Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Understanding the effect of freezing on rock mass behaviour as applied to the Cigar Lake mining method Roworth, Megan Rose 2013

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

Item Metadata

Download

Media
24-ubc_2013_fall_roworth_megan.pdf [ 9.68MB ]
Metadata
JSON: 24-1.0073990.json
JSON-LD: 24-1.0073990-ld.json
RDF/XML (Pretty): 24-1.0073990-rdf.xml
RDF/JSON: 24-1.0073990-rdf.json
Turtle: 24-1.0073990-turtle.txt
N-Triples: 24-1.0073990-rdf-ntriples.txt
Original Record: 24-1.0073990-source.json
Full Text
24-1.0073990-fulltext.txt
Citation
24-1.0073990.ris

Full Text

UNDERSTANDING THE EFFECT OF FREEZING ON ROCK MASS BEHAVIOUR AS APPLIED TO THE CIGAR LAKE MINING METHOD  by  Megan Rose Roworth  B.A.Sc., The University of Waterloo, 2005  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in The Faculty of Graduate Studies (Mining Engineering)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  July 2013  © Megan Rose Roworth, 2013  Abstract The objective of this research is to determine how ground freezing affects weak rockmass behaviour with application to the Cigar Lake mine. Cigar Lake mine is a prospective high grade uranium property in northern Saskatchewan where artificial ground freezing will be implemented to support the weak rock associated with the orebody and minimize the potential for a significant water inflow while mining the ore. The deposit comprises a mixture of massive pitchblende, clay and sand and is overlain by thick zones of sandy clay, unconsolidated sand, and altered sandstone. Above and below the orebody, the rockmass shows variations in porosity and permeability due to fracturing and alteration. Artificial ground freezing can be an effective approach to successfully manage and control underground excavations in weak rock mass conditions. Numerous mining and civil projects use artificial freezing worldwide; however, uncertainties remain with respect to understanding and predicting the behavior of frozen rock mass. Previous studies of frozen ground have largely focussed on the behaviour of soil, or in the few studies involving rock, the rock matrix. Of particular interest here is the behaviour of frozen discontinuities present in the weak rock mass and its influence in combination with the matrix on the overall frozen rock mass strength. A comparison of the Cigar Lake mine rockmass and mining operations with that of the McArthur River mine, an unconformity uranium deposit in northern Saskatchewan also utilizing artificial ground freezing will provide the basis for the increase in rockmass quality from unfrozen to frozen conditions. Improving in situ and laboratory characterization methods and developing a better understanding of rock behaviour at sub-zero temperatures is the key focus of this research. A material testing program including unconfined compressive strength, direct shear, and four-point beam experiments was completed using frozen Cigar Lake rock samples. These results are then discussed with respect to the behaviour of the frozen material encompassing the mined out cavities in order to ensure cavity stability during mining. The influence of freezing on the rockmass quality is found to be significant for very weak rocks and decreases exponentially with increasing rockmass strength.  ii  Preface Chapter 8 is based on the paper "Developments in Empirical Approaches to Mining in Frozen Rock Masses" prepared by UBC graduate students Sheila Ballantyne and Megan Roworth, Cristian Caceres, and Rimas Pakalnis for presentation at the 47th US Rock Mechanics / Geomechanics Symposium held in San Francisco in June 2013.  iii  Table of Contents Abstract ........................................................................................................................................... ii Preface............................................................................................................................................ iii Table of Contents ........................................................................................................................... iv List of Tables ................................................................................................................................. ix List of Figures ................................................................................................................................ xi Glossary ....................................................................................................................................... xiv Acknowledgements ....................................................................................................................... xv 1.  2.  Introduction ............................................................................................................................. 1 1.1  Thesis Outline .................................................................................................................. 2  1.2  Research Objective ........................................................................................................... 3  1.3  Location and Background ................................................................................................ 5  1.4  Cigar Lake Mining Method .............................................................................................. 7  Literature Review .................................................................................................................... 9 2.1  Properties of Frozen Ground .......................................................................................... 10  2.1.1  Artificial Ground Freezing Background ................................................................. 10  2.1.2  Ice Mechanical Properties ....................................................................................... 11  2.1.3  Frozen Soil Mechanical Properties ......................................................................... 14  2.1.4  Frozen Intact Rock Properties................................................................................. 26  2.1.5  Creep Behaviour in Weak Rock ............................................................................. 34  2.2  Thermal Properties ......................................................................................................... 37  2.3  Frozen/Unfrozen Interface Behaviour ............................................................................ 38  2.4  Mining in Permafrost ..................................................................................................... 39  2.4.1  Case Studies in Frozen Underground Mines .......................................................... 40  2.4.2  Case Studies in Frozen Soil and Ice Deposits......................................................... 43  2.4.3  Ground Control of Frozen Placer Deposits............................................................. 44  2.5  Weak Rock Mass Behaviour .......................................................................................... 46  2.5.1  Rock Mass Classification Systems ......................................................................... 47  2.5.2  Modification of Rock Mass Classification Systems for Frozen Ground ................ 53 iv  2.5.3 2.6  Failure Mechanisms in Frozen Stratified Ground .......................................................... 57  2.6.1  Beam Theory ........................................................................................................... 58  2.6.2  Voussoir Analogue.................................................................................................. 60  2.7  Span Design of Underground Excavations .................................................................... 60  2.7.1 2.8 3.  4.  Rock Mass Strength ................................................................................................ 54  Critical Span Empirical Chart ................................................................................. 61  Applicability of Hoek-Brown Parameters to Frozen Ground ........................................ 63  Methodology.......................................................................................................................... 65 3.1  Assessment of Existing Information .............................................................................. 65  3.2  Conceptual Model of Failure Mechanisms .................................................................... 66  3.3  Material Properties Sampling Program .......................................................................... 67  3.3.1  Sample Collection ................................................................................................... 67  3.3.2  Sample Integrity During Drilling ............................................................................ 68  3.4  Classification Systems in Frozen Weak Rock................................................................ 69  3.5  Laboratory Testing to Establish Influence of Freezing .................................................. 70  3.5.1  Unconfined Compressive Strength Testing ............................................................ 72  3.5.2  Four Point Beam Testing ........................................................................................ 72  3.5.3  Direct Shear Testing ............................................................................................... 73  Cigar Lake Geology, Hydrogeology, and Historical Geotechnical Data .............................. 74 4.1  Regional Geology ........................................................................................................... 74  4.2  Formation of the Cigar Lake Deposit and Mineralization ............................................. 74  4.3  Local Geology ................................................................................................................ 75  4.3.1  Alteration ................................................................................................................ 75  4.3.2  Faulting and Structures ........................................................................................... 77  4.4  Geotechnical Site Investigations .................................................................................... 79  4.5  Geotechnical Zones ........................................................................................................ 80  4.5.1  Mineralization/Ore .................................................................................................. 82  4.5.2  Clay Altered Sandstone........................................................................................... 83  4.5.3  Sand/Highly Friable Sandstone and Fractured Sandstone ...................................... 85  4.5.4  Altered Basement .................................................................................................... 87 v  4.6 5.  In-Situ Stress Measurements .......................................................................................... 90  Back-Analysis of Historical Data .......................................................................................... 91 5.1  Comparison of Cigar Lake and McArthur River Mines ................................................ 91  5.2  Cigar Lake Mine, Jet Boring Trial in 2000 .................................................................... 93  5.2.1  Geology ................................................................................................................... 94  5.2.2  Instrumentation ....................................................................................................... 96  5.2.3  Influence of Freezing on Weak Altered Rockmass ................................................ 96  5.3 Rock Mass Classification Comparison of Frozen to Unfrozen Conditions at the McArthur River Mine.............................................................................................................. 103 6.  7.  Cigar Lake Geotechnical Material Properties Based on 2009 Drilling ............................... 109 6.1  Cigar Lake Geotechnical Domains .............................................................................. 109  6.2  Historical Geotechnical Drilling .................................................................................. 112  6.3  2009 Material Properties Drilling Program .................................................................. 112  6.4  Geotechnical Logging .................................................................................................. 114  6.4.1  Rock Quality Designation ..................................................................................... 114  6.4.2  Rock Strength........................................................................................................ 116  6.4.3  Joint Condition ...................................................................................................... 116  6.5  Interpretation of the Lithology and Rock Mass Characterization ................................ 117  6.6  Summary of 2009 Surface Freeze Drill Holes for Laboratory Testing Samples ......... 119  Frozen Laboratory Testing .................................................................................................. 125 7.1  Unconfined Compressive Strength Testing.................................................................. 125  7.1.1  Sample Collection ................................................................................................. 125  7.1.2  Sample Preparation and Setup .............................................................................. 126  7.1.3  Equipment ............................................................................................................. 126  7.1.4  Discussion of Results ............................................................................................ 130  7.1.5  Results ................................................................................................................... 152  7.2  Four-Point Beam Testing ............................................................................................. 153  7.2.1  Sample Preparation ............................................................................................... 156  7.2.2  Frozen Beam Testing Cement Mixture Samples .................................................. 157  7.2.3  Frozen Beam Testing Cigar Lake Drill Core Samples ......................................... 159 vi  7.2.4 7.3  8.  Results ................................................................................................................... 160  Frozen Direct Shear Testing ......................................................................................... 164  7.3.1  Sample Preparation ............................................................................................... 164  7.3.2  Test Procedures ..................................................................................................... 165  7.3.3  Results ................................................................................................................... 165  Influence of Freezing on a Weak Rock Mass ...................................................................... 167 8.1  Rock Mass Classification Schemes .............................................................................. 167  8.1.1  Intact Rock Strength ............................................................................................. 167  8.1.2  Joint Condition Ratings......................................................................................... 172  8.1.3  Water ..................................................................................................................... 175  8.2  Case Studies ................................................................................................................. 176  8.3 Comparison of Unfrozen to Frozen 2009 Surface Freeze Drilling Rock Mass Classification ........................................................................................................................... 181 8.3.1 9.  Discussion ............................................................................................................. 182  Failure Mechanism of Frozen Weak Rock Masses ............................................................. 186 9.1  Mohr-Coulomb Criterion ............................................................................................. 187  9.2  Hoek-Brown ................................................................................................................. 190  9.3  Frozen Material Properties ........................................................................................... 191  10.  Conclusions ...................................................................................................................... 193  10.1  Cigar Lake Rock Mass Highly Variable ...................................................................... 193  10.2  Frozen Laboratory Testing ........................................................................................... 193  10.3  Intact Rock Strength and Rock Mass Quality .............................................................. 194  11.  Recommendations ............................................................................................................ 196  11.1  General ......................................................................................................................... 196  11.2  Laboratory Testing ....................................................................................................... 196  11.3  In Situ Testing .............................................................................................................. 197  11.4  Developing Empirical Relationship Unfrozen to Frozen Rock Mass .......................... 198  11.5  Numerical Modelling ................................................................................................... 198  References ................................................................................................................................... 199 Appendix A: X-Ray Diffraction Testing .................................................................................... 210 vii  Appendix B: 2009 Unconfined Compressive Strength Testing .................................................. 211 Appendix C: Four Point Beam Testing ....................................................................................... 212 C1 - Concrete .......................................................................................................................... 213 C2 - Cigar Lake Drill Core ...................................................................................................... 214 Appendix D: Direct Shear Testing .............................................................................................. 215  viii  List of Tables Table 2.1: Values of Parameters in Primary Creep Law Equations, from Andersland and Ladanyi (2004) .............................................................................................................................. 34 Table 2.2: Summary of Creep Testing, after EBA (1990) and Golder (1986)......................... 35 Table 2.3: Cigar Lake Creep Parameters from Historical Testing ........................................... 36 Table 2.4: Summary of Relevant Mines in Permafrost ............................................................ 40 Table 2.5: Soviet Classification of Frozen Intermediate Roof Materials Up to 15 m Thick and Stable Spans after Extraction, after Emelanov et al. (1982) ......................................................... 45 Table 2.6: 1976 Rock Mass Rating Classification Scheme, from Bieniawski (1976) ............. 49 Table 2.7: Q Rating Parameters, from Barton et al. (1974) ..................................................... 51 Table 4.1: Results of Quantitative Phase Analysis (wt.%) ...................................................... 77 Table 4.2: Mineralization/Ore Unfrozen Material Properties (Golder, 2002) ......................... 82 Table 4.3: Mineralization/Ore Frozen Material Properties (Golder, 2002) ............................. 83 Table 4.4: Clay Unfrozen Material Properties ......................................................................... 84 Table 4.5: Clay Frozen Material Properties ............................................................................. 85 Table 4.6: Altered Sandstone Unfrozen Material Properties ................................................... 87 Table 4.7: Altered Basement Unfrozen Material Properties .................................................... 88 Table 4.8: Summary of Metapelite Basement Strength (Itasca, 2008) .................................... 89 Table 4.9: Altered Basement Frozen Material Properties ........................................................ 90 Table 5.1: Comparison of McArthur River and Cigar Lake Mine........................................... 91 Table 5.2: Cigar Lake Jet Boring Trial Dimensions ................................................................ 97 Table 5.3: Cigar Lake Jet Boring Trial Span Compared to Rock Strength.............................. 98 Table 5.4: Average Increase Between Frozen Face Mapping and Unfrozen Core Logging (Mawson, 2012) .......................................................................................................................... 108 Table 6.1: Summary of Rock Formations and Rock Descriptions Used for the 2009 Geotechnical Logging of Samples .............................................................................................. 110 Table 6.2: Summary of 2009 Surface Freeze Holes for Geotechnical Sampling .................. 113 Table 6.3: Field Strength of Geotechnically Logged 2009 Drillholes ................................... 116 Table 6.4: Joint Roughness of Geotechnically Logged 2009 Drillholes ............................... 117 Table 6.5: Joint Alteration of Geotechnically Logged 2009 Drillholes ................................. 117 Table 6.6: Unfrozen RMR76 and Q' of Geotechnically Logged 2009 Drillholes ................... 118 Table 6.7: Summary of Surface Freeze Borehole Field Strength, RQD, and RMR .............. 119 Table 7.1: Summary of Frozen UCS Testing on Bleached Sandstone .................................. 132 Table 7.2: Summary of Frozen UCS Testing on Hematized Sandstone/Clay ....................... 134 Table 7.3: Summary of Frozen UCS Testing on Graphitic Metapelite Basement ................. 136 Table 7.4: ISRM Field Strength Estimates, after Brown (1981)............................................ 138 Table 7.5: Summary of Unfrozen Bulk Densities .................................................................. 148 Table 7.6: Summary of Cement Mixture Samples for Four-Point Beam Testing ................. 158 ix  Table 7.7: Table 7.8: Table 8.1: Table 8.2: Table 8.3: Table 8.4: Table 8.5: Table 8.6: Table 8.7: Table 8.8: Table 8.9: Table 8.10: Table 8.11: Table 9.1: Table 9.2:  Summary of Drill Core Samples for Frozen Four-Point Beam Testing ............... 159 Summary of Frozen Direct Shear Testing Results on Drill Core ......................... 165 RMR Classification for Intact Rock Strength (Bieniawski, 1976) ....................... 167 Descriptions of Rock Strength and Approximate UCS (ISRM, 1981) ................. 168 RMR Classification for RQD (Bieniawski, 1976) ................................................ 170 RMR Classification for Joint Spacing (Bieniawski, 1976) ................................... 170 Jn Number for the Q Rock Mass Classification (Barton et al., 1974) .................. 171 RMR Classification for Joint Condition (Bieniawski, 1976)................................ 173 Q System Classification for Joint Roughness (Jr) (Hoek, 1980) .......................... 174 Q System Classification for Joint Alteration (Ja) (Hoek, 1980) ........................... 175 RMR Classification for Water (Bieniawski, 1976)............................................... 176 Average Increase Between Frozen Face Mapping and Unfrozen Core Logging . 179 Case History Summary of Frozen Rock Mass Conditions and Span.................... 180 Summary of UCS Failure Angles ......................................................................... 189 Frozen Material Properties .................................................................................... 192  x  List of Figures Figure 1.1: Location of the Cigar Lake Uranium Deposit, after Fayek et al. (2002) .................. 5 Figure 1.2: Cross-Section of Cigar Lake Orebody and Underground Development .................. 6 Figure 2.1: Schematic Stress-strain Curves for Low (10-7 s-1), Intermediate, and High Strain (10-3 s-1) Rates, after Schulson (1999) .......................................................................................... 12 Figure 2.2: Tensile and Compressive Strengths of Equiaxed and Randomly Oriented Fresh Water Ice of About 1 mm Grain Size vs. Strain Rate, after Schulson (1999) .............................. 13 Figure 2.3: Typical Ductile Stress-Strain Curve for Polycrystalline Ice Under a Constant Strain Rate ............................................................................................................................................... 14 Figure 2.4: Shear Stresses and Strain Curves for Frozen and Unfrozen Sands, after Youssef and Hanna (1988) .......................................................................................................................... 17 Figure 2.5: Variation of Angle of Friction and Cohesion for Frozen Sand with Low Ice Content, after Harris (1995) .......................................................................................................... 18 Figure 2.6: Frozen Soil Strength vs. Temperature, after Schultz and Hass (2005) ................... 19 Figure 2.7: Effect of Moisture Content on the Unconfined Compressive Strength of Frozen Sand at -12oC and a Strain Rate of 2.2 x 10-6 s-1, after Andersland and Ladanyi (2004) ............. 20 Figure 2.8: Idealized Creep Curve............................................................................................. 22 Figure 2.9: Frozen Soil Frost Heave Behaviour, after Shultz and Hass (2005) ........................ 26 Figure 2.10: Strength of Granite, Limestone, and Sandstone in Uniaxial Compression, after Mellor (1971) ................................................................................................................................ 28 Figure 2.11: Summary of Uniaxial Test Results for Unfrozen and Frozen Sandstone, after Yamabe and Neaupane (2001) ...................................................................................................... 29 Figure 2.12: Axial Stress vs. Axial Strain for Unfrozen and Frozen Sandstone, after Yamabe and Neaupane (2001) ................................................................................................................... 30 Figure 2.13: Strength of Granite, Limestone, and Sandstone in Uniaxial Tension, after Mellor (1971) ............................................................................................................................................ 32 Figure 2.14: Scale Effects, Intact Rock to Jointed Rock Mass, after Wyllie and Mah (2007) 54 Figure 2.15: GSI Values for Blocky Rock Masses, after Marinos and Hoek (2000) .............. 56 Figure 2.16: Four Point Beam Bending Load Test .................................................................. 60 Figure 2.17: Critical Span Curve, after Lang (1994) ............................................................... 62 Figure 2.18: Weak Rock Mass Critical Span Curve, after Ouchi et al. (2004) ....................... 62 Figure 2.19: McArthur River Stability Graph with Ground Support, after Pakalnis (2012) ... 63 Figure 4.1: Athabasca Basin and Cameco Corporation Active Mining Projects ...................... 74 Figure 4.2: Cigar Lake Deposit and Alteration Limits, after Jefferson et al. (2007) ................ 76 Figure 4.3: Stereonet Plots of Structural Data from 1999 Underground Drilling, from Baudemont (2000) Data ................................................................................................................ 79 Figure 4.4: Cigar Lake Geotechnical Zones .............................................................................. 81 Figure 5.1: Jet Boring Cavity Geology and Schematic of Surveyed Trial Cavities, after xi  Cameco (2000) .............................................................................................................................. 95 Figure 5.2: Cavity 1, Jet Boring Survey of Ore Cavity, UCS Based on Ore Grade .................. 99 Figure 5.3: Cavity 2, Jet Boring Survey of Ore Cavity, UCS Based on Ore Grade .................. 99 Figure 5.4: Cavity 3a, Jet Boring Survey of Ore Cavity, UCS Based on Ore Grade .............. 100 Figure 5.5: Cavity 4, Jet Boring Survey of Ore Cavity, UCS Based on Ore Grade ................ 100 Figure 5.6: Jet Boring Cavity Span on the McArthur River Critical Span Curve with Ground Support, after Pakalnis (2012) .................................................................................................... 101 Figure 5.7: 510L RMR Values and Diamond Drill Hole Trajectories .................................... 104 Figure 5.8: Combined Results of Core RMR vs. Drift RMR .................................................. 105 Figure 5.9: 510-8240 Drift RMR Compared to Rock Core RMR........................................... 106 Figure 5.10: 8220N Drift RMR Compared to Rock Core RMR ........................................... 107 Figure 6.1: Geological Variability of Material at the Cigar Lake Mine, after MDH (2008) .. 109 Figure 6.2: Borehole ST791-05, from 433.45 to 442.4 m ....................................................... 111 Figure 6.3: Rock Quality Designation Plots of Geotechnically Logged 2009 Drillholes ....... 115 Figure 6.4: 2009 Surface Freeze Holes for Laboratory Testing .............................................. 120 Figure 6.5: Cross Section North 10,032, Through Surface Freeze Holes, Unfrozen RMR76 . 121 Figure 6.6: Cross Section East 10,800 Through Surface Freeze Holes, Unfrozen RMR76 ..... 122 Figure 6.7: Cross Section East 10,790 Through Surface Freeze Holes, Unfrozen RMR76 ..... 123 Figure 6.8: Cross Section East 10,796 Through Surface Freeze Holes, Unfrozen RMR76 ..... 124 Figure 7.1: Inside Cold Room, Triaxial Cell Setup. Left Triaxial Cell is a Sample Freezing Waiting to be Tested. Right Triaxial Cell is a Sample Undergoing Testing. ............................. 127 Figure 7.2: Triaxial Cell Filled with Mineral Oil, Sitting on Load Cell. Displacement LVDT Sensor Seen to Top Right of Cell. Load is Applied by the Top Load Conducting Rod ............. 127 Figure 7.3: Syringe Pump Controlling Loading Rate and Measuring Load............................ 128 Figure 7.4: Glycol Transfer Unit Circulating Glycol in Copper Coils Outside of Triaxial Cell. Glycol Circulating at Half a Degree Celsius Below Ambient Room Temperature. ................... 128 Figure 7.5: Cross Section of Frozen High Moisture Content Hematized Sandstone Showing Little to No Ice Lensing Present after 24 hours Freezing at -10oC ............................................. 130 Figure 7.6: Frozen UCS vs. Total Strain of Bleached Sandstone Samples ............................. 133 Figure 7.7: Frozen UCS vs. Total Strain of Hematized Sandstone/Clay ................................ 135 Figure 7.8: Frozen UCS vs. Total Strain of Graphitic Metapelite Basement .......................... 137 Figure 7.9: Frozen UCS vs. Unfrozen ISRM Rock Strength, All Data................................... 139 Figure 7.10: Frozen UCS vs. Unfrozen ISRM Rock Strength, Good Data, Samples That Failed Through Joints or Bedding Removed .............................................................................. 140 Figure 7.11: Plot of All Samples, Frozen UCS vs. Applied Strain Rate, T=-10oC ............... 141 Figure 7.12: Plot of All Samples, Frozen UCS vs. Applied Strain Rate, T=-20oC ............... 142 Figure 7.13: Frozen UCS vs. Strain Rate of All 2009 Samples, by Failure Mode ................ 143 Figure 7.14: Influence of Freezing and Strength Gain for Weak Cigar Lake Rock .............. 145 Figure 7.15: Influence of Temperature on Frozen UCS, 2009 Data, by Failure Mode ......... 146 xii  Figure 7.16: Influence of Temperature on Frozen UCS, All Data, by Rock Type ................ 147 Figure 7.17: Frozen UCS vs. Unfrozen Bulk Density ........................................................... 149 Figure 7.18: Frozen UCS vs. Porosity, by Material Type ..................................................... 150 Figure 7.19: Frozen UCS vs. Porosity, by Failure Mode ...................................................... 151 Figure 7.20: Frozen UCS vs. Moisture Content, 2009 Data .................................................. 152 Figure 7.21: Four-Point Beam Test Apparatus ...................................................................... 154 Figure 7.22: Frozen Tensile Strength vs. Moisture Content, Cement Samples by Mixture .. 161 Figure 7.23: Frozen Tensile Strength vs. Moisture Content, Cement by Joint Presence ...... 162 Figure 7.24: Frozen Tensile Strength vs. Moisture, Drill Core Samples by Joint Presence.. 163 Figure 7.25: Plot of Direct Shear Testing Results on Drill Core ........................................... 166 Figure 8.1: Empirical Support Design, after Grimstad and Barton (1993) ............................. 178 Figure 8.2: Case Studies Frozen RMR vs. Cavity Span on the McArthur River Rock Mass Critical Span Curve, after Pakalnis (2012) ................................................................................ 180 Figure 8.3: Comparison of an Unfrozen RMR to Frozen RMR, after Bieniawski (1976) ...... 183 Figure 8.4: GSI Values for Blocky Rock Masses with Unfrozen and Frozen RMR, after Marinos and Hoek (2000) ........................................................................................................... 184 Figure 8.5: Cross Section North 10,032, Unfrozen and Frozen RMR76.................................. 185 Figure 9.1: Mohr-Coulomb Failure Envelope ......................................................................... 187 Figure 9.2: Example of UCS Failure Angle ............................................................................ 188  xiii  Glossary Bulk Density:  Measure of the weight of the soil or rock per unit volume. Measured in grams/cm3 or kilograms/m3.  Cohesion:  Measure of internal bonding of the material. Part of the shear strength used to describe the strength of a material to resist deformation due to shear stress. Measured in kPa or Pa.  Flexural Strength:  Defined as the material's ability to resist deformation under load, the highest stress that a material can experience within the material at its moment of rupture. Also termed modulus of rupture.  Internal Friction:  Internal friction is caused by contact between particles of the material. Part of the shear strength used to describe the strength of a material to resist deformation due to shear stress. Measured in degrees.  Hoek-Brown:  Failure criterion for isotropic rock material and masses.  Modulus of Elasticity:  Mathematical description of an object’s tendency to be deformed elastically when a force is applied to it. Defined as the slope of the stress-strain curve in the elastic deformation region.  Modulus of Rupture:  Defined as the material's ability to resist deformation under load, the highest stress that a material can experience within the material at its moment of rupture. Also termed flexural strength.  Mohr-Coulomb:  Mathematical model that relates the shear strength to the stress of a material element, equation: τ = c +σ tan (θ). Materials behaving according to the theory are referred to as Mohr-Coulomb material.  Poisson's Ratio:  Ratio of the amount of lateral strain to axial strain.  Stress:  Internal resistance offered by a unit area of a material from which a member is made to an externally applied load. Measured in kPa, MPa or N/m2.  Tensile Strength:  Defined as the maximum tensile stress that a rock can sustain. Rocks placed in tension (outward pulling force) will fail at a much lower value than in compression. Units of stress are in kPa, MPa or N/m2.  UCS:  Unconfined Compressive Strength. The maximum force that can be applied to a sample without breaking it. Units of stress are in kPa, MPa or N/m2.  Young's Modulus:  Modulus of elasticity measuring the stiffness of a rock material. Defined as the ratio, under small strains, as the change in stress with strain. Values reported in this thesis are calculated at 50% of the UCS value. xiv  Acknowledgements I would like to acknowledge those who provided financial, technical and personal support during the study of research program. It is with their help throughout that this project was completed. Special thanks to Cameco Corporation and NSERC for their sponsorship of this research project; specifically Kerry McNamara, Scott Bishop, and Ken Gullen. Special thanks to my supervisors, Rimas Pakalnis and Erik Eberhardt. I also thank Lukas Arenson, BGC Engineering, Stephen Gamble and David Sego (University of Alberta) who assisted with the frozen laboratory testing providing invaluable comments and usage of the cold room at the University of Alberta.  xv  1.  Introduction  The uranium deposits in the Athabasca Basin in northern Saskatchewan are typically located at the unconformity between the basement rock and an overlying porous sandstone layer. Above and below the unconformity, the rock mass shows high variability in porosity and permeability due to intense fracturing and alteration around the orebody. The porous nature of the sandstone combined with a 450 meter hydrostatic head of groundwater and poor rock conditions have resulted in large inflows and flooding of the Cigar Lake Mine, a prospective high grade uranium property. Geotechnical challenges to mine the Cigar Lake orebody include groundwater control and supporting the weak ground overlying and below the orebody. To mitigate the potential for groundwater inflow, the Cigar Lake project plans to implement artificial ground freezing along with the non-entry mining method of jet boring. Although artificial ground freezing has been used for ground support and water control for many decades, the influence of artificial freezing on a weakly jointed rock mass at depth is not well understood. This introduces uncertainty, which impacts the safety and economic viability of the mines. Natural ground freezing occurs seasonally in many areas of the world and can adversely affect the performance of the ground and adjacent structures as the freezing of pore water to ice causes a phase change expansion of approximately nine percent of the pore water volume. Freezing results in a significant strength increase of the ground due to ice bonding in saturated soils and rock masses. Artificial ground freezing is typically a last resort excavation support alternative to cut-off walls and grouting that involves the use of refrigeration pipes underground to convert in situ pore water into ice. Artificial ground freezing to provide groundwater control and excavation support is typically applied in shaft sinking and less commonly in deep underground mines. McArthur River uranium mine, located 30 km southwest of Cigar Lake, is the only mine in Canada to currently use ground freezing to create a permeability barrier between mine workings and potential water inflow sources. The geological setting at Cigar Lake is similar to the McArthur River mine in that the sandstone overlying the basement rocks of the deposit contains significant water at high hydrostatic pressure; however, McArthur River currently does not rely upon frozen ground for primary ground support only to control water.  1  Geotechnical boreholes to characterize the Cigar Lake orebody and surrounding area have been completed from the mid-1980s to present. Initial samples of the weak rock overlying the orebody have been collected to establish frozen strength parameters. A material properties data collection program was completed by Cigar Lake mine in early 2009 to address data gaps from historical geotechnical drilling and provide an understanding of the shear strength, time dependent behaviour, and thermal properties of weak frozen rock under pressure. Four PQ (3”) boreholes were cored through the orebody for material sampling as part of the surface freeze drilling test program. Frozen laboratory testing was completed by the author on the weak rock overlying the orebody to understand the failure mechanism and strength relationship with varying temperatures and strain rates. The key focus of the laboratory testing is to improve in situ and laboratory characterization methods and provide a better understanding of weak rock behaviour at sub zero temperatures. 1.1  Thesis Outline  This thesis consists of nine chapters that include a description of the Cigar Lake mine operation and development history, a literature study of the current research, a back analysis of historical excavations in frozen ground at the McArthur River mine and Cigar Lake mine, results of the frozen laboratory testing on Cigar Lake material, and subsequent analysis of the influence of freezing on a weak and jointed rock mass. Chapter 2 reviews the current research in the mechanical behaviour of frozen soil and rock, mining in frozen ground, and various methods in understanding the failure mechanisms of a frozen jointed weak rock. Chapter 3 outlines the methodology of the research to understand the influence of freezing on a weak and altered/fractured rock mass at depth. Chapter 4 details the regional geology, hydrogeology regime and geomechanical properties of the Cigar Lake mine rock types relevant to the research. Chapter 5 details the back analysis of a jet boring trial in frozen ground at the Cigar Lake mine and a comparison of unfrozen drill core and frozen face mapping at the McArthur River mine to establish the influence of freezing on rock mass rating (RMR) classification values.  A 2  comparison of the geotechnical parameters between Cigar Lake and McArthur River mines, both operated by Cameco Corporation, is included in order to provide recommendations on artificial ground freezing design. Geotechnical core logging and laboratory testing for freeze wall design has been minimal at both the Cigar Lake and McArthur River mine sites. Chapter 6 presents the current understanding of the geomechanical properties of the Cigar Lake orebody, the material overlying and beneath the orebody, a strongly altered sandstone, and altered basement metapelite, respectively. In 2009 a surface freeze drilling campaign was completed at the Cigar Lake mine where select boreholes were sampled as part of a geotechnical laboratory testing program for this research. Chapter 7 discusses the frozen Unconfined Compressive Strength, frozen four point beam, and frozen direct shear testing completed on the 2009 surface freeze drilling material to provide an understanding of the gain in strength due to freezing of a weak rockmass, how a weak jointed frozen rock mass fails, and to develop a model of the gained shear strength of a frozen joint Chapter 8 presents the interpretation of case history data of mines in permafrost or artificially frozen ground and the laboratory testing from the Cigar Lake mine to understand and predict the behaviour of openings in frozen rock masses using the empirical approaches of the rock mass rating (RMR) system. Chapter 9 discusses the potential failure mechanisms in a frozen jointed weak rock mass and the summary of geotechnical parameters as inputs for numerical modelling frozen weak rock. The conclusions summarize the thesis findings including the gain in strength of freezing on a weak rock mass, the behaviour of a weakly jointed rock mass under tensile stresses, and the development of a frozen rock mass rating vs. span based on available case histories.  1.2  Research Objective  The objective of this research is to determine how freezing affects the behaviour of a weak and jointed rock mass with direct application to the Cigar Lake mine. Cigar Lake’s orebody and the adjacent surrounding rockmass will undergo bulk freezing prior to mining, at an approximate depth of 430 to 450 m below surface. The weakest and most challenging material identified at 3  Cigar Lake Mine is the dense clay to very weak and altered sandstone directly above the orebody, which will form the back of the jet bored cavity. Above this zone is a heterogeneous and permeable material (sandstone origin) comprising soft to indurated sandy clay, unconsolidated sand, and variably altered sandstone. The potential for high and uncontrolled groundwater inflow events are mitigated through ground freezing; however, fracturing of the ice cap above the orebody will be catastrophic, creating a direct conduit to high pressure water. The orebody will be mined by jet boring, a non-entry mining method using pressurized water to excavate cavities. In order to design the freeze wall (ice cap) and ensure stability of the jet bored stability, a better understanding of cavity failure mechanisms in frozen weak rock is required. The behaviour and stability of the mined out cavities once mining commences is a function of the frozen rock mass overlying the orebody. Potential failure mechanisms of an excavated ore cavity include the separation between unfrozen and frozen material in the back of the cavity and severe cracking of the ice matrix. The behaviour of frozen soil is well documented with extensive research in the mechanical and creep relationships with varying grain sizes, moisture, and temperature. However, limited information exists on the behaviour and failure mechanisms of frozen weak rock at great depth as the majority of frozen ground research is based on permafrost regions in surficial soil. The influence of freezing on a jointed weak rockmass at depth has not been cited in the literature to date. This research will provide a better understanding of how a weak frozen jointed rock behaves in order to assist with the freezing design for jet bored cavities at the Cigar Lake mine, such as the thickness of frozen ground above the orebody (ice cap) and stable cavity dimensions for varying ground conditions. The Cigar Lake Mine (Cameco Corporation), a major sponsor of this research, will integrate the results into the day-to-day mining and management of operations. The work will be used to develop, evaluate and forecast safe, environmentally favorable mining strategies at depth for the life of mine plan. Prevention of inflows is one of Cameco Corporation's greatest challenges going forward that other companies in the industry share.  4  1.3  Location and Background  The Cigar Lake project, one of the world’s largest undeveloped uranium mines, is operated by Cameco Corporation and located 660 km north of Saskatoon, about 40 km inside the eastern margin of the Athabasca Basin region, as shown in Figure 1.1. The Athabasca Basin region supplies 20% of the world’s uranium with the majority operated by Cameco Corporation. The Cigar Lake deposit is a high grade uranium mineralization with proven and probable reserves of more that 226.3 million pounds U308 at an average grade of 20.7% (Cameco, 2007).  Figure 1.1:  Location of the Cigar Lake Uranium Deposit, after Fayek et al. (2002)  Discovered in 1981, the orebody is located at a depth of 450 m between the Athabasca sandstone formation and the underlying Precambrian basement rocks. The deposit is approximately 1,950 m long, 20 to 100m wide, and ranges up to 12m thick, with an average thickness of about 5m (Figure 1.2).  5  Figure 1.2:  Cross-Section of Cigar Lake Orebody and Underground Development  Project construction at Cigar Lake began in early 2005 and is anticipated to be completed in 2013. During test mining and mine construction, Cigar Lake project experienced several inflow events due to poor ground conditions and high water pressures. •  In October 1999, a rock collapse lead to a water inflow of 40 m3/hr on the 465 mine level near No. 1 Shaft. The inflow at this location was manageable, but the collapse was believed to be approximately 3 m from the unconformity with the potential of becoming a more significant inflow problem (MDH, 2008).  •  During the sinking of Shaft 2 in April 2006, a water inflow occurred resulting in shaft flooding; this event is believed to be a result of the reactivation of ancient fault structures (Baudemont, 2007).  •  In October 2006, a collapse in the vicinity of the 944 Drift East and the 773 Launch Chamber on the 465 mine level caused an inflow event that flooded the mine completely. The October 2006 inflow is located at the southern margin of the mineralized zone, 6  where sand locally comes in contact with the primary mineralized zone or clay cap attributed to a combination of water pressures in unconsolidated material (near the unconformity) and disturbance made to the east-west trending fault system close to the unconformity (Baudemont, 2007). •  During mine dewatering in August 2008 a ground fall occurred at the 420L near Shaft No. 1 during remediation from the 2006 inflow event. Water inflow associated with this ground fall resulted in the mine flooding to ground surface.  1.4  Cigar Lake Mining Method  This section describes the Cigar Lake mining method based on input from the Cigar Lake technical services team in 2009. At Cigar Lake, mining will be conducted from the 465 m production level which is located 10 m below the uranium deposit. Artificial ground freezing will be implemented to support the weak rock surrounding the orebody to minimize the potential for a large water inrush while mining the ore, and stop radon migration. Two strategies are being considered to freeze the ore zone prior to mining. The first option is bulk freezing where vertical freeze holes from the 480 m level up through the orebody will be drilled. Installing freeze pipes from surface to the 465 m production level is the second option. The ground freezing system consists of an ammonia refrigeration plant on surface, a surface and underground brine piping system and in-situ freeze pipes. Calcium chloride brine at -30oC is delivered underground through pipes from a surface refrigeration plant. Jet boring is the proposed plan to mine out the Cigar Lake orebody and considered a unique and novel non-entry mining method not applied in any other mine worldwide. The jet boring system (JBS) developed by Cameco Corporation involves the following steps: •  artificial ground freezing of the orebody and surrounding rock,  •  development of access crosscuts below the orebody,  •  installation of cased pilot holes up through the ore,  •  ore extraction with rotating high pressure water jets, and 7  •  cavity backfilling with concrete.  The cutting of ore with high pressure water produces a slurry to be pumped in pipelines. Ore extraction with rotating high pressure water jets is expected to produce cavities fairly circular in shape measuring 4 to 5 m in diameter and heights varying with ore thickness (3 to 12 m). Underground mining tests of the JBS were completed in 1992 providing the design basis for the field trial in 1999 and 2000. In 2000, four cavities were excavated in frozen waste rock, just below the ore as part of the second JBS test program. The study area was frozen through near vertical freeze pipes installed through the orebody with calcium chloride circulating at -40C through the freeze pipes. Several cavities were jet bored and surveyed to determined potential cavity sizes. The cavities were noted to be stable for several days after excavation.  8  2.  Literature Review  This section is a compilation of studies investigating the mechanical and thermal behaviour of frozen soils, frozen hard rock masses, mining within naturally frozen soils and rocks, and rock mass classification systems. This literature review is presented in its entirety as no applied information exists on frozen ground with respect to its application on weak rock masses and their design related to a mining environment. The behaviour of frozen soil is well documented with extensive research in the mechanical and creep relationships with varying grain sizes, moisture contents, and temperatures. However, the behaviour and failure mechanisms of frozen jointed weak rock at depth (> 100 m) is not well understood as the majority of frozen ground research is based on permafrost in surficial soil. Limited to no research on the mechanical and thermal properties of weak frozen rock was available at the time of preparing this thesis. Given the lack of mines operating at depth under artificially frozen environments, research into mines operating in permafrost environments where the ground (hard rock and soils) is frozen is reviewed here. Key questions to address as part of this literature review include the following: • What is the influence of freezing on joints and fractures in a rock mass? • How does weak frozen material behave under pressure? • How do frozen material properties compare to unfrozen geotechnical properties? • Do we understand potential failure mechanisms such as separation between unfrozen and frozen material, cracking of the ice matrix, or failure as a weak rock mass? • What failure criteria for frozen, jointed, and weak rock masses have been established, if any?  The following topics below are discussed in this literature review to address the key questions outlined above. • Artificial ground freezing history and uses in the mining industry; • Mechanical and thermal properties of ice, soil, and rock; • Behaviour of the interface between unfrozen and frozen ground; • Excavations in frozen ground, including the performance, dimensions and behavior of the cavity; • Behaviour and failure mechanisms of unfrozen weak rock; and • Rock mass classification systems and the influence of freezing on the input parameters. 9  2.1  Properties of Frozen Ground  2.1.1 Artificial Ground Freezing Background Frozen ground is defined as soil or rock below 0oC in temperature and is independent of the water and ice content within the soil or rock matrix (Andersland and Ladanyi, 2004). As the temperature drops below 0oC, soil and weak rock masses become impervious to seepage and increase in strength as ice bonds together adjacent particles providing structural support. Artificial ground freezing (AGF) involves the use of refrigeration systems underground to convert in situ pore water into ice. Benefits of AGF are that the ground remains undisturbed as it is non-invasive and can be used in any soil formation regardless of structure, grain size, permeability or groundwater flow velocity. AGF is versatile in soil and rock as long as there is sufficient moisture for ice bonding and the regional groundwater flow is nominal. Artificial ground freezing was first applied to support vertical openings in South Wales, Australia in 1862 and patented by H. Poetsch in Germany in 1883 (Harris, 1995). Artificial ground freezing is typically considered for excavation support in deep, difficult, disturbed or sensitive ground or when complete groundwater cut-off is critical (Schmall et al., 2005). Ground freezing has historically been used in shaft sinking through wet loose soils and recently for temporary support or as an aid to recovery due to collapsed soils in other areas such as underpinning, mining, deep excavations, and groundwater cut-offs. Artificial ground freezing for deep excavation support has been applied in shaft sinking up to depths of 900 m in Saskatchewan for difficult ground conditions and rock/soil interfaces producing large water inflows (Harris, 1995). The primary objective of ground freezing is to remove heat from the ground until the temperature is below the freezing point of the groundwater system. Continuous energy is required to maintain and establish a freeze wall that is achieved through two options; a refrigerated brine or liquid nitrogen system. The conventional freezing system is mechanical refrigerated calcium chloride brine circulating through a closed circuit pipe system and returns to the refrigeration plant for cooling. The chilled brine is typically circulating at -25oC to -40oC to chill the strata to -5oC. Liquid nitrogen, the alternative, is allowed to evaporate and freeze within tubes installed underground to cool the ground. Liquid nitrogen systems are commonly used for 10  rapid freezing as the system is more efficient than refrigerated chilled brine. Chilled brine refrigeration plants are cost effective for long periods while liquid nitrogen as a refrigerant is only viable for short term stabilization. Freeze wall growth and complete cut-off is typically monitored in the ground with temperature probes. Closure of a freeze wall can be inhibited by high groundwater velocity layers, undissolved contaminants, saline pore fluids, and dissolved solids. Schmall et al., (2005a), Shultz et al. (2005), and Harris (1994) summarize applications of ground freezing projects in difficult highly sensitive ground or under high groundwater velocity. Effective groundwater flow velocities in excess of 2 m/day is considered a threshold value on a chilled brine freeze system as the high flow rate demands an excessive heat load (Schmall et al., 2005). The critical groundwater velocity depends on freeze pipe spacing, coolant temperature, soil permeability, shape and size of the design frozen mass. Remedial measures in difficult ground include increasing freeze pipe spacing or reducing ground permeability through grouting. Catastrophic failures of ground freezing projects have been rare, but partial failures due to an unfrozen zone are not uncommon. Leakages in freeze walls due to higher than anticipated groundwater velocities were fixed with additional freeze pipes and grouting around the leaking zone (Schultz and Hass, 2005). Material properties relevant for a ground freezing structural analysis are the strength and deformation properties as frozen earth behaves visco-elastically and is subject to time-dependent deformation under constant stress. 2.1.2 Ice Mechanical Properties Pure ice is a crystalline structure that is obtained by the freezing of water. The mechanical behaviour of ice is dependent on strain rate, temperature, porosity, grain size and structure. Pure ice is typically polycrystalline with random crystal orientation whose response to a deviatoric stress can be represented by a power law creep equation. For short periods of loading polycrystalline ice behaves elastically with little recoverable deformation at high loading rates. Under sustained loading, micro cracking may occur under low stresses with the cracks dominating at high loading rates. When ice is loaded at small strain rates the maximum stress 11  remains the same initially and then decreases with an increase in confining pressure. Ice failure modes are mainly dependent on the applied strain rate. The ductile-brittle transition in ice occurs at lower strain rates under tension as the applied stress opens the cracks directly. Under compression, the required tensile stress is generated locally through crack sliding. Figure 2.1 illustrates the typical ice response to loading regimes under low (I), intermediate (II), and high (III) strain rates (Schulson, 1999).  Figure 2.1: Schematic Stress-strain Curves for Low (10-7 s-1), Intermediate, and High Strain (10-3 s-1) Rates, after Schulson (1999) Note: I – low stress loading regime, allows for creep and a sustained load II – intermediate loading regime, the ice will fail at a higher tensile and compressive strength than under a low stress environment, but in a brittle manner III – high strain rate loading regime, the ice will fail in a brittle, with a higher compressive strength, though no change in tensile strength  Figure 2.2 plots the results of several tests of ice under tensile and compressive loading conditions. With increasing axial strain rate, samples in compression will gain strength. However, the tensile strength of ice remains constant under varying strain rates. At low strain rates, there is little to no compressive strength of ice. At higher strain rates, ice has a high 12  compressive strength.  Change in Compressive Strength with Strain Rate  Change in Tensile Strength with Strain Rate  Low Strain Rate  High Strain Rate  Figure 2.2: Tensile and Compressive Strengths of Equiaxed and Randomly Oriented Fresh Water Ice of About 1 mm Grain Size vs. Strain Rate, after Schulson (1999)  Typically, the tensile strength of ice varies from 0.7 to 3.1 MPa and the compressive strength varies from 5 to 25 MPa over the temperature range -10 to -20°C. The ice compressive strength increases with decreasing temperature and increasing strain rate, but ice tensile strength is relatively insensitive to these variables (Petrovic, 2003). The implications of this are relevant to the Kupol mine, discussed in Section 2.4.1, where the mine operates at just above freezing (1oC) and is still able to confine the dead weight of the frozen back relying on cohesive strength. Ice samples in uniaxial compression show a small volume increase during testing. When subjected to 13  shear stresses at low hydrostatic pressures, polycrystalline ice showed ductile yielding at low strain rates. The strength of ice is dependent on the load path experienced by the ice, as illustrated in Figure 2.3. The yielding and failure of polycrystalline ice under a triaxial state with high hydrostatic pressures causes weakening and eventual melting. Ice generally behaves in a ductile manner up to a strain rate of 10-4 above which ice goes through a transition to completely brittle failure above a strain rate of 10-2 (Michel, 1978).  σ  Peak Strength  Initial Yield Point  To residual strength  1% Failure Strain, εt  ε (%)  Figure 2.3: Typical Ductile Stress-Strain Curve for Polycrystalline Ice Under a Constant Strain Rate  2.1.3 Frozen Soil Mechanical Properties Frozen soil is a four phase mixture comprising soil particles, ice, water, and voids. Voids can be filled with air, ice, and/or unfrozen water. Frozen soil mechanical behaviour closely reflects that of ice although unfrozen water may be present in the frozen matrix. Andersland and Ladanyi 14  (2004) consider the most important characteristic distinguishing the mechanical behaviour of frozen soils from unfrozen soils to be the ice and water composition that constantly varies with temperature and applied stress. The mechanical properties of a frozen soil at a given temperature can vary from brittle to plastic depending on the unfrozen water content (Andersland and Ladanyi, 2004). The initial freezing temperature for cohesionless soils is close to 0oC and for fine grained soils the temperature depression can be up to 5oC as the pore water does not freeze uniformly at the same temperature. The rate at which soil freezes is dependent upon its thermal properties, moisture content, and temperature. Generally sands and quartz rich soils will convert all water to ice several degrees below 0oC; however, clay rich material will keep unfrozen water in the matrix well below 0oC. A significant amount of unfrozen water can still exist in fine grained soils below the initial freezing temperature as thin liquid like layers on the particle surfaces. The unfrozen water content will affect the thermal and mechanical properties of the frozen soil. Strength and stiffness decrease with increasing unfrozen water content. Unfrozen water content is influenced by mineralogy, temperature, and salinity of the pore water. Tice et al. (1976) developed experimental unfrozen water content parameters for various soil types. Determining the freezing temperature of the four phase solid, water, ice and gas mixture for soils was studied by Miller (1980) who highlighted the influence of the unfrozen water content on the freezing temperature. 2.1.3.1 Compressive and Shear Strength of Frozen Soil As with unfrozen soil, the strength of frozen soil depends on interparticle friction, particle interlocking and cohesion. In frozen soils, the bonding of particles by ice is the major stabilizing factor. Ting (1983) indicates that three mechanisms control the strength of frozen soils; ice strength, soil strength, and interaction between the ice matrix and soil skeleton. The soil skeleton and ice matrix yield at different strengths when sheared in compression under low confining pressures. Typically two yield points at 1 and 10% axial strain are present, corresponding to the peak strength of the ice and soil, respectively. The ice strength dominates at low strains where cracking of the ice matrix occurs at less than 1% strain (Sayles, 1988), which is before maximum 15  compression of the frozen sample. The short term strength of frozen soil represents the instantaneous strength and is significantly higher than the long term strength due to the brittle to viscoelastic response of ice under varying load times. The short term strength of frozen soil is measured as the total stress at a constant rapid deformation rate. Long term strength of frozen soil is a measure of its time dependent creep behaviour and is determined using uniaxial creep tests at constant deformation rate and various percentages of loading stress. The long term strength is typically 1.5 to 2.0 less than the short term compressive strength (Andersland and Ladanyi, 2004). Stress-strain behaviour of a frozen soil depends on soil type, mineralogical composition, ice content, temperature, and strain rate. Strain rate and temperature have less influence on the friction angle of a frozen sample, than on the cohesion (Andersland and Ladanyi, 2004; Jessberger et al. 2003). Typically the friction angle will decrease with sub zero temperatures and the cohesion will increase significantly, especially in non cohesive soils and very weak rock masses where the ice is bonding together the particles. Youssef and Hanna (1988) compared the stress-strain behaviour of unfrozen and frozen sands. Frozen sands have higher shear strengths than unfrozen sands due to the interlocking nature of the water in the matrix converted to ice. Figure 2.4 shows that at a temperature of -5oC, freezing results in a shear strength increase by a factor of 2.5. At higher strain levels the friction angle approaches that of unfrozen sand while cohesion approaches zero.  16  Frozen – exhibits first peak, near that of ice  Figure 2.4: Shear Stresses and Strain Curves for Frozen and Unfrozen Sands, after Youssef and Hanna (1988)  Nater et al. (2008) developed a correlation of the effective angle of internal friction (φ’) and cohesion (c) with temperature dependent parameters, for example defining the volumetric ice content (wi), where the strength of frozen soils depends on the temperature. Nater et al. (2008) observed that the effective angle of internal friction decreases with the volumetric ice content, whereas the cohesion increases with increasing ice content. The correlations are based on laboratory tests carried out on undisturbed samples of alpine permafrost soils. Figure 2.5 depicts the change in friction angle and cohesion with decreasing temperature after Harris (1995).  17  Cohesion Friction  With decreasing temperatures, the cohesion of a frozen sand will increase. The friction angle will increase slightly, then begin to decrease with colder sub-zero temperatures.  Figure 2.5:  Variation of Angle of Friction and Cohesion for Frozen Sand with Low Ice Content, after Harris (1995)  2.1.3.1.1 Influence of Strain Rate Frozen soil stress-strain behaviour is strongly affected by strain rate. For lower strain rates a sample exhibits plastic flow followed by small elastic deformation and as the strain rate increases the strength increases and failure mode changes from ductile to brittle. Soil strength dominates at larger strain rates influencing the long term frozen soil strength. The strain rate at which transition to brittle behaviour occurs is higher for clays than gravels presumed due to the greater unfrozen water contents (Andersland and Ladanyi, 2004). Cohesive strength of frozen soils increases with strain rate. The ice matrix under normal pressure and temperature is more rigid than the soil skeleton where it reaches peak strain under much lower strains. 2.1.3.1.2 Influence of Temperature The strength of frozen ground becomes greater at lower temperatures, but decreases with the applied loading time. In general, a decrease in temperature results in a significant increase in the strength of frozen soil, but the brittleness also increases (Sayles and Haines, 1974; Haynes and 18  Karalius, 1977; and Haynes, 1978).  Figure 2.6 shows the uniaxial strength (UCS) versus  temperature for typical soil types and for pure ice (Schultz and Hass, 2005). The average temperature where frozen soil exhibits linear behaviour usually ranges between -5 and -25oC.  Note: 1. Fine Sand 2. Silty Sand 3. Medium Sand 4. Clay 5. Pure Ice 6. Pure Ice  Figure 2.6:  Frozen Soil Strength vs. Temperature, after Schultz and Hass (2005)  2.1.3.1.3 Influence of Ice Content Studies on the frozen soil mechanics of sand-ice mixtures were performed by Goughnour and Andersland (1968), Kaplar (1971), Hooke et al. (1972), and Baker (1979).  The studies  concluded that up to a grain volumetric content of 40%, pore ice governs frozen behaviour; at 40% by volume sand content particle contact is established; between 40% and 60% friction governs; above 60% dilatancy adds to shear strength. Interparticle friction and dilatancy influences the strength at mixtures greater than 40% sand by volume. At lower concentrations the strength of the sand and ice mixture was only slightly higher than that of pure ice. High and 19  low ice contents tend to reduce the strength of the frozen ground, which peaks between 25 and 45% moisture content (Andersland and Ladanyi, 2004). Sayles and Carbee (1980) studied the effect of silt concentration on the behaviour of ice-silt mixtures. At silt concentrations greater than 50% displayed a strain-hardening where at concentrations less than 50% the mixture is dominated by ice. 2.1.3.1.4 Influence of Saturation Based on the research by Kaplar (1971) and Baker (1979) the strength of a frozen soil is dependent on the degree of saturation, as the peak strength increases as the soil increases in saturation content, reaching a peak frozen strength at approximately 30% water content (refer to Figure 2.7). The lowest frozen compressive strengths are associated with completely dry and fully saturated conditions. When a soil is completely dry, the strength is that of an unfrozen soil as there is no added gain in pore water freezing and strength. With saturation increasing beyond 40%, fine sand has a compressive strength of approximately 60% of its maximum strength, rapidly decreasing to the strength of frozen ice.  Figure 2.7: Effect of Moisture Content on the Unconfined Compressive Strength of Frozen Sand at -12oC and a Strain Rate of 2.2 x 10-6 s-1, after Andersland and Ladanyi (2004)  20  2.1.3.2 Uniaxial Tension The amount of data on tensile testing of frozen soils is more limited than that on compression testing. However, in general, the behaviour of frozen soil in uniaxial tension is more brittle compared to uniaxial compression tests under similar conditions, but tensile strength is less sensitive to temperature and strain rate (Haynes et al., 1985; Bragg and Andersland, 1982). The failure strain in tension of the ice rich silt was approximately one order of magnitude lower than that in compression (Zhu and Carbee, 1984; 1987). For frozen ice rich silt, the tensile strengths remain constant up to the plastic-brittle transition, beyond which the tensile strengths decreased. Sayles (1991) defined a peak tensile strength with a power law based on the uniaxial compression values for a sandy silt, fine sand, and gravelly sand at temperatures of -1.1 to -6.7oC and strain rates between 10-1 and 10-5 h-1. Yuanlin and Carbee (1985) studied the strain rate effect on the tensile strength of silt and concluded that for ductile behaviour both the tensile and compressive strength were substantially influenced by the strength of the ice matrix which was similar in both tension and compression under the same testing conditions. 2.1.3.3 Creep Behaviour When a frozen specimen is subjected to a load it will respond with an instantaneous deformation and a time-dependent deformation, termed creep. Frozen soils are more susceptible to creep and relaxation due to the presence of ice and unfrozen water where the strength is a function of temperature. The creep response of ice varies with different soils given the potential of ice lens formation. Frozen soil samples will creep under constant axial stress. During creep, the ice content, temperature, time and strain rate will have significant effect on the strength of the frozen ground. Creep strength of a frozen soil is defined as the stress level that can be resisted up to a finite time at which instability occurs. Long-term strength of the frozen material will generally decrease with time and is normally set at the time to reach inflection point on creep curve. Frozen soil generally has a decrease in strength and stiffness from 40 to 60% of the initial value due to creep (Shultz and Hass, 2005). The significance of creep behaviour to the study of frozen ground at the Cigar Lake mine is that the opening for a jet bored cavity or for a tunnel development through frozen ground could squeeze beyond the allowable limit for deformation. Understanding the creep behaviour of the ground in addition to its shear strength behaviour is important. 21  The basic creep curve (see Figure 2.8) comprises three stages, (1) primary (strain-hardening), where the creep rate is decreasing, (2) secondary (linear), where the creep rate is constant, and (3) tertiary (strain-softening), where the creep rate is increasing. Initially, the creep rate decreases with time, thereafter the strain rate increases with time. Eventually, cracks develop in the ice matrix and specimen fails. An increase in axial stress and decrease in temperature cause a decrease in time to failure. The total strain a specimen undergoes consists of the initial and delayed elastic strains and irrecoverable creep strain.  Figure 2.8:  Idealized Creep Curve  Sufficient laboratory testing has established the creep behaviour of frozen soils. The non-linear stress-strain behaviour of frozen soil has been described by Vyalov (1965), Ladanyi (1962), Klein (1978), and Sayles and Haines (1974). Modelling creep behaviour can be done either theoretically based on the quantified physical processes or empirically based on curve fitting. Laboratory testing to monitor creep behaviour has well defined boundary conditions with reasonably uniform stress and strain fields applied to the samples. However, strain rates applied during in situ testing are often higher than those applied in the field or laboratory. In situ testing methods such as pressuremeter testing minimize the effect of sample stress relief and quantify the material properties on a larger scale. The pressuremeter test involves placing an inflatable 22  packer at depth and measuring the volumetric strain and applied pressure to estimate the deformation modulus of the material. The pressuremeter provides an in situ estimate of the shear modulus (G), short term and long term stress-strain relationships, and shear strength parameters. Ladanyi and Johnston (1973) performed pressuremeter testing of frozen ice rich silty soils to establish long term strength parameters. Dusseault and Fordham (1993) note that creep is not typically associated with competent unfrozen sandstone though high porosity poorly cemented sandstones which are the expected rock overlying/comprising the Cigar Lake orebody, may undergo creep due to loading induced grain packing. The transient creep observed in these poor quality sandstones weakens the bonds causing structural collapse. Dusseault and Fordham (1993) comment that there is no widely accepted method of interpretation and analysis for hard and soft rock creep data as the mechanisms and processes equations of the transient state are not clear. Rocks that are most likely to creep are softer, more sensitive, soluble rocks and are often difficult to sample and prepare for laboratory testing.  2.1.3.4 Influence of Hydrostatic and Confining Pressure on Freezing The Cigar Lake orebody is approximately 10 m thick at a depth of 430 m below ground surface. A hydrostatic pressure of 5 MPa is expected on the frozen mass above the orebody. Based on conversations with Cigar Lake Mine, the design freezing thickness above the orebody is anticipated to be two times the thickness of the orebody. The presence of groundwater and in situ stresses will exert hydrostatic pressure on the frozen ground overlying the orebody, resulting in a combined mechanical and thermodynamic effect. The isothermal compression governs the stress and the thermodynamic effect leads to pressuremelting phenomena. Pressure-melting depresses the freezing point of ice that results in water migration toward lower stress regions. When a hydrostatic confining pressure is applied to a frozen granular mass, pressure melting will occur locally at grain-to-grain contacts. A pressure of approximately 13.5 MPa is required to depress the freezing point by 1oC according to the equation, dT/dp = -0.743 K/MPa (Andersland and Ladanyi, 2004). To summarize, a system will require to be lowered by one degree beyond the design temperature to account for 1 atm 23  pressure. Under low confining pressures the stress strain behaviour is brittle in tension and strain softening in compression. The addition of confining pressure in frozen soils suppresses dilation and ice cracking with a noticeable increase in soil strength and decrease in strain softening (Andersland and Ladanyi, 2004). At high confining pressures a second yield occurs. For the second yield, the failure envelope shows a friction angle close to that of unfrozen soil suggesting that the first yield is related to the ice matrix strength and the second yield represents the frictional resistance and residual strength. For clays, the effect of confining pressure on frozen specimens has been noted to be less significant. Sayles (1973) completed triaxial compression tests on saturated Ottawa sand to evaluate the influence of confining pressure under a constant rate of strain and the rate of loading on strength deformation under a constant load. Triaxial tests completed at a constant strain rate of 0.03%/min showed two peaks representing the strength of ice and the second as the internal granular friction. Cohesion and friction were found to be independent of each other after a strain of 0.02%. Chamberlain et al. (1972) found that dilatancy was suppressed at confining pressures higher than 50 MPa on frozen sand mixtures. Chamberlain completed high pressure triaxial compression tests at confining pressures ranging between 3.5 to 280 MPa. Samples were fully saturated and frozen rapidly to -10oC and tested at a strain rate of 6%/min. Three distinct stress regions were observed; a low pressure region of constant or increasing shear stress a mid-pressure region of decreasing shear stress and a high pressure region of slightly increasing shear stress. At confining pressures greater than 52.5 MPa, dilation is completely suppressed indicating crushing of individual soil particles. Pressure melting is suggested to become critical at these confining stresses given the suppression of dilation. Ma et al. (1998) and Wang et al. (2008) describes the strength loss of frozen soil under increasing confining pressure due to pressure melting of pore ice, particle crushing and microcrack growth. The strength of a frozen soil increases to a maximum value with increasing confining pressure as Chamberlain et al. (1972) described above, but decreases beyond confining pressures of approximately 15-45 MPa. 24  Golder (2001) states that ice lens formation is not expected above a confining stress of 1 MPa. 2.1.3.5 Frost Heave Frost heave is the expansion of frozen ground due to the phase change of water to ice in frost susceptible soils. Forces are transmitted from the soil to the overlying foundation and can subject it to large uplift forces (Andersland and Ladanyi, 2004). Heaving results from ice segregation during freezing. In frost susceptible soils, additional water can migrate from unfrozen soil into the frozen zone under a temperature induced suction gradient. Ice lenses form in all soil types by the addition of water during slow moving or stationary freezing fronts. Normally, in coarse non frost susceptible soils such as sands and gravels the pores will fill with ice and excess pore water will drain into the unfrozen areas. Heaving pressures also vary and depend mainly on the type of soil and its moisture content. In general, coarse sands and clean gravels do not heave, while fine sand and silts are very susceptible to heaving. Clays also are very susceptible to heaving although they normally heave slowly but often with tremendous pressures. Silts show a high rate of heave but have much lower heaving pressures than clays. High freezing rates in sands allow excess pressures to build; high freezing rates in silts develops suction and ice lensing parallel to the freezing front; low freezing rates in clay can have reticulate ice lenses which are preferential flow pathways. The highest frost heave (as seen in Figure 2.9) occurs in soil with a permeability of 1x10-6 to 1x10-7 m/s, values typical for silt or silty clay (Shultz and Hass, 2005). The highest frost pressure can occur in clayey soils. The hydraulic conductivity for water in frozen soils is small but not zero and follows Darcy’s law (Burt and Williams, 1976; Lunardini et al., 1982; and Arteau, 1984).  25  Figure 2.9:  Frozen Soil Frost Heave Behaviour, after Shultz and Hass (2005)  2.1.4 Frozen Intact Rock Properties Limited information exists on the behaviour and failure mechanisms of frozen weak rock at depth. The majority of previous research centers on the freezing and thawing of soils, with a smaller number of studies involving massive good quality rock samples. Comparing these, the strength of frozen rock behaves in a similar fashion to frozen soil where the strength depends on interparticle friction, particle interlocking and cohesion. When the sample undergoes freezing the failure mode transitions from plastic to a brittle behaviour due to the conversion of water to ice. Rockmass properties vary with rock temperature and are related to the proportion of ice and unfrozen water. As the temperature drops, mineral grains shrink and the formation of ice in pore spaces contributes directly to the strength of the material. The porosity of a rockmass is considerably lower than a typical soil specimen, and therefore the 26  water content has a reduced influence on the gain in compressive strength with freezing. However, for a weaker, jointed rock mass, such as the Cigar Lake orebody and surrounding host rock, there are more voids and open conduits for water to fill, yielding greater opportunity for strength increase with freezing. Strength values for frozen rock mostly focus on massive, good quality rock, with little jointing or alteration. These studies were performed to support the design and construction of liquid nitrogen storage caverns (i.e. for temperatures below -200oC). It must be emphasized that the strengths involved (> 30 MPa) are not representative of the Cigar Lake material tested as part of this research; unfrozen weak rock typically has zero tensile strength and a compressive strength less than 25 MPa. This research is intended to build on the current knowledge of the influence of freezing on a weakly jointed rock. 2.1.4.1 Compressive and Shear Strength The strength criterion for isotropic rock is commonly defined by the Mohr-Coulomb shear criteria, which is comprised of a cohesion and frictional component. Strength is defined as, the largest stress (load per unit area) a rock can sustain until failure, and can be quantified in the lab on a small cylindrical sample (intact strength) or for a rockmass in the field (rock mass strength). The uniaxial compressive strength (UCS) is a common description or rock strength, with strength then increasing as a function of confining pressure. This relationship (strength as a function of confining pressure) is described by the Mohr-Coulomb relationship. Initial work by Mellor (1971, 1973) measured the uniaxial compressive and tensile strengths of water saturated and air dry granite, limestone, and sandstone rock core from temperatures of 25 to -195oC. Mellor observed that the compressive strengths increase with decreasing temperature. Freezing was noted to increase rock strength by a factor of 4 in porous rock and by a factor of 1.8 in crystalline rock. Figure 2.10 shows compressive strength results, where the gain in strength with decreasing temperature is evident up to -50oC, beyond which little gain in strength is noted. Strength variation can be related to pore-size distribution and freezing characteristics.  27  Figure 2.10: Strength of Granite, Limestone, and Sandstone in Uniaxial Compression, after Mellor (1971)  Further research by Kumar (1968), and Yamabe and Neaupane (2001) indicate a significant strength increase in several rock types with sub-zero temperature. Young’s modulus increases with a decrease in temperature; however, a further decrease in temperature from -10 to -20 C has no effects at all on the Young’s modulus (Yamabe and Neaupane, 2001).  28  Figure 2.11: Summary of Uniaxial Test Results for Unfrozen and Frozen Sandstone, after Yamabe and Neaupane (2001)  2.1.4.1.1 Influence of Strain Rate Gunzel (2008) performed a series of constant strain and constant stress direct shear tests with artificial samples simulating ice-filled rock joints. In constant stress tests, the ice-filled joints show a parabolic relationship between normal stress and shear stress unlike the linear relationship usually found in mineral filled rock joints (Barton, 1974). Unfrozen UCS tests are typically undertaken at strain rates of 10-5 to 10-4 s-1 according to ISRM standards (Brady and Brown, 2006). Very fast or very slow strain rates will influence the peak 29  strength of rock in the same manner as ice, the mode of failure will be brittle under fast loading and ductile under slow loading (<10-8 s-1). Figure 2.12 illustrates the influence of freezing and strain rate on saturated sandstone. With increasing axial strain rates, the gain in strength of frozen sandstone is substantially higher than lower axial strain rates.  Frozen  Unfrozen  Figure 2.12: Axial Stress vs. Axial Strain for Unfrozen and Frozen Sandstone, after Yamabe and Neaupane (2001)  2.1.4.1.2 Influence of Initial Moisture Content When freezing occurs in a partially saturated rock, water will initially migrate toward large 30  empty pores. Inada et al. (1997) and Inada and Kinoshita (2003) completed Brazilian tensile, uniaxial tension, and uniaxial compression tests on tuff, granite, andesite, and sandstone samples at temperatures from 20o to -160oC. At 15°C, the strength for saturated rock samples is lower than that of the strength for dry samples. However, at -160°C the strength for saturated samples is greater than the dry strength due to the conversion of the water in the pores to ice. Also, the tuff having a higher porosity and thus higher moisture content than the granite specimen saw a larger strength increase with sub-zero temperatures. A large change in the tangential Young’s modulus with temperature is not seen in the dry specimens; however, for saturated specimens, Young’s modulus increases significantly with decreasing sub-zero temperatures. Sammis and Biegel (2004) comment on Mellor (1971, 1973) testing data and explain the failure behaviour using a damage mechanics model. In compression, the saturated samples show a stronger increase in strength than the air dry samples. Both air-dry and saturated granite samples strengthen at approximately the same rate. The micromechanical damage model (Ashby and Sammis, 1990) can be used to explain the strength increase difference between porous and crystalline rocks as the rock undergoes freezing. Failure occurs on the sliding of pre-existing cracks in rocks which induces fracture damage and ultimate failure. The damage mechanics explanation of this behaviour is that for saturated samples the frozen water inhibits sliding on fractures and strengthens the sample. The flow strength of ice increases as the temperature falls below the freezing point thus increasing the apparent coefficient of friction and strengthening the samples. For air dry limestone and sandstone samples there is not enough adsorbed water in the pores to provide significant strengthening. For granite, a non-porous crystalline rock, the pre-existing microcracks are narrower therefore there is little difference between air-dry and saturated specimens.  2.1.4.2 Influence of Freezing on Uniaxial Tension Mellor (1973) observed that tensile strength increases with decreasing sub-zero temperatures. Dry rocks gain tensile strength at an average rate of approximately 2 x10-3 MPa/oC with decreasing temperature regardless of rock type and the saturated samples for granite, sandstone 31  and andesite samples show more strengthening at low temperatures than air-dry samples. Figure 2.13 shows tensile strength results by Mellor (1973) where the gain in tensile strength from unfrozen to frozen conditions is significant, but little change in tensile strength with decreasing temperatures beyond -10oC is evident.  Figure 2.13: Strength of Granite, Limestone, and Sandstone in Uniaxial Tension, after Mellor (1971) Dutta and Kim (1993) focussed on testing of tensile failure in their study of limestone and granite samples. Brazilian tensile specimens under quasi-static and dynamic loading were tested between 24 and -40oC. The tensile strength was found to be more sensitive to loading rate than temperature. The samples showed a slightly higher average tensile strength in the frozen specimen compared to that at room temperature. The average tensile strength of the rock samples 32  increased by 0.1% per drop in degree Celsius. Also, the frozen tensile strength of wet specimens increased more than dry specimens at below freezing temperatures. Inada and Kinoshita (2003) explained this noting that for granite, the tensile strength fails along the largest crack which is too large to be saturated compared to the smaller micro-cracks responsible for compressive failure. 2.1.4.2.1 Influence of Temperature Rockmass properties vary with temperature and are related to the proportion of ice and unfrozen water. As the temperature drops, mineral grains shrink and the formation of ice in pore spaces contributes directly to the strength of the material. As noted in Figure 2.11, the compressive strength increases with decreasing temperature, substantially from unfrozen to -50oC, after which the gain in strength is minimal. Chislov (1991) studied the effect of low temperatures on the strength of tuffaceous shales in a highly fractured orebody and concluded that by increasing the ambient temperature from -2 to above 0oC, the rock strength decreased by 20%. Walder and Hallet (1985) present a mathematical model for the breakdown of porous granite and marble by the growth of ice in cracks. The model predicts crack growth rates indicating that sustained freezing is most effective in producing crack growth from temperatures between -4 to 15oC. At higher temperatures, thermodynamic limitations prevent ice pressure from building up significantly and at lower temperatures the migration of water for sustaining crack growth is inhibited. Glamheden and Lindblom (2002) measured frozen rock mass properties and completed numerical modelling for an unlined hard rock cavern measuring 7m diameter and 15 m high in Gothenburg, Sweden. The chamber is located approximately 70 m below ground surface and 30 m below the water table. The rock mass is a medium to fine grained, strong to very strong, nonweathered, gneissic granodiorite. The Q-value is approximately 15, the RMR89 is approximately 75, and the GSI is 67 to 69. After lowering the cavern temperature to -40oC, laboratory testing showed that the tensile strength increases with decreasing temperature and Young’s modulus and Poisson’s ratio marginally increase at decreasing temperature. 33  2.1.5 Creep Behaviour in Weak Rock Creep parameters are determined through loading the sample and testing at specified percentages of the uniaxial short-term compressive strength. The generalized creep equation defines the total strain, ε, composed of the instantaneous strain, εo, and creep strain ε(c). The time dependent frozen compressive strength is calculated following the power law approximations of Hult (1966) and Ladanyi (1972). For the portion of the creep curve at and beyond the inflection point but before tertiary creep, the total strain is defined as, ε = ε(i) + ε (c)mint Where: ε(i) = lumped primary creep of defined by the intersection on the strain axis and is expressed by the power law, εk(σ/σkθ)κ (σkθ is a temperature dependant total deformation modulus) (c) ε min is the rate of steady state creep with time and is defined by the power law, εc(σ/σcθ)n , (σcθ is the temperature-dependent creep modulus) The primary creep law, Andrade's empirical creep law, defines the creep strain as 𝜀 (�) = 𝐴𝜎 � 𝑡 � and re-written by Ladanyi and Johnston (1974) as 𝜎� � 𝜀� 𝑡 � (�) 𝜀� = � � � � 𝑏 𝜎��  Where n, b, and 𝜎�� are three experimentally determined coefficients from creep testing.  Based on similar material creep testing results in Andersland and Ladanyi (2004), typical values of n and b for clay and sand are listed below in Table 2.1 Table 2.1: Frozen Soil Type Bat-Baioss Clay  Values of Parameters in Primary Creep Law Equations, from Andersland and Ladanyi (2004) Source Vyalov, 1962  b  n  0.45  2.50  𝝈𝒄𝜽 (MPa) 0.18  LL, PL  Gs  e, void Ratio  51, 24  2.73  1.045  34  Frozen Soil Type Ottawa Sand  Source Sayles, 1968  b 0.45  n 1.28  𝝈𝒄𝜽 (MPa) 1.05  LL, PL -  Gs 2.65  e, void Ratio 0.587  Two samples of intermediate clay from Golder (1986) laboratory testing program were taken for constant stress creep tests. Each test involved the determination of the steady state strain rate developed when applying two different stresses under unconfined conditions. Testing was conducted at -5oC. Golder (1986) noted that the samples at the highest stress level (1000 kPa) exhibited classical creep behaviour. EBA (1990) completed four frozen creep tests on intermediate clay from boreholes U-8 and U-221 at a temperature of -20oC. Two of the samples from borehole U-8 failed before steady state creep was measured. Table 2.2presents the interpreted steady state creep rate achieved under the applied stress levels; though, EBA (1990) commented that none of the samples achieved a true steady state creep. Table 2.2:  Sample No. Golder, G8-2 Golder, G8-2 Golder, G-41 Golder, G-41 EBA, Hole 221, Depth 454.7 m EBA, Hole 221, Depth 454.7 m  Summary of Creep Testing, after EBA (1990) and Golder (1986)  Moisture Content (%)  Bulk Unit Weight (kg/m3)  Test Temp. (oC)  Applied Stress (kPa)  Steady State Creep Rate (%/min)  Steady State Creep Rate (min-1)  Time to Failure (hours) Onset of Tertiary Creep  18.0  2,142  -5  500  3.56x10-5  3.56x10-7  >120  18.0  2,142  -5  1000  6.40x10-5  6.40x10-7  67  23.6  2,023  -5  500  2.52x10-5  2.52x10-7  >167  23.6  2,023  -5  1000  2.99x10-4  2.99x10-6  67  23.6  1,993  -20  2500  2.24x10-5  2.24x10-7  >72  23.6  1,993  -20  3000  1.42x10-5  1.42x10-7  >170  Mellor and Cole (1981) suggest that the peak stress from a constant strain rate experiment 35  corresponds to the point at which the minimum strain rate occurs on a typical constant stress (creep) strain-time curve. The point on the creep curve and constant strain rate curve, therefore, measures the material (behaviour) under a similar condition although the path to achieve this condition differs. Analyzing EBA (1990) and Golder (1986) frozen test creeping data can be compared to evaluate the flow law of frozen soils. Applying the simplified flow law for frozen soil,  Where 𝜀 � = strain rate σ = applied stress B = temperature dependent coefficient n = exponent (temperature dependent)  𝜀 � = 𝐵𝜎 �  The calculated values of B and n from Golder (1986) and EBA (1990) testing at temperatures of -5oC and -20oC are summarized below. Table 2.3:  Cigar Lake Creep Parameters from Historical Testing  Testing Temperature (oC) -5 -20  B (min /kPa) 1.95 x 10-5 5.13 x 10-14 -1  n 7.08 15.57  Though there are only two sets of creep testing completed to date on Cigar Lake material, the results show the effect of temperature on deformation properties. The colder temperature (-20oC) significantly reduces the deformation rate by several orders of magnitude compared to the warmer temperature (-5oC). For example, an applied stress of 3MPa, would lead to a deformation of 1 x 10-6 min-1 at -20°C while the same stress applied to materials at -5°C would yield a deformation rate of 5 x 10-2 min-1. Given the lack of creep testing on Cigar Lake material at the design freezing temperature of 12oC, the author recommends undertaking creep testing and a test jet boring trial similar to that in 2000; however at the planned design frozen ground temperature of -12oC. 36  2.2  Thermal Properties  The thermal characteristics of the ground are important for thermal analysis to verify the freeze hole layout and ensure an adequately thick freeze wall forms. A ground freezing thermal analysis requires input data referring to geometry, thermal boundary conditions, and material characteristics. The response of a soil to temperature changes is influenced by its thermal properties: thermal conductivity, heat capacity, thermal diffusivity, latent heat, and thermal expansion (Andersland and Ladanyi, 2004). Thermal properties vary depending on the water content. The specific heat, defined as, the amount of heat required to change the temperature of a substance by a given amount, depends on mineral composition and is defined as the ratio of its heat capacity to that of water. Thermal conductivity, defined as a material's ability to conduct heat, depends upon porosity, dry density, degree of saturation, and temperature. Cooling a rock mass shrinks the mineral matrix and induces changes in thermal rock parameters, (Glamheden and Lindblom, 2002). Lindblom (1977) and Aoki et al. (1989) evaluated the decreasing linear thermal expansion coefficient with decreasing temperature. Mellor (1973) evaluated the mean linear expansion coefficient of rock specimens between -10oC and -100oC and -90oC and -160oC to 4.13 x10-6/oC and 3.52 x10-6/oC, respectively. This correlates well with tests done by Kuriyagawa (1980). Lindblom’s test was performed under varying load conditions, but the Mellor and Aoki testing was not. Differences in the results may be due to different test procedures, as Mellor used dilatometers and Lindblom used a strain gauge glued to the rock samples to measure thermal strain. Kuriyagawa et al. (1980) and Aoki et al. (1989) reported that the thermal conductivity at -100oC is up to 10 to 20% greater than at 20oC, with no major difference between dry and wet specimens. Park et al. (2004) completed laboratory tests on dry granite and sandstone from -160 to 40oC. DSC (Differential Scanning Calorimeter) for specific heat, a transient hot-wire method for thermal conductivity, and the strain gauge method for thermal expansion coefficient. Results show thermal conductivity changed little with decreasing temperature. Specific heat and thermal expansion coefficient decreased with decreasing temperature. 37  Frost propagation in a saturated material occurs with a phase change of water to ice and heat transfer due to conduction. In frozen soil, the amount of heat transferred by conduction increases with increasing dry density and degree of saturation. The long term behaviour of frozen ground will be influenced by a thermal gradient from the freeze pipe to the excavation face. As of 2009, all of the previous thermal modelling for Cigar Lake used material properties based on calibrated values obtained from modelling at McArthur River in similar ground types. Measured data from actual rock samples was available from previous McArthur River testing and was used to verify that the trends developed in the dataset were reasonable (Stead and Szczepanik, 1996). Assumptions were made regarding the degree of similarity of ground and amount of water stored within the rock in these ground types. Newman (2007) carried out thermal analyses of the actively freezing production ore zone at Cigar Lake in an attempt to calibrate thermal properties and water contents at different elevations below the ore, within the ore, and just above the ore. Newman (2009) developed a spreadsheet which incorporates the theoretical relationships developed by de Vires (1963) and Johansen (1975) for volumetric heat capacity and thermal conductivity respectively. The calibrated values from the 2007 Cigar Lake model were used as target final property values in the theoretical relationships while other parameters such as rock density, quartz content, and porosity were altered so that the estimated properties matched the calibrated properties. 2.3  Frozen/Unfrozen Interface Behaviour  The separation between the unfrozen and frozen boundary is considered a potential failure mechanism in the back of a jet bored cavity. The back of the jet bored cavity is in horizontally bedded altered sandstone. The potential for the frozen ground to separate at the unfrozen/frozen interface warrants additional testing to be completed on Cigar Lake material, though was outside of the scope of this research. Direct shear testing of the unfrozen and frozen boundary of frozen soils has been completed by Goto et al. (1988) and Thomson and Lobacz (1973). The shear strength at the frozen/unfrozen interface was found to be greater than the shear strength of completely unfrozen soil. The weakest zone lies in the unfrozen zone adjacent to the frozen/unfrozen boundary as it is free from 38  the influence of the suction force at freezing front.  2.4  Mining in Permafrost  Artificial ground freezing to provide groundwater control and excavation support is typically applied in shaft sinking and less commonly in deep underground mines. Mining in the permafrost regions of Canada, Alaska, and the Russian Arctic where the ground is perennially frozen poses technical challenges similar to excavation in artificially frozen ground. Arctic mines are within the continuous to discontinuous permafrost regions where the ground is below 0oC year round in depths up to several hundred meters. Giegerich (1992) reviewed the technical challenges of the Black Angel, Polaris, and Red Dog mines located in the Arctic region of North America. Udd and Betournay (1999) summarize the current literature on the stabilities of openings in frozen ground for mines located in the Arctic regions of North America and Europe. All report a significant loss of strength when the host rock or ore rose above 0oC due to ice melting. Udd and Betournay (1999) conclude that openings in frozen ground allowed for larger excavation spans than under above freezing conditions. Mines operating in permafrost benefit from increased roof stability and reduction of groundwater. However, when the ground temperature increases to greater than -2C significant strength loss occurs causing instability especially where the host rock has been highly altered or decomposed into a soil-like material. Table 2.4 lists mines operating in permafrost regions. The majority of the mines in permafrost conditions operate without ground control issues as long as the openings remain frozen.  39  Table 2.4:  Summary of Relevant Mines in Permafrost  Mine  Location  Mining Method Open pit  Ore asbestos  In Situ Temperature -4.5 to -7oC  Asbestos Hill  Quebec, Canada  Black Angel  Greenland  Zinc  -12oC  NWT, Canada  Room and pillar Open pit  Jericho  Diamond  -  Julietta  Russia  Longhole  Gold silver  and  -  Kupol  Russia  Open pit and  Golder and silver  Lupin  NWT, Canada  Nanisivik  NWT, Canada  Polaris  NWT, Canada  Raglan  Quebec, Canada  Red Dog  Alaska, USA  Longhole open stoping Room and pillar room and pillar and sub-level longhole open stoping with backfill Open pit, cut and fill, and longhole Open pit  Lead  -  Schefferville  Quebec, Canada  Open pit  Iron  -  Shkolnoye/Matrosov  Russia  Shrinkage Stoping  Spitsbergen Store Norske  Norway  Room pillar  and  Comments  Source  Within an increasing Young’s modulus and compression and shear wave velocities a decrease in fragmentation after blasting was noted  Udd and Betournay (1999)  Giegerich (1992) No published data on ground conditions See increase in RMR by 18% due to permafrost  Wardrop (2005)  -  See increase in RMR due to permafrost  Pakalnis (2012)  Gold  -7oC  -  Lead zinc  -10 to -12oC  1-2% ice  Lead zinc  -2oC  Ore exhibits little strength after thawing Pillar stages left open too long cause cracks to form in adjacent pillar stage 5% ice See increase in RMR by 1060%% due to permafrost  Andres (1999) Giegerich (1992)  Ore prevent the ice from thawing it was mined in the winter Significant strength loss in ore above freezing 10% ice See increase in RMR by 13% due to permafrost  Giegerich (1992)  Rock strength properties not significantly influenced by permafrost At thawing bounding water inflow and instability major issue  Myrvang, (1988) and Wandinger (1999)  -  -  Coal  -4oC  Wardrop (2005)  Udd and Betournay (1999) Wardrop (2005)  2.4.1 Case Studies in Frozen Underground Mines Wardrop (2005) in a report prepared for the Kupol mine (Bema Gold Corporation now Kinross) studied the benefit of permafrost to improving ground conditions and its effect on excavation design. The report examines the current exploration core logging data of Kupol and compares the data with other mines in permafrost to establish a base case minimum ground support. Wardrop (2005) states that whether a rock mass is frozen or not, the ground conations after excavation depend more on the characteristics of the fracturing i.e. the block size, shape, and infilling material, than on the intact material properties. For frozen ground the maximum unsupported 40  span is 16 m, though where frozen ground conditions cannot be guaranteed, the recommended stope span is 5 to 6 m (Wardrop, 2005). Pakalnis (2012) visited the Kupol mine and commented the following: •  Areas visited including the 455 level noticed significant improvement from the unfrozen RMR76 of less than 25 observed in the drill core compared to the frozen face RMR76 of 60. Spans excavated were typically 6 m.  •  The freezing assists the overall stability in the operation and should be considered as augmenting the ground support in place, but not replacing the support.  The following summarizes the improvement in rock mass quality due to freezing at several Russian underground mines in permafrost. Caution should be used when comparing the data from case studies, as the improvement in RMR from unfrozen to frozen conditions assessed by Wardrop (2005) assumed that the increased span opened in frozen conditions is relatable to a frozen RMR by the Grimstad and Barton (1993) chart. Better practice is to assess the frozen RMR conditions in the field with face mapping and to compare the unfrozen RMR conditions using geotechnical core logging. Note that the Russian case studies presented in Wardrop (2005), did not observe the unfrozen RMR conditions at the exposed face.  2.4.1.1 Shkolnoye/Matrosov Mine The Shkolnoye/Matrosov Mine is located in northeastern Siberia, Russia. The mine is entirely located within the permafrost zone. Wardrop (2005) states the following: •  •  The average hanging wall conditions without benefit of permafrost are classified as good rock mass quality according to Barton’s Q’ and Bieniawski's RMR, where Q’ = 17.8 and RMR = 70. Based on the empirical support design chart (Grimstad and Barton, 1993), relating Q and excavation span and to recommended support requirements, with an ESR value of 5 for temporary mine openings, the resulting maximum span is 35 m. However, the mine has 50 x 50 m shrinkage stope panels that are stable. The exceeded maximum span predicted by empirical methods is attributed to permafrost. 41  •  A back analysis of the minimum rock mass condition required to support a 50 m stable span relates to a minimum increase in rock mass quality of 13% from the unfrozen RMR value.  2.4.1.2 Julietta Mine The Julietta Mine is located in the Magadan region of Russia. The mine is entirely located within the permafrost zone. Wardrop (2005) states the following: • •  •  The ground conditions without benefit of permafrost are classified as poor rock mass quality according to Barton’s Q’ and Bieniawski's RMR, where Q’ = 3.4 and RMR = 55. Based on the empirical support design chart (Grimstad and Barton, 1993), relating Q and excavation span and to recommended support requirements, with an ESR value of 1.6 for permanent mine openings, the resulting maximum span is 5.6 m. However, the mine had 8m stable spans on the 745m and 850m levels. The exceeded maximum span predicted by empirical methods is attributed to permafrost. A back analysis of the minimum rock mass condition required to support a 8 m stable span relates to a minimum increase in rock mass quality of 18% from the unfrozen RMR value.  2.4.1.3 Raglan Mine The Raglan Mine is located in the Nunivak region of northern Quebec, Canada. The mine is entirely located within the permafrost zone. Wardrop (2005) states the following: •  •  •  •  KW 1475 Stope - the ground conditions without benefit of permafrost are classified as poor rock mass quality according to Barton’s Q’ and Bieniawski's RMR, where Q’ = 1.5 and RMR = 47. The excavation was stable in frozen conditions up to a span of 50 m. A back analysis of the minimum rock mass conditions required to support a 50 m stable span relates to a minimum increase in rock mass quality by 70-80%. C 1460 L Cut - the ground conditions without benefit of permafrost are classified as fair to good rock mass quality according to Barton’s Q’ and Bieniawski's RMR, where Q = 10 and RMR = 65. The excavation was stable in frozen conditions up to a span of 40 m. A back analysis of the minimum rock mass conditions required to support a 40 m stable span relates to a minimum increase in rock mass quality by 13-18%. Q 1350 Cut - the ground conditions without benefit of permafrost are classified as fair rock mass quality according to Barton’s Q’ and Bieniawski's RMR, where Q = 7.5 and RMR = 62. The excavation was stable in frozen conditions up to a span of 35 m. A back analysis of the minimum rock mass conditions required to support a 35 m stable span relates to a minimum increase in rock mass quality by 13%. The estimated difference between the frozen and unfrozen rock quality is a factor of 15 or more. 42  •  Permafrost provides a greater percentage of improvement for weaker ground conditions than for stronger ground conditions. This relationship decreases exponentially with improving ground conditions  2.4.2 Case Studies in Frozen Soil and Ice Deposits Russian and U.S. researchers have examined the stability of underground excavations in frozen soil deposits and ice. Underground of mining frozen gravel deposits in Alaska, Yukon, and Russia have all been developed using variations of the room and pillar method and are typically less than 100 m in depth. The properties of frozen gravel and silt depend on many variables including in-situ temperature, ice content, particle size and composition, and stratification. Nelson (2001) commented that the most important characteristic of frozen placer materials is their tendency to creep and to exhibit considerable deformation before failure. 2.4.2.1 Fox Tunnel, Alaska The Fox tunnel located near Fairbanks, Alaska was excavated in warm ice-rich silt. The test area geology is comprised of 15 to 20 m of silt overlying 1.5 to 4.5 m of Wisonconsin gravel and schist bedrock. Rooms measuring 4.6 x 15.2 x 2.4 m at 15 to 20 m depth were excavated in frozen gravel with successive slabs taken off the back. Gravels are several meters thick in the back and sidewalls. Excavations were noted to deform considerably by plastic flow when kept at the original ground temperature of -1.1oC to -0.6oC due to the high ice content of the silt. Roof subsidence was monitored to measure the flowing of the overlying silt. Pettibone (1973) concluded that the creep of the frozen silt could be reduced with circulation of cold air to cool the tunnel walls. Microseismic monitoring to detect unstable roof conditions did not monitor any noises during removal of the jacks supporting the roof. Pettibone (1973) theorized that the deformation observed was due to creep of ice matrix and not fracturing. Weerdernburg and Morgenstern (1984) analysed the in situ deformation behaviour of the Fox Tunnel in Alaska showing that the flow law for polycrystalline ice does not yield an upper bound to the observed room closure measurements. The tunnel closure is believed to be from creep and plastic yielding.  2.4.2.2 Dome Creek Drift, Alaska The Spokane Research Center of the U.S. Bureau of Mines conducted a ground stability analysis 43  to understand the behaviour of underground openings in permafrost at a small underground placer mine, the Dome Creek Drift Mine, northeast of Fairbanks Alaska. Mining induced stresses and displacements in frozen gravels were monitored over nine months from 1993 to 1994 in a retreat room and pillar section. The depth to bedrock near the instrumented site is less than 46 m. Biaxial stressmeters, two-point horizontal extensometers, two-point vertical extensometers, string potentiometers, manual closure point stations, convergence meters, and temperature sensors were installed at various locations in a retreat room and pillar section of the mine. Measurements in the Dome Creek Drift Mine showed that roof to floor closure depended on the width of the entry, proximity to active mining, and elapsed time. The roof usually moved as a unit creeping slowly into the entry until slabs developed along silt layers or other planes of weakness. Closure occurred slowly and predictably (Seymour et al., 1996 ). The overlying frozen gravels exhibited mass flow behaviour slowly creeping in to the mine openings until roof slabs separated under their own weight or along planes of weakness such as interbedded silt horizons. 2.4.2.3 Greenland The U.S. Army Cold Regions Research and Engineering Laboratory (CRREL) monitored three excavations in glacial ice in the Greenland Ice Cap to assess the feasibility of tunnels and rooms in ice for storage (Abel, 1961; Russel, 1961). The second tunnel excavation closed due to unpredicted excessive deformations of the openings attributed to warming of the ice. 2.4.3 Ground Control of Frozen Placer Deposits Bandopadhyay et al. (1996) completed finite element analyses of roof to floor convergence for simulated entries in frozen gravel accounting for both material characteristics and heat transfer. The relationships between opening convergence, time and span were investigated based on a finite-element analysis of the thermo-elasto plastic creep of frozen gravels. Analysis shows that a linear relationship exists between span and roof-floor convergence at different times. Roof floor convergence was modelled at air temperatures of -0.6oC and -2.8oC. The higher the air temperature, the larger the roof deflection as a greater difference between the air and ground temperatures result in more intensive heat exchange. A higher temperature within the rockmass means a larger decrease in Young’s modulus and greater increase in Poisson’s ratio. With increasing rock temperature the modulus of elasticity of the rockmass decreases and the rockmass shows plastic behaviour. 44  Soviet researchers defined stability classes for excavations in frozen placer materials (Nelson, 2001). The stability class includes recommended stable spans and pillar sizes from shallow excavations in frozen gravel and silt material. Summarized below in Table 2.5. Implications of the classification system for the maximum span of excavations in frozen soil to the Cigar Lake mine is to provide a basis for maximum spans given the lack of available data for mining in frozen weak rock. Table 2.5: Soviet Classification of Frozen Intermediate Roof Materials Up to 15 m Thick and Stable Spans after Extraction, after Emelanov et al. (1982)  Stability Class I. Highly stable  II. Stable  III. Medium stable  IV. Poorly stable  V. Unstable  Composition Alluvial coarse grained deposits Alluvial and lacustrine deposits Sandy and loamy deposits Homogeneous silty and clayey deposits Alluvial and lacustrine sediments with interbedded fine layers  Alluvial and lacustrine coarse grained deposits Gravelly sand deposits Homogeneous silty and clayey deposits Interbedded clay and gravel deposits  Ground ice Alluvial and lacustrine coarse grained deposits Gravelly sand deposits Homogeneous silty and clayey deposits Interbedded clay and gravel deposits Loess silt and clay Ice rich silt Ground ice Plastic alluvial Plastic alluvial with sandy matrix Unconsolidated and cemented by ice Ground ice  Temperature, (oC) < -6 -6 to -3 < -3 < -4 < -3  -3 to -2 -4 to -3 -2 to -1 < -6  Ice Content, (wet wt.%) < 25 < 25 < 25 25 to 50 < 25 for coarse grained 25 to 50 for fine grained < 25  -2 to -1  25 to 50 25 to 50 < 25 for coarse grained 25 to 50 for fine grained > 60 < 25  < -6 -2 to -1 -2 to -1 -3 to -1.5 -6 to -3 -6 to -3 > -1.5 > -1 Any > -3  25 to 60 25 to 50 25 to 50 25 to 50 > 50 > 60 < 50 < 50 <3 > 60  Thickness of affected strata (m) for Monolithic Roof 14-20 13-16  Maximum span of openings for Room and Pillar Monolithic Roof 35-45 25-35  10-13  30-25  7-10  10-15  4-7  6-10  45  2.5  Weak Rock Mass Behaviour  The Cigar Lake orebody is located at an unconformity between sandstone and basement metapelite rock comprising very weak rock to soil like material. Above and below the unconformity, the rock mass shows variability for tens of meters in porosity and permeability due to fracturing and alteration processes. Rock mass classification and geotechnical domains of the ground conditions at the Cigar Lake mine were completed previously by the mine and their consultants though are lacking detail on the properties of the weakest material, typically the ore and overlying clay altered sandstone. Understanding the behaviour of unfrozen weak ground is in itself a challenge, and therefore this section focuses on the behaviour of weak rock and establishing failure criteria, and modifying these classification systems for frozen weak rock. The process for designing excavations in hard rock masses is well established in geotechnical literature. Excavation through weak rock masses requires a more thorough design as squeezing and/or instability are common. Weak rock masses result from processes such as alteration and faulting creating low strength, sheared, and crushed material with a loss of any interlocking structure which may have existed. Weak rocks are often overstressed at low stress levels as a result of their low strength and high deformability. These characteristics can lead to yielding, slabbing, spalling, ravelling, and squeezing conditions. Several different authors have defined conditions under which they would consider a rock mass to be weak: •  Hoek (1999) defines weak rock as that where the in-situ uniaxial compressive strength (UCS) is less than about one third of the in situ stress acting upon the rock mass.  •  The ISRM (1981) defines a rock mass with a UCS between 0.25 to 25 MPa as being weak to extremely weak.  •  Robertson (1988) defines a weak rock as any rock mass where the Mohr-Coulomb effective shear strength parameters are less than c’=0.2 MPa and φ’=30o, which is equivalent to a UCS strength of less than 0.7 MPa.  •  Pakalnis (2008) defines a weak rock mass as that with a rock mass classification value (RMR76) less than 45.  46  The material providing support of a jet bored cavity above the Cigar Lake orebody is considered a weak rock mass ranging from a dense/indurated clay to weak and altered sandstone. The following sections outline rock mass classification schemes and their application to weak rock masses. 2.5.1 Rock Mass Classification Systems Rock mass classifications systems are useful as a quick assessment of the rock mass conditions for support design and stability assessment. A rock mass rating is determined by assigning numerical values to features that are considered to influence its behaviour, and combining these into an overall rating. Rating values have subsequently been correlated with the observed stable spans of unsupported excavations, stand-up times of unsupported spans, support requirements for various spans, cavability, and pit slope angles (Brady and Brown, 2006). Terzaghi (1946) was the first to develop a rock mass classification system originally for the estimation of loads to be supported in the design of steel arches for tunnel construction. Terzaghi’s classification terms are very subjective descriptions of the rock mass. The two most common rock mass classifications systems are the CSIR Rock Mass Rating (RMR) by Bieniawski (1976, 1989) and the NGI Tunnelling Index (Q-System) by Barton et al. (1974). More recently, the GSI system (Hoek et al., 1995) was developed as a visual extension of the RMR based on geological observations of the size and shape of intact rock blocks (blockiness) and surface condition of the discontinuities. Weak rock masses are complex and have highly variable properties in stiffness, strength, and failure modes that lead to difficulties in applying classification systems. Classifications such as RMR and Q were created for jointed rock masses whose behaviour is controlled by discontinuities and do not specifically address unique characteristics of weak rocks such as overstressing or deterioration. In poor rock conditions, even though the rock masses have similar rock mass classification values, the failure modes, and rock support requirements were very different due different degrees of interaction between the intact rock and discontinuities (Mathis and Page, 1995). Comparing the RMR and Q-system, both methods incorporate geological, geometric and design/engineering parameters in arriving at a quantitative value of their rock mass quality (Hoek, 2007). Rock mass classification values are dependent on input parameters such as the intact rock strength, rock quality designation (RQD), joint spacing, joint alteration, and 47  groundwater condition. Both RMR and Q can be adjusted to account for the relative gain in strength from unfrozen to frozen weak rock for subsequent use in design for an ice cap overlying a mined out cavity. 2.5.1.1 Rock Quality Designation The Rock Quality Designation (RQD) created by Deere (1964), is a quantitative index of rock mass quality based upon rock core recovery by diamond drilling. RQD is defined as the percentage of core recovered as intact pieces of 100 mm or more in length relative to the total length of the core run. Mechanical breaks due to drilling, handling, or high stress are ignored as only natural core breaks are considered in this calculation. Core with an estimated unconfined compressive strength less than 1 MPa (ISRM rock hardness less than R1) are not be included in the RQD and should be assigned an RQD of zero. 2.5.1.2 Geomechanics Classification System (RMR) Bieniawski (1974, 1986) introduced a geomechanics’ classification system for rock masses based on experiences in South African tunnelling projects in 1973 with revisions in 1976 and 1989. The Rock Mass Rating (RMR) is the sum of six rock mass rating parameters: uniaxial compressive strength, RQD, joint spacing, joint condition, groundwater condition, and joint orientation. Ratings are assigned to each of the weighted parameters and the sum of these ratings defines the RMR and rock mass quality. RMR values range from zero to 100, indicating extremely poor rock to extremely good rock, respectively. Table 2.6 lists the parameters and their assigned rating values for Bieniawski's 1976 version. The biggest difference between RMR76 and RMR89 is in the joint condition description and ratings, but there are also slight changes to the UCS and joint spacing values and ratings.  48  Table 2.6:  1  1976 Rock Mass Rating Classification Scheme, from Bieniawski (1976)  Parameter point load strength index Strength of intact rock uniaxial material compressive strength Rating  2  Drill core quality RQD Rating  3  Spacing of Joints Rating  Condition of Joints 4  Rating  Groundwater 5  Inflow per 10 m tunnel length Ratio General Condition s Rating  Range of Values For this low range uniaxial compressive test is performed  > 8 MPa  4-8 MPa  2-4 MPa  1-2 MPa  > 200 MPa  100-200 MPa  50-100 MPa  25-50 MPa  10-25 MPa  3-10 MPa  1-3 MPa  15  12  7  4  2  1  0  90-100%  75-90%  50-75%  25-50%  < 25%  20  17  13  8  3  > 3m  1-3 m  0.3-1 m  50-300 mm  < 50 mm  30  25  20  10  5  very rough surfaces hard joint wall rock not continuous no separation  slightly rough surfaces hard joint wall rock separation < 1 mm  Slightly rough surfaces separation < 1mm soft joint wall rock  Slickenside d surfaces or Gouge < 5 mm thick or Joints open 1-5 mm Continuous Joints  Soft gouge > 5 mm or Joints open > 5 mm Continuous joints  25  20  12  6  0  None  < 25 l/min  25-125 l/min  > 125 l/min  0  0.0 - 0.2  0.2 - 0.5  > 0.5  Completely Dry  Moist only  Water under moderate pressure  Severe Water Problems  10  7  4  0  The RMR76 (Bieniawski, 1976) classification system is calculated as follows: RMR76 = P1 + P2 + P3 + P4 + P5 Where:  P1  is the strength of intact rock material (rating = 0 to 15);  P2  is the drill core quality, Rock Quality Designation, RQD (rating = 3 to 20);  P3  is the spacing of joints (rating = 5 to 30);  P4  is the condition of joints (rating = 0 to 25); and,  P5  is the groundwater (rating = 0 to 10). 49  The rock mass conditions can be classified follows: Class I – Very Good Rock (RMR > 80); Class II – Good Rock (60 < RMR < 80); Class III – Fair Rock (40 < RMR < 60); Class IV – Poor Rock (20< RMR < 40); and Class V – Very Poor Rock (RMR < 20). 2.5.1.3 Rock Tunnelling Quality Index, Q The Rock Tunnelling Quality Index (Q) was developed by Barton et al. (1974) for the determination of rock mass characteristics and tunnel support requirements based on hard rock tunnels in Scandinavia. The Q rating varies on a logarithmic scale from 0.001 (exceptionally poor) to greater than 400 (exceptionally good). The Q rating is based on six parameters: RQD, number of joint sets (Jn), joint roughness (Jr), joint alteration/infilling (Ja), water (Jw), and stress reduction factor (SRF). Use of the Q system for mining applications will give conservative answers (Potvin, 1980) as it was designed and is used for civil applications. Q rating is calculated using the following equation:   RQD   J r   J w   ×   ×  Q =   J  n   J a   SRF  Where each parameter relates to: RQD/Jn = measure of the block size Jr/Ja = roughness and frictional characteristics of joint walls or infilling; shear strength Jw/SRF = two stress parameters; active stress  Table 2.7 lists the input parameters for Jn (number of joint sets), Jr (joint roughness parameter), and Ja (joint alteration).  50  Table 2.7:  Q Rating Parameters, from Barton et al. (1974) Description  Massive, no or few joints One joint set One joint set plus random Two joint sets Two joint sets plus random Three joint sets Three joint sets plus random Four or more joint sets, random, heavily jointed, "sugar coated" Crushed rock, earth-like  Jn 0.5 to 1.0 2 3 4 6 9 12 15 20  Infill Thickness  Description  Jr  Discontinuous joints Rough and irregular, undulating Smooth, undulating Slickensided, undulating Rough or irregular, planar Smooth, planar Slickensided, planar Zones containing clay minerals thick enough to prevent wall contact Sandy, gravelly, or crushed zone thick enough to prevent wall contact  4 3 2 1.5 1.5 1 0.5  none  Tightly healed, hard, non softening, impermeable filling Unaltered joint walls, surface staining only  < 1mm  Description  Slightly altered joint walls, non-softening mineral coatings, sandy particles, clay-free disintegrated rock  > 2mm but < 5mm  < 2mm  Silty or sandy clay coatings, small clay fraction (non-softening)  1  Ja 0.75 1 2 3  Softening or low friction clay mineral coatings, I.e. kaolinite, mica, chlorite, talc, gypsum, graphite, and small discontinuities of swelling clay (discontinuous coatings, 1-2mm or less in thickness)  4  Sandy particles, clay-free disintegrated rock  4  Strongly overconsolidated, non-softening clay mineral fillings (continuous <5mm thick) Medium or low over consolidated, softening clay mineral fillings (continuous <5mm thick) Swelling clay fillings (continuous > 5mm thick) Values of Ja depend upon percent of swelling clay-sized particles, and access to water. Zones or bands of disintegrated or crushed rock and clay * Strongly over consolidated, non-softening clay * Medium / low over consolidation, softening clay * Swelling clay (i.e. montmorillonite)  ≥ 5mm  1  Zones or bands of silty clay or sandy clay, small clay fraction, non-softening. Thick continuous zones or bands of clay *Strongly over-consolidated, non-softening clay *Medium / low over-consolidation, softening clay. *Swelling clay (i.e. montmorillonite)  6 8 812 6 8 812 5 10 13 624 624  The classification ratings for the Q’ values are as follows: Class I – Very Good Rock (40 < Q’ < 100); Class II – Good Rock (10 < Q’ < 40); Class III – Fair Rock (4 < Q’ < 10); Class IV – Poor Rock (1 < Q’ < 4); and Class V – Very Poor Rock (Q’ <1). 51  2.5.1.4 External Factors and RMR’ and Q’ Calculations RMR’ and Q’ are modified versions of the RMR and Q that assumed dry conditions and exclude the SRF term (RMR does not have a stress parameter). This is done for the purpose of assessing the rock mass ratings in the absence of external factors, where these may be accounted for in separate calculations. For example, groundwater and in situ stresses are sometimes better accounted for using numerical modelling methods, but RMR and Q may still be required to estimate the rock mass properties to provide model input; pore pressures and in suit stresses are not properties of the rock mass. RMR76’ is calculated using the first four terms; the rock mass is treated as if it were completely dry and a groundwater rating of 10 is assigned. Very favourable joint orientations should be assumed and the Adjustment for Joint Orientation value should be 0. The Q’ value was defined according to the following formula, without any correction for external influences such as stress or water conditions (i.e. Jw = 1 and SRF = 1).  Q' =  RQD Jr × Jn Ja  Again, RMR’ and Q’ should only be used where the design procedure specify their use. Where water pressures or high in situ stresses are present, these should be accounted for either empirically or numerically. 2.5.1.5 Discussion Milne (2007) discusses issues with rock mass classification systems that arise when the same rock mass can yield different classification values depending on subjectivity in assessing the joint orientation, stress conditions, drift orientation, depth, and excavation history. The Q-system can differentiate between more than 60 conditions of joint surfaces making repeatability an issue. As previously noted, typically groundwater and stress factors are omitted to obtain rockmass properties for the purpose of numerical modelling and analysis. These are accounted for explicitly in the design calculations. Both the Q-system and RMR system were not developed to specifically address weak rock conditions, though both have been modified over the years to account for a wider range of rock 52  mass conditions. This was one of the objectives of Marinos and Hoek (2002) in their development of the Geological Strength Index (GSI) system to visually classify rock masses.  2.5.1.6 Relating Q and RMR Based upon 111 case histories, Bieniawski developed a relationship between RMR and Q (Bieniawski, 1976).  The Q-value is related to Bieniawski’s RMR value using one of the following equations:  RMR = 9 ln(Q) + 44 or,  Q = 10  RMR − 44 21  2.5.2 Modification of Rock Mass Classification Systems for Frozen Ground Rock mass classifications are used to estimate rock mass behavior, excavation stability, and provide ground support guidelines (Milne et al., 1998). Establishing unfrozen and frozen rock mass rating values for various material types can be used to understand the influence of freezing on the empirical rock mass rating and stable open span relationships for underground cavities. The two main rock mass classification systems RMR and Q were developed for unfrozen rock masses. Both systems have similar input parameters for rock strength, RQD, joint condition/alteration, joint spacing, and water. When a rock mass undergoes freezing, some of these parameters will be influenced by freezing and others will not. Specifically, the influence of water freezing in joints and whether this can be treated as a healing of a joint is an obvious starting point. Building on this, the time span the excavation is expected to remain open, long term (months) or short term (days), will influence whether a frozen joint should be counted for as increasing in the rock mass quality. The development of a frozen rock mass rating system and its application as an empirical approach for ground control in frozen ground is discussed in detail in Section 8.  53  2.5.3 Rock Mass Strength The strength of an intact rock sample compared to that of a jointed rock mass varies considerably due to scale effects. The intact rock is the strength of a point sample measured by Unconfined Compressive Strength (UCS) testing, a sample typically measuring 2” in diameter by 6” in length. Compared to the rock mass strength, which encompasses the discontinuities, and is influenced by spacing, infilling, and the compressive strength of the rock. Figure 2.14 (after Wyllie and Mah, 2007) depicts the transition due to scale effects from intact rock to the rock mass strength with increasing sample size and influence of jointing.  Figure 2.14: Scale Effects, Intact Rock to Jointed Rock Mass, after Wyllie and Mah (2007)  Hoek and Brown (1980) developed a shear strength criterion for the rock mass based on a backanalysis of fractured rock masses for the design of underground excavations in hard rock. The criterion was initially based on the properties of the intact rock, and then included the properties and characteristics of the joints in the rock mass. The generalized Hoek-Brown Failure Criterion (Hoek, 2006) for jointed rock masses is defined by: 𝜎��  =  𝜎��  𝜎�� + 𝑠� + 𝜎�� �𝑚� 𝜎��  �  54  Where 𝜎�� = Maximum effective principal stress at failure 𝜎�� = Minimum effective principal stress at failure 𝑚� = Value of the Hoek-Brown constant m for the rock mass s and a = Constants which depend on the rock mass 𝜎�� = Uniaxial compressive strength of the intact rock pieces Estimating the strength of the rock mass, an interlocking matrix of discrete blocks, with laboratory testing has been found to not be practical, and needing to rely on visual observations (Hoek, 2006). Marinos and Hoek (2000) developed the Geological Strength Index (GSI), a visual assessment tool for jointed rock masses to estimate the rock mass strength (Figure 2.15).  55  Figure 2.15: GSI Values for Blocky Rock Masses, after Marinos and Hoek (2000)  The GSI provides a system for estimating the reduction in rock mass strength for varying geological conditions (Hoek, 2006). The GSI value is related to the degree of fracturing and the condition of the fractures. Higher GSI values represent very good quality rock masses where low GSI values represent very poor quality rock mass conditions. 56  The influence of freezing on jointed weak rock mass will be investigated in this thesis with frozen UCS, direct shear, and four-point beam testing. However, the overall gain in strength due to freezing is believed to have a greater impact on the rock mass, which can be estimated using the GSI chart and Rock Mass Rating (RMR) system.  2.6  Failure Mechanisms in Frozen Stratified Ground  Failure of an underground rock excavation is influenced by stress, structure, and the rock mass. In weak rock, stress induced failures are not a concern due to the yielding nature of the rock mass. Weak rock mass failure is typically due to the overall degradations of the rock mass and mobilization of friction (in contrast, failure of strong brittle rock is driven by cohesion loss). The failure surface through weak rock is a complex combination of failure through soft intact rock along weak joints and through soil like weathered zones. Robertson (1988) states that where the RMR is greater than 40 the stability will be determined by the orientation and strength along discontinuities, and when the RMR is less than 30, failure may occur through the rock mass at any orientation. Structural features can control the stability of excavations at shallow depths and in de-stressed areas. Structurally controlled failures occur when features such as joints, bedding, or faults intersect to form blocks or wedges that can slide or fall due to gravity. The Cigar Lake orebody is hosted in sandstone with joints parallel to bedding and random subvertical fractures due to cross-jointing and faults. Jet bored cavities will be excavated in frozen, medium strong, pitchblende rock overlain by several meters of frozen very weak, jointed, sandstone to dense clay. Potential failure mechanisms of an excavated cavity include the separation between unfrozen and frozen material in the back of the cavity and cracking of the ice matrix due to the larger stresses on the roof beam. Ice-filled rock joints are a potential plane of weakness in the frozen rockmass depending on the aperture and infilling of the joint prior to freezing. Parallel laminations and stratifications can be a dominant factor controlling stability of roofs in large excavations. In stratified ground the only load acting on the detached strata is the beams own weight. Underground openings in bedded rocks can expect to develop an arch structure in the back of the opening and at a small scale the immediate roof deflects downward as a beam.  57  2.6.1 Beam Theory The traditional approach to understand stability in stratified ground is to model the immediate roof as if it were a beam. Beam theory assumes that the immediate roof can be represented by a series of equal width beams, with a length equal to the room span. The stable roof span is designed for the allowable tensile stress of the roof strata. A beam is capable of carrying loads in bending as it applies loads transverse to its longest dimension. Beam bending induces failure by flexure as the rock mass can separate at bedding planes due to deflection. Simple beam testing is commonly used to determine the first crack strength and flexural strength of concrete or fibre reinforced concrete. Two loading methods are practiced on beams supported on two outer points, i) third-point loading, termed center point loading by the ASTM, and ii) four-point loading, termed third-point bending by the ASTM. In third-point beam bending the entire load is applied at the center of the span and the maximum stress concentrates in the center part of the beam. Four-point beam bending applies two concentrated loads on top of the beam with the maximum stress located at each point load. Four-point loading calculates the flexural strength assuming that the fracture initiates at the center of the beam. If fracture occurs outside the maximum moment region greater than 5% of the span length the strength results are considered to be invalid. Under third point or center point loading the location of the fracture is not an issue as fracture at a location other than mid-span corresponds to a lower extreme fibre stress than exists at mid-span as the bending moment varies linearly from zero at the support to maximum at mid-span. Four-point beam bending is recommended for testing frozen weak rock behaviour because Goodman (1988) states that fourpoint testing yields better reproducibility of results than three-point loading. A beam section is expected to crack for the first time when the stress reaches the value of the modulus of rupture. Mechanical properties of the beam can be characterized by peak load, first crack load associated with crack deflection and residual flexural load. The flexural strength also termed ‘modulus of rupture’ is the maximum tensile stress on the bottom of the specimen corresponding to peak load and is calculated using simple elastic beam theory. Typically the flexural strength is two to three times the rock specimen’s tensile strength under four-point loading (Goodman, 1989). If the material is homogeneous, tensile strength and flexural strength would be equivalent. 58  No ASTM or ISRM standard exists on beam testing of cylindrical rock core. Related standards included ASTM standards for concrete and fibre reinforced concrete and an ISRM standard on notched rock core specimens under four-point loading to estimate the fracture toughness.  •  ASTM C 78-02 - Standard Test Method for Flexural Strength of Concrete (Using Simple Beam with Third-Point Loading).  •  ASTM C 293-08 - Standard Test Method for Flexural Strength of Concrete (Using Simple Beam with Center-Point Loading).  •  ASTM D 1635 - Standard Test Method for Flexural Strength of Soil-Cement Using Simple Beam with Third-Point Loading.  •  ASTM C 1018-97 - Standard Test Method for Flexural Toughness and First-Crack Strength of Fiber-Reinforced Concrete (Using Beam with Third-Point Loading).  Concrete beam testing methods for the purpose of establishing the beams ability to resist slab failure under bending follow ASTM C 78 or ASTM C 293-08 which determines the flexural strength of concrete using a simple beam with third point loading. Modification of standard concrete beam testing is part of the first phase of frozen beam laboratory testing to gain an understanding on a controlled frozen sample prior to testing the rock collected from Cigar Lake. The rock core sampled from the 2009 diamond drilling of surface freeze holes at Cigar Lake will be tested with a four point beam apparatus to determine the failure mechanisms of frozen jointed weak rock mass. The failure of a rock beam through four-point loading allows for a simple and repeatable flexural test. Four-point flexural loading on a rock beam with the bottom of the core supported on points near the ends and the top of the core loaded from above yields better reproducibility of results than three-point loading (Goodman, 1989). The tests setup is illustrated in Figure 2.16. The modulus of rupture for four-point loading of cylindrical rock specimen with loads applied at L/3 from each end and reactions at the ends is defined as TMR = 16PmaxL / 3πd3 (Goodman, 1989). 59  Figure 2.16: Four Point Beam Bending Load Test  Where Pmax = maximum load L = length between load reactions on the lower surface d = core diameter  2.6.2 Voussoir Analogue The failure of underground openings in stratified ground has been observed to not fail acting as a simple beam, but rather composed of individual rock blocks (Sofianos, 1996). These blocks formed by transverse discontinuities cutting bedding are termed “voussoirs”. The development of tensile cracking or discontinuities normal to the beam inhibits the tensile capacity of the beam creating a compression arch from the abutments to a highpoint at midspan. Different voussoir beam models and failure criteria have been proposed by Brady and Brown (2006), Sofianos (1996), and Diederichs and Kaiser (1999). Voussoir beam theory states that in a confined situation the ultimate strength of a beam is larger than its elastic strength. A beam will develop a compressive arch carrying its own weight and transmitting it to the abutments with an assumed linearly varying load distribution, resulting in a stronger beam assuming Voussoir conditions exist.  2.7  Span Design of Underground Excavations  Failure of a rock mass is influenced by the size of the opening, structures, and rock mass strength. Empirical relationships relating rock mass quality and underground span opening have been developed based on past performance in underground mines and excavations. The term 60  “critical span” used by design methods/graphs refers to the largest circle that can be drawn within the boundaries of the excavation when viewed in plan.  2.7.1 Critical Span Empirical Chart The critical span curve (Figure 2.17) developed by Lang (1994) provides a relationship between span and the RMR rock mass quality to evaluate the back stability in cut and fill mines. The graph is divided into three areas: stable, potentially unstable, and unstable. These are characterized as follows: 1) Stable Excavations a. No uncontrolled falls of ground b. No observed movement in the back c. No extraordinary support measures implemented 2) Potentially Unstable Excavations a. Extra ground support has been installed to prevent potential falls of ground b. Movement in the back of 1mm or more in 24 hours has been observed (Pakalnis, 2002) c. Increase in the frequency of popping and cracking indicating ground movement 3) Unstable Excavations a. Area has collapsed b. Support was not effective in maintaining stability  61  Figure 2.17: Critical Span Curve, after Lang (1994) Figure 2.17 is a simple and useful tool that aids in the design of underground man-entry openings later updated by Wang (1999). The updated span design curve chart has uncertainties below RMR76 values of 50 and above RMR76 values of 80 due to the lack of data in the very poor quality and good to excellent quality rock masses. Ouchi (2005) updated the critical span curve after Wang (1999) to include additional data points in weak rock, specifically for RMR76 less than 50. These are shown at the lower RMR76 range, marked by “green lines” in Figure 2.18, where points in the previously defined unstable one were shown in mining operations remain stable with only local support.  Figure 2.18: Weak Rock Mass Critical Span Curve, after Ouchi et al. (2004)  Pakalnis (2012) with the support of Cameco’s McArthur River mine, updated the Critical Span Curve based on McArthur River mine openings with ground support in unfrozen ground. Figure 2.19 shows the updated potentially unstable zone (solid black lines) compared to the original potentially unstable zones (dashed red lines) based on the data sets from Ouchi (2005), and Wang (1999). 62  The McArthur River mine updated span curve was developed based on observations of the rock mass quality and span in supported ground excavations including failures such as the Bay 12 failure. The stability graph shifted the potentially unstable zone to the left to match the span and area. Observations of the influence of freezing on the RMR will be plotted on the same McArthur River stability graph as McArthur River has similar rock mass conditions as Cigar Lake. Although the potentially unstable zone at McArthur River is based on empirical observations with ground support, the frozen unstable/stable curve will be developed from a different approach.  Figure 2.19: McArthur River Stability Graph with Ground Support, after Pakalnis (2012)  2.8  Applicability of Hoek-Brown Parameters to Frozen Ground  Application of Hoek-Brown brittle parameters to frozen ground was investigated by Yang et al. (2012) and noted it to be applicable in low stress environments though did not correspond well in high confining stress environments. Yang et al. (2012) discovered that the Mohr-Coulomb strength criterion fit the low confining stress range of the frozen soil specimens as the frozen soil has a linear relationship with confining pressure. However, for frozen soil under high confining 63  stresses, the strength relationship with confining pressure exhibited a non linear relationship. Based on experimental testing Yang et al. (2012) found the strength of frozen soil increased with confining pressure up to a limit; however, the strength decreased with further increase of confining pressure beyond this limit. At high confining pressures, the non linear strength of frozen soils is attributed to pressure melting and crushing of the ice crystals. Frozen soils therefore tested in the low stress range can be expected to have higher friction values than those at high confining pressures. In order for Yang et al. (2012) to describe the non-linear strength characteristic of frozen soil better, the Hoek-Brown criterion (Hoek et al., 2002) was modified by incorporating a new parameter to account for the effect of pressure melting and crushing phenomena. The new formulation was presented as:  � 𝜎� 𝜎� − 𝐴𝜎� = �𝑚 + 1� 𝜎� 𝜎�  incorporating a new parameter to account for the nonlinear strength characteristic of frozen soil where m, n and A are constants for materials determined by the Levenberg-Marquardt fitting method. The laboratory data for the testing range of confining pressures was not included in this research and therefore quantifying what high stress environment was applied to this testing is difficult to compare with the conditions at the Cigar Lake mine.  64  3.  Methodology  This section outlines the process followed to understand the influence of freezing on a weak and altered/fractured rock mass at depth. The Cigar Lake orebody is located at the unconformity between metamorphic basement rocks and sandstone at a depth of approximately 430 m. Regional faulting and alteration processes in northern Saskatchewan have created a series of uranium deposits in the Athabasca basin along this unconformity. The alteration surrounding the orebody during uranium mineralization created a highly heterogeneous and permeable zone of poor ground comprising soft to moderately indurated sandy clay, unconsolidated sand and altered rock (sandstone above the orebody and metapelite basement below). Cigar Lake mine construction commenced in 2005; however, the underground levels were flooded from 2006 to 2010 due to several water inflow events due to loss of ground. The geotechnical data collection program carried out for this research was initially planned to sample material from both surface and underground drilling. However, the underground levels were inaccessible after the last inflow event in the summer of 2008 limiting material sampling to surface diamond drilling. The author believes that sampling the Cigar Lake material underground in an unfrozen and frozen state, combined with underground in-situ testing is essential for understanding the behaviour of frozen weak rock.  3.1  Assessment of Existing Information  A review of the literature on frozen soil, rock, and mining within frozen ground provides detailed information on the sub-zero behaviour conditions of soil (sand, clay, mixed gravels) and hard rock storage caverns for liquid nitrogen under extremely cold conditions (up to -196 oC). Limited research data was found on the behaviour of frozen weak rock, especially at depth. Laboratory testing of rock core in a sub-zero environment that was sampled in an unfrozen state from both the Cigar Lake project and McArthur River mine has been completed on a small scale over the past 20 years. However, the previous research did not address or adequately give insight into the failure mechanisms and behaviour of a cavity in frozen weak rock. 65  Existing geotechnical site investigations, hydrogeological reports, geological mapping, and diamond drill hole information, can be reviewed in order to: 1. define geological/hydrogeological variability and types of materials to be encountered. 2. identify mechanical and thermal material property data gaps in previous site investigations relevant to ground freezing design and stability. 3. quantify the percent clay/silt of the matrix and clay mineralogy to establish how the frozen material will behave. 4. develop a database of creep parameters from geotechnically similar materials. Developing a database of frozen strength, creep, and thermal parameters from geotechnically similar materials is ongoing to complement the current Cigar Lake laboratory database. The clay cap and clay ore zone would be compared to ice poor materials of similar plasticity and grain size gradation. For loosely unconsolidated zones of material such as altered sandstone/sand, assuming known creep parameters of ice rich sand will be conservative.  3.2  Conceptual Model of Failure Mechanisms  Cigar Lake mine intends to mine the uranium ore through the process of jet boring, a non entry mining method. Jet bored cavities are developed by a high pressure water nozzle rotating in a pilot hole from the top of the cavity to the base. Cavity dimensions are expected to be the height of the ore (ranging 5 to 15 m in height) with diameters that will vary depending on the ground conditions and excavation sequencing. The behaviour and stability of frozen material over the mined out cavities once mining commences is a function of the frozen rock mass. The stability of the frozen cavity will depend upon the excavated span, rock mass strength, length of exposure, thermal regime, and ground mass ice content. Failure can occur due to wedge fallout, slab failure, gravity driven caving, and beam failure. A material data collection and laboratory testing program was undertaken here focusing on the influence of ice in increasing the strength of weak rock and the influence of freezing on rock joints with and without infilling. Frozen unconfined compressive strength (UCS) and frozen 66  beam testing is explored in a series of laboratory tests to determine the failure mechanism of a typical frozen weak rock overlying the Cigar Lake orebody. Only after gaining an understanding of frozen rock mass behaviour, the stability and stand up time of a jet bored cavity can be assessed. Verification of the conceptual model of frozen weak rock masses will be compared with current mining practices in frozen ground at the McArthur River mine and historical field trials at the Cigar Lake mine. 3.3  Material Properties Sampling Program  Cigar Lake Mine undertook a diamond drill core sampling program in 2009 to address data gaps from historical geotechnical drilling and laboratory testing and to better define the highly variable nature of the altered zone over the orebody. This will provide additional information of the unfrozen and frozen geotechnical properties of a weak and jointed rock mass. A diamond drilling contractor was retained in 2009 to complete a surface freeze drilling program of eight boreholes, located approximately 150 m north of Shaft 1 at Cigar Lake. From the surface freeze boreholes, four PQ (3”) holes were cored through the orebody and used for material sampling part of this research. 3.3.1 Sample Collection The local geological formations within the target sampling depths of the orebody have a known history of zero to poor recovery due to the material’s loose, cohesionless, and friable nature. The following discusses the methodology to core quality samples from a diamond drill ensuring maximum core recovery and minimal sample disturbance. Ground freezing is expected to be from the base of the orebody to a minimum of 20 m above the orebody. To characterize the behaviour of the frozen material, the target sampling and testing zone is approximately 30 m above the orebody to 15 m below the orebody. The top elevation of the orebody was estimated on a hole by hole basis from the current site geological model to establish the target depth to commence core retrieval. While coring through the orebody, the overlying clay cap or known zone of soil like material, a clay face injection bit was used to cut back water flow and reduce the risk of washing away the sample. Metal liners or ‘splits’ are standard for triple tube coring and are sufficient for drilling and sampling competent sandstone. However, instead of metal splits, acrylic tubing was placed inside the core barrel when drilling 67  within friable sandstone to soil like material to limit core removal handling and disturbance. The 1.5 m long acrylic tubes were sealed on either end at the drill rig and stored inside the Cigar Lake core logging warehouse prior to shipment for laboratory testing.  3.3.2 Sample Integrity During Drilling Sample disturbance is the difference between the in situ and lab measured material properties and soil structure. The sampling technique, stress release, handling, and preparation can all cause sample disturbance. Stress relief occurs during coring samples that are subject to high in situ stresses at great depths. The rapid unloading of confining stress can permanently damage the structure of brittle rocks and sensitive clays. The difference in the shear strength of soils reconsolidated to the in situ stress for laboratory testing is not considered an issue. However, cemented soils and brittle rocks can be problematic materials exhibiting lower strengths after coring. The direction of coring also influences the stress path of the sample during unloading. For anisotropic strata, the effect of coring horizontally compared to vertically introduces the need to consider the directionality of stress path unloading. The surface freeze pipe drillholes with sampling for geotechnical testing will be drilled vertically through horizontally bedded sandstone. Core samples in the lab will therefore be loaded perpendicular to bedding for strength testing. There is the possibility of drilling through a titled fault block that should be detected if the bedding angle observed is steeper than the regional bedding. Sample disturbance during handling will underestimate the pre-consolidation pressure and initial void ratio. Tube sampling strains on soft clays can damage the microstructure, reduce the mean effective stress and cause water content redistribution. Actions that were taken to minimize sample disturbance at the drill rig during the surface freeze sampling program by Cigar Lake mine include the following: •  When drilling though soils or friable rock, the use of an acrylic liner instead of metal splits will limit sample expansion during core retrieval and remove the need for unnecessary sample handling from the core barrel.  68  •  Cutting back on water flow while sampling soil like material will minimize the potential to wash away loose or soft zones, when using a face injection bit.  •  Ensuring all samples are consistently handled, preserved, and tested according to the same procedures will limit the issue of testing samples not at in situ stress after undergoing stress relaxation.  3.4  Classification Systems in Frozen Weak Rock  As the Cigar Lake mine and shaft were flooded up to surface at the time of this research, direct observation and monitoring of the influence of freezing on a weak rock mass was not possible. Instead, comparing rock mass classification systems, Rock Mass Rating (RMR) and Q-system, for unfrozen and frozen Cigar Lake weak rock will provide pre-mining input into strength implications with respect to the design of the ice cap to overlie each mined out cavity. The Cigar Lake orebody and surrounding material is a heterogeneous mixture of fractured and altered rock that has weakened to clay and sand. The influence of freezing on the rock mass rating (RMR) specifically the unfrozen to frozen correlation between rock mass rating (RMR) and span for weak rock in underground mines is based on the work of Ouchi et al. (2004), Pakalnis (2002), and Lang (1994). When a groundmass freezes, the rock mass strength will increase due to pore water converting to ice. This increase in strength can be attributed to an increase in the UCS and the freezing of the joint walls if there is infilling present. The degree to which freezing influences the RMR input parameters is expected to vary under different temperatures, moisture content, clay content and initial rock mass strength. Unconfined compressive strength and triaxial tests on unfrozen and frozen drill core samples will be able to assess the influence of freezing on the rock hardness parameter. Four point beam testing and shear strength testing are planned to determine the influence of freezing on the joint condition parameter and cohesion. However, an important parameter that is not addressed in unfrozen rock mass classification is creep or the decrease in rock mass strength over time due to steady state loading. The creep of frozen rock masses over a long period of time may result in strength loss, similar to that seen for a block of ice under an instantaneous load or a constant load applied over a long period of time. 69  3.5  Laboratory Testing to Establish Influence of Freezing  Structural and thermal calculations are required for the design of a ground freezing project. Strength and deformation properties of the unfrozen and frozen soil, which are time and temperature dependent, are necessary for the structural design of a soil or rock mass support structure. Thermal characteristics are also important for thermal analysis to verify the freeze hole layout and ensure an adequate frozen ground thickness. Thermal analyses are not within the scope of this research. The most important input parameters for the analysis of frozen material overlying an excavated cavity are the unfrozen and frozen elastic modulus and shear strength (cohesion and friction) parameters. A better geotechnical understanding of the material surrounding the ore body, the clay cap and altered basement frozen strength and creep behaviour is required as these materials control the stability of an excavated cavity. Limited geomechanical information is published on the shear strength, time dependent behaviour, and thermal properties of frozen rock or soil at great depths. Laboratory testing of the samples collected in unfrozen conditions from the 2009 Surface Freeze Drilling program (for the purpose of installing freeze pipes), was completed on the weak rock overlying and beneath the orebody within a controlled cold temperature room environment. The key focus of the laboratory testing is to improve in situ and laboratory characterization methods and provide a better understanding of weak rock behaviour at sub zero conditions with varying temperatures and strain rates. Any rock core retrieved containing greater than 2% U3O8 by the mine geologists was deemed unsafe to handle by laboratory personnel. Therefore no laboratory testing was completed on any samples from the orebody. Unconfined compressive strength (UCS), four point beam testing, direct shear testing, X-Ray diffraction, and moisture content testing was completed on samples from the altered sandstone (clay cap) overlying the orebody and altered metapelite basement rock below the orebody. Thermal properties of the rock core were not part of the scope of this research. The University of Alberta’s geotechnical laboratory is equipped with several cold rooms that can accommodate triaxial cells for UCS and triaxial testing in a sub zero environment. Frozen UCS 70  testing was undertaken at the University of Alberta cold room and all remaining testing was completed at the University of British Columbia geomechanics laboratory. To determine the shear strength of frozen soil, triaxial compression tests must be completed. The triaxial test is suitable for all types of soil and rock, and has the following key advantages; i) drainage conditions can be controlled, ii) pore water pressure measurements can be made, and iii) the two loading directions can be controlled independently. However, triaxial testing of the collected rock core was not feasible at the University of Alberta cold room due to the lack of drill core samples for testing and the triaxial cell available for testing could not accommodate axial loads greater than 20 MPa. There are limitations to reproducing in situ freezing conditions in the lab environment, as the Cigar Lake orebody is located at approximately 430 m depth. How the samples freeze, the rate of freezing and ice lens growth will influence the frozen strength, though to what degree is an uncertainty. Applying a high confining pressure on the samples as it freezes similar to that experienced underground was not an option during testing. The samples for UCS testing were frozen rapidly to the desired testing temperatures with no confining pressure to prevent ice lens growth. Rock specimens for testing were cut in half to examine the ice lens growth in the laboratory freezing environment. The main parameter that will affect the freezing rate and ice lens formation is the water content in the ore region, as this region has the potential to have both low conductivity and high water content. McArthur River established the in situ moisture content through a back analysis spanning several years of measured ground temperature vs. time profile and thermal properties. At Cigar Lake, as the layer of frozen altered sandstone overlying the orebody will be subjected to hydrostatic pressure (in situ stresses and water in the sandstone), shear stresses (shear zone caused by fracturing and squeezing ground around ore zone) and a creep regime (presence of ice and squeezing environment). In order to optimize the design of the frozen material over the orebody and rock mass frozen strength, the creep behaviour and shear strength is required to predict the stability of the proposed jet bored cavities. For the purpose of design, the increase in strength due to freezing to needs to be addressed under both short term (several hours to days) and long term (several days to weeks) loading. The loss in strength due to creep behaviour 71  however, is not part of this research. 3.5.1 Unconfined Compressive Strength Testing Unconfined compressive strength (UCS) is the load per unit area at which a soil or rock sample will fail in uniaxial compression. The unconfined compressive strength is an input parameter into Bieniawski's Rock Mass Rating (RMR) classification system relating the intact strength to the overall rock mass behaviour. The UCS of a frozen sample will vary with the temperature and applied strain rate. Ideally several series of UCS testing at temperatures ranging from unfrozen (0oC) to -20oC and varying strain rates of the applied load would be undertaken. However, the availability of intact samples from the 2009 surface freeze drilling sampling was limited to less than 5 m of core from the rock overlying and beneath the orebody. UCS testing of the Cigar Lake samples were therefore reduced to testing at two temperatures, -10oC and -20oC at one applied strain rate. 3.5.2 Four Point Beam Testing Four point beam testing was undertaken on a suite of pre-mixed cement and altered sandstone material (clay cap overlying the orebody) to identify the influence of freezing on a frozen joint. Three point and four point flexural testing is typically used in the laboratory to measure the modulus of elasticity in the bending moments of concrete, wood, steel or other materials. Bending tests are simple and quick to complete, but are influenced by the applied strain rate and specimen geometry. The flexural strength is equivalent to the tensile strength assuming the beam is homogeneous without defects or flaws. The beam will fail at the midpoint, developing a crack due to tension as the beam fails under tensile stresses before compressive stresses with this loading regime. Four point beam testing was completed on cement and sand mixtures having strengths similar to those for the altered sandstone overlying the orebody; the cement mixture samples were prepared to contain a single smooth, planar joint with no infilling in the center of the beam. Testing various cement mixture samples with joints provides the basis for understanding how a frozen beam fails under tension using a controllable sample material. In an unfrozen state the degree of jointing and infilling material in a rock mass will control the failure. No research or data was located by the author on how a frozen jointed weak rock mass 72  fails. Failing a rock specimen in tension, produces a crack at the midpoint of the beam. If the frozen joint is weaker than the intact rock, ideally the beam will fail along the joint. If the frozen joint is stronger than the intact rock, the beam will fail through the solid beam material at the midpoint of the beam. The increased cohesion of a joint undergoing freezing will be influenced by the type and thickness of infilling and the degree of moisture on the joint surface. A smooth and planar joint with no infilling and no moisture will not have sufficient cohesion to bond the joint surfaces together. 3.5.3 Direct Shear Testing Determining the shear strength of rock joints is significant to understanding rock mass behaviour. The rock mass fabric is influenced by jointing, bedding, foliation, faulting and potentially other factors all which have distinct shear strength components. The shear behaviour of rock joints is determined in the laboratory with a direct shear apparatus that applies a constant normal load during uniform shearing. Cohesion and friction angle of the joint surface are determined by a linear regression of the shear and normal stresses applied. The Cigar Lake orebody is hosted in a flat lying sedimentary basin in an area of historical faulting. Predominant joint sets are parallel to the main faults and along bedding planes. Away from the ore body the joint sets are typically rough, planar, and with trace amounts to little infilling. However, the intense alteration surrounding the orebody has degraded the sound rock mass infilling the joints with thick seams of clay and sand. The freezing of a rockmass is believed to have a significant influence on the shear strength behaviour, specifically the cohesion. Direct shear testing on natural joint surfaces and intact rock specimens was completed to develop a model of shear strength gained along a frozen joint. Testing of intact rock specimens was carried out to determine the intact shear strength of recognizable shear planes/planes of weakness; testing of shearing resistance along the jointed/fractured specimens was carried out to determine the lower bound residual strength.  73  4.  Cigar Lake Geology, Hydrogeology, and Historical Geotechnical Data  This section summarizes the regional geology, hydrogeology and geomechanical properties of the Cigar Lake mine rock types. 4.1  Regional Geology  The Cigar Lake deposit is located along a major east-northeast trending 30 km long trough in the Athabasca basin. The Athabasca basin covers approximately 100,000 km2 in northern Saskatchewan, Canada (see Figure 4.1) and is filled with sandstones, conglomerates, shales, and dolomites of the middle Proterozoic Athabasca Group. Similar to other major uranium deposits of the Athabasca basin, the Cigar Lake deposit is located at the unconformity separating sandstones of the Athabasca group from metasedimentary gneisses and plutonic rocks of the Wollaston Domain. The sandstone units of the Athabasca Group host most of the uranium mineralization and lie unconformably over the basement metasedimentary gneisses.  Cigar Lake  Figure 4.1: 4.2  Athabasca Basin and Cameco Corporation Active Mining Projects  Formation of the Cigar Lake Deposit and Mineralization  The Cigar Lake deposit is flat lying, approximately 1950 m long, 20 to 100 m wide, and ranges up to 16 m thick, with an average thickness of about 6 m. The unconformity related deposit is a typical sandstone hosted orebody structurally associated with a one kilometre east-west basement corresponding to a graphitic shear zone. Unconformity related uranium deposits are believed to have formed through an oxidation-reduction reaction at a contact where oxygenated fluids meet 74  with reducing fluids and the unconformity provides that contact (Jefferson et al., 2007). The Cigar Lake deposit is referred to as an “Egress type” unconformity associated uranium deposits which typically develop alteration halos in the siliclastic strata overlying the deposit. 4.3  Local Geology  Geological and structural interpretations are on-going by Cameco Corporation. Local and regional geological interpretations have been completed by Bruneton (1986, 1993), Baudemont (2000), Fouques et al. (2000), Portella and Annesley (2000), Jefferson et al. (2007) Golder Associates (1986, 2001), and MDH Engineering Solutions (2008). The Cigar Lake orebody is located at an unconformable contact between the overlying Manitou Falls Formation of the Athabasca Group sandstones and the metamorphic basement rocks of the Pre-Cambrian shield. Above the unconformity, sediments consist of a basal conglomerate overlain by sandstone of the Manitou Falls Formation a 450 m thick quartz arenite with local conglomerate layers. At the unconformity, sand is interpreted to form a continuous subhorizontal layer along the southern margin of the deposit establishing a hydraulic connection. The presence of sand above the unconformity is due to dissolution/desilification of the sandstone at the time of deposit formation. Dissolution has created a depressed zone on top of the deposit with bedding dipping shallowly at 5 to 15 degrees (Baudemont, 2000). Sand rich zones are characterized by high porosity, high permeability, and very poor rock strength. Above and below the unconformity, the rock mass shows variations in porosity and permeability due to fracturing and alteration. Zones of intense faulting and alteration pose geotechnical challenges during mining including control of groundwater and ground support of weak rock.  4.3.1 Alteration Several alteration events have created intense fracturing, massive quartz dissolution in the sandstone and extensive clay alteration around the Cigar Lake orebody. Alteration zones are characterized by well developed concentric zones in the sandstone and basement rocks surrounding the ore deposit. This alteration halo in the sandstone is centered on the deposit and reaches up to 300 m in width and height. In the basement rocks, this zone extends in the range of 200 m in width and as much as 100 m in depth below the deposit. Alteration is associated with 75  the loss of cohesion in the sandstone and the enrichment in clay content (Hoeve and Quirt, 1984). Percival et al. (1993) subdivided the alteration zones from the outermost to innermost with increasing alteration towards the orebody, listed below (refer to Figure 4.2). •  An outermost alteration zone consists of altered Manitou Falls sandstones characterized by dissolution textures, lower quartz contents and slightly higher clay contents than the overlying sandstones.  •  Underlain by a clay rich alteration halo around the deposit characterized by 10-30 % by weight clay and averaging 1 to 5 m thick with a maximum thickness 10 m.  •  The clay cap directly over the orebody (illite with some kaolinite and sudoite) is known for its high relative portions of clayey material commonly mixed with sand, silt or clayrich sandstone. Encapsulating the orebody is a hematite-rich clay zone (Bruneton, 1997).  Figure 4.2:  Cigar Lake Deposit and Alteration Limits, after Jefferson et al. (2007)  The degree of alteration of the sandstone or metapelite can be related to the clay mineralogy. X76  ray diffraction (XRD) testing was completed by the University of British Columbia Department of Earth and Ocean Sciences lab on two samples from the 2009 surface freezing drilling program of altered sandstone; bleached sandstone and hematized clay from boreholes ST786-07 and ST801-04, respectively. Details of the XRD testing are provided in Appendix A and summarized below in Table 4.1. Table 4.1:  Results of Quantitative Phase Analysis (wt.%) Sample 18  Sample 19  Bleached Sandstone  Hematized Clay  ST786-07  ST801-04  Mineral  Ideal Formula  427.3 m  434.7 m  Illite  K0.65Al2.0(Al0.65Si3.35O10)(OH)2  95.3  82.9  Kaolinite  Al2Si2O5(OH)4  3.0  Rutile?  TiO2  1.0  0.8  Alunite?  K2Al6(SO4)4(OH)12  0.7  0.5  Hematite  α-Fe2O3  13.4  Pyrite  FeS2  2.4  Total  100.0  100.0  Both the bleached sandstone and hematized clay samples are predominantly illite, though the bleached sandstone contains trace amount of kaolinite which is not present in the hematized clay sample. The influence of clay minerals on the frozen behavior and freezing rate has not been directly assessed though the salinity, unfrozen water content, and plastic limit of the clay material will have a greater influence on the freezing rate. The two samples submitted for XRD testing were non plastic.  4.3.2 Faulting and Structures The Cigar Lake deposit comprises several folding events and later faulting. Regional compression has resulted in the reactivation of the Hudsonian faults post Athabasca deposition and the development of large scale NE-SW trending reverse faults. The crystalline basement has been subjected to multiple deformation events resulting in complex fold patterns. Evidence exists 77  of a high-grade diagensis (changes in the sandstone mineralogy due to low temperatures and pressures) throughout the Athabasca sediments which overly the crystalline basement rock. Faulting through the Athabasca sandstone has mechanically disintegrated and fractured the sandstone to sand. Sections affected by faulting are marked by strong bleaching, hydrothermal silification and perched mineralization. In the basement, clay-alteration appears to be strictly fault-controlled producing local squeezing clay and high-pressure water. These weak sand/alteration zones are responsible for the ground falls and subsequent inflows at Cigar Lake. Baudemont (2000) interpreted a limited number of oriented drillholes identifying the vertical evolution of the regional fracture and fault system and characterizing the post-Athabasca fault structures. Nine geotechnical holes with core orientation were logged in sub-horizontal and inclined geotechnical drillholes from the 210, 420, and 480 level in 1999 (Baudemont, 2000). The recorded data is presented in the stereoplot in Figure 4.3. Although the data only covers a 200 by 300 m wide section of the Cigar Lake mine, the following can be concluded with respect to the local structures: •  Two conjugate sets of steeply dipping faults are predominantly oriented striking to 85 and 285 degrees, and are characterized by a conjugate set of normal to strike slip faults  •  Basement foliation by underground mapping is consistent with the oriented core (strike/dip of 090/70)  •  200 m and more above the orebody, evidence of faulting is scarce and fracture frequency low  •  The orebody located in an east-west trending high is interpreted as an uplift horst (100130 m wide and 20-30 m high) bounded by a system of normal faults.  •  Intense graphite and pyrite enrichment is associated with the Cigar Lake shear development.  78  Figure 4.3:  4.4  Stereonet Plots of Structural Data from 1999 Underground Drilling, from Baudemont (2000) Data  Geotechnical Site Investigations  The Cigar Lake deposit was delineated by a major surface drilling program from 1982 to 1986, followed by several small drilling programs for geotechnical and infill holes to 1998. Underground diamond drilling was undertaken from 1989 to 2006 to determine ore and waste rock characteristics in advance of development and mining. During 2006, several hundred freeze and temperature monitoring holes were drilled as part of establishing the ground freezing system, though the freeze holes were drilled by percussion methods so no core was retrieved. Geotechnical boreholes drilled to characterize the geomechanical and thermal properties of the orebody and surrounding area completed from the mid-1980s to present, are outlined below. •  Boreholes drilled in 1983 and 1984 were drilled to characterize ground formations near the orebody and obtain test samples for uniaxial compression, triaxial compression, slake durability, porosity, water and clay content, Atterberg limits, and permeability testing. 79  • • • • • • • •  Samples were collected in 1985 and 1986 boreholes for unfrozen and frozen UCS testing and unfrozen triaxial testing (Golder Associates). In 1996, insitu temperature profiles were logged from surface to the orebody in several boreholes (Golder Associates). From the 1990 drilling program, frozen samples were collected for creep and UCS testing (EBA). Unfrozen graphitic metapelite was collected in 1994 for UCS, triaxial, and creep tests. In 1999, underground drilling for core orientation was completed on the 210, 420, and 480 level (Baudemont, 2000). UCS and porosity testing was completed on unfrozen rock in 2000 (U of Saskatchewan). The 2007 drill program assisted the development of a site geological and hydrogeological model (MDH). 2009 surface drill program with sampling for testing frozen UCS, direct shear, and fourpoint beam (by the author).  Rock mechanics data for underground deposits are initially collected from drill core, a point sample of the rock mass. In weaker rock, the sample is often disturbed with the amount of disturbance a function of the rock mass quality, drilling, and sample handling. Golder (2002) reviewed all collected geotechnical drillholes information commenting on a lack of consistency between various data sets in the Cigar Lake rock types. The majority of boreholes drilled in the beginning of the exploration program were also not specifically for geotechnical purposes and therefore lacking completeness of the geotechnical database. Unfrozen and frozen Unconfined Compressive Strength (UCS) testing was completed on a suite of samples in the 1980s and the relevant samples for this research are included in the discussion section of the UCS testing.  4.5  Geotechnical Zones  The deposit and host rocks consist of three principal geological and geotechnical zones: the deposit itself, the overlying sandstone, and the underlying metamorphic basement rocks. Artificial ground freezing of the Cigar Lake orebody is expected to intersect five major material types (see Figure 4.4): mineralization/ore (indurated clay to claystone), altered sandstone (dense clay to weak sandstone), sand/highly friable sandstone, fractured sandstone, and altered basement. 80  Athabasca Sandstone Manitou Falls Formation  Massive High Grade Ore Alteration  Figure 4.4:  Unconformity Basement Rock Graphitic Metapelite  Cigar Lake Geotechnical Zones  The orebody geotechnical properties are thought to be relatively consistent across the orebody; however, above the orebody, ground conditions are highly variable ranging from extremely weak and altered sandstone to a hard indurated clay. Golder (2001) has noted a significant variability in the mechanical properties that exists within the east-west trending altered shear zones. The following geological interpretations are based on geotechnical site investigation reports by Golder Associates (1986, 2002), JD Smith Engineering (1983), and MDH (2008). The following unfrozen material properties are summarized from multiple interpretations of UCS and triaxial testing. Frozen material properties consist of laboratory work completed by Golder (1986) and EBA (1991). Frozen unconfined compressive strength and triaxial testing was completed on clay cap and orebody samples at temperatures ranging from -2 to -20oC. Note that moisture content data is limited and highly variable therefore influencing reliability of the mechanical and thermal properties. 81  4.5.1 Mineralization/Ore The ore deposit is located at approximately 430 to 450 m depth, is approximately 2000 m long, 250 m wide and up to 16 m thick with an average thickness of 5.5 m. The orebody is crescent shaped in cross section and follows the paleo-topography of the unconformity. The massive high-grade ore is formed by metal oxides, arsenides and sulphides in a matrix of generally well indurated greenish clay, or claystone. The orebody consists of a mixture of massive pitchblende, pitchblende-rich clay, pitchblende-impregnated sandstone, clay, silt, and sand. It is capped by a layer of similarly indurated clay that is variably 1-5 m thick. Based on several reports, Table 4.2 and Table 4.3 list the unfrozen and frozen geotechnical properties of the mineralization/ore. Table 4.2:  Mineralization/Ore Unfrozen Material Properties (Golder, 2002)  Source  S.G.  Moisture Content (%)  Friction Angle, φ  Cohesion, c (MPa)  Elastic Modulus E (GPa)  Poisson’s ratio  Calculated UCS (MPa)  Ore (assumed same as clay cap)  Golder, 1995  Variable, function of metal. Range 1.8-4. Avg 2.69  Highly variable, 5-30%  -  -  0.1 – 15 (highly variable)  -  For R<1, 5 to 15, else up to 60  Ore (assumed same as clay cap)  Golder, 2002  mean  -  -  25  1.6  1.0  0.30  5.0  mean, stnd. dev  -  25.7 ± 15.2  -  -  11.4 ± 8.1  0.18 ± 0.06  17.8 ± 22.5  range  -  3.2 - 61.4  -  -  0.07 22.3  0.13 - 0.32  0 - 65.8  number of tests  -  25  -  -  8  8  14  Description  Ore  CLMC, 1989  82  Table 4.3: Material Ore (Indurated Clay) Ore (Intermediate Clay)  Mineralization/Ore Frozen Material Properties (Golder, 2002)  Unit Weight, (kN/m3)  Friction Angle, φ  Cohesion, c (MPa)  Elastic Modulus, E (MPa)  Poisson’s ratio  Calculated UCS, (MPa)  Test Temperature, (oC)  26.2  30  0.87  1000  -  10  -5  20.5  -  -  -  -  2  -5  4.5.2 Clay Altered Sandstone Hydrothermal alteration associated with the ore deposition has made a clay rich alteration halo around the deposit, averaging 1 to 5 m thick with a maximum thickness of 10 m. The clay cap, directly overlying the orebody, is known for its high relative portions of clayey material commonly mixed with sand, silt or clay-rich sandstone. Geological log descriptions suggest this zone is typically associated as having a clay parent material (geotechnical classification of >35% by weight) or elevated portions of the clayey material (geotechnical classification of 20% to 35% by weight). Clay minerals within the clay/ore zone are predominantly illite and kaolinite with some chlorite. Original geological logs of exploration boreholes also indicate regions of prominent core loss; within this unit, likely attributed to the influence of preferential flow pathways due to faults. Fractures control the permeability in clay/ore zone and dip steeply to the south at 1 m spacing. Above this cap there is a highly heterogeneous, highly permeable zone from 20 to 50 m thick consisting of soft to moderately indurated sandy clay, unconsolidated sand and variably altered sandstone. The clays are divided into three types based on the degree of hydrothermal alteration in the sandstone. After Kennard (1998), the clays are defined as: Soft clay: occur as layers along bedding planes above the massive clay zone and as veins in steeply dipping faults. These clays are mostly comprised of illite. Intermediate clay: clays most commonly encountered in mineralized areas, represent a transition between the "soft" and harder "indurated" clays. Intermediate clays comprise the 83  clay cap overlying the orebody and form a matrix for mineralization in the massive ore zone. These clays are comprised of various forms of iron and magnesium rich illite and chlorite. Indurated clay: hard clays located in the massive clay cap and are composed of various forms of iron and magnesium rich illite and chlorite. Table 4.4 and Table 4.5 summarize the unfrozen and frozen properties of the clay based on testing by Golder (1986) on several boreholes intersecting the clay, EBA (1990), Kennard (1998), and Golder's data reinterpretation for numerical modelling (2002).  Table 4.4: Moisture Content, (%)  Material Source  Clay Cap  Intermediate Clay Indurated Clay  Clay Unfrozen Material Properties  Elastic Modulus, E (GPa)  Poisson’s ratio, v  UCS, (MPa)  CLMC (1989)  Kennard (1998)  Golder (2002)  CLMC (1989)  Kennard (1998)  Golder (2002)  CLMC (1989)  Kennard (1998)  Golder (2002)  mean, stnd. dev  59.7 ± 15.4  0.88 ± 0.80  0.07 13  1.0  0.18  0.13 – 0.32  0.30  1.2 ± 4.3  0 – 0.32  5.0  range  3.5 - 67.7  0.25 2.5  -  -  0.18  -  -  0-15.4  -  -  No. of tests  33  8  -  -  1  -  -  13  -  -  -  0.15  -  -  -  -  0 – 0.5  0.188  0.24 – 9.1  0.355  -  0.06 – 0.42  -  -  0–7  3  84  Table 4.5:  Material  Clay Frozen Material Properties  Test Temperature, (oC)  Avg. Frozen Unit Weight, (kN/m3)  -10  22.68 26.24  Source  Indurated Clay  Intermediate Clay  Soft Clay  Elastic Modulus, E (GPa) Golder (2002)  -5 -2 -20 -10 -5 -2  22.55 19.4 19.8 17.0  -5  18  0.15  Average UCS, (MPa) Golder (2002) 11.34 10.19 6.0 4.4 1.5 2.3 0.6 1.3  4.5.3 Sand/Highly Friable Sandstone and Fractured Sandstone The sandstone at Cigar Lake is divided into three geotechnical domains according the quality and degree of fracturing of the rockmass, including: • • •  Competent sandstone Fracture sandstone Sand/Highly friable sandstone  Directly overlying the orebody, the highly friable sandstone zone comprises unconsolidated sand or zones of no core recovery, and is representative of the unconsolidated material above the sandstone bedrock. Sand often comes in contact with the clay cap along the south margin of the deposit. The highly friable sandstone is unaltered or weakly to strongly clay altered sandstone zone with an RQD less than 70%. This zone extends southward and northward from the primary mineralized zone for up to 75 to 100 m along the unconformity (not systematically present). No geotechnical data or site specific laboratory testing is provided due to the poor to no recovery of material in this zone. Geotechnical properties of the frozen sand will behave similarly to that tested by Sayles (1968) and discussed in the literature review. Sandstone above the orebody has been subjected to various degrees of hydrothermal alteration from less altered (at a distance above the orebody) to extremely altered (immediately adjacent to 85  the orebody). The fractured sandstone forms a halo around the highly friable sandstone, typically extending up to approximately 50 to 75 m above the unconformity and may extend upward along select faults for more than 200 m. This material is comprised of a varying mixture of fractured and densely fractured sandstone with an RQD between 70% and 90%. Increasing clay content and a decrease in the cementation of the sandstone causes a gradational strength decrease with proximity to the orebody. It has been a concern of previous consultants and the mine that the geological complexities of the deposit and the lack of proper description in the uniaxial compression test data sheets prevents a qualitative analysis of the tests between altered and unaltered units. Cigar Lake Mine has classified the sandstone into three geotechnical categories RM1 – Consists of a mix of R0 (extremely weak) to R1 (very weak rock), UCS ranging from 0 to 5 MPa. Comprises sand to highly fractured (RQD <25%) sandstone RM2 – consists of weak rock (R2) with an estimated intact rock strength between 10 and 30 MPa, and an RQD average of about 30-70% Joint surfaces are typically planar, often slightly moist, smooth to rough, often graphitic. The surfaces are moderately altered, with smears of clay and mud and frequently slickensided. RM3 – is poor to fair quality rock with an average strength of 25 to 50 MPa (R3, medium strong) No historical test data dividing the rock mass properties of the altered sandstone as per the divisions above were completed. However, Table 4.6 below summarizes the historical altered sandstone testing.  86  Table 4.6: Source  Altered Sandstone Unfrozen Material Properties Moisture Content (%)  Friction Angle, φ  Cohesion, c (MPa)  Elastic Modulus, E (GPa)  Poisson’s ratio, v  Calculated UCS, (MPa)  5.3 ± 1.3  -  -  14.5 ± 7.7  0.17 ± 0.07  36.3 ± 19.3  2.9 - 9.6  -  -  0.07 - 31  0.05 - 0.37  1.4 - 83.7  89  -  -  71  71  71  mean  -  35  2.0  5.0  0.25  7.7  mean, stnd. dev  -  41 ± 2  1.32 ± 0.67  -  -  5.8  mean  -  26  0.4  -  -  1.3  mean  -  45  5.4  -  -  26.1  -  36  3.6  -  -  14.1  Description mean, stnd. dev range number of tests  CLMC, 1989 Golder, 2001 Geosciences, 1988  Kennard, 1998  Kennard, 1995  Kennard, 1995  All tests Low confining stress (high clay content) High confining stress (low clay content) All tests  4.5.4 Altered Basement The metamorphic basement rocks consist mainly of graphitic metapelitic gneisses and calcsilicate gneisses. Graphite-and pyrite-rich “augen gneisses”, occur primarily below the Cigar Lake orebody. The mineralogy and geochemistry of the graphitic metapelitic gneisses suggest that they were originally carbonaceous shales (Bruneton, 1993). Basement rock mass conditions vary considerably within short distance from good to extremely poor. In general, basement alteration does not show a strong correlation spatially with that of altered sandstone above the unconformity. The lower and upper basement varies from east to west with three lithostratigraphic units identified by geology and alteration. The upper basement geology shows significantly more clay alteration, especially along the margins of the graphitic units.  The main basement unit comprise pelites, with many interlayered units of various  composition and tends to be less altered than the others. Graphitic metapelites are associated with mineralization and are moderately to strongly gneissic and banded. Clay alteration peaks in this unit, especially along fault zones. Gneissic layering dips steeply 60 to 90 degrees to the 87  south. Arkoses are restricted to the southern margin of the primary mineralization zone. The Cigar Lake Mine classified the altered basement into three geotechnical categories as it can range from very weak and deformable to competent rock: RM1 – Highly altered metapelite, predominantly within shear zones, that can be described as a graphitic silty sand, occasionally with a low clay content and displaying slight to low plasticity. RM1 rock contains ISRM strength grades R0 and R1. Zones of RM1 rock are subject to squeezing and creep based on previous excavations. Shear zones containing the weak RM1 material may be up to 10-15 m wide in the north-south direction (Golder, 2001) RM2 – Fractured and moderately to strongly altered metapelite containing some clayey silt, estimated intact rock strength between 10 and 30 MPa, and an RQD average of about 30-70%. Joint surfaces are typically planar, often slightly moist, smooth to rough, often graphitic. The surfaces are moderately altered, with smears of clay and mud and frequently slickensided RM3 – Weakly to moderately altered strong metapelite with a rock strength ranging from 25 to 50 MPa (corresponding to R3, moderately strong). Table 4.7 summarizes the unfrozen altered basement properties after Golder (2001) and CLMC (1989). Table 4.7: Material  RM1 RM2 RM3 Basement, no alteration noted  Moisture Content (%)  Friction Angle, φ º  Cohesion, c (MPa)  Elastic Modulus E (GPa)  Poisson’s ratio, v  UCS (MPa)  mean  -  34  0.32  1.5  0.40  1.2  mean  -  45  0.42  3.1  0.30  2.0  mean  -  57  0.69  14.3  0.20  4.7  11.9 ± 7.8  -  -  5.8 ± 4.6  0.16 ± 0.07  11.8 ± 12.1  72  -  -  0.01 - 13.7  0.06 - 0.35  0 - 45.5  1.9 - 26.8  -  -  33  28  48  Source Golder (2001) Golder (2001) Golder (2001)  CLMC (1989)  Altered Basement Unfrozen Material Properties  mean, stnd. dev range number of tests  88  Itasca Consultants (2008) sampled metapelite basement material from geotechnical boreholes 274 and 276 in 2008. A total of 3 UCS and 36 triaxial tests were conducted, along with bulk density, moisture content for index testing. Table 4.8 summarizes the unfrozen testing completed on the metapelite basement. Table 4.8:  Summary of Metapelite Basement Strength (Itasca, 2008)  3  Avg. Strength, (MPa) 71.3  Std. Dev., (MPa) 9.0  Avg. Modulus, E (GPa) 25.9  Std. Dev., (GPa) 9.2  13  98.9  8.1  19.9  5.6  16  108.2  8.2  17.5  6.1  7  140.5  10.2  23.7  4.3  19.6  6.0  No. of tests Uniaxial testing Testing at 2 MPa Confinement Testing at 6 MPa Confinement Testing at 10 MPa Confinement All tests  Golder (2001) noted that RM1 altered basement rock samples tested at -15oC with confining stresses of up to 5 MPa exhibited a higher cohesion but a significantly reduced friction angle by ten degrees compared to unfrozen rock. The strength of the rock was noted to be dependent on the effective stress state at the time of freezing. Table 4.9 summarizes the frozen testing data completed on altered basement rock from Golder (2001). Segregation potential tests on frozen altered basement material (Golder, 2001) demonstrated that RM1 altered basement rocks tend to form ice lenses at low stress in the laboratory; however, they believe there is little potential for ice lens formation above 1 MPa stresses.  89  Table 4.9:  Material Basement RM1 Basement RM2  4.6  Altered Basement Frozen Material Properties  Density  Moisture Content, (%)  Friction Angle, φ  Cohesion, c (MPa)  Elastic Modulus, E (GPa)  Intact UCS (MPa)  Test Temperature (oC)  2100  15  7.5  0.5  1  1.1  -15  -  9  40  0.4  2  1.7  -15  In-Situ Stress Measurements  Golder (2002) completed borehole hydrofracture tests in the sandstone above the ore. The minimum principal stress was 87% of the overburden pressure and the maximum principal stress was 115% of the overburden pressure for a rock mass with a saturated density of 2,500 kg/m3. The exact depth and testing methodology of hydrofracture testing was not included in the Golder (2002) report.  90  5.  Back-Analysis of Historical Data  This section discusses the mining experience in frozen ground at Cigar Lake mine and McArthur River, both owned and operated by Cameco Corporation. 5.1  Comparison of Cigar Lake and McArthur River Mines  This section compares the geotechnical parameters between Cigar Lake and McArthur River mines, both operated by Cameco Corporation, in order to provide recommendations on data collection and data management for artificial ground freezing design. Both McArthur River and Cigar Lake are unconformity related deposits mining in areas of very weak rock with artificial ground freezing. Geotechnical core logging and laboratory testing for freeze wall design has been minimal at both the mine sites. McArthur River differs from Cigar Lake mine in terms of geology, extraction methods, support and freeze pipe configuration. Table 5.1 presents a comparison of the McArthur River and Cigar Lake min with regards to the mine design, geology, hydrogeology, and ground freezing design.  Table 5.1: Mine Design  •  •  • •  •  •  Comparison of McArthur River and Cigar Lake Mine  McArthur River Mine McArthur River initiated ground freezing in 1999 to reduce the risk of potential water inflow adjacent to drifts. The orebody is surrounded on three sides by fairly dry competent ground and the other sides by highly fractured sandstone, with significant amounts of flowing sand and clay regions. The frozen wall barrier was designed to permit drainage of water to reduce water pressure. The wall also required to provide structural support of weak clay/ore near mining cavities (GeoAnalysis, 2000). Production began in early 2000 within Zone 2 is in a steeply dipping orebody situated almost entirely in dry basement metapelite aligned parallel to the subvertical P2 fault zone near the contact with water saturated sandstone. Development is driven entirely with the basement metapelite, consisting of an upper  •  •  •  •  • •  •  Cigar Lake Mine As is the case at McArthur River, development will take place entirely in the basement metapelite. The Cigar Lake orebody will be frozen prior to mining due to the relatively low rock strength and proximity of the overlying sandstone aquifer. Planned production will use jet boring technology to enable mining from below the ore zone. A zone of intense clay alteration that is not present at McArthur River caps the Cigar Lake orebody. Freezing at Cigar Lake will incorporate the entire ore zone within each production panel. Freeze holes will either be drilled vertically through the ore zone from the 480 Level or will be drilled subhorizontally from the 465 Level from above and below the ore limits. It is anticipated that the minimum thickness  91  •  Geology  •  •  •  •  Hydrogeology  •  •  •  •  McArthur River Mine level to create the freeze wall and for set up of the production raisebore and a lower level for retrieval of raisebore cuttings. The freeze wall for the production area is positioned in a U-shape with the crest of the freeze wall primarily in the water saturated sandstone and two walls extending into the basement metapelite.  Cigar Lake Mine of frozen ground above the ore zone will be at least 10 meters.  The orebody is located 550 m to 620 m below surface where the groundwater pressure is approximately 5.5 MPa. The ore zone can be divided into a high grade pelite/pitchblende matrix and a low grade clay and sand rich quartzite matrix below the high grade zone. The properties of the rock mass vary considerably, particularly with increasing levels of alteration. Although extensive testing has been conducted to determine rock strengths, limited testing has been performed to determine the other mechanical properties of the rock. The hanging wall and the lower footwall of the P2 fault zone are composed of basement rocks. The hanging wall contains primarily a pelitic gneiss sequence, whereas the lower footwall basement rock is dominated by quartzites.  •  Groundwater is present largely to the footwall of the zone but can be present in appreciable quantities in the low grade quartzite zone. The source of the groundwater is the more permeable sandstone unit. There is evidence of considerable vugginess in the quartzite unit which as most times is water bearing. The most troublesome unit is immediately to the footwall of the high grade pelite ore. A saturated clay bearing unit varies in thickness from 1-8 m is present. This unit is often adjacent to water bearing units, which can provide the necessary motive force to mobilize this clay unit.  •  Post-mineralization fracturing is the dominant control of hydraulic conductivity as fractures cut the otherwise impervious clay/sandstone core of the deposit acting as conduits for water, sand and soft clay.  •  The highest hydraulic conductivity occurs in the sandstones with the altered sandstone being greater than that of the unaltered sandstone. Within the sandstone formation, the hydraulic conductivity measurements ranged from 7 x 10-10 m/s to greater than 5 x 10-6 m/s, with the majority of the measurements between 10-9 m/s and 10-8 m/s.  •  Within the basement rock masses, the hydraulic conductivity is entirely fracture controlled and two to three orders of magnitude below that of sandstone, typically  The hydraulic conductivity of small volumes of rock is difficult to determine due to the  •  •  •  The Cigar Lake orebody is a flat lying structure with a crescent shaped profile. The orebody is located at an approximate depth of 430 - 450 m at the unconformity between the Athabasca sandstone formation and the underlying basement rocks. The deposit is approximately 1,950 m long, 20 to 100m wide, and ranges up to 12m thick, with an average thickness of about 5m. Above and below the unconformity, the rock mass shows variability in porosity and permeability due to fracturing and alteration processes  92  McArthur River Mine presence of non-uniformly spaced and sized fractures and clay/sand infill material. • •  •  Ground Freezing Design  •  •  Cigar Lake Mine due to the tightness of the fracturing and the clay and chlorite alteration of the fracture surface, particularly in the graphitic metapelite. The basement rock has typical hydraulic conductivity values from 10-11 to 10-10 m/s.  However, on a large scale, it is possible to define and measure a bulk conductivity. Golder Associates (1995) completed long term inflow testing (3 days) resulted in wide spread impacts both horizontally and vertically in the sandstones suggesting that the sandstone is well fractured both in the horizontal and vertical and the fractures are well interconnected. From the analysis, the vertical hydraulic conductivity of the sandstone is approximately 3 x 10-4 cm/s, about 3 times greater than that of the horizontal hydraulic conductivity  At McArthur River, only thermal parameters of unfrozen and frozen materials have been directly measured to date. Frozen compressive strength, triaxial creep testing has been completed on indurated clay and altered sandstone material from three boreholes at Cigar Lake.  •  •  • •  5.2  At Cigar Lake, mining will be conducted from the 465 m production level which is located 10 m below the deposit. Artificial ground freezing will be implemented to support the weak rock associated with the orebody, minimize the potential for a large water inrush and stop radon migration. Jet boring is the proposed plan to mine out the Cigar Lake orebody. The cutting of the ore with high pressure water is expected to produce cavities fairly circular in shape measuring 4 to 5 m in diameter.  Cigar Lake Mine, Jet Boring Trial in 2000  At the Cigar Lake mine, four cavities in frozen waste rock, just below the orebody, and four cavities within the ore were excavated as part of a jet boring test program in mid-2000. The purpose of the jet boring trial study was to determine the potential cavity sizes, production cycle times, and cavity stand-up time to backfilling. The test mine area excavated in 2000 is located near 10700E and 10000N on the mine grid. The factual report of the test mine geology, ground conditions, and jet boring results are presented in “2000 Jet Boring Systems Test – Final Report” 93  (Cameco, 2000). This section will discuss the test results and interpretation of the four jet bored cavities in the orebody. The study area was frozen from the production level (480 level) below the ore through near vertical freeze pipes installed up into the orebody (~430 level) with calcium chloride circulating at -40oC. The area was allowed to freeze to -20oC prior to mining. An intermediary level (460 level) above the freeze level was mined for the trial study to drill the pilot holes up into the ore body and develop the test cavities. After the pilot hole for the test cavity was lined with casing, a drill string with a nozzle was inserted in the casing and while rotating from the top of the planned cavity down, pressurized water jet opened the cavity. The upper part of the cavity was noted to grow laterally while jetting occurred lower down, though no uncontrolled sloughing was observed. The ore slurry left the cavity by gravity and was pumped away from the mining area, resulting in fairly circular cavities 2 m in diameter and up to 5 m in height. The cavities were left open for several days before backfilling with concrete.  5.2.1 Geology The geology in the test mine area comprises three rock types, the basement (altered metapelite), ore zone and clay cap, and directly above is the altered sandstone. Figure 5.1 shows the typical geology encountered in the jet boring trial study and the cavity dimensions.  94  Figure 5.1:  Jet Boring Cavity Geology and Schematic of Surveyed Trial Cavities, after Cameco (2000)  The basement rock in the test trial comprised moderately to extremely clay altered graphitic metapelite. Immediately below the orebody, the first 1 to 5 m of the basement is defined as a medium strong clay. The orebody, overlying the basement, varied from 4 to 6.5 m in thickness in the test zone and was located at the unconformity. The orebody comprised three distinct zones varying in hardness and mineralization. Including: • • •  Massive high grade mineralization (less than 10% by volume of the orebody), a very hard, heavy rock with an average UCS of 50 MPa. Altered and friable sandstone (less than 20% by volume of the orebody), a very weak to weak rock with a UCS ranging from 1 to 25 MPa. Clay/Claystone (approximately 70% by volume of the orebody) is an intermediate to indurated sandy clay to claystone with a UCS ranging from 5 to 15 MPa.  The strength of the ore zone tested in the jet bored cavities was estimated based on the ore grade, measured by gamma probing. The ore strength generally increases with the grade based on past 95  experience at the Cigar Lake project by Cameco (2000). The percent ore grade: • • •  Between 0% and 15% relates to a UCS of less than 15 MPa, Between 15% and 25% relates to a UCS between 15 and 40 MPa, and An ore grade greater than 25% is comparable to a UCS greater than 40 MPa.  The test zone cross-cut at the orebody level has an average UCS of 10 MPa (ranging from 0.25 to 35 MPa). Overlying the orebody is the sandstone typically altered in the first few meters to a sandy indurated clay. The altered sandstone encountered is fractured and extremely altered, typical of this rock type at the site.  5.2.2 Instrumentation Temperature probes were installed within the row of freeze pipes at the top and midpoint of the orebody. When test mining commenced in September 2000, the rock mass temperature of the ore zone was measured to be -20oC. The area reached -10oC within the first four months of freezing, typical of the freezing times experienced at the McArthur River mine. The base of the orebody was observed to be approximately 6oC warmer (-14oC) than the midpoint of the ore and underlying and overlying rock masses. Cameco (2000) attributes this temperature fluctuation due to a higher pore water content and clay content at the unconformity, (the base of the orebody) overlying the basement. Ground freezing of the test mine area was assumed to be complete as no water was observed during jet boring or the drilling of temperature monitoring probes. Geotechnical instruments to measure the rock mass behaviour and ground support response to the jet boring of frozen ground included pressure cells to monitor ground loading on the cross cut support, a tape extensometer to measure convergence of the 713 cross cut, instrumentation on drill holes in the 480 level, and caliper surveys to measure convergence of the cased test holes.  5.2.3 Influence of Freezing on Weak Altered Rockmass Jet boring testing in four frozen ore cavities was undertaken over several weeks in September 2000. The cavities excavated at approximately 435 m depth were within the orebody extents ranging from 4 to 6 m thickness. The orebody, as noted above is a highly variable rock mass, with a strength ranging from very weak to medium strong rock (UCS from 1 to 40 MPa), fractured and containing wide zones of rock altered to clay. The dimensions and estimated ISRM 96  rock strength of each cavity is noted below in Table 5.2.  Table 5.2:  Cavity No. 1 2 3a 4  Volume, (m3) 65.5 30.1 62.7 79.0  Maximum Span in Cavity, (m) 6.0 3.5 5.5 5.5  Cigar Lake Jet Boring Trial Dimensions Average Cavity Radius, (m) 4.0 3.0 5.0 4.5  Cavity Height, (m) 4.8 3.0 5.0 5.5  Average Grade 25% 8% 8% 19%  Avg UCS based on Average Grade, (MPa) 15 – 40 < 15 < 15 15 – 40  Estimated Unfrozen RMR76 < 35 < 35 < 35 < 35  Cameco (2000) included several cross-sections of the caliper surveys in each ore cavity. The cavities (1, 2, 3A, and 4) were surveyed with a laser range finder after mining completion at 300 mm vertical increments. The ore grade plotted along the vertical scale of each cavity was estimated over 50 cm intervals from the cavity gamma survey. Based on the relationship discussed earlier relating ore grade to rock strength, the rock strength has been estimated and drawn on each cavity included in Figure 5.2 to Figure 5.5. Table 5.3 summarizes the span compared to the estimated rock strength for each cavity.  97  Table 5.3: Cavity  1  2 3A  4  Cigar Lake Jet Boring Trial Span Compared to Rock Strength  Surveyed Elevation (m) Top to Bottom 24.0 - 22.2 22.2 - 21.2 21.2 - 19.6 19.6 - 19.5 19.5 - 19.2 24.2 - 21.6 22.5 - 22.3 22.3 - 21.6 21.6 - 17.7 23.8 - 22.3 22.3 - 21.5 21.5 - 20.5 20.5 - 19.9 19.9 - 18.6  Average Excavated Span by Jet Boring (m) 5.0 (back of cavity) 4.4 3.2 3.2 3.0 (base of cavity) 2.4 - 3.8 4.4 (back of cavity) 4.4 5.0 (base of cavity) 3.0 (back of cavity) 4.0 3.6 4.0 4.0 (base of cavity)  Estimated Rock Strength (MPa) < 15 15 - 40 > 40 15 - 40 < 15 < 15 < 15 15 - 40 < 15 < 15 15 - 40 > 40 15 - 40 < 15  Figure 5.2 to Figure 5.5 are after cross-sections drawn in the report “2000 Jet Boring Systems Test – Final Report” by the Cigar Lake Mining Corporation (CLMC, 2000).  98  Figure 5.2:  Cavity 1, Jet Boring Survey of Ore Cavity, UCS Based on Ore Grade  Figure 5.3:  Cavity 2, Jet Boring Survey of Ore Cavity, UCS Based on Ore Grade  99  Figure 5.4:  Cavity 3a, Jet Boring Survey of Ore Cavity, UCS Based on Ore Grade  Figure 5.5:  Cavity 4, Jet Boring Survey of Ore Cavity, UCS Based on Ore Grade  100  5.2.3.1 Increase in Strength and Rock Mass Rating With Freezing Based on the empirical relationship between span and RMR for weak rock masses (Section 2.6), the stable span for unsupported ground given an RMR less than 35 is no greater than 3 m (Figure 2.18 after Ouchi, 2008). However, the average cavity diameter of the four cavities jet bored in the frozen ore measured 4 to 6 m in width (refer to Table 5.2). The four cavities were left open for several days with no deterioration or ground instabilities noted before backfilling with a cement concrete. Figure 5.6 plots the unfrozen to estimated frozen RMR on the McArthur River developed critical span rock mass curve to show the gain in strength with freezing during the jet boring trial. Note the McArthur River span rock mass curve is for excavations with ground support.  Jet Boring Trial Unfrozen Frozen  Figure 5.6:  Jet Boring Cavity Span on the McArthur River Critical Span Curve with Ground Support, after Pakalnis (2012)  The influence of freezing on weak rock is clearly shown to increase the rock mass conditions from an estimated unfrozen RMR of less than 35 of the jet bored cavities to approximately 50 (based on the stable unsupported line for a 5 m span). This increase in the frozen rock mass strength is attributed to the increase in cohesion and UCS of the weak rock as the pore water 101  freezes. A detailed discussion of the frozen lab testing in Sections 7 presents the observed influence of freezing on a weak rock mass based on unconfined compressive strength and four-point beam testing. This is corroborated by the Wardrop (2005) report on the increase of span opening in frozen ground of several Russian underground mines for similar rock in unfrozen ground. 5.2.3.2 Creep Behaviour Time dependent deformation of a rock or soil without changes in the stress state is defined as creep. Factors influencing the time-dependent behaviour of a rock include its mineralogy, fabric, moisture content, porosity, stress, applied strain rate, and temperature. Frozen soils are susceptible to creep and relaxation due to the presence of ice and unfrozen water. The creep response of ice varies with different soils given the potential of ice lens formation. The basic creep curve comprises three stages; (1) primary (strain-hardening), where the creep rate is decreasing, (2) secondary (linear), where the creep rate is constant, and (3) tertiary (strainsoftening), where the creep rate is increasing. Cavity 3a exhibited creep in the lower grade ore, attributed to the slightly higher ground temperature and higher pore water content of this clay rich zone (Cameco, 2000). The convergence measured from a borehole calliper survey, occurred within the first four hours after drilling the pilot hole. This creep behaviour was expected in the clay rich weak rock and frozen ground, and is consistent with previous experience by Cigar Lake mine. The total convergence in this section of Cavity 3a is up to 8 cm, 21% of the hole diameter after 10.5 hours and is within the lowest grade of the ore zone, interpreted to be the highest clay rich portion. The test hole was drilled vertically upward from the cross cut intersecting the unconformity at a height of 17.5 to 18 m above the cross cut. The orebody extended to 23.5m along the hole, a thickness of 6 to 6.5 m. Convergence readings taken at 30 second intervals measured inward displacement only within the lower half of the ore zone from 16 to 21 m above the cross cut. A second borehole, No. 2 drilled in the 1991 test mining also displayed convergence in the pilot hole measuring up to 11% displacement equal to a closure of 38 mm. This borehole was drilled as part of the initial test mining trials of boxhole boring and jet boring studies.  102  Results from both boreholes (No. 2 in 1991 and Cavity 3a in 2000) showed that convergence occurred after the first four hours of drilling and none after that time. Given that the creep rate accelerated for the first four hours and then remained constant before backfilling the cavity, the convergence occurred through the primary strain hardening and secondary linear portion of the creep curve. Creep testing of the collected rock core from the Cigar Lake mine in 2009 was not completed as part of the frozen lab testing program. Typical creep rates for dense clay to sandy clay should be established with unfrozen and frozen creep testing. The rock mass is very poor and weak and will squeeze/creep under unfrozen conditions due to the weak rock mass and under frozen conditions due to the flow of ice over time. When frozen soil deforms its structure changes continuously with varying influence by density, ice content, temperature, and confining pressure.  5.3  Rock Mass Classification Comparison of Frozen to Unfrozen Conditions at the McArthur River Mine  The increase in the RMR76 value from an unfrozen to frozen state was recently assessed by Pakalnis and Mawson at Cameco’s McArthur River Mine (Mawson, 2012a and Mawson, 2012b). Ground freezing has been used at the McArthur River mine since the early 2000’s as a barrier from the porous water bearing Athabasca sandstone and less for increasing the rock mass strength of very weak rock. However, Cameco noted an increase in the competency of the rock in areas which were frozen on the 510L at McArthur River mine. Mawson (2012a and 2012b) compared assumed frozen RMR76 values from geological face mapping of the 510L with unfrozen RMR values from geotechnical core logging of five unfrozen core logs in the same area. The data presented in this section from the 510L at McArthur River; headings 8240N and 8220N, and core logs from diamond drill holes 2903, 2907, 2917, 2037 and 2573. The trajectory of the drill holes can be seen in Figure 5.7.  103  Figure 5.7:  510L RMR Values and Diamond Drill Hole Trajectories  The 510L is considered high risk mining because it is located in close vicinity to the water bearing unconformity and in some cases actually passes through the unconformity on this level. The ground in the vicinity of the unconformity was frozen prior to development to ensure that any water bearing features would be sealed off. Heading 8225N was the first drift which was mined through the unconformity, though no face mapping with RMR76 calculations were done in this drift. The next heading mined in frozen conditions was the 8240N, and RMR76 calculations were done with the face mapping for the length of the drift. Following this, the 8220N slash was 104  developed off of 8225N, again with RMR76 values being recorded for the length of the drift. The results of face mapping data were compared to corresponding unfrozen core logging data. In general unfrozen face mapping data showed a slight increase in RMR76 parameters from the core logging data; this can be attributed to scale, orientation and differences in mapping techniques as opposed to core logging techniques. Within the sample set which was analyzed, the RMR76 parameter which was most greatly affected by ground freezing was the joint condition parameter. The average increase from unfrozen core to frozen mapping was over 10 points but was as much as 15 points. Figure 5.8, Figure 5.9, and Figure 5.10 plot the unfrozen to frozen RMR values along the 5108240N and 510-8220N drifts. Ground improvement due to freezing appears to increase with decreased ground competency.  Figure 5.8:  Combined Results of Core RMR vs. Drift RMR  105  Figure 5.9:  510-8240 Drift RMR Compared to Rock Core RMR  106  Figure 5.10: 8220N Drift RMR Compared to Rock Core RMR  Pakalnis and Mawson (Mawson 2012a, and Mawson 2012b) showed that the RMR76 increases by an average of 38, for rock mass with RMR unfrozen of approximately 40 or less. Table 10 summarizes the average increase for each of the five parameters in the RMR system.  107  Table 5.4: Average Increase Between Frozen Face Mapping and Unfrozen Core Logging (Mawson, 2012) Parameter  Average increase in RMR76 value (unfrozen to frozen) ±8 Rock Strength ±7 RQD ±11 Joint spacing ±11 Joint condition 0 Water TOTAL average ±38 RMR76 increase The sample size is small; however this study is a good basis for future studies. These studies should use data from both unfrozen drill core and frozen excavated faces. Comparing the influence of freezing from unfrozen to frozen conditions is recommended to be with frozen face mapping and unfrozen drill core rather than comparing the frozen span and frozen rock mass conditions in the previous Wardrop (2005) studies.  108  6.  Cigar Lake Geotechnical Material Properties Based on 2009 Drilling  This section discusses the geotechnical domains that will be used to assist in designing the jet bored cavities based on the geotechnical drilling and material properties from the previous section including the 2009 surface freeze drilling campaign boreholes. 6.1  Cigar Lake Geotechnical Domains  As discussed in Section 0, the Cigar Lake orebody is highly variable and overlain by varying degrees of altered sandstone. Figure 6.1 illustrates the highly variable nature of the material surrounding the ore with cross-sections from MDH (2008). A (South)  A’  Cross-Section A A’  A Figure 6.1:  B (South)  B’  Cross-Section B B’  B  Geological Variability of Material at the Cigar Lake Mine, after MDH (2008) 109  For the 2009 surface freeze drill program, the rock descriptions applied in this research were modified from previous nomenclature. The samples were logged by alteration not by stiffness as the soft, intermediate and indurated clays are not located with consistent spatial order from the orebody. Understanding the rock mass quality with vertical distance away from the orebody will be the focus for defining the geotechnical zones over grouping by lithology. The alteration and fracturing of the rock overlying the orebody is highly variable and inconsistent between boreholes. Golder (2002) noted that there does not appear to be any trend relating rockmass conditions above the ore nor was there any general pattern in the drillholes indicating the location of intermediate or indurated clay over the orebody. Based on the 2009 boreholes, it was noted that the rock types generally followed the following lithology sequence. • • • •  Competent good quality sandstone of the Athabasca Formation overlying the orebody. With decreasing rock mass quality from approximately 30 to 40 m above the ore. Increasing fracturing and alteration of the sandstone occurs to within 10 to 20 m of the orebody. The outer 10 to 20 m of the orebody comprises a highly altered (bleached sandstone) where the rock mass is very poor quality, white, sandy clay to friable sandstone. Directly overlying the orebody, lays the "clay cap", though as mentioned previously is not a massive continuous clay cap over the orebody. Instead the material overlying the orebody is termed hematized sandstone, referring to the iron oxidation alteration process. The hematized sandstone is typically extremely to very weak sandstone or a dense sandy clay. The hematized sandstone is stiffer and contains more of the sandstone rock fabric than the bleached sandstone. The hematized sandstone is not present or a continuous layer over the orebody, in some areas the bleached sandstone directly overlies the orebody.  Table 6.1 presents the rock descriptions applied in this research to the Cigar Lake material. Table 6.1:  Summary of Rock Formations and Rock Descriptions Used for the 2009 Geotechnical Logging of Samples  Rock Type  Origin / Formation  Description  Sandstone  Manitou Sandstone Manitou  White to pinkish grey, fine grained, medium strong, fresh to slightly weathered, RQD 60-100%. White to pinkish grey, fine grained, medium strong,  Altered  Falls Falls  Average Thickness 400 m 25 m  110  Rock Type  Origin / Formation  Description  Sandstone Bleached Sandstone Clay Hematized Sandstone Clay Ore  /  Sandstone Manitou Sandstone  /  Manitou Sandstone  slightly weathered, increasing fracturing RQD 40-70%. White, hydrothermal bleaching, massive clay to mixed sandstone and clay, soft clay to extremely weak rock, moderately to highly weathered. Zones of core loss Red to greyish red, close proximity to ore, intermediate/indurated clay to weak rock, structural fabric and jointing still present Greyish green, very weak to medium strong, slightly to moderately weathered, clay banding, increasing rock hardness with ore content, Graphitic metapelite, green, extremely weak to very weak, clay and pebble (gritty) mixture, moderately weathered. RQD 70-90% Graphitic metapelite, green, strong, fresh to slightly weathered. RQD 80-100%  Altered Basement Basement  Faulting Hydrothermal Alteration Pre-Cambrian  Pre-Cambrian  Falls  Falls  /  Average Thickness 5 - 10 m  5m  3 - 10 m  5m  -  Bleached Sandstone Hematized Sandstone  Ore  Figure 6.2:  Borehole ST791-05, from 433.45 to 442.4 m  The clay cap from here on will be represented by the hematized sandstone and bleached sandstone, the material intersected in all 2009 surface freeze drillholes both overlying the orebody. Caution should be exercised on relying in this material to be entirely overlying the orebody as Itasca (2009) commented on the discontinuous and heterogeneous nature of the "clay cap" comprising very weak sandstone to stiff/very stiff clay. 111  6.2  Historical Geotechnical Drilling  The purpose of this section is to compare previous geotechnical summaries of the material overlying the orebody with the data collected from the drilling program. Prior to the 2009 drilling program there was insufficient geotechnical data to characterize the hematite-rich “clay cap” material overlying the orebody. The material directly overlying the orebody has commonly been described as a massive clay rich zone averaging 1 to 5 m thick with a maximum thickness 10 m. However, discontinuous zones of intermediate clay, indurated clay and very weak to weak sandstone are present. The historical geomechanical database provided by Cigar Lake Mine contains the geotechnical parameters (recovery, RQD, strength, weathering, and lithology) of 48 boreholes drilled in the 1980's and 1990's. Joint condition and joint alteration were not routinely logged and therefore the historical boreholes are not appropriate for calculating Rock Mass Rating (RMR) parameters, to establish the degree of alteration and fracturing around the ore body. RQD data was the only parameter routinely collected. However, this data is extremely suspect as high RQD values were given to intervals of very weak rock (S6 to R1); as previously noted, core with a UCS of less than 1 MPa (less than R1) are not supposed to be included in the RQD and should have been assigned a RQD of zero. NQ boreholes (48 mm core diameter) were drilled and logged by Cameco's geologists or technicians. It was noted by the author that the level of accuracy of the geotechnical parameters especially RQD percentages did not reflect the rock strength or recovered core length for the same drill interval. Drill runs with a strength of less than R1 (soil like) were often recorded as 100 % RQD. Using RQD % alone from the historical drilling may imply the ground over the orebody is stronger than it actually is. Reviewing all collected geotechnical drillhole information there is a lack of consistency between various data sets as the majority of boreholes drilled in the beginning of the exploration program were not specifically logged for geotechnical purposes and therefore lack completeness.  6.3  2009 Material Properties Drilling Program  Geotechnical boreholes to characterize the orebody and surrounding area have been completed 112  from the mid 1980s to present. A diamond drilling contractor was retained from February to April 2009 to complete a surface freeze drilling program located approximately 150 m north of Shaft 1, at the Cigar Lake Mine site. Prior to the 2009 surface freeze hole drilling program there was insufficient geotechnical data to characterize the hematite-rich “clay cap” material overlying the orebody. Eight PQ size drillholes were logged geotechnically to 450 m depth and samples collected continuously above, within, and below the orebody to better define the rock mass overlying the orebody. Table 6.2 lists the boreholes drilled and used for frozen laboratory testing part of this research. Table 6.2:  Summary of 2009 Surface Freeze Holes for Geotechnical Sampling  Borehole ID  Easting  Northing  SF79106 SF79107 SF80104 SF80105 ST78607 ST79105 ST79605 ST80103  10791.0  10027.5  10791.0  Borehole Dip/Dip Direction  90/NA  Bleached Sandstone Top Thick. (m) (m) 407 24  Lithology Intersections Hematized Orebody Basement Sandstone Top Thick. Top Thick. Top Thick. (m) (m) (m) (m) (m) (m) 431 0.3 431.3 12 443.3 -  10032.5  90/NA  400  30  430  3.2  433.2  2.6  435.8  -  436.5  10801.0  10027.5  90/NA  -  -  432  5.1  437.1  3.3  440.4  -  438.4  10801.0  10032.5  90/NA  400  29  429  5  434  4.8  438.8  -  437.3  10786.0  10020.0  90/NA  410  21.6  -  0  431.6  8.7  440.3  -  439.3  10791.0  10022.5  90/NA  422  12  434  1.15  435.15  8.35  443.5  -  439.25  10796.0  10030.0  90/NA  410  20.5  430.5  2.3  432.8  6  438.8  -  437.9  10801.0  10022.5  90/NA  422  12  434  3.2  437.2  4.5  441.7  -  440.2  Unconformity Depth (m) 435.3  Three predominant material types overly the orebody: (1) intermediate clay, (2) indurated clay / very weak sandstone, (3) weak sandstone. The weakest material is intermediate clay locally up to several meters thick. From the surface freeze drillholes, the orebody ranges from 3 to 15 m thick with an average of 6 m. The hematized sandstone (clay altered sandstone) typically directly over the orebody ranges from 2 to 5 m thick. The material (hematized and bleached sandstone) overlying the orebody is clay rich comprising discontinuous zones of very weak sandstone to stiff/very stiff clay. The highly altered zone above the orebody averages 10 m and extends up to 113  15 m thick. The extremely altered zone commonly thought as massive clay several meters above the orebody is not consistent between drillholes. The purpose of the material properties data collection program was to address data gaps from historical geotechnical drilling and provide an understanding of the shear strength and time dependent behaviour of weak frozen rock under pressure. From the eight boreholes drilled, samples were collected from four boreholes. Acrylic liners were placed inside the core barrel instead of metal splits to minimize sample handling and disturbance on surface. The 1.5 m long acrylic tubes were sealed on either end at the drill rig and stored inside the Cigar Lake core warehouse prior to shipment for laboratory testing. 6.4  Geotechnical Logging  Detailed geological/geotechnical logging and digital photographing of the 2009 drill core was undertaken under the direction of Cameco at the Cigar Lake Mine core shack. Soil classification was based on the Unified Soil Classification System and the rock core logging comprised the following: • • • • • • •  Total core recovery (%) Rock quality designation (RQD %) Detailed geology (rock type, colour, mineralogy, texture, weathering, etc.) Fractures (count, type, infill, roughness, alteration, aperture, angle, etc.) Bedding ISRM estimate of rock strength Calculation of NGI-Q and RMR76  The following sections discuss some of the input parameters (RQD and strength) logged for the rock mass classification in order to develop cross-sections from the surface freeze drilling campaign to illustrate the benefit of ground freezing to increasing the rock mass quality discussed in Section 6.5. 6.4.1 Rock Quality Designation Rock quality designation was recorded for all drilled boreholes of the 2009 surface freeze drilling campaign per 1.5 m drill interval. Figure 6.3 summarizes the Rock Quality Designation (RQD) values for the historical boreholes where the data was verified against other geotechnical parameters (strength, and fracture spacing) and the 2009 surface freeze drilling campaign. 114  Figure 6.3:  Rock Quality Designation Plots of Geotechnically Logged 2009 Drillholes  115  6.4.2 Rock Strength Intact rock strength is defined as the load per unit area at which a UCS sample fails and can be estimated by using standard field identification methods such as a knife or hammer, point load testing apparatus, or directly in the laboratory with a UCS load frame. Table 6.3 summarizes the unfrozen field strength of the holes that were geotechnically logged from the 2009 surface freeze drilling program. The field strength of the rock core in 2009 was measured by the geologist with a knife or hammer. No point load testing was completed on the rock core. Table 6.3:  Field Strength of Geotechnically Logged 2009 Drillholes  Lithology Sandstone Altered Sandstone (Bleached) Altered Sandstone (Hematized) Ore Altered Basement Basement  Average 16 12  Field Strength (MPa) Minimum 0 0.5  Maximum 37.5 37.5  4  0.5  25.0  8 4 13  0 0.5 0  37.5 25 37.5  6.4.3 Joint Condition The structural data collected by the Cameco geologists in the borehole logging is summarized below in Table 6.4 and Table 6.5. The majority of the identified joint surfaces in the sandstone are rough and planar with increasing alteration towards the orebody. Within 30 m from the orebody, the joint surfaces are typically coated with silty sand infill 2 to 5 mm thickness. Within several meters above the orebody, the sandstone rockmass has altered to sandy silty clay. Directly below the unconformity, lies the metapelite basement, which also shows increasing alteration within proximity of the orebody. Joint surfaces decrease in alteration and infilling 15 to 20 m below the orebody.  116  Table 6.4: Lithology Sandstone Altered Sandstone (Bleached) Altered Sandstone (Hematized) Ore Altered Basement Basement Table 6.5: Lithology Sandstone Altered Sandstone (Bleached) Altered Sandstone (Hematized) Ore Altered Basement Basement  6.5  Joint Roughness of Geotechnically Logged 2009 Drillholes Average 2.6 2.5  Joint Roughness (Jr) Minimum 0 1  Maximum 3 3  2.0  1  3  1.8 2.5 2.4  0 1 0  3 3 4  Joint Alteration of Geotechnically Logged 2009 Drillholes Average 4–8 10 – 15  Joint Alteration (Ja) Minimum 1 1  Maximum 15 15  6 – 15  1  15  4–8 8 – 15 4–8  4 3 1  15 18 15  Interpretation of the Lithology and Rock Mass Characterization  An assessment of the overall rock mass quality was completed for the surface freeze drillholes from the geotechnical database as recorded by Cameco geologists. Both the Rock Mass Rating (RMR) (Bieniawski, 1976, 1989) and Q-System (Barton et al., 1974) were calculated. The following presents the results of the assessment of Q and RMR (1976 and 1989) per drill run interval. Note that a typical drill run interval was 1.5 m (5 ft). For both RMR calculations, a groundwater rating for dry conditions has been assumed for the purpose of assessing the geomechanical characteristics of the rock mass in the absence of 117  external factors. For certain design applications, it may be necessary to adjust the rock mass quality to account for the expected water conditions. The Cigar Lake rock mass around the orebody is generally medium strong to strong, blocky with preferential joints along bedding, and fair to good quality. Poor rock zones (shown in red in Figure 6.5 to Figure 6.8) are generally very weak to weak and associated with faulted areas and high degrees of alteration. Faults encountered to date can be described as poor to good quality, depending on the relative intensity of fracturing and infilling within the fractures. Table 6.6 summarizes the measured rock mass classification values for main lithologies observed in the 2009 surface freeze drilling program.  Table 6.6:  Unfrozen RMR76 and Q' of Geotechnically Logged 2009 Drillholes  Lithology Sandstone (below 400 m elev.) Altered Sandstone (Bleached) Altered Sandstone (Hematized) Ore Altered Basement Basement (to end of hole)  Rock Mass Rating (RMR76) Average Minimum Maximum 29 9 50 27 10 42 25 13 41 30 3 39 29 11 41 37 18 58  Lithology Sandstone (below 400 m elev.) Altered Sandstone (Bleached) Altered Sandstone (Hematized) Ore Altered Basement Basement (to end of hole)  Q' Minimum 0.06 0.01 0.01 0.01 0.01 0.01  Average 2.4 1.8 0.4 1.8 1.6 6.6  Maximum 45.0 25.4 3.0 4.0 3.5 50.0  118  6.6  Summary of 2009 Surface Freeze Drill Holes for Laboratory Testing Samples  Samples for laboratory testing were collected from the following boreholes listed below in Table 6.7. This table lists the recorded field strength, rock quality designation and rock mass rating for the ore and material overlying the orebody in each borehole.  Table 6.7:  Summary of Surface Freeze Borehole Field Strength, RQD, and RMR  Weighted UCS, MPa (field strength) 0-5m above ore  5 - 10 m above ore  10 - 15 m above ore  Weighted RQD 5 - 10 0-5m m above above ore ore  Weighted RMR 10 - 15 m above ore  0-5m above ore 23  5 - 10 m above ore -  10 - 15 m above ore 32  Hole  CrossSection  ST786-07  10775  -  9.3  -  26.2  -  45  -  68  ore -  ST791-05  10800  37.5  21.2  15.8  17.1  69  47  52  51  36  27  25  23  SF791-06  10800  9.5  2.1  12.6  5.0  85  68  12  24  28  22  16  23  SF791-07  10800  4.0  12.5  10.1  20.1  72  79  67  66  27  30  24  23  -  22  41  ore  ore  ST796-05  10800  3.0  -  15.0  9.0  51  -  54  52  29  ST801-03  10800  2.0  7.0  3.0  10.3  93  70  37  49  33  30  22  24  SF801-04  10800  -  -  -  12.0  -  -  -  59  -  -  -  34  SF801-05  10800  5.0  0.6  3.7  4.6  57  61  83  44  26  19  31  21  10  9  10  13  71  62  51  52  30  25  23  28  AVERAGE  NOTE: Boreholes ST786-07, ST796-05, and SF801-04 have not been logged in clay cap as core in acrylic tubes.  119  •  From approximately 400 m below ground surface to the top of the orebody, the rock mass quality decreases from an approximate RMR76 of 50 to an average RMR76 of 30 along with an observed strength decrease in field hardness from R2.5 (37.5 MPa ) to R1 (1 to 5 MPa).  •  There are no clear rock mass quality transition zones between boreholes or with depth as anomalous zones of very poor or medium strong rockmass are present.  •  Comparing the Rock Mass Rating (RMR), field strength, and Rock Quality Designation (RQD) transitioning upwards from the orebody with the decrease in alteration away from the orebody is not very helpful to establish trends in the geotechnical properties given the scatter of data.  •  The transition of alteration from the orebody may not be a vertical gradient with distance away from the orebody, but rather a mixture of materials controlled by faulting.  Figure 6.4 shows the cross-section locations and boreholes selected for laboratory testing. Figure 6.5 to Figure 6.8 plot cross-sections of the calculated unfrozen rock mass rating (RMR76) in the 2009 surface freeze drillholes and nearby historical drillholes. The purpose of these sections is to apply the relationship between the unfrozen and frozen RMR, developed in Section 8 and illustrate the gain in strength that is possible due to freezing conditions.  N 10,032  E 10,800  Borehole Geotechnical Logging Laboratory Testing  Figure 6.4:  for and  2009 Surface Freeze Holes for Laboratory Testing 120  ST786-07  ST791-07  ST796-05  ST801-05  Moderately/Highly Altered  RMR76 < 20 RMR76 20 – 35 RMR76 35 - 45  Extremely Altered  ore unconformity  Figure 6.5:  Cross Section North 10,032, Through Surface Freeze Holes, Unfrozen RMR76  121  ST801-03  ST801-04  ST801-05  Moderately/Highly Altered  Extremely Altered  RMR76 < 20 RMR76 20 – 35 RMR76 35 - 45 unconformity ore  Figure 6.6:  Cross Section East 10,800 Through Surface Freeze Holes, Unfrozen RMR76  122  ST791-05  SF791-06  SF791-07  Moderately/Highly Altered  RMR76 < 20 RMR76 20 – 35 RMR76 35 - 45  Extremely Altered  ore  Figure 6.7:  unconformity  Cross Section East 10,790 Through Surface Freeze Holes, Unfrozen RMR76  123  130  SF796-05  109  RMR76 < 20 RMR76 20 – 35 RMR76 35 - 45  ore unconformity  Figure 6.8:  Cross Section East 10,796 Through Surface Freeze Holes, Unfrozen RMR76  124  7.  Frozen Laboratory Testing  This section discusses the frozen Unconfined Compressive Strength (UCS), frozen four-point beam, and frozen direct shear testing on the core collected from the 2009 surface freezing drilling program. 7.1  Unconfined Compressive Strength Testing  Frozen Unconfined Compressive Strength (UCS) testing was completed to determine the influence of freezing on the short term strength of the Cigar Lake weak rock mass. Frozen soils are stronger than unfrozen ground due to the bonding of ice; however, how much stronger is a function of the temperature, moisture content, material, and applied strain rate. As with unfrozen soil, the strength of frozen soil depends on interparticle friction, particle interlocking and cohesion. The frozen strength varies with many factors, but those controlled in the laboratory testing are temperature, applied loading rate, and application of freezing. The unfrozen UCS is also a parameter in Bieniawski's Rock Mass Rating system (Bieniawski, 1976 and 1989) (refer to Section 2.5.1). Establishing unfrozen and frozen UCS values for the various Cigar Lake material types will be used to understand the influence of freezing on the empirical data of rock mass rating values vs. opening span for underground cavities.  7.1.1 Sample Collection Samples were collected with a diamond drill from surface as part of Cigar Lake’s surface freeze drill program in February and March 2009. All samples were drilled PQ (83 mm) with acrylic tubes inside the core barrel to eliminate sample disturbance at the drill rig. Samples were collected in the unfrozen state and left in the acrylic tubes to preserve moisture. After drilling, the acrylic tubes were capped, and shipped off to the laboratory, and stored in a moisture and humidity controlled environment prior to testing. The target sampling zone from the material properties drill program were the materials above and below the orebody to be influenced by the bulk freezing. As described in Section 4, altered Athabasca sandstone unconformably overlies altered metapelite basement. The orebody is a highly altered uranium rich heterogeneous mixture of pitchblende, pitchblende-rich clay, 125  pitchblende-impregnated sandstone, clay, silt, and sand. The University of Alberta laboratory was not equipped to handle and test samples greater than 2% U3O8, therefore only the altered sandstone and metapelite basement material was tested. 7.1.2 Sample Preparation and Setup UCS tests were completed at the University of Alberta Civil Engineering cold room between June and July 2009. Assistance for setting up the laboratory testing procedures and use of the laboratory equipment were provided by Lukas Arenson (BGC Engineering) and Steve Gamble (University of Alberta, Cold Room Lab Manager). 7.1.3 Equipment Samples were trimmed with a knife to measure approximately 75 mm in diameter by 150 mm in length to maintain a length to diameter ratio of 2:1 and placed inside a rubber membrane inside the triaxial cell. The triaxial cell was then filled with mineral oil around the sample. The temperature of the mineral oil was controlled with glycol circulating in copper rings. Outside the triaxial cell are rings of copper with glycol circulating at half a degree lower than ambient temperature. The load cell sits underneath the triaxial cell with a maximum capacity of 5000 lb. A displacement transducer is attached to the top of the load conducting rod to measure axial displacement. A LC-5000 single syringe pump was used to apply the required load to the sample up to a maximum load of 20 MPa. Load and displacement data is recorded at user specific time intervals, typically 15s. Figure 7.1 to Figure 7.4 shows the setup and equipment for frozen UCS testing at the University of Alberta cold room.  126  Figure 7.1: Inside Cold Room, Triaxial Cell Setup. Left Triaxial Cell is a Sample Freezing Waiting to be Tested. Right Triaxial Cell is a Sample Undergoing Testing.  Figure 7.2: Triaxial Cell Filled with Mineral Oil, Sitting on Load Cell. Displacement LVDT Sensor Seen to Top Right of Cell. Load is Applied by the Top Load Conducting Rod 127  Figure 7.3:  Syringe Pump Controlling Loading Rate and Measuring Load  Figure 7.4: Glycol Transfer Unit Circulating Glycol in Copper Coils Outside of Triaxial Cell. Glycol Circulating at Half a Degree Celsius Below Ambient Room Temperature.  128  7.1.3.1 Temperature and Strain Rate for UCS Testing The target design freeze temperature prior to mining the Cigar Lake orebody is -12oC (personnel communication with Cigar Lake mine). Previous UCS testing was undertaken by Golder (1986) and EBA (1990) of the clay cap and orebody material. The historical UCS testing was conducted at temperatures of -2, -5 and -20oC. Results of the previous data are summarized with the current data in Figure 7.16. EBA (1990) suggested additional frozen UCS testing be completed of the soft and intermediate clays at -5, -10, and -20oC to establish the relationship between frozen strength with temperature. Two sets of UCS testing at -10oC and -20oC were completed at three strain rates (varying from 0.01%/min to 0.1%/min) on the three main rock types drilled: hematized sandstone/clay (more altered), bleached sandstone (less altered), and altered metapelite basement. Samples were loaded to failure or approximately 10% axial strain if the load remained constant during the test. Samples were also tested at strain rates varying from 0.01%/min to 0.1%/min to understand the effect of applied strain rate on the frozen material. Strain rates above 1%/minute will induce brittle behaviour resulting in higher strength data than that expected in the field. Strain rates below 0.01%/minute can possibly exhibit creep behaviour due to the long loading time on the sample (several days).  7.1.3.2 Freezing Samples Prior to Testing Samples were frozen for a minimum of 24 hrs inside the triaxial cell of the cold room, simulating all around freezing as is expected to occur at the Cigar Lake mine. As the samples are high moisture content (20-35% by weight), freezing from all around was considered to be a potential problem as cracks could be created in the center of the sample due to the 9% volume expansion during freezing; however, frozen sample cross-sections were examined and noted to be uniform with no ice expansion cracks. The basis for freezing the samples a minimum of 24 hrs was selected based on previous experience by the University of Alberta staff, and measuring the time for a UCS sample to reach -10oC to -20oC from ambient to be on average 12 hrs.  129  Figure 7.5: Cross Section of Frozen High Moisture Content Hematized Sandstone Showing Little to No Ice Lensing Present after 24 hours Freezing at -10oC  7.1.4 Discussion of Results The results of the UCS testing for the bleached sandstone, hematized sandstone, and the altered basement are presented in Table 7.1, Table 7.2 and, Table 7.3 respectively. The moisture content of each sample was averaged from sample trimmings prior to testing (unfrozen) and after sample testing (frozen). Young's modulus was measured at 50% of the UCS based on manually measured vertical displacements. Specific gravities of select samples from the UCS testing were measured using the gas pyncometer method according to ASTM 5550. Weaker rock samples (unfrozen strength less than 2 MPa) with low moisture content failed on obvious shear plans, such as bedding or pre-existing joints. Samples tested with unfrozen moisture contents greater than 30% did not fail on pre-existing shear planes but rather on the friction plane. Samples were loaded to failure, or approximately ten percent total strain. After the sample is loaded past 15-20% strain the results are not considered reliable due to the breakdown and cracking of the ice bonding. Three strain rates (0.1%/min, 0.06%/min, and 0.03%/min) were 130  applied to each rock type set (graphitic metapelite basement, bleached sandstone, and hematized sandstone/clay) for temperatures at -10oC and -20oC. The strain rate was controlled by the rate of the applied load; however, the measurements were collected manually with an LVDT (Load Value Displacement Transducer) attached to a screwdriver on the top of the loading plate. There are inconsistencies and missing data with the measured strain rate over time using the screwdriver with LVDT. The jumps or missed data are averaged over these portions. When a frozen specimen is subjected to a load it will respond in instantaneous deformation and a time-dependent deformation. Creep of a jet bored cavity is a concern as the stand-up time and time-deformation properties of this material is not fully defined. The conditions under which creep would be expected were not present during the UCS testing. Graphs of the UCS testing for each rock type and testing method are presented in the following sections. Individual data files for each test completed are included in Appendix B. Note the rock strength index term of R0.5 is applied in this research to define the unfrozen rock strength of the UCS samples. This term applies to rock that did not fit either the ISRM R0 (indented by thumbnail) or R1 (crumbles under firm blows with point of geological hammer) term, as the matrix of these very weak rock masses was still present many samples could not be indented by a thumbnail but be sliced with a knife with ease. 7.1.4.1 Bleached Sandstone UCS Results Table 7.1 and Figure 7.6 present the UCS testing data and UCS strain plots for the bleached sandstone. The Manitou Falls formation sandstone overlying the unconformity at Cigar Lake transitions from competent, slightly weathered sandstone to highly altered, friable, sand and clay within proximity of the orebody. The bleaching of the sandstone occurred with hydrothermal alteration and degraded the rock mass quality. The bleached sandstone samples were collected in the 15 to 20 m above the orebody, though the bleached sandstone occurs for tens of meters above the orebody. The bleached sandstone rockmass varies considerably from slightly to moderately weathered sandstone to soft clay.  131  Table 7.1: Sample ID  6 7 8  9  16 17 18  ID ST78607 ST78607 ST78607  SF801-04  ST78607 ST78607 ST78607  Summary of Frozen UCS Testing on Bleached Sandstone  Depth (m)  Unfrozen Strength (MP) (1)  Test Temp (oC)  Strain Rate (%/min)  Avg. Moisture Content (by Wt)  427.55  0.5  -10  0.14  427.73  2  -10  424.9  3  -20  S.G.  Bulk Density (g/cm3)  Porosity  UCS (MPa)  E (MPa)  35.6  2.71  1.36  0.50  2.12  922  0.01  38.1  2.68  1.34  0.50  1.57  1158  0.11  34.2  2.70  1.48  0.45  1.35  2346  5946  428.76  20  -10  0.47  10.0  2.70  2.19  0.19  Did not fail (>20 MPa)  426.9  3  -20  0.10  33.2  2.71  1.58  0.42  4.48  1325  427.1  3  -20  0.06  30.0  2.71  1.54  0.43  5.03  1872  427.3  0.5  -20  0.01  43.0  2.68  1.31  0.51  3.67  3322  22  SF801-04  432.35  2  -10  0.5  30.7  2.64  1.50  0.43  2.25  1195  23  SF801-04  432.55  2  -10  0.04  30.9  2.70  1.54  0.43  2.38  968  Note: 1.  The unfrozen strength was assessed with a pocket knife  132  Unfrozen Strength R0 ~ 0.25 to 0.5 MPa R0.5 ~ 0.5 to 1 MPa R1 ~ 1 to 5 MPa  T=-20oC  T=-10oC  Figure 7.6:  Frozen UCS vs. Total Strain of Bleached Sandstone Samples  Note: (1). R0 and R0.5 refer to the field strength (R0 to R6) assessed while trimming the samples  133  7.1.4.2 Hematized Sandstone UCS Results Table 7.2 and Figure 7.7 present the UCS testing data and UCS strain plots for the hematized sandstone/clay. This material directly overlies the orebody in the majority of the 2009 surface freeze drilling boreholes and is typically 2 m thick (ranging from 0.5 to 5 m). The alteration processes of the orebody have created a hematite rich dark red, dense clay to highly altered sandstone. The sandstone fabric and jointing are still present in this material though the strength of this sandstone borders on soil like, easily indented with a thumb or sliced with a knife.  Table 7.2: Sample ID  1 3 4 5 19 20 24  ID ST79106 SF80104 SF80104 SF80104 SF80104 SF80104 SF80104  Summary of Frozen UCS Testing on Hematized Sandstone/Clay  Depth (m)  Unfrozen Strength (MP) (1)  Test Temp. (oC)  Strain Rate (%/min)  Avg. Moisture Content (by Wt)  432.25  2  -10  0.06  435.15  0.5  -10  435.25  0.5  435.5  S.G  Bulk Density (g/cm3)  Porosity  UCS (MPa)  E (MPa)  23.2  2.81  1.94  0.31  4.81  1352  0.15  20.6  2.85  1.91  0.33  2.08  3540  -10  0.01  20.7  3.01  1.93  0.36  1.33  1198  2  -10  0.05  15.9  3.09  2.14  0.31  6.54  2685  434.7  0.5  -20  0.15  22.8  3.01  1.83  0.39  3.39  2055  435  2  -20  0.03  20.9  3.01  1.87  0.38  4.16  1830  432.75  2  -20  0.14  28.2  2.70  1.63  0.40  5.71  1845  Note: 1. The unfrozen strength was assessed with a pocket knife  134  Unfrozen Strength R0 ~ 0.25 to 0.5 MPa R0.5 ~ 0.5 to 1 MPa R1 ~ 1 to 5 MPa  T=-20oC  T=-10oC  Figure 7.7:  Frozen UCS vs. Total Strain of Hematized Sandstone/Clay  Note: (1). R0 and R0.5 refer to the field strength (R0 to R6) assessed while trimming the samples  135  7.1.4.3 Graphitic Metapelite Basement UCS Results Table 7.3 and Figure 7.8 present the UCS testing data and UCS strain plots for the altered graphitic metapelite basement. The basement rock present below the orebody (starting ~ 440 m level) is highly altered due to the formation of the orebody, though alteration in the basement does not correspond spatially with alteration of the overlying sandstone. Within the first few meters of the orebody, the basement rock comprises soft clay in a pebbly matrix to slightly weathered, medium strong metapelite. The rock from the 2009 surface freezing drilling core samples was highly fractured leaving a limited number of samples that were competent for testing. From approximately 10 m away from the orebody, the basement rock samples were too strong for frozen UCS testing at the University of Alberta cold room given the 20 MPa load limit of the testing apparatus.  Table 7.3: Sample ID  11 12 13 26 27 28  ID SF80104 SF80104 SF80104 SF80104 SF80104 SF80104  Summary of Frozen UCS Testing on Graphitic Metapelite Basement  Depth (m)  Unfrozen Strength (MP) (1)  Test Temp. (oC)  Strain Rate (%/min)  Avg. Moisture Content (by Wt)  441.28  3  -10  0.13  441.47  3  -10  441.9  10  442.85  S.G  Bulk Density (g/cm3)  Porosity  UCS (MPa)  E (MPa)  22.0  2.67  1.69  0.37  2.80  240  0.04  26.1  2.67  1.65  0.38  3.38  433  -10  0.56  15.8  2.64  1.81  0.31  7.96  5346  2  -20  0.15  25.0  2.64  1.69  0.36  6.60  3217  443.05  3  -20  0.05  25.0  2.60  1.61  0.38  3.10  3862  443.2  3  -20  0.02  25.0  2.60  1.61  0.38  4.07  1332  Note: 1. The unfrozen strength was assessed with a pocket knife  136  T=-20oC  T=-10oC  Figure 7.8:  Frozen UCS vs. Total Strain of Graphitic Metapelite Basement  7.1.4.4 UCS vs. Unfrozen Rock Strength Classification Rock strength is based on some general field tests which can be related to a range of UCS values. The strength of the pieces can be estimated using a pocket knife or rock hammer. The samples for the frozen UCS testing were initially assessed based on their unfrozen rock strengths determined through field index testing to estimate the gain in strength due to freezing.  137  Table 7.4:  Grade R0  Description Extremely weak rock  R1  Very weak rock  R2  Weak rock  R3  Medium strong rock  R4  Strong rock  R5  Very strong rock  R6  Extremely strong rock  ISRM Field Strength Estimates, after Brown (1981)  Field Identification Indented by thumbnail. Crumbles under firm blows with point of geological hammer, can be peeled by a pocket knife. Can be peeled by a pocket knife with difficulty, shallow indentations made by firm blow with point of geological hammer. Cannot be scraped or peeled with a pocket knife, specimen can be fractured with single firm blow of geological hammer. Specimen requires more than one blow of geological hammer to fracture it. Specimen requires many blows of geological hammer to fracture it. Specimen can only be chipped with geological hammer.  Approx. Range of Uniaxial Compressive Strength MPa 0.25 – 1.0 (>2.5 on Pocket Penetrometer) 1.0 - 5.0 (Pocket Penetrometer does not indent) 5.0 – 25  25 – 50 50 – 100 100 - 250 >250  Figure 7.9 and Figure 7.10 plot the unfrozen ISRM strength vs. frozen UCS value of all samples and of samples that failed in shear (not on pre-existing joints or bedding), respectively. The weakest rock samples (R0 and R1) are expected to have the greatest gain in strength due to freezing. However, given the high variability of the samples tested, no trend between the unfrozen and frozen strengths can be established from this data set.  138  Unfrozen Strength R0 ~ 0.25 to 0.5 MPa R0.5 ~ 0.5 to 1 MPa R1 ~ 1 to 5 MPa  Figure 7.9:  Frozen UCS vs. Unfrozen ISRM Rock Strength, All Data  139  Unfrozen Strength R0 ~ 0.25 to 0.5 MPa R0.5 ~ 0.5 to 1 MPa R1 ~ 1 to 5 MPa  Figure 7.10: Frozen UCS vs. Unfrozen ISRM Rock Strength, Good Data, Samples That Failed Through Joints or Bedding Removed 7.1.4.5 UCS vs. Strain Rate The applied strain rate directly influences the failure load of a UCS sample. Frozen material will be strongest under an instantaneous load compared to an applied failure rate taking minutes, hours or days. Frozen material exhibits creep behaviour with the rate typically related to the available pore water converting to ice. As discussed in the literature review, at low rates of deformation the frozen material is ductile and cracks do not form. At higher rates of loading, the material forms microcracks and the failure is brittle. Plotting the UCS of each specimens applied strain rate on a log scale should ideally show a linear trend. Figure 7.11 and Figure 7.12 plots the 2009 results for freezing temperatures of -10 and -20oC, 140  respectively. Figure 7.13 plots the failure mechanism of the UCS samples, combining all rock types and frozen test temperature. No linear trend between the applied strain rate and UCS is evident, which is attributed to the varying degrees of alteration of the same rock type, the samples failing in different manners, and the limited data set. A slight increase in the UCS was noticed with increasing applied strain rate, though no correlation in the applied strain rate with the UCS or mode of UCS failure could be established due to the small data set and highly variable nature of the samples. No apparent trend on the types of failures in the frozen UCS samples could be established by rock type.  Unfrozen Strength R0 ~ 0.25 to 0.5 MPa R0.5 ~ 0.5 to 1 MPa R1 ~ 1 to 5 MPa  Figure 7.11: Plot of All Samples, Frozen UCS vs. Applied Strain Rate, T=-10oC Note: (1). R0 and R0.5 refer to the field strength (R0 to R6) assessed while trimming the samples  141  Unfrozen Strength R0 ~ 0.25 to 0.5 MPa R0.5 ~ 0.5 to 1 MPa R1 ~ 1 to 5 MPa  Figure 7.12: Plot of All Samples, Frozen UCS vs. Applied Strain Rate, T=-20oC Note: (1). R0 and R0.5 refer to the field strength (R0 to R6) assessed while trimming the samples  142  Figure 7.13: Frozen UCS vs. Strain Rate of All 2009 Samples, by Failure Mode  7.1.4.6 UCS vs. Temperature The gain in strength due to freezing is a function of temperature, with higher strengths typically achieved under decreasing temperatures. The influence of temperature on strength, discussed in the literature review, is a function of the unfrozen water content, where at temperatures just below freezing there is water that has not converted to ice in the pores therefore the strength is lower than at colder temperatures, but will decrease with decreasing temperatures. The conversion of water to ice is a function of material type, porosity, salinity and confining pressures. The target design freezewall temperature of the Cigar Lake orebody prior to jet boring is -12oC (personal communication with Cigar Lake mine staff). Historical testing by EBA (1990) and Golder (1986) was completed at temperatures of -5oC and -20oC. Given the limited number of 143  samples available from the 2009 surface freeze drilling program, test temperatures of -10 and 20oC were used to compare with historical testing data. Based on the UCS testing results (Figure 7.6, Figure 7.7, and Figure 7.8), a gain in strength of approximately 5 MPa is evident from -10oC to -20oC in all rock types. The extremely weak to very weak (R0 to R1) rocks are expected to have the largest strength gain with freezing due to the higher moisture content in very weak rock samples. Medium strong rocks (R3, 50 MPa) and greater are not expected to show significant gain in strength with freezing due to the reduced moisture content and lack of available pore water to convert to ice. The strength of ice, though a function of strain rate and temperature, is typically on the order of 20 to 35 MPa. Very weak rocks, with compressive strengths of 1 to 5MPa, will almost double their strength due to the conversion of water to ice. Beyond unfrozen rock strengths of 40 MPa (R3), the upper bound strength of ice, little to no strength gain is expected with freezing. Based on the testing completed at temperatures of -10oC, Figure 7.14 below establishes the relationship between the estimated unfrozen rock strength and measured frozen strength. Note that no samples greater than 25 MPa were tested in the 2009 laboratory testing program.  144  Figure 7.14: Influence of Freezing and Strength Gain for Weak Cigar Lake Rock From the 2009 testing (Figure 7.15), UCS samples tested at temperatures of -10oC and -20oC exhibited brittle, elastic perfectly plastic and strain softening behaviour. Samples tested at -10oC failed between 1 and 8 MPa with the samples failing in strain softening behaviour comprising the weakest material tested (unfrozen strength of R0 to R0.5, equivalent to 0.5 to 1 MPa). The samples failing in a brittle manner comprise the strongest material tested (unfrozen strength R1 to R2, equivalent to 1 to 5 MPa). Samples tested at -20oC failed between 1 and 7 MPa; however, the majority of the specimens failed elastic perfectly plastic with only a couple exhibiting brittle or strain softening behaviour. The change in UCS failure mode with a decrease in temperature is attributed to polycrystalline ice behaving brittle with colder temperatures, though this was not evident in the 2009 lab testing. The majority of the samples tested at -20oC failed elastic perfectly plastic compared to the 145  samples tested at -10oC that failed as strain softening. This is attributed more to the samples tested at -10oC having a lower unfrozen strength than the samples tested at -20oC, and may not be due to a change in failure mechanism with temperature.  Unfrozen Strength R0 ~ 0.25 to 0.5 MPa R0.5 ~ 0.5 to 1 MPa R1 ~ 1 to 5 MPa  Figure 7.15: Influence of Temperature on Frozen UCS, 2009 Data, by Failure Mode Note: (1). R0 and R0.5 refer to the field strength (R0 to R6) assessed while trimming the samples  Figure 7.16 summarizes the effect of temperature on the UCS strength with the historical UCS testing of Cigar Lake material along with the 2009 samples from the surface freeze drill program. The upper and lower bound lines are drawn based on visual assessment of the data. 146  Freezewall Design Temperature o T=-12 C UCS ~ 3 MPa  Unfrozen UCS o T=0 C UCS ~ 0.5 to 2 MPa  Figure 7.16: Influence of Temperature on Frozen UCS, All Data, by Rock Type  7.1.4.7 UCS vs. Bulk Density The bulk density was calculated by measuring the samples moisture content prior to freezing the samples for testing. Bulk densities for the hematized clay, altered basement, and bleached sandstone are summarized in Table 7.5 below. The influence of bulk density on the unfrozen strength is not documented and was not evident in any trends of the frozen strength. The frozen bulk density is expected to be slightly lower than unfrozen based on the measured frozen moisture contents and sample weight, though was not recorded for each UCS sample.  147  Table 7.5:  Summary of Unfrozen Bulk Densities Bulk Density (g/m3) Max  Min  Average  No. Samples  Bleached Sandstone  2.19  1.31  1.54  9  Hematized Sandstone/Clay  2.14  1.63  1.89  7  Graphitic Metapelite Basement  1.81  1.61  1.68  6  Material Type  Based on Figure 7.17, no trend between the bulk density of the sample and the UCS can be established. The hematized clay/sandstone has the highest bulk density due to the iron rich alteration of the sandstone.  148  Unfrozen Strength R0 ~ 0.25 to 0.5 MPa R0.5 ~ 0.5 to 1 MPa R1 ~ 1 to 5 MPa  Figure 7.17: Frozen UCS vs. Unfrozen Bulk Density Note: (1). R0 and R0.5 refer to the field strength (R0 to R6) assessed while trimming the samples  7.1.4.8 UCS vs. Porosity The porosity of the samples was determined by measuring the specific gravity with the pyncometer and bulk density from the samples moisture content. The porosity for the hematized clay, altered basement, and bleached sandstone are summarized in the table below. The relationship between porosity and unfrozen UCS is that the UCS generally increases with decreasing porosity. The lower the porosity the higher the specimen’s strength due to the dense packing of particles filling the void spaces and increasing the volume change under an applied load. Porosities higher than 0.2 are generally classified as weak rock, as comparable in the material tested. From all the material tested, a significant decrease in the frozen UCS strength 149  from an average of 5 MPa at a porosity of 0.30 to an average of 2 MPa at a porosity of 0.50. With decreasing porosity, there is a general increase in the frozen strength data. Bleached sandstone has a higher porosity (ranging from .42 to 0.52) compared to the rest of the material types tested. Figure 7.18 and Figure 7.19 plot the measured porosity to the frozen UCS by rock type and failure mode.  Unfrozen Strength R0 ~ 0.25 to 0.5 MPa R0.5 ~ 0.5 to 1 MPa R1 ~ 1 to 5 MPa  Figure 7.18: Frozen UCS vs. Porosity, by Material Type  150  Figure 7.19: Frozen UCS vs. Porosity, by Failure Mode  7.1.4.9 UCS vs. Moisture Content The moisture content of the UCS specimens was measured from trimmings collected during the preparing of the samples for UCS testing prior to freezing in the cold room. A higher moisture content relates to a higher porosity and a lower unfrozen UCS. The general trend noted in the unfrozen moisture content, is that the samples with the lowest moisture content have a higher UCS.  151  Unfrozen Strength R0 ~ 0.25 to 0.5 MPa R0.5 ~ 0.5 to 1 MPa R1 ~ 1 to 5 MPa  Figure 7.20: Frozen UCS vs. Moisture Content, 2009 Data  7.1.5 Results In summary, the following observations were noted from the frozen UCS testing of Cigar Lake material: •  Samples frozen to T = -10oC failed at an average UCS of 2MPa and total strain of 2-3%. The unfrozen strength of these samples ranged from 0.5 to 1 MPa. Overall, the approximate strength gain was 2 MPa.  •  Samples frozen to T = -20oC failed at an average UCS of 5 MPa and a total strain of 46%. The unfrozen strength of these samples ranged from 0.5 to 1 MPa. Overall, the approximate strength gain was 4 MPa.  •  Samples tested at T = -10oC typically exhibit strain-softening behaviour compared to 152  those frozen at T = -20oC where they exhibit elastic/plastic behaviour. •  The UCS of the material tested (altered sandstone and basement) did not appear to be strain rate dependent.  •  No creep was observed under the testing regime.  •  Some samples were observed to fail in a brittle manner even though the strain plots do not really support the failure mode.  •  The strength increase in the frozen UCS was several MPa comparing the historical samples tested by Golder (2002) and EBA (1996) at -5oC to the current samples at -10 to -20oC.  Limitations of laboratory testing with the provided setup that were not resolved; • •  •  7.2  Cannot freeze samples at 4.5 MPa confining pressure Cannot freeze samples at the same freezing rate in the lab as expected in the field. Samples were generally frozen in 24 hrs in the cold room compared with 6 months to a year that is expected for a full freeze front to form around the Cigar Lake orebody Concern of the representativeness of the samples frozen in the laboratory given the potential for ice lenses to develop  Four-Point Beam Testing  The freezing of a very weak rock is expected to add tensile strength to the rock mass due to the bonding of ice in joints; however, no data exists to support this gain of tensile strength. The failure of a frozen weak rockmass is proposed to be investigated through four-point beam testing. Four-point loading allows for a simple and repeatable flexural test. The purpose of the beam testing of frozen specimens is to understand how frozen weak rock fails where a frozen joint is present or as a frozen weak rock mass. Four-point beam testing was undertaken on a suite of premixed concrete and altered sandstone drill core from the 2009 surface freeze drill sampling program. The four-point beam apparatus is shown in Figure 7.21.  153  Figure 7.21: Four-Point Beam Test Apparatus  Prior to testing the core from Cigar Lake, frozen sand/cement mixtures were tested to refine the freezing and testing procedure be used as well as to test the behaviour of a frozen joint using a controlled material for the matrix. Measurement of the load at first crack, peak load and deformations on the core midspan and ends were recorded. The traditional approach to understand stability in stratified ground is to model the immediate roof as if it were a beam. Beam theory assumes that the immediate roof can be represented by a series of equal width beams, with a length equal to the room span. A beam is capable of carrying loads in bending and applies loads transverse to its longest dimension. Three point and four point 154  flexural testing is typically used in the laboratory to measure the modulus of elasticity in the bending moments of concrete, wood, steel or other materials. Bending tests are simple and quick to complete, but are influenced by the applied strain rate and specimen geometry. The beam will fail at its midpoint, developing a tensile crack as the beam fails under tensile stresses that develop from its underside (relative to its flexure), before the compressive stresses that develop on its top side approach the compressive strength. The flexural strength is equivalent to the tensile strength assuming the beam is homogeneous without defects or flaws. Beam theory relates flexure resulting from applied forces without considerations of shearing forces. Assumptions of simple beam theory include: the beam is symmetrical across its axis, and there is a fixed relationship between stress and strain as a beam behaves the same in tension as in compression. Flexural strength is determined by loading a beam with a span length at least three times the depth. The flexural strength is expressed as a modulus of rupture in psi or MPa. The modulus of rupture for four-point loading of cylindrical rock specimen with loads applied at L/3 from each end and reactions at the ends is defined as TMR = 16PmaxL / 3πd3 (Goodman, 1989). Where : Pmax = maximum load L = length between load reactions on the lower surface d = core diameter  In an unfrozen state the degree of jointing and infilling material in a rock mass will control the failure. No research or data was located by the author on how a frozen jointed weak rock mass fails. Failing a rock specimen in tension, produces a crack at the midpoint of the beam. If the frozen joint is weaker than the rock mass ideally the beam will fail along the joint. If the frozen joint is stronger than the rock mass the beam will fail as a solid beam through the midpoint of the beam. The increased cohesion of a joint undergoing freezing will be influenced by the type and thickness of infilling and the degree of moisture on the joint surface. A smooth and planar joint with no infilling and no moisture will not have sufficient cohesion to bond the joint surfaces together. 155  The following sections outline the samples and method of preparation for the four point beam testing. The tensile strength of each beam is based on the modulus of rupture. The modulus of rupture is calculate for each beam at the peak force at failure and center point deflection recorded using a Linear Variable Displacement Transducer (LVDT) on the same axis as the two outer roller pins.  7.2.1 Sample Preparation The following sample preparation and testing procedures were developed for the frozen fourpoint beam testing: Sample Size • The core diameter of the 2009 surface freeze holes is approximately 83 mm (3.25”) • According to Goodman (1988), for 3” diameter core samples, the test span length should be 9” and the beam length prepared to 12”. Freezing • Both the concrete beam samples and the Cigar Lake drill core samples were placed inside a large freezer in the University of British Columbia Rock Mechanics lab. • The freezer temperature was set to a temperature of -12oC (the design freeze temperature of the jet bored cavities); however, the temperature inside the freezer fluctuated considerably. • The samples were stored inside a Styrofoam container to minimize the influence of the freezer door opening during the samples’ 24 hr period inside the freezer. • Both the cement mixture and drill core samples were rotated once during their freezing period to eliminate the effect of a freezing front, where the sample will freeze faster from the side closest to the freezer walls. • Metal clamps were placed around the PVC containing the cement mixture and drill core samples while in the freezer to control the 9% phase change expansion of water to ice.  Applied Loading Rate • The influence of loading rate on a frozen beam is important however the current setup involves applying the load manually using a hand pump. •  Frozen UCS testing is to be undertaken at low strain rates (0.01%-0.1%/min). The current 156  loading rate of the four-point beam testing by hand will apply a strain rate that is significantly greater and should be addressed with future testing. Temperature •  The test temperature was it ambient room temperature with the option to surround the test apparatus with insulation if necessary. However, the samples were tested and failed within 60 seconds. The option to include insulation around the beam test apparatus was not pursued.  7.2.2 Frozen Beam Testing Cement Mixture Samples Four point beam testing was completed on cement mixture having strengths similar to the altered sandstone overlying the orebody. Results of the four-beam testing on cement mixture samples are included in Appendix C-1. The samples were prepared as solid cores that contained a single smooth, planar joint with no infilling in the center of the beam. Beam testing of cement mixture samples prior to testing Cigar Lake material helped to establish the correcting testing methodology with a number of control samples. The prepared cement mixtures followed testing method ASTM C 78 which determines the flexural strength of concrete using a simple beam with 3-point loading where half the load is applied at each third of the span length and the maximum stress is present over the center 1/3 of the beam, or ASTM C 298-08 where the entire load is applied at the center span and the maximum stress is only present at the center part of the beam. Batches of cement and sand mixtures were prepared at various proportions, moisture content, and joint condition. Forty cement mixture samples were prepared by mixing Portland cement, sand and water in a 5 gallon bucket and pouring into a 12” x 3” cylindrical PVC mold. Four types of mixtures were prepared each with different moisture contents and without or with the presence of a joint through the axial center plane of the cement mixture sample:  157  • • • •  100 % cement mixed with aggregate; 50:50 sand:cement; 33:66 sand:cement; and 40:60 sand:cement.  The beams were allowed to cure for 3 hours prior to placing in the freezer for a period of 24 hours. The cement mixture samples are listed below in Table 7.6. Table 7.6: Test No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27  Sample No. 1 2 3 4 1 2 3 4 1 2 1 2 3 1 2 3 4 1 2 3 4 1 2 3 1 2 3  Summary of Cement Mixture Samples for Four-Point Beam Testing Batch No. 1 1 1 1 2 2 2 2 3 3 4 4 4 5 5 5 5 6 6 6 6 7 7 7 8 8 8  Mix Design 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement Cement w/ Aggregate Cement w/ Aggregate Cement w/ Aggregate Cement w/ Aggregate Cement w/ Aggregate Cement w/ Aggregate Cement w/ Aggregate Cement w/ Aggregate Cement w/ Aggregate Cement w/ Aggregate  Joint (y/n) No No No No Yes No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No No No Yes Yes Yes Yes no  Freezing Temp. (oC) -12 -12 -12 -12 -13 -13 -13 -13 -11 -11 -11 -11 -11 -12 -12 -12 -12 -12 -12 +20 +20 -12 -12 -12 -12 -12 -12  Moisture Content (%) 14.7 14.7 14.7 14.7 17.5 17.5 17.5 17.5 18.7 18.7 18.7 18.7 18.7 12.1 12.1 12.1 12.1 12.1 12.1 12.1 12.1 10.8 10.8 10.8 13.9 13.9 13.9  Peak Pressure (kPa) 2650 2440 3350 3950 560 1830 2800 990 490 n/a(1) n/a(1) n/a(1) n/a(1) 1730 520 2320 680 880 n/a(1) n/a(1) n/a(1) 1900 800 800 n/a(1) n/a(1) 1100  Tensile Strength (MPa) (2) 1.79 1.51 2.27 2.48 0.37 1.19 1.86 0.66 0.31 n/a(1) n/a(1) n/a(1) n/a(1) 1.11 0.33 1.49 0.44 0.56 n/a(1) n/a(1) n/a(1) 1.21 0.49 0.51 n/a(1) n/a(1) 0.68  158  Test Sample Batch No. No. No. 28 4 8 29 1 9 30 2 9 31 3 9 32 4 9 33 1 10 34 2 10 35 3 10 36 4 10 37 1 11 38 2 11 39 3 11 40 4 11 Note: 1. Sample failed on handling  Mix Design Cement w/ Aggregate 33/66 Sand/Cement 33/66 Sand/Cement 33/66 Sand/Cement 33/66 Sand/Cement 40/60 Sand/Cement 40/60 Sand/Cement 40/60 Sand/Cement 40/60 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement 50/50 Sand/Cement  Joint (y/n) Yes Yes Yes Yes No Yes no yes Yes Yes No Yes Yes  Freezing Temp. (oC) -12 -11 -11 -11 -11 -12 -12 -12 -12 -5 -5 -5 -5  Moisture Content (%) 13.9 28.8 28.8 28.8 28.8 18.5 18.5 18.5 18.5 16.7 16.7 16.7 16.7  Peak Pressure (kPa) n/a(1) 700 900 n/a(1) 1000 800 1300 1300 1500 1900 1600 1100 800  Tensile Strength (MPa) (2) n/a(1) 0.48 0.63 n/a(1) 0.63 0.58 0.78 0.91 0.91 1.34 1.08 0.71 0.54  2. Based on an assumed Young’s Modulus of 0.5 GPa  7.2.3 Frozen Beam Testing Cigar Lake Drill Core Samples Frozen four point beam testing was completed on seven samples of altered sandstone overlying the Cigar Lake orebody. Samples without and with a single joint through the core sample were selected. Results of the four-beam testing on drill core samples are included in Appendix C-2. The samples are listed below in Table 7.7. Table 7.7:  Summary of Drill Core Samples for Frozen Four-Point Beam Testing  Test No. Sample No. Hole ID Depth (m) 1 1 SF791-06 429.5 2 2 SF801-04 431.2 3 2 SF801-04 431.2 4 3 SF801-04 433.5 5 3 SF801-04 433.5 6 4 SF801-04 431.4 7 5 SF796-05 432.05 Note: 1. Sample failed on handling 2. Based on an assumed Young’s Modulus of 0.5 GPa  Unfrozen Sample Strength R 0.5 R2 R1 R2 R 1.5 R 0.5 R1  Joint (y/n) No Yes No No Yes Yes Yes  Moisture Content (%) 34.0 11.9 11.9 28.7 28.7 35.5 17.9  Peak Pressure (kPa) 680 970 1090 1690 n/a(1) 760 n/a(1)  Tensile Strength (MPa) (2) 0.30 0.43 0.49 0.76 n/a(1) 0.34 n/a(1)  159  7.2.4 Results Based on the frozen four-point beam testing the following can be concluded on the strength of frozen joints with trace to little infill and tight aperture: • • • • • •  For unfrozen rock strengths less than 2 MPa (based on field strength assessments), the frozen joint is as strong as the frozen rock mass. For unfrozen rock strengths greater than 2 MPa (based on field strength assessments), the joint was observed to be weaker than the frozen rock mass For cement mixture and rock drill core samples greater than 30% moisture, a frozen joint is as strong as the frozen rock mass. For the cement mixture beam testing, with increasing sand content, an increase in tensile strength was observed Failures along joints with varying moisture or unfrozen strengths were not repeatable in the laboratory. The frozen tensile strength of the cement mixture samples (~0.5 to 2 MPa) is slightly higher comparing to similar unfrozen materials such as paste backfill (~0.2 MPa, Hughes (2008).  Figure 7.22, Figure 7.23, and Figure 7.24 plot the tensile strength of the beam calculated from the modulus of rupture versus the moisture content for the cement beams and rock drill core.  160  Figure 7.22: Frozen Tensile Strength vs. Moisture Content, Cement Samples by Mixture  161  Figure 7.23: Frozen Tensile Strength vs. Moisture Content, Cement by Joint Presence  162  Figure 7.24: Frozen Tensile Strength vs. Moisture, Drill Core Samples by Joint Presence  163  7.3  Frozen Direct Shear Testing  Determining the shear strength of rock joints is significant to understanding rock mass behaviour. The freezing of a rockmass is believed to have significant influence on the shear strength behaviour, specifically the cohesion. Direct shear testing on natural joint surfaces and intact rock specimens was undertaken to assisting with developing a model of the gained shear strength of a frozen joint. Direct shear testing includes intact rock specimens to determine the breaking strength (intact cohesion) of the rock, those with recognizable planes of weakness to determine the shearing resistance along these planes, or jointed/fractured specimens to determine the lower bound residual strength. 7.3.1 Sample Preparation The following testing and sample preparation procedures were developed for the frozen direct shear testing: Sample Size • The core diameter of the 2009 surface freeze holes is approximately 83 mm (3.25”) Freezing • After preparing the Cigar Lake drill core in the direct shear mould, the samples were placed inside a large freezer in the University of British Columbia Rock Mechanics lab. • The freezer temperature was set to a temperature of -12oC (the design freeze temperature of the jet bored cavities); however, the temperature inside the freezer fluctuated considerably. • The samples were stored inside a Styrofoam container to minimize the influence of the freezer door opening during the samples 24 hr period inside the freezer. • Samples were rotated once during their freezing period to eliminate the effect of a freezing front, where the sample will freeze faster from the side closest to the freezer walls.  Temperature •  The test temperature was at ambient room temperature with the option to surround the test apparatus with insulation if necessary. However, the samples were tested and failed within 60 seconds. The option to include insulation around the beam test apparatus was not pursued. 164  7.3.2 Test Procedures Direct shear samples were trimmed and placed in a mould using sand and Portland cement and tested according to ASTM D 5607-95. Samples of the altered sandstone (hematized and bleached) overlying the orebody were selected for testing. The unfrozen strength of the samples was approximately 1 MPa based on field strength assessments. The moisture content of the samples ranged from 15 to 30%.  7.3.3 Results From the 5 samples selected, only one sample contained a natural joint; the other four were intact specimens used to obtain the breaking strength (by loading the sample to failure). Table 7.8 presents the summary of frozen direct shear testing completed on the Cigar Lake drill core. Detailed results of each test are included in Appendix D. Table 7.8:  Sample No. 1 2 3 4 5  Borehole SF79106 SF80104 SF80104 SF80104 SF79605  Summary of Frozen Direct Shear Testing Results on Drill Core  Depth (m) 429.5 431.2 433.5 431.4 432.05  Description Bleached sandstone Bleached sandstone Hematized Sandstone Bleached Sandstone Hematized sandstone  Test Type Breaking Strength Breaking Strength Joint Plane Breaking Strength Breaking Strength  Angle of Joint  Peak Failure Load (kPa)  Normal Force Applied (kg)  Shear Stress (kPa)  Normal Stress (kPa)  Moisture Content (%)  -  15,320  25  1.69  0.46  34.0  -  14,780  5  1.67  0.12  11.95  55o  6,990  5  0.75  0.12  28.74  -  14,950  45  1.77  0.85  35.46  -  14,160  25  1.56  0.46  17.93  Figure 7.25 plots the normal load applied and calculated shear stress of each test. The frozen cohesion (at T=-10oC) backs out to approximately 1.6 MPa, which is considered a little high due to the frictional component and uneven break plane of this test. However, no triaxial testing has been completed on frozen Cigar Lake material to compare this value to. This cohesion value, is within the expected range for frozen rock. Additional testing is recommenced, give the small data set and lack of testing on the influence of temperature to the cohesion. 165  Estimated Cohesion ~ 1.6 MPa Frozen at T=-10oC  Figure 7.25: Plot of Direct Shear Testing Results on Drill Core  166  8.  Influence of Freezing on a Weak Rock Mass  This section presents the interpretation of case history data of mines in permafrost or artificially frozen ground and the Cigar Lake mine laboratory testing to understand and predict the behaviour of openings in frozen rock masses. 8.1  Rock Mass Classification Schemes  The two most common rock mass classifications in North America are the previously discussed Q (Barton, 1974) and RMR systems (Bieniawski 1976 and 1989). Each of these classifications consist of geotechnical parameters that, when combined, yield a number to describe the rock mass quality. Each system is discussed in Section 2.5.1 (Rock Mass Classification Schemes), and a discussion of how ground freezing affects its individual parameters is provided below. 8.1.1 Intact Rock Strength Intact rock strength is the first parameter in the RMR system; it is not considered directly in the Q system. The RMR input is based on the UCS of the intact rock. Table 8.1 presents the 1976 Rock Mass Rating Classification intact rock strength parameter ratings for the six UCS ranges. Table 8.1:  RMR Classification for Intact Rock Strength (Bieniawski, 1976)  Parameter Strength of intact rock material UCS (MPa)  Range of Values  > 200  100-200  50-100  25-50  10-25  3-10  1-3  RMR Rating  15  12  7  4  2  1  0  ISRM, R  R6  R5  R4  R3  R2  R1  R0  The UCS of a rock is divided into six strength categories, and can be estimated through standard field identification and laboratory testing methods, as shown in Table 8.2 (Barton, 2002). The UCS can also be estimated through the use of point load testing.  167  Table 8.2:  Descriptions of Rock Strength and Approximate UCS (ISRM, 1981)  Grade  Description  Approximate Range of Uniaxial Compressive Strength (MPa)  R0  Extremely weak rock  0.25 - 1.0  R1  Very weak rock  1.0 - 5.0  R2  Weak rock  5.0 - 25  R3  Medium strong rock  25 - 50  R4  Strong rock  50 - 100  R5  Very strong rock  100 - 250  R6  Extremely strong rock  >250  The strength of intact rock is defined through the above ratings. It has been observed that freezing increases the strength of intact rock, and therefore the RMR value, particularly for extremely weak to weak rock (R0 to R2). Figure 7.16 plots the UCS value for all Cigar Lake samples at the range of temperatures tested. Note, samples at T = -2oC and T = -5oC are from historical testing at the Cigar Lake mine (EBA, 1990, and Golder, 1986). An average gain in strength of approximately 1 MPa is achieved by reducing temperatures from -5 to -10oC, and almost 2 MPa from -10 to -20oC in all rock types. An interesting correlation appears when the rocks are grouped based on their initial, unfrozen strengths. The extremely weak to very weak (R0 to R1) rocks have the largest strength gain with freezing due to the higher moisture content in very weak rock samples. Medium strong rocks (R3, 50 MPa) and greater are not expected to show significant gain in strength with freezing due to the reduced moisture content and lack of available pore water to convert to ice. The strength of ice, though a function of strain rate and temperature, is typically on the order of 20 to 35 MPa (Andersland and Ladanyi, 2004). Very weak rocks, with compressive strengths of 1 to 5 MPa, will almost double their strength due to the conversion of water to ice. In contrast, unfrozen rock strengths of approximately 40 MPa correspond to the upper bound strength of ice, and little to no strength gain is therefore expected with freezing. Figure 7.14 shows the relationship between unfrozen rock strength (shown from R0 to R4) and 168  ISRM UCS rock strength upper and lower bounds (Barton, 2002), and the UCS gained for the corresponding unfrozen rock strength when frozen (red line). All tests were completed at -10oC. No samples greater than 25 MPa were tested in the 2009 laboratory testing program. The increase gained in RMR values is thus highly dependent on the unfrozen strength of the rock. In R0 to R2 unfrozen rock, RMR may be increased by as much as 7 points when frozen. For example, an unfrozen rock within an R0 strength would have an RMR rating of 0, if the same rock is R4 when frozen, the RMR rating would become 7, a 7 point increase. For unfrozen rock strengths higher than 50 MPa (R3 or greater), the RMR is not affected with respect to the intact rock strength parameter. 8.1.1.1 RQD, Joint Spacing, and Number of Joint Sets When a rock mass undergoes freezing, geologically speaking, the discontinuities healed with ice in the frozen rock mass still exist. However, geotechnically speaking these healed discontinuities are no longer considered in the design and are not counted in the rock mass classification. Joints are typically assumed to have zero tensile strength. If the ice-healed discontinuities are strong enough to withstand gentle twisting by hand, they should no longer be considered a discontinuity with zero tensile strength in the design. This section discusses the effect of freezing on RQD, joint spacing, and joint set input parameters to both the RMR and Q rock mass classification system.  8.1.1.2 Rock Quality Designation (RQD) RQD was developed for geotechnically quantifying drill core soundness; however, it can be visually estimated in mapping excavation faces by relating it to the number of joints in a cubic meter (Palmstrom, 1982).  RQD = 115 − 3.3 * Jv Where: Jv = number of joints in one cubic meter. Rocks that are not strong enough to withstand gentle hand pressure are not considered intact rock. For example, a very weak rock that may appear to have no discontinuities (RQD=high) 169  should be assigned a RQD of zero (0) as all of the rock would break into pieces smaller than 10cm if gentle pressure were applied. RQD is the second parameter in the RMR classification system. The ranges and ratings used in the RMR 1976 system are shown in the following table. Table 8.3:  RMR Classification for RQD (Bieniawski, 1976)  .  Range of Values  Drill core quality (RQD)  90-100%  75-90%  50-75%  25-50%  <25%  Rating  20  17  13  8  3  In the Q system, RQD is the first index, entered from 0 (worst) to 100 (best). 8.1.1.3 Joint spacing Joint spacing is the third input parameter in the RMR system. It is the average spacing between discontinuities either in a core run, or in face mapping, as the average block size. Table 8.4 lists the ranges and ratings for spacing of joints. Table 8.4:  RMR Classification for Joint Spacing (Bieniawski, 1976)  Parameter  Range of Values  Spacing of Joints  >3m  1-3m  0.3-1m  50-300mm  <50mm  Rating  30  25  20  10  5  8.1.1.4 Number of Joint Sets Joint number (Jn) is the fourth input parameter in the Q-system. It is rated based on the number of joint sets in a geotechnical group of rock ranging from 0.5 (best) to 20 (worst). Table 8.5 lists how various joint set descriptions relate to the Jn number.  170  Table 8.5:  Jn Number for the Q Rock Mass Classification (Barton et al., 1974) Number of Joint sets Intact rock (no joints) 1 set 1 set + random 2 set 2 set + random 3 set 3 set + random 4 set 4 set + random Earthlike, crushed rock  Jn rating 0.5 1 2 3 4 6 9 21 15 20  8.1.1.5 Effect of Freezing on RQD, Joint Spacing, and Joint Number Freezing can have a significant impact on increasing the frozen RMR and Q vales by simply reducing the number of discontinuities through healing the discontinuities with ice. There is more of an impact on weak and/or highly fractured rock, as these units are more heavily jointed and thus greater opportunity for healing of joints through freezing. A rock with an RQD of zero (0) could improve up to one hundred (100) through freezing, by making very weak rock sound and intact by healing of all the joints. In the RMR system this would result in an increase from as low as three (3) to as high as twenty (20). Similarly, a rock with joint spacing less than fifty millimeters (<50mm) could have a spacing of >3m once frozen, resulting in an increase in RMR from five (5) to thirty (30). The Jn in the Q system could be improved from twenty (20) to point five (0.5), assuming the entire rock mass remains frozen. While handling the frozen cement mixture beams, for the four-point beam testing, it was noted that the beams with a frozen joint could not be twisted or broken along that joint with mild hand pressure. The aperture of the joints was tight (<1 mm), and the joint surface was smooth with no infilling. The moisture within the sample during freezing is attributed to healing of the joints and thus it is clear that significant gains in rock mass quality can be made in the reduction of open joints through freezing.  171  8.1.2 Joint Condition Ratings Discontinuities are commonly described by their roughness, planarity, aperture, and infill material. Each of these parameters controls the friction angle of a discontinuity, and in the case of infill, the cohesion. In the rock mass characterization of a core run or tunnel face, the critical discontinuity or discontinuity set (i.e. with lowest friction and cohesion) is described, for a geotechnical zone. Both the RMR and Q system incorporate joint condition parameters. This section discusses the effect of freezing on joint condition input parameters to both the RMR and Q rock mass classification system. 8.1.2.1 Joint Condition The fourth input parameter of the RMR system is joint condition. It is a qualitative description of the discontinuities that relates to known frictional and cohesive strengths of joints. Table 8.6 describes the category and corresponding RMR rating for joint condition.  172  Table 8.6:  RMR Classification for Joint Condition (Bieniawski, 1976) Condition of Joints Very rough surfaces Not continuous No separation Hard joint wall rock  Rating  25  Slightly rough surfaces Separation < 1mm Hard joint wall rock  20  Slightly rough surfaces Separation > 1mm Soft joint wall rock  12  Slickensided surfaces or gouge < 5mm thick or joints open 1-5mm Continuous joints Soft gouge >5mm thick or joints open > 5mm Continuous joints  6  0  Roughness can be estimated using the joint roughness coefficient (JRC) chart (Barton, 1974). Determining separation of joints in drill core can prove to be difficult and requires experienced judgment by the logger. Similarly, it must be considered that infill on joints may be washed away through the drilling process. 8.1.2.2 Joint Roughness (Jr) In the Q system, joint condition is divided into joint roughness (Jr) and join alteration (Ja). Joint roughness in the Q systems is based on JRC, infill, and planarity, as shown in Table 8.7.  173  Table 8.7:  Q System Classification for Joint Roughness (Jr) (Hoek, 1980) Infill & JRC  Planarity  Jr  Slickensided  Planar  0.5  Slickensided  Undulating  1.5  Slickensided  Discontinuous  2.0  No infill, smooth (JRC <10)  Planar  1.0  No infill, smooth (JRC <10)  Undulating  2.0  No infill, smooth (JRC <10)  Discontinuous  3.0  No infill, rough (JRC >10)  Planar  1.5  No infill, rough (JRC >10)  Undulating  3.0  No infill, rough (JRC >10)  Discontinuous  4.0  Gouge-filled  Planar  1.0  Gouge-filled  Undulating  1.0  Gouge-filled  Discontinuous  1.5  8.1.2.3 Joint Alteration (Ja) The second part of the joint condition description in the Q system is joint alteration (Ja). It is often split into two groups: filled and unfilled. Table 8.8 lists the classification ratings for the joint alteration parameter in the Q system. The dilatant or contractile coefficient of friction for joints can be estimated through Jr/Ja (Barton et al., 1974).  174  Table 8.8:  Q System Classification for Joint Alteration (Ja) (Hoek, 1980) Alteration  Ja  Unfilled, staining only  1  Unfilled, slightly altered joint walls  2  Minor silt or sand coatings  3  Minor clay coatings  4  Sand or crushed rock filled  4  Stiff clay filling less than 5mm thick  6  Soft clay filling less than 5mm thick  8  Swelling clay filling less than 5mm thick  12  Stiff clay filling more than 5mm thick  10  Soft clay filling more than 5mm thick  15  Swelling clay filling more than 5mm thick  20  8.1.2.4 Effect of Freezing on Joint Condition Freezing improves the discontinuity considerations of rock mass characterization primarily by healing them, as discussed in the previous section. Healed joints should not be considered in design if the ice can withstand gentle hand pressure and the frozen state is expected to be constant (Robertson, 1988). Open joints, however, may be worse in frozen state than unfrozen. Ice could reduce the cohesion and friction below that of the original intact material. There is no change in RMR and Q for this parameter under freezing. A joint would need substantial strength to reduce the likelihood of a wedge failing along a frozen joint. A frozen joint can be treated as healed if it has a strength that approaches that of the intact rock material.  8.1.3 Water The influence of temperature on strength, is a function of the unfrozen water content, where at temperatures just below freezing there is water that has not converted to ice in the pores therefore the strength is lower than at sub-zero temperatures. The conversion of water to ice is a function of temperature, material type, porosity, salinity and confining pressures. When a rock mass 175  undergoes freezing, the degree of unfrozen water decreases as water in the pores converts to ice, creating a barrier to flowing water. Table 8.9 describes the categories and rating for water in the 1976 RMR system. Water ratings in the Q system (Jw) are not considered in this discussion.  Table 8.9:  RMR Classification for Water (Bieniawski, 1976) Water  Rating  Completely dry  10  Moist only (interstitial water)  7  Water under moderate pressure  4  Severe water problems  0  Typically RMR and Q calculations do not include water as discussed earlier given that groundwater is treated separately for the rock mass behaviour. Frozen ground is also considered impermeable as water is assumed to be converted to ice. Thus there is no change in the water parameter rating from unfrozen to frozen in the RMR’ and Q’ calculations for this comparison. For certain design applications, it may be necessary to adjust the rock mass quality to account the expected groundwater conditions.  8.2  Case Studies  The previous section has established that a gain in rock mass strength can be expected when the rock mass undergoes freezing, especially if the unfrozen state involves very weak to weak rock. Empirical data from case studies in the literature review also shows that rock mass ratings of weak rock are increased by up to 70%. However, caution should be used when comparing the data from the case studies below, as the method of recording the unfrozen and frozen RMR (i.e. from core logging or field mapping), varies between the sites. 176  The improvement in RMR from unfrozen to frozen conditions assessed by Wardrop (2005) assumed that the increased span opened in frozen conditions is relatable to a frozen RMR by the empirical Grimstad and Barton (1993) chart. Better practice is to assess the frozen RMR conditions in the field is with face mapping and to compare the unfrozen RMR conditions using geotechnical core logging. Figure 8.1, the Grimstad and Barton (1993) chart, shows the relation of the Q system of rock mass classification to the span and support requirements of an underground excavation, termed the equivalent dimension of the excavation, De. . Where:  De =  Excavation span, diameter or height (m) Excavation Support Ratio ( ESR)  ESR = Excavation Support Ratio (ranging from 3-5 for temporary mine openings to 1.6 for permanent mine openings)  177  Figure 8.1:  Empirical Support Design, after Grimstad and Barton (1993)  Increase in the RMR76 from an unfrozen to frozen state was recently assessed by Pakalnis and Mawson at Cameco’s McArthur River Mine (Cameco, 2012). Four unfrozen core logs were studied and compared to frozen face mapping of two drifts in the same area. Pakalnis and Mawson showed that the RMR was increased by an average of 38.Table 8.10 summarizes the average increase for each of the five parameters in the RMR system based on one hundred plus observations.  178  Table 8.10:  Average Increase Between Frozen Face Mapping and Unfrozen Core Logging Parameter  Average increase (frozen – unfrozen)  Strength  8  RQD  7  Joint spacing  11  Joint condition  11  Water  n/a  TOTAL average RMR increase  38  Freezing the rock mass has an effect of increasing rock quality through gains in strength, reductions in joint spacing (healing of joints), increases of joint quality condition, and removal of water. This translates into an overall RMR (and Q) increase where in some documented cases would be up to 40 points in the RMR rating for weak porous moist rocks. The biggest gain due to freezing of the RMR parameters is the RQD and joint spacing, compared to the intact rock strength parameter. This leads to the idea that the influence of freezing a weak discontinuous rock, has a significant effect on the rock mass, but less so on the intact rock, which was initially thought to control the excavation design. Table 8.11 and Figure 8.2 summarize the case histories of underground mine openings in permafrost and artificially frozen ground. The gain in strength of the RMR76 ranges from 13% to 68% from the unfrozen RMR76 value. The dashed green line represents the proposed unstablestable line for frozen RMR vs. cavity span.  179  Table 8.11:  Case History Summary of Frozen Rock Mass Conditions and Span Unfrozen (logged from core) RMR76  Frozen Span (m)  Equivalent Frozen RMR76  Percent Improvement in RMR76  17.8  70  50  79  13  3.4  55  8  63  15  2.8  52  5  68  31  0.15  47  50  70  68  Mine  Location  Source  Shkolnoye/Matrosov  Wardrop (2005)  Julietta Mine  745m L and 850 mL  Mining Method Shrinkage Stope  Wardrop (2005)  Longhole  Raglan Mine  Katinniq Ramp  Wardrop (2005)  Raglan Mine  KW 1475 Stope  Wardrop (2005)  Long hole  Raglan Mine  C 1460 L Cut  Wardrop (2005)  Cut and Fill  10  65  40  73  13  Raglan Mine  Q 1350 Cut  Wardrop (2005)  Cut and Fill  7.5  62  35  70  13  Kupol Mine  455 Level  Pakalnis (2012)  Long hole  -  25  24  60  140  Kupol Mine  530 Level  Pakalnis (2012)  Long hole  -  25  5  65  160  Cigar Lake  Cavity 1  Cameco (2000)  Jet Boring  -  30  6  50  65  Cigar Lake  Cavity 2  Cameco (2000)  Jet Boring  -  30  3.5  50  65  Cigar Lake  Cavity 3a  Cameco (2000)  Jet Boring  -  30  5  50  65  Cigar Lake  Cavity 4  Cameco (2000)  Jet Boring  -  30  4.5  50  65  McArthur River  510-8240 N  Cameco (2012)  Roadheader  -  20  7  55  35  McArthur River  510-8220 N  Cameco (2012)  Roadheader  -  30  7  65  35  Q’  Suggested Frozen Unsupported Unstable/Stable Limit Based on the Unfrozen RMR76  Suggested Upper Limit Frozen RMR76 Unfrozen RMR76 Frozen RMR76 (refer to Table 8.11)  Figure 8.2:  Case Studies Frozen RMR vs. Cavity Span on the McArthur River Rock Mass Critical Span Curve, after Pakalnis (2012) 180  Implications of quantifying the observed increase span or gain in rock mass rating value to the Cigar Lake mine are that: •  A larger span can be opened up with reduced ground support  •  The weaker rocks (RMR76 < 45) see substantial gain in strength, up to an additional 30 points on the RMR76 scale.  8.3  Comparison of Unfrozen to Frozen 2009 Surface Freeze Drilling Rock Mass Classification  Applying the interpretation of case history data of mines in permafrost or artificially frozen ground and the laboratory testing from the Cigar Lake mine to the input parameters of the rock mass classification (RMR) system can be summarized as follows. Strength In R0 to R2 unfrozen rock, RMR may be increased by as much as 7 points when frozen. In unfrozen rock strengths higher than 50 MPa, the intact rock parameter remains the same. RQD A rock with an RQD of zero (0) could improve up to one hundred (100) through freezing, by making very weak rock sound and intact by healing all the joints. In the RMR system this would result in an increase from as low as three (3) to as high as twenty (20). Joint Spacing A rock with joint spacing less than fifty millimeters (<50mm) could have a spacing of >3m once frozen, resulting in an increase in RMR from five (5) to thirty (30). Joint Condition Freezing improves the discontinuity considerations of rock mass characterization primarily by healing them, however, the type of infilling and surface roughness will not change when undergoing freezing. No change in the joint condition rating is expected for tight aperture joints; however for joints that are slightly open freezing will heal the joints. The change in RMR for this parameter is zero.  181  Groundwater Frozen ground is considered impermeable as all the water is converted to ice and dry conditions are often considered in unfrozen rock mass classification calculations as groundwater is considered separately. Typically the rockmass characterisation treats the presence of water as a negative attribute. However in frozen ground, water acts as a bonding agent between the particles and is the cause of strength increase under freezing conditions, thus improving ground conditions. In freezing ground, therefore, water is a positive parameter. The influence of the moisture content on the ground conditions (i.e. dry to saturated) was not the focus of this research. Groundwater will be left out of frozen RMR calculations in this research and there is no change assumed in the groundwater parameter for the frozen RMR' until further studies address this topic. An additional moisture content parameter is proposed to be included in the frozen RMR calculations to address the gain in strength with increasing water content from dry to partially saturated under freezing conditions. 8.3.1 Discussion From the 2009 geotechnical drilling program, an average RMR76 from the borehole logging (Section 6.5) is outlined below (in red). The influence of freezing based on interpretation of case history data and the expected increase in each RMR’ parameter is outlined in green. The unfrozen RMR’76 value is 40 and the estimated frozen RMR’76 value is 74, an overall increase of 83 percent. Future studies should use data from both drill core and excavated faces and the results separated to deal with any potential bias. Calculating the frozen RMR based on drillcore without mapping the face can lead to overestimating the expected frozen ground conditions.  182  Figure 8.3:  Unfrozen Frozen Comparison of an Unfrozen RMR to Frozen RMR, after Bieniawski (1976)  In addition to the RMR system, the influence of freezing can be illustrated using the GSI system, below in Figure 8.4. The unfrozen rock mass of the altered sandstone directly overlying the orebody collected from the Cigar Lake 2009 drill program is described as disturbed with poor to fair joint surface conditions, correlating to a GSI of 25 to 40. Based on the observations of the frozen rock samples in the laboratory, the influence of freezing on the joint surface condition does not change; however, the structure of the rock mass due to healing of the joints and increase in rock strength under frozen conditions has the potential to modify the structure to be intact to massive, an increase of the GSI from 60 to 80.  183  Unfrozen Figure 8.4:  Frozen  GSI Values for Blocky Rock Masses with Unfrozen and Frozen RMR, after Marinos and Hoek (2000)  Figure 8.5 visually depicts the expected gain in rock mass classification values based on Figure 8.3 and Figure 8.4 above, from the surface freeze drill hole sections as calculated in Section 6.5.  184  UNFROZEN ST786-07  ST791-07  FROZEN  ST796-05  unconformity  ST801-05  ST791-07  ST796-05  ST801-05  ore  Unfrozen RMR76 < 20 Unfrozen RMR76 20 – 35 Unfrozen RMR76 35 – 45  Figure 8.5:  ST786-07  Frozen RMR76 60 – 70 Frozen RMR76 70 – 80 Frozen RMR76 80 – 90  Cross Section North 10,032, Unfrozen and Frozen RMR76  185  9.  Failure Mechanism of Frozen Weak Rock Masses  This section is a summary of the geotechnical inputs for numerical modelling, including the Mohr-Coulomb parameters cohesion and friction and the Hoek Brown parameters, for the HoekBrown failure criterion and how they would be influenced by freezing. The mechanical behaviour of frozen ground differs from unfrozen behaviour due to the ice and water composition, which varies with temperature and applied stress. The behaviour of frozen soil is well documented with extensive research in the mechanical and creep relationships with varying grain sizes, moisture, and temperature. Limited information exists on the behaviour of frozen weak rock as the majority of frozen ground research is based on permafrost regions in surficial soil. As the temperature drops in a rockmass, mineral grains shrink and the formation of ice in pore spaces contributes directly to the strength of the material. The water that changes phases converts to ice increasing in volume by 9%. In the case of the Cigar Lake Project, the frozen material over a jet bored cavity will be subjected to hydrostatic pressure (in situ stresses and water in the sandstone), shear stresses (shear zone caused by fracturing and squeezing ground around the ore zone) and a creep regime (presence of ice and squeezing environment). The behaviour and stability of frozen material over the mined out cavities once mining commences is a function of the frozen rock mass. Failure can occur due to wedge fall, slab failure, gravity driven caving, and beam failure. There is the potential for high and uncontrolled groundwater inflow events that are mitigated through artificial ground freezing. Assuming an ice cap thickness of 10 m above the jet bored cavity, a hydraulic gradient (i) of 45 will be present at the back of the cavity (450 m head at 10m from the cavity, assuming 0 m of head at the back of the cavity). The pressure jets will thaw the cavity walls, creating unfrozen strengths. The ice cap thickness design must keep stresses acting uniformly around the cavity, withstand hydrostatic pressure at 450 m depth, not crack, and remain stable prior to backfilling, possibly up to 3 weeks from mining from top down.  186  9.1  Mohr-Coulomb Criterion  The Mohr-Coulomb shear strength of frozen rock or soil is defined through triaxial compression tests on frozen samples. The Mohr-Coulomb strength criterion assumes that a failure of the rock material occurs through the development of a shear plane. When failure occurs, the stresses developed on the shear plane define a strength envelope.  Figure 9.1:  Mohr-Coulomb Failure Envelope  The Mohr-Coulomb relationship suggested that the shear strength of rock is made up of two parts, a constant cohesion (c) and a normal stress-dependent frictional component, τ = c + σn tanφ. Where: c = cohesion φ = internal friction angle  In a shear stress-normal stress plot, the Coulomb shear strength criterion τ = c + σn tanφ is represented by a straight line, with an intercept c on the τ axis and an angle of φ with the 187  σn axis. This straight line forms the strength envelope. Extrapolating the linear Mohr-Coulomb strength envelope, the unconfined compressive strength (UCS, σc) can be derived by c and φ as: σc = 2c cos φ / 1 - sin φ  The angle of failure of the sample, defined as Β, is related to the internal friction angle where: Β = 45+ φ / 2 From the frozen UCS samples completed on the 2009 surface freeze drilling boreholes, discussed in Section 7, the angle of failure ranges from 50 to 60 degrees for samples that did not fail along bedding.  Figure 9.2:  Example of UCS Failure Angle  188  Table 9.1:  Test Temp (oC) -20 -20 -10 -10 -10  Material Type Hematized Sandstone Hematized Sandstone Altered Basement Bleached Sandstone Bleached Sandstone  Summary of UCS Failure Angles Average Moisture Content (by Wt)  UCS (MPa)  Angle of Failure Not Along Bedding or Joint (Β)  Friction Angle Based on Failure Angle (φ)  UCS Sample ID  Borehole  Depth (m)  Strain Rate (%/min)  19  SF801-04  434.7  0.15  22.8  3.39  60  30  20  SF801-04  435  0.03  20.9  4.16  55  20  11  SF801-04  441.28  0.13  22.0  2.80  50  10  6  ST786-07  427.55  0.14  35.6  2.12  60  30  7  ST786-07  427.73  0.01  38.1  1.57  55  20  Based on this relationship, the friction angle (φ) of the frozen rock samples can be back calculated to approximately 15 degrees and does not appear to be dependent upon temperature from the samples tested. Additional testing would confirm if there indeed is a difference in the friction angle between -20 to -10oC. It should also be noted that the angle of failure, especially under triaxial loading conditions where axial splitting dominates, is also significantly influence by sample end effects. From the samples tested as part of this research, the frozen friction angle does not appear to be affected by temperature or applied strain rate. Very weak rock samples (unfrozen strength less than 2 MPa) typically failed on obvious shear plans, such as bedding or pre-existing joints. Samples tested with unfrozen moisture contents greater than 30% did not fail on pre-existing shear planes but rather on the friction plane. Jessberger et al. (2003) states that it is typical practice to assume in frozen soils that the angle of internal friction is neither influenced by temperature nor loading distribution and that only cohesion is temperature dependent. However, this assumption is not always true and the angle of friction is based on the angle of internal friction for the average freeze wall temperature using the allowable long term compressive stress.  189  9.2  Hoek-Brown  The Hoek-Brown failure criterion was developed to design underground excavations in hard rock masses by Hoek and Brown (1980). Hoek and Brown linked Bieniawski's Rock Mass Rating (RMR) and later the Geological Strength Index (GSI) a visual tool for field mapping to define failure criteria through research of the brittle failure of intact and jointed rock. The Hoek–Brown criterion is an empirical equation for non-linear strength material developed through curve fitting of triaxial test data. The generalized Hoek-Brown criterion is defined as:  𝜎��  =  𝜎��  𝜎�� + 𝑠� + 𝜎�� �𝑚� 𝜎��  �  where mb is a reduced value of the material constant mi and is given by: mb=mi exp (  GSI-100 ) 28-14D  s and a are constants for the rock mass given by the following relationships: 𝐺𝑆𝐼 − 100 𝑠 = 𝑒𝑥𝑝 � � 9 − 3𝐷 𝑎=  1 1 ���� ��� + (𝑒 �� -𝑒 � ) 2 6  The rock mass uniaxial compressive strength is defined by: 𝜎� = 𝜎�� ∙ 𝑠 �  where 𝜎�� is set to zero in the failure criteria equation above.  The Hoek-Brown failure criterion was initially not developed for very poor quality rock masses and included the disturbance "D" parameter to force the tensile strength to zero. D ranges from 0 for TBM tunnels to 1.0 for very poor blasting. GSI refers to the Geological Strength Index (Marinos and Hoek, 2000) and is equivalent to RMR76 or RMR89 minus 5. 190  Using RocLab 1.0 (Rocscience, 2012), the rock mass parameters can be derived by scaling the laboratory derived intact rock properties using the rock mass characteristics quantified using GSI. This was done here for the frozen Cigar Lake based on the following assumptions: •  • •  • •  An intact Hoek-Brown mi parameter of 8 was assumed based on the value recommended in RocLab values for similar claystone/sandstone rock. Where possible, the mi value should be derived from triaxial testing A frozen UCS of 5 MPa was adopted based on the average value for frozen altered sandstone at -10oC. A GSI value of 50 based on the assumed increase of 20 points in the RMR from unfrozen to frozen was selected, where the unfrozen RMR76 of the altered sandstone overlying the orebody was 30. A disturbance factor, D = 0 was selected. A failure envelope stress condition at 450 m depth was assumed.  The latter assumption is required for converting the rock mass Hoek-Brown values to MohrCoulomb rock mass values. Because the Hoek-Brown failure envelope is non-linear, the linear Mohr-Coulomb values are estimated by fitting a straight line to the non-linear curve at the required minimum principal stress (determined here based on the depth of mining). The MohrCoulomb friction and cohesion values derived for the above assumptions are 19o and 0.5 MPa, respectively.  9.3  Frozen Material Properties  The uniaxial compressive strength of the frozen material is as important as the modulus of elasticity (E) for structural design of frozen ground. The UCS of frozen soils is typically defined as a function of applied strain rate as the shapes of the σ1-e1 curves will vary for the same material. Frozen UCS test results of the same material from the 2009 surface freeze drilling program, were highly variable within close proximity of the orebody due to the varying alteration of the rock mass. A range of values is suggested for the orebody and clay cap given its heterogeneous mixture of materials. 191  Based on the results of the UCS testing in Section 7 where the unfrozen rock strength was 0.5 to 3 MPa, and historical testing from EBA (1999), Table 9.2 lists the recommended frozen rock mass material properties. Note that limited strength testing was completed on the altered graphitic metapelite basement material, as the focus of this research was the altered sandstone material overlying the jet bored cavities. Table 9.2: Material  Bleached Sandstone (intermediate clay) Hematized Sandstone (indurated clay)  Ore  Altered Graphitic Metapelite Basement  Temp. (oC)  Frozen Material Properties  +20 -5 -10 -20 +20 -5 -10 -20 +20 -5 -10 -20 +20 -5 -10  Peak Strength (MPa) 7 2±2 4±3 1.2 ± 4.3 2.5 4.5 5 10 No data No data 2 No data 5±3  Residual Strength (MPa) 0.5 2 0.5 2 -  Friction (o )  Cohesion (MPa)  E (GPa)  Porosity  0.5  35 15 20 15 20 25 10 45 -  2 0.76 1.4 0.9 1.5 1.6 0.87 0.42 -  5 1 2 1 2 2 1 3.1 2±1  1  -  -  2±1  0.25 0.4 – 0.5 0.4 – 0.5 0.3 – 0.4 0.3 – 0.4 0.3 0.3 0.3 0.35 – 0.4 0.35 – 0.4  -20  8±4  192  10.  Conclusions  The purpose of artificial ground freezing at the Cigar Lake mine is to ensure stability of the jet bored cavities during mining and minimize groundwater inflow. The majority of this study focussed on the gain in strength due to freezing of a very weak, altered and jointed rock mass sampled directly above and below the Cigar Lake orebody. Although, well defined trends in the data were not established, it is clear there is a significant gain in strength of the rock mass due to freezing. The influence of freezing was initially thought to be controlled by the gain in the intact strength (UCS) from the unfrozen to frozen properties of the rock. However, the healing of joints under freezing conditions was found to add tensile strength under short term loading conditions significantly improving the rock mass quality from a very poor to a good quality rock mass when frozen. The Cigar Lake rock mass is not intact but a blocky to very blocky/disintegrated rock mass with discontinuities. The benefit of freezing at the Cigar Lake mine is the addition of joints taking on properties of the rock matrix, changing from having zero tensile strength and cohesion in unfrozen conditions. 10.1  Cigar Lake Rock Mass Highly Variable  The layer that will control the stability of the jet bored cavity is the clay altered sandstone directly over the ore, typically consisting of a hematized sandstone or if not present the bleached sandstone. Both of these layers can be completely altered to a dense clay or very weak rock (< 5 MPa unfrozen strength). There are no clear rock mass transition zones between boreholes or with depth as anomalous zones of very poor or medium strong rockmass are present. The transition in alteration from the orebody may not vary as a vertical gradient with distance away from the orebody, but rather a mixture of materials controlled by faulting. 10.2  Frozen Laboratory Testing  Improving in situ and laboratory characterization methods and a better understanding of the rock behaviour at low temperatures was the key focus of this research. Frozen Unconfined Compressive Strength (UCS), frozen direct shear, and frozen beam tests were completed on drill core material from the Cigar Lake project. The effect of freezing on a frozen weak rock mass can 193  be summarized as follows: •  • • • • • • • • •  10.3  The UCS failure changes from strain-softening to elastic/plastic with decreasing temperature, and the gain in strength from unfrozen is double for unfrozen material weaker than 5 MPa. Gain in strength of the material tested from -10oC and -20oC is minimal. The material tested is not strain rate dependent at temperatures of -10oC and -20oC. Samples tested at -20oC can withstand higher strain until failure compared to samples tested at -10oC. The residual strength of the material at -10oC and -20oC should behave the same for each material type, independent of temperature. From the frozen UCS testing, the sample failure mode observed from -10 to -20oC was the influence of ice taking over, becoming elastic perfectly plastic. The joint is always the weakest link, though dependent on the loading direction. For samples greater than 30% moisture, a frozen joint is as strong as a the frozen rock mass. The benefit of freezing a weak jointed rock mass is the addition of the tensile strength. For unfrozen rock strengths less than 2 MPa (based on field strength), the frozen joint is as strong as frozen rock mass. For unfrozen rock strengths greater than 2 MPa (based on field strength), the joint was observed to be weaker than the frozen rock mass  Intact Rock Strength and Rock Mass Quality  Freezing the rock mass has an effect of increasing rock mass quality through gains in strength, reductions in joint spacing (healing of joints), increasing joint quality condition, and the conversion of water to ice. This translates into and overall RMR (and Q) increase where in some documented cases would be up to 40 points in the RMR rating for weak porous moist rocks. At the McArthur River mine, the largest increase in rock mass classification values, and ground conditions, were observed in drillcore core that would have been classified as the poorest ground (RMR less than 40), while more competent ground tended to have more comparable RMR values between core logging and face mapping. From the back analysis of the Cigar Lake jet boring trial in 1999, the influence of freezing on weak rock is clearly shown to increase the rock mass conditions from an estimated unfrozen RMR of less than 35 of the jet bored cavities to approximately 50 (based on the stable 194  unsupported line for a 5 m span). This increase in the frozen rock mass strength is attributed to the increase in cohesion and UCS of the weak rock as the pore water freezes.  195  11.  Recommendations  This section discusses the proposed recommendations for future work based on the outcomes of this research. 11.1  General  Geotechnical descriptions of the Cigar Lake material including the "clay cap" or altered hematized and bleached sandstone overlying the orebody are heterogeneous and should be described by a range of values and not one point value. 11.2  Laboratory Testing  Additional UCS and direct shear tests are suggested along with triaxial testing varying strength, mineralogy, and moisture content to gain a better understanding on the frozen shear strength behaviour of a weak and jointed rock. Improvements to the UCS testing completed with the 2009 surface freeze drill core include: •  Better measurements of the vertical displacements of the loaded UCS sample. The vertical displacement of the top of the sample was measured with a screwdriver connected to an LVDT. The vertical displacement recording was not always consistent as the screwdriver did not always move with the loading platen.  •  Testing of weak samples was biased due to the ability to trim and prepare the core. Half of the samples collected could not be trimmed as they were too friable.  •  Freezing the samples under a confining load to simulate the conditions expect at Cigar Lake.  •  Additional UCS testing to evaluate the post peak characteristics of the frozen sample during failure  A series of direct shear tests from unfrozen, open and frozen, and healed with ice should be tested with varying roughness and infill. Significant gains in rock mass quality can be made in the reduction of open joints through freezing and future work should focus on this aspect, investigating the controlling factors on the healing of joints.  196  11.3  In Situ Testing  Design and construction of a freeze wall requires reliable strength and deformation material properties. The majority of material properties are from laboratory testing; however, the effect of sample disturbance prior to lab testing is an issue to address. In situ testing methods are recommended to minimize the effect of sample stress relief and quantify the material properties on a larger scale. In situ testing can be carried out in materials that cannot be sampled without considerable disturbance and with a larger volume of soil tested than in the laboratory. However, strain rates applied during in situ testing are often higher than applied in field or laboratory. Laboratory testing has well defined boundary conditions with reasonably uniform stress and strain fields applied on the samples. In situ testing methods can minimize the effect of sample stress relief, quantify material properties on a larger scale, and reduce the concern of relying upon data from samples in zones of poor core recovery. In situ tests recommended include the following: •  Permeability testing: packer testing, falling head or slug test  •  Strength and deformation testing: pressuremeter testing, downhole shear wave velocity, pocket penetrometer  •  Moisture content and temperature: resistivity probe  In situ testing methods must be done in an open uncased hole. Given the high risk of hole collapse in the target sampling area, in situ methods were not selected at Cigar Lake mine due to the high risk of hole collapse. Geophysical methods by downhole surveys in an open borehole or from surface can provide the properties of the surrounding rock mass such as porosity, moisture content, density, and contrasts in conductivity over larger areas than a drillhole. Geophysical methods to measure the in situ properties of the frozen and unfrozen Cigar Lake material are suggested including: • •  Downhole seismic survey, where an active nuclear source probe is placed down an open borehole to measure the insitu density and rock modulus. Downhole gamma and conductivity survey to measure the in situ density relatable to the 197  porosity 11.4  Developing Empirical Relationship Unfrozen to Frozen Rock Mass  These unfrozen to frozen rock mass relationships are based on a limited data set. Quantifying the change in rock mass from unfrozen to frozen conditions is recommend to be based on unfrozen drill core and compared with the face mapping of frozen excavations to establish a detailed relationship. The expected gain in the rock mass condition from unfrozen to frozen greatly depends on the unfrozen strength, blockiness, joint infilling, and temperature.  11.5  Numerical Modelling  It is proposed that once future laboratory testing confirms changes in rock properties due to freezing, numerical modeling approaches can be applied to assess the stability of mining excavations under varied conditions. It would be particularly beneficial to determine which constitutive model best represents the stress-strain behavior of frozen rock masses. It is important to find whether the behavior is strain softening, creep, or fully coupled thermal-fluid models. The weak highly variable soil and rock from the Cigar Lake mine is expected to demonstrate complex non linear behaviour, what constitutive laws apply is suggested for future work.  198  References Andersland, O.B. and B. Ladanyi. 2004. Frozen Ground Engineering. 2nd Ed. John Wiley & Sons, Inc., Hoboken, N.J. Andres, Dale. 1999. “The evolution of pillar mining at the Polaris Mine”. Proceedings of the 5th International Symposium on Mining in the Arctic. Yellowknife, Canada. June 1998. pp 125-134. Aoki, K., Hibiya, K. & Yoshida, T. 1989. Storage of refrigerated liquefied gases in rock caverns: Characteristics of rock under very low temperature, Nilsen (ed.), Proc. Intern. Conf. On storage of gas in rock caverns, Trondheim, pp. 237–243. Arenson, L., M.M. Johansen, S.M. Springman. 2004. “Effect of volumetric ice content and strain rate on shear strength under triaxial conditions for frozen soil samples.” Permafrost and Periglacial Processes. Vol 15. pp. 261-271. Arenson, L., and S. Springman. 2005. "Mathematical descriptions for the behaviour of ice-rich frozen soils at temperatures close to 0oC". Canadian Geotechnical Journal. Vol 42. pp 431-442. Baker, T.H.W. 1979. Strain rate effect on the compressive strength of frozen sand.” 1st International Symposium on Ground Freezing, Bochum, Germany. Edited H.L. Jessberger. pp. 223-231. Bandopadhyay, S., X. Wang., and M.G. Nelson. 1996. “Roof span stability in an underground placer mine in the arctic.” Proceedings of the fourth international symposium on Mining in the Arctic. Longyearbayen, Svalbard, Norway, 27-30 July, 1996. Barton, N.R., R. Lien, and J. Lunde. 1974. Engineering classification of rock masses for the design of tunnel support. Rock Mechanics. Volume 6. Pp 189-239. Baudemont, 2000. Core orientation from underground geotechnical holes, Tentative modeling of post-Athabasca faulting. Report prepared for COGEMA Resources. Issued January 2000. Baudemont, D. And Rafini S. 2000. Cigar Lake deposit, Eastern Athabasca. Core orientation from underground geotechnical holes. Tentative modelling of post-Athabasca faulting. Internal report, Areva. 28p. Bieniawski, Z.T. 1973. Engineering classification of jointed rock masses. Trans. South African Institute Civil Engineering. Volume 15. Pp 335-344 Bieniawski, Z.T. 1976. Rock mss classifications in rock engineering. Exploration for Rock Engineering. Vol 1, pp. 97-106. A.A. Balkema: Cape Town Bieniawski, Z.T. 1989. Engineering Rock Mass Classifications. Wiley: New York. Brady, B.H.G. and E.T. Brown. 2006. Rock Mechanics for Underground Mining. 3rd ed. Springer, Netherlands. 199  Bruneton, P. 1993. “Geological environment of the Cigar Lake uranium deposit.” Canadian Journal of Earth Sciences. Vol. 30. pp 653-673. Bruneton, P., 1986. Geology of the Cigar Lake uranium deposit (Saskatchewan, Canada) in C.F. Gilboy and L.W. Wigrass, eds., Economic Minerals of Saskatchewan, Geological Society of Saskatchewan, Special Publication, no. 8, p. 99-119. Cameco Corporation, 2007. “Cigar Lake Project, Northern Saskatchewan Canada, NI 43-101 Technical Report.” Issued March 30, 2007. Cameco, 1996. “Design ground freezing numerical analysis: pond 2, panels one and two” McArthur River project. Issued October 1996 Cameco, 1997. “Cigar Lake and McArthur River Freezing hole drilling programs – Past experiences and proposed methods” McArthur River project Cameco, 1998. “pod two panel three preliminary extraction plan.” McArthur River project. Prepared by Mine Engineering Department, McArthur River project. Issued September 1998. Cameco, 2000. “2000 Jet Boring Systems Test – Final Report”. Cigar Lake Mining Corporation. Chamberlain, Edwin. 1973. Mechanical properties of frozen ground under high pressure. Permafrost second international conference. 295-305. Chislov, A.I. 1991. “Peculiarities of rock massif behaviour in a zone of transition from subzero to above zero temperatures.” Int. Arctic Technology Conference, SPE, Anchorage, AK, May 2931, p. 593-602. Czurda, K.A. and M. Hohmann. 1997. “Freezing effect on shear strength of clayey soils. Applied Clay Science. Vol 12. pp. 165-187. Deere, D.U. 1964. Technical description of rock cores for engineering purposes. Rock Mechanics and Rock Engineering. Vol. 1. pp. 17-22. Deere, D.U. and D.W. Deere. 1988. Rock Quality Designation (RQD) after Twenty Years. U.S. Army Corps of Engineers. Contract Report GL-89-1. Diederichs, M.S. and P.K. Kaiser. 1999. “Stability of large excavations in laminated hard rock masses: the voussoir analogue revisited.” International Journal of Rock Mechanics and Mining Sciences. Vol 36. pp. 97-117. Dusseault, M.B. and C.J. Fordham. 1993. "Time-dependent behaviour of rocks". Journal of Comprehensive Rock Engineering. Volume 3. pp 119-149. Dutta, P.K. and K. Kim. 1993. High-Strain-Rate tensile behaviour or sedimentary and igneous rocks at low temperatures. U.S. Army Corps of Engineers Cold Regions Research and Engineering Laboratory. CRREL Report 93-16. 200  Dwivedi, R.D., A.K. Soni, R.K. Goel, A.K. Dube. 2000. “Fracture toughness of rocks under subzero temperature conditions. International Journal of Rock Mechanics & Mining Sciences. Vol 37. pp. 1267-1275. EBA Engineering Consultants Ltd.. 1990. 'Stress Analysis Ground Freezing Design - Cross-Cut Drift - 420 M Level.' Report EB-90-01. Economopolous, J.N., A.I. Sofianos., and N.J. Koronakis. 1994. Voussoir beam response of bedded limestone roofs in Greek underground mining excavations. Laboratory of Mining Technology, National Technical University of Athens, Greece. Fayek, M., J. Janeczek, and R. Ewing. 1997. “Mineral chemistry and oxygen isotopic analyses of uranite, pitchblende and uranium alteration minerals from the Cigar Lake deposit, Saskatchewan, Canada.” Journal of Applied Geochemistry, Vol. 12, pp. 549-565. Fayek, M., T.M. Harrison, R.C. Ewing, M. Grove, C.D. Coath. 2002. O and Pb isotopic analyses of uranium minerals by ion microprobe and U-Pb ages from the Cigar Lake deposit. Journal of Chemical Geology. Volume 185, issue 3-4, May 2002, p. 205-225. Fouques, J.R., Fowler, M., Knipping, H.D., Schimann, K., The Cigar Lake uranium deposit; discovery and general characteristics, in Evans, E.L. (ed), Uranium deposits of Canada: Canadian Institute of Mining and Metallurgy, Special v. 33, p. 218-229. Geo-Analysis Ltd. 2000a. “Analysis and calibration of freezing model for recent test mining area.” Memorandum issued to CLMC, November 2000. Geo-Analysis Ltd. 2000b. “Artificial ground freezing of the McArthur River Uranium ore deposit.” Geoconcept, 1995. “Ground Freezing studies, 1994-1995.” Prepared for Cameco Corporation. Issued July 1995. Geo-Slope. 2004. “Ground freezing modelling for CLMC.” Memorandum issued to Cameco Corporation, August 2004. Geo-Slope. 2007. “Ground freezing modelling for Cigar Lake mine.” Memorandum issued to Cameco Corporation, October 2007. Giegerich, H.M. 1992. “Keynote address: Mining in the Arctic”. Proceedings of the 2nd International Symposium on Mining in the Arctic. Fairbanks, Alaska, July 1992. pp 3-9. Glamheden, R. And U. Lindblom. 2002. “Thermal and mechanical behaviour of refrigerated caverns in hard rock.” Tunnelling and Underground Space Technology. Vol 17. pp 341-353. Grimstad, E. and Barton, N. 1993. Updating the Q-System for NMT. Proc. int. symp. on 201  sprayed concrete - modern use of wet mix sprayed concrete for underground support, Fagernes. 46-66. Oslo: Norwegian Concrete Assn. Golder Associates, 1995. “Geotechnical characterization and stability assessment of the proposed mining plan for the McArthur river project.” Prepared for Cameco Corporation. Issued July 1995. Golder Associates, 1995. “Underground flow testing program, McArthur river project” Prepared for Cameco Corporation. Issued August 1995. Golder Associates, 1997. “Hydraulic conductivity testing vent shaft pilot holes MAC 258 and MAC 259, McArthur River project.” Prepared for Cameco Corporation. Issued February 1997. Golder Associates, 1999. “Hydraulic conductivity testing shaft no. 3 pilot hole MAC 260, McArthur River project.” Prepared for Cameco Corporation. Issued February 1999. Golder Associates. 1985. 'Geotechnical Model Above and Below Clay-Ore Zone.' Golder Associates. 1985. 'Preliminary Evaluation of the Freezing Properties of the Clay-Ore Zone.' Report GO-85-09. Golder Associates. 1986. '1986 In situ testing program Cigar Lake Property Main Text and Appendix A, A geotechnical report on drillholes WDG1-193A, WDG1-195 and WDG1-196.' Golder Associates. 1986. 'Additional Laboratory Testing of Clay-Ore Zone Samples’. Golder Associates. 1986. 'Borehole Deformation and Hydrofracturing Test Analysis of Holes WDG1-193A, WDG1-195 and WDG1-196 in the Cigar Lake Orebody.' Report GO-86-06. Golder Associates. 1986. 'Laboratory Testing of Samples Borehole WDG1-187A.' Report GO86-03. Golder Associates. 2001. 'Numerical Analyses To Estimate Stresses in Crosscut Linings At Cigar Lake Mine.' Report . Golder Associates. 2002. 'Production Horizon Analysis and Validation Study.' Report Goodman, 1989. Introduction to Rock Mechanics. John Wiley and Sons, New York. Goto, S., K. Minegishi, M. Tanka. Direct Shear Test at a frozen/unfrozen interface. 1988. 5th International Symposium on Ground Freezing. Jones & Holden eds. 1988 Balkema Rotterdam Goughnour, R.R. and O.B. Andersland. 1968. “Mechanical properties of sand-ice system.” ASCE Journal of the Soil Mechanics and Foundations Division. Vol 94. Pp. 923-950. Gunzel, F. 2008. “Shear strength of ice-filled rock joints. 9th International conference on permafrost, Fairbanks Alaska. pp 581 202  Harris, J.S. 1995. Ground Freezing in Practice. Thomas Telford Services, Ltd., London. Hoek, E. and Brown, E.T. 1980. Underground Excavations in Rock, London, Instn Min. Metall. Hoek, E., Carranza-Torres, C., and Corkum, B. 2002. "Hoek-Brown Failure Criterion - 2002 Edition." Hoek, 2006. Practical Rock Engineering. John Wiley and Sons, New York.. Hoek, E. 1999. Support for very weak rock associated with faults and shear zones distinguished lecture for the opening of the International Symposium on Rock Support and Reinforcement Practice in Mining, Kalgoorlie, Australia, 14-19 March, 1999. Hoeve, J., and Quirt, D. 1984. Mineralization and host-rock alteration in relation to clay mineral diagenesis and evolution of the Middle-Proterozoic Athabasca basin, northern Saskatchewan, Canada. Saskatchewan Research Council, Technical Report 187. Hoeve, J., Rawsthorn, K., and Quirt, D. 1981. Uranium metallogenic studies: clay mineral stratigraphy and diagenesis in the Athabasca Group. In Summary of investigations. Saskatchewan Geological Survey, Miscellaneous Report 81-4, pp. 76-89. Hutchinson, D.J. and M.S. Diederichs. 1996. Cablebolting in Underground Mines. BiTech Publishers Ltd., Canada. Inada, Y. And N. Kinoshita. 2003. “A few remarks on storage of low temperature materials in rock caverns.” ISRM 2003 - Technology roadmap for rock mechanics, South African Institute of Mining and Metallurgy. Inada, Y., Kinoshita, N., Kohmura, Y., Matsumoto, T., 1999. Consideration on Thermal Behaviour of Rock Mass Around Openings Affected by Thermal Hysteresis of Low Temperatures. A.A. Balkema, Rotterdam. Inada, Y., Kinoshita, N., Seki, S., 1995. Thermal behaviour of rock mass around openings affected by low temperature. Proceedings of the Eighth International Congress on Rock Mechanics, pp. 721– 724. Inada, Y., N. Kinoshita, A. Ebisawa, S. Gomi. 1997. “Strength and deformation characteristics of rocks after undergoing thermal hysteresis of high and low temperatures.” International Journal of Rock Mechanics and Mining Sciences. Vol. 34:3-4. Paper No. 140. Inada, Y., Yagi, N., 1980. Mechanical characteristics of rocks at very low temperatures. J. Soc. Mater. Sci. 29, 1228–1233. Japan (In Japanese with subtitles in English). Ishizuka, Y., Kinoshita, N., 1988. Analysis of crack propagation around LPG storage cavern using fracture mechanics. Shimizu Tech. Res. Bull. (Tokyo) 7, 55–64. Ishizuka, Y., Kinoshita, N., Okuno, T., 1985. Stability analysis of a rock cavern with LPG 203  storage under thermal stresses. Proceedings of the Fifth International Conference on Numerical Methods in Geomechanics, vol. 2. A.A. Balkema, Rotterdam, pp. 1233–1240. Jefferson, C.W., Thomas, D.J., Gandhi, S.S., Ramaekers, P., Delaney, G., Brisbin, D., Cutts, C., Quirt, D., Portella, P., and Olson, R.A., 2007, Unconformity associated uranium deposits of the Athabasca Basin, Saskatchewan and Alberta, in Goodfellow, W.D., ed., Mineral Deposits of Canada: A Synthesis of Major Deposit-Types, District Metallogeny, the Evolution of Geological Provinces, and Exploration Methods: Geological Association of Canada, Mineral Deposits Division, Special Publication No. 5, p. 273-305. Jessberger, H.L. R. Jagow-Klaff, and B. Braun. 2003. “Ground Freezing” in Geotechnical Engineering Handbook, Volume 2: Procedures. Ed. Ulrich Smoltczyk. Ernst & Son, pp. 117-167. Kaplar, C.W. 1971. “Some strength properties of frozen soil and effect of loading rate. U.S. Army Cold Regions Research and Engineering Laboratory, Hanover, New Hampshire. Special Report 159. Kennard, D. 1998. Numerical Modelling Analysis of Excavations in Weak Rock at the Cigar Lake Test Mine. M.A.Sc. Thesis. University of Saskatchewan, Department of Geological Sciences. Konrad, J.M. 2002. “Prediction of freezing-induced movements for an underground construction project in Japan. Canadian Journal of Geotechnical Engineering. Vol 39, pp. 1231 – 1242. Kumar, A.1968. The effect of stress rate and temperature on the strength of basalt and granite. Geophysics, 33: 501-510. Kuriyagawa, M., Matsunaga, I., Kinoshita, N. & Hibi, K. 1980. Rock behavior of underground cavern with the storage of cryogenic liquefied gas. Proceeding International Symposium on subsurface space, Stockholm, p. 665. Ladanyi, B. 2006. “ Creep of frozen slopes and ice-filled rock joints under temperature variation.” Canadian Journal of Civil Engineering. Vol. 33, pp. 719-725. Lang, B., 1994. Span Design for Entry Type Excavations. M.A.Sc. Thesis, University of British Columbia, Vancouver, BC. Li, N., P. Zhang, Y. Chen, G. Swoboda. 2003. “Fatigue properties of cracked and saturated frozen sandstone samples under cyclic loading.” International Journal of Rock Mechanics and Mining Sciences. Vol 40. pp. 145-150. Lindblom, U. 1977. Experimental and analytical research related to LNG storage in rock caverns. Proc. 1st International Symposium on storage in excavated rock caverns, Stockholm., pp. 257–263. Marcellus, R.W. and D.N. Huff. 1991. “The role that fracture plays during ice failure.” Canadian Geotechnical Journal. Vol 28: pp 752-759. 204  Marinos, P and Hoek, E. 2000. GSI – A geologically friendly tool for rock mass strength estimation. Proc. GeoEng2000 Conference, Melbourne. Martin, C.D., P.K. Kaiser, D.R. McCreath. 1999. “Hoek-Brown parameters for predicting the depth of brittle failure around tunnels.” Canadian Geotechnical Journal. Vol 36: pp. 136-151. Mathis, J.I. and C.H. Page. 1995. “Drifting in very poor rock – experience and analysis.” 101st Annual Northwest Mining Association Convention, Spokane, Washington, December 6-8, 1995. Mawson, S. 2012a. Interoffice Memo: Comparison of RMR parameter ratings for frozen and unfrozen ground. Saskatoon: Cameco Corporation. August 23, 2012. Mawson, S. 2012b. Interoffice Memo: Comparison of RMR parameter ratings for frozen and unfrozen ground. Saskatoon: Cameco Corporation. November 5, 2012. MDH Engineered Solutions, 2008. “Draft – Cigar Lake Conceptual Geological Model, M1225190007.” Report prepared by MDH and Triple Point Consulting for Cameco Corporation. Issued February 2008. Mellor, M. and D.M. Cole. 1981. Deformation and Failure of Ice Under Constant Stress or Constant Strain Rate. Cold Regions Science and Technology. Volume 5, p. 201-219. Mellor, M. 1970. Phase composition of pore water in cold rocks. U.S. Army Corp. of Engineers. CRREL Research Report 292. Hanover, New Hampshire. Mellor, M. 1971. Strength and deformability of rocks at low temperatures. USA Cold Regions Research and Engineering Laboratory, Research Report 294. Mellor, M. 1973. Mechanical properties of rocks at low temperatures. In Permafrost: The North American Contribution to the 2nd International Conference on Permafrost, Yakutsk, 13-28 July, Washington D.C.: Natural Academy of Sciences. Miller, R.D. (1980). The adsorbed film controversy. Cold Regions Science and Technology, 3 (1), pp. 83-86. Milne, Hadjigeorgiou, Pakalnis, 1998. Rock mass characterization for underground hard rock mines. Tunnelling and Underground Space Technology. 13:4, 383-391. Milne, D. 2007. Problems with rock classification for empirical and numerical design. Proceedings of the International Workshop on Rock Mass Classification in Underground Mining. Information Circular 9498. pp. 111-118. Min, K., J. Rutqvist, C. Tsang, L. Jing. 2005. “Thermally induced mechanical and permeability changes around a nuclear waste repository – a far-field study based on equivalent properties determined by a discrete approach.” International Journal of Rock Mechanics and Mining Sciences. Vol 42. Pp 765-780. 205  Murphy, Patrick. 1999. “Freeze hole drilling issues at the McArthur River underground Uranium mine.” Thesis submitted to the department of mining engineering in accordance with the requirements for the degree of bachelor of science. Queen’s university. Myrvang, A.M. 1988. “Rock mechanics related to coal mining in permafrost on Spitzbergen.” 5th International Conference on Permafrost. Trondheim, Norway, August 1988. pp. 1435-1440. Nater, P., L.U. Arenson., S.M. Springman. 2008. “Choosing geotechnical parameters for slope stability assessments in alpine permafrost soils.” 9th international conference on permafrost. Pp 1261Nelson, M.G. 2001. “Underground Mining Frozen Placers” Underground Mining Method: Engineering Fundamentals and International Case Studies. Ed W.A Hustrlid and R.L. Bullock. SME. Ouchi, A.M., R. Pakalnis, and T.M. Brady. 2004. “Update of Span Design Curve for Weak Rock Masses.” Pakalnis, R., 2002. Empirical Design Methods – UBC Geomechanics Update. NARMS-TAC 2002. Pakanlis, R., 2012. Site Visit Exit Report, McArthur River Cameco, Technical Review. June 2012. Palmstrom, A. 1982. The Volumetric Joint Count – A Useful and Simple Measure of the Degree of Rock Jointing. Proc. 41st Int. Congress Int. Ass. Engng. Geol. Delphi. 5: 221-228. Park, C., J.H. Synn, H.S. Shin, D.S. Cheon, H.D. Lim, S.W. Jeon. 2004. “An experimental study on thermal characteristics of rock at low temperatures.” International Journal of Rock Mechanics and Mining Sciences. Vol 41. Pp 367-368. Petrovic, J.J. 2003. “Review mechanical properties of ice and snow.” Journal of Materials Science. Volume 38, Number 1, pp 1-6. Pettibone, H. 1973. “Stability of an underground room in frozen gravel.” Permafrost: North American contribution to the Second International Conference, 13-28 July 1973, Yakutsk, U.S.S.R. pp 699-706. Phukan, A. And S. Takasugi. 1982. “Excavation Resistance of Artificially Frozen Soils”. Proceedings of the 3rd International Symposium on Ground Freezing. 22-24 June 1982. Special Report 82-16. Pp. 35Portella. P., Annesley I.R., 2000. Paleoproterozoic tectonic evolution of the eastern subAthabasca basement, northern Saskatchewan: Integrated magnetic, gravity and geological data. Geol. Assoc. Can. – Mineral. Assoc. Can, Program Abstr. 206  Rhoads, J. L. (1994) Basic Explanation of Creep Processes. NE-161 Report. Department of Nuclear Engineering, University of California, Berkeley. Robertson, A. 1988. “Estimating weak rock strength.” AIME Annual General Meeting, Tucson, Arizona. Ruzicka, V., and LeCheminant, G.M. 1986. Developments in uranium geology in Canada, 1985. In Current research, part A. Geological Survey of Canada, Paper 86-l A, pp. 531 -540. Sammis, C.G. and R.L. Biegel. 2004. “Mechanics of strengthening crystalline rock at low temperatures: a preliminary assessment.” 26th Seismic Research Review – Trends in Nuclear Explosion Monitoring. pp. 475-485. Sanger, F.J. and F.H. Sayles. 1979. Thermal and rheological computations for artificially frozen ground construction. Engineering Geology 13, 311-337. Sayles, F. 1973. “Triaxial and Creep Tests on Frozen Ottawa Sand.” Permafrost: North American Contribution to the Second International Congress. Pp 384-391 Schmall, P.C., A.B. Corwin, and L.P. Spiteri. 2005a. “Ground freezing under the most adverse conditions: Moving groundwater.” Rapid Excavation and Tunnelling Conference Proceedings. Ed. J.D. Hutton and W.D. Rogstad. Pp 360-368. Schmall, P.C., A.B. Corwin, D. Maishman. 2005b. “Defining the Niche of Ground Freezing.” Rapid Excavation and Tunnelling Conference Proceedings. Ed. J.D. Hutton and W.D. Rogstad. Pp 996-1005. Schultz, M.S. and H. Hass. 2005. “Application of Ground Freezing for Underground Construction.” Rapid Excavation and Tunnelling Conference Proceedings. Ed. J.D. Hutton and W.D. Rogstad. pp 1062-1080. Schmitke, B.W. 2004. “Cigar Lake’s Jet Boring Mining Method.” World Nuclear Association Annual Symposium. September 8-10, 2004, London. Schulson, Erland. 1999. “The structure and mechanical behaviour of ice.” JOM, 51 (2) (1999), pp. 21-27. Seymour, J.B. D.R. Tesarik, and R.W. McKibbin. 1996. “Stability of Permafrost Gravels in an Alaskan Underground Placer Mine. Proceedings of the Fourth International Symposium on Mining in the Arctic, Longyearbyen, Svalbard, Norway, 27-30 July, 1996. pp 147-158. Sleptsov, A.E. 1992. “Basic tendencies of underground mining strategy development in the Arctic zone.” Proceedings of the 2nd International Symposium on Mining in the Arctic. Fairbanks, Alaska, July 1992. pp 97-101. Sofianos, A.I. 1996. “Analysis and Design of an Underground Hard Rock Voussoir Beam Roof.” International Journal of Rock Mechanics and Mineral Science. Vol 33, No. 2. pp. 153-166. 207  Thomson, S. And E.F. Lobacz. 1973. “Shear Strength at a Thaw Interface” Permafrost: North American Contribution to the Second International Congress. Pp 419Tsytovich, N.A. 1960. “Bases and Foundations on Frozen Soil.” Special Report 58, Highway Research Board. Udd, John E. And Marc C. Betournay. 1999. “Some ground control considerations for mining in permafrost.” Proceedings of the 5th International Symposium on Mining in the Arctic. Yellowknife, Canada. June 1998. pp 43-51. Ushakov, G.S. 1974. “Calculation of the long term strength of the roofs of rectangular rooms in perennially frozen ground.” Institute of Physicotechnical Problems of the North, Yakutsk. Translated from Fiziko-Tekhnicheskie Problemy Razrabotki Poleznykh Iskopaemykh, No. 4, pp. 34-40. Vyalov, S.S. 1962. The strength and creep of frozen soils and calculations for ice-soil retaining structures. U.S. Army, CRREL, Hanover, New Hampshire, Translation 76. Walder, J. and B. Hallet. 1985. “A theoretical model of the fracture of rock during freezing.” Geological Society of America Bulletin. Vol. 96. pp. 336-346. Wang, D., W. Ma, and Z. Wen. 2008. “The freezing process deformation of soil under high confining pressure.” Pp 1917-1920 Wardrop, 2005. Report on “Assessment of Permafrost Cases”. Issued to Bema Gold Corporation January 2005. Weerdenburg, P.C. and N.R. Morgenstern. 1984. “Underground Cavities in Ice-Rich Frozen Ground.” 4th International Conference on Permafrost. Washington, DC, July 1983. pp. 13841389. Wyllie and Mah. 2007. Rock Slope Engineering, Civil and Mining. 4th Ed. Spon Press, New York. Yang, Yugui, Gao, Feng, Lai, Yuanming. 2012. "Modified Hoek-Brown criterion for nonlinear strength of frozen soil". Cold Regions Science and Technology. doi: 10.1016/j.coldregions.2012.10.010 Yamabe, T. and K.M. Neaupane. 2001. “Determination of some thermo-mechanical properties of Sirahama sandstone under subzero temperature condition.” International Journal of Rock Mechanics and Mining Sciences. Vol 31. pp. 1029-1034. Youseff, H. And A. Hanna. 1988. “Behaviour of Frozen and Unfrozen Sands in Triaxial Testing.” Transportation Research Record 1190, TRB, National Research Council, Washington D.C., pp 57-64. Yuanlin, Z. And D. Carbee. 1985. Strain rate effect on the tensile strength of frozen silt. Fourth 208  International Symposium on Ground Freezing, Sapporo.  209  Appendix A: X-Ray Diffraction Testing  210  QUANTITATIVE PHASE ANALYSIS OF TWO POWDER SAMPLES USING THE RIETVELD METHOD AND X-RAY POWDER DIFFRACTION DATA.  Megan Roworth – Rimas Pakalnis Mining Engineering Dept. – UBC 5th Floor, 6350 Stores Road Vancouver, BC V6T 1Z4  Mati Raudsepp, Ph.D. Elisabetta Pani, Ph.D. Jenny Lai, B.Sc. Dept. of Earth & Ocean Sciences 6339 Stores Road The University of British Columbia Vancouver, BC V6T 1Z4  October 27, 2009  EXPERIMENTAL METHOD The core samples 18 and 19 were reduced to the optimum grain-size range for quantitative X-ray analysis (<10 μm) by grinding under ethanol in a vibratory McCrone Micronising Mill for 7 minutes. To avoid preferred orientation of the platy illite crystals, the ground samples were suspended in a 0.5% aqueous solution of polyvinyl alcohol (PVA) and sprayed from an airbrush into a heated chamber (150°C). As the spray falls in the heated chamber, spheres of randomly orientated crystals a few tens of micrometers in diameter are formed. Step-scan X-ray powder-diffraction data were collected over a range 3-80°2θ with CoKa radiation on a Bruker D8 Focus Bragg-Brentano diffractometer equipped with an Fe monochromator foil, 0.6 mm (0.3°) divergence slit, incident- and diffracted-beam Soller slits and a LynxEye detector. The long fine-focus Co X-ray tube was operated at 35 kV and 40 mA, using a take-off angle of 6°.  RESULTS The X-ray diffractograms were analyzed using the International Centre for Diffraction Database PDF-4 using Search-Match software by Siemens (Bruker). X-ray powder-diffraction data of the samples were refined with Rietveld program Topas 4 (Bruker AXS). The results of quantitative phase analysis by Rietveld refinements are given in Table 1. These amounts represent the relative amounts of crystalline phases normalized to 100%. The Rietveld refinement plots are shown in Figures 1–2. To avoid preferred orientation of the platy illite crystals, the ground samples were suspended in a 0.5% aqueous solution of polyvinyl alcohol (PVA) and sprayed from an airbrush into a heated chamber (150°C). As the spray falls in the heated chamber, spheres of randomly orientated crystals a few tens of micrometers in diameter are formed.  Table A.1. Results of quantitative phase analysis (wt.%) Mineral  Ideal Formula  Illite  K0.65Al2.0(Al0.65Si3.35O10)(OH)2  Kaolinite  Al2Si2O5(OH)4  3.0  Rutile?  TiO2  1.0  0.8  Alunite?  K2Al6(SO4)4(OH)12  0.7  0.5  Hematite  α-Fe2O3  Pyrite  FeS2  Total  18  19  95.3  82.9  13.4 2.4 100.0  100.0  RP-MR_SstCore-18_spray-D8.raw Illite 2M1 Kaolinite Rutile? Alunite?  8,000  7,000  95.32 % 2.96 % 0.99 % 0.74 %  6,000  5,000  4,000  3,000  2,000  1,000  0  -1,000  -2,000  -3,000 10  15  20  25  30  35  40  45  50  55  60  65  70  75  Figure A.1. Rietveld refinement plot of sample “18” (blue line - observed intensity at each step; red line - calculated pattern; solid grey line below – difference between observed and calculated intensities; vertical bars, positions of all Bragg reflections). Coloured lines are individual diffraction patterns of all phases.  80  RP-MR_SstCore-19_spray-D8.raw_1 Illite 2M1 Rutile? Hematite Pyrite Alunite?  9,000  8,000  82.94 % 0.78 % 13.38 % 2.41 % 0.49 %  7,000  6,000  5,000  4,000  3,000  2,000  1,000  0  -1,000  -2,000  10  15  20  25  30  35  40  45  50  55  60  65  70  75  Figure A.2. Rietveld refinement plot of sample “19” (blue line - observed intensity at each step; red line - calculated pattern; solid grey line below – difference between observed and calculated intensities; vertical bars, positions of all Bragg reflections). Coloured lines are individual diffraction patterns of all phases.  80  Appendix B: 2009 Unconfined Compressive Strength Testing  211  Sample ID:  After Sample Trimming  1  Borehole:  ST791-06  From (m):  432.25  To (m):  432.40  Test Date:  24-Jun-09  Tested by:  M. Roworth  Failure Mode: Geology:  After Failure                       Shear Hematized Clay R0.5  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -10  0.07  23.2  1.9  2.81  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 1352  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  60.80  2903.3  121.00  2.0  14.0  4.8  UCS (psi) 697.4  Unconfined Compressive Strength Test 6  Axial Stress (MPa)  5  4  3  2  1  0 0  1  2  3  4  5  Displacement (mm)  6  7  8  After Sample Trimming  Sample ID:  After Failure                       3  Borehole:  SF801-04  From (m):  435.15  To (m):  435.35  Test Date:  30-Jun-09  Tested by:  M. Roworth  Failure Mode:  Shear  Geology:  Hematized Clay R0  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -10  0.22  20.6  1.9  2.85  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 3540  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  80.34  5068.9  142.30  1.8  9.5  2.1  UCS (psi) 302.3  U Unconfined fi d Compressive C i Strength S h Test T 2.5  Axial Stress (MPa)  2  1.5  1  0.5  0 0  1  2  3  4  5  6  Displacement (mm)  7  8  9  Sample ID:  After Sample Trimming  4  Borehole:  SF801-04  From (m):  435.25  To (m):  435.45  Test Date:  01-Jul-09  Tested by:  M. Roworth  Failure Mode: Geology:  After Failure                       Shear Hematized Clay R0  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -10  0.02  20.7  1.9  3.01  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 1198  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  81.37  5200.2  136.28  1.7  6.1  1.3  UCS (psi) 192.7  Unconfined Compressive Strength Test 1.4  1.2  Axial Stress (MPa)  1  0.8  0.6  0.4  0.2  0 0  0.5  1  1.5  2  Displacement (mm)  2.5  3  3.5  Sample ID:  After Sample Trimming  5  Borehole:  SF801-04  From (m):  435.50  To (m):  435.70  Test Date:  02-Jul-09  Tested by:  M. Roworth  Failure Mode:  After Failure                       Shear  Geology:  Hematized Clay R0.5  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -10  0.08  15.9  2.1  3.09  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 2685  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  82.66  5366.4  152.68  1.8  29.8  6.5  UCS (psi) 948.6  Unconfined Compressive Strength Test 7  6  Axial Stress (MPa)  5  4  3  2  1  0 0  0.5  1  1.5  2  2.5  3  Displacement (mm)  3.5  4  4.5  Sample ID:  After Sample Trimming  6  Borehole:  ST786-07  From (m):  427.55  To (m):  427.75  Test Date:  02-Jul-09  Tested by:  M. Roworth  Failure Mode:  After Failure                       Shear  Geology:  Bleached sandstone R0  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -10  0.22  35.6  1.4  2.71  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 922  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  81.60  5229.2  154.00  1.9  9.7  2.1  UCS (psi) 308.0  U Unconfined fi d Compressive C i Strength St th Test T t 2.5  Axial Stress (MPa)  2  1.5  1  0.5  0 0  1  2  3  4  5  Displacement (mm)  6  7  8  Sample ID:  After Sample Trimming  7  Borehole:  ST786-07  From (m):  427.73  To (m):  427.93  Test Date:  03-Jul-09  Tested by:  M. Roworth  Failure Mode:  After Failure                       Shear  Geology:  Bleached sandstone R0.5  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -10  0.02  38.1  1.3  2.68  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 1158  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  83.20  5436.7  162.85  2.0  7.1  1.6  UCS (psi) 227.2  Unconfined Compressive Strength Test 1.8 1.6  Axial Stress (MPa)  1.4 1.2 1 0.8 0.6 0.4 0.2 0 0  1  2  3  4  5  6  Displacement (mm)  7  8  9  Sample ID:  After Sample Trimming  8  Borehole:  ST786-07  From (m):  424.90  To (m):  425.10  Test Date:  July 15,2009  Tested by:  M. Roworth  Failure Mode:  After Failure                       Shear  Geology:  Bleached sandstone R0.5  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -20  0.00  34.2  1.5  2.70  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 2346  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  82.90  5397.6  162.49  2.0  7.1  1.3  UCS (psi) 195.4  Unconfined Compressive Strength Test 1.6 1.4  Axial Stress (MPa)  1.2 1 0.8 0.6 0.4 0.2 0 0  0.2  0.4  0.6  0.8  1  1.2  Displacement (mm)  1.4  1.6  1.8  2  Sample ID:  After Sample Trimming  9  Borehole:  SF801-04  From (m):  428.76  To (m):  428.96  Did Not Fail  Test Date:  July 9 2009  Tested by:  M. Roworth  Failure Mode:  After Failure                       Shear  Geology:  Bleached sandstone R2  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -10  0.75  10.0  2.2  2.70  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 5946  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  82.06  5289.2  160.23  2.0  77.7  17.0  UCS (psi) 2,469.4  Unconfined Compressive Strength Test 18 16  Axial Stress (MPa)  14 12 10 8 6 4 2 0 0  0.5  1  1.5  2  2.5  3  Displacement (mm)  3.5  4  4.5  5  After Sample Trimming  Sample ID: Borehole:  SF801-04  From (m):  441.28  To (m):  441.48  Test Date:  06-Jul-09  Tested by:  M. Roworth  Failure Mode: Geology:  After Failure                       11  Shear altered GrMp R0.5  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -10  0.20  22.0  1.7  2.67  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 240  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  81.88  5265.6  154.34  1.9  12.8  2.8  UCS (psi) 406.1  Unconfined Compressive Strength Test 3  Axial Stress (MPa)  2.5  2  1.5  1  0.5  0 0  5  10  15  Displacement (mm)  20  25  Sample ID:  After Sample Trimming  12  Borehole:  SF801-04  From (m):  441.47  To (m):  441.67  Test Date:  07-Jul-09  Tested by:  M. Roworth  Failure Mode:  After Failure                       Shear  Geology:  altered GrMp R0.5  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -10  0.06  26.1  1.7  2.67  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 433  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  81.71  5244.2  158.37  1.9  15.4  3.4  UCS (psi) 490.6  Unconfined Compressive Strength Test 4 3.5  Axial Stress (MPa)  3 2.5 2 1.5 1 0.5 0 0  2  4  6  8  10  Displacement (mm)  12  14  16  Sample ID:  After Sample Trimming  13  Borehole:  SF801-04  From (m):  441.90  To (m):  442.10  Test Date:  July 9 2009  Tested by:  M. Roworth  Failure Mode:  After Failure                       Shear  Geology:  altered GrMp R2  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -10  0.93  15.8  1.8  2.64  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 5346  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  82.06  5288.7  165.49  2.0  36.3  8.0  UCS (psi) 1,154.9  U Unconfined fi d Compressive C i Strength S h Test T 9 8  Axial Stress (MPa)  7 6 5 4 3 2 1 0 0  1  2  3  4  5  Displacement (mm)  6  7  8  Sample ID:  After Sample Trimming  16  Borehole:  ST786-07  From (m):  426.90  To (m):  427.10  Test Date:  12-Jul-09  Tested by:  M. Roworth  Failure Mode: Geology:  After Failure                       Shear Bleached sandstone R0.5  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -20  0.16  33.2  1.6  2.71  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 1325  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  83.50  5476.4  162.82  1.9  20.4  4.5  UCS (psi) 649.8  Unconfined Compressive Strength Test 5 4.5 4  Axial Stress (MPa)  3.5 3 2.5 2 1.5 1 0.5 0 0  2  4  6  8  Displacement (mm)  10  12  14  Sample ID:  After Sample Trimming  17  Borehole:  ST786-07  From (m):  427.10  To (m):  427.30  Test Date:  12-Jul-09  Tested by:  M. Roworth  Failure Mode: Geology:  After Failure                       Shear Bleached sandstone R0.5  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -20  0.09  30.0  1.5  2.71  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 1872  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  84.95  5667.4  162.25  1.9  22.9  5.0  UCS (psi) 729.4  Unconfined Compressive Strength Test 6  Axial Stress (MPa)  5  4  3  2  1  0 0  2  4  6  8  Displacement (mm)  10  12  14  Sample ID:  After Sample Trimming  18  Borehole:  ST786-07  From (m):  427.30  To (m):  427.50  Test Date:  13-Jul-09  Tested by:  M. Roworth  Failure Mode: Geology:  After Failure                       Shear Bleached sandstone R0  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -20  0.02  43.0  1.3  2.68  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 3322  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  84.22  5570.8  163.44  1.9  16.7  3.7  UCS (psi) 532.3  Unconfined Compressive Strength Test 4 3.5  Axial Stress (MPa)  3 2.5 2 1.5 1 0.5 0 0  1  2  3  4  Displacement (mm)  5  6  7  Sample ID:  After Sample Trimming  19  Borehole:  SF801-04  From (m):  434.70  To (m):  434.90  Test Date:  13-Jul-09  Tested by:  M. Roworth  Failure Mode: Geology:  After Failure                       Shear Hematized Clay R0  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -20  0.25  22.8  1.8  3.01  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 2055  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  81.61  5230.5  161.97  2.0  15.5  3.4  UCS (psi) 492.0  Unconfined Compressive Strength Test 4 3.5  Axial Stress (MPa)  3 2.5 2 1.5 1 0.5 0 0  1  2  3  4  Displacement (mm)  5  6  7  Sample ID:  After Sample Trimming  20  Borehole:  SF801-04  From (m):  435.00  To (m):  435.20  Test Date:  14-Jul-09  Tested by:  M. Roworth  Failure Mode: Geology:  After Failure                       Shear Hematized Clay R0.5  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -20  0.05  20.9  1.9  3.01  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 1830  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  78.89  4888.4  162.16  2.1  19.0  4.2  UCS (psi) 603.1  Unconfined Compressive Strength Test 4.5 4  Axial Stress (MPa)  3.5 3 2.5 2 1.5 1 0.5 0 0  2  4  6  8  Displacement (mm)  10  12  14  Sample ID:  After Sample Trimming  22  Borehole:  SF801-04  From (m):  432.35  To (m):  432.55  Test Date:  15-Jul-09  Tested by:  M. Roworth  Failure Mode:  After Failure                       Shear  Geology:  Bleached sandstone R0.5  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -10  0.00  30.7  1.5  2.64  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 1195  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  84.68  5631.9  151.39  1.8  19.0  2.2  UCS (psi) 325.9  Unconfined Compressive Strength Test 2.5  Axial Stress (MPa)  2  1.5  1  0.5  0 0  1  2  3  4  5  6  Displacement (mm)  7  8  9  10  Sample ID:  After Sample Trimming  23  Borehole:  SF801-04  From (m):  432.55  To (m):  432.75  Test Date:  14-Jul-09  Tested by:  M. Roworth  Failure Mode: Geology:  After Failure                       Shear Bleached sandstone R0.5  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -10  0.00  30.9  1.5  2.70  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 968  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  84.46  5603.1  159.75  1.9  19.0  2.4  UCS (psi) 345.0  Unconfined Compressive Strength Test 2.5  Axial Stress (MPa)  2  1.5  1  0.5  0 0  1  2  3  4  Displacement (mm)  5  6  Sample ID:  After Sample Trimming  24  Borehole:  SF801-04  From (m):  432.75  To (m):  432.95  Test Date:  14-Jul-09  Tested by:  M. Roworth  Failure Mode: Geology:  After Failure                       Shear Hematized Clay R0.5  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -20  0.22  28.2  1.6  2.70  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 1845  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  84.82  5650.5  152.10  1.8  26.0  5.7  UCS (psi) 828.0  Unconfined Compressive Strength Test 6  Axial Stress (MPa)  5  4  3  2  1  0 0  2  4  6  8  Displacement (mm)  10  12  14  Sample ID:  After Sample Trimming  26  Borehole:  SF801-04  From (m):  442.85  To (m):  443.05  Test Date:  11-Jul-09  Tested by:  M. Roworth  Failure Mode:  After Failure                       0 tca  Geology:  altered GrMp R0.5  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -20  0.24  25.0  1.7  2.64  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 3217  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  81.85  5261.7  154.88  1.9  30.1  6.6  UCS (psi) 957.5  Unconfined Compressive Strength Test 7  6  Axial Stress (MPa)  5  4  3  2  1  0 0  1  2  3  4  5  6  Displacement (mm)  7  8  9  Sample ID:  After Sample Trimming  27  Borehole:  SF801-04  From (m):  443.05  To (m):  443.25  Test Date:  July 10 2009  Tested by:  M. Roworth  Failure Mode: Geology:  After Failure                       Shear, 58 deg tca altered GrMp R1  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -20  0.08  25.0  1.6  2.60  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 3862  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  83.57  5485.6  143.18  1.7  14.1  3.1  UCS (psi) 449.4  Unconfined Compressive Strength Test 3.5  3  Axial Stress (MPa)  2.5  2  1.5  1  0.5  0 0  1  2  3  4  Displacement (mm)  5  6  Sample ID:  After Sample Trimming  28  Borehole:  SF801-04  From (m):  443.20  To (m):  443.40  Test Date:  July 10 2009  Tested by:  M. Roworth  Failure Mode: Geology:  After Failure                       Shear, 50 deg tca altered GrMp R1  Temperature  Strain Rate  Moisture Content  (C)  (%/min)  %  -20  0.02  25.0  1.6  2.60  Diameter, (φ)  Area, (A)  Height, (h)  Ratio  Peak Load  2  Bulk Density  S.G.  E (MPa) 1332  σ  (mm)  (mm )  (mm)  h/φ  (kN)  (MPa)  84.22  5570.8  144.94  1.7  18.5  4.1  UCS (psi) 589.6  U Unconfined fi d Compressive C i Strength S h Test T 4.5 4  Axial Stress (MPa)  3.5 3 2.5 2 1.5 1 0.5 0 0  1  2  3  4  Displacement (mm)  5  6  Appendix C: Four Point Beam Testing C1 - Concrete C2 - Cigar Lake Drill Core  212  C1 - Concrete  213  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  25‐May‐09 1 1 1 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  2650 6 5 130 90  Beam Length Beam Diameter  325 mm 74.2 mm  Moisture Content  14.7 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 50/50 Sand/Concrete Joint  No Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  5.36 5 36 MPa 1.79  PHOTOGRAPHS Before Test  After Test No Photo  No Photo Failed 30 mm from center  Pressure vs Deflection Mid Span  5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  Deflection (mm)  6  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  25‐May‐09 1 1 2 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  2440 5 5 130 120  Beam Length Beam Diameter  314 mm 76.4 mm  Moisture Content  14.7 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 50/50 Sand/Concrete Joint  No Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  4.52 4 52 MPa 1.51  PHOTOGRAPHS Before Test  After Test No Photo  No Photo Failed 27 mm from center  Pressure vs Deflection Mid Span  5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  Deflection (mm)  6  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  25‐May‐09 1 1 3 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance  3350 7 6 130  Test Duration  Beam Length Beam Diameter  335 mm 74.1 mm  Moisture Content  14.7 %  Applied Strain Rate  kPa kN mm mm  Mix Design 50/50 Sand/Concrete Joint  No Modulus of Rupture, TMR  90 s  6.80 MPa  ~Tensile Strength  2.27  PHOTOGRAPHS Before Test  After Test No Photo  No Photo Failed 38 mm from center  Pressure vs Deflection Mid Span  5000  Pressure (kPa)  4000  3000  2000  1000  0 0  1  2  3  4  5  Deflection (mm)  6  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  25‐May‐09 1 1 4 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Test Duration  3950 8 4 130 90  Beam Length Beam Diameter  310 mm 76 mm  Moisture Content  14.7 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 50/50 Sand/Concrete Joint  No Modulus of Rupture, TMR  7.43 MPa 2.48  ~Tensile Strength  PHOTOGRAPHS Before Test  After Test No Photo  No Photo Failed 15 mm from center  Pressure vs Deflection Mid Span  5000  Pressure (kPa)  4000  3000  2000  1000  0 0  1  2  3  4  5  Deflection (mm)  6  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  25‐May‐09 1 2 5 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  560 1 2 160 30  Beam Length Beam Diameter  310 mm 74.5 mm  Moisture Content  17.5 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 50/50 Sand/Concrete Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  1.12 1 12 MPa 0.37  PHOTOGRAPHS Before Test  After Test  Failed at joint Pressure vs Deflection Mid Span  5000  Pressure (kPa) Press  4000  3000  2000  1000  0 0  1  2  3  4  5  Deflection (mm)  6  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Bottom Roller Span Top Roller Span  25‐May‐09 1 2 6 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  1830 4 8 130 90  Beam Length Beam Diameter  330 mm 75 mm  Moisture Content  17.5 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 50/50 Sand/Concrete Joint  No Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  3.58 3 58 MPa 1.19  PHOTOGRAPHS Before Test  After Test  Pressure vs Deflection Mid Span  5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  Deflection (mm)  6  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  25‐May‐09 1 2 7 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection  Beam Length Beam Diameter Moisture Content  17.5 %  Applied Strain Rate  2800 kPa 6 kN 6 mm  Crack Distance Crack Distance Test Duration  3150 mm 74.5 mm  Mix Design 50/50 Sand/Concrete Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  160 mm 90 s  5.59 5 59 MPa 1.86  PHOTOGRAPHS Before Test  After Test  Did not fail at joint  Pressure vs Deflection Mid Span  5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  Deflection (mm)  6  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  25‐May‐09 1 2 8 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance  990 2 2 130  Test Duration  Beam Length Beam Diameter  350 mm 74.5 mm  Moisture Content  17.5 %  Applied Strain Rate  kPa kN mm mm  Mix Design 50/50 Sand/Concrete Joint  Yes Modulus of Rupture, TMR Modulus of Rupture, T  90 s  1 98 MPa 1.98  ~Tensile Strength  0.66  PHOTOGRAPHS Before Test  After Test No Photo Failed at Joint  Pressure vs Deflection Mid Span  5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  Deflection (mm)  6  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  2‐Jun‐09 1 3 9 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  490 1 1 150 45  Beam Length Beam Diameter  315 mm 75.4 mm  Moisture Content  18.7 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 50/50 Sand/Cement Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  0.94 0 94 MPa 0.31  PHOTOGRAPHS Before Test  After Test  Failed at joint  Pressure vs Deflection Mid Span  5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  Deflection (mm)  6  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  18‐Jun‐09 1 5 14 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  1730 4 4 155 120  Beam Length Beam Diameter  310 mm 75.4 mm  Moisture Content  12.1 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 50/50 Sand/Concrete Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  3.33 3 33 MPa 1.11  PHOTOGRAPHS Before Test  After Test  Failed through joint  Pressure vs Deflection Mid Span  5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  Deflection (mm)  6  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  18‐Jun‐09 2 5 15 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  520 1 1.5 150 120  Beam Length Beam Diameter  300 mm 75.4 mm  Moisture Content  12.1 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 50/50 Sand/Concrete Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  1.00 1 00 MPa 0.33  PHOTOGRAPHS Before Test  After Test  Failed at joint  Pressure vs Deflection Mid Span  5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  Deflection (mm)  6  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  18‐Jun‐09 3 5 16 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  2320 5 4 150 210  Beam Length Beam Diameter  310 mm 75.4 mm  Moisture Content  12.1 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 50/50 Sand/Concrete Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  4.47 4 47 MPa 1.49  PHOTOGRAPHS Before Test  After Test  Failed at joint  Pressure vs Deflection Mid Span  5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  Deflection (mm)  6  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  18‐Jun‐09 3 5 17 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  680 1 1.5 170 45  Beam Length Beam Diameter  360 mm 75.4 mm  Moisture Content  12.1 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 50/50 Sand/Concrete Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  1.31 1 31 MPa 0.44  PHOTOGRAPHS Before Test  After Test  Failed at joint  Pressure vs Deflection Mid Span  5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  Deflection (mm)  6  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  18‐Jun‐09 1 6 18 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  880 2 1 155 120  Beam Length Beam Diameter Moisture Content  310 mm 75.4 mm #N/A  %  Applied Strain Rate  kPa kN mm mm s  Mix Design Concrete w/ Agg Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  1.69 1 69 MPa 0.56  PHOTOGRAPHS Before Test  After Test  Failed through joint  Pressure vs Deflection Mid Span  5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  Deflection (mm)  6  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  18‐Jun‐09 2 6 19 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  1560 3 3 150 120  Beam Length Beam Diameter Moisture Content  300 mm 75.4 mm #N/A  %  Applied Strain Rate  kPa kN mm mm s  Mix Design Concrete w/ Agg Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  3.00 3 00 MPa 1.00  PHOTOGRAPHS Before Test  After Test  FAILS THROUGH MID SECTION FAILS THROUGH MID SECTION DID NOT FAIL AT JOINT  Pressure vs Deflection Mid Span  5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  Deflection (mm)  6  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  29‐Jul‐09 1 7 22 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  1900 4 8.7 221 42  Beam Length Beam Diameter Moisture Content  294 mm 75.62 mm 10.8 %  Applied Strain Rate  kPa kN mm mm s  Mix Design Concrete w/ Agg Joint  No Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  3.63 3 63 MPa 1.21  PHOTOGRAPHS Before Test  After Test  Pressure essu e vs s Deflection e ect o Mid Span  5000  Pressure essure (kPa)  4000  3000  2000  1000  0 0  1  2  3  4  5  6  Deflection (mm)  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  29‐Jul‐09 2 7 23 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  800 2 7.5 155 15  Beam Length Beam Diameter Moisture Content  295 mm 76.62 mm 10.8 %  Applied Strain Rate  kPa kN mm mm s  Mix Design Concrete w/ Agg Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  1.47 1 47 MPa 0.49  PHOTOGRAPHS Before Test  After Test  Pressure essu e vs s Deflection e ect o Mid Span  5000  Pressure essure (kPa)  4000  3000  2000  1000  0 0  1  2  3  4  5  6  Deflection (mm)  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  29‐Jul‐09 3 7 24 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  800 2 7.5 185 34  Beam Length Beam Diameter Moisture Content  297 mm 75.42 mm 10.8 %  Applied Strain Rate  kPa kN mm mm s  Mix Design Concrete w/ Agg Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  1.54 1 54 MPa 0.51  PHOTOGRAPHS Before Test  After Test  FAILS THROUGH MID SECTION FAILS THROUGH MID SECTION  Pressure essu e vs s Deflection e ect o Mid Span  5000  Pressure essure (kPa)  4000  3000  2000  1000  0 0  1  2  3  4  5  6  Deflection (mm)  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  31‐Jul‐09 3 8 27 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  1100 2 7.7 195 17  Beam Length Beam Diameter Moisture Content  347 mm 76.36 mm 13.9 %  Applied Strain Rate  kPa kN mm mm s  Mix Design Concrete w/ Agg Joint  no Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  2.04 2 04 MPa 0.68  PHOTOGRAPHS Before Test  After Test  Pressure essu e vs s Deflection e ect o Mid Span  5000  Pressure essure (kPa)  4000  3000  2000  1000  0 0  1  2  3  4  5  6  Deflection (mm)  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  10‐Aug‐09 1 9 29 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  700 1 7.6 135 12  Beam Length Beam Diameter Moisture Content  254 mm 73.98 mm 28.8 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 33/66 Sand/Concrete Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  1.43 1 43 MPa 0.48  PHOTOGRAPHS Before Test  After Test  Pressure essu e vs s Deflection e ect o Mid Span  5000  Pressure essure (kPa)  4000  3000  2000  1000  0 0  1  2  3  4  5  6  Deflection (mm)  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  10‐Aug‐09 2 9 30 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  900 2 7.7 161 19  Beam Length Beam Diameter Moisture Content  255 mm 73.07 mm 28.8 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 33/66 Sand/Concrete Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  1.90 1 90 MPa 0.63  PHOTOGRAPHS Before Test  After Test  Pressure essu e vs s Deflection e ect o Mid Span  5000  Pressure essure (kPa)  4000  3000  2000  1000  0 0  1  2  3  4  5  6  Deflection (mm)  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  10‐Aug‐09 4 9 32 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  1000 2 7.6 136 14  Beam Length Beam Diameter Moisture Content  295 mm 75.78 mm 28.8 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 33/66 Sand/Concrete Joint  No Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  1.90 1 90 MPa 0.63  PHOTOGRAPHS Before Test  After Test  FAILS THROUGH MID SECTION FAILS THROUGH MID SECTION  Pressure essu e vs s Deflection e ect o Mid Span  5000  Pressure essure (kPa)  4000  3000  2000  1000  0 0  1  2  3  4  5  6  Deflection (mm)  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  13‐Aug‐09 1 10 33 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  800 2 7.7 125 30  Beam Length Beam Diameter Moisture Content  285 mm 72.3 mm 18.5 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 40/60 Sand/Concrete Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  1.75 1 75 MPa 0.58  PHOTOGRAPHS Before Test  After Test  Pressure essu e vs s Deflection e ect o Mid Span  5000  Pressure essure (kPa)  4000  3000  2000  1000  0 0  1  2  3  4  5  6  Deflection (mm)  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  13‐Aug‐09 2 10 34 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  1300 3 8 174 55  Beam Length Beam Diameter Moisture Content  299 mm 76.99 mm 18.5 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 40/60 Sand/Concrete Joint  no Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  2.35 2 35 MPa 0.78  PHOTOGRAPHS Before Test  After Test  Pressure essu e vs s Deflection e ect o Mid Span  5000  Pressure essure (kPa)  4000  3000  2000  1000  0 0  1  2  3  4  5  6  Deflection (mm)  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  13‐Aug‐09 3 10 35 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  1300 3 7.9 125 65  Beam Length Beam Diameter Moisture Content  289 mm 73.35 mm 18.5 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 40/60 Sand/Concrete Joint  yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  2.72 2 72 MPa 0.91  PHOTOGRAPHS Before Test  After Test  Pressure essu e vs s Deflection e ect o Mid Span  5000  Pressure essure (kPa)  4000  3000  2000  1000  0 0  1  2  3  4  5  6  Deflection (mm)  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  13‐Aug‐09 4 10 36 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  1500 3 7.9 145 85  Beam Length Beam Diameter Moisture Content  262 mm 76.84 mm 18.5 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 40/60 Sand/Concrete Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  2.73 2 73 MPa 0.91  PHOTOGRAPHS Before Test  After Test  Pressure essu e vs s Deflection e ect o Mid Span  5000  Pressure essure (kPa)  4000  3000  2000  1000  0 0  1  2  3  4  5  6  Deflection (mm)  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  28‐Aug‐09 1 11 37 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  1900 4 7.5 155 135  Beam Length Beam Diameter Moisture Content  262 mm 73.02 mm 16.7 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 50/50 Sand/Concrete Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  4.03 4 03 MPa 1.34  PHOTOGRAPHS Before Test  After Test  Pressure essu e vs s Deflection e ect o Mid Span  5000  Pressure essure (kPa)  4000  3000  2000  1000  0 0  1  2  3  4  5  6  Deflection (mm)  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  28‐Aug‐09 2 11 38 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  1600 3 7.5 120 105  Beam Length Beam Diameter Moisture Content  315 mm 74.12 mm 16.7 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 50/50 Sand/Concrete Joint  No Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  3.24 3 24 MPa 1.08  PHOTOGRAPHS Before Test  After Test  Pressure essu e vs s Deflection e ect o Mid Span  5000  Pressure essure (kPa)  4000  3000  2000  1000  0 0  1  2  3  4  5  6  Deflection (mm)  7  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID Batch Test # Top Roller Span Bottom Roller Span  28‐Aug‐09 3 11 39 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection Crack Distance Crack Distance Test Duration  1100 2 8 115 80  Beam Length Beam Diameter Moisture Content  257 mm 75.44 mm 16.7 %  Applied Strain Rate  kPa kN mm mm s  Mix Design 50/50 Sand/Concrete Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  2.12 2 12 MPa 0.71  PHOTOGRAPHS Before Test  After Test  Pressure essu e vs s Deflection e ect o Mid Span  5000  Pressure essure (kPa)  4000  3000  2000  1000  0 0  1  2  3  4  5  6  Deflection (mm)  7  8  9  10  C2 - Cigar Lake Drill Core  214  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID  20‐Nov‐09 1  Beam Length Beam Diameter  Test # Top Roller Span Bottom Roller Span  1 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection C k Di t Crack Distance Test Duration  680 1 1 130 90  310 mm 85 mm  Moisture Content  %  Applied Strain Rate  kPa kN mm mm s  Joint  No Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  0.91 0 91 MP MPa 0.30  PHOTOGRAPHS Before Test  After Test No Photo  No Photo  Pressure vs Deflection 5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  6  7  Mid Span Deflection (mm)  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID  21‐Nov‐09 2  Beam Length Beam Diameter  Test # Top Roller Span Bottom Roller Span  2 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection C k Di t Crack Distance Test Duration  970 2 5 130 90  310 mm 85 mm  Moisture Content  11.9 %  Applied Strain Rate  kPa kN mm mm s  Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  1.30 1 30 MP MPa 0.43  PHOTOGRAPHS Before Test  After Test No Photo  No Photo  Pressure vs Deflection 5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  6  7  Mid Span Deflection (mm)  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID  21‐Nov‐09 2  Test # Top Roller Span Bottom Roller Span  Beam Length Beam Diameter  3 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection C k Di t Crack Distance Test Duration  1090 2 4 130 55  310 mm 85 mm  Moisture Content  11.9 %  Applied Strain Rate  kPa kN mm mm s  Joint  No Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  1.47 1 47 MP MPa 0.49  PHOTOGRAPHS Before Test  After Test No Photo  No Photo  Pressure vs Deflection 5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  6  7  Mid Span Deflection (mm)  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID  22‐Nov‐09 3  Test # Top Roller Span Bottom Roller Span  Beam Length Beam Diameter  4 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection C k Di t Crack Distance Test Duration  1690 4 4.3 130 90  310 mm 85 mm  Moisture Content  28.7 %  Applied Strain Rate  kPa kN mm mm s  Joint  No Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  2.27 2 27 MP MPa 0.76  PHOTOGRAPHS Before Test  After Test No Photo  No Photo  Pressure vs Deflection 5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  6  7  Mid Span Deflection (mm)  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID  22‐Nov‐09 3  Beam Length Beam Diameter  Test # Top Roller Span Bottom Roller Span  5 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection C k Di t Crack Distance Test Duration  0 0 0 130 90  310 mm 85 mm  Moisture Content  28.7 %  Applied Strain Rate  kPa kN mm mm s  Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  0.00 0 00 MP MPa 0.00  PHOTOGRAPHS Before Test  After Test No Photo  FAILED ON HANDLING  Pressure vs Deflection 5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  6  7  Mid Span Deflection (mm)  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID  23‐Nov‐09 4  Beam Length Beam Diameter  Test # Top Roller Span Bottom Roller Span  6 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection C k Di t Crack Distance Test Duration  760 2 2.7 140 60  310 mm 85 mm  Moisture Content  35.5 %  Applied Strain Rate  kPa kN mm mm s  Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  1.02 1 02 MP MPa 0.34  PHOTOGRAPHS Before Test  After Test No Photo  No Photo  Pressure vs Deflection 5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  6  7  Mid Span Deflection (mm)  8  9  10  FOUR‐POINT BEAM BENDING TEST Test Date Sample ID  24‐Nov‐09 5  Beam Length Beam Diameter  Test # Top Roller Span Bottom Roller Span  7 75 mm 229 mm  Peak Pressure Peak Force Mid Span Deflection C k Di t Crack Distance Test Duration  0 0 0 130 90  310 mm 85 mm  Moisture Content  17.9 %  Applied Strain Rate  kPa kN mm mm s  Joint  Yes Modulus of Rupture, T Modulus of Rupture TMR ~Tensile Strength  0.00 0 00 MP MPa 0.00  PHOTOGRAPHS Before Test  After Test No Photo  FAILED ON HANDLING  Pressure vs Deflection 5000  Pressure (kPa) Pressu  4000  3000  2000  1000  0 0  1  2  3  4  5  6  7  Mid Span Deflection (mm)  8  9  10  Appendix D: Direct Shear Testing  215  DIRECT SHEAR TEST - Breaking Strength  Sample 1 Sample 2 Sample 3 Sample 4 Sample 5  Borehole SF791-06 SF801-04 SF801-04 SF801-04 SF796-05  Depth 429.5 431.2 433.5 431.4 432.05  Description Bleached Bleached Hematized w joint Bleached Hematized  Peak (kPa) 15320 14780 6990 14950 14160  Normal (kg) 25 5 5 45 25  Shear Stress (kPa) 1.69 1.67 0.75 1.77 1.56  Sample 1  Sample 2  Sample 3  Sample 4  Sample 5  UBC Geomechanics Lab  Normal Stress (kPa) 0.46 0.12 0.12 0.85 0.46  Moisture 34.00 11.95 28.74 35.46 17.93  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0073990/manifest

Comment

Related Items