = delP(i3) - delP(i2) DERIVl =(deIPl/DlntI)»Dlnt2 DERIV2 = (deIP2/Dlnt2)*DlntI DERIV = (DERTV1+DERIV2)/Dlnt3 IF (DERIV .GT. 0.0) then WRITE (2,500) T(i2), ti2, DELP(i2), DERIV WRITE (*,500) T(i2), ti2, DELP(i2), DERTV ENDIF 1000 CONTINUE 500 FORMAT (IX, E15.6, 3X, E15.7, 3X, E15.7, 3X, E15.7) STOP END 282 APPENDIX I: MODIFICATION OF HSIEH AND NEUMAN (1985a) SOLUTION TO ACCOUNT FOR STORAGE AND SKIN IN MONITORING INTERVALS In this appendix we modify the cross-hole solution of Hsieh and Neuman (1985a) to account for storage and skin effects in monitoring intervals by following an approach originally proposed by Black and Kipp (1977). Mass balance in the monitoring interval can be expressed as q^ = c„^ (i-i) dt where is volumetric flow rate into the interval, = Vjp^^ is storage coefficient associated with an interval of volume V, and p^^ is pressure in the interval Hvorslev (1951) expresses volumetric flow rate into the observation interval as (1-2) where F is a shape factor [L], is permeability [L'] which we take to represent a skin, and Ap = /; - pow where p is pressure in the rock outside the skin. Substituting (1-2) into (I-l) and rearranging yields ^p = ^ (1-3) Fk. dt This can be rewritten as (1-4) 283 where tg = C^^lFk/\s basic time lag [7] as defined by Hvorslev. The latter is a characteristic constant of the monitoring interval, which reflects its response time (lag) to changes in pressure within the rock. If tg is known, one can use (1-4) to calculate pressure in the rock outside the skin, based on measurements of pressure within the monitoring interval. Note that when the latter is stable, the two pressures are equal unless (s is infinite due to an impermeable skin. The basic time lag can be determined by means of a pressurized slug test in the monitoring interval according to a graphical method proposed by Hvorslev (1951). For known p, (1-4) is an ordinary differential equation in terms ofpow. In the case where pow = /? at / = 0, its solution is (1-5) Now suppose that p is given by the solution of Hsieh and Neuman (1995a) described in Chapter 5. Dimensionless pressure and time are defined in Chapter 5 as p^ = A7q}kRlOn and ^kfpj^fpP^ where R is distance between the centroids of the injection and monitoring intervals. Introducing a new variable ^ l^kp-q then transforms (1-5) into (1-6) where u= 1/4/^ , Q = Ak^tg/Sf~pR^ is a dimensionless well response time, = (f>lp is a gas storage factor defined in equation 2.1-23, and Po(^) ^^ts as the kernel of a Fredholm 284 integral equation of the 2"** kind. For the case of point-injection/line-observation, the solution oiHsieh and Neuman (1985a, p. 1658, eq. 27) reads Pd = y J -exp[- (l - P:;)y\ • [erf[y""-(p, +1/;3,)] - erf[y"^{p, - l/A)]}'^' ^ »=i/4Jo y (1-7) Substituting this into (1-6) gives (5-46) and (5-47). 285 APPENDIX J: RECORD OF BOREHOLE TELEVISION (BHTV) SURVEY FROM ALRS A record of borehole television survey (BHTV) from boreholes at the ALRS is provided to document discernible downhole features such as fractures, drill rings, cavities, and mineralizations. A. Caster from COLOG (Golden, Colorado) surveyed the boreholes under the supervision of E. Hardin from the University of Arizona. Depths are provided in feet and inches with respect to the lower lip of each borehole casing (LL marker-see Chapter 4) for consistency with other data available at the site. No attempt is made to quantify geometrical properties (aperture width, orientation, etc.) of each fracture due to the moderate to poor quality of borehole videos. In some descriptions, the orientation of each fracture is given by the depth at which the 3 prongs (0, 120, and 240 wands) of the survey camera intersect the fracture. Borehole VI Borehole length; Date of investigation by COLOG; Date of investigation by author; Single-hole pneumatic test data; none Video corrupted Borehole V2 Borehole length; 92.7' Date of investigation by COLOG; 9/12/89 Date of investigation by author; 1/24/96 Smgle-hole pneumatic test data; \-m nominal scale (Guzman et al., 1996) Note; found steel tape at the bottom of the borehole Downhole features high angle fracture 7'7.5" fracture 9'2.0" 286 smooth surface from 10' to 25' high angle fracture top limb 25'2.5", bottom limb 26' 1.0" high angle fracture top limb 26'4.5", bottom limb 27'5.0" near horizontal fracture 28'5.25" high angle fracture top limb 29'7.0", bottom limb 30'9.0" high angle fracture top limb 34'3.0", bottom limb 35'9.25" fracture intersecting third of the borehole at 37'10.25" (top limb?), 38'4.75" (bottom limb?) borehole intersect W2 at 44'8.0" couple of hairline fractures at 62.5 to 63.0 ft. hairline fracture at 78' 82' start of zone with numerous hairline fractures relatively smooth surface from 85' - 92.7' bottom of borehole Borehole V3 Borehole length; 100' Date of investigation by COLOG; 9/12/89 Date of investigation by author: 1/24/96 Single-hole pneumatic test data; none Downhole features high angle fracture 1T fracture 19'0.25" fracture 34' fracture 37' fracture 39'7.5" major horizontal fracture 40'2.75" through 40'1.25" fracture 42.4.75" fracture 43'1.5" fracture 45'8" fractures 46.5' hairline fractures 59'7.5" fracture 66'7.75" fairly smooth surface 70' - 87' Borehole W1 Borehole depth; 52' Date of investigation by COLOG; 9/12/89 Date of investigation by author; 1/24/96 Single-hole pneumatic test data; none Downhole features fracture 10'7.25" 287 fracture 13'3.6" fracture I5'3.25" series of fractures: 1st 17'5.7", 2nd \T1.V\ 3rd ir9.35" high angle hairline fracture 18'7.3" smooth surface (no visible fractures, drill rings) 19' to 39' fracture 0 deg. wand at 39'4.4" - fracture intersecting perpendicular to borehole fracture 40'8.4" fracture 120 wand 4r3.25" fracture 5r2.0" Borehole W2 Borehole length; 102' Date of investigation by COLOG: 9/14/89 Date of investigation by author: 1/26/96 Single-hole pneumatic test data: none Downhole Features end of survey 6.0' relatively smooth surface 6.0' - 16' fracture at 16.39' smooth surface below fracture to right above next fracture fracture at 37.74' fracture at 42.53' fracture at 43.81' loss of video from 49' to 58' fracture at 59'5.2" drill hole intersection at 68'7.3" fracture at 72'5.25" fracture at 80'2.25" fracture at 93'0.4" possible fracture at 95'4.6" 96' level a lot of mud on wall at the bottom of the well - covered with mud with mudcracks - COLOG analyst thought initially that bottom of borehole was being intersected with numerous large fractures Borehole W2A Borehole length: 98.9' Date of investigation by author: 1/26/96 Date of investigation by COLOG: 7/15/93 (Black and white) Single-hole pneumatic test data: \-m nominal scale (Guzman et al., 1996) Downhole Features 7.1' casing 288 relatively smooth surface 7.1' through 12' series of drilling rings/fractures - camiot tell from video drill ring 12' break in wall 28.3' - 28.8' high angle fracture 52' inclusions 64', 64.9' fracture 79.6' fracture 84' bottom of hole 98.9' Borehole W3 Borehole length; 149 to 150' Date of investigation by COLOG; 9/14/89 Date of investigation by author; 1/26/96 Single-hole pneumatic test data; see Table 6.1 in this dissertation Downhole Features 14ft end of tape fracture at 19.9' fracture at 22.36' fracture with 0 wand 29.65', 120 wand 29.74', 240 wand 29.85' fracture at 36.46' fracture at 36.69' fracture at 37' fracture at 41.25' note: measuring scale changes smooth from 41' to 61' fracture at 6r4.5" large fracture at 63'4.3" large open fracture at 63'8.5" fracture at 64'0.6" fracture at 64'2.3" 65' start of breakout zone - end of breakout zone 64'6.25" fracture 65'6.1" break in wall 66'1" opening - break 78'7.8" possible fracture 84'9.0" fracture with 0 wand 93'2.64", 120 wand 93'5.2", 240 wand 93'1.5" drilling ring at 108' fracture at 112'6.35" fracture at 0 wand 123'10.5", 120 wand 124.0', 240 wand 124'1.1" drilling ring 139'0.7" 289 standing water I49'8.5" camera submerged in water!! Borehole XI Borehole length: 63' 6.0" Date of investigation by COLOG; 9/12/89 Date of investigation by author: 1/24/96 Single-hole pneumatic test data: 3-/n nominal scale (Rasmussen et ai, 1990) Downhole Features 7.0 ft. fracture 120 wand 8'6.77" 240 wand at same depth, 0 wands at 8'5.65" 240 wand at 8' 11.25", 0 and 120 wands at 8' 10.25" 240 wand at 10'6.25" 120 wand at 10'7.8", 0 deg wand at 10' 5.4" all wands on same line 11' 7.1' fracture at 14.0' all wands on same line 15' 6.4" 120 wand 16' 2.9", 240 wand at 15'11.1", 0 deg wand??? 15' 7 1" large fracture 120 degree wand 17'0.9", 0 deg wand 16' 10" fracture 240 wand 17'6.2", 0 wand 17'7.3" unknown feature (possibly a fracture) 20' 10.65" fracture 2r7.0 ",2r4.5" fracture 26'3.5" drilling ring at 26.5' drilling rings at 27' - 28' fracture 32'6.5'to 32'5.7" high angle fracture 35'2.5" - 34'11.4" high angle fracture 32'3.7" - 31 '9.75" fracture 36'5.1" to 36'8.2" drilling rings at 38' fracture 39'8.25" drilling rings (smeared) 43.5' another series of drilling rings 48' hairline fracture 50'3.25" drilling ring? 53' drilling ring 58' fracture 120 wand 60', 0 and 240 wands 59' 11" Borehole X2 Borehole length: 100' Date of investigation by COLOG: 9/13/89 Date of investigation by author: 1/25/96 Single-hole pneumatic test data: 3-/n nominal scale {Rasmussen et al., 1990); \-m and 2-m nominal scale (Guzman et al., 1996) 290 Downhole Features large fracture at 7.0' fracture at 7'5.0" fractures at 8'7.3" fractures at 9'0.8" several fractures at 12'4.9" break out zone - large fractures 13' - - 120 wand 12' 10", 240 wand 18'5.0" hairline fractures 14' possible hariline fracture 17'1.9" fracture with 0 wand 18'1.9", 120 wand 18'8.6", 240 wand 18'5.0" mineralized zone at 22' fracture at 33' hairline fracture at 37'6.8" series of drilling rings 41' unknown feature 46'2.0" unknown feature 46'7.75" fracture? series of drilling rings 47' fracture 50.0' unknown feature 52'10.5" small fracture 55'8.0" opening (cavity) 55' 11.0" hairline fracture 57' hairline fracture 58'3.0" hairline fracture 58'8.8" two possible hairline fractures near 61' level fracture with 0 wand 63'6.6", 120 wand 62'7.75", 240 wand 63' 1.8" drilling ring at 64' level possible fracture at 66'2.25" small fracture at 66'9.1" medium size fracture with 240 wand 75' 11.25" two small fractures at 76'4.5" fracture at 77'2.0" relatively rough surface 77' - 92' series of drilling rings between 91' and 92' Borehole X3 Borehole length: 150' Date of investigation by COLOG; 9/14/89 Date of investigation by author: 1/26/96 Single-hole pneumatic test data: 3-/w nominal scale {Rasmussen et al., 1990) 291 Downhole Features fractures at 5.78' fracture at 6.87' fi:acture at 7.71' hairline fractures at 8' high angle fracture at 15.46' drilling rings at 16' fracture at 16.55' fracture at 22.2' possible fracture at 23.7' possible hairline fracture at 24' mineralized zone at 25' bands of different mineralogy possible hairline fracture at 26' hairline fracture at 28' several in this area fracture at 29.2' Owand 29.37', 120 wand 28.84' possible fracture at 35.37' drilling ring at 36' series of 6 to 7 fracture at 37.38' fracture at 38.12' drilling ring at 42' fracture at 43.2' fracture with 0 wand 44', 120 wand 44.54', 240 wand 44.07' fracture with 240 wand at 52'0.25" drilling ring at 55' fracture Avith 0 wand 60'8.75", 120 wand 61' 1.6", 240 wand 61'!" hairline fractures just passed below 62' level possible drilling rings? above 68' fracture with 0 wand 69'4.7", 120 wand 69'6.0", 240 wand 69' fracture at 84'5.75" drilling ring at 87' large fracture 105'5.8" part of last fracture set - multiple and intersect large opening at 105'0.2" possible fracture at 111 '3.0" fracture at 114'0.2" fracture at 114'3.0" fracture at 115'7.0" (could be part of fracture at 117'0.5") fracture at 117'7.5" hairline fracture at 125'9.5" drilling ring at 130' fracture at 138'1.8" small fractures at 142' (intersecting each other on borehole face) fracture at 143' 11.0" fracture at 147'5.6" 292 bottom 150' Borehole Y1 Borehole length: 56.5' Date of investigation by COLOG: 9/13/89 Date of investigation by author: 1/25/96 Single-hole pneumatic test data: 3-/ii nominal scale {Rasmussen et al., 1990) Downhole Features casing 4.5' fracture at 7'3.75" two small hairline fractures at 10'3.2" fracture at higher angle 0 wand 10'11.4" 120 wand 10'8.3" 240 wand 10'9.5" fracture all wands at same level H'1.5" fracture will all wands at the same level 14'3.5" hairline fracture 15'2.5" filled fracture' 18'6.7" fracture 0 wand 20'7.75", 120 wand 20'10.75", 240 wand 20'8.4" unknown feature at 24'2.0" (possibly a fracture?) hairline fracture 28'7.3" hairline fracture 29' 1.0" drilling ring 30.0' drilling ring 35.0' drilling ring 45.0' fractureOwand 48'5.6", 120 wand 48'10.1", 240 wand 48'9.1" relatively smooth borehole surface from 49' to the bottom of the hole 56'6" Borehole Y2 Borehole length: 100.5' Date of investigation by COLOG: 7/15/93 Date of investigation by author: 1/24/96 Single-hole pneumatic test data: 3-m nominal scale {Rasmussen et ai, 1990); 0.5, 1, 2, and 3-m nominal scale; 20-m nominal scale {Illman et al., 1998) casing depth: 4.7' Downhole Features rough surface with some fractures 6-10', 10' some fractures filled with some minerals (silica?) 11-15' cavities with minerals, fracture at 12.2' (horizontal - high angle in reality because borehole is slanted 45 deg.) 18.2' filled fracture 21-30' no major feature: relatively smooth surface 293 30.8' series of small near horizontal fractures 31-35' relatively smooth 42.7' single, near horizontal fracture 46-50' relatively smooth borehole surface 52.4' massive cavitation (high k zone - undetermined number of fractures) 52.4 - 52.9' smooth surface 53-55' seris of fractures 58.5' high angle fracture 58.8' high angle fracture 64.1' near horizontal fracture 71' relatively smooth surface with some hairline fractures 76 - 80' relatively smooth surface 81-85' sealed high angle fracture 92.5 - 93' horizontal and near vertical fractures 96 - 100' zone moderate angle fracture 100' Borehole Y3 Borehole length; 150' Date of investigation by COLOG: 9/13/89 Date of investigation by author: 1/25/96 Single-hole pneumatic test data: 3-m nominal scale {Rasmussen et al., 1990) Downhole Features fracture at 8.59' fracture with 120 wand 9.13', 240 wand 9.44' fracture at 10.8' zone of several fractures at 14.58' several fractures at 16.03' fracture at 16.47' fracture at 18.4' hairline fracture at 21.43' fracture at 23.41' minral bands at 24' layer of tuff with different lithology unknown feature at 28' (drilling ring or fracture) possible fracture at 28.45' fracture at 31.44' fracture at 33.4' fracture at 34.49' fracture at 38.45' cavity at 39.20' - break in wall: lithologic weathering? vug? fracture 40.5' (measurements above this level are in feet only - no inches) fracture at 43'7.5" break in wall (cavitation) 47'4.25" 294 break in fracture wall?? 50'1.2" fracture at 52'5.0" fracture with 0 wand 53'3.8", 120 wand? 53'2.25", 240 wand 52'9.50" fracture at 52'5.0" fracture with 0 wand 57'9.3", 120 wand 57'7.5", 240 wand 57'4.0" fracture at 60' fracture at 0 wand 65'6.8". 120 wand 65'9.75". 240 wand 65'3.5" fracture with 120 wand 67'5.2", 240 wand 66'7.0" - 2nd fracture with 0 wand 67'3.5", 120 wand 67' fracture with 240 wand at 77' with lower limit of fracture at 76'8.5" fracture with 0 wand 78'8.8", 120 wand 78'6.5", 240 wand 78' 1.0" fracture with 0 wand 81'!.0", 120 wand 81240 wand 80' 1.0" possible fracture at 86' 10.0" possible fracture at 104'8.25" drilling ring at 109' drilling ring at 119'2.6" possible fracture at 129' due to twisted tape all depths should be 7.0ft shallower - depths adjusted already for this document Borehole Z1 Borehole length; 55.0' Date of investigation by COLOG: 9/13/89 Date of investigation by author; 1/26/96 Single-hole pneumatic test data; 3-m nominal scale (Rasmussen et al., 1990) Downhole Features fracture with 240 wand 7'6.5" fracture with 240 wand 8'8.25" cavitation at 11 '3.5" (break out zone) several fractures near surface of borehole 5' - 10' hairline fractures 12' high angle fracture 12'8.7" hairline fractures 14' - 15' fracture with 0 wand 16'9.5", 120 wand 17'0.0", 240 wand 17'0.25" several hairline fractures 17'3.3" possible fracture 18'3.25" cavity? at 21' fracture with 240 wand 25' 1.9" fracture at 25'5.9" possible fracture at 28* 1.5" possible fracture at 28'5.3" 295 fracture with 240 wand 30'2.0" fiacture with 0 wand 31*1.5", 120 and 240 wands at 31 '4.2" fracture with 120 wand 31'6.5", 240 wand 31 '8.2" - note another fracture - total of 3? borehole breakoff at 32' fracture with 0 wand 32'2.0" small fracture 34' fracture with 0 wand 34'6.75", 120 wand 34*10.25", 240 wand 34*8.6" fracture with 0 wand 36*2.0*', 120 wand 36*6.0*', 240 wand 36'3.25" hairline fracture at 37*5.25** hairline fracture at 39*0.5" fracture with 0 wand? 40*4.9**, 120 and 240 wands at 40*1.0" possible fracture at 40*5.5** fracture with 240 wand 41* high angle fracture at 41*7.5*' fracture with 120 wand 42*0.5", 240 wand 41*10.6" small fracture or a small cavity? at 42*7.3*' small fracture at 45*0.5** fracture with 0 wand 48*7.35", 120 wand 48*2.75", 240 wand48'5.1" fracture with 120 wand 50*3.7** fracture at 51*10.6" fracture with 0 wand 54*1.5", 120 wand 54'3.2", 240 wand 54'5.6" Borehole Z2 Borehole length; 100.0* Date of investigation by COLOG: 9/13/96; also 7/15/93 (black and white) Date of investigation by author; 1/25/96 Single-hole pneumatic test data; 3-/n nominal scale {Rasmussen et al., 1990); \-m nominal scale (Guzman et al.., 1996) Downhole Features fracture at 20*3.75** hairline fractures above 26* level possible fracture at 27* 1.0** fracture at 30*6.8** possible hairline fractures at 32* fracture with 0 wand 39*, 120 wand 39*2.5**, 240 wand 39*4.3** hairline fracture at 40* possible two drilling rings slightly above the 41* level fracture at 41*6.2** fracture zone at roughly 42.5* to 45' (approximately 20 fractures - counted) uneven (ragged) fracture zone 44*5.4" several intersecting fractures ~ end of zone at 44* - note high angle fracture 296 intersecting fractures wand pos. 0 wand 47'5.7", 120 wand 47'4.25", 240 wand 47'7.25" large fracture 0 wand 47'9.8", 120 wand 47'10.25", 240 wand 47'7.25" fracture 48'9.5" (horizontal - with respect to borehole) fracture with 0 wand?? 48'2.3", 120 wand 48'5.7", 240 wand 48'3.25" fracture at 50*6.35" fracture at 51' fracture at 5r4.5" fracture with 0 wand 51*8.5", 120 wand 5 I'll.2", 240 wand 51*9.5" - same level with 120 wand of second fracture hairline fracture 52*8.25** hairline fracture 54*3.0** hairline fracture 55*6.5** hairline fracture 56*0.85** hairline fracture 59*2.79** at 60*8.8** level several fractures comprise a broken (cavity/cavernous) zone fracture with 0 and 240 wands 62*6.25*' and 120 wand 62*7.75" hairline fracture 63*7.25** hairline fracture 65*7.7** drilling rings at 71* top of intersecting fracture at 73*5.75** - large fracture with an estimated aperture of quarter of an inch intersecting another fracture large fracture 0 wand 74*1.5**, 120 wand 74*4.75**, 240 wand 74*0.6** at 75' level a possible third fracture fracture with 120 wand 75'5.0", 240 wand 75'4.25" drilling ring at 81' hairline fracture at 86* hairline fracture at 91 * 1.7* * large fracture with 0 wand 91 * 10.25**, 120 wand 91* 11.25**, 240 wand 92*2.0** fracture with 0 wand 95*1.5**, 120 wand 95*0.0**, 240 wand ??? fracture with 0 wand ?? 97'1.1", 120 wand 97'4.0", 240 wand 97*5.5** Borehole Z3 Borehole length: 149.0* Date of investigation by COLOG: 9/14/89 Date of investigation by author: 1/26/96 Single-hole pneumatic test data: 3-/w nominal scale (Rasmussen et al., 1990) Downhole Features drilling ring at 6* fracture at 6.8* drilling ring at 8* fracture with 0 wand 31.9*, 120 wand 31.71', 240 wand 31.52' 297 fi^cture 0 wand 36.75', 120 wand 36.8', 240 wand 36.38' fracture at 38.73' drilling ring at 39' drilling ring right below 44' fracture at 51'5.0" fracture and breakout zone (from Y2?) 55'6.0" drilling ring 56' drilling ring and fracture 64'3" fracture at 72'8" fracture (lower) with 120 wand 77'7.0" upper fracture 120 wand 77'3.75" drilling ring at 79.5' fracture at 0 wand and 240 wand approx. same spot 81*9 I" 120 wand 81 '5.75" fracture with 0 wand 85'2.6", 120 wand 84'H", 240 wand 84'7.25" fracture intersecting drilling ring (smearing?) hairline fracture 86'3.75' hairline fracture 86'10.1" large fracture 0 wand 9r2.5", 120 wand 9r4", 240 wand 90'8.9" possible fracture at 93' ( possible mineral band or a filled fracture) drilling ring at 95' fracture with 0 wand 99'1.25", 120 wand 99', 240 wand 98'7.5" drilling ring at 100' drilling ring at 105' fracture intersects drilling ring at 110'3.0" with drilling ring at 110' drilling ring at 115' drilling ring at 120' drilling ring at 125' fracture with 0 wand I26'4.5", 120 wand 126'3.r', 240 wand 126'0.0" drilling ring at 130' fracture with 0 wand and 240 wands at I33'5" 120 wand I33'3.25" drilling ring at 135' drilling ring at 140' drilling ring at 145' fracture at 147'6.7" 298 REFERENCES Aganval, R. G., R. Al-Hussainy, H. J. Ramey Jr., An Investigation of Wellbore Storage and Skin Effect in Unsteady Liquid Flow: I. Analytical Treatment, AIME Trans., 279290, 1970. AI-Hussainy R. and H. J. Ramey Jr., The Flow of Real Gases Through Porous Media, JPT, 624-636, 1966. Alzaydi, A. A., C. A. Moore, and I. S. Rai, Combined Pressure and Difflisional Transition Region Flow of Gases in Porous Media, AlChE Journal, 24(\\ 35—43, 1978. Ando, K., Continuum stochastic modeling of flow and transport in a crystalline rock mass, M.S. thesis. The University of Arizona, Tucson, 1995. Aronofsky, J. S. and Jenkins, Unsteady Flow of Gas Through Porous Media; OneDimensional Case, Proceedings of the First U.S. National Congress of Applied Mechanics, American Society of Mechanical Engineers, New York, 1952. Aronofsky, J. S., Effect of Gas Slip on Unsteady Flow of Gas Through Porous Media, J. App. Phy., 25(1), 48-53, 1954 Aronofsky, J. S. and R. Jenkins, A Simplified Analysis of Unsteady Radial Gas Flow, Tram., AIME, 201, 149, 1954. Aronofsky, J. S. and O. D. Ferris, Transient Flow of Non-ideal Gases in Porous MediaOne Dimensional Case, J. AppL Phy., 25, 289, 1954. Aronofsky, J. S. and J. D. Porter, Unsteady Radial Flow of Gas Through Porous Media: Variable Viscosity and Compressibility, J. AppL Mech., 23, 128, 1956. Baehr, A. L., and M. C. Hult, M.C., Evaluation of Unsaturated Zone Air Permeability Through Pneumatic Tests, Water Resour. Res., 27(10), 2605-2617, 1991. Baehr, A. L. and C. J. Joss, An updated model of induced airflow in the unsaturated zone. Water Resour. Res., 31(2), 417-421, 1995. Barenblatt, G. I., Y. P. Zheltov, and I. N. Kochina, Basic concepts in the theory of seepage of homogeneous liquids in fissured rocks, J. AppL Math. Mech., English Transl., 24, 1286-1303, 1960. Bassett, R. L., S. P. Neuman, T. C. Rasmussen, A. Guzman, G. R. Davidson, and C. L. Lohrstorfer, Validation Studies for Assessing Unsaturated Flow and Transport Through Fractured Rock, NUREG/CR-6203, 1994. 299 Bassett, R. L., S. P. Neuman, P. J. Wierenga, G. Chen, G. R. Davidson, E. L. Hardin, W.A. niman, M. T. Murrell, D. M Stephens, M. J. Thomasson, D. L. Thompson, E. G. Woodhouse, Data Collection and Field Experiments at the Apache Leap Research Site, May 1995-1996, NUREG/CR-6497, 1997. Bear, J, Dynamics of Fluids in Porous Media, Dover Publications, Inc., 764 pp., 1972. Beckett, G. D. and D. Huntley, Characterization of Flow Parameters Controlling Soil Vapor Extraction, Groundwater, 52(2), 239-247, 1994. Benito, P. H., B. Faybishenko, and C. Doughty, Field study of pneumatic connectivity in unsaturated fractured basalt using cross-borehole interference injection tests. Fall AGU meeting poster session, H42B-07, San Francisco, CA, 1998. Berkowitz, B., J. Bear, and C. Braester, Continuum models for contaminant transport in fractured porous formations. Water Resour. Res , 24{^), 1225-1236, 1988. Bibby, R., Mass transport of solutes in dual-porosity media. Water Resour. Res., 17(A), 1075-1081, 1981. Birkholzer, J., H. Rubin, H. Daniels, and G. Rouve, Contaminant advection and spreading in a fractured permeable formation, 2, Numerical simulation, J. Hydrology, 144, 35-38, 1993. Black J. H. and K. L. Kipp Jr., Observation Well Response Time And Its Effect Upon Aquifer Test Results, J. Hydrology, 34: 297-306, 1977. Blasingame, T. A., J. L. Johnston, and W. J. Lee, Type Curve Analysis Using the Pressure Integral Method, Paper SPE 18799. 1989. Boardman C. R. and J. Skrove, Distribution in Fracture Permeability of a Granitic Rock Mass Following a Contained Nuclear Explosion, JPT, 619-623, 1966. Bourdet, D., J. A. Ayoub, and Y. M. Pirard, Use of Pressure Derivative in Well-Test Interpretation, SPEFE, 293-302, 1989. Bruce, G. H., D. W. Peaceman, H. H. Rachford, Jr., and J. D. Rice, Calculations of Unsteady-State Gas Flow Through Porous Media, 198, 79-92, 1953. Burcik, E. J., Properties of Petroleum Reservoir Fluids, International Human Resources Development Coorporation, 1957. 300 Butler Jr., J. J. and J. M. Healy, Relationship Between Pumping-Test and Slug-Test Parameters: Scale Effect of Artifact?, Groundwater, 36(2), March-April, 305-313, 1998 Cacas, M. C., E. Ledoux, G. de Marsily, A. Barbreau, P. Calmels, B. Gaillard, and R. Margritta, Modeling fracture flow with a stochastic discrete fracture network; Calibration and validation, I, The flow model. Water Resour. Res., 26(2), 479-489, 1990a. Cacas, M. C., E. Ledoux, G. de Marsily, A. Barbreau, P. Calmels, B. Gaillard, and R. Margritta, Modeling fracture flow with a stochastic discrete fracture network: Calibration and validation, 2, The transport model. Water Resour. Res., 26(3), 491-500, 1990b. Carrera, J., J. Heredia, S. Vomvoris, and P. Hufschmied, Modeling of flow in a small fractured monzonitic gneiss block, pp. 115-167 in Neuman, S.P. and I. Neretnieks, Hydrogeology of Law Permeability Environments, Hydrogeology, Selected Papers, 2, Verla Heinz Heise, Hannover, Germany, 1990. Carter, R. D., Performance Predictions for Gas Reservoirs Considering Two-Dimensional Unsteady-State Flow, .SPE/, 35-44, 1966. Carslaw, H. S. and J. C. Jaeger, Conduction of Heat in Solids, 2"** Ed., Oxford Science Publications, 510 pp., 1959. Chen G, S. P. Neuman, and P. J. Wierenga, 4. Infiltration tests in Fractured Porous Tuffs at the ALRS, in Bassett, R L., S. P. Neuman, P. J. Wierenga, G. Chen, G. R. Davidson, E. L. Hardin, W. A. Illman, M. T. Murrell, D. M Stephens, M. J. Thomasson, D.L. Thompson, E.G. Woodhouse, Data Collection and Field Experiments at the Apache Leap Research Site, May 1995-1996, NUREG/CR-6497, 1997. Chuang, Y., W. R. Haldeman, T. C. Rasmussen, and D. D. Evans, Laboratory Analysis of Fluid Flow and Solute Transport Through a Variably Saturated Fracture Embedded in Porous Tuff, NUREG/CR-5482, 328 pp., 1990. Cinco-Ley, H., H. J. Ramey, Jr., and F. G. Miller, Unsteady-State Pressure Distribution Created by a Well With an Inclined Fracture, paper SPE 5591 presented at the SPEAIME 50 Annual Fall Technical Conference and Exhibition, Dallas, Sept. 28-Oct. I, 1975. Cinco-Ley, H., Samaniego-V. F., and Dominguez, A. N., Transient Pressure Behavior for a Well with a Finite Conductivity Vertical Fracture, SPEJ, 253-264, 1978. Cook, P. J. and J. S. Y. Wang, Air permeability studies of Fractured Rock in Niches and Alcoves at the Yucca Mountain Sites, Fall AGU meeting poster session, H42B-08, San Francisco, CA, 1998. 301 Cornell, D. and D. L. Katz, Pressure Gradients in Natural Gas Reservoirs, Trans., AIME 198, 6\, 1953. CRC, Handbook of Chemistry and Physics, David R. Lide (editor-in-chief), 73"* ed., CRC Press, 1992-1993. Crump, K. S., Numerical Inversion of Laplace Transforms Using a Fourier Series Approximation, y. Ass. Comp. Mach., 23(1), 89-96, 1976. Culham, W. E., Pressure Buildup Equations for Spherical Flow Regime Problems, SPEJ, 545-555, 1974. Dachler, R., Gnmdwasserstromung (Flow of Groundwater), Julius Springer, Wien, 141 pp., 1936. De Hoog F. R., Knight, and A. N. Stokes, An Improved Method for Numerical Inversion ofLaplace Transforms, SIAMJ. Sci. Stat. Camp., 5(3), 357 - 366, 1982. Dershowitz, W., P. Wallman, and S. Kindred, Discrete Fracture Modeling for the Stripa Site Characterization and Validation Drift Inflow Predictions, Stripa Project Technical Report 91-16, SKB Swedish Nuclear Fuel and Waste Management Co., Stockholm, Sweden, 1991. DePaoli, D. W., J. H. Wilson, and C. O. Thomas, Conceptual Design of Soil Venting Systems, J. Env. Eng., 122(5), 399-406, 1996. Douglas, Jr. J., D. W. Peaceman, H. H. Rachford, Jr., Calculation of Unsteady-State Gas Flow Within a Square Drainage Area, Petro. Trans., AIME, 204, 190-195, 1955. Duguid, J. O. and P. C. Y. Lee, Flow in fractured porous media. Water Resour. Res., 13, 558-566, 1977. Dverstorp, B., J. Andersson, and W. Nordqvist, Discrete fracture network interpretation of field tracer migration in sparsely fractured rock, iVater Resour. Res., 28(9), 2327-2343, 1992. Dykhuizen, R. C., A new coupling term for dual porosity models. Water Resour. Res., 26, 351-356, 1990. Edwards, K. B., Air Permeability From Pneumatic Tests in Oxidized Till, J. Env. Eng., 120(1), 329-347, 1994. Edwards, K. B., Venting Test Analysis Using Jacob's Approximation, J. Env. Eng., 722(3), 232-234, 1996 302 Energy Resources Conservation Board, Theory and Practice Of The Testing Of Gas Wells, 3"* ed., 1975. Evans, D. D., Unsaturated Flow and Transport Through Fractured Rock - Related to High-Level Waste Repositories, NUREG/CR-3206, 231 pp., 1983. Evans, D. D. and T. C. Rasmussen, Unsaturated Flow and Transport Through Fractured Rock Related to high-Level Waste Repositories, Final Report - Phase III, NUREG/CR5581, pp. 75, 1991. Falta, R. W., Analytical Solutions for Gas Flow Due to Gas Injection and Extraction from Horizontal Wells, Groundwater, 33(2), 235-246, 1995. Ge, S. and M. Liao, An analytical evaluation for airflow to inlet wells in vapor extraction systems under leaky conditions. Water Resour. Res., 32(3), 743-748, 1996. Ge S., A New Analytical Treatment of Airflow to Inlet Wells in Soil Vapor Extraction Systems, GWMR, 133-136, 1998 Gerke, H. H. and M. T. van Genuchten, A dual-porosity model for simulating the preferential movement of water and solutes in structured porous media. Water Resour. Res., 29, 305-319, 1993a. Gerke, H. H. and M. T. van Genuchten, Evaluation of a first-order water transfer term for variably saturated dual-porosity models. Water Resour. Res., 29, 1225-1238, 1993b. Glover, R. E., Flow from a Test-Hole Above Groundwater level, in Theory and problems of Water Percolation, U.S. Bureau of Reclamation Engineering Monograph No. 8, 66-71, 1953. Gradshteyn, I. S. and I. M. Ryzhik, Table of Integrals, Series, and Products, Academic Press, Inc., 1204 pp., 1965. Green, R. T. and D. D. Evans, Radionuclide Transport as Vapor Through Unsaturated Fractured Rock, NUREG/CR-4654, 163 pp., 1987. Green L. and C. H. Wilts, Nonsteady Flow of Gas Through a Porous Wall, Proc., First U.S. National Congress Applied Mechanics, J. E. Edwards, Inc., Ann Arbor, Mich. 777, 1962. Gringarten, A.C., The Use of Source and Green's Functions in the Solution of Unstecufy Flow Problems in Reservoirs, Pub. ITTE 71-9, Dept. of Civil Engineering, U. of California at Berkeley, 1971. 303 Gringartea, A. C., and H. J. Ramey Jr., The Use of Source and Green's Function in Solving Unsteady-Flow Problems in Reservoirs, Soc. Pet. Eng. J., 285-296, 1973. Gringarten, A. C., H. J. Ramey, Jr., and R. Raghavan, Unsteady-State Pressure Distributions Created by a Well With a Single Infinite-Conductivity Vertical Fracture. SPEJ, 347-360; Trans., AIME, 257, 1974. Gringarten, A. C. and H. J. Ramey Jr., Unsteady-state pressure distributions created by a well with a single horizontal fracture, partial penetration, or restricted entry. SPEJ, 413426; Trans., AIME, 257., 1974. Gringarten, A C., Flow Test Evaluation of Fractured Reservoirs, Symp. On Recent Trends in Hydrology, Berkeley, CA, February 8-11, 1979. Gringarten, A. C., Flow Test Evaluation of Fractured Reservoirs, Geological Society of America, Special Paper 189, 237-263, 1982. Guo, W., Analytical solution of transient radial air flow to an extraction well, J. Hydrology, 191, 1-14, 1997. Guzman, A. G. Jind S. P. Neuman, Field Air Injection Experiments, pp. 52-94 in Rasmussen, T. C., S. C. Rhodes, A. Guzman, and S. P. Neuman, Apache Leap Tuff INTRAVAL Experiments: Results and Lessons Learned, NUREG/CR-6096, 1996. Guzman, A. G., S. P. Neuman, C. Lohrstorfer, and R. Bassett, pp. 4-1 through 4-58 in Bassett, R. L., S. P. Neuman, T. C. Rasmussen, A. G. Guzman, G. R. Davidson, and C. L. Lohrstorfer, Validation Studies for Assessing Unsaturated Flow and Transport Through Fractured Rock, WS^GICK-6102, 1994. Guzman, A. G., A. M. Geddis, M. J. Henrich, C. Lohrstorfer, and S. P. Neuman, Summary of Air Permeability Data From Single-Hole Injection Tests in Unsaturated Fractured Tuffs at the Apache Leap Research Site: Results of Steady-State Test NUREG/CR-6360, 1996. Hantush, M. S., Wells in homogeneous anisotropic aquifers. Water Resour. Res., 2(2), 273-279, 1966a. Hantush, M. S., Analysis of data from pumping tests in anisotropic aquifers, J. Geophys. Res., 7/(2), 421-426, 1966b. Hantush, M. S., and R. G. Thomas, A method for analyzing a drawdown test in anisotropic aquifers. Water Resour. Res., 2(2), 281-285, 1966. 304 Harleman, D. R. E., P. F. Mehlhom, and R. R. Rumer, Dispersion-permeability correlation in porous media, J. Hydrau. Div. Amer. Soc. Civil Eng., No. HY2, 89, 67-85 1963. Hetherington, C. R., D. T. MacRoberts, and R. L. Huntington, Unsteady Flow of Gas Through Porous Media, Trans., AIME, 146, 166-174, 1942 Hildebrand, Advanced Calculus for Applications, 2™* ed. Prentice Hall Inc, 1948. Hsieh, P. A., S. P. Neuman, and E. S. Simpson, Pressure Testing of Fractured Rocks - A Methodology Employing Three-Dimensional Cross-Hole Tests - Topical Report, NUREG/CR-3213, 1983. Hsieh, P. A. and S. P. Neuman, Field Determination of the Three-Dimensional Hydraulic Conductivity Tensor of Anisotropic Media, 1. Theory, Water Resour. Res., 2/(11): 16551665, 1985a. Huang, C. and D. D. Evans, A 3-Dimensional Computer Model to Simulate Fluid Flaw and Contaminant Transport Through a Rock Fracture System, NUREG/CR-4042, 109 pp., 1985. Hvorslev J., Time Lag and Soil Permeability in Ground-Water Observations, Bulletin No, 36, Waterways Experiment Station, Corps of Engineers, U.S. Army, 50 pp., 1951. Illman, W. A., Neutron Probe Logging and Calibration - A Progress Report on the Theory, Methodology, Calibration and its Application to Apache Leap Tuff, internal report submitted to Drs. S. P. Neuman and P. J. Wierenga, Department of Hydrology and Water Resources, July 22, 1996. Illman, W. A., D. L. Thompson, G. Chen, and S. P. Neuman, Pneumatic Testing of Fractured Porous Tuffs at the ALRS (Subtask 1.1), p. 3-1 through 3-43 in, Bassett, R. L., S. P. Neuman, P. J. Wierenga, G. Chen, G. R. Davidson, E. L. Hardin, W. A. Illman, M. T. Murrell, D. M. Stephens, M. J. Thomasson, D. L. Thompson, and E. G. Woodhouse, Data Collection and Field Experiments at the Apache Leap Research Site - A/qy 19951996, NUREG/CR-6497, 1997. Illman W. A., D. L. Thompson, V. V. Vesselinov, G. Chen, and S. P. Neuman, Singleand Cross-Hole Pneumatic Tests in Unsaturated Fractured Tuffs at the Apache Leap Research Site: Phenomenology, Spatial Variability, Connectivity and Scale, NUREG/CR-5559, 188 pp., 1998. Jenkins, R., and J. S. Aronofsky, Unsteady Radial Flow of Gas Through Porous Media, J. ofApp. Mech, A.S.M.E. 20(2), 1953. 305 Johnson, P. C., C. C. Stanley, M. W. Kemblowski, D. L. Byers, and J. D. Colthart, A Practical Approach to the Design, Operation, and Monitoring of In Situ Soil-Venting Systems, GmiR, 70(2), 159-178, 1990. Joseph, J. A., and L. F. Koederitz, Unsteady-State Spherical Flow With Storage and Skin, SPEJ, pp.804 - 822, 1985. Joumel, A. G., Nonparametric Estimation of Spatial Distributions, Math. Geol., 15(3), 445-468, 1983. Joumel, A. G., Fundamentals of geostatistics in five lessons. Short course in geology, 8, Am. Geophys. Union, Washington, D.C., 1989. Kaluarachchi, J. J., Analytical solution to two-dimensional axisymmetric gas flow with Klinkenberg effect, J. Em. Eng., 72/(5), 417-420, 1995 Katz, D. L., D. Cornell, R. Kobayashi, F. Poettman, J. A. Vary, J. R. Elenbaas, and C. F. Weinaug, Handbook of Natural Gas Engineering, Mc-Graw-Hill Book Co. Inc., New York, N.Y, 802 pp., 1959. Kazemi, H., Pressure transient analysis of naturally fractured reservoirs with uniform fracture distribution. Trans. Sac. Petrol. Engin. AIME, 246, 451-462, 1969. Kidder R. E., Unsteady Flow of Gas Through a Semi-Infinite Porous Medium, J. App. Mech., 24, 329, 1957. Kirchhoff, G, Vorlesungen tiber de theorie der Warme, Barth, Leipzig, 1894. Klinkenberg, L. J., The permeability of porous media to liquids and gases. Am. Pet. Inst., Drilling and Production Practice, 200-213, 1941. Knudsen, M., The Kinetic Theory of Gases, Methuen, London, 1950. Kostner, A., Geostatistical and numerical analysis of flow in a crystalline rock mass, M.S. thesis. The University of Arizona, Tucson, 1993. Lane, H. S., W. J. Lee, A. T. Watson, An Algorithm for Determining Smooth, Continuous Pressure Derivatives From Well-Test Data, SPEFE, 493-499, 1991. LeCain, G. D., Pneumatic testing in 45-degree-inclined boreholes in ash-flow tuff near Superior, Arizona, U.S. Geological Survey Water-Resources Investigations Report 954073, 1995. 306 LeCain, G. D., Air-injection testing in vertical boreholes in welded and nonwelded tuf^ Yucca Mountain, Nevada, U.S. Geological Survey Milestone Report 3GUP610M, Denver, Colorado, 1996a. LeCain, G. D. and J. N. Walker, Results of air-permeability testing in a vertical borehole at Yucca Mountain, Nevada, Radioactive Waste Management, 2782-2788, 1996b. Le Cain, G. D., Results from Air-Injection and Tracer Testing in the Upper Tiva Canyon, Bow Ridge Fault, and Upper Paintbrush Contact Alcoves of the Exploratory Studies Facility, August 1994 through July 1996, Yucca Mountain, Nevada, Water-Resources Investigations Report 98-4058, 28 pp., 1998. Loo, W. W., K. Frantz, and G. R. Holzhausen, The application of telemetry to large-scale horizontal anisotropic permeability determinations by surface tiltmeter survey, GWMR, ^(4), 124-130, 1984. Louis, C., Introduction a I'Hydraidique des Roches (in French), Bureau de Recherches Geologiques et Minieres, Orleans, France, 1974. Massmann, J. W., Applying Groundwater Flow Models In Vapor Extraction System Design,/. Env. Eng., 775(1), 129-149, 1989. Massmann, J. W. and M. Madden, Estimating Air Conductivity And Porosity From Vadose-Zone Pumping Tests, J. Env. Eng., 720(2), 313-327, 1994. Matthews, C. S., 1961, Analysis of Pressure Buildup and Flow Data, JPT, pp. 862-870, 1961. MacRoberts, D.T., Effect of Transient Conditions in Gas Reservoirs, Trans., AIME 186, 36-38, 1949. McTigue D. F., C. L. Stein, W. A. Illman, M. J. Martinez, and T. D. Burford, Heat and Vapor Transport Beneath and Impermeable Barrier, Abstract and talk presented at the 1993 Fall Meeting, San Francisco, CA, 1993 McWhorter, D. B., Unsteady Radial Flow of Gas in the Vadose Zone, J. Cont. Hyd., 5, 297-314, 1990. Mendoza, C. A., and E. O. Frind, Advective-dispersive transport of dense organic vapors in the unsaturated zone, I, Model Development, Water Resour. Res., 26(3), 379-387, 1990. 307 Mishra, S., G. S. Bodvarsson, and M. P. Attanayake, Injection and FallofFTest Analysis to Estimate Properties of Unsaturated Fractured Fractures, in Flow and Transport Through Unsaturated Fractured Rock, eds. D. D. Evans and T.J. Nicholson, American Goephysical Union, Geophysical Monograph 42, 1987. Moench, A. F., Double-porosity models for a fissured groundwater reservoir with fi^cture skin. Water Resour. Res., 20, 831-846, 1984. Montazer, P. M., Permeability of Unsaturated, Fractured Metamorphic Rocks Near An Underground Opening, Ph. D thesis, Colorado School of Mines, Golden, Colorado, 1982. Morelli G. and D. J. LaBrecque, Advances in ERT Inverse modeling. Euro. J. Env. Eng. Geoph., /, 171-186, 1996. Muskat M., The Flow of Compressible Fluids Through Porous Media and Some Problems in Heat Conduction, Physics, 5, 71-94, 1934. Muskat, M., TJie Flow of Homogeneous Fluids, J.E. Edwards, Inc., Ann Arbor, Mich., 1946. National Research Council, Rock Fractures and Fluid Flow: Contemporary Understanding and Applications, National Academy Press, Washington, D C., 1996. Moran, J. H. and E. E. Finklea, Theoretical Analysis of Pressure Phenomena Associated with the Wireline Formation Tester, JPT, 899-908, 1962. Neuman, S. P., Analysis of pumping test data fi-om anisotropic unconfmed aquifers considering delayed gravity response. Water Resour. Res., //(2), 329-342, 1975. Neuman, S. P., Stochastic continuum representation of fi-actured rock permeability as an alternative to the REV and fi-acture network concepts. In; I.W. Farmer, J.J.K. Dalmen, C.S.Desai, C.E.Glass and S.P. Neuman (Editors), Rock Mechanics, Proc. Of the 28^^ U.S. Symposium. Balkema, Rotterdam, pp.533-561, 1987. Neuman, S. P., A proposed conceptual fi-amework and methodology for investigating flow and transport in Swedish crystalline rocks, Arbetsrapport 88-37, 39 pp., SKB Swedish Nuclear Fuel and Waste Management Co., Stockholm, September 1988. Neuman, S. P. and J. S. Depner, Use of Variable-Scale Pressure Test Data to Estimate the Log Hydraulic Conductivity Covariance and Dispersivity of Fractured Granites Near Oracle, Arizona, J. Hydrology, 102, 475-501, 1988. 308 Neuman, S. P. and D. A. Gardner, Determination of Aquitard/aquiclude hydraulic properties from arbitrary water level fluctuations by deconvolution. Groundwater, 27(1), 66-76, 1989. Odeh, A. A., Unsteady-state behavior of naturally fractured reservoirs, SPEJ, 60-66, 1965. Onur, M., and A Reynolds, A New Approach for Constructing Derivative Type Curves for Well Test Analysis, 5!PEFj?, 197-206, 1988. Papadopulos, I. S., Nonsteady flow to a well in an infinite anisotropic aquifer, in Proceedings of the Dubrovnik Symposium on Hydrology of Fractured Rocks, pp.21-31, International Association of Scientific Hydrology, Debrovnik, Yugoslavia, 1965. Papadopulos, I. S. and Cooper Jr., H. H., Drawdown in a Well of Large Diameter, Water Resour. Res., i(l); 241-244, 1967. Peters, R. R. and E. A. Klavetter, A continuum model for water movement in an unsaturated fractured rock mass. Water Resour. Res., 2-/, 416-430, 1988. Peterson, D. W., Dacitic Ash-Flow Sheet Near Superior and Globe, Arizona, Ph.D. Dissertation, Stanford University, Palo Alto, CA, 1961. Philip, J. R., Approximate Analysis of the Borehole Permeameter in Unsaturated Soil, Water Resour. Res., 21(1), 1025-1033, 1985. Present, R. D., Kinetic Theory of Gases, McGraw-Hill Book Company, Inc., 280 pp., 1958. Press, W. H., S. A. Teukolsky, W. T. Vetterling, B. P. Flannery, Numerical Recipes in FORTRAN - The Art of Scientific Computing, 2 ed., 963 pp., 1992. Pruess, K. and T. N. Narasimhan, A practical method to modeling fluid and heat flow in fractured porous media, SPEJ, 25(1), 14-26, 1985. Pruess, K., J. S. Y. Wang, and Y. W. Tsang, On thermohydrologic conditions near highlevel nuclear waste emplaced in partially saturated fractured tuff: 2. Effective continuum approximation. Water Resour. Res., 26, 1249-1261, 1990. Pruess, K., A general purpose numerical simulator for multiphase fluid and heat flow, Lawrence Berkeley Laboratory Report LBL-29400 UC-251, May, 1991. Raghavan R., A. Uraiet, and G. W. Thomas, Vertical fracture height: Effect on transient flow behavior, SPEJ, 265-277, 1978. 309 Raghavan, R., Well Test Analysis, Prentice Hall Petroleum Engineering Series, PrenticeHall Inc., 558 pp., 1993. Rasmussen, T. C. and D. D. Evans, Unsaturated Flow and Transport Through Fractured Rock-Related to High-level Waste Repositories, NUREG/CR-4655, 474 pp., 1987. Rasmussen T. C. and D. D. Evans, Fluid Flow and Solute Transport Modeling Through Three-Dimensional Networks of Variably Saturated Discrete Fractures, NUREG/CR5239. 193 pp., 1989. Rasmussen, T. C., D. D. Evans, P. J. Sheets, and J. H. Blanford, Unsaturated Fractured Rock Characterization Methods and Data Sets at the Apache Leap Tuff Site, NUREG/CR-5596, 125 pp., 1990. Rasmussen, T. C. and D. D. Evans, Nonisothermal Hydrologic Transport Experimental Plan, NUREG/CR-5880, 41 pp., 1992. Rasmussen, T. C., D. D. Evans, P. J. Sheets, and J. H. Blanford, Permeability of Apache Leap Tuff: Borehole and Core Measurements Using Water and Air, Water Resour. Res., 29(J), 1997-2006, 1993. Rasmussen, T. C., S. C. Rhodes, A. Guzman, and S. P. Neuman, Apache Leap Tuff INTRAVAL Experiments - Results and Lessons Learned, NUREG/CR-6096, 1996. Scheidegger, A.E., The physics of flow through porous media, 3"* ed.. University of Toronto Press, 352 pp., 1974. Schrauf, T. W. and D. D. Evans, Relationship Between the Gas Conductivity arui Geometry of a Natural Fracture, NUREG/CR-3680, 131 pp., 1984. Scott, D. S., and Dullien, F. A. L., AIChEJ., 8, 293, 1962a. Shan, C., Falta, R.W., Javandel, I., Analytical Solutions for Steady State Gas Flow to a Soil Vapor Extraction Well, Water Resour. Res., 28(4), 1105-1120, 1992. SKB, Annual Report 1992, SKB Tech. Rep. 92-46, Swedish Nuclear Fuel and Waste Management Co., Stockholm, Sweden, May 1993. Snow, D. T., Three-hole pressure test for anisotropic foundation permeability, Felsmech. Ingenieurgeol., ^(4), 298-314, 1966. 310 Theis, C. V., The relation between the lowering of the piezometric surface and rate and duration of discharge of a well using groundwater storage. Trans., Amer. Geophy. Union, 2, 519-524, 1935. Tidwell, V. C., T. C. Rasmussen, and D. D. Evans, Saturated hydraulic conductivity estimates for fractured rocks in the unsaturated zone, in Proceedings of International Conference and Workshop on the Validation of Flow and Transport Models for the Unsaturated Zone, edited by P.J. Wierenga, New Mexico State University, Las Cruces, N.M., 1988. Tongpenyai, Y. and R. Raghavan, The effect of wellbore storage and skin on interference test data, ^r, 151-160, 1981. Trauz, R_ C., Rock Fracture Aperture and Gas Conductivity Measurements In-Situ, Topical Report, Contract No. NRC-04-81-224, 1984. Tsang, C. F. and S. P. Neuman, Introduction and general comments on INTRAVAL Phase 2, Working Group 2, Test Cases, in The International Intraval Project, Phase 2, NEA/OEC-D, Paris, France, in press, 1995. Tsang, Y. W., C. F. Tsang, F. V. Hale, and B. Dverstorp, Tracer transport in a stochastic continuum model of fractured media. Water Resour. Res., i2( 10): 3077-3092, 1996. van Everdingen, A. F., and Hurst, W., The Application of the Laplace Transformation to Flow Problems in Reservoirs, Trans. AIME, 186, 305-324, 1949. van Genuchten, M. T., A Closed-Form Equation for Predicting the Hydraulic Conductivity of Unsaturated Soils, Soil Sci. Soc. Am. J., 44, p. 892-898, 1980. Vasserman, A. A., Ya. Z. Kazavchinskii, and V. A. Rabinovich, Thermophysical Properties of Air and Air Components, Academy of Sciences of the USSR, Edited by A.M. Zhuravlev, Translated from Russian by Israel Program for Scientific Translations, Jerusalem 1971, 1966. Vickers, B. C., S. P. Neuman, M. J. Sully, and D. D. Evans, Reconstruction Geostatistical Analysis of Multiscale Fracture Apertures in a Large Block of Welded Tuff, Geoph. Res. Ut., 79(10), 1029-1032, 1992. Vogt, G. T., Porosity, Pore-Size Distribution and Pore Surface Area of the Apache Leap TufifNear Superior, Arizona using Mercury Porosimetry, M.S. Thesis, University of Arizona, 130 pp., 1988 311 Wang, J. S. Y., P. J. Cook, R. C. Trautz, R. Salve, A. L. James, S. Finsterle, T. K . Tokunaga, R. Solbau, J. Clyde, A. L. Flint, and L. E. Flint, Field Testing And Observation of Flow Paths In Niches; Phase I Status Report Of The Drift Seepage Test And Niche Moisture Study, Level 4 milestone SPC314M4 for WBS 1.2.3.3.1.2.4, Yucca Mountain Site Characterization Project, 1998. Warren, J. E. and P. J. Root, The behavior of naturally fractured reservoirs. SPEJ., 3, 245-255, 1963. Warrick, A. W. and A. Rojano, Effects of source cavity shape on steady, threedimensional flow of soil gases, accepted - Water Resour. Res., 1999. Way, S.C., and C.R. McKee, In-situ determination of three-dimensional aquifer permeabilities. Groundwater, 20(5), 594-603, 1982. Weber, D. D. and D. D. Evans, Stable Isotopes of Authigenic Minerals in VariablySaturated Fractured Tuff, NUREG/CR-5255, 1988. Weeks, E.P., Determining the ratio of horizontal to vertical permeability by aquifer-test analysis. Water Ressour. Res., 5{\), 196-214, 1969. Yeh, T. C., T. C. Rasmussen, and D. D. Evans, Simulation of Liquid and Vapor Movement in Unsaturated Fractured Rock at the Apache Leap Tuff Site - Models and Strategies, NUREG/CR-5097, 73 pp., 1988, Zimmerman, R. W., G. Chen, T. Hadgu, and G. S. Bodvarsson, A numerical dualporosity model with semianalytical treatment of fracture/matrix flow. Water Resour. Res., 29(1), 2121-mi, 1993. Zyvoloski, G. A., Z. V. Dash, S. Kelkar, FEHM: Finite Element Heat and Mass Transfer, LA-11224-MS, Los Alamos National Laboratory, 1988. Zyvoloski G. A., B. A. Robinson, Z. V. Dash, L. L. Trease, Users Manual for the FEHMN Application, LA-UR-94-3788, Los Alamos National Laboratory, 1996. Zyvoloski G. A., B. A. Robinson, Z. V. Dash, L. L. Trease, Summary of the Models and Methods for the FEHM Application - A Finite-Element Heat- and Mass-Transfer Code, LA-13307-MS, Los Alamos National Laboratory, 1997. IMAGE EVALUATION TEST TARGET (QA-3) <0 w A C 150mm IIVMGE. Inc 1653 East Main Street Roctiester, NY 14609 USA Phone: 716/482-0300 Fax: 716/288-5989

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement