INVESTIGATION INTO THE TIME DEPENDENT DEFORMATION BEHAVIOUR AND FAILURE MECHANISMS OF UNSUPPORTED ROCK SLOPES BASED ON THE INTERPRETATION OF OBSERVED DEFORMATION BEHAVIOUR Kenneth George Mercer Doctor of Philosophy A thesis submitted to the Faculty of Engineering and the Built Environment, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Doctor of Philosophy. Johannesburg, 2006 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Declaration DECLARATION I declare that this thesis is my own, unaided work. It is being submitted for the Degree of Doctor of Philosophy in the University of the Witwatersrand, Johannesburg. It has not been submitted before for any other degree or examination in any other University. _________ _ __ _ _ __ _ _ __ _ __ _ _ __ _ _ __ _ __ _ _ __ _ _ __ _ _ December 2006 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes A bstract ABSTRACT Deformation has long been recognized as one of the most important parameters which define the state of rock masses. Deformational behaviour of excavated slopes and failures can be extremely complicated and up to now there have been no methods or models that have adequately addressed the range of behaviour that is possible during ex cavation of rock slopes in different geological environments. The principal objectives of this research project were: 1. To expand the general knowledge base of ti me and event dependent deformation behaviour in excavated unsupported rock slopes in different geological terrains. 2. To establish typical patterns of deformation behaviour and identify trigger mechanisms for different modes of failure in the context of different slope geometries, geology, structure and rates of mining. 3. To develop where possible a more fundamental characterisation of the different components of time dependent deformation, 4. Develop a predictive method whereby deformation can be forecast. The achievements of this research are significant. They include; 1. The collection of twelve international detailed mining case studies represents a comprehensive, independently assembled large open pit mining database. 2. A new Time and Event Dependent Deformati on Model has been developed which describes how deformation behaviour of excavated rock mass may be presented using deformation pathways. The model accommodates five principal stages of deformation ranging from primary and secondary rock mass creep modes through the onset-of-failure to collapse and post collapse or post mining recovery deformation behaviour. A very significant feature of the Model is the provision of changing deformation rate decay functions as a slope progresses towards failure. 3. A new statistical method has been developed to use in conjunction with the Model to enable forecasting of deformation behaviour in order to make predictions such as the time to collapse. The new method offers many advantages over existi ng empirical and semi-empirical methods. This research project has significantly expanded the frontiers of the knowledge of time and event dependent deformational behaviour of rock slopes an d has provided a Model for the interpretation and a new method for the forecasting of deformational behaviour. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Dedication Sovereign Lord, this research is dedicated to you, ?the maker of heav en and earth? (Acts 4:24) PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Acknowledgments Page 1 of 2 ACKNOWLEDGEMENTS This thesis was made possible by cont ributions from many different people. I would first like to thank my wife Bridget, Jaden, Mom and Dad, Stan and Brenda and the rest of my family for their continued support and encouragement through the many long and difficult months that it took to collect the data and write up this thesis. You guys made it all possible. With regard technical assistance, I would like to thank the following people for their time and effort they have made available to assist with this research in so many different ways; Anglo Gold Ashanti Mr Gareth Taylor, Manager Mining, East and West Africa Region Mr Frik Badenhorst, Manager Operations, Navachab Mine, Namibia Mr Tim Williams, Geotechnical Superintendent, Geita Mine, Tanzania Mr Tim Botha, Chief Surveyor Anglo American Platinum : Potgiete rs rust Platinum Mine, South Africa Mr Alan Bye, Mr Desmond Mossop, Chief Rock Engineer Mr Johan Scheepers, Chief Surveyor Ms Megan Little, Geotechnical Engineer Ms Kelly Lachenicht, Geotechnical Engineer BHPbilliton : Mt Keith Nickel Operations (Mt Keith Open Pit) Mr David Goodchild, Geomechanical Superintendent Mr Jason Summerville, Juni or Engineering Geologist BHPbilliton : Leinster Nickel Op erations (Harmon y Open Pit) Ms Jacqui Cahill, Geomenchanics Engineer Coldelco North : Chuquicamata Mine, Chile Mr Alex Calderon Rojo, Superintendente De Ing. Geotecnica DeBeers/Debsw ana : Orapa, Letlhaka ne and Venetia Mines Dr Alan Guest, Group Consultant - Geotechnical Mr George Kayesa, Senior Geotechnical Engineer Mr Felix Ramsden, Mine Geologist ? Geotechnical Mr Jan Janse van Rensburg, Mine Geologist ? Geotechnical Mr Dudley Mulville, Seni or Geotechnical Engineer Mrs Linda Hannweg, Geotechnical Engineer GroundProbe Mr David Noon, General Manager Mr Keith Rowley, Business Manager Africa Geomos/Leica Africa Mr Colin Thomson, Product Manager: Mining & Monitoring Systems PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Acknowledgments Page 2 of 2 Harmony (Kalgold Operat ion), South Africa Mr Branko Pieterse, Surveyor Mr Christo Pothas, Mining Manager Mr Dirk Venter, Technical Coordinator Kalgoorlie Consolidated Gold Mines (KCGM) Mr Adam Beer, Senior Geotechnical Engineer Maptec Africa, Vulcan Licence Mr Barry Venter, Technical Manager Rio Tinto : Palabora Mining Company, South Africa Mr David Pretorious, Technical Manager Mr Michael van der Heever, Specialist Geologist SRK Consulting Dr Oskar Steffen, Consultant - retired Mr Peter Terbrugge, Director Mr Alan Naismith, Principal Mining Engineer Mr Neil Marshall, Principal Geotechnical Engineer Ms Ren? Roux, Principal Engineering Geologist Universit y of The Witw at ersrand Professor T.R. Stacey, Ph D Supervisor (Department of Mining Engineering) Professor D. Mason (Department of Mathematics) Other Dr Horst Marker, retired PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Table of Contents Page 1 of 7 TABLE OF CONTENTS Declaration Abstract Dedication Acknowledgements Contents List of Figures List of Tables List of Symbols Definition of Terms 1 INTRODUCTION ................................................................................................................ .......... 1-1 1.1 Background ............................................................................................................... ............1.2 1.2 Objectives ............................................................................................................... ........... 1-2 1.3 Research Approach .............................................................................................................. 1-2 1.4 Literature Review and Development of a Literature Review Case Study Database ............ 1-3 1.5 Definition of the Research Topic Based on Findings of the Li terature Review.................... 1-3 1.6 Detailed Mining Operat ion Case St udies. ............................................................................. 1-4 1.7 Evaluation and Interpretation of Data Collected from Detailed Case Studies and Presentat ion of Findings ................................................................................... 1-4 1.8 Final Comments ............................................................................................................ ........ 1-4 2 REVIEW OF LITERATURE RELEVANT TO ROCK SLOPE DEFORMATIONAL BEHAVIOUR.. 2-1 2.1 Introduction 2-1 2.2 Definitions of Failure ............................................................................................................. 2-1 2.3 Deformation Behav iour Termi nology .................................................................................... 2-2 2.4 Nature of Intact Rock ..................................................................................................... ....... 2-2 2.5 Creep Behaviour of Intact Rock ............................................................................................ 2-2 2.6 Time Dependent Deformation Behaviour of Ro ck Join ts...................................................... 2-3 2.7 General Descriptions of the Defo rmation Behaviour of Rock Ma ss...................................... 2-5 2.8 Rock Mass Strain Hardening and So ftening......................................................................... 2-6 2.9 General Components of Rock Mass Deformation Result ing from Exca vation ..................... 2-6 2.9.1 Vertical (Upwards/Positiv e) Deformation and Rebound (Swelling) Resulting Fr om Unloading .......................................................................................... 2-6 2.9.2 Initial Horizontal and Vertical (Downwards) Deformation (Relaxation) Resulting fr om Excavation .......................................................................................... 2-8 2.9.3 Long Term Horizontal and Vertical (Downwards) Creep Deformation (Relaxation) Result ing from Exca vation...................................................................... 2-8 2.10 Classification of Inst ability and Failure Mechanisms ............................................................ 2-9 2.11 Modes and Mechanism s of Instab ility ................................................................................. 2-10 2.11.1 Evaluation of Modes of Inst ability ........................................................................ 2-11 2.11.2 Planar Instability Modes....................................................................................... 2-11 2.11.3 Toppling Instability Modes ................................................................................... 2-13 2.11.4 Wedge Instability Modes...................................................................................... 2-17 2.11.5 Rotational/Circular or Rock Mass Inst ability Modes ............................................ 2-19 2.11.6 Higher Order or Compound Instab ility Modes ..................................................... 2-21 2.12 Gravitational Creeping Behaviour of Large Scale Rock Mass (R ock Mass Creep)............ 2-21 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Table of Contents Page 2 of 7 2.13 Rock Mass Time to Failure Predictions .............................................................................. 2-25 2.14 Review of Displacement Monitoring Te chniques................................................................ 2-27 2.15 Summary of the Literature Review...................................................................................... 2-2 8 3 LITERATURE REVIEW CA SE STUDY DATA BASE ..................................................................... 3 - 1 3.1 Case Study 1, Brenda Open Pit, Canada, Northw all Failure ................................................ 3-2 3.2 Case Study 2, Cassiar Mine, C anada, North East Se ctor Failure ........................................ 3-5 3.3 Case Study 3, Luscar Mine, Pit 51-B- 2 Northwall Failure, Alberta C anada ....................... 3-10 3.4 Case Study 4, Telfer Gold Mine, Australia, Pit 1A Highwall Fa ilure ................................... 3-12 3.5 Case Study 5, Aft on Mine, Canada, Wall Inst abilities and Failures.................................... 3-15 3.6 Case Study 6, Jinchuan Mine, China, Southwest Slope Failu re......................................... 3-22 3.7 Case Study 7, Steep Rock (Hogarth) Mine , Canada, No 1. Zone Highwall Failure ........... 3-27 3.8 Case Study 8, Inspiration?s Mines, Arizo na, USA, General Instabilities and Failures ........ 3-30 3.9 Case Study 9, Nchanga Open Pit, Zambia, July 2004 Fa ilure ........................................... 3-34 3.10 Summary of the Literatu re Review Case Studies ............................................................... 3-40 4 DEFINITION OF RESEARCH TOPIC BASED ON FINDINGS OF LITERATURE REVIEW ......... 4-1 4.1 Specific Aspects of Time Dependent Deformatio n to be Addressed.................................... 4-1 4.1.1 Time Dependent Change In The Deformation Rate De cay Function......................... 4-1 4.1.2 The Usage of Acceleration to Predict the Onset Of Collapse Po int. .......................... 4-2 4.1.3 Time-to-colla pse Forecasts ........................................................................................ 4-2 4.2 Detailed Res earch Approach ................................................................................................ 4-2 4.2.1 Identification of Requi red Data for Co llection ............................................................. 4-5 4.2.2 Selection of De tailed Case Studies ............................................................................ 4-5 4.3 Detailed Case Study Data Presentat ion, Evaluation and In terpretation ............................... 4-9 4.4 Evaluation and Interpreta tion of Data Collected ................................................................... 4-9 4.5 Presentati on of Findings ..................................................................................................... 4-10 4.6 Discussion ............................................................................................................... ......... 4-10 5 CASE STUDY 1, NAVACHAB GOLD MINE, NAMIBIA................................................................. 5-1 5.1 Intr oduction .......................................................................................................................... 5-1 5.2 Regional Geology.................................................................................................................. 5-1 5.3 Strati graphy .......................................................................................................................... 5-1 5.4 Structur al Geology ................................................................................................................ 5-2 5.5 Geologic al Section ........................................................................................................ ........ 5-4 5.6 Rock Mass Classification and Geotechnical Characteristics ................................................ 5-4 5.7 Pit Configur ation Parameters................................................................................................ 5-5 5.8 Displacement Moni toring Systems........................................................................................ 5-5 5.9 History of Failure/s ...................................................................................................... .......... 5-6 5.10 Displacement Behav iour of Slopes leading up to March 2001 Fa ilure ................................. 5-6 5.11 Interpretation of Displacement B ehaviour, Failure Mechanism and Tri gger......................... 5-6 5.11.1 Horizontal Displacements ........................................................................................... 5-8 5.11.2 Vertical Displacements ............................................................................................. 5-1 0 5.12 Displacement B ehaviour of Non-failed Pit Wall Sect ors ..................................................... 5-10 5.12.1 East Slope............................................................................................................ ..... 5-11 5.12.2 West Slope............................................................................................................ .... 5-12 5.13 Discussion .............................................................................................................. .......... 5-12 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Table of Contents Page 3 of 7 6 CASE STUDY 2, BIBIANI GOLD MINE, GHANA .......................................................................... 6-1 6.1 Introdu ction .......................................................................................................................... 6-1 6.2 Regional Geology.................................................................................................................. 6-1 6.3 Structur al Geology ................................................................................................................ 6-1 6.3.1 Shearing and Faulting................................................................................................. 6-1 6.3.2 Folia tions/Bedding ..................................................................................................... . 6-4 6.3.3 Jo int Sets ............................................................................................................. ....... 6-4 6.4 Weatheri ng Profile................................................................................................................. 6-4 6.5 Grou ndwater ......................................................................................................................... 6-5 6.6 Geological Sections ....................................................................................................... ....... 6-5 6.7 Rock Classifica tion and Strengths ........................................................................................ 6 -5 6.8 Slope Configurat ion Parameters........................................................................................... 6-6 6.9 Displacement Moni toring Systems........................................................................................ 6-6 6.10 History of Failure/s ..................................................................................................... ........... 6-7 6.10.1 West Slope Failures................................................................................................... . 6-7 6.10.2 East Slope Failu res/Instabilit ies.................................................................................. 6-9 6.11 Deformation B ehaviour of Slopes ......................................................................................... 6 -9 6.12 Interpretation of Displacement B ehaviour, Failure mechanism and Tri gger....................... 6-10 6.13 Discussion .............................................................................................................. .......... 6-11 7 CASE STUDY 3, MT KEITH OPEN PIT, WESTERN AUSTRALIA ............................................... 7-1 7.1 Introduction 7-1 7.2 Regional Geology.................................................................................................................. 7-1 7.3 Stratigraphy and Weathering ................................................................................................ 7-1 7.3.1 Footwall S equence (East)........................................................................................... 7-3 7.3.2 Mt Keith Ultram afic Complex (MKUC) ........................................................................ 7-4 7.3.3 Hangingwall Pyritic Ch ert and Black Graphi tic Slat e .................................................. 7-4 7.3.4 Hangingwall Mafics and Felsics.................................................................................. 7-4 7.3.5 Cliffs Ultramafic........................................................................................................... 7-5 7.4 Structur al Geology ................................................................................................................ 7-5 7.4.1 1st Order Structures (Regional Stru ctures) ................................................................ 7-5 7.4.2 2nd Order St ructures (Folds, Major Shear Zones and Major Faults) ......................... 7-6 7.4.3 3rd Order Structures - Minor Fault and Joint Se ts...................................................... 7-7 7.4.4 4th Order Struct ures - Foli ation .................................................................................. 7-7 7.5 Representat ive Sections ................................................................................................... .... 7-8 7.6 Rock Mass Classifica tion and St rengths............................................................................... 7-9 7.7 Pit Configur ation Parameters................................................................................................ 7-9 7.8 Displacement Moni toring Systems........................................................................................ 7-9 7.9 Displacement Behaviour N on-failed Pit Wall Sectors ........................................................... 7-9 7.10 History of Failure/s ..................................................................................................... ........... 7-9 7.11 Discussion on the Deformation B ehaviour of the Mt Keith Failures.................................... 7-10 7.11.1 Phase 1 : Init iation of Inst ability ........................................................................... 7-10 7.11.2 Phase 2 : Pro pagation of Inst ability ..................................................................... 7-11 7.11.3 Phase 4 : Collaps e............................................................................................... 7-12 7.11.4 Phase 5 : Post Co llapse Rock Mass Behaviour .................................................. 7-12 7.11.5 Furthe r Discussion............................................................................................... 7-13 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Table of Contents Page 4 of 7 8 CASE STUDY 4, HARMONY OPEN PIT, LEINSTER NICKEL OPERATIONS, WESTERN AUSTRALIA .................................................................................................................. ........ 8-1 8.1 Intr oduction .......................................................................................................................... 8-1 8.2 Regional Geology.................................................................................................................. 8-1 8.3 Lithology and Weathering ..................................................................................................... 8-1 8.4 Grou nd water ........................................................................................................................ 8-3 8.5 Representat ive Sections ................................................................................................... .... 8-3 8.6 Structur al Geology ................................................................................................................ 8-3 8.7 Rock Classifica tion and Strengths ........................................................................................ 8 -5 8.8 Pit Configur ation Parameters................................................................................................ 8-5 8.9 Displacement Moni toring Systems........................................................................................ 8-5 8.10 Displacement Behaviour Non-failed Pi t Wall Sect ors ........................................................... 8-5 8.11 History of Failure/s ..................................................................................................... ........... 8-6 8.11.1 West Wall............................................................................................................ ... 8-7 8.11.2 East Wall............................................................................................................ .... 8-9 8.12 Discussion and Conclusions ............................................................................................... 8-12 9 DETAILED CASE STUDY 5, VENETIA, SOUTH AFRICA ............................................................ 9-1 9.1 Introduction 9-1 9.2 Regional Geology.................................................................................................................. 9-1 9.3 Lithology and Weathering ..................................................................................................... 9-2 9.4 Structur al Geology ................................................................................................................ 9-2 9.5 Grou ndwater ......................................................................................................................... 9-4 9.6 Representat ive Sections ................................................................................................... .... 9-4 9.7 Rock Classifica tion and Strengths ........................................................................................ 9 -4 9.8 Pit Configur ation Parameters................................................................................................ 9-4 9.9 Displacement Moni toring Systems........................................................................................ 9-6 9.10 Displacement Behaviour Non-failed Pi t Wall Sect ors ........................................................... 9-6 9.11 History of Failures ................................................................................................................. 9-6 9.11.1 South-ea st Slope Fa ilure ....................................................................................... 9-6 9.11.2 March 2000 No rth Slope Fa ilure............................................................................ 9-8 9.12 Discussion .............................................................................................................. ............ 9-8 10 CASE STUDY 6, ORAPA AND LETL HAKANE OPEN PITS, BOTSWA NA ............................... 1 0 - 1 10.1 Introduction ............................................................................................................ ............ 10-1 10.2 Regional Geology................................................................................................................ 10-1 10.3 Kimberlite Lithology and Weathering Profile ....................................................................... 10-2 10.4 Structural Geology Orapa ................................................................................................. .. 10-4 10.4.1 Basalt ............................................................................................................... .... 10-4 10.4.2 Kimberlite............................................................................................................. 10-4 10.5 Structural Geology Letlhakane............................................................................................ 10-6 10.6 Groundwater .............................................................................................................. ......... 10-7 10.7 Characterisation of the Country Ro ck Mass ....................................................................... 10-7 10.8 Representative Sections .................................................................................................. ... 10-8 10.9 Pit Configur ation Parameters.............................................................................................. 10-8 10.10 Displacement M onitoring Sy stems.................................................................................... 10-10 10.11 Displacement B ehaviour of Non-failed Pi t Wall Sector s ................................................... 10-11 10.12 History of Failures ............................................................................................................. 10-11 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Table of Contents Page 5 of 7 10.12.1 Orapa................................................................................................................ . 10-11 10.12.2 Letlhakane ......................................................................................................... 10 -12 10.13 Discussion ............................................................................................................. ......... 10-13 11 DEVELOPMENT OF A GENERALI SED TIME AND EVENT DEPENDENT DEFORMATION BEHAVIOUR MODEL .................................................................................... 11-1 1 1.1 Introduction 11-1 11.2 Development of the Time and Event Dependent Deformation Model ................................ 11-2 11.2.1 Deformation Behav iour Termi nology .................................................................................. 11-2 11.3 Generalised Time and Event Depen dent Rock Mass Defo rmation Model ......................... 11-3 11.3.1 Category 1, Pre-onset-of -failure Deformat ion Behaviour .................................... 11-3 11.3.2 Category 2, Pre-onset-of -failure Deformat ion Behaviour .................................... 11-3 11.3.3 Deformation Stages of th e Model ........................................................................ 11-4 11.3.4 Basic Defo rmation Behaviour .............................................................................. 11-4 11.3.5 Typical Sequentially Mined Deformation Behaviour for Rock Slopes Leading to Collaps e................................................................................. 11-8 11.3.6 Typical Sequentially Mined Deformation Behaviour for Rock Slopes with Partial Colla pse of the Slope...................................................................... 11-10 11.3.7 Typical Sequentially Mined Ho rizontal Deformation Behaviour for Non-Collaps ed Slopes ....................................................................................... 11-12 11.3.8 Typical Sequentially Mined Vertical Deformation Behaviour for Non-Collaps ed Slopes ....................................................................................... 11-12 11.3.9 Modes of Possible Deformation B ehaviour for Horizontal Displacements ........ 11-16 11.4 Su mmary ...................................................................................................................... 11-18 12 APPLICATION OF THE TIME AND EVENT DEPENDENT DEFORMATION MODEL USING CASE STUDY DATA EXAMPLES ............................................................................................... 12-1 12.1 Introduction ............................................................................................................ ............ 12-1 12.2 Discussion on Stage 1 : Prim ary Rock Mass Creep Mode ................................................. 12-1 12.3 Discussion on Stage 2 : Seco ndary Rock Mass Creep M ode ............................................ 12-4 12.3.1 Category 1 Pre-onset-of-f ailure Deformation Behaviour .......................................... 12-4 12.3.2 Category 2 Pre-onset-of-f ailure Deformation Behaviour ........................................ 12-12 12.4 Discussion on Stage 3, Onset- of-failure to Collapse ........................................................ 12-12 12.5 Discussion on Stage 4, Post Colla pse .............................................................................. 12-12 12.5.1 Disintegrat ion (S4?type 1) ................................................................................. 12-12 12.5.2 Post Collapse Recovery (S4?type 2 to S 4?type 6) .......................................... 12-16 12.6 Discussion on Stage 5, Post Mining/Re covery ................................................................. 12-19 12.7 Vertic al Rebound......................................................................................................... ...... 12-19 12.8 Su mmary ...................................................................................................................... 12-22 1 3 FORECASTING OF DE FORMATION BEHAVIOUR USING THE MODEL................................. 13-1 13.1 Introduction ............................................................................................................ ............ 13-1 13.2 The Problem........................................................................................................................ 13-1 13.3 A Proposed Method of Forecast ing Deformation Behaviour .............................................. 13-3 13.4 Derivation of Equations of Displacement Rate and Veloci ty............................................... 13-4 13.4.1 Derivation Usi ng Natural Logar ithms................................................................... 13-4 13.4.2 Derivation Using Log arithms to the Base 10 ....................................................... 13-5 13.5 Categories of Failures ......................................................................................................... 13-6 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Table of Contents Page 6 of 7 13.6 Forecasting of Collapse Times fo r Stage 3 Deformat ion Behaviour................................... 13-7 13.6.1 Identification of t he Onset-of-failu re Point ........................................................... 13-7 13.6.2 Cu rve Fitting ........................................................................................................ 13-8 13.6.3 Selection of Limiting Deformat ion Rate at the Poin t of Collapse......................... 13-8 13.6.4 Di fferentiation....................................................................................................... 13-8 13.6.5 Selection of Appr opriate Deformation Data......................................................... 13-8 13.7 An Example Illustrating the Use of this Forecasting Method for St age 3 Deformation....... 13-9 13.7.1 Exampl e Selectio n ............................................................................................... 13-9 13.7.2 Analysis and Resu lts ........................................................................................... 13-9 13.7.3 Discussion on Example ..................................................................................... 13-15 13.8 Forecasting of Deformati on Behaviour for Stage 2........................................................... 13-16 13.9 An Example Illustrating the Use of this Forecasting Method for Sta ge 2 Deformation..... 13-16 13.10 Furthe r Validat ion.............................................................................................................. 13-17 13.11 Discussion on Fo recasting Method................................................................................... 13-19 13.12 Future Development.......................................................................................................... 13-23 14 CONCLUSIONS ................................................................................................................ ........ 14-1 14.1 General review of Time Dependent Deformation ............................................................... 14-1 14.2 Literature Review Case Study Fi ndings.............................................................................. 14-2 14.3 Detailed Case Study Findings............................................................................................. 14-3 14.4 Development and Application of the Time and Event Dependent Deformation Model ...... 14-4 14.5 Forecasting of De formation Behaviour ............................................................................... 14-4 14.6 Concludi ng Comm ents........................................................................................................ 14-5 15 RECOMMENDATIONS FOR FURTHER RESEARCH................................................................. 15.1 16 REFERE NCES ................................................................................................................ ........ 16.1 17 BIBLIOGRAPHY ............................................................................................................... ......... 17.1 APPENDICES Appendix 1, Case Study 1 (Section 5) Mine/Location: Navachab Gold Mine, Namibia Appendix 2, Case Study 2 (Section 6) Mine/Location: Bibiani Gold Mine, Ghana Appendix 3, Case Study 3 (Section 7) Mine/Location: Mt Keith, Western Australia Appendix 4, Case Study 4 (Section 8) Mine/Location: Leinster Nickel Operations, Western Australia Appendix 5, Case Study 5 (Section 9) Mine/Location: Venetia Mine, South Africa PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Table of Contents Page 7 of 7 Appendix 6, Case Study 6 (Section 10) Mine/Location: Orapa/Letlhakane Mine, Botswana PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes List of Figures Page 1 of 7 LIST OF FIGURES Section 2 Figures 2.1 Idealised Joint Shear Beha viour (Amadei and Curr an,1980) ..........................................2.5 Figures 2.2 Time Dependency of Shear Stress - Idealised Shear Deformation Relationship for a Rock Joint (Amadei and Curran,1980) .............................................................................2.5 Figure 2.3 Typical Displacement Behaviour as Determined by Broadbent and Ko (1980)...................2.7 Figure 2.4 Characteristic Time Depe ndent Behaviour (M artin, 1993)...................................................2.7 Figure 2.5 Rock Failure Types Based on Structure/ slope Characteristics and Associated Typical Displacement Curves (Bro adbent and Zavodni , 1981) ........................................................2.9 Figure 2.6 Classification of Slope Failure Mechanisms (Martin a nd Kaiser, 1984) .......................... 2.10 Figure 2.7 Stereogram Illustrating Kine matic Conditions for Plane Failure ....................................... 2.12 Figure 2.8 Planar Failure Mechani sms (Piteau and Ma rtin, 1982 )..................................................... 2.14 Figure 2.9a to 2.9f Failure Mechanisms for Thin Sl abs on a Slope (Piteau and Martin, 1982) ...................... 2.15 Figure 2.10a to 2.10c Primary Toppling Mec hanisms (Goodman and Bray, 1 976)............................................. 2.16 Figure 2.11a to 2.11d Secondary Toppling Mech anisms (Goodman an d Bray, 1976) ........................................ 2.16 Figure 2.12 Stereogram Illustrating Kine matic Conditions for Toppling Fa ilure................................... 2.17 Figure 2.13 Illustration of We dge Failure Geometry in 3D (Piteau and Ma rtin, 1982) ......................... 2.18 Figure 2.14 Stereogram Illustrating Kine matic Conditions for Wedge Failure ..................................... 2.18 Figure 2.15a Rotation al Shear Failure ................................................................................................... 2.20 Figure 2.15b Shear Failure Along an Irregular Surface ......................................................................... 2.20 Figure 2.15c General Mec hanisms of Rotational Ro ck Mass Fail ure.................................................... 2.20 Figures 2.16a To 2.16i Creep Deformation in Rock Slopes (Ter-Stepanian, 1966) ........................................... 2.23 Figure 2.17a to 2.17b Large Scale Deformation in Rock Slopes (Zis chinsky,1966 ) ........................................... 2.24 Figure 2.18 Typical Displacement Rate vs Ti me Record of a Large-scale Rock Failure Proceeding to Collapse (Zav odni and Broadbent, 1980).................................................. 2.26 Section 3 Figure 3.1a Brenda Mine, Northwall Failure, Angle and Distances to Target Reflectors, Summer 1978 (Blackwe ll and Calder , 1981) .......................................................................3 .3 Figure 3.1b Brenda Mine, Northwall Failure, Cross Section Through Failure (Calder and Blac kwell, 1980) ...............................................................................................3.4 Figure 3.1c Brenda Mine, Northwall Failure, Relative Movements of Blocks (Calder and Blac kwell, 1980) ...............................................................................................3.4 Figure 3.2a Cassiar Open Pit Failure, Stereogra phic projection of average orientation of hanging wall structures (Martin, 1993) .................................................................................3.5 Figure 3.2b Cassiar Open Pit Failure, Relative movement Versus Time for Prism 24 (Martin and Mehr, 1993)....................................................................................................... 3.7 Figure 3.2c Cassiar Open Pit Failure, G eometry and Mechanism (Martin, 1993) ..................................3.8 Figure 3.2d Cassiar Open Pit Failure, Pit Geometry and Prism Vector Movements (Martin an d Mehr, 1993) .....................................................................................................3 .9 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes List of Figures Page 2 of 7 Figure 3.3a Luscar Mine 51-B-2 Pit Northwall Failure , Nov 1979, Section through Northwall Failure (Cruden and Maso umzadeh, 1987) .................................................................................. 3.11 Figure 3.3b Luscar Mine 51-B-2 Pit Northwall Failu re, Nov 1979, Time Series Plot for Prism 26B (Johnson, 1982) ............................................................................................................... . 3.11 Figure 3.4a Telfer Gold Mine, Pit 1A Highwa ll Failure, Typical geological cross section (Dancker t, 1994) .............................................................................................................. . 3.13 Figure 3.4b Telfer Gold Mine, Pit 1A Highwa ll Failure, Section through July 1989 Failure (Dancker t, 1994) .............................................................................................................. . 3.14 Figure 3.4c Highmont Mine, Southeast wall Failure , Prism dispacements prior to October 1992 failure (Dan ckert, 1994)..................................................................................................... 3 .14 Figure 3.5a Afton Mine, Geology and Position of Failures (Stewart and Re id, 1988) ......................... 3.17 Figure 3.5b Afton Mine, Northwest Sector Slide - Rate of Movement Plot (Stewart and Re id, 1988) .................................................................................................. 3.17 Figure 3.5c Afton Mine, Northwest Sect or Slide - Piezometric Levels (Stewart and Re id, 1988) .................................................................................................. 3.18 Figure 3.5d Afton Mine, Northwest Sect or Slide - Section Through Failure (Stewart and Re id, 1988) .................................................................................................. 3.18 Figure 3.5e Afton Mine, Ci rcular Failure - Secti on Through Failure (Stewart and Reid, 1988) ......... 3.19 Figure 3.5f Afton Mine, Toppling Failure - Location Map showing Crack Pattern and Monitor Positions (Modified after Reid and Stewart, 1986) ............................................. 3.19 Figure 3.5g Afton Mine, Toppling Failure - Pris m P32 Vertical Displacement and Location (Stewart and Re id, 1988) .................................................................................................. 3.20 Figure 3.5h Afton Mine, Toppling Failure - Prism P30 Vertical Displacement (Stewart and Re id, 1988) .................................................................................................. 3.20 Figure 3.5i Afton Mine, Toppling Failure - Degree of Activity within Instability (Reid and Stew art, 1986) .................................................................................................. 3.21 Figure 3.5j Afton Mine, Toppling Failure - Cumulative Displacement for Prisms P1, p30 and P40 (Reid and Stew art, 1986) .................................................................................................. 3.21 Figure 3.5k Afton Mine, Toppling Failure - Rate of Displacement for Prisms P1, p30 and P40 (Reid and Stew art, 1986) .................................................................................................. 3.21 Figure 3.5l Afton Mine, Toppling Failure - Cumu lative Displacement for Extensometers E24 and E25 (Reid and Stewart, 1986).................................................................................... 3.21 Figure 3.6a Jinchuan Mine South West Slope Failure,Slope Deformation Mechanism 1st Stage (Sijing, 1980) ................................................................................................................. ..... 3.23 Figure 3.6b Jinchuan Mine South West Slope Failure, Slope Deformation Mechanism 2nd Stage (Sijing, 1980) ................................................................................................................. ..... 3.23 Figure 3.6c Jinchuan Mine South West Slope Failu re, Slope Deformation Mechanism Tertiary Stage (Sijing, 1980) ................................................................................................................. ..... 3.23 Figure 3.6d Jinchuan Mine Sout h West Slope Failure, Displacem ent of Slope (Sijing,1980) ............. 3.24 Figure 3.6e Jinchuan Mine South West Slope Fa ilure, Slip Displacement of Fault F14 (Sijing, 1980) ................................................................................................................. ..... 3.24 Figure 3.6f Jinchuan Mine South West Slope Failure, Isoclines of Displacement Rate and Direction of Horizontal Displacement, March to May 1975 (Sijin g,1980) ........................ 3.25 Figure 3.6g Jinchuan Mine South West, Slope Failure Decelerated Displacement of Slope (Sijing, 1980) ................................................................................................................. ..... 3.25 Figure 3.6h Jinchuan Mine South West, Slope Fa ilure, Accelerated Displacement of Slope, Feb to Oct 1976 (Sijing,1980 ) ........................................................................................... 3.25 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes List of Figures Page 3 of 7 Figure 3.6i Jinchuan Mine South West Slope Failure, Isoclines of Displacement Rate and Direction of Horiz ontal Displacement, March to May 1975 .............................................. 3.26 Figure 3.7a Steep Rock Mine, Hogarth Pit Failure, Schematic Sketch through Failure Illustrating Relative Displacements (Modified afte r Brawner, Stacey and Stark, 1976) ..................... 3.28 Figure 3.7b Steep Rock Mine, Hogarth Pit Failure, Extensometer Displacements till 10th June 1975 (Brawner and St acey, 1979 ) ............................................................................................. 3.29 Figure 3.7c Steep Rock Mine, Hogarth Pit Failu re, Vector Displacements till 10th June 1975 (Brawner and St acey, 1979 ) ............................................................................................. 3.29 Figure 3.8a Inspiration Mines, Thornton No rthwest Sector, Cross Section Through Slope (Savely and Kast ner, 1981) .............................................................................................. 3.31 Figure 3.8b Inspiration Mines, Thornton Northwes t Sector, Stereographic Projections of the Structures (Savely and Kastner, 1 981) ............................................................................. 3.32 Figure 3.8c Inspiration Mines, Thornton Northw est Sector , Relationship between Rainfall, Tons Mined and Extensometer Moveme nt (Savely and Ka stner, 1981) ......................... 3.33 Figure 3.9a Nchanga Mine, July 2004 Failure , Geological Cross Section through Slope (Naismith and We ssels, 2005) .......................................................................................... 3.36 Figure 3.9b Nchanga Mine, July 2004 Failure, North Wall Monitoring Points (Naismith and We ssels, 2005) .......................................................................................... 3.36 Figure 3.9c Nchanga Mine, July 2004 Failure, Pris m Displacements from 10/12/2003 to 7/7/2004 (Naismith and We ssels, 2005) .......................................................................................... 3.37 Figure 3.9d Nchanga Mine, July 2004 Failure, Displacement Rates from 10/12/2003 to 19/5/2004 (Naismith and We ssels, 2005) .......................................................................................... 3.37 Figure 3.9e Nchanga Mine, July 2004 Failure, Radar Image of Wall and Thermal Movement Image (Naismith and We ssels, 2005) .......................................................................................... 3.38 Figure 3.9f Nchanga Mine, July 2004 Failure , Displacement Rates on Section 21E (Naismith and We ssels, 2005) .......................................................................................... 3.38 Figure 3.9g Nchanga Mine, July 2004 Failure, Displacement Rates and Accumulated Displacement for 4 days Prior to Colla pse (Naismith and We ssels, 2005) ...................... 3.39 Section 4 Figure 4.1 Characteristic Time Dependent B ehaviour (Modified afte r Martin, 1 993) ...........................4.3 Figure 4.2 Postulated Change in the Deformat ion Rate Decay Function for Increasing Slope Hei ght (h i) ............................................................................................................. .....4.4 Figure 4.3 Postulated Change in the Co nstant ?A? with Slope Hei ght ..................................................4.4 Figure 4.4 Postulated Change in the Co nstant ?b? with Slope Hei ght...................................................4,4 Section 5 Figure 5.1 General Layout of Pit and Li thology Contacts (April 20 05)..................................................5.2 Figure 5.2 Geological Cross Section............................................................................................ .........5.4 Figure 5.3 March 2001 Failure, Accumulat ed 2D Horizontal Di splacements .......................................5.7 Figure 5.4 March 2001 Failure, 2D Horizontal Velocities......................................................................5. 7 Figure 5.5 March 2001 Failure, Accumula ted Vertical Displacements .................................................5.8 Figure 5.6 March 2001 Failure, Horizontal Vector Movements in Time Peri od to 1 Feb 2001 .............5.9 Figure 5.7 March 2001 Failure, Horizontal Vect or Movements during Progressive Behaviour Time Period 1 Feb 2001 to 8 March 2001 (C ollapse)..........................................................5.9 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes List of Figures Page 4 of 7 Section 6 Figure 6.1 Bibiani Open Pit, Location of Mine and Regional Geology (Minproc Engineers, 1995) ......................................................................................................................... ...........6.2 Figure 6.2 Bibiani Open Pit, Vulcan Model, Orthogonal View of the Open Pit Showing Reef and Dyke System............................................................................................................... ..6.3 Figure 6.3 Bibiani Open Pit , Section 487 North East, (Modified after Australian Mining Consultant s, 1997) ............................................................................................................ ...6.6 Section 7 Figure 7.1 Structural Map of the Mt Keith O pen Pit at the 379RL Level (Haywood, 2004) ..................7.2 Figure 7.2 Figure 7.2: Stereonet Pr ojections of Foliation Poles in Major Litho-structural Domains Hayward, 2004) ................................................................................................................ ....7.8 Figure 7.3 Mt Keith Open Pit, Vulcan M odel, East-West Cros s Sectio n...............................................7.8 Figure 7.4 Failure #4 : 2004-08-11, Pre-collapse Ra dar Monitoring Data Slowing Deformation Propagation on Displacement Time Gr aph....................................................................... 7.11 Figure 7.5 Progressive Dama ge Evolution in Mt Keith Fresh Rock Fa ilures ..................................... 7.13 Section 8 Figure 8.1 Harmony Pit , Regional Geology, (Western Mining Cor poration Resources Limited, 1999) ................................................................................................................ ......8.2 Figure 8.2 Harmony Pit , Local Area Geology, (Western Mining Corporation Resources Limited, 1999) ................................................................................................................ ......8.2 Figure 8.3 Harmony Pit, Geological and Pit Cr oss Section, (Western Mining Corporation Resources Limi ted, 1999) ....................................................................................................8. 4 Figure 8.4 Harmony Pit, Equal Area Stereog raphic Projection of Structures .......................................8.4 Figure 8.5 Harmony Pit, Illustration of the West Wall Failure Mechanism for the 1C and Komatsu Failures .................................................................................................................8.9 Figure 8.6 Illustration of the Ea st Wall Failure Mechanism................................................................ 8.12 Section 9 Figure 9.1 Venetia Open Pit, Regional Geology and Structures, (Modified after Basson and Barnett, 2005) ............................................................................................................ ...9.3 Figure 9.2 Venetia Open Pit , North-South Vert ical Section Through the Pit Illustrating the Geology..........................................................................................................................9.5 Section 10 Figure 10.1 Orapa and Letlhakane Open Pits, Regio nal Geology , (Modified after Smith, 1984) ....... 10.3 Figure 10.2 Orapa and Letlhakane Open Pits , Gener alised Summary of the Lithostratigraphy of the Host Roc ks (Basson, 2006) .................................................................................... 10.3 Figure 10.3 Orapa Open Pit, Structural Interpretati on of Total Magnetic Im age (Basson, 2006) ........ 10.5 Figure A10.4 Orapa Open Pit, North - S outh Section Profile for September 2006 (Orapa Geotec Dept, 2006) .............................................................................................. 10.9 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes List of Figures Page 5 of 7 Figure A10.5 Letlhakane Open Pit, North - S outh Section Profile for September 2006 (Letlhakane Geot ec Dept, 2006)....................................................................................... 10.9 Section 11 Figure 11.1 Generalised Time and Event Dependent Rock Mass Deformation Model An Illustrative Deformation Pattern fo r Horizontal Displace ment Behaviour .................... 11.5 Figure 11.2 Generalised Time and Event Dependent Rock Mass Deformation Model An Illustrative Deformation Pattern for Ho rizontal Displacement Rate Behaviour............ 11.6 Figure 11.3 Generalised Time and Event Dependent Rock Mass Deformation Model Typical Horizontal and Vertical Deform ation Behaviour Patterns for Sequentially Mined Rock Slopes Leadi ng to Colla pse .......................................................................... 11.9 Figure 11.4 Generalised Time and Event, Dependent Rock Mass Deformation Model - Typical Horizontal and Vertical Deformation Behaviour Patterns for Sequentially Mined Rock Slopes Leading to Partial Collapse of t he Slope ............................................................. 11.11 Figure 11.5 Generalised Time and Event Dependent Rock Mass Deformation Model - Typical Horizontal Deformation Behaviour fo r Sequentially Mined Non-Collapsed Rock Slopes ........................................................................................................................ ..... 11.13 Figure 11.6 Generalised Time and Event Dependent Rock Mass Deformation Model - Typical Vertical Deformation Behaviour for Sequent ially Mined Non-Collapsed Rock Slopes ... 11.14 Figure 11.7 Generalised Time and Event Dependent Rock Mass Deformation Model, Illustrative Vertical Deformation Behaviour for Instantaneous Mined Non-Collapsed Rock Slopes ................................................................................................................... . 11.15 Figure 11.8 Generalised Time and Event Dependent Rock Mass Deformation Model, Illustrative Vertical Deformation Behaviour for Rapi dly Mined Non-Collapsed Rock Slopes........... 11.15 Figure 11.9 Generalised Time and Event Dependent Rock Mass Deformation Model, - Modes of Possible Deformation Behaviour for Horizontal Di splacement ....................................... 11.17 Section 12 Figure 12.1a Generalised Time and Event Dependent Rock Mass Deformation Model Idealised S1-type2 Mode of Deformation, Horizontal Displace ment Behaviour ............... 12.2 Figure 12.1b Generalised Time and Event Dependent Rock Mass Deformation Model Idealised S1-type2 Mode of Deformation, Hori zontal Displacement Rate Behaviour ...... 12.2 Figure 12.2a Navachab Open Pit , Prism G-C9 Stage 1, S1-type1 Mode of Deformation, Horizontal Displace ment Behaviour ................ 12.3 Figure 12.2b Navachab Open Pit , Prism G-C26 Stage 1, S1-type2 Mode of Deformation, Horizontal Displace ment Behaviour ................ 12.3 Figure 12.2c Navachab Open Pit , Prism G-P229P Stage 1, S1-type2 Mode of Deformation, Horizontal Displace ment Behaviour ................ 12.3 Figure 12.3a Generalised Time and Event Dependent Rock Mass Deformation Model Stage 2, S2-type1 Mode of Deformation, Horizontal Displace ment Behaviour ................ 12.5 Figure 12.3b Generalised Time and Event Dependent Rock Mass Deformation Model Stage 2, S2-type1 Mode of Deformation, Hori zontal Displacement Rate Behaviour ....... 12.5 Figure 12.4a: Generalised Time and Event Dependent Rock Mass Deformation Model Stage 2, S2-type2 Mode of Deformation, Horizontal Displace ment Behaviour ................ 12.6 Figure 12.4b Generalised Time and Event Dependent Rock Mass Deformation Model Stage 2, S2-type2 Mode of Deformation, Hori zontal Displacement Rate Behaviour ....... 12.6 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes List of Figures Page 6 of 7 Figure 12.5a Generalised Time and Event Dependent Rock Mass Deformation Model Stage 2, S2-type3 Mode of Deformation, Horizontal Displace ment Behaviour ................ 12.7 Figure 12.5b Generalised Time and Event Dependent Rock Mass Deformation Model Stage 2, S2-type3 Mode of Deformation, Displacement Ra te Behaviour......................... 12.7 Figure 12.6 Illustration of Stage 2 Mode of Defo rmation Behaviour (Broad bent and Ko, 1980) ......... 12.8 Figure 12.7 Illustration of Stage 2 Mode of Deformation Behaviour Jinchuan Mine South West Slope Failu re (Modified afte r Sijing,198 0)............................. 12.8 Figure 12-8 Nchanga Mine, July 2004 Failure Stage 1,2 and 3 Mode of Deformation Behaviour from 10/12/2003 to 19/5/2004 (Modified after Naismith and Wessels, 2005) ................................................................... 12.9 Figure 12-9 Harmony Open Pit East Wall Failure, Stage 2, S2-type2 Mode of Horizont al Deformation Behaviour..................................... 12.10 Figure 12.10 Venetia Open Pit July 2003 Failure Stage 2, S2-type3 Mode of Deformation Behaviour ....................................................... 12.11 Figure 12.11 Harmony Pit, 1C Failure, October 2001 Illustration of Category 2, Stage 2 Mode of Deformation Behaviour .............................. 12.13 Figure 12.12 Harmony Pit, Komatsu Failure, February 2003 Illustration of Category 2, Stage 2 Mode of Deformation Behaviour .............................. 12.13 Figure 12.13 Letlhakane 2005 Failure Category 2, Stage 2 Mode of Deformation Behaviour .................................................... 12.13 Figure 12.14 Bibiani West Wall October 2002 Failure Category 2, Stage 2 Mode of Deformation Behaviour .................................................... 12.15 Figure 12.15 Mt Keith, 2001-Dec to 2003-Nov SE Wall Failure 544RL Stage 4, S4-type 2 Post Collapse Hori zontal Prism Deformation Behaviour (MKO Geotec Dept, 2006) .............................................................................................. 12.17 Figure 12.16 Harmony Pit, March 2006 East Wall Failure Stage 4, S4-type 3 Post Collapse Horiz ontal Prism Deformation Behaviour ................. 12.17 Figure 12.17 Venetia Open Pit, 11 Septem ber 2004 Extension to SE Slope Failure Stage 4, S4-type 2 Transitioning into S4-type 6 Post Collapse Horizontal Prism Deformation Behaviour ......................................................................................... 12.18 Figure 12.18 F Stage Southeast Failure Evolution Stage 4, Complex S4 Post Collapse Horizontal Prism Deformation Behaviour (MKO Geotec Dept, 2006) .............................................................................................. 12.18 Figure 12.19a Generalised Time and Event Dependent Rock Mass Deformation Model Stage 5, S5-type 1or S5-type 2 Modes of Horizontal Deformation Behaviour ............... 12.20 Figure 12.19b Generalised Time and Event Dependent Rock Mass Deformation Model Stage 5, S5-type 1or S5-type 2 Modes of Ho rizontal Deformation Rate Behaviour....... 12.20 Figure 12.20 Harmony Pit Prism Survey 2D Movement Magnitude, Zone 5 April 2004 to March 2006 Stage 5, S5-type 1 Post Mining Horiz ontal Prism Deformation Behaviour..................... 12.21 Figure 12.21 Navachab Open Pit East Wall Selected Prism Survey Vertical Movement Magnitude August 1999 to March 2006 ..... 12.21 Figure 12.22 Navachab Open Pit Prism Survey Vertical Movement Ma gnitude April 2005 to March 2006 ........................ 12.24 Figure 12.23 Harmony Open Pit West Wall Selected Prism Survey Vertical Movement Magnitude April 2004 to February 2006c ... 12.24 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes List of Figures Page 7 of 7 Section 13 Figure 13.1 Harmony Pit, East Wall Failure, Deformation Rate for Prism 422 .................................... 13.2 Figure 13.2 Harmony Pit, East Wall Failure Pris m Survey 2D Movement Magnitude, Selected Zone 2 Prisms ................................................................................................................... 13.2 Figure 13.3 Sequential Curve Fitting for Displace ment Magnitudes, Harmony Pit, East Wall Failure, Prism 422 ........................................................................................................... 1 3.11 Figure 13.4 Displacement Rate Forecasts to 1m/day, Harmony Pit, East Wall Failure, Prism 422 ..................................................................................................................... ... 13.11 Figure 13.5 Sequential Curve Fitting for Displa cement Magnitudes, Harmony Pit, East Wall Failure, Prism 75 ..................................................................................................... 13. 12 Figure 13.6 Displacement Rate Forecasts to 1m/ day, Harmony Pit, East Wall Failure, Prism 75 .... 13.12 Figure 13.7 Harmony Pit, East Wall Failure, Prog ressive Forecasts for Date of Collapse ................ 13.13 Figure 13.8 Harmony Pit, East Wall Failure, Collapse Forecast at 14/12/ 2005................................. 13.13 Figure 13.9 Harmony Pit, East Wall Failure, Collapse Forecast at 11/1/ 2006................................... 13.14 Figure 13.10 Harmony Pit, East Wall Failur e, Collapse Foreca st at 23/ 1/2006................................... 13.14 Figure 13.11 Harmony Pit, East Wall Failure, Acceleration Forecast for Prism 422 ........................... 13.15 Figure 13.12 Harmony Pit, East Wall Failur e, Stage 2 Forecasts for Prism 422 ................................. 13.17 Figure 13.13 Mt Keith #2 Failure, Time, Forecast Time for Date of Collapse...................................... 13.18 Figure 13.14 Letlhakane July 2005 Failure, Time, Forecast Time for Da te of Collapse...................... 13.18 Figure 13.15a Mt Keith #2 Failure, Curve Fit for Deformation Data Using Order 4 Log Polynomial Fit.................................................................................................................. 13.21 Figure 13.15b Mt Keith #2 Failure, Residuals Using Order 4 Log Po lynomial Fit ................................. 13.21 Figure 13.15c Mt Keith #2 Failure, Error B ounds for Order 4 Log Po lynomial Fit................................. 13.21 Figure 13.16a Mt Keith #2 Failure, Curve Fit for Deformation Data Using Order 7 Log Polynomial Fit.................................................................................................................. 13.22 Figure 13.16b Mt Keith #2 Failure, Residuals Using Order 7 Log Po lynomial Fit ................................. 13.22 Figure 13.16c Mt Keith #2 Failure, Error B ounds for Order 7 Log Po lynomial Fit................................. 13.22 Figure 13.17 Harmony Pit, East Wall Failure, Change in 2D Deformation Rate for Prism 422 During Prog ressive Behaviour ....................................................................... 13.19 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes List of Tables Page 1 of 1 LIST OF TABLES Table 3.0 Summary of Literature Surv ey Open Pit Case St udy Database..........................................3.1 Table 3.2 Estimated Rock Mass Strengths (After Martin and Ca rew, 1986 and Martin, 1993) ...........3.6 Table 3.17 Nchanga Open Pit Li thology ............................................................................................. 3.34 Table 5.1 Mean Orientations of 3 rd Order Structures (Roux, 200 6) .....................................................5.3 Table 5.2 Summary of Mean Rock Mass Ratings (Roux, 200 6)..........................................................5.5 Table 5.3 Summary of Intact Rock Strengths (Roux, 2006) ................................................................5.5 Table 5.4 Pit Configurati on Parameters for 2005 ..............................................................................5.5 Table 6.1 West and Ea st Wall Joint Sets....................................................................................... ......6.5 Table 6.2 Survey ed Slope Angles................................................................................................ ........6.6 Table 7.1 Summary of Lithological Domains and Lithological Codes (Mining One , 2003 and Goodchild, 2006)) ............................................................................................................. ................................... 7.3 Table 7.2 Summary of Join t Sets (Haywood, 2004) ............................................................................7.7 Table 7.3 Overall Pit Configur ation Parameters (July 2006) ...............................................................7.9 Table 8.1 Overall Pit Configuration Pa rameters at Abandonment (2006) ........................................8.5 Table 9.1 Venetia Overall Pit Confi guration Parameters, October 2006 .............................................9.5 Table 10.1 Orapa Overall Pi t Configuration Parameters , September 20 06 ................................... 10.10 Table 10.2 Letlhakane Overall Pit Configur ation Parameters, Se ptember 2006 ............................ 10.10 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes List of Symbols Page 1 of 2 LIST OF SYMBOLS BRMR Bieniawski Rock Mass Rating (Bieniawski, 1976) c? Cohesion intercept based on effective stress ci Cohesion ? intact rock cm Cohesion ? rock mass D Hoek Brown disturbance factor Ei Young?s modulus of intact rock EM Modulus of deformation of rock mass EMs Static modulus of deformation of rock mass EMd Dynamic modulus of deformation of rock mass EDM Electronic distance measurement FS or FOS Factor of safety GSI Geological strength index (Hoek,1999) Gi Shear modulus ? intact rock Gm Shear modulus ? rock mass IRS Intact rock strength (Unconfined uniaxia l compressive strength of intact rock) JRC Barton?s Joint roughness coefficient (Barton,1973) JSC Barton?s Joint compre ssive strength (Barton,1973) Km Bulk modulus ? rock mass Ki Bulk modulus ? intact rock lps litres per second mi Hoek-Brown constant for rock mass MRMR Mining rock mass rating (Laubscher,1990) OOC Onset of collapse PN Pit north orientation r Correlation coefficient for st atistical evaluation (statistics) RQD Rock quality designation PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes List of Symbols Page 2 of 2 SDH Slope development history SMP Survey monitoring point or slope management programme TN True north orientation UCS Unconfined/Uniaxial compressive strength x horizontal direction of movement y horizontal component of movement z vertical direction of movement ? Normal stress ?ci Uniaxial compressive strength of intact rock (also intact rock strength) ?tm Rock mass tensile strength ?cm Rock mass compressive strength ?? Friction angle based on effective stresses ? Friction angle based on total stresses ? i Friction angle - intact rock ? m Friction angle ? rock mass ? Shear stress of a material ? Dilation angle ? Poisson?s ratio u Pore pressure ? Strain ?& Rate of strain PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Definition of Terms Page 1 of 3 DEFINITION OF TERMS Banded Ironstone Formation (BIF) A layered deposit of chemical sedimentary rocks containi ng at least 15 percent (by weight) iron in the form of sulfide, oxide, hydroxide, or ca rbonate minerals (Geology.com, 2006). Boudins/Boudinage A geological structure formed by the folding of a mix ed series of competent and incompetent beds whereby the more brittle beds become fractured over the crests of the folds into long ?pole shaped? pieces slightly separated from each other called boudins. The spaces between adjacent boudins may become infilled with quartz or the incompetent material from above or below which has flowed in. (Read and Watson, 1970) Cause of Damage These are the mechanical conditions such as stre sses and blasting that make the rock mass deform permanently or fracture Damage Mechanism This is the manner in which a rock mass is damaged or the processes that take place in the rock mass causing it eventually to fail. Damage Evolution This is the accumulation of the damage mechanisms durin g the different stages of mining. The evolution can be gradual or sudden depending on the causes and the mechanical properties of the rock mass. Deformation Pathway A deformation pathway shows the continuous deformation time relationship of a particular monitoring point on a slope in the context of the five deformation sta ges of the Deformation Behaviour Model as discussed below. Dilation General term used to describe the general loosening up of the rock mass in and around an open pit as a result of unloading of the rock which results in both vertical and horizontal stress relaxation. Events Events are used to describe individual mining related cuts/pushbacks/blasts/excavations which directly result in a change in the global stress field with in the rock slope and consequently trigger a deformation response from the rock mass. By their very nature mining events are time independent in the sense that they occur as and when they are planned. For that reason events are reflected in the deformation model as occurring at relatively random time intervals. Events are shown as blue arrows. Epoch A period in time marked by special even ts (the Concise Oxford Dictionary). The term is used in this work to refer to a period of time during which a series of individual survey measurements has been obtained. Facies The characteristics of a rock mass t hat reflect its depositional environment. These characteristics enable the rock mass to be distinguished from rocks deposited in adjacent environments (Geology.com, 2006). PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Definition of Terms Page 2 of 3 Felsic A term used to describe an igneous rock that has a large percentage of light-colored minerals such as quartz, feldspar, and muscovite. Also used in reference to the magmas from which these rocks crystallize. Felsic rocks are generally rich in silicon and alum inum and contain only small amounts of magnesium and iron. Granite and rhyolite are examples of felsic rocks. (See mafic to contra st.) (Geology.com, 2006). Geotechnical Zone or Design Sector Refers to a major geological unit that behaves and responds in a particular manner to local mining activities These zones are often defined by major rock types Initial Response Contrary to Martin?s (1993) usage of the term as disc ussed in Section 2.4.2, the term ?initial response? is used here to the immediate deformation response following a change in the stress field also referred to as strain hardening. Laccolith An igneous intrusion that has been forced between two layered rock units. The top of the intrusion is arched upwards and the bottom of the intrusion is nearly flat (Geology.com, 2006). Mafic A term used to describe an igneous rock that has a large percentage of dark-colored minerals such as amphibole, pyroxene and olivine. Also used in referenc e to the magmas from which these rocks crystallize. Mafic rocks are generally rich in iron and magnesium. Basalt and gabbro are examples of mafic rocks. (See felsic to contrast.) (Geology.com, 2006). Major Discontinuities or Major Structures Those structures as above which ar e sufficiently well developed and have great enough continuity so that shear failure along them will not involve shearing of any intact rock material (Piteau, 1970). Regolith A general term used in refernce to unconsolidated rock, alluvium or soil material on top of bedrock. Regolith may be formed in place of, or transported in from adjacent lands (Geology.com, 2006). Schistosit y The parallel arrangement of platy or prismatic minerals in a rock that is caused by metamorphism in which directed pressure plays a significant role (Geology.com, 2006). Stratigraphy A study of sedimentary rocks units, including their geographic extent, age, classification, characteristics and formation (Geology.com, 2006). Lithology The study and description of rocks, including their mineral composition and texture. Also used in reference to the compositional and textural characte ristics of a rock (Geology.com, 2006). Onset-of-failure (OOF) The onset-of-failure point was used by Zavodni and Broadbent (1980) as the point defining the transition from the ?regressive stage? to the ?p rogressive stage?. In that sense t he meaning is retained. It is further PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Definition of Terms Page 3 of 3 defined for the purposes of this work as the point in time beyond which displacements will continue to accelerate to collapse unless remedial measures are implemented. Rock Mass The rock mass is defined as an aggregate of blocks of so lid rock material which contains structural features which constitute mechanical discontinuities. The rock mass 'refers to any in-situ rock with all inherent geomechanical anisotropies' (John, 1962). Rock Mass Creep Mode This refers to the regressive (creep) behaviour char acteristics of the rock mass after a mining event and initial response (strain hardening) has occurred. Rock Mass Damage Rock mass damage in an engineering context is any frac ture and crack generation or extension, shearing and degradation of in situ fractures and discontinuities, that causes a reduction in the rock mass strength. Rock Material or Intact Rock This is the consolidated aggregate of mineral particle s forming the solid material in between structural discontinuities. 'The properties connected with it re fer to rock material not subject to geomechanical anisotropies such as jointing' (John, 1962). Structural Discontinuities or Planes of Weakness These include all geological features t hat separate the solid blocks of the rock mass, such as joints, faults, etc. These features have appreciably lower strength than the rock material and constitute mechanical discontinuities in the rock mass (Piteau, 1970) Slope Fatigue This is a term which describes the process whereby rock mass damage is accumulated. SSR Slope stability radar Time Dependency As has been shown throughout this research, the very nature of the non-linear rock mass response is time dependent from the point in time when the change in the stress field occurred. The time dependent response behaviour which is triggered by a specific event is strongly influenced by the individual rock mass characteristics at the time of the event. The Model A truncated term used for the Time and Event Dependent Deformation Behaviour Model. Ultrabasic Rock/Ultramafic An igneous rock with a very low silica content and rich in minerals such as hypersthene, augite and olivine. These rocks are also known as ultramafic rocks. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 1 : Introduction Page 1-1 1 INTRODUCTION 1 . 1 Background Deformation has long been recognized as one of the most important parameters which define the state of rock masses. M ?ller (1974) suggested that deformation rather than stress be used as the basis for the assessment of stability of rock masses. The obvious advantag e of this is the relative ease with which deformation can be measured compared with stress. Hoek (1999) pointed out that the evaluation of trends is of great value to geotechnical engineers because if these trends can be shown to be consistent over a wide range of conditions then it is likely that a basic law is at work, which if isolated can possibly be formulated mathematically. Monitoring of deformation movements has long been recognised as an important engineering practice to ascertain the ongoing performance of any engineering structure, especially in situations where margins of safety are low and early warning of impending failure is required in order to save lives and equipment and implement remedial works to save the structure or engineering works (Jennings & Steffen, 1979). It has been shown in literature (Martin, 1993) and co nfirmed by the Author?s own experience that deformation behaviour of both excavated rock slope s and failures occurring within excavated rock slopes are both event and time dependent. Deformations are event dependent in the sense that mining or excavation of rock slopes by human activi ty induces relatively rapid changes in the stress field within a rock mass resulting in elastic or quasi-elastic deformation responses whenever the excavation occurs. Deformations are also time depen dent in the sense that a change in a stress field within a rock mass can result in significant def ormations occurring over a relatively long period of time, especially as a result of creep occurring both within the intact rock and the structures making up the rock mass as a whole. Although survey (or prism) displacement monitoring systems are now in widespread use throughout the open pit mining industry it is the Author?s expe rience that most usage of deformation data takes the form of identifying overall basic trends and patterns which are used to assist in decisions regarding mining operations. With the exception of attempting to ?calibrate? numerical models, very little attempt is made to quantitatively use deformation measurements. In has often been the Authors experience that where ?calibration? of numerical models has been attempted there often arises significant confusion as to what deformation measurements should be used, especially in situations where it is apparent that the creep component of the deformation is significant. At worst, the Author has witnessed deformation measurements having been applied blindly to numerical models without any attempt to understand the behaviour patterns and relevant components of deformation movement. The work described in this research effort was undertaken in order to research the time and event dependent deformation behaviour of excavated rock slopes throughout a wide range of geological terrains and slope configurations, with the overall objective of expanding the knowledge base in this important area of monitoring of open pit mining operations. Additionally, the research work focussed on possibly identifying potential predictive deformation indicators of instability or failure and evaluating and developing their practical usage for mining operations. It is known that the complex displacement behav iour and patterns observed in everyday slope monitoring is a superposition of different patterns of displacement resulting from different components of rock mass behaviour, such as elastic and inelastic rebound, stress relief, relaxation, creep from PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 1 : Introduction Page 1-2 structures and intact rock, effectively forming ?patte rns within patterns?. The research therefore also seeks to build, where possible, greater understandi ng of these individual displacement components. It should be emphasised that this research is focussed towards furthering and applying better knowledge and understanding of rock mass deformati on behaviour in a practical sense in a mining environment. As such, further assessments of the degree of stability of the associated failures and instabilities by limit equilibrium methods or rigorous numerical modelling methods, do not therefore form a focus of this research. In order to evaluate the time dependent deformation b ehaviour of an excavation, it is necessary not only to understand the general geological and structural terrain within which the excavation is situated, but also other influencing factors such as the type and properties of intact rock, structures within the rock mass, slope height, slope geometry and mining rates. Transient influencing factors include groundwater, seismic loading and blasting, undermining by mechanical equipment, precipitation, storm water control and to a lesser extent climate (in re gions where winter freezing occurs). In addition, instabilities caused by underground mining and caving can form significant destabilising influences. This requirement for considerable supporting know ledge in order to evaluate and contextualise individual deformation behaviour makes this type of research a daunting but immensely rewarding task to undertake. 1.2 Objectives The overall objectives of this rese arch project are set out as follows: 1. To expand the general knowledge base of ti me and event dependent deformation behaviour in excavated unsupported rock slopes in different geological terrains. 2. To establish typical patterns of deformation behaviour and identify trigger mechanisms for different modes of failure in the context of different slope geometries, geology, structure and rates of mining. 3. To develop where possible a more fundamental characterisation of the different components of time dependent deformation, which include rebound, stress relief response, strain hardening/softening, creep and failure initiation, in the context of the different geological terrains and structures. 4. To identify, if possible, relatively simple potent ial predictive deformation indicators of instability or failure and to evaluate and develop their practical usage for mining operations. 5. To develop, if possible, a displacement/behav iour criterion that would enable displacement behaviour to be correlated to a measure of stability 1 . 3 Research Approach In order to study and characterise instabilit y mechanisms and time dependent deformation in excavated rock slopes, the overall research approa ch was intentionally broad, but still achievable within the limits of time and financial resources available. As required deformation monitoring data was obtained within individual mining operations, the research was of necessity centred around individual PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 1 : Introduction Page 1-3 mining case studies. Due to the sometimes considerable amounts of data made available, considerable effort has had to be made to summarise and present only relevant data as concisely as possible. The thesis is presented in three principal sectio ns. The first is presentation of the findings of a literature review and the development of a literature review case study database. The second section of the work involves the presentation of detailed ca se studies where, in all cases, the Author had an opportunity to inspect the pits and collect comprehensive data. The final section of the thesis presents the findings of the evaluation of the deformation behaviour from all the case studies and draws conclusions. The approach to each of these three sections is discussed in further detail below. 1 . 4 Literature Review and Developmen t of a Literature Review Case Study Database A comprehensive literature review is presented in Section 2. The literature review commences with a review of the definition of failure and overall deformation behaviour terminology. This is followed by a review of the basic physical characteristics of rock mass which directly influence deformation behaviour. These characteristics include a brief review of what rock is and the nature of intact rock, creep behaviour of intact rock as well as creep b ehaviour of rock joints. This is followed by the classification of instability and failure mechanisms as well as a summary of general modes and mechanisms of instability. Previous literature concerning descriptions of the nature and behaviour of rock mass in excavations and slopes is present ed together with associated deformational components and features. A significant body of early literature was found which concerned the gravitational behaviour of large scale rock mass (rock mass creep) in Alpine regions. Further literature is presented on rock mass time to failure predictions and displacement monitoring techniques which are commonly used to monitor deformation are reviewed. Case studies of published slope failures associated with mining operations were specifically identified for review. The principal objective of these case study reviews was to obtain actual deformation data of failures that have occurred and to review this data in the context of the modes of instability and failures that were reported. In Section 3 a selection of case studies are indivi dually summarised. The selection criteria were based on achieving a variation of failure modes, the individual study reporting on a relatively large or significant failure within an open pit, as well as the literature containing good time-dependent deformation data and associated geological and structural data. 1 . 5 Refinement and Definition of the Research To pic Based on Findings of the Literature Review Section 4 presents the refined research approach based on all the literature review findings. Specifically, the following aspects are dealt with; ? What information is required to be collected from the detailed mining operation case studies. ? What detailed mining operation case studies are available and which would be the most suitable based on availability and quality of the required information. ? How the interpretation and evaluat ion of the information will be undertaken in the final sections. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 1 : Introduction Page 1-4 ? How this information will be pr esented in the detailed case study write ups and in the chapters where the information is interpreted. 1.6 Detailed Mining Operat ion Case Studies. In Section 5 to 10 six detailed case studies are presented. Not all the case studies which were selected in Section 4 for investigation are presented for reasons discussed later. 1 . 7 Evaluation and Interpre t at i o n of Data Collected from Detailed Case Studies and Presentation of Findings In Section 11 a time and event dependent deformation model is presented. The development of this model was facilitated by the extensive review of deformation behaviour and failure characteristics of both literature and detailed case studies. The model incorporates both a description of horizontal and vertical deformation behaviour patterns from the start of mining of the rock slope through to collapse or post mining recovery and/or stabilisation. Section 12 illustrates the application of the model using case study data and Section 13 presents a statistical based approach which has been developed to enable successive forecasts of deformation rate and acceleration to be made on a real time basis. 1.8 Final Comments In the literature there is sometimes considerable variation in terminology describing time dependent deformation behaviour. Terminology can vary liter ally from one mining operation to another which inevitably leads to unnecessary confusion. A special effort has therefore been made to standardise terminology and most of the terminology used has been defined under the Section ?Definition of Terms?. Although no specific objectives were set in terms of the number of case studies and failures to be studied, for information purposes the statistics on this research are summarised below; ? Number of literature survey case studies reviewed: 30+ ? Number of literature survey case studies failure events reviewed: 30 ? Number of literature survey case studies presented: 9 ? Number of detailed case studies investigated: 12 ? Number of detailed case studies presented: 6 ? Number of detailed case study multi-bench scale failure events reviewed: 53 ? Number of detailed case study multi- bench scale failure events presented: 27 ? Total number of literature and detailed case studies reviewed: 42 ? Total number of literature and detailed case studies presented: 15 ? Total number of failure events reviewed (literature + detailed case studies): 83 ? Total number of overall failure events re corded (including those reviewed): 240+.* * Note : Mt Keith Operations alone had a database at the time of this research of 141 failures. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-1 2 REVIEW OF LITERATURE RELEVANT TO ROCK SLOPE DEFORMATION BEHAVIOUR 2.1 Introduction The literature review commences in Sections 2.2 and 2.3 with a review of the definition of failure and overall deformation behaviour terminology. This is followed by a review of the basic physical characteristics of rock mass which directly influence deformation behaviour. These characteristics include; ? A brief review of what rock is and the nature of intact rock (Section 2.4), ? Creep behaviour of intact rock (Section 2.5), ? Creep behaviour of rock joints (Section 2.6). ? Rock mass strain hardening and softening (Section 2.7) Previous literature concerning descriptions of the nature and behaviour of rock mass in excavations and slopes is presented together with associated deformational components and features (Section 2.8 and 2.9). This is followed by the classification of instability and failure mechanisms as well as a summary of general modes and mechanisms of instability in Sections 2.10 and 2.11. Time dependent deformation of rock slopes has been known from as early as 1807 (Zay, 1807), but started to be documented in more detail in the early 1900?s (Lapworth, 1911; Lugeon and Oulianoff, 1922; Heim, 1932; Ampferer, 1939). A lot of the early research work focused on the creep behaviour of Alpine slopes in Europe and is presented in Section 2.12. Further literature is presented on rock mass time to failure predictions (Section 2.13). In the final section displacement monitoring techniques which are commonly used to monitor deformation are reviewed in detail. 2.2 Definitions of Failure In the literature review as well as in the Author?s experience, ?failure? is a loosely applied term that has been used to describe almost every degree of instability and type of behaviour possible. For the purposes of this research only the terms ?collapse?, ?functional failure? and ?instability? are used and are defined as follows. ?Collapse? is defined as the complete overall loss of rock mass integrity and structure. ?Functional failure? is defined as a situation where a slope cannot perform the function for which it was intended. This implies that it does not necessarily involve overall collapse although localised sections of the structure may have collapsed. Examples would include haul roads, ramps or conveyor ramps (where sufficient misalignment of the conveyor renders it inoperable (Goldberg & Frizzell, 1989)). ?Instability? is defined as any other deformational movement or behaviour that does not involve collapse and/or functional failure. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-2 2.3 Deformation Behaviour Terminology Zavodni and Broadbent (1980) divided the displacement rate of slopes into two categories or distinctive phases termed ?progressive? and ?regressive? stages. Regressive behaviour was defined as a ?stage during which failing rock mass will re-stabilise if some disturbance(s) external to the rock and structure is (are) removed; the overall average velocity in this stage can slightly accelerate, remain constant or decelerate?. Similarly, progressive behaviour was defined as a ?stage during which the failure will displace at an accelerating rate to the point of total collapse unless active control measures are taken?. This behaviour is illustrated in Figure 2.18. Martin (1993) describes progressive failure rather more broadly as being accompanied by increasing displacement and movement rates of a rock mass in response to continued deformation. Slope fatigue is a ?loose term? which is used to describe the accumulation of rock mass damage in such a way as to continuously weaken the slope. Zavodni and Broadbent (1980) used the term slope fatigue factor which was defined by them as a ?factor that is related to the number and size of external impulses?. 2.4 Nature of Intact Rock Rocks are natural aggregates of one or more minerals as well as amorphous particles of mineral or organic matter that are cemented together. The different particles can themselves each have different mechanical behaviour (Cristescu and Hunsche,1998). Rocks can be formed in igneous processes by the cooling and consolidation of molten magma, sedimentary processes as a result of consolidation of sediments or loose material derived from the weathering of other rocks or in metamorphic processes as a result of chemical or physical changes occurring due to changing situational conditions. Rock materials can be far from homogeneous, continuous and isotropic, being influenced by interconnected pores and microfissures which result in nonlinear load and deformation response, reduction in tensile strength, a stress dependency and variability in material properties when tested (Goodman, 1989). Rocks along with other geo-materials are also unique in that they display dilatency at failure (Cristescu and Hunsche,1998). Idealised constitutive stress strain models have been developed for rocks which include perfectly elastic, elastic with hysteresis, linear elastic, elastic plastic, elastic strain hardening and softening models (Franklin and Dusseault, 1989). Where non recoverable plastic strain occurs this can usually be attributed to deformation of crystal grains of the material, rearrangement of the mineral aggregate, intergranular slip and/or initiation and extension of fractures within the material (Martin, 1993). 2.5 Creep Behaviour of Intact Rock The creep behaviour of intact rock is an enormously complex topic that has been the subject of considerable research, with over 600 references on the general topic of rock creep having been found. The creep behaviour of intact rock is a long term phenomenon which is considered to have a minimal influence in the limited time frames of the required life of a hard rock mining slope. It could nevertheless have a more significant influence on slope behaviour in regions with deep weathering profiles. The characteristic behaviour of intact rock creep has therefore been reviewed superficially. The following description of the behaviour of intact rock creep has been summarised from Cristescu and Hunsche (1998). PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-3 On a change in differential stress (load change) creep deformation of intact rock initially responds with a high deformation rate termed a transient creep that decreases continuously until it reaches a stationary creep rate which is also termed a steady state creep rate. This process reflects a dual behaviour of strain hardening and recovery and occurs whenever there is a change in differential stress. Hardening predominates after an increase in stress and the initially high rate of creep decreases with time. This is also known as normal transient creep. Tertiary creep is characteristed by an increasing strain rate caused by increasing internal damage of the rock material and usually ends in creep rupture (failure or collapse).Tertiary creep can occur after a period of steady state creep or directly after transient creep. Deformation mechanisms for creep are most commonly influenced by stress and temperature, but can also be influenced by water, which is termed hydraulic weakening (not to be confused with the effective stress principle) and is often the focus of studies in crustal and lithospheric deformation. The creep deformation behaviour of different rock types varies considerably. For example, uniaxial creep curves for Dolomite show practically only transient creep behaviour after each sequential loading increment which is followed relatively suddenly by failure (tertiary creep) at approximately 0.25% axial strain and a ?1 = 125MPa. In contrast, schist reflects an increased steady state creep rate after each load increment eventually transitioning into failure (tertiary creep) at approximately 0.7% axial strain and ?1 = 34.3MPa (Cristescu and Hunsche, 1998). Almost all creep behaviour is loading rate dependent as well as loading history dependent. Volumetric creep is generally only of concern at high confining pressures with superimposed loading. This is however generally not applicable to excavation through insitu rock mass. 2.6 Time Dependent Deformation Behaviour of Rock Joints The time dependent deformation behaviour of discontinuities in rock is complex and has historically not been well researched but is of immense importance in the understanding of time dependent deformation behaviour of rock masses. For this reason the subject has been reviewed in more detail. Noteworthy contributions to this field have been made by Bieniawski, (1970), Wawersik (1974), Schneider(1977), Dieterich (1972), Dieterich (1978) and Amadei and Curran (1980). The study by Amadei and Curran (1980) is particularly noteworthy. They undertook triaxial and direct shear tests on unfilled and clean joints using four rock suites namely sandstone, limestone, marble and granite. Their findings are summarised as follows; ? In general the creep deformation of the intact rock was found to be negligible in comparison to the creep occurring along the joints in the stress ranges considered. ? Typical creep behaviour can be described as an initial elastic deformation, followed by a steady- state transient creep phase. ? Tertiary creep (rapid acceleration or failure) was never observed in the experiments. ? The relationship between shear displacement ?u and time can be expressed by the following equation; ?u (t) = A.log (t+1) + B + C.t (2.1) PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-4 Where A, B and C are constants for a given loading. The frictional displacement mechanisms on the joint surfaces reported by Amadei and Curran (1980) were as follows; Sandstone : brittle fracture of the quartz grains and wear grooves due to ploughing Granite : ploughing Marble : ploughing due to the presence of a small amount of quartz accompanied by plastic deformation Limestone : plastic deformation Parameters such as rock type, surface conditions, stress state, degree of interlocking and testing conditions all influence joint behaviour and were not able to be easily differentiated. It was noted that two joints with identical surface conditions but with different degrees of initial interlocking can give significantly different results in a friction experiment and Dieterich (1972) showed that the static coefficient of friction (?) is itself time dependent and can be represented by the equation; ?(t) = ?0 + ?. log (t+1) (2.2) Where ?0 is the initial coefficient of friction and ? is a constant for different rock types. This implies that the peak shear strength (?p) is also a function of time. Dieterich (1978) suggests that adhesion and/or ?asperity ploughing? are the dominant frictional processes controlling this behaviour. Amadei and Curran (1980) illustrated the quasi-static shear behaviour of a discontinuity at a constant normal stress ?n as shown in Figure 2.1. For any given normal stress (?n) creep displacements can occur along the joint for any value of shear stress (?) less than ?p and, intuitively, creep deformation can be expected to increase for any given ?n as ? increases. This implies that the coefficients A, B and C in Equation 2.1 are functions of the ratios of ?/?n or ?/?p. ?p0 is the initial peak shear strength determined from a quasi-static shear test. The results of the research showed that there was a limit in terms of the ratio ?/?p0 above which creep along the joint becomes the dominant component of deformation, and below which creep displacement along the joint is negligible and creep of the intact rock dominates. The value of ?/?p0 determined for marble was approximately 0.5. From Figure 2.1 it is also apparent that assuming the peak and residual shear strengths and stiffnesses (k1 and k2) are constant, the magnitude of the creep deformation ?uf necessary for failure is represented by the distance x-y. Amadei and Curran (1980) went on to show that the time dependency of the static coefficient of friction implies that the peak and residual as well as the stiffnesses (k1 and k2) are also time dependent and that the creep deformation ?uf necessary for failure changes accordingly as shown in Figure 2.2. In summary the long term stability of rock joints is influenced by two interrelated processes. Whilst creep deformation ?u is occurring the magnitude of ?uf necessary for failure is also changing. Failure occurs when ?u = ?uf. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-5 2.7 General Descriptions of the Deformation Behaviour of Rock Mass Knowledge of the deformation behaviour of a rock mass is of considerable practical importance. The presence of structural discontinuities can significantly influence the behaviour of a rock mass. The degree of influence of discontinuities is dependent upon their genetic type, nature, frequency, orientation, infill, continuity and interaction of different sets of structures within the rock mass. The presence of discontinuities results in a discontinuous rock mass medium and the mechanical behaviour of the rock mass being both anisotropic and heterogeneous (Piteau, 1970b). Further factors influencing rock mass behaviour are groundwater, mineralogy, lithology and weathering (Piteau, 1970b). Deere et al (1967), Broadbent and Zavodni (1981) and Patton (1970) have also confirmed the importance of identifying major structures such as faults and shear zones as the principal controlling factors in slope stability and deformation behaviour. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-6 Research into the time dependent deformation behaviour of fractured rock has to a certain extent been focussed towards underground mining or tunnelling in relatively high stress environments. Bieniawski (1970) undertook uniaxial compression tests and concluded that while initially affected by time, insitu fractured rock will finally reach a long term stable state. From these gradual stabilisation trends long term design data may be derived. Broadbent and Ko (1971) detail the findings of a considerable field scale research effort by Kennecott Copper Corporation into stability research. Their research was the first to identify the acceleration and decay type movements resulting from mining operations as shown in Figure 2.3 and they went on to develop a basic rheological model to describe the pattern of displacements. Kaiser and Morgenstern (1979 and 1981) studied the time-dependent deformation of jointed rock near failure using coal and derived a phenomenological model for underground jointed rock using frictional and rupture elements. Zavodni and Broadbent (1980) examined the displacement records of 13 failures in open pit porphyry copper mines and derived composite displacement rate vs time plots for the data, from which they derived an empirical equation for a line fit on a log-normal chart (see Section 2.13). Martin (1993) established an idealised characteristic time dependent deformation behaviour as illustrated in Figure 2.4 and further characterised magnitude of displacement, displacement rates, strain hardening and softening and or progressive failure which are discussed in further detail in the following Section. 2.8 Rock Mass Strain Hardening and Softening From a homogeneous material constitutive model point of view strain hardening and softening has been well documented and can be described as the hardening or softening of friction, cohesion, dilation or tensile strength after the onset of plastic yield (Itasca, 2001). Martin (1993) described strain hardening of the rock mass as a ?locking up of the rock mass through dilation? and suggested that progressive deformation results in mobilization of available shear strength within the rock mass or along failure surfaces and usually occurs during or just after the ?initial response?. The strain hardening results in decreasing deformation movements as a result of the mobilization of shear strength thereby increasing overall stability. Martin (1993) goes on to describe strain softening of the rock mass as a reduction in the shear strength of a rock mass or failure surface as a result of increasing displacement or strain. 2.9 General Components of Rock Mass De formation Resulting from Excavation Deformational components resulting from excavation are discussed in detail below. It should be noted that in general it can sometimes be difficult to categorise overall deformation behaviour which may involve several different behavioural components with each producing differing geological features. For example features associated with valley formation such as up warping and bulging may be attributed to both positive vertical rebound as well as valleyward (inward) movement resulting from large scale rock mass creep deformation. 2.9.1 Vertical (Upwards/Positive) Deformation and Rebound (Swelling) Resulting From Unloading Jennings and Steffen (1979) described rebound as an expansion of a geological material as a result of the reduction of effective stress. The effects of rebound have been studied especially closely in the PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-7 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-8 civil and construction industries where movements of excavations can significantly influence construction of the foundations of structures, especially dams. On a larger scale, this type of deformation has also been relatively well documented geologically for the rebound of river valleys and canyons. Feld (1966) documented the movement in the walls and floor of trenches excavated for the Niagara Falls power station (219mm vertically and 63.5mm horizontally) which created unstable rock conditions. Hanna and Little (1991) estimated rebound potential at a dam site in British Columbia and Matheson and Thomson (1973) documented the up warping of valley rims adjacent to valleys (also termed valley flexure) using several different sites. The up warping was created by the elastic rebound of the rock into which the valleys had eroded. Matheson and Thomson (1973) concluded that magnitude of rebound could be suitably predicted by elastic mathematical models and that the relative magnitude ranges from 3% to as much as 10% of the depth of the valley. In sedimentary rocks up warping can be accompanied by shearing between or along strata contacts. Ferguson(1967), Ferguson and Hamel (1981) and Martin (1993) reported on the occurrence of residual stress within sedimentary rocks in various locations throughout North America. They described features such as sub-vertical tension fractures sub-parallel to the valley walls and arching and compression type faulting in the floor of the valleys which were created after valley formation by inward movement of the valley walls resulting from high insitu residual stresses. The frequency of the fractures is related to the natural discontinuities and rock strength. The fractures have resulted in unusually wide zones of loosened rock up to 300m behind the valley walls. 2.9.2 Initial Horizontal and Vertical (Downwar ds) Deformation (Relaxation) Resulting from Excavation After an excavation has taken place the rock mass in the slope can be expected to respond relatively rapidly to the change in the stress field generated by the excavation (blast) and removal of supporting rock material. Elastic deformation is the linear relationship between applied stress and observed strain. In the excavated rock slope environment this linearity is used in the sense of a physically non- recoverable or reversible strain occurring during excavation and Jennings and Steffen (1979) considered this behaviour to reflect ?elastic? behaviour even though the movements are not reversible. 2.9.3 Long Term Horizontal and Vertical (Downwar ds) Creep Deformation (Relaxation) Resulting from Excavation Martin (1993) reported that the rate of response following an individual mining event, generally decreases from an initial rate of several millimeters per day to no movement over a period of time that can extend from weeks to years. Martin (1993) showed that this decrease in response rate takes the form of a negative exponential relationship of the form; btAeR ?= (2.3) Where R : movement rate A,b: constants depending on rock mass properties, slope geometry, height etc. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-9 Martin (1993) rather confusingly uses the term ?Initial response? to refer to the entire time period from initiation of excavation up to the onset of progressive behaviour. During this time period Martin (1993) describes the slope as ?undergoing a period of time dependent adjustment? in response to changes in the local and global stress environment as a result of excavation (mining). 2.10 Classification of Instability and Failure Mechanisms Broadbent & Zavodni (1981) classified fundamental large scale rock failures into three distinct types based on structure orientation and discontinuity strengths. Each type was associated with a different characteristic displacement behaviour. The primary structures were further classified into ?simple control? where failure was defined by one or two surfaces and ?complex control? where the failure was defined by multiple structures. These three systems and their associated characteristic deformation behaviour are summarized in Figure 2.5. Martin and Kaiser (1984) classified slope failure mechanisms into three classes based on the kinematics of the instability as summarized below and illustrated in Figure 2.6. Class 1 : Instabilities characterised by rigid body motion along planar or circular failure surfaces, Class 2 : Instabilities characterised by local yielding of the rock mass to permit rigid body motion along an irregular, non-planar or non-circular failure surface. Class 3: Instabilities characterised by yielding of the unstable rock mass along internal shear surfaces to permit deformation along irregular failure surfaces, PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-10 2.11 Modes and Mechanisms of Instability Rock slope instability results from either; ? a failure through intact rock mass, or ? along pre-existing structural discontinuities, or ? along a failure path or surface made up of combinations of both failure through sections of intact rock as well as along section/s defined by pre-existing structural discontinuities which together make up a path of least resistance (Piteau & Martin, 1982). There are several fundamentally recognised modes of instability which include instability of the rock mass as a whole; wedge type instability, planar sliding and toppling type instabilities. The mechanisms causing instability in any one of these modes are usually complex and site unique. They may relate to some or all of the following factors such as physical, mechanical and shear strength properties of the intact rock and structural discontinuities, geology and hydrogeology as well as the slope geometry (Piteau & Martin, 1982). The most important of these are usually the discontinuities and hydrogeology, however the interrelationship between all the factors should be considered and correctly evaluated in order to identify the correct instability or failure mechanism. The analysis methods for the different mechanisms are not reviewed. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-11 2.11.1 Evaluation of Modes of Instabilit y Potential failure mechanisms or modes of instability should be evaluated from the point of view of the magnitude of the impact or potential consequences for the mining operations and safety. In this regard overall slope failure as well as localised, limited or smaller (bench scale) failures should be considered. Potential instability mechanisms are generally the simplest to evaluate in two dimensions (2D) or plane strain. In terms of the theory of elas ticity, the assumptions for 2D plane strain (in an x,z plane) are that; ? Cross section remains constant, ? Applied body forces must be ?in plane? ie parallel to the x,z plane ? Strain in the y (out of plane) direction is either zero or a constant, ? Anisotropic materials are orientated so that their two principal axis are ?in plane? In geotechnical terms, 2D plain strain assu mptions would also include the assumptions that; ? the potential failure surface is and remains parallel to the strike of the slope, ? lateral release planes or surfaces of separation exist which do not provide any resistance to failure. Evaluation of 2D potential failure mechanisms can involve both a single discontinuity or a single set of discontinuities which would result in planar, toppling or buckling type failure modes. An additional second set of discontinuities would possibly modify the failure path further to a step path mechanism (Piteau & Martin, 1982). Certain types of instability modes such as the wedge type instabilities which form by the intersection of one or more discontinuity sets have to be evaluated in three dimensions. Instability modes have been researched and docume nted by many authors, some of the early researchers being John (1962), Jennings and Steffen (1967), Jennings (1968), Hoek (1970) and Hoek and Bray (1981). There are several well known basic instability modes. These are summarised in the following sections. 2.11.2 Planar Instabilit y Modes In their simplest form, planar instabilities involve s liding of a rock mass along a single surface (Figure 2.8a). These types of failures are generally t he easiest to understand and analyse and may be applicable if instability occurs along well defined continuous, planar and non-undulating discontinuities. In order for failure to be kinematically possible the following conditions must be met. These are; 1. A continuous plane on which sliding occurs must strike parallel or subparallel to the strike of the slope (within ? 20? ), 2. The dip of the failure plane( ?fp) must be less than the average dip of the slope ( ?s) ie it must ?daylight? in the slope ( ?fp < ?s), 3. The dip of the failure plane ( ?fp) must be steeper than the angle of friction of this plane unless groundwater pressure exists in which case ?fp < ? fp, 4. Lateral relief boundaries which provide negligible resistance to sliding must be present. These conditions are illustrated in t he stereographic plot in Figure 2.7. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-12 Should the conditions for well defined continuous and non-undulating discontinuities not be fully satisfied, the planar instability surface develops both through intact rock and along discontinuities. The following possible alternative forms of planar failures are briefly described. Discontinuous Joints Failure along discontinuous joints occurs as a resu lt of failure or sliding along lower strength discontinuous joints/shears/bedding combined with shear failure through intact rock bridges between the joints/shears/bedding as illustrated in Figure 2.8b. Development of a Step Path Mechanism A step path instability mechanism can develop when the path of least resistance is formed by a combination of either; ? sliding along discontinuities, tensile separation of intact rock between discontinuities as well as shear failure through intact rock bridges along discontinuities, or ? sliding along discontinuities, tensile separation along steeply dipping cross cutting joints as well as shear failure through intact rock bridges along discontinuities These mechanisms are illustrated in Figures 2.8b and 2.8c. It should be noted that in all step path mechanisms the path of least resistance can be ex pected to involve the minimum of failure length through intact rock. ?Curvy?, ?Wavy? or Undulating Planar Slides The presence of waviness or undulations on a potenti al plane of instability results in a shallower angle of effective dip of the failure mass compared with the failure surface. The effect of waviness should only be taken into consideration if the integrity of the sliding rock mass can be assured. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-13 Planar Failure of Thin slabs In instances where slope angles are parallel the dip of bedded rocks and in particular where relatively low strength rocks (such as sandston es) are interbedded with weaker horizons such as coal seams, the result is the formation of long thin slabs of rock which can destabilise as a result of; ? planar sliding with cross joints ? compression failure of intact rock at base of slope ? hydrostatic uplift ? buckling These instability mechanisms are illu strated in Figures 2.9a to 2.9f. 2.11.3 Toppling Instabilit y Modes Toppling instabilities occur wherever a set of steeply dipping reasonably well developed discontinuities with moderate to close spacing occur sub-paralle l to the slope (Piteau & Martin, 1982). Hoek and Bray (1981) summarised the basic principles for sliding and toppling failure of an individual block and showed that the potential for toppling is essentially determined by the ratio of block width to height (b/h) and the dip of the base plane ( ?bp). Whenever the condition of b/h < tan ?bp then toppling can occur. It should be noted the basic principles of t oppling can be significantly influenced by external factors such as groundwater pressures, surcharge loading and support. Flexural toppling occurs when long thin columns of rock formed by well defined s ub-vertical discontinuities bend, or break in bending, as a result of loss of support or over steepening of the slope (Piteau and Martin, 1982). In addition to bending, toppling can also involve tensile and compressive failure of the rock columns as well as frictional sliding or interlayer slip between adjacent columns which must occur before large flexural deformations can develop. Goodman and Br ay (1976) differentiated between primary toppling and secondary toppling where they described primary toppling as occurring under the action of gravity and in situ stress and secondary toppling as occurring as a consequence of an influencing factor which could be an initial failure resulting from a different instability mode. Illustrations of primary and secondary toppling from Goodman and Bray (1976) are shown in Figures 2.10a to 2.10c and 2.11a to 2.11d. Goodman (1989) stated that if the steeply dipping la yers or joints forming rock ?columns? have an angle of friction of ? j then slip can only occur if the angle of applied compression makes an angle greater than ? j with the normal to the ?columns? (or colu mn poles). It the dip of the ?columns? is ? and the overall angle of the slope is ?, the precondition for inter-column slip is (90 ? - ?) + ? j < ?. In addition the strike of the layers or? columns? must be within 30 ? of the strike of the slope. Kinematic conditions required for toppling are illustrated in the stereographic plot in Figure 2.12. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-14 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-15 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-16 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-17 2.11.4 Wedge Instability Modes In the simplest form, wedge sliding instability takes place on an intersection of two planes and as such, the sliding block takes the shape of a four sided tetrahedron. Different combinations of wedge instability are (Piteau & Martin, 1982); ? Failure by sliding along both planes in a direction formed by the line of intersection of the two planes (see Figure 2.13) ? Failure by sliding along one plane and separation along another plane ? Failure by rotational sliding on one plane and separation across the other plane ? Failure by progressive ravelling of rock along planes forming the wedges in highly jointed rock. The presence of more than two discontinuities/joint sets can significantly increase the number of possible combinations and resulting lines of intersections of wedge failures, however the slope design will usually be influenced by only one or two ?critical wedges?. Evaluation of kinematically possible wedges is greatly helped by the use of stereographic plots as illustrated in Figure 2.14. This figure illustrates the line of intersection of two discontinuities and the minimum overall slope angle required to prevent undercutting of the wedges. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-18 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-19 2.11.5 Rotational/Circular or Rock Mass Instability Modes Rotational rock mass instability modes occur under the following conditions; ? the rock mass is sufficiently fractured, and/or ? the intact rock is sufficiently weak and ? the slope is sufficiently high. Instability surfaces and rock mass shear strength are made up of a combination of intact material and discontinuities. There are three principal types of failure; Rotational Shear Failures, Regular or Circular Failures These types of instabilities occur in relatively weak ductile homogeneous materials which can be considered to have relatively uniform strength properties such as soils and relatively soft unjointed or altered rock masses or highly jointed or fractured rock masses as illustrated in Figure 2.15a. It is theoretically possible that strong rock mass can fail via rotational mode provided the slopes are sufficiently high and the shear strength of the rock is exceeded. Rotational sliding can occur through a process of yielding and then rotation. When the materials are sufficiently plastic to yield without excessive loss of strength at locations of high strength, stresses become more evenly distributed along the potential sliding surface. Because of this yielding, the shear strength of the material is maintained even though some zones have become overstressed. Conditions for rotational sliding are; ? particle size is very small compared with the size of the slope ? particles are not interlocked The total driving force causing sliding is assumed to be greater than or equal to the total shear resistance along the surface of sliding. The strength of the material is generally assumed to be based on the Mohr-Coulomb failure criterion (in terms of effective stresses): ? = c? + (? - ?) tan?? (2.4) Shear Failure along an Irregular Surface Failure along an irregular surface may be partly controlled by lithology and/or a major structural feature as illustrated in Figure 2.15b. Shear failure along the section of the rock mass/structure interface or within the structure would be controlled by the shear strength of that particular interface or structure. Block Flow Block flow occurs when a rock mass is not sufficiently ductile to permit rotational sliding and conditions are not amenable to failure along discontinuities. The rock mass essentially disintegrates causing a flow of broken rock as illustrated in Figure 2.15c. Overall slope stability is controlled by interparticle structural equilibrium and kinematics. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-20 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-21 2.11.6 Higher Order or Compound Instability Modes In addition to the basic instability modes summarised above, there is a whole range of higher order or compound modes of instability which can involve interaction of any two or more of the basic instability modes. Examples of these include; ? Active-passive sliding blocks (Calder & Blackwell, 1980) detailed as a case study in Section 3.2. ? Ratchet mechanisms (Martin,1993) detailed as a case study in Section 3.4. ? Laccolith slides (Stewart & Reid, 1980) detailed as a case study in Section 3.9 ? Slide toppling mechanisms (Sijing,1980) detailed as a case study in Section 3.11 These mechanisms are described in further detail in the literature review case studies in Section 3. 2.12 Gravitational Creeping Behaviour of Large Scale Rock Mass (Rock Mass Creep) Rock mass creep is defined as the relatively slow downward and outward movement of a mass of earth material on a slope without the formation of a rupture or failure surface (Radbruch-Hall,1978). This excludes soil creep, debris flows and glacial action. Different terminology has been applied to large-scale gravitational creep of rock masses on slopes. These include "deep-seated large-scale rock slides " (Terzaghi, 1962), "depth creep" (Ter-Stepanian, 1969), "deep-seated creep" (Nemcok, 1972), "continuo us creep", in contrast to "seasonal creep", which occurs only in the top layer of the ground (Terzaghi, 1950), "deep -seated continuous creep" (Hutchinson, 1968), and "mass creep" (Terzaghi, 1953; Skempton and Hutchinson, 1969). Measured velocities of rock mass creep range from 1.78 cm per year (Huffman et aI., 1969) to 20 cm per day (Muller, 1968). According to Nemcok (1972), creep may cause deformation to depths of 250-300m. Rock mass creep often occurs in situations where relatively hard, rigid rocks overlie or are interbedded with softer, visco-plastic rocks. However, it may also occur in relatively homogeneous rocks. It is most common on steep high slopes, particularly in mountainous areas, but is also known to occur on flatter slopes of only a few degrees. (Radbruch-Hall,1978). Rock mass creeping instabilities of mountain slopes, especially in the Alpine region of Europe, have been a well known and well researched phenomenon. Causes of the landslides were researched in the early 1900?s, but it was Terzaghi (1950) who first accounted for the mechanisms involved. General mechanisms of deep creeping instabilities were summarized by Ter-Stepanian (1966). Ter-Stepanian (1966) showed that the stress distribution in large scale creeping instabilities is not uniform but is dependent on many factors such as geometry, rock types and orientation of lithology, residual stresses and groundwater. The rate of creep deformation is dependent on time and mobilized shear strength which is not low enough to allow shear failure. Ter-Stepanian (1966) categorized the creep deformation mechanisms in rocks into three categories, planar, rotational and general. Planar creeping mechanisms, which occur on long slopes where the varying strata dip parallel to the slope, can be divided into 5 subdivisions; creeping of slabs, consequent creeping, creeping away of blocks, outcrop creeping and terminal creeping as illustrated in Figures 2.16a to 2.16e. Rotational creeping mechanisms occur on relatively short slopes where the rock mass is relatively weak and homogeneous. Two subcategories of rotational creeping are asequent creeping and s-like creeping illustrated in Figures 2.16f and 2.16g. General creeping PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-22 mechanisms are more complex, but two subcategories, those of insequent creeping and compensating creeping were identified by Ter-Stepanian (1966). Insequent creeping takes place where overlying competent strata are deformed as a result of the large possibly plastic deformation movements of weak underlying strata (Figure 2.16h). Similarly a compensating creeping mechanism occurs due to vertical rebound effects in a weak underlying stratum caused by unloading due to river erosion. The rebound effectively breaks up overlying stronger rock mass causing further subsistence of the sides of the river valley (Figure 2.16i). Large scale deformation behaviour of rock slopes was investigated by Zischinsky (1966) in the Austrian Alps. Zischinsky (1966) noted that creep (or flow) of rocks was such a significant factor in the deformation behaviour of large slopes which he investigated that the presence of basal failure surfaces was not necessarily a requirement to ensure a high component of mobility. Zischinsky (1966) reported on 4 large scale instabilities which he had investigated, two of which are discussed below. Matrei-Glunzerberg (Eastern Tyrol) Instability This instability has a slope height of approximately 1100m to 1200m and an overall slope angle (OSA) of 24?, see Figure 2.17a. The rocks are phyllites, calc-phyllites and calc-schists. The regional orientation of the planes of schistosity (s-planes) is 50?/180? (dip/dipdir) very uniformly for several kilometers along strike. Below 1750m elevation the s-planes rotate plastically with a dip 30? to 60? in towards the slope, bending back towards the regional orientation at the base on the western side of the instability. In addition to the deformation of the s-planes there exist shear planes within the deformed rock mass towards the top of the instability. These planes do not cut through the instability but end within it. These shear planes are effectively a ?failure within a failure?. There are therefore two types of deformation occurring within the overall instability. The first is related to the development of limited length shear planes and the second is a viscous plastic deformation of the s-planes down the slope. This combined type of movement is characteristically termed ?sackung? which can be defined as a situation where ?the amount of continuous deformation is large in relation to the displacement along a basal sliding surface?. Millst?tter Alm (Radenthein, Carinthia) Instability This is a very similar mechanism to the Matrei instability discussed above and is illustrated in Figure 2.17b. This can be summarized as internal rotation of s-planes below a higher zone of relatively brittle deformation along shear planes dipping into the instability. It is of interest to note that in instabilities of these magnitudes the moving mass of material is usually restrained from moving within the base of the valley. Continual erosion by rivers of the material at the toe of these instabilities provides room for continuous but slow down slope deformation. Different mobility components ? towards the top of the slopes deformation was much more brittle (solid), whereas in the lower reaches of the slope the deformation is more viscous (plastic). Zischinsky (1966) attributed this to distribution of stress within the slope and to the mechanical properties of the rock. Zischinsky (1966) concluded that the viscous behaviour or continuous deformation behaviour of the rock mass can best be explained by a power function as follows; ba ?? .=& (2.5) Where?& is the rate of strain, ? is the shear stress, a and b are constants. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-23 This power function is typically used to describe an imperfect viscous liquid and describes the behaviour of other materials such as snow and ice. In terms of the discontinuous deformation behaviour (shearing), Zischinsky (1966) applied the Mohr-Coulomb failure criterion. Essentially it is proposed that the larger the shear stress (?) the larger the deformation however the deformation itself then results in the reduction of the shear stress. It was therefore concluded that the behaviour of a rock mass is equivalent to that of a viscous liquid and that other properties such as creep can be assigned to the rock mass. Total displacements or strain rates of the creeping instabilities were not reported but from scaled drawings it would appear that this is of an order of tens of metres. Details on the hydrogeology were also omitted. A further feature associated with long term creep deformation in valleys is that of cambering and valley bulging. This results from the inwards movement of valley walls especially where stronger strata overly weaker units, which results in the squeezing together of the rock mass in the floor of the valley. This causes buckling, heave and deformation of strata at the centre of the valley. Lapworth (1911) documented this behaviour in several dam sites in England. Hollingsworth et al (1943) who made a detailed study of gravitational creep in the Northhampton Ironstone Field, Rutland and South Lincolnshire, described ?cambering? as the slight convex arching of strata on ridges, and ?valley bulges? and the anticlinal uprises on valley floors and fissures parallel to the slopes as ?gulls? or tension fractures. These features were created by inward (valleyward) movement of soft strata underlying more competent rocks. It was reported that these structures can have over 30m of structural relief affecting areas as large as several square kilometers. There are a considerable number of documented occurrences of other localized rock mass creep and associated natural landslip features but detailing them is beyond the scope of this research. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-24 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-25 Dodds (1966) describes a fairly unique study undertaken of the frictional stability of rock slabs along stress relief fractures and open joints in a canyon wall in south-western Iran. The failures and failed material consisted of a singe rock unit termed the upper Bakhtiari Conglomerate which is a rock made up of a fluvial deposit of boulders, cobbles and gravel with intercalated sandstone lenses. Calcite is the principal cementing agent in the rock. The stress fractures developed as a result of elastic rebound of the rock resulting from rapid erosion and excavation. A plot of both the horizontal and vertical movement during a typical failure event lifetime showed that each of the failures did not proceed to collapse smoothly, but that the process was intermittent. Series of smaller failures were separated by ?quiet? periods where little movement took place, but where energy of the failure mass was used to ride up and shear off protrusions along the failure surface. When the shearing resistance of the protrusions was less than the shearing force then localised failure occured. The rock mass then moved again until it came into contact with another protrusion. The fracture was thus observed to open and close as protrusions were systematically sheared off. This process gradually resulted in an accelerating loss of shearing resistance with time, finally leading to total collapse. It was concluded that the amount of opening and slippage a rock slab can undergo before collapse is a function of the size of the irregularities on the failure surface, the inclination of the surface and the yield point of the rock. The most critical factor was the rate of movement, which showed that once a peak rate of 0.1mm/day had been reached, final collapse was imminent. The coarse composition of the conglomerate enabled observations of macroscopic failure processes, which in most rocks occur only microscopically. 2.13 Rock Mass Time to Failure Predictions Zavodni and Broadbent (1980) present the findings of a research effort made into the study of the kinematics of large scale slope failures in open pit mines. They recognized two principal failure stages which they termed a regressive stage and a progressive stage. The point at which the slope changes from regressive to progressive was termed the onset-of-failure (OOF) point. The regressive stage is characterized by failure deformation that can re-stabilize should external disturbances be removed. The ?onset-of-failure? point defines the stage at which the velocity of the slope would start to undergo continued acceleration (progressive behaviour) finally leading to collapse if no further active remedial measures are implemented (Figure 2.18). They concluded that the rate of movement measurement should be used as it ?provided the most sensitive indication of slope movement?. Zavodni and Broadbent (1980) examined the displacement records of 13 failures in open pit porphyry copper mines which included planar fault and joint, wedge and rotational failures. They derived composite displacement rate vs time plots for the data and from this derived an empirical equation for a line fit on a log-normal chart of the form steCV .= (2.6) KV V o mp = (2.7) Where; PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-26 V=velocity, s=slope of line (days-1), C=constant, t=time (days), e=base of natural logarithm, Vmp=velocity at mid-point in the progressive failure stage, V0=velocity at the onset-of-failure point, K = constant. The following equations can be used to predict the velocity at the onset of collapse by substituting V0 =C at the onset-of-failure point; steVV .0= (2.8) The velocity at the collapse point can then be predicted using the following equation Vcol = K 2 V0 (2.9) The mean value of K was determined to be 7.21. Zavodni and Broadbent (1980) concluded that the onset-of-failure point is almost impossible to predict while the slope is in regressive behaviour. They went on to suggest that the onset-of-failure point is likely to be influenced by an aspect of ?slope fatigue? which is related to the number of external impulses, the breakup of the rock mass and associated strength reduction and soil pressures developing in the crest tension cracks. They also concluded that the OOF point will occur between 4 and 45 days prior to collapse and that efforts to link the mass of the failure to the time from Vo to Vcol were not successful (Broadbent & Zavodni, 1981). PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-27 Goldberg & Frizzel (1989) used this approach to estimate time-to-failure (tf) for instabilities in the Berkeley open pit. They showed that for the failures in the Berkeley pit which occurred in saturated coarse alluvium overlying coarse oxidized granitic rock, the constant K was equal to 6.5. Cruden and Masoumzadeh (1987) evaluated 4 different creep laws namely, Saito?s laws (Saito, 1969 and 1980), exponential laws (Varnes, 1982), power laws and Zavodni and Broadbent laws (Zavodni and Broadbent ,1980) for the Luscar Mine failure described in Section 3.3. They found that in the two stage creep model the power and exponential laws underestimated displacement rates close to failure and instead showed that a 3 stage creep model using a generalized Saito law was the most accurate descriptor of the deformation data for the Luscar case study. 2.14 Review of Displacement Monitoring Techniques A comprehensive monitoring program is considered essential for facilitating relatively safe mining conditions in open pit mining operations (Call, 1981). Common monitoring methods in use include automated high precision survey networks, laser scanning, wire extensometers, borehole inclinometers and more recently slope radar systems. The survey networks and radar systems are reviewed in further detail below. High Precision Survey Automated high precision survey network monitoring systems are now in common usage throughout the open pit mining industry. These systems incorporate robotic theodolite stations, prism networks and high precision GPS systems, and use conventional survey methods such as triangulation and trilateration to continuously calculate deformation movements. The survey data is automatically saved in electronic databases. These systems are flexible in that they can be used to monitor sections of a slope continuously or larger sections on a programmed cycle. They are cost effective and can provide close to real time monitoring of the pit slopes. Slope Stability Radar (SSR) Slope stability radar is a relatively new technology introduced to provide real time displacement monitoring of rock slopes. The technology has several advantages and disadvantages as summarised below. Advantages 1) 24 hours, 7 days a week continuous monitoring coverage gives real time warning of rock fall and collapse, thereby reducing risks to personnel and equipment operating in areas of concern or other designated high rock fall risk areas. This means that restrictions on working hours necessitated by survey monitoring and visual observation can be relaxed. 2) Accurate detection of movement over the complete area of concern with the capability to detect zones of instability that may not be covered by a prism monitoring system and may not be evident from visual observation. This means that equipment withdrawn from localized areas of instability could often be deployed elsewhere in the area of concern. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-28 3) Improved productivity through extending operating hours beyond daylight and reducing down time by eliminating subjectivity in allowing re-entry following a withdrawal period. 4) Improved understanding of the overall behaviour of the slope, in particular its time-related performance and response to blasting. This can possibly lead to improved slope designs with the potential for improved and early ore recovery. 5) Additionally the technology has no requirement for targets. Radars can be deployed rapidly to problem areas, alarms can be set to alert operators to deformation exceeding threshold values and the technology is now proven. Disadvantages 1) There is no indication of the accumulated amount of deformation that had taken place in all the time leading up to collapse. Each time an SSR scan commences the displacement automatically starts from zero and there is no indication of previous displacements. Any predictive criteria based on deformation therefore is inappropriate 2) Measurements taken are along an azimuth between the object and the SSR unit which may not be at right angles to the slope. 3) The in-pit locations of the SSR units may be undergoing deformation themselves. Measured displacements and displacement rates are therefore relative. SSR units cannot currently orientate themselves onto survey control networks around the pit and therefore fix their own grid positions. 4) Cost and reliability. 2.15 Summary of the Literature Review In fresh or hard rock slopes, instability occurs along structural discontinuities such as bedding, joints, geological contacts, faults etc. Failure seldom takes place through intact rock unless it is very soft or forms a rock bridge between discontinuities. Consequently the most important factor influencing rock slope behaviour is structural discontinuities. It may be concluded that the majority of rock failures are structurally controlled. Instability of highly fractured and weathered rock mass display characteristics similar to soil type ?mass failures?, however in most instances they are still influenced by relic structures. Rock type, in situ stress, position of phreatic surface and slope geometry also influence instability and failure behaviour. Jointed rock basically consists of time-dependent as well as time-independent resistance to deformation. However it is important to consider the relevance of the ?time factor? in considering any potential reduction in the rock or joint strength within the time frame of the required active mining life of the slope. In this regard the deformation of intact hard rock occurs over such a lengthy time scale in comparison to that of the limited required life of the slope that it can potentially be considered as insignificant. In contrast, the time dependent behaviour of discontinuities is considered significant where the potential creep deformation time leading to failure can be well within the required length of PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 2 : Literature Review Page 2-29 slope life. It has also been shown in studies on the time dependent deformation behaviour of discontinuities that the magnitudes of the peak and residual shear strengths of the joints are themselves time related. No two failures are the same. The characteristics of instabilities are site specific. Characteristic time dependent behaviour graphs for accumulated displacement and displacement rate were developed by Martin (1993). The problem is that this characteristic behaviour description offers no ability to predict failures or even characterise different patterns of behaviour leading up to failure. The Author concurs that this classical behaviour pattern is very generally typical of deformation and behaviour leading to collapse in larger excavated rock slopes and believes that it should be investigated in much further detail over a broader and much larger range of more common case studies. Various types of time dependent deformation of geological features have been recorded. These include large scale valley rebound and up warping of valley rims due to rapid erosion or glacier melting, deep seated continuous creep deformation of entire mountain sides commonly observed in the European Alps and rebound of excavations and foundations for large structures such as dams. Numerous small and medium scale landslides and slope instabilities have occurred throughout the world. In the next section, selected literature survey case studies are presented. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-1 3 LITERATURE REVIEW CASE STUDY DATABASE Published case studies of slope failures associated with mining operations were specifically identified for review as further sources of deformation data. The principal objective of these case study reviews was to obtain actual deformation data of failures that have occurred and to review the data in the context of the modes of instability and failure that were reported. The intention was for this information on the deformations associated with the failures to supplement data obtained from the detailed case studies in order to form a more comprehensive overall picture of deformation behaviour. Thirty literature case studies of documented failures which have occurred both in natural slopes and in a mining environment were assembled. From these thirty case studies a set of nine was selected to be individually summarised and presented in this Section. The selection was based on the following criteria; ? the case study should document a relatively large or significant failure within an open pit, ? the case study should containing well documented time-dependent deformation data and associated geological and structural data, ? the overall selection of case studies should cover a broad scope of possible deformational and failure mechanisms. No specific limit to the number of literature case studies used was originally set. In the case study descriptions in the following sections, the units used in the publications have been retained. Table 3.0 summarises the case studies selected together with their associated failure mechanisms. Table 3.0 : Summary of Literature Survey Open Pit Case Study Database No. Case Study No. Failures Reviewed Type of Failure 1 Brenda Open Pit, British Columbia, Canada 1 Active-passive sliding blocks 2 Cassiar Mine, Canada 1 Complex ratchet mechanism with sliding wedge and combination of sliding toppling in lower blocks. 3 Luscar Mine, Alberta Canada 1 Large sliding failure along bedding and through rock mass at toe. 4 Telfer Gold Mine, Western Australia 2 Step path failure and deep seated failure mechanism with shear failure across the bedding planes. 5 Afton, Canada (1984) 6 Wedge, rock mass, circular, laccolith slide and toppling type failures 6 Jinchuan Mine, China (1976) 1 Compound slide-toppling failure 7 Hogarth Open Pit, Ontario, Canada 1 Classic toppling failure 8 Inspiration?s Mines, Nevada, USA, (1953-1980) 16 Predominantly wedge and planar shear failures 9 Nchanga Failure, Zambia 1 Compound PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-2 3.1 Case Study 1, Brenda Open Pit, Canada, Northwall Failure Geology and Lithology Brenda Mine is situated in British Columbia, Canada. The orebody occurs in a fractured Jurassic rock mass referred to as the Brenda Stock which is approximately 3000 ft in diameter and 1000 ft deep. The Brenda Stock has been intruded into the Upper Triassic Nicola Group tuffs, breccias and argillites and includes four gradational units of quartz diorite plus an additional partly discordant unit termed the fine quartz diorite. These rocks have been extensively fractured by compressional forces with the mineralization restricted to fracture fillings (veins) within the diorite host rock (Soregaroli, 1974). Shearing, faulting and jointing occurred later in many stages, producing mineralized clay gouge zones up to several feet in thickness which have a major effect on the stability of the pit walls. The structural geology is relatively simple. The major structural joint system (J 1) has an orientation of 70?/180? (dip/dipdir) but is not undercut by t he pit slope. Bedding planes have an orientation of approximately 30?/0? (dip/dipdir). Two major infilled fault zones referred to as Gouge A and Gouge B strike parallel to the bedding at approximate orientations of 50?/180? and 70?/180? (dip/dipdir). The gouge zones are filled with a soft molybdenum-clay mixture with the consistency of a soft lubricant for which no reliable magnitude of friction or cohesion can be measured. OSA (not including haulage roads) is 45 degrees (Calder and Blackwell, 1980). The primary displacement monitoring system consisted of a Geodimeter 710 with prisms (Blackwell, Pow and Keast, 1975). The haul road leading to the shell 2 (cut 2) intersected gouge zone A on the north wall with an approximate orientation of 50?/180?(dip/dipdir). The intersection passed along the road but did not daylight in the pit wall. Large amounts of water were encountered during excavation through the gouge. Displacement monitoring at the time showed that that movement was being influenced by blasting. Mining proceeded normally until during Ju ne 1978 the outside of the haul road underwent a vertical differential displacement of 2 feet. Mining was temporarily stopped and in September 1978 the initial settlement was followed by a differential settlement of 2 feet following a period of heavy rainfall. A plot of the vertical movement for station 112 which was situated on the haul road (crest of failure) is shown in Figures 3.1a. Failure Mechanism The failure mechanism identified by Calder and Blackwell (1980) is illustrated in Figure 3.1b. The moving rock material is divided into an active and a passive block. The blocks are demarcated by three structures being Gouge A and B as well as the bedding plane which acts as the base sliding plane. It was clear from the monitored displacements that the active block was moving down the 50? dip of the Gouge A plane whilst the passive block was being pushed up the 30? bedding plane. Additionally, the magnitude of movement of the passive block was being controlled by the magnitude of movement of the active block as illustrated in Figure 3.1c. As the molybdenum-clay gouge fill was impermeable to water it was assumed that the full water pressure distributions were acting on the moving blocks. Calder and Blackwell (1980) concluded that the structure together with the water pressure and the effects of blast generated vibrations were the factors which controlled the failure. They calculated the factor of safety to be 0.78. Calder and Blackwell (1980) went on to develop a relationship between acceleration due to blasting and scaled distance as shown in Equation 3.1. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-3 022.1 9.295 ? ??? ???? ?= W DA 3.1 Where A : peak acceleration level (in/sec2) D : distance from the blast (ft) W : weight of explosives (ANFO) in lb D/W0.5 : scaled distance From this relationship Calder and Blackwell (1980) showed that there was a significant decrease in the FOS with decreasing scaled distance. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-4 Discussion Two principal remedial measures which were implemented, were dewatering/depressurisation and limiting the scaled distance of blasts. These measures limited movement and prevented collapse and enabled the recovery of all the ore reserve. At no time did the rock mass break up as a result of the continued movement. The sensitivity of the moving block to blast vibrations is most likely a result of the low FOS. Once again the progressive displacement behaviour which took place in June and September changed into a creeping and finally regressive state. This can be attributed to internal depressurisation of the slope as a result of large displacements. Very limited displacement data are given and therefore there are limited comments on displacement behaviour. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-5 3.2 Case Study 2, Cassiar Mine, Canada, North East Sector Failure Cassiar Mine is located in the middle of the north trending Cassiar Mountains in British Columbia, Canada and is bounded on the east side by the Liard Plain and on the west by the Stikine Plateau (Hewett, 1978). Geology and Lithology The orebody occurs in a sill of highly fractured and blocky serpentinized peridotite which intrudes the west limb of a syncline formed in Paleozoic argillites, dolomites, quartzites, and associated rocks of the Sylvester and Sandpile groups. The Sylvester Group rocks which form the hangingwall consist mainly of fine-grained, massive, reasonably competent argillite with minor interbedding of volcanic flows which occur at irregular intervals. Both the argillite and the volcanic rocks are generally hard, except in close proximity to fault zones where the strength may decrease considerably. The periodotite sill is roughly tabular, strikes north-south and dips 30 to 45 degrees to the east. The contact between the intrusion and the enclosing sediments is demarcated on the hanging wall by a 10- to 15-foot wide indurated zone of hard, tough hornfels alteration. The argillite and related rocks of the Sylvester Group in the hanging wall have a very uniform structural pattern and only minor variation occurs in orientation across the hanging wall. In general, the argillitic and volcanic rocks appear to have similar structural characteristics (Piteau and Martin,1977). Structure 1st order structure Several major faults exist which are mostly normal to the slope and sub-parallel to joint set A (JA). Their average orientation is 68?/356? (dip/dipdir) and most are continuous. Shears are related to faulting and have an average orientation of 65?/350? (Piteau and Martin,1977). 2nd order structure: Cross joints are prominent across bedding planes, whereas bedding joints are not well developed. Bedding strikes obliquely to the slope and dips away from the pit with an average orientation of 26?/135?. There are three principal joint sets. Their orientations as measured from their peak concentrations, are JA (72?/348?), JB (78?/168?) and JC (80?/253?) (Piteau and Martin,1977). The stereographic projection of all the structures are shown in Figure 3.2a PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-6 Groundwater Confined mostly to fault zones and related jointed/fractured rock. Annual rainfall averages 305mm together with 3950mm of snow. Strength properties of Rock The strength properties are summarised in Table 3.2. Table 3.2: Estimated Rock Mass Strengths (Aft er Martin and Carew, 1986 and Martin, 1993) Rock Mass Unit weight (kg/m3) Friction Angle (degrees) Cohesion kPa UCS (MPa) RQD (%) Argillite and volcanics 2720 32 690 80 73-96 Waste serpentine 2580 26 345 35-50 53-81 Ore bearing serpentine 2580 26 207 35-50 53-81 Shear zones 2450 26.5 69 < 8.5 < 50 Slope Stability Assessments The finding of initial stability assessments carried out by Piteau and Martin (Piteau and Martin,1977) concluded that due to the orientations of the structures the possibility of deep-seated failure of the hanging wall was small, however kinematic analyses revealed that bench scale wedge failures were of principal concern. However, a later review of the stability (Martin and Carew, 1986) showed that the Phase 10 or ultimate pit slope configuration was only marginally stable against deep seated rotational failures and would require both slope depressurisation and modification of slope geometry in order to ensure the stability of the slope to closure. Description of Failure (after Martin ,1993) In 1988 during the mining of the final slope, instability was identified in the north east sector of the pit. Mining was accelerated and ore stockpiled during the Winter of 1988/1989 when movement rates had slowed. During the following spring thaw and summer of 1989 movement rates increased significantly, with rates of greater the 125mm/day being recorded and differential displacements on the scarp of the failure greater than 5m. Figure 3.2b illustrates the typical movement response for a prism near the crest of the main area of instability during this period. The pit was closed and reopened briefly during the winter of 1990 to recover remaining mineable ore. Failure mass exceeded 17.8 million tons (5.9 million m3). Martin (1993) used the geological mapping and prism monitoring to interpret the failure mechanism. The initial failure is bounded by the Serpentinite dyke on the east, by the North Fault on the west and by sets of parallel faults termed A1, A2 and A3 on the south as illustrated in Figure 3.2c. Martin?s field investigations also revealed the presence of several steeply dipping cracks which were termed Cracks B, C, D, E and F. These cracks appear to have been caused by the failure of the rock mass along an irregular surface formed by the intersection of joint sets as described above. Failure mechanism The failure mechanism is complex and has been summarised as concisely as possible from Martin (1993) and illustrated in Figure 3.2c. The average orientation of the pit slope in the northeast sector of the pit was approximately 38?/240? (dip/dipdir). The upper section of the slope which formed the initial failure zone (or graben block) was not buttressed by the pit wall, and hence it was free to move at an oblique angle to the slope along a line of intersection formed between the Serpentinite Dyke and Fault A1 at a strike of 308? and a dip of 52?. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-7 This orientation is consistent with the movement vectors of prisms in the upper part of the slope which reflected an average azimuth of 271 to 315? and an average inclination of 40 to 54? a shown in Figure 3.2d. Further north westward movement was prevented below the 6060ft level by the pit wall to the west of the North Fault. As a result movement swung out of the wall along the north fault. Theoretically, a sliding wedge type failure of the initial block is not kinematically possible. Nevertheless the shearing along Cracks B, C, D and E, which form the boundaries of large blocks, was most likely initiated as a result of additional stress placed on the lower rock mass by the sliding movement of the initial block. These cracks together with possible squeezing of serpentinite in the toe facilitated the formation of a deep seated toppling mechanism. The progressively shallower plunge angles of the prism movement vectors in the lower section of the active failure area indicated that a combination of sliding and toppling occurred in the lower blocks. The plunge angles of prisms were steeper than anticipated because the blocks were not deforming as rigid blocks. Instead, internal crushing, shearing and squeezing along the edges of the blocks is likely to have occurred. The consistent patterns of movement of the prisms indicated that the postulated failure mechanism continued without significant changes. The increased rates of movement during spring thaw suggested that the failure blocks were sensitive to changes in the overall phreatic surface as well as the buttressing effects of debris accumulating on the lower slope. During spring thaw, recorded movement of surface extensometers indicated that slope movements varied on an approximate time cycle. Slope movements appeared to increase under the influence of high water pressure. Movement of the blocks resulted in drainage and associated depressurization of the slope, which reduced the movement rate until water pressures could again build up due to groundwater recharge. This resulted in a "ratchet mechanism" of movement rates which varied from minimal movement to movement rates in excess of 100mm/day in a cycle of three to five days. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-8 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-9 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-10 3.3 Case Study 3, Luscar Mine, Pit 51-B-2 Northwall Failure, Alberta Canada Geology and Stratigraphy The north wall failure of Pit 51-B-2 which occurred in November 1979 is a classic example of a planar sliding failure. The Luscar coal mine is situated in the Rocky Mountains of Alberta, Canada. The displaced mass was 245 metres in length, 106 metres high and 1.07 million cubic metres in volume. Sixteen displacement prisms were monitored using successive EDM readings. The stratigraphy of the Luscar mine is made up of Lower Cretaceous coal-bearing rocks consisting of interbedded sandstones and siltstones which dip into the pit with dip and dip direction of 38/204. Ground water elevations at the time of failure were approximately 50 metres below the crest of the slope (Cruden and Masoumzadeh, 1987). The instability in the north wall resulted from a weak zone dipping subparallel to the wall intersecting overturned beds at the toe of the wall. Moderate to steeply dipping sandstone strata, together with overlying siltstone and shale strata formed five moving blocks. Two major directions of movement were identified; nearly parallel to the strike of the bedding and down the dip direction of the bedding. The latter movement was much larger than the former. A common plane of movement dipped 28? to the west. The geology and geometry of the slope is illustrated in Figure 3.3a. No data regarding rock strength parameters, structures, classifications or monitoring records were included in the paper. Description of Failure Deformation Behaviour The following description of the deformation movements was summarised from Johnson (1982). An initial displacement of 30mm was first noticed on May 24, 1979 from the results of the weekly monitoring program in the pit. The weekly monitoring frequency was maintained for the next five months with the movement continuing at an average rate of 30mm per week. On October 29, there was a jump of 60mm in one day. Then on the morning of November 10, 1979, there was a further jump of 0.34 metres horizontally and 0.15 metres vertically from the previous day's readings. Following this, monitoring proceeded on a half hourly basis. By five o'clock that same day, the movement had some what stabilized to a rate of 0.06 metres per day. The total movement for the eight hour period was 2.65 metres horizontally and 1.52 metres vertically. Remedial action in the form of establishing three de-watering wells and a buttress berm was implemented immediately. The construction of the buttress took two months and was completed while under continuous monitoring of the wall. The wall movement continued at the rate of approximately 60mm per day for five weeks then slowed to 30mm per day and stopped moving on January 20, 1980 upon completion of the buttress. The total movement of the failure was 6.85 metres horizontally and 4.05 metres vertically over an eight month period. A time series plot of one of the prisms situated on the failure is shown in Figure 3.3b. Discussion There are several intriguing aspects to the description of this failure. ? Pre-collapse deformation occurred in intermittent sudden movements. ? Typical of a sliding failure the pattern of x and y displacement mirrored each other after commencement of sliding but not before. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-11 ? The ratio of the magnitudes of horizontal and vertical components of movement corresponded exactly to the average dip of the plane of movement. ? After the main 10th November slide (?failure?) of 2.65m horizontally and 1.52 m vertically the displacement behaviour suddenly decelerated to a creeping behaviour enabling remedial action to be taken by the mine staff which finally brought the movement under control. ? This sudden change to a creeping type behaviour was most likely a result of the relief of groundwater pressures in response to the changing internal geometry of the slope. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-12 3.4 Case Study 4, Telfer Gold Mine , Australia, Pit 1A Highwall Failure Geology, Lithology and Structure The Pit1A highwall failure at Telfer Gold Mine was described by Dankert (1994). Mining takes place in the Proterozoic sediments (reefs) of the Yeneena Basin which are associated with two doubly plunging antiforms or domes named the Main Dome and West Dome. The host rock consist of a suite of siltstone and sandstone members which dip approximately 35? to the east. A typical geological section is shown in Figure 3.4a. No structure/discontinuity orientations, rock mass strengths or phreatifc surface levels were provided. Description of Failures Following two independent geotechnical evaluations of pit wall stability, further geotechnical advice and a geotechnical review program, a decision was taken by the mine to deepen the pit and steepen the highwall bench face angles from 61? to 78? in order to achieve an average slope angle of 75?. It was also recommended by consultants to use cable bolting for additional reinforcement. Following approximately 2 years of further mining the highwall of Pit1A collapsed in July 1989 with little or no visible warning. The total mass of failed material was estimated to be between 150 000 to 200 000 tons. At the time of failure the displacement monitoring system had been newly installed and was not yet fully operational. A section through the failure is shown in Figure 3.4b. As part of the safety action plan the displacement monitoring system was expanded to include additional extensometers, an automated robotic survey system and additional prisms. In October 1992 the highwall experienced a second large failure. This failure was preceded by the following events: ? September 1991: slight increase in the rate of movement from <0.05mm/day to between 0.05 and 0.1mm/day. ? April 1992: tension crack developing 20m behind crest of highwall ? June 1992: More tension cracks up to 160m behind crest ? 5 October 1992: A rockfall of 600 tons ? 25 Oct 1992: A further rockfall of 200 tons adjacent to the 5 Oct fall. ? 29 October 2006: Rate of movement increase identified ? 31 October 1992: 180 000 ton failure occurs on full slope height Slope movements of prisms 19 and 20 which were located at the crest of the failure are shown in Figure 3.4c. Mechanism of Failure The mechanisms of the two failures differ. It appears that the base of the July 1989 failure was a step path failure behind the cable bolts along a relatively narrowly spaced joint set which dipped into the pit. The crest of the failure appears to have been initiated in the 450mRL to 480mRL bench which was not supported by anchors. It appears to be a relatively shallow failure mechanism. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-13 The October 1992 failure appears to have been initiated by subsidence at the 400m level as a result of underground operations which destabilised the entire slope. Cracking at 160m behind the crest of the highwall suggests a deep seated failure mechanism with shear through the bedding planes. Discussion The movements of the prisms for the Oct 1992 failure confirm that collapse occurred after very little increase in the rate of movement. Even though the pending Oct 1992 failure was identified and better monitored as a result of more visible signs of slope distress (small rock falls and cracks at the crest of slope) the slopes in both failures collapsed catastrophically. The step path failure mechanism was located behind the cable bolts and involved an unfavourably orientated closely spaced joint set dipping into the pit. Lack of further discontinuity and rock mass strength parameters prevents a more detailed analysis. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-14 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-15 3.5 Case Study 5, Afton Mine, Ca nada, Wall Instabilities and Failures The Afton copper deposit is located 13 km west of Kamloops, British Columbia. The mine is structurally and lithologically complex. The geology and the description of the failures were detailed by Stewart and Reid (1988). At the time of writing the publication the pit had reached 700 feet below natural ground level (Stewart and Reid, 1988) . The north wall of the pit has an OSA of 40? and the rest of the pit is 45?. This case study illustrates how many different types of failures can occur in a single pit. Geology and Stratigraphy The Afton orebody occurs within a Triassic volcanic and plutonic diorite body. Tertiary sediments and volcanics have been faulted into a position along side the diorite body. The northern third of the pit has been mined in sandstones, mudstones, shales, tuffs and arkoses (Tertiary rocks) which in turn have been over and in filled both conformably and unconformably by younger volcanics, the most common of which is dacite although andesites and latites occur as well. Intense folding and faulting of the rocks has added further complexity. Rock strengths vary greatly, with UCS?s ranging from 207MPa for dacite, to between 3.45MPa to 10.3MPa for mudstones (Stewart and Reid, 1988). The intrusion hosting the orebody is made up of diorites and micro diorites with some syenites and older volcanics. Abundant calcite, chlorite, epidote, feldspar saussuritization, and pisolitic alteration to chlorite in basalt all indicate that alteration has been significant. Various diorite subunits have been identified, with gray diorite generally being harder and more coarsely jointed than brown diorite. UCS values of the diorites range from 20.6MPa to 110.3MPa (Stewart and Reid, 1988). Faulting within the pit has been extensive with major through going faults usually forming lithological boundaries. Significant changes in fault strikes and dips occur over relatively short distances and clay gouge infilling, which can result in low friction angles, is common. A layout of the geology is shown in Figure 3.5a. Description of Failures and Failure Deformation Behaviour Northwest Sector Slide (Position 1 in Figure 3.5a) The 8th March 1984 failure measured approximately 200 000 tons. The failure was preceded by tension cracks on the crest of the pit. The instability was well monitored through to collapse. Figure 3.5b illustrates the rate of movements measured by an extensometer located on the crest of the slide. What is particularly interesting about this failure is that it was characterised by a continual gradual drop in groundwater all the way up until and after failure as shown in Figure 3.5c. The failure was a rotational slide failure as illustrated in Figure 3.5d. Haul Road North Wall (Position 2 in Figure 3.5a) The November 1984 failure consisted of a wedge type failure in tertiary sediments which was fault controlled and occurred directly below the haul road. The failure was stabilised by mechanical tiebacks. Arkose Wedges (Position 3 in Figure 3.5a) In the northeast corner of the pit multiple bench scale wedge type failures occurred in Arkose. The failures were structurally controlled in one direction by a fault and in the other by joints of limited continuity and orientation. The failures self stabilised. Laccolith Failure (Position 4 in Figure 3.5a) A dacite laccolith intrusion which was surrounded by wet soft mudstone started to slide out of the wall en masse after the bottom portion of the structure had been exposed by mining. Displacement PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-16 however reduced as the structure cracked internally. However despite slope flattening, dewatering and a buttress the structure continued to move especially after wet periods. Circular Arc Failure (Position 5 in Figure 3.5a) The steepening of the upper northeast corner of the pit activated a failure of over 1Mt in overburden and highly weathered sediments. The semi-circular failure was dimensioned by cracks 100ft behind the crest of the pit, a severed piezometer 130ft below surface and toe heave on the 2040ft bench. The failure is illustrated in Figure 3.5e. A FOS was calculated as 1.06. The failure ultimately self stabilised. Toppling Failure South Wall (Position 6 in Figure 3.5a) The south wall of the pit was considered as stable up until May 1984 when cracks were discovered behind the crest of the pit and on every bench to the base of the pit (see Figure 3.5f). According to Stewart and Reid (1988) the predominant discontinuity causing the mode of instability was orientated at 68?/210? dip/dipdir. An example of displacement movements for prisms situated on the benches is shown in Figure 3.5g which shows a slightly upward tending vertical displacement as the block rotates into the pit. Figure 3.5h illustrates a creeping pattern of movement for blocks which were kinematically able to displace downwards. Dewatering efforts in the form of drilling of drainholes was ineffective in improving stability. Reid and Stewart (1986) described the failure more fully and illustrated the extent of the instability (Figure 3.5i). They plotted the average rate of displacement over 30 days prior to failure and showed that the velocities were not all the same and that a central gully ?hotspot? existed within which 5 smaller failures took place prior to the main failure. The slope finally failed on 6th June. The cumulative displacement rates of prisms P1, P30 and P40 (positions shown in Figure 3.5f) are shown in Figures 3.5j and 3.5k. Cumulative displacement of extensometers E24 and E25 are shown in Figure 3.5l. Discussion The pit clearly illustrates the different types of failures that can occur in a structurally and lithologically complex pit. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-17 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-18 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-19 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-20 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-21 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-22 3.6 Case Study 6, Jinchuan Mine, China, Southwest Slope Failure Geology and Stratigraphy The Jinchuan Mine is situated in an igneous body of ultrabasic rocks surrounded by an older (Pre- Sinian) metamorphic series of migmatite, marble and schists. Strata strike parallel to the long axis of the pit. Several sets of faults and joints exist including a set of imbricated thrusts. The interbeds and lithological contacts form additional planes of weakness. The rock mass has a tabular blocky structure (Sijing, 1980). Description of Failure The failure mechanism developed in 4 stages (Sijing, 1980); 1) Slip Deformation of the Upper Part of the Slope During mining of the upper sections of the slope between 1964 and 1970 two sliding zones developed namely a ?drag sliding? zone and a main sliding zone. These sliding zones were controlled by fault planes which eventually opened up to over 0.2m. A large fault zone F34 below the sliding zones was then exposed. The compression of this fault zone provided the restraint release trigger which enabled the full mobilization of the sliding zones which then started slipping downwards applying further pressure on the lower slope (see Figure 3.6). 2) First Slide-toppling Deformation Between 1970 and 1974 the open pit was excavated a further 3 benches. A series of faults and strata contacts were exposed down the slope. Thrust from the sliding zones resulted in shearing along the faults and flexural bending in the rock (see Figure 3.6b). Resulting deformations were 7.00m horizontally and 7.18m vertically. 3) Second Slide-toppling Deformation Between 1974 and 1976 the open pit was excavated a further 2 benches. Continued sliding and toppling resulted in accelerated deformation with benches tilting over at an angle ranging from 15? to 20?(see Figure 3.6c). 4) Slide and Collapse of Overall Slope The general condition of the slope continued to deteriorate as a result of weathering and blasting. Ravelling occurred leading eventually to collapse. Failure Deformation Behaviour The deformation behaviour of the slope over its life is shown in Figure 3.6d. The curves show four distinct deformation stages. The first stage of the failure is characterised by linear creep of the rock mass which is followed by progressive behaviour associated with the first of the two major slide- toppling deformation events. The slope then decelerates and enters a stage of rapid creeping with horizontal movements of 4-10mm/day. Finally the slope begins to accelerate until failure. Figure 3.6e illustrates the deformation behaviour as measured on fault F14, the location of which is shown in Figure 4.6c. The azimuths of the resultant displacement vectors in the thrust sliding zone correspond to intersection of faults and fractures in the rock, which strike from 65?N E to 81?N E. This direction swings in the toppling zone to a direction that is at right angles to rock strata and strike of faults, as shown in Figure 3.6f. In addition the dip angles of displacement in the sliding zones, which were 30? to 40?in the drag PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-23 sliding and 49? in the sliding zone, correspond closely with the plunge angles of the fault intersections which were 33? and 53? respectively. Dip angles in the toppling zone were uniform at between 15? to 20?. Sijing (1980) goes on to show how the overall deformation rate both in terms of deceleration and acceleration of the failing rock mass can be described by an exponential relationship with time as shown in Figures 3.6g and 3.6h. Sijing (1980) then goes on to derive an empirical relationship to determine the ?time to reach the stage of accelerated deformation?. This is discussed further in Section 2.13 where results are compared with other time to failure prediction methods. Discussion Sijing (1980) gives a fairly comprehensive review of the displacement behaviour of a complex progressively developing slide-toppling failure in an open pit. Unfortunately many details such as changes in mining rates, the position of the groundwater etc are not discussed. The failure was clearly structurally controlled with deformation movements in the upper sliding zones controlled by fault intersection plunge directions and toppling normal to strata and faults. This is illustrated in the sterographic projection of the structures and slope in Figure 3.6i. It should be noted that orientations are approximate having been measured off sketch plots. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-24 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-25 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-26 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-27 3.7 Case Study 7, Steep Rock (Hogarth) Mine, Canada, No 1. Zone Highwall Failure The Hogarth Mine is situated in Ontario, Canada close to Atikokan. Between August 1974 and June 1975 the highwall displayed increasing signs of distress which culminated in a failure which started on the 23rd June. The geology and description of the failure were described in Brawner and Stacey (1979), Brawner, Stacey and Stark (1975) and Brawner, Stacey and Stark (1976). The failure is interesting in that it details a comprehensively monitored large scale classic toppling mode failure. Geology and Stratigraphy The Steep Rock ore zone is composed of a goethitic, soft iron ore horizon, which dips steeply to the west. The ore zone is overlain by pre-Cambrian volcanics (ash rock) and underlain by sediments (paint rock and carbonate). The Middle Arm section of the orebody is separated into several sections by faulting and folding. The major divisions, from south to north, are termed the Errington Zone, the Roberts Zone and Hogarth Zone. At the north end of the Hogarth Zone the ore terminates against the Bartley Fault, which strikes northeast and dips to the southeast at 85 degrees. Northwest of the fault the high wall is comprised of a hornblende-biotite metadiorite also referred to as ?diorite?. The diorite contains sheared basic dykes which parallel the fault. It also contains two well-developed vertical joint sets which strike parallel to and perpendicular to the major fault direction. The failure was restricted to the diorite highwall. A schematic drawing of the highwall is shown in Figure 3.7a. Failure Deformation Behaviour A crack behind the crest of the diorite highwall was first identified in October 1973. Further cracking occurred in August 1974 as mining moved closer to the toe of the slope. Comprehensive EDM triangulation and extensometer monitoring was implemented and a seismograph installed. Overall creep of the slope continued until January 1975 at wh ich time there was the onset of sub-freezing conditions (Figure 3.7b). Movements were erratic with accelerated movement correlated to periods of heavy rainfall. Vector movements as shown in Figure 3.7a reflect horizontal:vertical movement ratios of 3:1 with almost no movement at the toe, confirming a toppling instability mode. Blocks situated between the front and back cracks showed horizontal:vertical movement ratios of 1:1 indicating a slumping wedge style behaviour behind the front toppling block. All movement directions were on a east-north-easterly azimuth. With the onset of sub-freezing conditions from January 1975 to 10 th March 1975 almost all deformation ceased. With the onset of thawing conditions deformations started to accelerate. Heavy rainfall in May resulted in continued acceleration of the instability which was ravelling with small failures in the ash rock (volcanics). Deformation behaviour patterns continued with the front block toppling (rotating) and back blocks slumping. The following chronology summarises the events leading up to failure; 21st May: 100 cu.yds. slab failed off the front mass 10th June: wedge failed several 100 cu.yds. 11th June: further section failed off front bloc ks (Average movement 1ft/day prior to 10th June ) 20th June: heavy rainfall 23rd June: 5hrs prior to failure front block movement was 30.5ft horizontally and 4.8ft vertically since August 1974 (see Figure 3.7c). 23rd June: Failure in toppling mode ? front column/b lock rolled forward and remainder of the material slumped down and ravelled. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-28 Discussion No results of stability analyses or positions of groundwater tables were presented in the papers. Up until failure there was no evidence of the toe being pushed out. It is difficult to assess from the literature whether the actual mechanism was facilitated by; ? the front blocks rotating as a result of instability at their base causing the back blocks to slip down into the void behind, or ? the back blocks providing a driving force causing the front blocks to topple. Either way, groundwater in the form of melt water from thawing and heavy rainfall appears to have played the primary driving force in destabilising the slope. The fact that almost no movement took place in the frozen season suggests that mining activity had little impact on the failure. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-29 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-30 3.8 Case Study 8, Inspiration?s Mines, Ar izona, USA, General Instabilities and Failures The Inspiration?s Mines four open pits which include Live Oak, Red Hill, Thornton and Joe Bush Extension are located near Globe in Arizona, USA and have had a history of 16 recorded instabilities from 1953 to 1980. The rocks are predominantly deformed schists with tertiary intrusive porphyry rocks. The geology and the description of the instabilities were detailed by Savely and Kastner (1981). This case study illustrates how adverse structural geology of the rock mass can result in slopes that are highly sensitivity to mining activities and rainfall, even at relatively flat slope angles. Geology and Stratigraphy The predominant rock unit is a precambrian Pinal Schist which have been moderately to strongly deformed as a result of multiple episodes of intrusive activity, tectonism, and regional folding. The orientation of the schistosity ranges from 40?/140? to 45?/140? (dip/dipdir) with local variations. Three predominant types of tertiary intrusive rocks are also present which are a quartz monzonite porphyry and a coarser grained granite porphyry as well as a biotite granite. Contacts between the granitic rocks are generally gradational. Dacite and two conglomerates also occur in minor amounts. There is a prominent north-northeast striking system of normal faults which include the Barney, No.5, Keystone, Bulldog, and Miami faults. Dips on these faults are 25? to 55?SE with interpreted vertical displacements ranging from 30m to over 300m. A system of reverse faults has also been mapped in the southwest area of the Live Oak pit and this same orientation is probably present throughout the area. These faults have a general west-north-west strike with dips of 40? to 45? SW. Further faults include the northwest striking sub-vertical Joe Bush fault and the Pinto fault with dips ranging from 35? to 55?. Description of Failures and Failure Deformation Behaviour The mines have a recorded history of over 16 instability events in different pit sectors from 1953 to 1980. Some of the instabilities were continuous. The predominant failure mechanisms were wedge failures which formed by the intersection of major joint sets and combination of joints and faults, planar shears on faults as well as rotational type failure mechanisms through highly jointed rock mass. The general aggravating factor for the instabilities has been the high phreatic surface levels which suggests saturated slope conditions. Evidence for this was recorded in the immediate movement of slopes following rainfall, wet slope faces and flowing blast and development holes. Visible slope deformations associated with these failures have been toe heave and cracking. A more detailed study of the Thornton north west sector (cross section shown in Figure 3.8a) was presented by Savely and Kastner (1981). In this sector adverse structure orientations, which are illustrated in the sterograms in Figure 3.8b, caused large scale 3D wedge failures and wedge break back on benches. Slide wire monitoring shown in Figure 3.8c illustrates the sensitivity that the slope had to mining activities at the toe as well as to rainfall. Mining in early 1980 was delayed 2 months when movement rates exceeded 65mm/day in response to a combination of mining and heavy rainfall. Calculated FOS for the slope ranged from 0.9 to1.2 depending on phreatic surface and strength parameter assumptions etc. The stability of the slope was finally improved through the installation of an extensive network of drainage holes although FOS remained low. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-31 Discussion The adverse effects of structural geology of the rock mass and the extent to which the Inspiration?s slopes are affected by mining and rainfall, which translates into higher phreatic surfaces, are clearly apparent even with the relatively flat slope angles used. Mining activities were continued despite these ongoing instabilities and failures. The literature confirmed the importance of reliable displacement monitoring systems in slope management and safe mining practices. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-32 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-33 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-34 3.9 Case Study 9, Nchanga Open Pit, Zambia, July 2004 Failure The Nchanga Open Pit is situated near Chingola in Zambia. The pit has a history of ongoing instabilities. The failure occurred in the north wall (Pit 20) of the Nchanga Open Pit. The rocks are predominantly folded soft sedimentary units. The geology and the description of the failure is detailed by Naismith and Wessels (2005). This case study once again highlights the sensitivity of the slopes to mining activities and additionally to degradation resulting from weathering and rainfall. The slope reflects a classic blast, acceleration and recovery (decay) deformation pattern. The case study also illustrates the early use of SSR technology at Nchanga for monitoring slope deformation. Geology and Stratigraphy The failure itself took place in the following lithology (Wessels, 2005); Table 3.17 : Nchanga Open Pit Lithology Units RMR UCS (MPa) Shale with grit (SWG) 34-45 - Chingola Dolomite (CDol) <35 <10 Upper Banded Shale (UBS) - - Banded Sandstone Upper (BSSU) - - A typical cross section of the slope is shown in Figure 3.9a. Description of Failures and Failure Deformation Behaviour Deformation of this failure was initially measured only by prism survey monitoring. Instability of the slope commenced in September 2002 at the 165m bench level which was interpreted to be localised toppling failure as a result of adversely orientated cleavage planes in the shale and grit formation. Ongoing and improved survey monitoring revealed average displacement rates of approx 2mm/day. In July 2003 some prisms between pit section 23E and 24E and the 150m and 225m benches were showing increases in displacement at a time when mining at the base of the pit exposed the core of a tight fold axis (as shown in Figure 3.9a). The relationship between blasting and deformation continued to become more evident and by April 2004 it was noted that blasting was having a significant impact on deformation rates with instantaneous deformation rates of up to 150mm/day occurring following blasting. At this time accumulated deformation in some prisms was over 3m in total and a large tension crack had developed at the toe of the face of the 150m bench which had widened to 2.5m before measurements were halted (Wessels, 2005). The extent of the instability is shown in Figure 3.9b. Total accumulated deformation for selected prisms is shown in Figure 3.9c. Total rates of movement for the same selected prisms are shown in Figure 3.9d An SSR unit was installed on the 19th May 2004 and immediately showed the extent of the developing instability. The SSR system was able to detect a zone of instability much wider than was initially estimated, due to insufficient coverage by prisms, as well as ?two hot? spots within this zone which PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-35 were displacing with larger deformation rates. The SSR thermal deformation as well as camera images for the zone of instability showing the different rates of deformation are shown in Figure 3.9e. On the 22nd May 2004 a collapse involving 3 benches occurred on the western side of the deformation zone on pit section 21E. The SSR derived displacement rates in the 4 days leading up to the 22nd May failure are shown in Figure 3.9f. The classic blast, acceleration and 1-2 day recovery deformation pattern continued up to early July after which the pattern began decaying, ie average slope relaxation rates stabilised at an ever increasing rate of deformation. After a blast on the 12th July movement rates exceeded 20mm/hr and showed no signs of recovery or decay in rate. The slope deformation began to behave progressively, with deformation rates of over 100mm/hr being recorded just before collapse, which occurred on the 16th July 2004. The SSR derived displacement rates in the 4 days leading up to the July failure are shown in Figure 3.9g. The total volume of collapsed material was estimated to be 1.8 million m3. Discussion The July 2004 failure appears to have taken place in two 2 distinct stages as follows; Stage 1 Initial toppling type mode of instability in the SWG with crest of the instability on the 150mB bench rotating outwards, leading to the development of a major vertical crack (release surface). Stage 2 The ?released? wedge places additional pressure on the underlying rocks generating a shear failure through the CDol and DolSchist with the base of the failure possibly sliding on the UBS unit. The deterioration in recovery rates of the classic long term blast, acceleration and recovery (or decay) deformation pattern are clearly evident in Figure 3.9c in the months of April to May 2004. A correlation between the accumulated deformation of the prisms outside of the collapse limit and their location in Figure 3.9b is not completely clear and would require further knowledge of the geology in order to interpret further. A comparison of Figures 3.9c and 3.9d confirms the relationship between deformation rates and accumulated deformation, but also illustrates why the ?noise? associated with instantaneous survey deformation rates is not necessarily instructive in revealing the degree of instability. The deterioration in recovery rates of the classic long term blast, acceleration and recovery (or decay) deformation pattern is most likely a result of the reduction in the shear strength of the materials as a result of high strain rates induced by plastic flow of the Chingola dolomites towards the base of the slope. The extent that the effect of blasting, mining rate and change in adverse geology had on the rate of development of the failure, and resulting failure mechanism, was unfortunately not established. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-36 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-37 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-38 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-39 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-40 3.10 Summary of the Literature Review Case Studies The nine selected literature review case studies presented above have illustrated significant similarities in the characteristics of rock mass deformation in slopes. These include; 1. Specific events, especially mining (blasting) and high rainfall are often responsible for triggering a deformational response of a slope. This is illustrated in the following failure case studies; Brenda Mine (Figures 3.1a), Jinchuan Mine (Figure 3.6d), Hogarth failure (Figure 3.7b), Inspiration Mines (Figure 3.8c) and Nchanga (Figure 3.9d). 2. The initial part of the deformational response involves a sudden increase in the velocity of the slope as illustrated in the above case studies. The velocity of the slope is movement shown in Figure 3.9d of the Nchanga case study which was the clearest presentation of prism velocities. 3. Following on from the initial deformational response, there is a period of relatively rapid reduction or decay in the velocity of the rock mass, followed by a period of more gradual creep. This is termed the rock mass recovery function, once again illustrated in Figure 3.9d (Nchanga failure). 4. Very significantly, the closer to failure the slope develops, the more the deformational behaviour of the recovery functions start to change. The recovery functions start to take an increasingly longer time to stabilise. This characteristic is illustrated in the Jinchuan (Figure 3.6d and e), Inspiration Mines (Figure 3.8c) and Nchanga (Figure 3.9d) case studies. 5. A point in time is reached where the deformation rate of the unstable rock mass starts to accelerate continuously in an exponential manner until collapse occurs. This is shown in the deformation behaviour of all the case study failures. 6. Although differing in rates and magnitudes, the pattern of behaviour as described above tends to be reflected in slopes displaying different failure mechanisms. Thus can be seen by contrasting the deformational behaviour patterns of the active/passive sliding blocks of the Brenda failure (Figure 3.1a) with the slide/toppling failure of Jinchuan (Figures 3.6d and e) and the pure toppling of Hogarth pit, the Thornton North-west sector wedge slide (Figure 3.8c) and finally the complex mechanism of the Nchanga failure (Figure 3.9c). 7. Post collapse deformation can be extremely complex. Examples include the Cassiar pit case study where Martin (1993) describes a ratchet mechanism developing as well as the Luscar Mine failure where a partial recovery following collapse was followed by a further and final collapse. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 3: Literature Review Case Study Database Page 3-41 These deformational characteristics which are summarised above will be used, together with the findings from the detailed case studies, to develop a more comprehensive description of rock mass deformation behaviour of rock slopes. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 4: Definition of Research Topic Page 4-1 4 DEFINITION OF RESEARCH TOPIC BASED ON THE FINDINGS OF THE LITERATURE REVIEW 4.1 Specific Aspects of Time Dependent Deformation to be Addressed Based on the conclusions to the literature review, the following specific aspects of time dependent behaviour were identified to be addressed in the detailed case study phase of the research; ? The usage of the time dependent change in the deformation rate decay function to predict failure ? The usage of acceleration to predict the onset of the collapse point. ? A review of the remaining time-to-collapse forecasts for instabilities. These are discussed further below. 4.1.1 Time Dependent Change In The Deformation Rate Decay Function As has already been shown in the literature review, the only previous attempt to develop a characteristic time dependent model of rock slope deformation behaviour was by Martin (1993), as described in Section 2.7 and 2.8. While at the time a significant development, it nevertheless has several deficiencies. These include; ? An absence of predictive ability. This means that there are no indicators to suggest, from observation of the deformation behaviour of a rock slope, when progressive failure is likely to be initiated. ? Martin (1993) used a uniform negative exponential displacement rate decay function (equation 2.3) to describe the deformation rate decay (steady state creep) of an excavated rock slope. He uses this function to describe the time dependent adjustment that occurs for a particular rock slope resulting from all events leading up to the onset of failure point. He terms this entire period leading up to the onset-of-failure, the ?initial response?. The derived values of the terms ?A? and ?b? in equation 2.3 were therefore constants, which are a function of the rock mass properties. However Martin(1993) confirms that the constants A and b vary for different slopes, and attributes this to variations in the quality of the rock mass, slope geometry (height and angle) and the ultimate failure mechanism. It is evident to the Author that application of uniform A and b constants to describe the deformation decay rates applicable to every new excavation event is a significant oversimplification of the deformation behaviour. Examination of the literature review case studies, as well as the Author?s experience of typical deformation leading to failure, shows that the steady state creep decay rate evolves from a sharp and rapid decline in the initial stages of slope excavation, to creep rates that take increasingly longer to recover, and result in increasingly higher critical steady state creep rates. Eventually the steady state creep rate reaches a point where no further decrease in the steady state creep rate takes place and progressive behaviour is initiated. This is also termed the onset of failure point (Zavodni and Broadbent, 1980). Progressive behaviour is characterised by continuously increasing deformation rates. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 4: Definition of Research Topic Page 4-2 It is therefore evident that the displacement rate decay function cannot be constant and that it is a function of the slope height, slope angle and rock mass properties. This revised description of characteristic deformation behaviour is illustrated in Figure 4.1. A continuous change in the decay function infers changing values of the constants ?A? and ?b? and therefore a uniform change in the decay function as illustrated in Figures 4.3 to 4.5. ? The model does not adequately reflect the entire range of complex modes of time and event dependent deformation behaviour. The following aspects of deformation rate are therefore to be researched; 1. Can a new time dependent deformation model be developed with predictive characteristics? 2. How do the constants A and b in the steady state creep rate change with time? 3. Can a threshold terminal steady-state creep rate be identified for different rock masses above which tertiary creep or progressive behaviour of the slope is initiated? 4. Does a negative exponential function result in the best fit for all deformation data or are other functions more applicable for different rock masses or failure modes? 5. What are the factors which influence the creep rate and can this terminal creep rate be predicted? 4.1.2 The Usage of Acceleration to Predict the Onset Of Collapse Point. The modified characteristic time dependent deformation graphs have been extended further to include a hypothesized acceleration diagram as shown in Figure 4.1. It can be inferred from this acceleration diagram that a change from a negative acceleration pattern associated with normal transient creep (regressive behaviour) to a positive acceleration pattern associated with progressive behaviour may be a good indicator of the onset of collapse (OOC). This is required to be verified. 4.1.3 Time-to-collapse Forecasts Existing empirical methods to forecast the remaining time-to-collapse for developing instabilities will be reviewed. It is planned that a new method of forecasting will be developed in the research. The potential application of this method needs to be fully investigated. 4.2 Detailed Research Approach The detailed research approach was designed to collect and analyse information to address the three specific aspects of time dependent deformation as discussed above. The approach includes the following aspects; ? Identification of required data for collection, ? Selection of detailed case studies and collection of data, PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 4: Definition of Research Topic Page 4-3 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 4: Definition of Research Topic Page 4-4 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 4: Definition of Research Topic Page 4-5 ? Evaluation and interpretation of data collected from detailed case studies, ? Presentation of findings. 4.2.1 Identification of Required Data for Collection The following categories of data were identified for collection; ? Regional geology, ? Stratigraphy and weathering characteristics, ? Structural geology, ? Groundwater monitoring data, ? Rock mass classifications and strength, ? Monthly surveys of open pit, ? All data relating to displacement monitoring, including prism surveys, radar, extensometers and inclinometers. 4.2.2 Selection of Detailed Case Studies Having identified what data to collect, a review was made of open pit mining operations that had good monitoring systems. This information was collected from monitoring system vendors such as Leica, Groundprobe, Maptek and consultants and specialists working in the open pit mining field. The ?detailed? open pit case studies were originally selected based on several criteria. These were: ? The existence of reliable slope displacement monitoring systems and associated histories, an established structural geological and geomechanical database, implementation of environmental monitoring programs and availability of information. ? A history of monitored failures. ? The availability of comprehensive event/time displacement behaviour for as broad a range of rock mass types and geological terrains, as possible. Due to the large amounts of information made available to the Author on each case study, the data presented have been carefully selected. For non-failed slopes in oval to rectangular shaped pits, two representative sections have typically been selected (footwall and hangingwall) for evaluation. In circular shaped pits additional representative sections have been included. To assist in the interpretation of case studies, Vulcan 3D visualization software has been used wherever possible to develop a geological and structural database of the pit and surrounding environment. Twelve case studies were selected for detailed study as listed below. In all cases except Chiquicamata Mine, the pits were inspected by the Author. In the cases of Palabora and Potgietersrust Platinum they were inspected several times during the duration of the research work. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 4: Definition of Research Topic Page 4-6 Detailed Case Study 1 Mine/Owner: Navachab Gold Mine/Anglo Gold Ashanti Location : Namibia Number of pits: 1 Geology : A suite of steeply dipping meta-sediments and meta-volcaniclastic rocks situated on a fold limb and transected by fracture and shear zones. Rock mass type: Dolomitic enrichment Slopes reviewed: East face controlled by steeply dipping foliation planes. West face has structurally interactive slope performance Failures investigated: East slope planar failures, small west slope wedge failures. Detailed Case Study 2 Mine/Owner: Potgietersrust Platinum Mine/Anglo Platinum Location: South Africa Number of pits: 2 Rock mass type : Platinum reefs, igneous rocks Slopes reviewed: Brittle slopes demonstrating structural mechanisms Failures investigated: Small but rapidly developing planar failures on the east slope. West slope is bedding controlled. Detailed Case Study 3 Mine/Owner: Kalahari Gold/Harmony Gold Mining Location : South Africa Number of pits: 1 Rock mass type: Archean BIF units. Slopes reviewed: East and west Failures investigated: Multiple small sliding failures Detailed Case Study 4 Mine/Owner: Chuquicamata Mine/Coldelco North Location: Chile, South America Number of pits: 1 Rock mass type: Phorphyritic copper, granodiorite basement rocks. Slopes reviewed: West and east, both show brittle behaviour. Potential effects of future block caving likely. Failures investigated: Zone 3 failure. Detailed Case Study 5 Mine/Owner: Palabora Mine/Palabora Mining Company (PMC) Location: Phalaborwa, South Africa Number of pits: 1 Rock mass type: Porphyritic copper Slopes reviewed: All Failures investigated: Glimmerite failure. Major current north wall failure initiated by recent underground block cave mining. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 4: Definition of Research Topic Page 4-7 Detailed Case Study 6 Mine/Owner: Bibiani Gold Mine/Anglo Gold Ashanti Location: Ghana Number of pits: 1 Rock mass type: Strongly silicified phyllites and greywackes with thin subvertical foliation. Slopes reviewed: East and west slope Failures investigated: 11 in total. Detailed Case Study 7 Mine/Owner: Geita Gold Mine/ Anglo Gold Ashanti Location : Tanzania Number of pits: 5 (Kukuluma, Geita Hill, Nyankanga, Lone Cone North and South) Rock mass type: Archean BIF Units, metavolcanics Slopes reviewed: Numerous Failures investigated: Numerous saprolite (oxide) failures, hard rock slips associated with major structures. Detailed Case Study 8 Mine/Owner: Mount Keith/ BHPbilliton Location : Western Australia Number of pits: 1 Rock mass type: Ultramafics and felsics, curvilinear structures, highly serpentonised Slopes reviewed: Footwall sedimentary rocks, hanging wall basalts Failures investigated: 12 failures mostly batter scale fresh rocks and one large saprolite failure Detailed Case Study 9 Mine/Owner: Leinster Nickel Operations (LNO)/ BHPbilliton Location : Western Australia Number of pits: 3 (only the Harmony pit used) Rock mass type: Ultramafics a nd felsics, curvilinear structures Slopes reviewed: East and west slopes Failures investigated: 4 multi-bench scale failures in including the east wall failure at closure, saprolite failures and hard rock slips Detailed Case Study 10 Mine/Owner: Venetia Mine/DeBeers Location : South Africa Number of pits: 1 Rock mass type: Limpopo Belt Gneissic and metasedimentary packages Slopes reviewed: North and South Failures investigated: 2 large multibench scale failures Detailed Case Study 11 Mine/Owner: Orapa and Letlhakane Mines/DeBeers Location : Botswana Number of pits: 2 Rock mass type: Karoo Supergroup Slopes reviewed: All - circular pits PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 4: Definition of Research Topic Page 4-8 Failures investigated: 3 large multibench scale failures Detailed Case Study 12 Mine/Owner: Kalgoorlie Superpit/ Kalgoorlie Consolidated Gold Mines (KCGM), Location: Western Australia Number of pits: 1 Rock mass type: Archean Greenstone Slopes reviewed: All Failures investigated: 4 The information obtained on these detailed case studies was examined carefully, and some of the case studies were then excluded from the detailed case study write up for the following reasons, ? Potgietersrust Platinum Mine/Anglo Platinum There have been no significantly large multi-bench scale failures which have occurred. Small but rapidly developing planar failures occur relatively frequently on the east slope which are more a safety risk rather than an overall slope stability concern. The Gemos prism survey monitoring system was data sparse and had only been operating for one year. The SSR system was at the time newly introduced and laser scanning data were not made available. ? Kalahari Gold/Harmony Gold Mining Numerous small structurally controlled sliding and toppling failures on the east wall associated with J1 and J2 joint sets. There was up until the date of inspection only one significantly large failure, which was not monitored. ? Chuquicamata Mine/Coldelco North Although a lot of data and literature were collected and reviewed this was the only pit the Author did not have the opportunity to inspect personally. It was therefore decided not to use this case study for write up. ? Palabora Mine/Palabora Mining Company (PMC) There was only one significantly large instability which occurred in 1990 in the glimmerite, prior to the large scale failure initiated by the underground block caving operation. Deformation monitoring data on the glimmerite instability were not available. ? Geita Gold Mine/ Anglo Gold Ashanti - there were no failures in the Kukuluma, Geita Hill and Lone Cone North pits. - there was one unmonitored sliding failure in the Lone Cone South pit - relatively small laterite instabilities around the perimeter of the Nyankanga pit and sliding instabilities on south west slopes caused by moderate north dipping structures. Survey monitoring data were at the time of the site visit and data collection, generally of poor quality. ? Kalgoorlie Superpit/ Kalgoorlie Consolidated Gold Mines (KCGM), - Two Chaffers pit failures occurred in weathered rock (oxides), were multi-bench scale failures but were not monitored. - Two Brown Hill pit active/passive wedge failures occurred in fresh basalt, were multi-bench scale failures and occurred as a result of the Oroya footwall shear which runs parallel with the PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 4: Definition of Research Topic Page 4-9 slope of the wall. Prior to the failures, no movement had been recorded in the sections of the east wall where the failures occurred, with either the SSR or the prism monitoring system. It should be noted that all relevant behaviour data from the case studies which have been excluded from the write up, were still considered in the development of the research findings. All publications relating to the six detailed case studies which have not been presented in the write up, are listed in the bibliography. 4.3 Detailed Case Study Data Presentation, Evaluation and Interpretation The write ups on the detailed case studies will be presented in the following format; ? Introduction ? Regional Geology ? Stratigraphy ? Structural Geology ? Geological Section ? Rock Mass Classification and Geotechnical Characteristics ? Pit Configuration Parameters ? Displacement Monitoring Systems ? History of Failure/s ? Displacement Behaviour of Slopes leading up to Failures ? Interpretation of Displacement Behaviour, Failure Mechanism and Trigger ? Displacement Behaviour Non-failed Pit Wall Sectors ? Discussion 4.4 Evaluation and Interpretation of Data Collected The methods by which data collected from the detailed case studies and literature survey case study database will be analysed are summarised below. ? The usage of the time dependent change in the deformation rate decay function to predict failure The evaluation of data in order to study the change in the deformation rate decay function will be carried out using regression analysis on the prism survey data obtained. Regression analyses will be carried out using alternative functions to Martin?s (1993) negative exponential in order to evaluate their suitability. ? The usage of acceleration to predict the onset of collapse point. Prism survey data will be evaluated to determine the practicality of using acceleration characteristics. ? Review time-to-collapse forecasts of failures PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 4: Definition of Research Topic Page 4-10 A review will be made of actual versus predicted time-to-collapse data for different modes of failure. In particular the ?prediction methods? as established by Zavodni and Broadbent (1980), Saito (1969, 1980) and Varnes (1983) will be used to evaluate best fit criteria. 4.5 Presentation of Findings The interpretation and findings of the data collected in the detailed case studies and literature review case study database will be presented in the following sections; ? Section 11: Development of a generalised time and event dependent deformation model. ? Section 12: Application of the time and event dependent deformation model using case study data examples. ? Section 13: Forecasting of deformation behaviour using the Model. 4.6 Discussion Having defined the research topic and research approach, the following sections present the findings of selected detailed case studies. The detailed case studies which have been presented in the following sections were selected based on the following criteria; ? Comprehensive geological information, ? At least one large failure (preferably several), ? Comprehensive monitoring data with relatively little noise, ? A range of different failure mechanisms between the case studies. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 5 : Case Study 1, Navachab Open Pit Page 5-1 5 CASE STUDY 1, NAVACHAB GOLD MINE, NAMIBIA 5.1 Introduction The Navachab Gold Mine is located in Namibia outside of the town of Karabib and is operated by AngloGold. The geological terrain consists of a suite of steeply dipping meta-sediments and meta- volcaniclastic rocks situated on a fold limb and transected by fracture and shear zones. The pit was selected as a case study due to the excellent monitoring systems and data available, as well as there having been a significant and well documented failure. All figures and tables pertaining to this section which are not shown in the main body of the report are contained in Appendix 1. 5.2 Regional Geology The Navachab gold deposit occurs in the planar, NNE trending, steeply westward-dipping, western limb of the Karibib dome, which is located within the south central zone of the Pan-African Damara orogen. In this limb, quartzites of the Etusis Formation are overlain by a thin conglomeratic horizon (Chuos Formation), which in turn is overlain by fine-grained biotite-rich quartzitic schists, interbedded with metabasic and calc-silicate layers (Spes Bona Member). These are overlain by marbles and calc- silicates of the Okawayo Member, biotite schists and metabasites of the Oberwasser Member and dolomitic marbles and calc-silicates of the Karibib Formation (SRK Consulting, 2001a). The Navachab open pit is approximately 900m long by 300m wide and currently 190m deep. It is centered on the mineralised calc-silicate unit of the Okawayo Member. The pit is orientated along the NNE strike of the Karabib orebody with the long axis orientated at 30? to strike of the strata. The strata dip at an overall angle of approximately 70? WNW. The dominant compositional layering or bedding dips steeply to the WNW (average dip/dip direction = 74?/289?) but shallows intermittently to daylight on the east slope with dip angles of less than 50? (R oux and Terbrugge, 2006). The layout of the pit is shown in Figure 5.1. 5.3 Stratigraphy From east to west, or footwall to hanging wall, the following units are exposed inside the Navachab open pit (SRK Consulting, 2001a); Lower schist (LS) : A more than 150m wide unit of planar, fine-grained biotite-schist or metasandstone, locally interbedded with calc-silicate layers and more mafic hornblende-rich horizons. The schist contains a poorly preserved tectonic foliation or cleavage (Sl) that is sub-parallel to the compositional layering (SO). Calc-silicates (CM) : A 40 to 50m wide zone of well-layered calc-silicate constitutes the main ore body host rock. Main marker horizon (MDMV): A 3 to 8m wide unit of metabasic rock consisting of hornblende, pyroxene and plagioclase with additional biotite. The unit has a massive, poorly foliated appearance. Along its hanging wall it is highly deformed with thin lenses and layers of metabasite parallel to the main unit. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 5 : Case Study 1, Navachab Open Pit Page 5-2 Marbles (Lower MDM): A 30 to 40m wide unit of intensely deformed, grey-white dolomitic marbles ('striped marbles'). Marbles (Upper MDM): A 40 to 50m wide unit of more massive, grey-white marbles ('spotted marbles'). Upper schist (US) : A 150m wide unit of biotite schist similar in appearance to the LS unit. 5.4 Structural Geology Most structural features are shear zones or shear joints that can be linked to main shear zones transecting the pit. A number of shear joints occur parallel to bedding. Bedding planes are curved because of several folding events. A further large-scale structural discontinuity is represented by sets of pegmatite veins that transect the pit. The structural features are summarised in further detail below and illustrated in Figure A1.1 (Roux, 2005). a) 1st Order (Primary) Shear Zones and Pegmatites The Navachab pit is transected by a number of 315? trending fracture zones. The main NW trending fracture zones have accommodated sinistral displacement of bedding by as much as 50m PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 5 : Case Study 1, Navachab Open Pit Page 5-3 across a 200m wide zone that transects the northern sector of the pit. The main pegmatite dykes are intruded along these fracture zones. Four NW trending pegmatite zones, each 20 to 30m wide, can be seen to traverse the pit. These zones are mostly visible in the LS, but some penetrate the MDM, especially in the north-western comer of the pit (SRK Consulting, 2001b). The 1st order shear structure orientations are illustrated in the sterograhic plot in Figure A1.2. a) 2nd Order Shear zones Shear zones crosscutting the pit can be classified as north-trending sinistral fault segments connecting bedding parallel fault segments within a 350? trending shear envelope. Two significant shear zones have previously affected slope stability. The first is a near vertical brittle ductile fault accommodating sinistral, east-down displacement (with fault plane orientation = 80?/094? (dip/dip dir) and slickenfibre lineation). The shear zone has a maximum displacement of 2 to 3m in its hanging wall (SRK Consulting, 2001b). This second shear zone is a near vertical brittle ductile fault, characterized by a 5 to 10cm wide zone of slickensided surfaces enveloping highly fractured rock, accommodating sinistral, east-down displacement (with fault plane orientation = 78?/100? (dip/dipdir) and slickenfibre lineation = 33?/ 178? (dip/dipdir)) (SRK Consulting, 2001b). The orientations of the 2nd order shears for the west and east slopes are illustrated in the sterograms in Figures A1.3 and A1.4 respectively. b) 3rd Order Bedding and Joint Sets The orientations of bedding planes in the footwall LS unit are not constant, but vary due to; ? Open flat-lying (recumbent) folding with fold axis orientations = 20?/005? (dip/dipdir). Bedding planes in these folds are locally shallow to dips of approximately 40?. ? Open inclined folding with near vertical fold axes in kink zones centred on pegmatite veins. The bedding planes in these folds steepen to near vertical. The orientations and pole concentrations for the west and east slopes 3rd order structures were derived from mapping and core orientation data (Roux, 2006) for the relevant units and are illustrated in the stereograms shown Figures A1.4 to A1.8. The average orientations for the bedding and significant joint sets are summarised in Table 5.1 Table 5.1: Mean Orientations of 3 rd Order Structures (Roux, 2006) Structure West Slope East Slope JRC 0 JCS 0 (MPa) Bedding 67?/357? 55?/10? 2?1 150?45 J1 35 ?/339? 40?/336? 6?1 120?36 J2 33 ?/241? 36?/237? 9?1 69?21 J3 27 ?/147? 51?/47? 7?2 85?26 J4 38 ?/171? 37?/171? 4?2 120?20 J5 33 ?/294? 41?/306? 8?2 90?20- J6 34 ?/87? 31?/48? 6?2 90?27 JB Not present 57 ?/335 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 5 : Case Study 1, Navachab Open Pit Page 5-4 5.5 Geological Section The selected geological section and weathering profile is shown in Figure 5.2. The position of the geological section through the pit is situated approximately adjacent to the 2001 ramp failure on pit section 2900E. There are three main stratigraphic units in which the slopes are excavated. These are; West Pit Slope ? Upper schist (US) ? weathered and fresh zones ? Marble dolomitic marble (MDM) ? fresh zone only East Pit Slope ? Lower schist ? weathered and fresh zones 5.6 Rock Mass Classification and Geotechnical Characteristics Rock mass classifications and geotechnical characteristics for the three units are summarised in Tables 5.2 and 5.3. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 5 : Case Study 1, Navachab Open Pit Page 5-5 Table 5.2 : Summary of Mean Rock Mass Ratings (Roux, 2006) Rock Type RMR MRMR Oxidised US 35 22 US 51 42 MDM 56 46 MC 46 39 Oxidised LS 31 20 LS 47 39 Table 5.3 : Summary of Intact Rock Strengths (Roux, 2006) Rock Type Mean UCS (MPa) Std Dev (MPa) Youngs Modulus Ei (GPa) ? Oxidised US 30 15 N/A N/A US 169 79 N/A N/A MDM 84 16 43.0 0.178 MC 82 25 47.5 0.201 Oxidised LS 55 1 N/A N/A LS 204 105 80.0 0.254 5.7 Pit Configuration Parameters Pit configuration parameters are listed in Table 5.4 for the pit as at end March 2006. Table 5.4 : Pit Configuration Parameters for 2005 West Slope East Slope Height (m) 167 165 Overall slope angle (?) (OSA) 43 50 Bench stack angle (?) (BSA) 38-52 55-59 Bench face angle (?) (BFA) 41-62 61-62 Inter ramp angle (?) (IRA) 53 50 Spill/catch berm width (m) 6 11-13.5 Spill/catch berm height (m) 22 40 Bearing (?) 174 (local) 114.2 (Lo) 354 (local) 294.2 (Lo) 5.8 Displacement Monitoring Systems High precision manual survey system since start of pit. Leica/Geomos monitoring system deployed since May 2005. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 5 : Case Study 1, Navachab Open Pit Page 5-6 5.9 History of Failure/s West Wall No major failure recorded. Failures are limited to bench scale wedge failures. East Wall There have been three significant failures on the east wall (foot wall) of the pit. The first two occurred in 1997 and were termed the 1997 Central Footwall Failure and the 1997 East Slope Wedge Failure (Roux and Terbrugge, 2006). The central footwall failure occurred in a 40m high bench above the pit floor situated at 1050m, was bounded by 2 discrete joint planes and was triggered by a production blast. The east slope wedge failure occurred in a section of the wall which had been stable for 3 years and was also triggered by a production blast. 2001 Central Footwall Failure (main failure) Relatively minor instability of the central footwall was first recorded in June 1998 which involved a 15m high failure of approx 2500 tons. This was followed by a second failure in September 1999 at the same location with a height of approx 30m recorded in prism no ?C?. The ?main? failure occurred in March 2001 and involved the catastrophic failure of approximately 75 000 tons. The failure was situated in a 50m high stack (BSA=63?) from the edge of the ramp (situated at 1100m) to a 15m wide bench at the 1050m level. The failure scarp extended over a height of approximately 55m and a average width of approximately 107m with an maximum depth of approximately 6m as measured from the Vulcan model. 5.10 Displacement Behaviour of Slopes leading up to March 2001 Failure The prisms affected by the March 2001 failure were C2, C3, A, B, L, P163, P164, P114, P130, P138, P139 and S44. The magnitudes of displacements leading up to the failure are shown in Figures 5.3, 5.4 and 5.5. Resultant horizontal vectors up to 1/2/2001 are shown in Figure 5.6. Resultant horizontal vectors during the period 1/2/2001 to 8/3/2001 after the onset of collapse are shown in Figure 5.7. 5.11 Interpretation of Displacement Behaviour, Failure Mechanism and Trigger The Dominant Structural Features Relevant to March 2001 Failure The March 2001 failure occurred in a section of the hanging wall of the main shear zone on the east face of the pit where the bedding orientation became shallower from 68?/290? to 35?/310? (dip/dipdir). The change in orientation corresponds to a fold axis orientation of 005?120?. The failure is bounded on the southern side by a N-S trending shear zone (called the footwall fault) and on the north side by a steep NW trending fracture zone parallel to pegmatite veins. These two structures together formed cut-off/release surfaces either side of the failed mass. Apparent Failure Triggers An unusually high rainfall period occurred in the region between 23rd and 26th February 2001 approximately 2 weeks before the failure, followed by an additional 18mm on 8th March 2001 and a further 27mm on the 9th March (SRK Consulting, 2001a) PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 5 : Case Study 1, Navachab Open Pit Page 5-7 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 5 : Case Study 1, Navachab Open Pit Page 5-8 Displacement Pattern Identified Onset of collapse (OOC) was estimated to be around the 1 February 2001. Total time from OOC to failure was 34 days. Prism movement reflected an accelerated creep rate finally resulting in catastrophic collapse. 5.11.1 Horizontal Displacements Prism C3 displayed a continuous progressive behaviour trend from approximately the 30th November, a total of 72 days from OOC. Prisms P138 and P138 registered an onset of progressive behaviour also around this date, however this turned regressive until the final OOF. The earliest OOF was approx 9th January (prism C3) and the latest was the 4 th February (prism B) a difference of approximately 27 days. The total incremental displacement from the onset of failure to collapse was in the order of 21mm in the southern prisms and 8mm in the central prisms. Average velocities during failure were therefore 0.344mm/day on southern prisms and 0.235mm/day on the central prisms. There were no prisms along the northern edge of the failure. This differential deformational movement indicates that collapse occurred after a period of 61 days from OOF, during which failure propagated horizontally (south to north) across the failure surface. Failure was therefore initiated along the line of the central fault. Figure 5.6 shows a comparison of the directional vector movements before OOC, and between the OOC and final collapse (Figure 5.7). What is evident is an immediate uniform change in direction of movement after OOC of approximately 40? westwards in all prisms on the failure mass with the PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 5 : Case Study 1, Navachab Open Pit Page 5-9 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 5 : Case Study 1, Navachab Open Pit Page 5-10 exception of prism C2 which is the most easterly. In addition, prism P103 reflects a 127?eastward swing pointing almost 180? away from the direction of the collapsing rock mass. Of interest to note is the directional horizontal vector movements of the rock mass outside the limits of the failure mass are on average approximately 24? eastwards of the vector directions of the rock mass within the failure zone. This suggests a different displacement mechanism at work. The instantaneous prism velocities leading up to the March 2001 failure (Figure 5.4) show no trends or patterns, just periods of greater or lesser ?activity?. 5.11.2 Vertical Displacements Vertical displacements tended to mirror the pattern and sequence of horizontal movements, but with considerably smaller magnitudes of movement. From the pattern of movements it is clear that the OOC is associated with a significant change in rock mass behaviour. This is clearly evident from the reversal of vector movements before and after the OOC. The directional vector movements of the rock mass before OOC clearly follow the plunge of the bedding, however after OOC the direction movements clearly no longer follow the plunge of the final failure scarp (bedding) but swing significantly eastwards. Before OOC the displacement mechanism was clearly bedding controlled. This can only be accounted for by the development of a failure surface downwards along the failure scarp probably initiating along the central fault and propagating downwards and eastwards resulting in a cantilever behaviour of the rock mass which pivoted about the eastern limit of the failure. The swing of prism P103 to the opposite direction of the failing rock mass represents an elastic rebound effect of the non-failing rock mass directly above the failing rock mass as a result of downwards propagation of the crack. The propagation of the failure surface reached a point where the rock mass could no longer support its own weight and it resulted in failure. The OOC therefore clearly represents a change over from regressive displacement behaviour to progressive failure behaviour. 5.12 Displacement Behaviour of Non-failed Pit Wall Sectors Six critical sections of the pit were selected in order to study the displacement behaviour of prisms located close to these sections. On the east slope the chosen sections are 1E, 2 E and 3E and on the west slope 1W, 2W and 3W. The cumulative horizontal and vertical displacements for prisms situated along these critical sections are shown in the following Figures in Appendix 1; East Slope ? Figure A1.9 : Section 1E, Cumulative Horizontal Displacements ? Figure A1.10 : Section 1E, Cumulative Vertical Displacements ? Figure A1.11 : Section 2E, Cumulative Horizontal Displacements ? Figure A1.12 : Section 2E, Cumulative Vertical Displacements ? Figure A1.13 : Section 3E, Cumulative Horizontal Displacements ? Figure A1.14 : Section 3E, Cumulative Vertical Displacements PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 5 : Case Study 1, Navachab Open Pit Page 5-11 West Slope ? Figure A1.15 : Section 1W, Cumulative Horizontal Displacements ? Figure A1.16 : Section 1W, Cumulative Vertical Displacements ? Figure A1.17 : Section 2W, Cumulative Horizontal Displacements ? Figure A1.18 : Section 2W, Cumulative Vertical Displacements ? Figure A1.19 : Section 3W, Cumulative Horizontal Displacements ? Figure A1.20 : Section 3W, Cumulative Vertical Displacements The resultant horizontal vector directional movements of the non-failed pit slopes are shown in Figures A1.21 and A1.22 and the vertical movements in Figures A1.23 and A1.24. 5.12.1 East Slope There are two significant aspects of the displacement behaviour of the east slope; 1. Firstly, with reference to Figures A1.9, A1.11 and A1.13 it is clearly evident that the north- eastern sector of the east slope is undergoing significantly higher levels of deformation than the south-western sector for an equivalent height of wall. 2. Apparent in the accumulated horizontal and vertical movements is that there are up to four phases in the average rate of change of the accumulated displacements as shown in the horizontal prism displacements for Section 2E and 3E (Figure A1.11 and A1.13). 3. Secondly, horizontal vector direction movements (not magnitudes) of the north-eastern sector clearly show a consistent westerly strike of approximately 331? compared with the vector direction movements of the south-western sector which show a consistent easterly strike of approximately 3?, a difference of 32?. There are four distinct phases in the average rate of change of the accumulated displacements. These are shown in the horizontal prism displacements for Section 2E and 3E (Figure A1.11 and A1.13) as well as for prism groups 1 and 2B as shown in Figures A1.28 and A1.29. These phases are summarised as follows; Phase 1 ? Pre March 2001 failure. Phase 1 is generally characterised by a constant rate of creep. Phase 2 ? Post March 2001 failure to May 2005 (accelerated mining). Phase 2 is generally characterised by a slightly reduced constant rate of creep as compared with Phase 1 even though the pit is significantly deeper. Phase 3 ? Accelerated mining from May 2005 to February 2006-07-09 (end of mining). This phase is characterised by significantly higher constant creep rate of the slope with no signs of continual acceleration or progressive behaviour. Phase 4 ? Post end of mining. Reduction in creep rates after completion of mining. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 5 : Case Study 1, Navachab Open Pit Page 5-12 The transitional events which initiated the different phases of displacement behaviour described above are discussed in further detail below; Phase1/Phase 2 It was evident that at the time of the March 2001 failure the average directional strike of the north- western sector prisms changed as shown in Figure A1.25 and have been consistently trending with the new strike till present day. This change in direction is only apparent with the older prisms which are higher on the slope. This directional change in movement is further illustrated in Figures A1.26 and A1.27. The March 2001 failure therefore created a change in both directional strike as well as average rates of displacement. Phase 2/Phase 3 It is not yet apparent what is responsible for the May 2005 start of accelerated displacements, but is likely to be accelerated mining at the base of the pit. Phase 3/Phase 4 Phase 3 to phase 4 transition was effected by the completion of mining at the base of the current pit. Discussion The March 2001 failure resulted in a consistent directional change in vector directional movements for the north-eastern sector prisms above the level of the foot wall thrust fault as well as a net change in average rate of movements for all prisms on the slope. A possible reason for this pattern of behaviour could be stress relief deformation along the footwall thrust fault, resulting in dextral movement of the rock mass to the south of the fault relative to the rock mass to the north of the fault. 5.12.2 West Slope West slope displacement behaviour is characterised by relatively small directionally consistent movements normal to the strike of the slope. This is typical of situations where all dominant structures have favourably dipping orientations (into the slope). As illustrated in the stereogram in Figure A1.5 the following joint intersection planes form kinematically possible wedge failure mechanisms, 1. Joint sets J2 and J6 at a joint plane intersection strike of 169 ? (pit north) and dip of 15?. The prism 10-P146 wedge failure of 26th June 2002 had a vector directional strike of 167 ? and dip of 13?. 2. Joint sets J2 and J4 at a joint plane intersection strike of 213 ? (pit north) and dip of 44?. West slope displacements are in general considerably less than the eastern slope, principally as a result of a flatter OSA as well as favourable structural orientations (dipping into the slope). This confirms the hypothesis that the most significant component of the overall displacement occurring can be accounted for by displacement along discontinuities or structures. 5.13 Discussion None of the 1st or 2nd order geological structures form kinematically possible slope failure sliding planes or failure wedges. All failures recorded at Navachab have been 3rd order structurally controlled, with bedding planes and various sets of joints having played a role. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 5 : Case Study 1, Navachab Open Pit Page 5-13 It is important to note how confusing slope velocity charts can be. The March 2001 failure indicated a possible rate of movement pulse between August 1999 and June 2000, but the average rate at which creep continued was constant. The final rate of movement pulse, starting in approximately January 2001 corresponded with significant increases in deformational movement. Mining activity or the mining rate did not play a role in initiating or controlling the March 2001 failure, but does have a significant effect on the overall rate of displacement (creep) of the non-failed slope. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 6 : Case Study 2, Bibiani Open Pit Page 6-1 6 CASE STUDY 2, BIBIANI GOLD MINE, GHANA 6.1 Introduction Bibiani Mine is located in the Western Region of Ghana about 90km west of Kumasi and is operated by AngloGold Ashanti (Bibiani) Ltd (AABL). Geological terrain is predominantly metasediments (phyllites) with subvertical folliations. The pit was selected as a case study due to the considerable number of failures experienced, which were characterised on the west wall by extremely rapid accelerations, often with little or no warnings and on the east wall as relatively slow developing wedge type failures. All figures and tables pertaining to this section which are not shown in the main body of the report are contained in Appendix 2 of the report. A photograph of the open pit is shown in Figure A2.1. 6.2 Regional Geology The regional geology is described in the Feasibility Study Report (Minproc Engineers Limited, 1995) as follows; ?The Bibiani mine is situated on the Eburean tectonic province which forms part of the extensive West African Precambrian shield (Figure 6.1). The Eburean is dominated by three main sequences namely the Lower and Upper Birimian and the younger Tarkwa ian series. The lower Birimian series consists mainly of phyllites, schists, meta greywackes and in places, metavolcanics. This dominantly metasedimentary series is overlain and, in part, interbedded with dominantly metavolcanic units grouped as the Upper Birimian series. The metavolcanics include lavas and pyroclastics, but also contain fine to medium metasediments not unlike the Lower Birimian. ?The Birimian is overlain by a thick Tarkwaian sedimentary sequence largely derived from older Birimian rocks and granitoid intrusions. Structurally the Birimian units are intensively folded and faulted, whereas the Tarkwaian units display more broad scale folding and overall less tectonic disruption. Both Birimian and Tarkwaian units display low or medium grade regional metamorphic effects.? In the central and southern pit a porphyry dyke, which is up to 80m wide, strikes in an approximate north-easterly direction closely following the hanging wall of the reef. The dyke bifurcates in the north of the pit with the major branch of the dyke continuing north-eastwards and the other minor branch continuing northwards parallel to the reef? (Figure 6.2). 6.3 Structural Geology The structural geology was summarised from the following reports; Bibiani Goldfields Limited (1997), Goel (2003) and Marshall and Mercer (2006). 6.3.1 Shearing and Faulting There is considerable evidence of shearing within and at the margins of the orebody. The main shear planes exposed in the pit can be identified by the presence of evidence of directional shearing movement having taken place along the plane. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 6 : Case Study 2, Bibiani Open Pit Page 6-2 1st Order (Primary) Shears The feasibility study (Minproc Engineers, 1996) describes the primary or main regional shear zone as running between 035? are almost true north and the di p of the shear zone varies from 60?-70? on the southern end, to 75? towards the central shaft, to sub-vertical at northern end of pit. Due to the geometrical shape of the pit, the primary shear does not appear to currently influence the wall stability. 2nd Order Shears 2nd order planes of shearing and faulting are especially evident in the meta-sediments either side of the main porphyry dyke and main shear zone. The dip and strike of the 2nd order shears are different either side of the dyke and are considered to have significant influence on the stability of the pit walls. On the west slope the 2nd order shears are sub-vertical, east to south-east dipping and are slickensided. The shear zones appear to be curvi-planar, both down dip and along strike, and occur in the south of the pit at small angles to the bedding. In the northern part of the mine area the shears steepen and become more oblique cross cutting the bedding planes. The direction of the shearing was measured at a plunge angle of between 30? to 40? to the north (Marshall and Mercer, 2006). PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 6 : Case Study 2, Bibiani Open Pit Page 6-3 On the east slope a well defined/major west dipping shear zone is present. The nature of the shear zone indicates that significant movement has taken place along the direction of shear, with gouge having been present between the two surfaces. Unlik e the west shears, this shear zone is not slickensided. The dip of the shear zone ranges from subvertical to 48?. The shear orientation was measured on pit section 5760N adjacent to the ramp as 046?/247? (dip/dipdir, pit north) (Marshall and Mercer, 2006). The 2nd order shears are generally persistent and where exposed, appear to extend over a significant length of the pit. All shears identified have unfavourable directions and therefore appear to negatively influence the overall stability of the walls. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 6 : Case Study 2, Bibiani Open Pit Page 6-4 3rd Order Shears Relatively minor 3rd order shears are also evident, especially on the west wall of the pit. They follow the trend of the 2nd order shears, are not persistent, appear to extend over relatively short distances and appear to only influence bench scale stability. Stereonets Orientations of the shears were obtained from structural mapping undertaken by the mine, the Geotechnical Investigation Report (Marshall, 1997) and orientations obtained at 150mRL and below (Marshall and Mercer, 2006). No shear zones were identified in the Kinematic Stability Analysis Report (Goel, 2003). Due to the curvi-planar nature of the shears, the orientations tend to lie within a range of 58?/50? to 87?/110? (dip/dipdir, pit north). The upper end of the orientations are sub-parallel with the mean foliation orientations. Shear orientations are summarised in the stereonets in Figure A2.2 and A2.3. 6.3.2 Foliations/Bedding The foliation planes of the phyllites/greywackes are considered the dominant controlling discontinuity in the pit. Foliation thicknesses range from millimetre to centimetre order of magnitude. It is likely that the foliation is a result of intense folding, however the shearing and the absence of marker units make identification difficult. In general foliation dip angles are near vertical (70? to 90?) in the south of the pit, with dip directions varying between east and west. Towards the north of the pit the average dip angles decrease, with dip directions only to the west. The foliation strike is sub-parallel to both the east and the west walls, varying over the range 351? to 011? (p it north). Foliation orientations obtained from the structural mapping undertaken by the mine are summarised in Figure A2.4. It is clear from the foliation orientations that toppling type instability is likely to have a significant destabilising influence, especially in the north of the pit where dip angles reduce (Marshall and Mercer, 2006). 6.3.3 Joint Sets West wall structural mapping undertaken by the mine between 1997 and 2001 indicates a primary joint set with a mean dip/dipdir (pit north) of approximately 83?/ 186?. This joint set is therefore sub-vertical and cross cuts the west face of the pit almost at right angles. The spacing was estimated to be in the order of 50 to 60 metres (Marshall and Mercer, 2006). Its orientation is illustrated in Figure A2.5. A summary of all east and west wall joint sets was obtained from Marshall (1997). These are listed in Table 6.1. Joint spacing was recorded as being in the order of 2 to 3m. Stereographic projections for the West Wall are shown in Figure A2.6 and for the East Wall in Figure A2.7. 6.4 Weathering Profile The weathering profile is divided into weathered, partially weathered and fresh rock zones also termed oxide, transition and sulphide zones. The base of the weathering or oxide material is generally undulating and is a mirror image of the topography, being deeper under topographic highs and shallower below flat lying areas. At the highest section of the mine alongside Central Hill, complete weathering has occurred to a depth of approximately 85 metres, becoming shallower to the east and south at approximately 35 to 40m depth. In the lower lying areas the base of the oxide zone is shallower, with an average depth of 25m. The thickness of the transition zone is generally in the order PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 6 : Case Study 2, Bibiani Open Pit Page 6-5 of 10 to 15m around the pit rim, but becomes considerably deeper in the northern and southern end of the pit in the mineralised zone (Marshall and Mercer, 2005). Table 6.1 : West and East Wall Joint Sets Pit Wall Joint ID Dip (mean) Dip Dir (mean) (True North) Dip Dir (mean) (Pit North)* West JS1W 72 125 91 West JS2W 78 347 313 West JS3W 83 280 246 West JS4W 24 348 314 West JS5W 19 182 148 West JS6W 53 310 276 East JS1E 66 157 123 East JS2E 53 104 70 East JS3E 22 106 72 6.5 Groundwater Very little is known about the local groundwater profile which is likely to be considerably affected by the continued dewatering of the underground mine. The foliations, shears and joints are considered to create significant anisotropic permeability conditions within the rock mass. The southern end of the open pit is situated in an existing river course which has been dammed. Consequently, as a result of seepage water, the southern end of the pit is wet with visible water flowing down some sections of the wall. Some drain holes along the southern section of the west wall on the 150 mRL bench (and below) were flowing, and in general, the ground water level appeared to be close to the pit face (Marshall and Mercer, 2005) 6.6 Geological Sections The geological section is orientated normal to the pit wall and situated approximately along the centreline of the main sulphide failure. The representative section has a pit bearing of 90? which is normal to east pit face at this location. A pit layout, illustrating principal geology, representative sections and monitoring points, is shown in Figure 6.3. Mining voids, which are situated within the reef, dip sub-vertically under the east slope potentially destabilising the slope. These voids however are not considered to have a significant destabilising effect on the west slope. 6.7 Rock Classification and Strengths The main rock mass through which the open pit has been excavated comprises thinly to widely foliated, well jointed, strong silicified phyllites. Both the phyllite and greywacke rock masses have similar geotechnical characteristics. These are summarised in Table A2.1. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 6 : Case Study 2, Bibiani Open Pit Page 6-6 6.8 Slope Configuration Parameters Overall slope configuration parameters are summarized in Table 6.2. Table 6.2 : Surveyed Slope Angles Material RL(m) Slope Angle (0) Berm Width (m) Inter-ramp Angle (?) Oxide 312 ? 222 32 - 35 8 29 Transition 222 ? 204 55 - 60 7 ? 8 49 Fresh 204 - 78 58 - 68 3 - 6 50-53 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 6 : Case Study 2, Bibiani Open Pit Page 6-7 6.9 Displacement Monitoring Systems The deformation data was obtained from a high precision manual survey system. Surveys of prisms both on the west and the east walls was undertaken on a relatively regular basis. A layout of the prism monitoring network is shown in Figure A2.8. 6.10 History of Failure/s The west slope of the open pit has experienced seven failures in total. Four relatively significant failures/instabilities have occurred on the east slope (Bibiani Geotec Dept, 2001 to 2005) and (Marshall, 2004) which are detailed below. 6.10.1 West Slope Failures The west slope of the open pit has experienced 7 failures in total. The most significant failures are summarised below: Failure No W2 and W3 Failure name: 1st and 2nd Oxide Failures (see Figure A2.9) Date: September 2001 Preceding observations: The potential development of the oxide failure was first noted on the 6/3/2000. A slip plane was observed on the 276m-258m-batter face during the routine morning pit inspection. An extensive tension crack had developed on the berm and could be traced along the batter face of the 294m bench, but was found to be terminating along the face. Location and Extent: Failure in oxide material from crest (294) on Central Hill to 222 bench. Mechanism: Slip circle failure in oxides possibly influenced by residual structures(shear planes) in the oxide ? vertical scarp face on rear failure surface. This was confirmed in the mine?s slope report dated 04/04/02. Possible Triggers: Initial movement and failure along shear plane in underlying sulphides as reported earlier in the same month. Fairly steep slope/high batter angle for oxide material, Raised perched phreatic surface, as the failure occurred after a period of high rainfall. Remedial measures: Mining away failed material and flattening of the overall slope The cleanup of the failed material along the toe of the failure precipitated a second slip failure within the perimeter of the 1st failure (W3) (Marshall and Mercer, 2005). Failure No W5 Failure name: 1st Sulphide Failure (Figure A2.10) Date: 10th October 2002. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 6 : Case Study 2, Bibiani Open Pit Page 6-8 Location and extent: Material was displaced from 255m elevation to 123m elevation. The width of the failure was approximately 120m along the strike of the wall. The crest of the failure is about 2m from the crest of the intermediate berm (255mRL) and the toe is approximately on the 132m RL (Bibiani Geotec Dept., 2001 to 2005). Mechanism: The failure was structurally controlled by the slickensided shear plane. It was bounded on the southern side by the day lighting shear plane extending from the 222m bench to 132m bench. The failure did not completely follow the shear plane to base of pit. It is likely that an extension of this shear plane possibly defined the extent of the earlier September 2001 oxide failure. Possible triggers: Structural instability as described above, high rainfall and the ingress of runoff water. It was reported that due to September 2001 failure drainage channels had become silted up, resulting in runoff becoming trapped. This would have increased ingress of water into the rock mass (Marshall and Mercer, 2005). . Failure No W6 Failure name: 2nd Sulphide Failure, Main Sulphide Failure (Figure A2.11) Date: 23rd November 2003 Preceding observations: On the 15th October 2003 it was reported that, on the 273mRL bench at the north western segment of the pit within an oxide zone, there was a fracture trending radially across the face of the wall just beneath the berm. There was evidence of slip as the fracture was seen to cause vertical displacement downwards and the slip plane is day lighting. The fracture was continually widening. It was reported that just beneath the 273mRL, on the 240mRL berms, three sets of fractures had developed which were described in detail. Location and extent: The failure extended vertically from 258mRL down to the pit floor and onto the 78mRL to the north (burial of parked up excavators). The failure extended for an approximate strike length of 300m along the north-west of the pit wall. Mechanism: The failure appears to be composite in nature. The upper elevations were possibly bounded by a shear plane in the weathered zone. It is likely that this is an extension of the structurally controlling shear plane of the September 2001 and October 2002 failure. The lower sections of the failure, which are situated predominantly within the fresh rock, were of a toppling type mechanism resulting from the fall of overlying material. It is likely that it was controlled/influenced by the shear plane. Failure extended from 294 mRL bench to 75mRL bench. Failure occurred with very little warning. The survey monitoring system recorded inconsistent readings, with only slight increases in displacement in 4 prisms of only 2mm prior to failure. Doubt PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 6 : Case Study 2, Bibiani Open Pit Page 6-9 was expressed in the pit and failure inspection reports with regard to the overall accuracy of the survey system at the time of this failure. Possible triggers: Initial oxide failure from slip on the 273mRL bench. Propagation of structural instability along shear plane as described above. High rainfall over previous few days resulting in high water tables in both oxides and sulphides. Bulldozing of failed oxide material into pit further loading benches below (Marshall and Mercer, 2005). Failure No W7 Failure name: 3rd Sulphide failure Date: January 2004 Location and extent: The failure extended from approximately 5860N to 6030N. Mechanism: Likely that it was also controlled/influenced by shear plane. Possible triggers: Possibly an extension of the west wall failure of Dec 2001. Structural instability/toppling failure due to flatter foliations. High rainfall and seepage from the historical valley (Marshall and Mercer, 2005). 6.10.2 East Slope Failures/Instabilities Four relatively significant failures/instabilities have occurred on the east slope of the pit. The most significant failures are summarised below: Failure E3 (Pit closure events) Failure name: 2 Ramp failures (Figure A2.12) Date: January 2005 (exact dates of different failures not known) Location and Extent: Two significant failures have occurred immediately below the east wall ramp from pit section 5420N to 5570N. Mechanism: The failures are of a sliding type and are almost certainly structurally controlled by the prominent east wall shear. The failures tend to be delineated by crosscutting steeply dipping joint sets and are characterised by deep vertical cracks possibly extending to the plane of movement. Possible Triggers: Failure was possibly induced by ongoing stress relaxation as a result of the presence of underground voids immediately below ramp (Marshall and Mercer, 2005). . PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 6 : Case Study 2, Bibiani Open Pit Page 6-10 6.11 Deformation Behaviour of Slopes Due to several problems with the manual monitoring of the prisms there are unfortunately gaps in the monitoring data. West Wall Sulphide Failures Horizontal deformation behaviour leading up to the October 2002 failure is shown using the 2 months of data for prisms 168-1 and 186-1, in Figure A2.13a. The prisms were experiencing a creep of 0.74mm/day and 1.31mm/day, and displayed progressive behaviour only 1 to 2 days prior to catastrophic collapse. There was no discernable increase in vertical displacements prior to failure (Figure A2.13b). West Wall Oxides Failures The prism data for the oxides prior to the failures was considered to be insufficiently accurate. However, post collapse recovery deformation data for horizontal and vertical displacements was able to be obtained following the January 2004 failure from prisms installed shortly after the failure. This prism data is shown in Figures A2.14 and A2.15. General deformation behaviour associated with on going creep and relaxation of the non-failed sections of the West wall is shown in Figure A2.16. East Wall Ramp Failures General East wall deformation and deformation associated with the E3 ramp failures, is shown in Figures A2.17a and A2.17b. Unlike the sudden accelera tion shown by the west wall sulphide failures, the east wall failure displayed a considerably longer period of progressive behaviour before failure. This can be attributed to the different mechanism involved with the E3 failure which was a sliding failure in contrast to the brittle structurally controlled shear plane failures on the west wall. The remainder of the prisms reflect slow steady state creeping associated with normal relaxation of the pit wall. 6.12 Interpretation of Displacement Behaviour, Failure mechanism and Trigger 1) Almost all failures occurred within or immediately after the high rainfall season of August to November. This suggests that the stability of the rock mass was sensitive to raised groundwater/phreatic surfaces in and around the pit. 2) All west wall failures appear to be interconnected in that they are structurally controlled and bounded by a prominent sub-vertical east dipping undulating slickensided shear zone. It is likely that instability propagated northwards along this shear zone. 3) Classical topping failure modes are only evident towards the northern side of the west wall where the dip of the foliation planes is flatter. It is likely that, in the sections of the pit where toppling modes of failure are evident, toppling failure is not a primary failure mechanism, but has resulted from falling failed or bulldozed material. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 6 : Case Study 2, Bibiani Open Pit Page 6-11 4) East wall ramp failures are of a sliding type and are structurally controlled by the prominent west dipping shear zone and/or underground voids. The failures tend to be delineated by crosscutting steeply dipping joint sets and are characterised by deep vertical cracks possibly extending to the plane of movement. 5) Due to the east wall geometry and undercutting voids, none of the ramp failures are considered to be influenced by seasonally high phreatic surfaces/ groundwater tables. 6.13 Discussion In spite of limiting survey monitoring data, the rock mass behaviour at the Bibiani pit can be summarized as follows: 1. Catastrophic failure preceded by very rapid progressive behaviour. This can be attributed to the sub-vertical dipping foliations and shear zones. 2. Compound shear/toppling failures on the west wall and wedge/sliding failures on east wall. 3. Propagation of instability along shear zones. 4. Displacement behaviour showed evidence of dry season /wet season cycles. 5. Triggers were principally associated with failure in overlying oxides following significant rainfall events. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 7 : Case Study 3, Mt Keith Open Pit Page 7-1 7 CASE STUDY 3, MT KEITH OPEN PIT, WESTERN AUSTRALIA 7.1 Introduction The Mt Keith MKD5 nickel sulphide deposit is 80 km South of Wiluna and 90 km North of Leinster, at 27?14'50''S, 120?32'30''E. The ultramafic deposit is situated on the Norseman-Wiluna Greenstone Belt. The geological structure is vast and complex. The host rock is heavily sheared, contains low shear strength infill materials and the shears themselves bifurcate, anastromise and are wavy in all directions (Adams, 2006). It is alleged by several authors (Adams (2006), Dight (2006) and Steffen (2005)) that the stability of the Mt Keith Operations (MKO) open pit is significantly influenced by very high insitu natural rock stresses. Instabilities in the form of multi-bench to bench scale (and smaller) failures are common with several generally occurring on each mining bench. Larger multi-batter scale failures are also relatively frequent. Approximately 50% of all structure daylights on the batters, which is a major cause of batter scale instability and several batter scale failures have occurred, which involve two or more structural sets (Goodchild, 2006). The fresh rock failures characteristically have a relatively rapid onset of collapse. A total of 141 failures have been recorded and detailed by the mine between June 2003 and July 2006. Seven of these failures, which were selected based on relatively good (radar) monitoring data were reviewed in detail for this research. One of the failures was a relatively large overall wall failure in the weathered saprolites and the rest were batter or multi-batter scale fresh rock failures. A photograph of the open pit is shown in Figure A3.1. 7.2 Regional Geology The Mt Keith MKD5 deposit occurs in the Agnew-Wiluna greenstone belt of the Yilgarn Craton. In the vicinity of Mt Keith the greenstone belt is narrowly sandwiched between the Mt Keith Granodiorite to the west and the Perseverance Granitoid to the east and lies near the end of a wedge bounded by the regional NNW-trending Erawalla Fault to the west and the North trending Perseverance Fault to the east (Figure A3.2) (Haywood, 2004). The regional geology comprises volcanic and volcanoclastic rocks (felsic and mafic), sedimentary rocks (pelites, shales and cherts), basalts (pillowed, tholeiitic and high Mg) and komatiite sequences (olivine ad-, meso- and ortho-cumulates, and thin spinefex-structured flows) that have undergone mid- lower greenschist facies metamorphism. The olivine mesocumulates that host nickel sulphide mineralisation in the region are the thicker, lenticular parts of the komatiitic sequences and formed as channelised flows (MKO Geotec Dept, 2006). 7.3 Stratigraphy and Weathering The Mt Keith stratigraphic sequence consists of a mafic and footwall felsic sequence in the east overlain by the Mt Keith Ultramafic Complex (MKU C). The hanging wall to the west of the ultramafic contains a pyritic chert and black shale at the contact and is overlain by a mafic sequence and finally a second (Cliffs) ultramafic unit. The strigraphy is summarised in Table 7.1, illustrated in Figure 7.1 and described in further detail below. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 7 : Case Study 3, Mt Keith Open Pit Page 7-2 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 7 : Case Study 3, Mt Keith Open Pit Page 7-3 Table 7.1 : Summary of Lithological Domains and Lithological Codes (Mining One ,2003 and Goodchild, 2006)) Domain Code Footwall Sequence (East) Footwall Contact Dolerite FCD Footwall Siliceous Volcanics FSV Footwall Felsics Footwall Sediments FWS Footwall Dolerite FWD Undifferentiated Footwall Sediments UFW Orebody Sequence (Central) Mesocumulate MES Hangingwall Orthocumulate HWO Western Adcumulate WAD Pentlandite Adcumulate PEN Footwall Orthocumulate FWO Millerite Adcumulate MIL Heazlewoodite Adcumulate HZL MKD5 Hangingwall Sequence (West) Undifferentiated Hangingwall Mafics UHM Cliffs Ultramafics CUM Hangingwall Mafics HWM Hangingwall siliceous Volcanics HSV 7.3.1 Footwall Sequence (East) Regolith Four to ten metres of transported material unconformably overlies the saprolite horizon of the footwall lithologies. The depth of the saprolite horizon over mafic volcanics varies from 35 to 100 meters and to more than 120 meters over felsic volcaniclastic rocks. The saprolite has a significant clay content which is dominated by kaolinite, especially over felsic lithologies. The depths given are averaged as the variation of the stratigraphy, coupled with shearing, results in an undulating weathering profile over the sequence. Iron oxide staining is a common feature of the saprolite zone over the mafic rocks (MKO Geotec Dept, 2006). Fresh Lithology Zone The footwall lithologies to the MKUC are composed of felsic sedimentary rocks and intermediate to felsic intrusives, volcanics and volcaniclastic rocks interbedded with minor fine grained mafic rocks. Quartz, plagioclase (albite), carbonate, chlori te, sericite, epidote and leucoxene are important components of these units. Intense quartz-carbonate alteration, associated with cross-cutting structures, is a common feature near the ultramafic contact (MKO Geotec Dept, 2006). PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 7 : Case Study 3, Mt Keith Open Pit Page 7-4 7.3.2 Mt Keith Ultramafic Complex (MKUC) Regolith Regolith over the MKUC consists of a transported zone overlying the saprolite and then fresh rock. The boundary between the transported material and the upper saprolite is marked by collapsed ferruginous saprolite composed almost entirely of iron and silica. Below the saprolite is the saprock which shows a gradational change to fresh rock. The saprock is the dominant regolith horizon and the main expression of weathering is the oxidation of sulphides. Clay development in the saprolite is limited to smectite formation over the more aluminium rich peridotites (MKO Geotec Dept, 2006). Fresh Lithologies (MKD5 + Unit 104) The MKUC represents a series of ultramafic extrus ive flow units which display a gradational change in igneous texture as a result of flow and cooling characteristics. Overall the rock is medium grained, with colours dominated by shades of green and black. Rock types range from dunite to peridotite (MKO Geotec Dept, 2006). 7.3.3 Hangingwall Pyritic Cher t and Black Graphitic Slate Regolith These hangingwall units are generally resistant to weathering. The black graphitic slate is silicified close to the surface together with oxidation of the sulphides. The pyritic chert forms an ironstone gossan at the surface due to the oxidation of massive pyrite. These units maintain their structure close to the surface (MKO Geotec Dept, 2006). Fresh Lithology The lowermost part of the hangingwall sequence is made up of a unit of pyritic chert overlain by a package of volcanogenic sedimentary rocks with multiple fine units of black graphitic slate. The pyritic chert is dominated by fine microcrystalline silica and minor carbonate which is white to grey in colour, massive coarsely crystalline pyrite and pyrrhotite and black graphitic slate. Other minor sulphides include chalcopyrite and sphalerite. In the south the lower part of the pyritic chert is dominated by silica overlain by a mixed zone followed by massive sulphide. In the north massive sulphide is interlayered with silica throughout the unit (MKO Geotec Dept, 2006). The pyritic chert is overlain by laminated and foliated volcanogenic sedimentary rocks containing fine layers of black graphitic slate. The black graphitic slate consists of fine grained graphitic slates with smeared pyrite crystals on bedding planes. The remainder of the sediment package is dominated by fine grained felsic to intermediate sediments. The dominant minerals are plagioclase (albite), quartz, carbonate, sericite and chlorite (MKO Geotec Dept, 2006). 7.3.4 Hangingwall Mafics and Felsics Hangingwall Regolith Up to eight metres of transported, unconsolidated alluvial sedimentary rocks unconformably overlie the residual rock which has been stripped to clay saprolite. The depth of the saprolite horizon over the mafic volcanic rocks varies from 4 to 28 metres while the base of weathering varies from 12 to 28 metres. Iron oxides are common throughout the regolith profile of the mafic rocks. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 7 : Case Study 3, Mt Keith Open Pit Page 7-5 The regolith over mafic lithologies is not as well developed as over ultramafic rocks. The upper saprolite is the most well developed horizon while the saprock is more limited. The saprolite is dominated by clays, especially smectite and often a large kaolinite component in the upper saprolite (MKO Geotec Dept, 2006). Fresh Lithology The remainder of the hangingwall is made of mafic rocks. The general stratigraphy consists of a mafic basalt and mafic dolerite with volcanogenic sediments at the interface between the two units. The basalt is a fine grained mafic rock which consists of plagioclase (albite), amphibole (actinolite), quartz, chlorite, carbonate, epidote, sericite and leucoxene. The dolerite is a medium grained rock interpreted to be a quartz dolerite. It appears as a fine to medium grained groundmass containing quartz phenocrysts and patches of leucoxene. The main mineralogy consists of quartz, plagioclase (albite), amphibole (actinolite), chlorite, carbonate, epidote, sericite and leucoxene. Between these two units is fine grained chloritic volcanogenic sediment, most likely of mafic origin. Minor bands of black graphitic slate are found within this unit (MKO Geotec Dept, 2006). 7.3.5 Cliffs Ultramafic Cliffs Ultramafic Regolith The regolith development over the Cliffs Ultramafic Unit is similar to the mafic lithologies although the saprock is more well developed. The Cliffs Ultramaf ic Unit contains higher aluminium than the MKUC so smectite development in the saprolite is significant, but kaolinite is a minor component (MKO Geotec Dept, 2006). Fresh Lithologies The Cliffs Ultramafic Unit is an ultramafic unit consis ting of a spinefex textured flow top to the west of an orthocumulate to mesocumulate textured base. The ultramafic has been completely serpentinised and displays the same mineral assemblages and range of alteration types as the MKUC. The contact between the Cliffs Ultramafic unit and the underlying sediment is intensely sheared and talc-carbonate altered (MKO Geotec Dept, 2006). 7.4 Structural Geology A description of the structural geology of Mt Keith is summarised from Hayward (2004) and illustrated in Figure 7.1. 7.4.1 1st Order Structures (Regional Structures) The principal 1st order regional structures consist of the Erawalla Fault and the Perseverance Fault which are discussed briefly below. The NNW-trending Erawalla Fault defines the west boundary of the greenstone belt (Liu et al., 2002) and joins the Six Mile Fault further south. It dips steeply eastwards with a shallow 2?-15? plunging mineral lineation and dominantly sinistral strike-slip shear fabrics. This fault appears to form part of the province-scale Keith-Kilkenny Lineament, a major sinistral wrench (Hayward, 2004). The NNW to NS-trending Perseverance Fault defines the eastern margin of the greenstone belt (Liu et al., 2002) and is continuous over >200 km strike ext ent. It comprises a major, subvertical, 1 to 6 km PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 7 : Case Study 3, Mt Keith Open Pit Page 7-6 wide, high-strain shear zone bordering the eastern Perseverance Granitoid batholith, similar to the Waroonga Shear Zone west of Agnew. The Fault has a steep SSE-plunging elongation lineation indicative of dominant dip-slip rather than strike-slip shear, but does show an apparent sinistral shear sense in plan view where it has a more NNW-strike (Hayward, 2004). 7.4.2 2nd Order Structures (Folds, Major Shear Zones and Major Faults) Aeromagnetic imagery (Figure A3.3) shows that Mt Keith Granodiorite contains a swarm of prominent NNE-trending faults >20 km long, spaced around 1 km apart, but with little apparent displacement. These faults trend towards MKO, crosscut the Erawalla Fault, and are also expressed in the mine area. Major EW-trending faults cut earlier-formed NNW-trending shears with apparent small sinistral wrench displacement in plan view. In contrast, the granite batholiths east of MKO show a NW-trending shear zone with apparent sinistral shear in plan view which converges on the north end of MKD5 (Haywood, 2004). The same regional deformation patterns are mirrored in the lower, 2nd and 3rd order structures found locally within the deposit as discussed below. Folds The Mt Keith deposit sits on the west limb of a regional very tight, south-east plunging anticline, the axis of which is missing due to granitoid emplacement to the east. The consistent steep-west dip of the axial plane schistosity here suggests that these regional folds were slightly overturned to the east There are no mine-scale folds visible within the deposit. Only two generations of minor mesoscopic folds occur within the pit hangingwall and footwall schists: (i) recumbent folds, and (ii) steeply W- plunging kink folds (Haywood, 2004). North-South Strike Faults The most dominant set of talc-magnesite-altered faults strike NS in the southern half of MKD5 to NNE- SSW in the northern half, in each case oriented (sub) parallel to major lithological domain contacts. They comprise brittle ductile faults to ductile shears with strike extents between 800m to >1200m long, spacings of 100-300m, and dips steeply 65-85? W to WNW (270-300?), with local steep east dips. Talc-magnesite alteration along these faults within MKD5 ranges from 0-10m wide. Figure A3.4 shows NS strike faults located within the north wall of stage E pit. These faults are more abundant in the north half of MKD5 where several appear to converge at depth and southward as splays from the Hangingwall Shear Zone (Haywood, 2004). East-West Cross Link Faults Several 200-800m long EW-trending faults with 65-85? dips were identified in pit mapping. These faults are spaced 100-300m at the south end of MKD5, but increase in spacing further northward. They are confined between, and naturally terminate against, the north-south strike faults. Most of these faults are associated with subvertical, EW-trending, sheeted joints that localise broad, 2-10m wide, talc- magnesite replacement zones. Figure A3.5 shows examples of EW Cross Link Faults within the SE wall of Stage E pit (Haywood, 2004). Northwest Wrench Faults NW trending wrench faults are most commonly developed in the southern half of the Mt Keith Mine area. They typically dip 70-85? SW, are spaced 80-700m apart, and exhibit strike extents >500m. These faults offset both the NS strike and EW cross link faults with sinistral wrench displacements of 3 to 50m. Figure A3.6 shows examples of NNW Wrench Faults (Haywood, 2004). PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 7 : Case Study 3, Mt Keith Open Pit Page 7-7 North-Northeast Wrench Faults NNE dextral wrench faults are very common throughout the Mt Keith Mine area and regionally. Although their orientation and location partially overlap those of the more NNE-striking components of the NS strike fault set, they are distinguished from the latter by their more brittle nature, lack of attendant talc-magnesite alteration, consistent dextral shear component and overprinting relationships. The NNE wrench fault set ranges in strike between NS and NE and typically exhibits reverse-dextral oblique slip with displacements of 5-100m in the mine area. Most are subvertical and spacing typically ranges of 50-150m apart. Their frequency is greatest in the central part of MKD5 (Haywood, 2004). NNE Conjugate Reverse Faults A set of NNE-trending conjugate reverse faults with subordinate flat thrust faults is also widely developed throughout the Mt Keith Mine area. This fault set is not associated with significant talc- magnesite alteration in MKD5 and these have small strike extents. These are relatively narrow brittle faults which dip moderately 30-70? ESE and WNW, and are typically spaced 10-20m apart with strike lengths of 200-500m. Some moderately ESE-dip and WNW-dipping reverse faults were observed to offset and even locally overturn the Footwall Fault zone. These footwall reverse faults present a significant risk for major inter- batter planar failure along the east pit wall. Also of concern is the observation that this set of faults is most commonly associated with groundwater flow into the pit (Haywood, 2004). 7.4.3 3rd Order Structures - Minor Fault and Joint Sets Minor faults with length of 10-100m are very common throughout the MKD5 complex and are typically spaced 3-15m apart. Principal Joint Set Orientations There are at least nine distinct joint and minor fault sets in the Mt Keith Mine area based on orientation clusters: their essential characteristics are summarised in Table 7.2. A more comprehensive table of minor fault and joint set characteristics and orientations is presented in Table A3.1. Table 7.2 : Summary of Joint Sets (Haywood, 2004) Group Minor Fault +Joint Sets Association 1 E-W- trend subvertical 1 Flat Typical of strongly schistose units in hanging and footwalls 2 NS-trend subvertical 2 WNW trend - subvertical 2 NE- trend subvertical Widespread and commonly related to major shears 3 Moderate ESE dip 3 Moderate WNW to NW dip Conjugate set dominantly within MKD5 4 Moderate ENE dip 4 Moderate WSW dip Conjugate set dominantly within MKD5 7.4.4 4th Order Structures - Foliation Both axial planar and shear zone cleavages are generally strongly developed within the MKO mine area, but show coarse partitioning within different litho-structural domains. The dominant cleavage throughout the Agnew Wiluna greenstone belt and within the MKO mine area, strikes NNW to N and is PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 7 : Case Study 3, Mt Keith Open Pit Page 7-8 generally subvertical to steeply W-dipping. Orientation data of the major litho-structural domains is displayed in Figure 7.2. 7.5 Representative Sections A typical section through the pit showing major lithology generated from the Vulcan model is shown in Figure 7.3. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 7 : Case Study 3, Mt Keith Open Pit Page 7-9 7.6 Rock Mass Classification and Strengths Rock mass strengths and classifications were summarised from the Geotechnical Status Report (Mining One, 2003) and are listed in Table A3.2. 7.7 Pit Configuration Parameters Pit configuration parameters are listed in Table 7.3 for the pit as at end June 2006. Table 7.3 : Overall Pit Configurat ion Parameters (July 2006) West Slope (Hangingwall) East Slope (Footwall) Height (m) 325m 325m Overall slope angle (?) (OSA) 33? (before G Stage cut) 24.8? (current) 31? Bench stack angle (?) (BSA) 40 ? (upper) 45.7? (lower) 39.1? (upper) 42.6? (lower) Bench face angle (?) (BFA) 54.5 ? 56? (fresh) Inter ramp angle (?) (IRA) 40 ? (upper) 42.9? (lower) 36? (upper) 39? (lower) Spill/catch berm width (m) Varies (5-10m) Varies(5-10m) Spill/catch berm height (m) 30m (fresh) 30m (fresh) Bearing (?) 90 ? 90? Pit coordinate system orientation is 11? 00? west of true north, 12? 00? west of magnetic north 7.8 Displacement Monitoring Systems The primary slope monitoring system consists of three SSR units An Autoslope/Leica survey monitoring system is used to monitor prism movements. 7.9 Displacement Behaviour Non-failed Pit Wall Sectors All pit sectors have experienced failures. 7.10 History of Failure/s A total of 141 failures have been recorded and detailed by the mine between June 2003 and July 2006. Seven of the larger, most comprehensively documented and monitored failures were selected for detailed review. A list of these seven failures is as follows; ? Failure #1 : 2001-Dec to 2003-Nov SE Wall Failure 544RL (prism monitoring) ? Failure #2 : 2003-12-04 NE Corner 319RL (with radar) ? Failure #3 : 2005-08-04 East Wall (with radar) ? Failure #4 : 2004-08-11 (with radar) ? Failure #5 : 2005-10-06 (with radar) ? Failure #6 : 2006-06-24 Failure 139 (with radar) ? Failure #7 : F-Stage East Wall Slip (prism monitoring) PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 7 : Case Study 3, Mt Keith Open Pit Page 7-10 All information and deformation data pertaining to the failures are contained in Appendix 3 and summarised below. It should be noted that not all the information regarding the failures is complete. Occasionally some data was omitted in the original MKO database. A discussion and interpretation of the failure deformation behaviour is presented in the following section. 7.11 Discussion on the Deformation Behaviour of the Mt Keith Failures The Mt Keith failure database essentially reflects 2 types of failures; 1. A soft rock (weathered rock/saprolite) failure. 2. Rapidly developing fresh rock failures where the potential instability/failure mass is bounded by failure planes or discontinuity boundaries (joints, shears or bedding) This summary therefore focuses on reviewing the characteristics of damage evolution of the second category of failure types at Mt Keith. The general characteristics of these failures are; a) They are generally relatively small failures making them difficult to detect with an ordinary prism network. b) Most of these types of failures occur during or shortly after mining, but some have occurred a significant time after mining (Failure #3 and #4). c) They generally do not reflect a continuous period of specific macro event (mining) influenced deformation behaviour prior to the onset of failure. In reality the onset of failure can occur with little or no prior warning. d) It was generally found that this category of failures has well defined ?failure planes? or discontinuity boundaries (joints, shears or bedding), as shown in the photographs of the failures. e) They are often characterised by visible propagation of instability/failure across the rock mass. The following section describes the phases of damage evolution observed in this category of failures. 7.11.1 Phase 1 : Initiation of Instability As has been mentioned this category of failures most often occurs during or immediately after mining. However the failures often occur in the absence of relatively large scale mining events. Proximal causes of the initiation of instability, leading to the onset of collapse, usually relate to micro events and continuous stress redistribution in the rock mass as a result of mining activities. The categories of possible events leading to these types of failures are; Dynamic Initiation Events By far the biggest initiator of failure surface propagation is blasting induced damage which results in vibrations. Similarly, these types of failures often seem to occur adjacent to ramps which may be PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 7 : Case Study 3, Mt Keith Open Pit Page 7-11 subject to vibrations to from haul trucks or in mining faces possibly bumped by shovels or wheel loaders during clean up operations. Static Initiation Events Static initiation events can be related to the general ongoing relaxation and dilatancy of the rock mass resulting from ongoing stress redistribution. This results in visible crack development in the rock mass and propagation of failure surface/s along existing discontinuities and often bedding. The continuous stress redistribution which takes place can lead to the toes of wedges becoming overstressed. 7.11.2 Phase 2 : Propagation of Instability Propagation of instability is associated with rapid increases in localised deformation most likely as a result of an advancing failure surface, and can be seen in the pre-collapse displacement time graphs. This is illustrated in detail in Figure 7.4 using Failure #4 as an example. The video of failure #4 (Figure A3.15) shows the developing instability which starts with small rocks fall and failure of the toe (at the bottom right of the instability) progressing towards the left generating further minor rockfalls and finally collapse. The displacement time graph in Figure 7.4 confirms the right to left time lag, using the difference in the magnitude of deformation being experienced in the corresponding radar monitoring pixels. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 7 : Case Study 3, Mt Keith Open Pit Page 7-12 It is of importance to note that the left hand sections of the failed rock mass (shown by pixels #2 and #7) experienced almost no deformation before failure. Clearly this section of stable rock mass could not resist (support) the actively failing rock mass on the right, and was finally forced to fail. This is termed a para-sympathetic failure and occurs when a collapse of a section of the rock mass triggers or ?drags with it? adjacent sections of the surrounding rock mass, which is combined into a single near simultaneous collapse. In terms of changing shear strength parameters during propagation of the instability it is considered likely that the sections of the rock mass on the plane of failure behind the advancing failure surface are characterised by residual shear strengths, whilst the rock mass in front of the advancing failure surface remains at peak shear strength, both in the intact rock and on the discontinuity boundary. In summary, the radar images confirm that propagation of instability is associated with rapid increases in localised deformation most likely as a result of an advancing failure surface. The propagation of instability is characterised in the advanced stages of deformation by minor rock falls (puffs of dust) visibly falling from the face of the instability, as progressively larger and larger deformations occur in the rock mass (see photos in Figure A3.9 and Figure A3.15 as examples). These minor rock falls are often seen to be moving progressively across the instability, which is likely to be associated with the movement of the advancing failure surface. 7.11.3 Phase 4 : Collapse Collapse is considered to occur when either; ? the advancing failure surface approaches or passes a discontinuity in the intact rock or rock mass which run across (normal) to the plane of failure surface propagation that is weaker than the intact rock or the rest of the rock mass. This would form an immediate relief surface triggering collapse. ? the tensile capacity of the intact rock holding the sliding mass is exceeded, triggering collapse. ? The remaining surface with peak shear strength parameters gets too small and active forces of the sliding rock exceed the resisting forces and collapse occurs. ? Any combination of the above. All the progressive deformation behaviour monitored by the radars have shown an exponential increase in acceleration to collapse. Finally the rock mass constituting the instability collapses and in most cases disintegrates with the exception of the failure #7 which has a relatively flat failure surface and has demonstrated post collapse sliding controlled by blasting, mining and rainfall events. 7.11.4 Phase 5 : Post Collapse Rock Mass Behaviour Post collapse behaviour of non-failed rock mass is characterised by; ? An immediate localised elastic rebound of the rock mass which originally carried the collapsed material. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 7 : Case Study 3, Mt Keith Open Pit Page 7-13 ? Elastic relaxation (forwards movement) generated by the loss in lower structural support from the failed section of rock mass may initiate further propagation of failure in rock mass directly above ? In some instances the failure surface has continued to propagate further into the rock mass causing further deformation and collapse (see Figures A3.11 and A3.12). These are termed extended failures. 7.11.5 Further Discussion The radar monitoring data provides conclusive evidence of progressive damage evolution in fresh rock failures. These failures generally do not reflect a continuous period of specific macro event (mining) influenced deformation behaviour prior to the onset of progressive behaviour. It is also evident that although the evolving deformation behaviour is similar, the exact mechanism is unique to every failure. Based on the behaviour described above, a generalised time dependent deformation relationship showing progressive damage evolution of the Mt Keith failures is illustrated in Figure 7.5. Note that the different deformation pathways illustrated in Figure 7.5 from the OOF to collapse represent the different acceleration patterns of locations across the localised unstable rock mass, as a result of the developing failure surface. At collapse the pathways join together as the whole instability collapses simultaneously. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 8 : Case Study 4, Harmony Open Pit Page 8-1 8 CASE STUDY 4, HARMONY OPEN PIT, LEINSTER NICKEL OPERATIONS, WESTERN AUSTRALIA 8.1 Introduction Leinster Nickel Operation (LNO) is situated 80km south of Mt Keith on the eastern margin of the Agnew-Wiluna greenstone belt which is related to the Perserverence Ultramafic Complex. It therefore has a similar ultramafic and felsic geological terrain to Mt Keith, however the structures differ and consequently the wall stability and type of failures tend to differ. Besides different structure orientations the joint and shear zone infilling is less. The Leinster operation has three open pits namely Harmony pit, Rocky?s Reward and 11 Mile Well. The larger and more significant wall instabilities have occurred in the Harmony pit whereas the other pits are limited to relatively small circular failures in the deeply weathered saprolite. The pit closed on 15th August 2005 due to a developing overall east wall failure. By that time the mine had achieved almost full ore recovery, with the exception of approximately 2.5m of the final bench. This case study examines the displacement behaviour of 4 large multi-bench scale failures, namely two structure controlled weathered rock (saprolite) failures, a fresh rock wedge failure and a near full height wall failure through both weathered and fresh rock. 8.2 Regional Geology The regional geology and structures are similar to those at Mt Keith as shown in Figure 7.1 in Section 7. LNO lies within the Yilgarn Shield of Western Au stralia, which consists mainly of granite and granite- gneiss cut by narrow north-north-west trending belts of Archean age metasedimentary and metavolcanic rocks, generally referred to as "greenst one belts'. The Leinster nickel deposits are within the Agnew-Wiluna greenstone belt, near the eastern margin, all related to the volumetrically large Perseverance Ultramafic Complex which thickens dr amatically at Leinster. The Perseverance deposit is the largest individual nickel sulfide deposit associated with komatiite flows (ultramafic lavas) in the world, and is hosted by an unusually thick individual flow sequence, interpreted as having been formed by a large ultramafic lava river which thermally eroded through the felsic footwall stratigraphy and deposited sulfide mineralisation at its base (Western Mining Corporation Resources Limited, 1999). Mineralogy and geometry of the sulfide deposits display both primary (magmatic) and secondary (structural) genetic characteristics. The structure of the region is dominated by north-north-west trending strike faults, with massive sulfide mineralisation at Leinster (1A, F2 Shoots), Rocky's Reward and Harmony postulated to have been re-mobilised to at least some extent. The most prominent structure is the Perseverance Fault, which, together with associated faults, locally forms the eastern boundary of the greenstone belt as shown in Figure 8.1. 8.3 Lithology and Weathering The stratigraphy at Harmony strikes 340? true (appro ximately LNO grid north-south, see Figure 8.2) and dips steeply to the west. The rocks hosting the ultramafic are massive, homogeneous metasediments and felsic to intermediate volcanics and volcaniclastics. Thin, discontinuous west- dipping mylonite zones have been identified in both the footwall and hangingwall, predominantly PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 8 : Case Study 4, Harmony Open Pit Page 8-2 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 8 : Case Study 4, Harmony Open Pit Page 8-3 immediately below, and parallel to, the eastern contact of the mineralised ultramafic unit. These mylonites are commonly sulfidic (pyrite-pyrrhotite), and are particularly well developed to the north, frequently logged as "black shales". See Figures 8.2 and 8.3 for typical cross-section and plan views. A discrete, folded, crystic volcaniclastic unit has been identified to the west, protruding into the proposed pit. Several relatively unmineralised ultramafic units lie to the east, largely outside the pit. The ultramafic hosting the mineralisation is 20-60m thick. Most of the primary ultramafic textures were destroyed during several metamorphic and alteration events. The ultramafic rocks are divided into two types, based on the stratigraphic position, mineral assemblages and MgO content. The dominant ultramafic rock type is high in MgO and occurs on the western ultramafic/sediment contact. It is a thick unit (20-25m), with a dominant mineralogy of talc, magnesite and chlorite and some magnetite. Chlorite is predominantly associated with structures and mineralisation. Serpentinisation increases at depth, imparting greater amounts of antigorite, lizardite, and calcic amphibole. Carbonate alteration is pervasive throughout the ultramafic and is particularly intense adjacent to shear zones. The area is deeply weathered, with intense, pervasive oxidation extending to between 80 and 105 metres below the surface. Up to nine metres (typic ally two metres) of transported soil, colluvium and conglomerate covers most of the proposed pit area, with around ten metres of iron and silica- cemented hardpan immediately below. The regolith may be divided into eight zones. At the top, below the topsoil, is a transported conglomerate unit 2-3m thick, underlain by a variably silicified mottled clay/duricrust of approximately 10m thickness. Beneath these units, from top to bottom, are an in situ silicified clay zone, a pallid plasmic clay zone (only present in the felsic lithologies), a clay saprolite zone, a ferruginous clay saprolite zone, followed by saprolite and saprock zones. The base of weathering is deepest in the south, extending to 120 metres below the surface, while in the north it is shallower, extending to 80m depth base (Western Mining Corporation Resources Limited, 1999). 8.4 Ground water The groundwater regime surrounding the open pit is strongly influenced by the anisotropic structure within the rockmass. The phreatic surface within the eastern slope is strongly drawn down whereas in contrast the phreatic surface within the west slope remains high and steep. This suggests that only the west dipping subverical J1 foliations are the dominat ing influence on the flow regime, while the other east and west dipping structures, especially in the west slope do not significantly influence ground water flow (Goodchild, 2006). 8.5 Representative Sections A typical section through the pit is shown in Figure 8.3. 8.6 Structural Geology The major geological structures are summarised in Table A4.1 and illustrated in the steronet in Figure 8.4 below. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 8 : Case Study 4, Harmony Open Pit Page 8-4 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 8 : Case Study 4, Harmony Open Pit Page 8-5 8.7 Rock Classification and Strengths Geotechnical logging of cores from the felsic unit indicate RMR values ranging from 8.75 to 29.21 and RQD values ranging from 6 to 12 (SRK Consulting, 2002). 8.8 Pit Configuration Parameters Pit configuration parameters are listed in Table 8.1 for the pit as at abandonment in August 2005. Table 8.1 : Overall Pit Configuration Parameters at Abandonment (2006) West Slope East Slope Weathered Fresh Weathered Fresh Height (m) Overall slope angle (?) (OSA) 21 53-55 33 -35 53-55 Bench stack angle (?) (BSA) - - - - Bench face angle (?) (BFA) - - - - Inter ramp angle (?) (IRA) <23 60-70 38-39 60-70 Spill/catch berm width (m) 5 6 5 6 Spill/catch berm height (m) 10 20 (double) 10 20 (double) Bearing (?) (pit cords) 90 varies 270 varies Pit coordinate system orientation is 19? 35? west of true north, 20? 50? west of magnetic north 8.9 Displacement Monitoring Systems Prism monitoring of both the east and west wall using Autoslope and Leica ?jiggers? (automated total stations). 2 radars monitored the slopes since 2002. The radars were continually repositioned according to mining locations and requirements (Cahill, 2006). 8.10 Displacement Behaviour Non-failed Pit Wall Sectors The east wall and sections of the west wall have failed, however the south and central section of the west wall has remained relatively stable. Prisms on the south central section of the west wall (Zone 5) have decelerated towards the end of the pit life and remain in only a slight creeping mode as shown in Figure A4.1. The re-establishment of a ramp following the Komatsu failure (described below) resulted in a ?nose? being left in the mid section of the west wall (Zone 6) which although unfailed remains in a creeping behaviour mode as shown in Figure A4.2. The whole of the west wall is underlain by a zone of east dipping J3 structures. Mining between the 370 and 365RL bench during January 2004 resulted in daylighting of the J3 structures and consequently significantly accelerated movements. This resulted in cracks on higher berms and behind the crest immediately above. The J3 structures ther efore appear to control the deformation in this sector (Western Mining Corporation Resources Limited, 2002). Continued relaxation of the rock mass after closure in this area has maintained relatively high post closure creep rates. Less accurate elevational data in the prism monitoring databases make further deformation analysis unproductive. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 8 : Case Study 4, Harmony Open Pit Page 8-6 8.11 History of Failure/s The chronology of pit wall instabilities was summarised from the Harmony Open Pit Review Study (Western Mining Corporation Resources Limited, 2002). The first significant in-pit failure was structural and occurred in November 2000 on the east wall of Stage 1B. This was safely cleaned ?back to hard? and had little impact on production. No further movement of this area was recorded. The next failure experienced in-pit occurred in February 2001 on the west wall of Stage 1A, adjacent to the main haulage ramp, which was in the vicinity of the later Stage 1C failure. This failure was the first significant movement along a J3 structure and si gnificantly impacted production. Due to the relationship between the location of this failure and the Stage 1B pit, a re-design was necessitated which deferred the remediation of the slip to a later date. In June 2001, signs of an imminent failure of the Stage 1C west wall became evident. Again, failure was induced by an ubiquitous J3 structure. The devel opment of this failure was closely monitored over several months and large-scale ground movement occurred in October 2001. This slip effectively quarantined the excavation of the Stage 1C ramp below the 430mRL level and deferred the delivery of some Stage 1C ore. Remedial works were carried out from October 2001, with the excavation of the Stage 1C ramp only recommencing in March 2002. In September 2001, signs of imminent structural failure became evident on the Stage 2 east wall; both at the surface and on mid-wall batters. This failure was a result of weak rock mass conditions imposed on intersecting J2, J4 and J5 structures. Major ground movement occurred in October 2001 on the 480-460mRL batter. By the end of December 2001, the failure had propagated from the surface to the 440mRL level. As a result of this failure, the lower levels of Stage 2 were re-designed, again deferring the presentation of some ore to later mining phases. Consultants were requested to conduct stability reviews and they recommended that both east and west wall angles be significantly reduced in the weathered rock mass (from surface to approximately 420mRL in Stages 2 and 3). Additionally, it was recommended that depressurisation holes be installed on the west wall. Excavation to the limit of the ultimate Stage 3 crest was delayed as long as possible whilst further geomechanical recommendations were awaited from consultants. However, when further delay was impractical, this limit was dug during January 2002. Remedial works on the October 2001 Stage 1C failure were completed in late March 2002. Horizontal depressurisation / dewatering drilling was commenced at the base of the pit in mid April. However drilling was prevented from continuing on 24th April due to movement in the west wall between 1A to 1C. At this time, a cutback commenced on the west wall of 1A / 1C to stabilise this section of wall. Blasting of caprock occurred on 29th April 2002. Excavation continued until mid May 2002, when a major failure occurred in the west wall of Stage 1C. Remedial action taken post mid May 2002 in the failure zone included; ? Side-casting of surface material into the pit to buttress the toe of the failure. ? Establishment of a new west wall crest and trimming the wall back as mining progressed. ? Excavation of failed rock by truck once sufficient buttressing and width had been established. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 8 : Case Study 4, Harmony Open Pit Page 8-7 After approximately six weeks of cleaning up the failure, a further structural failure occurred in the west wall of Stage 1C behind the area being worked. This movement was triggered by excavation at the toe of the failure zone, removing the stabilising buttress. This failure created a section of wall in the west wall of Stage 1C that had a vertical face to surface of approximately 30m height. This face posed a danger to future remedial work due to the potential for toppling failure. Remedial work was recommenced only after safety was assured and there were no further failures. The Komatsu Ramp failure occurred during the excavation of phase 3 pit after the walls of the initial phase 3 pit had already been flattened following consultant?s recommendations. The failure necessitated redesign of the ramp and additional flattening. 8.11.1 West Wall This section deals in more detail with the two relatively large failures which occurred on the west wall, namely the 1C failure and Komatsu Ramp failure. A few relatively small batter scale hard (fresh) rock failures also occurred which were identified before failure, but were not large enough or sufficiently closely monitored to warrant further discussion. The mechanisms of both the two large failures involved the same structures so they are both discussed together in further detail below. As discussed above the 1C failure was reactivated several times due to on going mining and cleanup operations. The monitoring of two of these reactivation events is also discussed. When The initial Stage 1C failure occurred between the 29th September 2001 and 21st October 2001. Photographs of the failure area are shown in Figures A4.3 to A4.4 (Cahill, 2006). Cleanup efforts induced further failures in December 2001, May 2002 and July 2002 (see radar data below) and the failure remained active through to August 2004 when further rock falls from the failure, as a result of continued creeping movement, necessitated the temporary closure of northern sections of the pit (Colley, 2004). Other significant movements occurred on the 20th April 2003 and the 18th and 22nd June 2003 which were triggered by mining activities. The Komatsu Ramp failure occurred on the 8th October 2002 during the excavation of the later Phase 3 pit (Cahill and Lee, 2006). Locations The Stage 1C failure was situated in the northwest wall and the Komatsu failure occurred in the southwest wall as illustrated in Figure A4.3 and A4.4. Both failures occurred in the relatively weak weathered felsic (saprolite) rock. Structure, Mechanism and Trigger The principal failure mechanism for both the failures was an active ?passive sliding block created by joint orientations. The steeply dipping foliations also facilitated secondary toppling type behaviour along the foliation planes. The action of the mechanism resulted in a passive wedge being pushed out of the base of the slope, which manifested itself in some places as floor heave (Goodchild, 2006). The contributing structures were; J3: which acted as the principal east dipping sliding plane for the active wedge J4 and J6: which acted as secondary sliding planes for the passive wedges PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 8 : Case Study 4, Harmony Open Pit Page 8-8 J1 foliations: acted as perpendicular release su rfaces (tension cracks) and facilitated secondary toppling type behaviour. J2: subvertical east-west cross cutting joints which acted as lateral release surfaces for both failures. J7: subhorizontal structures which also possibly acted as secondary sliding planes for the passive wedges The mechanism is illustrated in Figure 8.5. Failure 1C was a deep seated failure in comparison to the Komatsu failure (Goodchild, 2006). It was established during mining in the fresh rock that joint set J3 formed an arc shape on the western wall, dipping approx 40/080 in the south west and 40/100 in the north west, as illustrated in Figure A4.4. No further failures of this mechanism were experienced in the fresh rock as the J3 joint set was too rough to fac ilitate failure with its orientation in fresh rock (Goodchild, 2006). Photographs of the development of the failure are shown in Figure A4.5. The trigger appears to be ongoing mining and failure cleanup activities. Displacement Behaviour of Slopes Leading up to Failure October 2001 1C Failure Prism data of this failure are shown in Figure A4.6 December 2001, May 2002 and July 2002 Reactivation of 1C Failure Prism data of these failures are shown in Figure A4.7 20th April 2003 Reactivation of 1C Failure Prism data of these failures are shown in Figure A4.8 18th and 22nd June 2003 Reactivation of 1C Failure Radar monitoring data for these failures are shown in Figures A4.9, A4.10, A4.11 and A4.12. February 2003 Komatsu Failure Prism data of this failure is shown in Figure A4.13 Summary and Interpretation of Deformation Behaviour The time dependent deformation behaviour of the West Wall 1C and Komatsu failures was characterised by; ? A rapid acceleration in the initial failures, with very little characteristic increased movements prior to failure (Figure A4.6 and A4.13). 29 days in the case of the 1C failure and 9 days for the Komatsu failure. ? Extensions or reactivation of the failures displayed extreme sensitivity once again, accelerating to failure with relatively little warning. ? Radar wall displacement images of the reactivation of the 1C failure showed both characteristic propagation of failure from a ?hot spot? across the final failure mass (Figure A4.10), as well as a ?mass? type failure (Figure A4.10) where no apparent propagation takes place. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 8 : Case Study 4, Harmony Open Pit Page 8-9 ? Initiation of failure in the 1C Failure was from the toe upwards towards the crest over a time period of approximately 1 month (Figure A4.5). This is likely to be a result of destabilisation of the overall slope caused by the outward movement of the passive wedge at the base of the failure. The characteristic patterns of deformation and sensitivity to mining activities of the slopes confirm both overall structure control of the weathered rock mass (saprolite) and very low residual shear strengths of the structures following initiation of instability. 8.11.2 East Wall The stability of the north west wall has generally been poor. The failures on this wall have been small to medium in size and occur on irregular and wavy foliation with break out and undercutting occurring on J6 structures. Instabilities are often not always id entified by prisms, as prism coverage is poor due to difficult access onto the benches and the lower section of wall, which is prone to failure, is not covered by prisms and is not visible from the surface theodolite position. In the fresh rock (ultramafic and felsic), almost all batters have either been lost as a result of steeply west dipping J1 (foliation) structures and possibly some influence from flatter J4 structures, or are stacked with failed material. Two significant failures, which have been well monitored, are discussed in further detail. These two are the localised north-east failure and the overall east wall failure. Northeast Failure When The northeast rock slide occurred on the 16th November 2003. Location The failure occurred in the transitional to fresh ultramafic rock located behind the working bench in the north eastern corner of the pit. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 8 : Case Study 4, Harmony Open Pit Page 8-10 Structure, Mechanism and Trigger The failure mass was approximately 40m high. The mechanism appears to be a vertical wedge failure bounded principally by J1 and J2 structures. The tr igger was a rainfall event of 6.5mm on 15/11/2003 followed by a more significant 19mm on 16/11/2003. The failure involved approximately 2000 tons of material. Displacement Behaviour of Slopes Leading up to Failure There were no direct prism movements recorded for this failure, however prism data for Zone 1 prisms (in the north east area) are shown in Figure A4.14. Radar monitoring data of this failure are shown in Figures A4.15 and A4.16 Summary and Interpretation of Deformation Behaviour The time dependent deformation behaviour of the North West fresh rock failure was characterised by; ? General insensitivity of the overall rock mass to rainfall, however this small localised failure appears to have occurred directly after rainfall, suggesting sensitivity of cracked berms to filling with rainfall water. ? Extremely rapid acceleration to failure in the order of approximately 4 to 5 hours. ? The sequence of wall displacement images suggests that the failure propagated upwards from the northern ledge of the previous failure in a northwards direction, most likely along J1 foliations. ? The radar displacement graphs confirm that the collapse occurred in at least two distinct events as shown in the combined graph sub figure in Figure A4.16. This deformation behaviour is considered typical of relatively small localised structure controlled rock slides (wedges) in fresh rock. East Wall Failure The east wall failure is discussed in the context of the overall slope. The overall instability mechanism has over time, generated smaller scale failures, including two east wall ramp failures before and after closure. When During mining of the south east sector of the pit, the east wall has continued to experienced relatively large rates of deformation. In August 2005 a relatively small failure occurred in the south east corner of the pit which led to the pit abandonment as it was realised that a much larger overall failure was developing on the east slope (Cahill and Lee, 2006). Progressive behaviour continued up until collapse initiation in February 2006. A partial collapse occurred in March 2006, approximately 7 months after abandonment. The failure mass re-stabilised after total 2D vector displacements of 3 to 7m. The failed section of the pit is however continuing to show creeping movement although at much reduced rates. This has resulted in ongoing cracking and sloughing. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 8 : Case Study 4, Harmony Open Pit Page 8-11 Location Laterally, the physical failure itself extends from the south-eastern corner to the eastern ?nose?, although cracking along the pit crest suggests that the instability now extends almost the entire length of the east wall of the pit. Vertically, the east wall failure extends from the crest through the fresh rock and daylights approximately 20 to 30m above the final pit floor level (which was underwater at the time the pit was inspected). The extent of the failure is shown in Figure A4.17. Structure, Mechanism and Trigger The primary mechanism of the east wall failure and the east wall ramp failure is clearly that of a planar mode facilitated by the following controlling structures; J4: acts as the principal west dipping sliding plane for the overall failure J1 foliations: acted as sub-vertical release surfaces. J2: sub-vertical east-west cross cutting joints, which continue to act as lateral release surfaces both in the north and south of the failure. J7: subhorizontal structures which are likely to have only a minor influence with regards secondary sliding planes. J5 structures in the wall are also unfavourably orientated and form wedges in combination with J2 (Goodchild, 2006). There are numerous examples of these wedges on this slope. Displacement Behaviour of Slopes Leading up to Failure The time dependent deformation behaviour of the east wall instability has been continuously monitored by a network of prisms up until present time. Prism data of this failure are shown in Figure A4.18 and A4.19. Resulting vector deformation movements are illustrated in Figure A4.20 Radar monitoring data for a localised failure are shown in Figures A4.21 and A4.21 Note mesh on lower slope under ramp ? interfering with radar images. Summary and Interpretation of Deformation Behaviour The east wall failure is clearly structurally controlled. The resultant vector directions are parallel to the line of intersection of the J2 and J4 structures to within 3? (Goodchild, 2006). The classic sliding mechanism is illustrated in Figure 8.6. It was noted by geotechnical mining staff that the overall failure movement itself appears sensitive to mining, but insensitive to rainfall and groundwater changes (Cahill, 2006). The east wall failure is significantly different from the other Harmony failures in that, although it was structurally controlled and displayed similar sensitivity, the failure re-stabilised after partial collapse (ie it returned to regressive behaviour), despite the failed rock mass not encountering any passive resistance. This can only be attributed to dissipation in pore pressures as a result of the large internal displacements which occurred during the partial collapse. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 8 : Case Study 4, Harmony Open Pit Page 8-12 8.12 Discussion and Conclusions All failures are strongly structurally controlled, even in the highly weathered saprolite. In general the sensitivity of the rock mass and the rapid acceleration to failure, as displayed in the time dependent deformation figures can be attributed to adverse structural orientations in both the east and west wall, and to low residual shear strengths of infilling in both weathered and fresh discontinuities. Failure triggers in the Harmony pit are principally rate and depth of mining related for the larger failures and blasting and rainfall related for the smaller more localised hard rock failures. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 9, Case Study 5, Venetia Open Pit Page 9-1 9 DETAILED CASE STUDY 5, VENETIA, SOUTH AFRICA 9.1 Introduction Venetia Diamond mine lies within the Central Zone of the Limpopo Belt which is situated along the northern border of South Africa. The pit (Appendix 5, Figure A5.1) is situated in the core of a northward converging shallowly eastward plunging synclinal fold. The fold combines 23 different types of country rocks forming Gneissic and Metasedimentary packages as well as crosscutting lithologies and features. There are at least 14 distinctly separate kimberlite bodies in the Venetia cluster with only the two largest K1 and K2 being actively mined. Of geological interest are the unusual irregular shapes of the kimberlite pipes which are almost certainly a result of the highly complex nature of the country rock into which the pipes have intruded. The mine has experienced one large overall slope failure, several other multi-bench failures and almost continuously occurring bench scale planar failures, especially on the south wall. 9.2 Regional Geology The Limpopo terrain is one of the most intensely deformed geological domains in South Africa. It is a belt of high grade metamorphic rocks comprising a Central Zone with northerly striking fold axes, flanked to the north and the south by marginal zones consisting of sheared, reworked granite- greenstone assemblages. The oldest rocks in the Central Zone are considered to be supracrustal gneisses with an estimated age of approximately 3.3 billion years (Cairncross and Dixon, 1995). The Limpopo Belt is considered to have been created by the compressive interaction between the Zimbabwe and Kaapvaal Cratons during the Achaean and Protereozoic. The rocks in the Central Zone are characterised by high grade metamorphism and tectonism (Mulville, 2006) Where the Venetia mine is situated the country rocks are considered to have undergone numerous phases of shearing and folding. Mapping undertaken by the mine has shown that the open pit is situated in the core of a northward verging shallowly eastward plunging synclinal fold (D1). As a consequence of this fold, three primary structural domains have been delineated. These are the northward dipping southern limb of the fold, the axis of the fold and the steeply northward dipping northern limb of the fold. Interpretations from more recent pit mapping and aerial photographs have shown that the overall fold shape is controlled by the interaction between at least three fold axes of the same trend stepping towards the west and the north. In the vicinity of the open pit this synclinorium plunges at about 31? towards the east- northeast (Barnett, 2003). On the southern limb of the fold, bench-scale interference folds are particularly common. These folds vary extensively in trend from north-south through north-east to east-west. The cross-folds become tighter towards the western hinge zone. They vary in character from open northward plunging (parallel to the northward dip of the southern limb) folds to tight northeast plunging folds (up to 40? from the horizontal). The northward plunging folds are likely to have formed from east-west compression (D2) and the northeast plunging folds as coeval fault -parallel folding during D2 transpressional dextral strike-slip motion on northeast striking faults (Barnett, 2003). PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 9, Case Study 5, Venetia Open Pit Page 9-2 9.3 Lithology and Weathering Within the country rock, there are two packages of rocks. The first package referred to as the "Gneiss Package" is made up predominantly of biotite gnei ss, biotite schist, quartzo-feldspathic gneiss and amphibolitic gneiss (Figure 9.1). These rocks are strongly sheared and deformed, with clear mineral lineations and definite schistose and gneissic fabrics, and are located in the core of the fold through which most of the kimberlites have intruded. The second package is known locally as the "Metasedimentary Package" and comprises limestone and marble interbedded with metapelite and calcsilicate. Fuchsitic quartzite is also common, particularly adjacent to the tectonic contact between the two packages (Barnett, 2003). The lithologies of the "Metasedimentary Package" ha ve less strongly developed mineral lineations and have schistose fabrics only on definite shear zones. Layering in the "Gneiss Package" is only para- conformable to the layering in the "Metasedim entary Package" and appears not to be persistent along strike. From these relationships Barnett (2003) suggests that the "Metasedimentary Package" was thrust onto/into the "Gneiss Package" package pr ior to the main east-west folding event. Pegmatite dykes and a 20 to 60m thick sill of dolerite cross-cut all the rock types except the kimberlites, the dolerite dyke being situated at approximately 250m below surface. There are 14 individual kimberlite pipes in the Ventia cluster with only the two largest, K1 and K2 significantly contributing to financial returns. Basson (2005) developed an updated country rock model as shown in Appendix 5, Figure A5.2, and divided the country rock geology into five main zones summarised below; Zone 1: Eastern fold zone, eastern sections of K1and incorporating K4, K5, K6, K16 and K17 Zone 2: Western high strain zone incorporating the geology around K2 and K8 Zone 3: Prolate zone geology to the south of K7 and K12 Zone 4: Central low strain zone geology around the western portion of K1 Zone 5: Late kinematic shear zone bounded by the Lezel and Tina faults. This "Gneiss Package" is weathered to a depth of 20 to 30m below surface and the weathering profile of the "Metasedimentary Package" to as deep as 60m (Barnett, 2003). The weathering profiles of all the kimberlite pipes are considered to be very deep, however there was no specific data thereon (Hannweg, 2006). 9.4 Structural Geology Faults and Shears Cross-cutting the D1 folding are north-east striking faults that have dextral strike-slip shear senses. One of these faults called the Lezel fault has a displacement varying along its length from 150m to 200m. These faults are interpreted by Barnett (2003) as forming transpressional, dextral strike-slip faults with east-west compression that have been reactivated on at least three occasions. There are other minor faults varying in strike from east-west to northwest. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 9, Case Study 5, Venetia Open Pit Page 9-3 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 9, Case Study 5, Venetia Open Pit Page 9-4 The Gloudina Fault follows near to the axis of the fold and strikes approximately east-west. Barnett?s (2003) interpretation of the origin of these faults is again a dextral transpressional orogeny, but with north-northwest compression. Equal angle steronet plots of the poles to shear zones as mapped by Basson (2005) are shown in Figure A5.3. Jointing Both the faults and joints have been reactivated by a number of phases of deformation which has made it difficult for geologists to establish a reliable genetic mode. The orientations of some of the joint sets do imply that they have been formed and reactivated at the same time as the faults during the latter brittle phase of the Limpopo Belt's development. None of the sets are axial planar to any folding. Even though many of the joints show signs of shearing, they most likely originated as joints. Joint set J0 has been designated as layering parallel and is the most penetrative and well developed joint set, and is the controlling factor in almost all cases of instability (Barnett, 2003). The north east striking joints (J2) are the most well developed besides the J0?s, followed by the north to north-northeast striking joint set (J1). Besides the layering-parallel joint set (J0), there ar e no characteristic joint surface features that enable the sets to be distinguished from each other Barnett (2003). The joint sets as presented by Barnett (2003) are summarised in Appendix 5, Table A5.1. Equal angle steronet plots of the poles to gneissic and schistocity banding (J0) as mapped by Basson (2005) are shown in Appendix 5, Figure A5.4. 9.5 Groundwater The groundwater regime in the country rocks surrounding the open pit has to date not yet been defined and is currently the focus of an ongoing investigation (Hannweg, 2006) 9.6 Representative Sections A representative section through the pit is shown in Figure 9.2. 9.7 Rock Classification and Strengths Average Intact Rock Properties derived from laboratory tests are summarised in Appendix 5, Table A5.2, and average rock mass ratings are summarised in Appendix 5, Table A5.3. 9.8 Pit Configuration Parameters Pit configuration parameters are listed in Table 9.1 for the pit as at September 2006. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 9, Case Study 5, Venetia Open Pit Page 9-5 Table 9.1 : Venetia Overall Pit Configuration Parameters, October 2006 South Slope North Slope Overall height (m) 192m 128m Overall slope angle (?) (OSA) 36 ? 47? Bench stack angle (?) (BSA) 41 ?(upper), 40.8?(lower) 56? Bench face angle (?) (BFA) 37 ? 55? Inter ramp angle (?) (IRA) 39 ?(upper), 37?(lower) 56? Spill/catch berm width (m) varies varies Spill/catch berm height (m) 12m 12m Bearing (?) (pit cords) n/a n/a Pit coordinate system is Lo 29/Clake 1880 System PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 9, Case Study 5, Venetia Open Pit Page 9-6 9.9 Displacement Monitoring Systems Prism monitoring at Venetia use Gemos software and Leica TCA2003 automated total stations. 1 radar system has been installed in the pit since 2005. 9.10 Displacement Behaviour Non-failed Pit Wall Sectors The overall pit wall deformation behaviour was reviewed using the prisms located along the crest of the pit walls. Figure A5.7 (Appendix 5) illustrates the behaviour of a selection of prisms located around the perimeter of the south wall. As can be seen the wall is behaving regressively, reflecting a steady state creep conditions in between intermittent deformational responses to mining activities. 9.11 History of Failures South Domains The south wall of the pit has been characterised by localised bench scale planar instability resulting from metamorphic fabric parallel layering or gneissic banding and schistocity commonly referred to as J0 jointing (Barnett,2003). These structures di p northwards at an intermediate angle of 49? in Domain 4 and 38? in Domain 5 which are effectively approximately 2? steeper than the average inter-ramp wall angles for Domain 4 and 5. This localised bench scale instability on the south wall is shown in Appendix 5, Figure A5.7. North Domains Toppling is the dominant mode of bench scale failure on the north wall of the pit as a result as result of the slightly overturned northern limb of the fold made up of biotite gneiss and schist which has resulted in a steep sub-vertical (average 82?) north dipping J0 structures. Bench scale wedge failures based on other joint sets also occur from time to time. A few notable multi-bench scale failures have occurred in the pit. These include the; ? March 2000 North slope failure (triggered by 162mm rainfall) ? July 2003 South-east slope failure (trigg ered by mining at the toe on bench 6) ? September 2004 South-east slope extension failure (trigger unknown) ? May 2005 South-west face ramp failure (triggered by blasting) These are discussed in further detail below. 9.11.1 South-east Slope Failure When Original collapse occurred on the 2nd July 2003 An extension to this failure occurred on the 11th September 2004. Location South-east Wall Cut 3 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 9, Case Study 5, Venetia Open Pit Page 9-7 Structure, Mechanism and Trigger Figure A5. 8 illustrates the pit intersections with the Lezel and Tina faults (shown as bright red and blue lines to the west of the failure). Two other minor faults or shear zones appear to bound the extent of instability observed (the approximate locations of these faults are shown in yellow). It would appear that the layering has flattened out locally under the influence of this faulting. The failure occurred on layering which daylights into the pit in Bench 6. Based on the country rock geological model of the mine, the failed slope is located in Biotite Schist, which was confirmed by field observation. A large Amphibolite inclusion was also observed in the eastern most fault / shear zone (Keyter, 2003). The mining of Bench 6 immediately below this area, together with a high phreatic surface as a result of rainfall, possibly triggered the instability (Keyter, 2003). Displacement Behaviour of Slopes Leading up to Failure Cracking was first observed at the eastern end of the failure area on the 9th June 2003. At the time, the cracking was again attributed to near surface wedge instability as a result of jointing striking diagonally across the bench face in this area. Extensive cracking and subsidence was first observed on the pit rim of the K1 south face on the morning of the 2 July 2003. Movement had been observed on Prisms M11 and M12 since mid April (Figure A5.9). These movements were initially attributed to near surface instability associated with slabbing of metamorphic layering in the immediate area of the prisms. The zone of the instability extended for a distance of about 175m along the south rim of the pit, with a width of 30 to 40m behind the crest in the centre of the area of subsidence, and included a total of 6 benches below (Keyter, 2003). An additional five prisms were installed in the area affected. Approximately 1 hour prior to the failure the magnitude of vertical subsidence observed was estimated to be approximately 300mm, and horizontal crack widths estimated to range between 50 and 150mm. At the base of the failure bulging along the top of Bench 6 was noticed accompanied by continuous ravelling and audible fracturing of the rock. Following collapse of the slope more cracking was observed towards the back of the failure (Keyter, 2003). 11th September 2004 Southeast Slope Failure Extension Movement of the south pit rim of K1 directly to the east of the July 2003 South-east failure (Figure A5.12a and b) was first identified at the beginning of June 2004. The area affected extended from about 25m from the July 2003 south-east failure towa rd the east along the southern rim of the pit, with a width of 5 to 7m behind the crest and a total of 6 benches affected below. It appears that the failure plane is along a Biotite Shist shear zone. The failed slope is in an Amphibolite bouden structure, with Biotite Shist on the boundaries of the failure. The failure appears to be bound by a shear zone to the east and is an extension of the July 2003 failure to the west. The failure volume was estimated to be approximately 183 000 tons (Naidoo, 2004) Crack widening increased gradually from 1mm/day horizontally from June to 21mm/day horizontally on the 10th of September. Vertical movement accelerated from 1mm/day in June to 245mm/day on September 10th. The graphs of rate of movement are shown in Figures A5.12c and A5.12e An inspection on the 11th of September revealed that the failure had subsided approximately 1m (Figure A5.10). Later that day at about 1:30pm, a blast taken below the southern ramp triggered further subsidence of the failure to approximately 3.45m. In the following two weeks subsidence of the PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 9, Case Study 5, Venetia Open Pit Page 9-8 failure continued to occur, eventually dropping to over 6m from the crest (Figure A5.11) (Naidoo, 2004). Summary and Interpretation of Deformation Behaviour The South slope failures are clearly sliding/planar mode failures with the failure surface controlled by the locally flattened Biotite Schist layering. The Ju ly 2003 failure was triggered by the mining of bench 6 which is situated at the toe, together with a high phreatic surface. The deformation behaviour characteristics of the two failures are different in that the July 2003 failure was event driven by mining at the toe, whereas the September 2004 failure behaved as a localised intact rock failure in which failure results by a process of progressive damage evolution rather than being event driven. 9.11.2 March 2000 North Slope Failure When 19th March 2000 Location North wall Cut 2 (shown in Appendix 5, Figure A5.13). Structure, Mechanism and Trigger This was an 96 high multibench scale wedge failure with an estimated failure mass of 96 000tons. The wedge failure mechanism was effectively formed by the intersection of the following discontinuities; ? Sub-vertical J0 structur e with an orientation of 48?/128? ( dip/dipdir) steepening up to 80?/128 ? ( dip/dipdir) on the upper reaches of the failure. ? The eastern lateral release surface was a shear splay (fault) from the larger east-west striking Gloudina Fault with an orientation of 48?/217? ( dip/dipdir) The trigger mechanism was a combination of high rainfall (160mm over 2 days together with in excess of 600mm in previous two months) and poor limit blasting. Displacement Behaviour of Slopes Leading up to Failure Not monitored Summary and Interpretation of Deformation Behaviour No comment possible 9.12 Discussion The range of failures experienced at Venetia extends from regularly occurring bench scale planar failures through to large multi-bench scale failures. All failures are controlled by combinations of the larger structures and gneissic and schistostic banding occurring within the complex geological terrain. The deformation behaviour of failures in the Venetia open pit illustrate both direct event controlled failure deformation, as was shown in the SE July 2003 failure, as well as rapidly developing failures with minimal prior direct mining event driven influence, as illustrated in the September 2004 extension to the SE failure. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 10: Case Study 6, Orapa and Letlhakane Open Pits Page 10-1 10 CASE STUDY 6, ORAPA AND LETLHAKANE OPEN PITS, BOTSWANA 10.1 Introduction Orapa and Letlhakane open pit diamond mines are located in central Botswana, approximately 220 km west of Francistown and 500km north of Gaborone and are situated within the Central Kalahari sub- basin of the Karoo Supergroup. There are over 50 kimberlite volcanic pipes located in this area with the Orapa (designated A/K1) kimberlite pipe being the largest of the group with a surface area of 118 ha. The Letlhakane pipes (designated D/K1 and D/K2) are situated approximately 30km from Orapa and have surface areas of 13ha and 5ha. Although they constitute two separate and large open pits the geological country rock terrain is similar and therefore both pits are considered together as one case study. Both pits are considered to be uninfluenced by large scale structures. It should however be noted that the kimberlite rock types within the pipes vary considerably. This case study examines the failure mechanisms and deformation behaviour of 3 large multi-bench scale failures which occurred within the two pits. Photographs of the Orapa and Letlhakane open pits are shown in Appendix 6, Figures A6.1 and A6.2 respectively. 10.2 Regional Geology Orapa and Lethlakane open pits are situated within the Central Kalahari sub-basin of the Karoo Supergroup in Botswana as shown on Figure 10.1. The sub-basin is bounded in the east and northwest by outcrop or sub-outcrops of pre-Karoo formations. The southern limit is the limit of the Karoo strata south of the Zoetfontein Fault, and the northern limit is the crest of the basement high through the Makgadikgadi Pans. Southwestwards the Central Kalahari Sub-basin connects with the Gemsbok Sub-basin. The Central Kalahari Sub-basin is widely covered by Kalahari beds, except at its eastern margin, where most of the lithostratigraphy has been established. Despite the fact that the sub-basin is extensive and includes representatives of all the Karoo groups overlain by the Kalahari beds, the total thickness of sediments in the sub-basin is generally less than 500 m (Reeves, 1978). The Orapa and Letlhakane mines fall into a northern belt of generally shallower proximal Karoo facies, apparently reflecting the underlying cratonic basement (Smith, 1984). All lithostratigraphic units are all horizontal or have very shallow dips. The lithostratigraphy formations of the northern belt are illustrated in Figure 10.2 and summarised as follows: Pre-Karoo Basement The pre-Karoo basement below the Northern Belt essentially comprises Archean gneisses and schists forming part of the Rhodesian Craton (Reeves, 1978). The cratonic basement rises to form the crest of a ridge beneath the Makgadikgadi Pans (Reeves, 1978). Further south, where subsidence was greater, the Karoo succession has similarities with that in the southern and western parts of the sub- basin. Ecca Group Mea Arkose Formation (Unit 160) At Orapa the Mea Arkose locally rests on the pre-Karoo basement (Unit 170). The top of the Mea Arkose is less easily recognised but can be most usefully taken as the junction of the predominantly sandstone succession with an overlying banded grey and carbonaceous mudstone succession over 2m thick, which commonly includes coal seams. The Mea Arkose in the northern belt is commonly a PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 10: Case Study 6, Orapa and Letlhakane Open Pits Page 10-2 pale, coarse to medium grained, ill-sorted sandstone. The sandstone is commonly feld-spathic and some bands are pebbly or garnetiferous. Sandstone units fine up into thin grey siltstones and mudstones which appear to be impersistent, as are lenses of coal and carbonaceous shale up to 1 m thick (Smith, 1984) Tlapane Formation (Unit 150, Carbonaceous mudstones interbedded with sandstones) This formation is of a fairly uniform nature and is established over most of the Central Kalahari Sub- basin and is essentially one of carbonaceous mudstones and coal seams with subordinate sandstones and grey-brown mudstones (Smith, 1984). Beaufort Group Tlhabala Formation (Unit 140, Grey mudstones) The Tlhabala Formation is the term used for the non-carbonaceous division of the Tlapane Mudstone (Stansfield 1973). It can be further subdivided into the Masekangwe Member and the Kautse Member (Smith, 1984). Lebung Group (Unit 120 and 130, Red sandstones) The Lebung Group includes all the clastic 'red be d? formations deposited unconformably on the Tlhabala or older formations, and is in turn unconformably overlain by the Stormberg Lava Group. The Lebung Group can be divided into two separate formations. The lower one, the Mosolotsane Formation (Unit 130), comprises coarse to fine grained sandstones and mudstones interpreted as having been deposited sub-aqueously. The upper formation, the Ntane Sandstone (Unit 120), is a more uniform fine grained sandstone generally considered to involve aeolian deposition. Both these formations are sufficiently uniform in character to be recognised across the central Kalahari Sub-Basin (Smith, 1984). Stormberg Lava Group (Unit 110, Basalts) The Stormberg Lava Group outcrops in the northern belt in the vicinity of Orapa. The basalts have a variable thickness lying unconformably on the Ntane Sandstone Formation and locally, west of Orapa, on inliers of Mosolotsane Formation ('Toatshaa s andstone' and 'mudstone'). The inliers may have developed by pre-Stormberg Lava Group faulting followed by local erosion of the Ntane Sandstone prior to the volcanic eruptions. The Lava Group comprises a series of basalt flows, each up to 30m thick. In each lava flow, an upper and lower amygdaloidal zone and a massive central zone, can commonly be distinguished (Smith, 1984). 10.3 Kimberlite Lithology and Weathering Profile Orapa Pipes The Orapa pipe consists of two closely spaced kimberlite pipes, that merge into one body at approximately 250 m depth to form a single oval-shaped kimberlite outcrop at surface. The southern pipe truncates, and is therefore younger than the northern pipe. The kimberlite rocks are highly variable in character (Dirks, 2003). The majority of the southern pipe is a sediment filled crater with Basalt Breccia as well as epiclastic (talus slope, debris flow and crater lake), pyroclastic and volcaniclastic deposits that all belong to crater facies kimberlite. The dips of the bedding planes vary from 40 to 0 degrees, with the steepest beds occurring close to the margins of the kimberlite pipes The units originated as talus slope (dip > 20 degrees), de bris flow (dip < 20 degrees) or subhorizontal crater floor deposits (Dirks, 2003). PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 10: Case Study 6, Orapa and Letlhakane Open Pits Page 10-3 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 10: Case Study 6, Orapa and Letlhakane Open Pits Page 10-4 Diatreme facies kimberlite (STKB) has been drilled to 350 m in the southern pipe and (NTKB) to 450m depth in the northern pipe. The diatreme kimberlite but is currently not exposed in the Orapa pit (Dirks, 2003). Letlhakane Pipes The Letlhakane pipe does not have any sedimentary infill having most likely been removed by erosion and is therefore all diatreme kimberlite (Gibson, 1998). The weathering profiles of all the kimberlite pipes themselves are exceptionally deep with changes in the degree of alteration being recorded down to approximately 300m (Gibson, 1998). Weathering depth of the country rock (basalts) are generally considered to be approximately only 30m deep (Ramsden, 2006). 10.4 Structural Geology Orapa A variety of fracture/discontinuity sets can be identified in the open pits. Dirks (2003) reported that the nature and orientation of these fractures is strongly dependent on the host lithologies. Basson (2006) gives a structural interpretation based on a total magnetic image as shown in Figure 10.3. A summary of the principal fracture zones (i.e. those with strike lengths greater than 50 m) was derived from Dirks (2003) and Basson (2006) and is given below for the Basalts and the Kimberlites. The orientations of the large scale structures are summarised in Appendix 6, Table A6.1 and for the joint sets in Table A6.2. 10.4.1 Basalt Three discontinuity types identified in the basalt by Dirks (2003) were shear fractures or shear joints, planar joints and laminated carbonate veins.. These discontinuity types may occur in isolation, but commonly occur in combination with each other. At Letlhakane it was observed that shear zones grade along strike into zones containing planar joints, and both fracture types are intruded by laminated carbonate veins. At Orapa, outcrop in the basalt is too limited and mostly of too poor a quality (deeply weathered top benches offer the only outcrop in Basalt in most parts of the pit) to establish the same lateral relationships with certainty. However, at Orapa, like at Letlhakane, NW trending shear joints and related master joints and carbonate veins provide the most prominent larger-scale (i.e. strike lengths >200 m) discontinuities in the country rock (Dirks, 2003). 10.4.2 Kimberlite The Orapa kimberlite is dealt with in somewhat greater detail as it is where the two major failures have occurred. In the Kimberlite there are few continuous through-going structures and penetrative jointing is not common. In contrast to the coarse-grained breccias, finer grained pyroclastic, epiclastic and volcaniclastic kimberlite deposits in the pit are well-layered (Dirks, 2003). Bedding planes provide regional (>200 m) discontinuities in the ki mberlite and have contributed to large-scale failures in the pit. Planar joints occur within the kimberlite lithologies although systematically orientated sets appear rare. Of the joint sets only shear fractures have a regional expression, and have an important influence on pit PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 10: Case Study 6, Orapa and Letlhakane Open Pits Page 10-5 stability as they played a major role in the major failures that have occurred in the pit. Shear fractures can be sub-divided into two types depending on the host kimberlite lithology. These two types are: 1. Curviplanar, discrete slickensided surfaces developed in fine- to medium-grained, bedded volcaniclastic, epiclastic and pyroclastic kimberlite, 2. Curviplanar to irregular diffuse fracture zones in Basalt Breccia Kimberlite. The orientations of the shear fractures are highly variable depending on their locations in the pit relative to the kimberlite crater morphology (Dirks, 2003). Bedding Orientations Many of the kimberlite lithologies of the crater facies are well bedded. In general terms, bedding planes dip crater inward, i.e. towards the pit floor at angles between 0 and 45 degrees. However, significant variations occur. Basalt Breccias are mainly concentrated along the east and west margins of the pit. Bedding planes in the north pit consistently dip south wards; i.e. towards the centre of the south pit, rather than towards the centre of the north pit (see map). As a result, pit walls in the north pit generally transect the bedded deposits at near-right angles, favourable for the overall stability of these slopes. The northern side of the north pit is composed of the more massive and poorly bedded northern Pyroclastic Kimberlite, which appears to be dipping inward at low angles. Close to the neck area in the pit, bedding planes are horizontal along a roughly east-west axis (Dirks, 2003). PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 10: Case Study 6, Orapa and Letlhakane Open Pits Page 10-6 In the south pit, bedding plane orientations are more variable, and marginal units generally dip towards the centre of the pit. Directly south of the neck area, on either side of the pit (i.e. near grid blocks K23 and X23), well-bedded talus slope deposits occur. The bedding planes in these units define 200m scale inward curving open synformal structures (spoon shaped) with the core of the synform plunging 30-40 degrees towards the centre of the pit. Dips on bedding planes vary from 35- 40 degrees at the top of the pit near the crater rim, to 15-20 degrees near the toe of the talus slope deposits, bedding planes are very well developed in these deposits, which are overlain by Basalt Breccia across very discrete lithological surfaces (Dirks, 2003). Towards the south-west in the southpPit, bedding planes generally dip from 30-15 degrees towards the centre of the pit. Along the south-east side of the pit, dip angles tend to be lower and directed more towards the North rather than towards the pit centre. Along the south side of the south pit, dip angles are low and near the southern contact of the kimberlite pipe, bedding planes locally dip shallowly south suggesting the existence of a ?role over? anticline against a bounding fault that would have constituted the south or south-south-east margin of the crater rim (Dirks, 2003). Shear fractures in fine- to medium-grained, bedded vo lcaniclastic, epiclastic and pyroclastic kimberlite The finer grained bedded kimberlite units, especially where close to the crater rim, i.e. in regions where the dip of the bedding planes exceeds 15 degrees, contain extremely discrete curvi-planar surfaces that are characterised by highly polished, strongly striated planes. These fault planes are commonly less than 0.5 cm thick, very sharp surfaces with carbonate deposition along them (Dirks, 2003). Shear fractures in Basalt Breccia Kimberlite In the massive, coarse-grained, Basalt Breccia Kimberlites, many of the discrete curvi-planar fractures described above cannot be traced. In general, fracture planes within the finer grained, bedded kimberlite units that occur interbedded with the Basalt Breccia, curve to become with the contact zone (Dirks, 2003). 10.5 Structural Geology Letlhakane No major (1st or 2nd order) geological structures (shear zones/faults) have been identified which cross the pit. The major structural feature in the pit is a 400 m wide, NW (305?) trending shear domain into which the D/K1 and D/K2 kimberlite pipes have intruded. This shear domain is characterised in the basalt by parallel sets of 330? trending, near-vertical and mainly dextral shears with a typical strike length of 750m and a spacing of 50m Dirks (2001). The shear zones in the basalt are paralleled by extension veins with abundant openings. The dominant joint sets are summarised in Appendix 6, Table A6.3 and Table A6.4. The structures within the basalt are paralleled within the sandstone in several ways. Some shears appear to deflect into an altered zone parallel to the basalt/sandstone contact, whereas in other places master joint planes in the basalt are parallel in the sandstone as 5-10 m wide fracture zones, with a great number of (sub-) parallel fracture/joint planes. Strain on vertical shear zones within basalt is continued into sandstone along 5-10m wide fracture zones consisting of moderately to steeply west and east dipping conjugate fault sets accommodating normal displacements (Dirks, 2001). PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 10: Case Study 6, Orapa and Letlhakane Open Pits Page 10-7 The red mudstone units contain the ubiquitous north-west trending joint planes. Dirks (2001) indicated that emplacment of the kimberlites most likely took place during shearing along the major shear domain and that consequently no thorough going fractures are found in the kimberlite. 10.6 Groundwater The Ntane sandstone formation is considered to be the main aquifer of the upper Karoo sediments at Orapa and Letlhakane (Bush, Hofmann, and Van Rensburg, 1995). The low yield primary and high yield secondary aquifer have the most significant zone occurring at the contact with the basalt flows. The Mosolotsane formation, especially its coarse fluviatile arkoses, is also believed to form important aquifer horizons. The massive mudstones of the Tlhabala unit form a highly impermeable zone which acts as an aquiclude (Jakubec et al,2000). Very little is k nown about groundwater conditions below the Tlhabala unit. From core logging two possible aquifers were defined in coarse arkose units, one at the top of the unit with a 10 m head, and a second with a 30 m head in the middle of the unit (Jakubec et al,2000). There are forty two dewatering boreholes around the pit to reduce the impact of groundwater on the pit slopes. Twenty-eight of the boreholes were terminated in the Tlhabala formation. 10.7 Characterisation of the Country Rock Mass Jakubec et al (2000) report on the geotechnical asse ssment of the different rock types found at Orapa based on rock-mass classification, results of laboratory test work and field observations in the A/K1 pit. The results of their laboratory testing are summarised in Appendix 6, Table A6.5. The average values of the geotechnical parameters determined by Jak ubec et al (2000) are summarised in Appendix 6, Table A6.6, and Appendix 6, Table A6.7 shows the values of their derived intact rock and natural joint parameters. Jakubec et al (2000) establis hed the following rock mass characterisation: 1. Basalt lavas (Unit 110) - most of the basalt encountered is strong to very strong and has good rock-mass quality , similar to the basalt at Letlhakane Mine. Except for some small sections of weathered basalt close to the sandstone contact and on the top of individual lava flows, the basalt is not susceptible to weathering. The joints are mostly closed and filled with calcite. There are occasional sandstone lenses on top of the lava flows (Jakubec et al, 2000). 2. Massive sandstone (Unit 120). The massive sandstone unit underlying the basalt is medium strong and has a fair-quality rock mass. The sandstone has variable thickness and is not susceptible to weathering (Jakubec et al, 2000). 3. Laminated sandstones (Unit 130) . Layers of massive sandstone, which is a medium strong rock, are intercalated with weak to very weak mudstones. The amount of mudstone increases towards the base of the unit. In general, the unit has a fair to good rock-mass quality. While the sandstones are not susceptible to weathering, the mudstone units will disintegrate very quickly. As a result of sub-horizontal bedding, the strength of the rock in the vertical direction is higher than in the horizontal direction (Jakubec et al, 2000). 4. Massive mudstones (Unit 140) . The mudstones are competent medium to strong rocks which have a fair rock-mass quality if not exposed to weathering. The joints are mostly closed and filled with calcite. Except for occasional siltstone-sandstone horizons, which are also generally stronger, the whole unit is extremely susceptible to weathering and will disintegrate rapidly if exposed to weathering agents (Jakubec et al, 2000). PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 10: Case Study 6, Orapa and Letlhakane Open Pits Page 10-8 5. Carbonaceous mudstones-coal (Unit 150). There are three major carbonaceous horizons. Lower, weak horizons are richer in coal, while the upper, medium strong horizon is siltier, with variable amounts of sulphates and sulphides, which make the unit more susceptible to weathering (Buhmann and Atanasova, 1997). In general the horizons are highly jointed and are described as poor rock-mass quality (Jakubec et al, 2000). 6. Mea Arkose (Unit 160). No parameters are available for this sandstone unit, but it is thought to be similar to the sandstones of Units 120 and 130 (Rams den, 2006). 7. Granitic basement (Unit 170) . Tonalite granite is strong to very strong and has good rock mass quality, with some weaker amphibolitic zones. Very little is known about joint directions, and the rocks of this unit are not susceptible to weathering (Jak ubec et al, 2000). 8. Pyroclastic kimberlite. In general, the unit consists of weak to very weak rocks of the crater facies. These kimberlitic rocks are above 600 mamsl and are described as of poor to fair rock mass quality. These rocks are highly susceptible to weathering. From in pit measurements, the weathering rate of intact kimberlite is 50 to 200 mm per year (Jakubec et al, 2000). 9. TKB kimberlite (Unit 220). The kimberlitic diatreme facies below 600 m amsl are medium strong to strong rocks with fair to good rock mass quality. Like the crater-facies kimberlite, the diatreme- facies types of kimberlite rock are very susceptible to weathering (Jakubec et al, 2000). 10.8 Representative Sections Representative sections through the pits are as follows; ? Orapa Pit ? North South Section shown in Figure 10.4 ? Letlhakane Pit ? North South Section shown in Figure 10.5 10.9 Pit Configuration Parameters Currently the pit at Orapa mine is oval shaped with approximate dimensions of 1770 m by 1150 m. The pit rim is at a level of 950 m (NW) to 962 m (SE) above mean sea level and the bottom is currently at 756m. The pit is divided into a North Pit and a South Pit separated by a necked area. Design bench heights are 10 m within kimberlite, and 15 m high within the basalt. The overall pit slope in the kimberlite unit is approximately 50?, but variability exists dependent on rock type, with steeper angles attained in the massive Basalt Breccia units compared to fine to medium grained layered, volcaniclasic, debris flow and talus slope deposits, which are typically deeply weathered. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 10: Case Study 6, Orapa and Letlhakane Open Pits Page 10-9 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 10: Case Study 6, Orapa and Letlhakane Open Pits Page 10-10 Slope angles in the North Pit are slightly steeper than in the South Pit due to the current cut back. Optimal slopes for the basalts have been assessed at approximately 69 degrees (Dirks, 2003). The September 2006 Orapa and Letlhakane pit configuration parameters are listed in Table 10.1 and 10.2. Table 10.1 : Orapa Overall Pit Configuration Parameters, September 2006 East Slope West Slope Weathered Fresh Weathered Fresh Height (m) 181m 174m Overall slope angle (?) (OSA) 35 ? (all in kimberlite) 28? (kimberlite + basalt) Bench stack angle (?) (BSA) 42 ? (upper) 51? (lower) 56? (upper basalt) 38? (lower kimberlite) Bench face angle (?) (BFA) 90 ? 90? Inter ramp angle (?) (IRA) 35 ? (upper) 48? (lower) 55? (upper basalt) 30? (lower kimberlite) Spill/catch berm width (m) 7m (basalt), 10m (kimberlites), 12m (mudstones) 7m (basalt), 10m (kimberlites), 12m (mudstones) Spill/catch berm height (m) 15m (all) 15m (all) Section Bearing (?) (pit cords) 90 ? 90? Pit coordinate system orientation is Lo 25 Table 10.2 : Letlhakane Overall Pit Configuration Parameters, September 2006 North Slope South Slope Weathered Fresh Weathered Fresh Height (m) 256m 272m Overall slope angle (?) (OSA) 45 ? 34? (including current cut) Bench stack angle (?) (BSA) 61 ? (upper) 47?, 56? (lower) 44? (upper), 54? and 42? (lower) Bench face angle (?) (BFA) 40? (in mudstone) to 66? (in Basalt) 40? (in mudstone) to 66? (in Basalt) Inter ramp angle (?) (IRA) 61 ?(upper), 45?, 51? (lower) 43? (upper), 49?, 42? (lower) Spill/catch berm width (m) varies from 6.8m in basalt to 11m in the kimberlite varies from 6.8m in basalt to 11m in the kimberlite Spill/catch berm height (m) 7m in mudstones, 14m rest 7m in mudstones, 14m rest Section Bearing (?) (pit cords) 180 ? 180? Pit coordinate system orientation is Lo 25 10.10 Displacement Monitoring Systems Prism monitoring takes place throughout both Orapa and Letlhakane Pits using Gemos software and Leica TCA2003 automated total stations. 1 radar system has been installed in each pit since July 2006. The radars are continually repositioned according to mining locations and requirements PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 10: Case Study 6, Orapa and Letlhakane Open Pits Page 10-11 (Ramsden, 2006). Due to the recent installation of the radars, no failures or noted instabilities have yet been monitored by them. 10.11 Displacement Behaviour of Non-failed Pit Wall Sectors Prism layouts for Orapa and Letlhakane pits are shown in Appendix 6, Figures A6.3 and A6.4. Deformation behaviour for the stable walls in the Orapa pit are illustrated in the following Figures contained in Appendix 6, Prisms on the east wall instability : Figure A6.5 Selected prisms on south east wall Figure A6.6 Selected prisms from rest of pit Figure A6.7 Deformation behaviour for the stable walls in the Letlhakane pit are illustrated in the following Figures contained in Appendix 6, Selected prisms on the south wall: Figure A6.8 Selected prisms on the north wall: Figure A6.9 With the exception of the SE zone instability (further details from Felix?) the pit wall show very little deformational movement. 10.12 History of Failures 10.12.1 Orapa A/K1 Failure Illustrations and photographs of the A/K1 Failure are shown in Appendix 6, Figure A6.10. When 8th September 2000 and reactivation on the 21st and 22nd September 2000 Location East wall saddle Cut 1 extending from the south ramp to base of pit. Structure, Mechanism and Trigger 3 bench failures were recorded in this area between 1995 and 2000 (Orapa Geotec Dept, 2000). The failure occurred within the Basalt rich kimberlite and bedded talus deposit. The suspected failure surface was the slickensided contact between the bedded talus deposit and the basalt rich Kimberlite which effectively created a planar sliding surface (Ramsden, 2006). The trigger was high rain fall which resulted in a high phreatic surface within the kimberlite (Orapa Geotec Dept, 2000). Displacement Behaviour of Slopes Leading up to Failure Prism database lost. No deformation behaviour evaluation undertaken. Summary and Interpretation of Deformation Behaviour No further comment. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 10: Case Study 6, Orapa and Letlhakane Open Pits Page 10-12 Cut 2 South East Ramp Access Failure Illustrations and photographs of the Cut 2 South East Ramp Failure are shown in Appendix 6, Figure A6.11. When 23rd December 2004 Location South wall Cut 2 (SE Ramp) Structure, Mechanism and Trigger The failure occurred on the slickensided contact between the basalt and the southern volcaniclastic kimberlite (SVK). The mechanism was that of planar sliding. The trigger was that of a high phreatic surface resulting from increased rainfall (Orapa Geotec Dept, 2005). Displacement Behaviour of Slopes Leading up to Failure 5 single bench failures were recorded in area around the ramp prior to the main failure. In December 2004 cracks were noted on the highwall and on 21st December a crack developed on the ramp. Failure occurred on the 23rd December 2004. The prism database has been lost. No deformation behaviour evaluation was therefore undertaken (Orapa Geotec Dept, 2005). Summary and Interpretation of Deformation Behaviour No further comment. 10.12.2 Letlhakane South West Corner Sandstone Failure Illustrations and photographs of the South West Corner Sandstone Failure are shown in Appendix 6, Figure A6.12. When 14th July 2005. Location South West corner of the open pit, Cut No:4, E11-G11, 878m level to F10-G12, 808m level Structure, Mechanism and Trigger Height (m): 70m (5 benches) Strike Length (m): 183m Width (m): 47m (at the base) Volume (m3): 233 000m3 Failure occurred in the fine-medium grained Mosolotsane Sandstone with weak Red Mudstone lenses prevalent in the bottom sequence. The controlling structures were steep NW-SE trending highly persistent joints with rough slightly undulating surfaces, no joint infill and joint spacing of 0.3 to 3.0m. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 10: Case Study 6, Orapa and Letlhakane Open Pits Page 10-13 The mechanism of failure was that of toppling failure with the vertical release surface observed to be along a large crack parallel to these joints. The failure occurred below the water table and no rainfall immediately prior to failure was recorded. Last mining in the area occurred on the 23rd June 2005 (Janse van Rensburg, 2006). Deformation Behaviour of Slopes leading up to Failure The following summary of the deformation behaviour leading up to the failure were obtained from Kayesa (2005). Accelerated deformation behaviour of prisms S34 and S38 in the cracked area had been noted since 8 July 2005. Rapid acceleration was noted at 10:30 on the 14th July 2005 and collapse occurred at 11:56. Prisms were calculated to be moving in north-easterly direction (0390-0430) at an angle of 60-100 downwards. Slope failure occurred in the direction predicted by the prism movement. Maximum prism deformation since installation was calculated to be 742mm on prism S34. Immediately prior to failure a deformation rate of approximately 243mm/day was recorded for prism S34 compared to 20mm/day for the previous day. Vertical prism movement leading up to the failure was negligible. Summary and Interpretation of Deformation Behaviour It seems unlikely that a classical ?toppling? mechanism, as described by witnesses, could have developed on horizontal strata even with subvertical release surfaces (joints). It is more likely that the proximal trigger was a bearing failure of the base of the unstable rock mass through the weaker mudstone units would have facilitated an overall rotational behaviour causing the upper rock mass to appear to rotate outwards and ?topple?. This is confirmed by the negligible vertical movement recorded up until collapse. Examination of the failure surface was not possible as the failed material has been left in place (to be mined out in the next cut). 10.13 Discussion The deformation behaviour and failure mechanisms of the two open pits in near identical host rock differ significantly. Due to the size of the Orapa kimberlite pipe most mining in cut 1 and 2 appear to have taken place within the kimberlite facies with only the latter stages of cut 2 mining into the basalt and exposing Ntane sandstones. Consequently the two large near surface failures experienced have been largely controlled by the slickensided volcaniclastic kimberlite/basalt contacts forming the perimeter of the pipe and facilitated by both the bedding planes found within the kimberlite beccias and the weakened strength characteristics of the highly weathered kimberlites. In contrast the July 2005 Letlhakane failure took plac e in fresh host rock nose (sandstone underlain by mudstone) and was most likely triggered by a bearing failure through the underlying mudstone unit and release surfaces formed by subvertical jointing in the sandstone. The deformational behaviour of the slopes in the two open pits reflect very stable conditions for different reasons. In the Orapa pit the reasons relate to the flat overall slopes in the kimberlite facies and the relatively shallow depth of the pit. In the significantly deeper Letlhakane pit the reason is most likely the arching effect resulting from the near circular nature of the open pit. Current instability of the south-eastern section of the Orapa pit relates directly to a major S-E trending structure cross cutting the pit. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-1 11 DEVELOPMENT OF A GENERALISED TIME AND EVENT DEPENDENT DEFORMATION BEHAVIOUR MODEL The deformation behaviour model developed in this research is dealt with over three separate sections. In this section the different attributes of the model are presented and described. In Section 12 the application of the model is demonstrated using deformation data collected in the literature case study database (Section 3) as well as the six detailed case studies (Sections 5 to 10). In Section 13 a statistical based approach is presented which has been developed to enable successive forecasts of deformation rate and acceleration to be made on a real time basis. 11.1 Introduction In this section a characteristic time and event dependent model of rock slope deformation behaviour is presented. The development of this model was facilitated by the extensive review of deformation behaviour and failure characteristics of both literature and detailed case studies. The model incorporates both a description of horizontal and vertical deformation behaviour patterns from the start of mining of the rock slope through to collapse or post mining recovery and/or stabilisation. It must be noted once again that deformation behaviour of rock slopes varies considerably and that no two rock slopes can be expected to behave in the same way. As discussed earlier, deformation behaviour is influenced by a large number of factors some of which are slope geometry, height, intact rock type/s as well as the nature, orientation, spacing, infill and persistence of discontinuities making up the rock mass. Additionally, other transient factors such as the position of phreatic surface, weathering, seismicity and blasting practices can also have a significant influence. The particular emphasis of the model has been on: a) Identifying predictive trends which can be used to characterise deformation behaviour of rock mass close to failure, in particular, changing deformation rates. b) Representing deformation behaviour development across an unstable rock mass. c) Representing post collapse deformation behaviour. d) Representing associated vertical deformation behaviour. An initial objective was to link different instability modes and mechanisms with particular types of deformation behaviour. However an important finding of this research is that the pre-collapse deformation rate behaviour is largely independent of the slope failure mechanism and mode. Therefore, the real potential for prediction is based on changing deformation rate patterns rather than identifying modes of instability or making comparisons of instantaneous deformation rates. As part of the development of the Model, the following questions regarding changing deformation rates, as identified in Section 4.1, have been researched and will be addressed in the following three sections. 1. Does a negative exponential function result in the best fit for all deformation data, or are other functions more applicable for different rock masses or failure modes? PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-2 2. Can a threshold terminal steady-state creep rate be identified for different rock masses above which tertiary creep or progressive behaviour of the slope is initiated? 3. Where and how do earlier time-to-collapse forecasting theories fit in with the new Model? 11.2 Development of the Time and Event Dependent Deformation Model The primary objective of the development of a time and event dependent deformation model is to provide a framework whereby different displacement behaviour patterns can be directly compared and quantitatively evaluated. 11.2.1 Deformation Behaviour Terminology Terminology used to describe different rock mass deformation behaviour has varied considerably in the past and has consequently resulted in inconsistent usage. The deformation behaviour terminology used in this thesis maintains the meanings most commonly used by practioners and for which they are best understood. These are defined below for clarification. ? Events Events are used to describe individual mining related cuts/pushbacks/blasts/excavations which directly result in a change in the global stress field within the rock slope, and consequently trigger a deformation response from the rock mass. By their very nature, mining events are time independent in the sense that they occur as and when they are planned. For that reason events are reflected in the deformation model as occurring at relatively random time intervals. Events are shown as blue arrows. ? Onset-of-Failure Point The onset-of-failure point (OOF) was used by Zavodni and Broadbent (1980) as the point defining the transition from the ?regressive stage? to the ?progressive stage?. In that sense the meaning is retained. It is further defined, for the purposes of this research, as the point in time beyond which displacements will continue to accelerate to collapse, unless remedial measures are implemented. ? Time Dependency As has been shown throughout this research, the very nature of the non-linear rock mass response is time dependent from the point in time when the change in the stress field occurred. The time dependent response behaviour, which is triggered by a specific event, is strongly influenced by the characteristics of the individual rock mass at the time of the event, as discussed in Section 11.1 above. ? Initial response Contrary to Martin?s (1993) usage of the term, as discussed in Section 2.4.2, the term ?initial response? is used here to refer to the immediate accelerated deformation response following a rapid change in the global stress field usually as a result of a mining event. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-3 ? Deformation Rate Decay Function This term refers to the recovery of the deformation rate from the accelerated movements caused by the initial response down to a point where no further creep takes place. The decay function is usually divided into a transient creep phase and a steady state creep phase. The mechanism within the rock mass causing the deformation rate decay is that of strain hardening. ? Rock Mass Creep Mode This refers to the regressive (creep) behaviour characteristics of the rock mass after an event and initial response (strain hardening) has occurred. ? Deformation Pathway A deformation pathway shows the continuous deformation time relationship of a particular monitoring point on a slope, in the context of the five deformation stages of the Deformation Behaviour Model, as discussed below. ? The Model A truncated term used for the Time and Event Dependent Deformation Behaviour Model. 11.3 Generalised Time and Event Dependent Rock Mass Deformation Model Based on the evolution of slope behaviour leading up to the onset-of-failure, the pre-onset-of-failure deformation characteristics in rock slopes can be differentiated into two broad categories, which are discussed below. The categorization of the pre-onset-of-failure deformation behaviour is principally concerned with the presence or absence of event driven deformation behaviour. 11.3.1 Category 1, Pre-onset-of -failure Deformation Behaviour The most significant deformation behavioural characteristics of Category 1 pre-onset-of-failure deformation is that the deformation usually reflects a continuous period of specific, macro event (mining) driven, deformation behaviour prior to the onset-of-failure, ie they are event driven. This is because the stability of the slope is controlled or being maintained by the remaining material at the base of the slope which acts as a buttress. The rate at which the slope progresses to the onset-of- failure is therefore, to an extent, controlled by the rate at which mining removes the buttressing material or resisting forces. Category 1 deformation behaviour evolution, leading to the onset-of-failure, generally occurs over a relatively long period of time compared with Category 2 deformation evolution. 11.3.2 Category 2, Pre-onset-of -failure Deformation Behaviour The most significant characteristic of Category 2 pre-onset-of-failure deformation is that the deformation does generally not reflect a continuous period of specific macro event (mining) influenced deformation behaviour prior to the onset-of-failure. The onset of failure can occur with little or no prior warning in the way of changing deformation patterns and can often be attributed to micro events, time dependent decay of the static coefficient of friction and stiffness of joints and/or crack propagation. Category 2 deformation behaviour typically occurs in the form of smaller, hard rocks failures, which develop relatively rapidly as ?hot spots? on the wall of the pit. The resulting failures usually occur after mining at the base of the instability has been completed and the location of the instability may PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-4 therefore often be situated up on the pit wall above the floor of the pit. Category 2 pre-onset-of-failure deformation behaviour evolution appears to be largely ?automated? in the sense that once instability is initiated it usually progresses to collapse without requiring any further ?assistance?. The resulting failures generally occur over a relatively short span of time. The rate at which instability develops is largely a function of the characteristics of the rock mass. 11.3.3 Deformation Stages of the Model Time and event dependent deformation behaviour in the Model can be characterised, both in terms of horizontal and vertical displacements and displacement rates, into five distinct stages of deformation behaviour as follows; ? Stage 1 : Pre-collapse, Primary Rock Mass Creep Modes ? Stage 2 : Pre-collapse, Secondary Rock Mass Creep Modes ? Stage 3 : Post-onset-of-failure to Collapse Behaviour Modes ? Stage 4 : Post-collapse Behaviour Modes ? Stage 5 : Post-mining/Recovery Behaviour Modes The characteristic deformation behaviour of each stage is described in the sections which follow. 11.3.4 Basic Deformation Behaviour The model is illustrated initially as simply as possible by using the horizontal deformation pathway of a single hypothetical point on a continuously excavated slope, which eventually collapses. The conceptual deformation pathway for the point is plotted using a deformation time relationship, where the deformation can be expressed in terms of horizontal or vertical displacements or displacement rates. The deformation pathway in terms of horizontal displacement is illustrated in Figure 11.1 and in terms of deformation rate in Figure 11.2. Both Figures should be read in conjunction with each other. Stages 1 and 2 deformation patterns encompass the time frame from initiation of mining activities up until the onset-of-failure (OOF) or alternatively, mining is completed. All the patterns for these two stages are characterised by regressive behaviour, ie all deformation behaviour where failure (collapse) has not been initiated. Stage 1 deformation behaviour is characterised in between mining events by constant, steady state rock mass creep rates. Stages 1 (Horizontal Deformation Behaviour) In general, the deformation rates after each mining event for Stages 1 and 2 are characterised by an initial acceleration spike, followed by a reduction in the deformation rate, which returns ultimately to zero. This overall reduction in the deformation rate is also referred to as a deformation rate decay PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-5 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-6 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-7 function. The horizontal displacement magnitude after each event correspondingly shows an initial rapid increase in total deformation over a relatively short period of time, followed by a change over to a more gradual and ?steady state? increase in deformation with time, corresponding to the latter phase of the decay rate function. This pattern of behaviour following a particular mining event has been subdivided into three phases. In terms of deformation rate, Phase 1 is an initial (sudden) response, characterised by rapid acceleration usually resulting from a sudden change in the stress field following a blast. This is immediately followed by transient creep (Phase 2), which is characterised by a rapidly reducing deformation rate. Phase 3 is termed a steady state creep, which is characterised by a slow and relatively constant deformation of the rock mass over a long period of time, gradually reducing to a zero deformation rate over a very long period of time. This Stage 1 behaviour suggests a two stage linear (bilinear) decay rate function for the transient and steady state creep phases. The nature of this two stage decay function is discussed in further detail in Section 12. Stage 1 deformation is therefore characterised by 3 phases of deformation, with the deformation decay rate function covering phases 2 and 3 which can be approximated by a two stage linear function. The overall behaviour is regressive. Stage 2 (Horizontal Deformation Behaviour) The principal difference between Stage1 and 2 is that it was recognised that, in general, the more stressed a rock slope becomes, the more the characteristics of the deformation rate decay function change. Unlike the three well defined phases associat ed with Stage 1, Phase 3 (steady state creep) deformation behaviour is no longer achieved within the time frame between events. The average deformation rates associated with deformation decay (transient creep) remain increasingly higher for longer and longer. In effect the decay function changes from an approximated two stage linear function into a negative exponential relationship of the form shown in equation 2.1 (Section 2.5.2), as reported by Martin (1993). The changing nature of the deformation rate decay function translates into a change in the negative exponential relationship constants as illustrated in Figure 4.2 to 4.4 (Section 4). The full nature of the change in these constants will be discussed in Section 12. It is possible that cracks may occur at the crest of the slope before the onset of Stage 3. In this case the rock mass behind the crack may start to exhibit deceleration and elastic rebound. This possibility is catered for by the dashed deformation pathway which results in the rock mass entering into a Stage 5 recovery mode of deformation. Stage 2 deformation is therefore characterised by 2 phases of deformation with the deformation decay rate function for phase 2 becoming a continuously changing negative exponential relationship. The overall behaviour is still regressive. Stage 1 to 2 Transition In available deformation data there appears to be no clear cut transition point in the change in behaviour from Stage 1 to 2. Therefore provision is made in the model for a transitional stage between Stage 1 and 2 where, to some degree, characteristics of both Stage 1 and Stage 2 can be manifest. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-8 Stage 3 (Horizontal Deformation Behaviour) A deformation pathway enters Stage 3 when the rock mass passes the onset-of-failure point (OOF). This stage is characterised by a continuous acceleration in the magnitude of the deformation, and the deformation rate, of the unstable rock mass, up until the point of collapse. At the same time, the stable rock mass behind the failure mass starts to undergo a corresponding elastic rebound as shown by the dashed deformation pathway. This rebound is usually associated with large scale downward developing crack formation and is characterised by negative horizontal deformation rates (in a direction directly opposite to that of the unstable rock mass) that peak at or just after the point of collapse. The overall behaviour of the failure mass is progressive. Stage 4 (Horizontal Deformation Behaviour) Post collapse refers to the behaviour of the rock mass after the deformation pathway crosses the point of collapse. The post collapse behaviour of the rock mass can be complicated but for the initial illustration, a classic disintegration of the rock mass is shown. After the point of collapse the deformation rate can still marginally increase as the collapsed material accelerates under gravity but ultimately reduces to a zero within a short space of time as the collapsed material rapidly comes to rest. The deformation magnitude is shown to accelerate directly off the graph. The rebound (dashed line) deformation rate of the non-failed rock mass may follow a complicated seesaw pattern whereby the non-failed rock mass, after fully rebounding away from the failure then rebounds to a lesser extent back again in the direction of the failure, finally establishing a steady state creep rate corresponding to the new slope geometry, and similar to the steady state creep established after a mining event. Stage 5 (Horizontal Deformation Behaviour) The Stage 5 deformation behaviour of the non-failed rock mass usually relates to the stabilisation and recovery of the deformation patterns following mining or a failure. It is by definition a time frame where the rock mass is only influenced by time dependent behaviour (deformation rate decay functions) set up during stages 1 to 4. 11.3.5 Typical Sequentially Mined Deformation Behaviour for Rock Slopes Leading to Collapse The next level of complexity in describing the deformation behaviour in the Model is that of illustrating the deformation behaviour associated with sequential mining. Figure 11.3 shows typical horizontal and vertical deformation behaviour associated with each new mining cut. Horizontal Deformation Behaviour There are several aspects of sequential horizontal deformation behaviour leading to collapse that are apparent compared to the deformation of a single point. These are; PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-9 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-10 a) Different sections of the same slope can be in different stages of regressive behaviour (Stage 1 and 2) at the same time. This is related to the fact that the deeper the monitoring points (towards the base of the pit excavation) the less opportunity they have had to undergo horizontal displacement. Lower monitoring points could therefore conceivably be in a Stage 1 creep mode while monitoring points higher up the slope could be in a Stage 2 creep mode. There may therefore effectively be a time lag in the horizontal behaviour of the lower monitoring points. b) Not all sections of the initial instability enter Stage 3 simultaneously. This is related to the concept of damage evolution where instability may be shown to propagate across a rock mass. c) All points involved in the initial instability do however collapse (enter Phase 4) simultaneously. This infers that the Phase 3 acceleration characteristics of the different monitoring points differ. Should this unstable rock mass not disintegrate, the deformation pathways of all the points merge into a single pathway, and any further reported deformation behaviour of the collapsed rock mass is reflected as a single deformation pathway. Vertical Deformation Behaviour The vertical deformation behaviour for a consecutively mined slope (deformation pathways [a] to [g]) shows that the magnitude of positive vertical elastic rebound increases with depth at an established relationship. However once the positive rebound has occurred, monitoring points tend to displace vertically downwards over the long term. The rebound relationship/function is discussed further in Section 12. Of particular importance is that the vertical deformation behaviour does not appear to be as strongly influenced by, or as sensitive to mining events lower on the slope as horizontal deformations are. There are two likely reasons for this; a) The first is that the physical removal of blasted material by mining operations, which leads to the reduction in vertical stresses and consequently to vertical rebound, takes place over a considerably longer time frame compared with the relatively sudden change in the global stress field influencing horizontal deformations. b) A further reason for the apparently low sensitivity of vertical deformations is the lower accuracy of vertical measurements compared with horizontal measurements. For these reasons, existing vertical deformation pathways are plotted as having near continuous deformation behaviour in subsequent mining events. 11.3.6 Typical Sequentially Mined Deformation Beha viour for Rock Slopes with Partial Collapse of the Slope Collapse of part of the slope would obviously lead to sections of the slope displaying both collapse and long term stability. The method by which the Model can be used to illustrate this dual deformation behaviour is shown in Figure 11.4. As discussed in Section 11.3.4, a key feature of the model is that it shows that deformation of sections of the non-failed slope adjacent to the failure almost always reflect PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-11 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-12 a rebound response to the failure. This may be followed by a sympathetic change in direction of the resultant vector movement in towards the failure. This can be attributed to a further change in the global stress field as a result of the removal of the direct or lateral support of the adjacent failed material. 11.3.7 Typical Sequentially Mined Horizontal Deformation Behaviour for Non-Collapsed Slopes Typical horizontal deformation behaviour for a sequentially mined rock slope (deformation pathways [a] to [g]) where no final collapse takes place, is illu strated in Figure 11.5. In this case deformation pathways pass from Stages 1 and 2 straight into Stage 5 as there is no onset of failure or collapse. The horizontal deformation rate behaviour illustrated is similar to the deformation rate behaviour described above (Section 11.3.5). Once again the deformational pathways shown in Figure 11.5 illustrate how it is possible that different elevational zones on any given slope can exhibit different stages of deformation simultaneously. 11.3.8 Typical Sequentially Mined Vertical Defo rmation Behaviour for Non-Collapsed Slopes The vertical deformation behaviour is characterised by two important deformation patterns; 1. A monitoring point on a mined slope typically reflects a pattern whereby there is an initial positive (upwards) movement of the monitoring point as the rock experiences rebound followed by a long term gradual negative (downwards) deformation component as the rock slope relaxes. This negative deformation which takes place over the long term is most likely a function of creep both within the intact rock and along discontinuities forming the rock mass. 2. The magnitude of positive vertical elastic rebound increases with depth, which can be intuitively expected. Therefore each consecutive (deeper) mining level can be expected to experience larger rebound magnitudes in comparison with those levels higher up. These vertical deformation behaviour patterns for sequentially mined rock slopes where no final collapse takes place are simplistically illustrated in Figure 11.6. It should however be noted that a considerable variation in this pattern of vertical deformation does occur. The rate of mining and nature of the rock mass play a significant influence on the rebound behaviour. Rapid mining in hard rock masses can result in a stack of several benches all exhibiting positive vertical rebound simultaneously, but at differing rates, whereas slower mining in a softer (weathered) rock mass can result in upper prisms reflecting overall negative vertical velocities, whilst lower prisms exhibit positive vertical rebound velocities. The extremes of these vertical deformation patterns have been illustrated in Figures 11.7 and 11.8. Figure 11.7 illustrates vertical deformation behaviour upon a hypothetical instantaneous removal of rock material from an excavation. It is important to note, in this illustration, that the vertical rebound zone of influence is not limited to the area immediately within the pit perimeter, but can also extend to an area some distance away from the pit, as illustrated by the dotted line. Consequently, even the crest (rim) of the pit which was at natural ground level can be expected to experience a limited amount of rebound. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-13 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-14 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-15 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-16 Figure 11.8 illustrates the situation in a hard rock environment, where consecutive mining of deeper levels occurs at a relatively high mining rate. This situation results in the overall rebound velocity of the whole wall, being greater than the negative vertical deformation component occurring as a result of creep and rock mass relaxation. The negative vertical deformational component does, however, dominate once mining stops and the magnitude of the positive vertical rebound reduces. It should be noted that the deformation behaviour illustrated in Figures 11.7 and 11.8 assumes no failure has developed. Once a slope enters a late Stage 2 or progressive behaviour associated with Stage 3, the vertical deformations associated with the failure will override the rebound behaviour and will dominate the vertical behaviour from then onwards. 11.3.9 Modes of Possible Deformation Be haviour for Horizontal Displacements In addition to the relatively simplified and straightforward general deformation behaviour discussed above, detailed deformation behaviour of rock slopes associated with different failure mechanisms can be expected to exhibit a considerable range of complexity and variation. For this reason a range of deformation pathways occurs, which can nevertheless still be accommodated in the Model. This range is illustrated in Figure 11.9 and discussed below and in further detail in Section 12. Alternative deformation pathways are given codes for ease of reference. Stage 1 and 2 Whilst the examples above illustrated a continuous progression of the deformation pathways towards the onset-of-failure, as influenced by each subsequent mining event, in reality different failure modes and mechanisms which influence the sensitivity of the rock mass may cause a considerable variation in the deformation pathways (the detailed deformation behaviour). For example, an unstable rock mass may progress to the onset-of-failure point from Stage 1 or 1 to 2 transition as the result of only one or two mining events as shown by the S2?type 1 and S2?type 2 pathway. Stage 3 Progressive deformation pathways starting from the onset-of-failure to collapse may show a considerable variation in time from almost instantaneous brittle failures (S3?type 3 pathway) which occur within minutes to hours, to failures that develop over several weeks to months or even more (S3?type 1 pathway). In the case of the latter, further mining events (cuts) are possible, however deformation pathways (S3?type 2 pathway) are likely to be characterised by the same initial response and increased transient creep responses as those experienced by mining events in Stages 1 and 2. However, these new responses are likely to accelerate the overall progressive behaviour and correspondingly reduce the remaining time to collapse. A deformation pathway may also reflect near constant but high creep rates (S3?type 4 pathway) where normal mining cannot continue safely and which result in final collapse. Stage 4 Post collapse behaviour patterns essentially describe the behaviour of the rock mass after collapse. In post collapse scenarios where the rock mass has maintained its integrity (ie not disintegrated), complex deformational behaviour of the collapsed rock mass can be expected to occur. Six principal post collapse deformation modes have been identified. These are (with reference to Figure 11.9); ? S4?type 1 : Disintegration, ? S4?type 2 : Partial recovery and gradual deceleration to creeping, PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-17 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-18 ? S4?type 3 : Full recovery, velocity almost completely stops, ? S4?type 4 : Partial recovery followed by another final collapse, ? S4?type 5 : Ratchet mechanism, ? S4?type 6 : High creeping rate and probably accelerating. Deformation modes may alternate between any one of type 2 to 6, in which case the overall mode is referred to as complex. The time frame this period encompasses can range from the order of minutes, to years, which could typically occur in the case of creeping rock masses or ratchet mechanisms. The deformation pathways in Figure 11.9 illustrate that whilst entering this stage as progressive behaviour, deformation can change into regressive behaviour again, and can ultimately flip flop between regressive and progressive behaviour any number of times, depending on transient factors such as changing phreatic surface, changing geometry and seismicity. Stage 5 A post mining or a recovery mode of deformation can be expected to be entered whenever no further mining events or collapses are influencing deformation behaviour. Long term stabilisation of the rock mass is expected as shown in the S5?type 1 and S5?type 2 pathways. However, according to long term creep deformation theory of rock masses and joints, as discussed in Section 2.6 and 2.8, eventual failure of rock is possible under constant stress conditions as a result of the phenomenon of the time dependency of the peak and residual static coefficient of friction of joints, as well as the stiffness of the rock. Consequently the S5?type 3 deformation pathway, which shows a transition from Stage 5 back into Stages 3 and 4, is considered possible. 11.4 Summary In this Section a time and event dependent deformation model has been developed and a description and illustrations have been given as to how deformation behaviour of excavated rock mass may be presented using deformation pathways. The model accommodates five principal stages of deformation ranging from primary and secondary rock mass creep modes through the onset-of-failure to collapse and post collapse or post mining recovery deformation behaviour. The model has been shown to be sufficiently general as to enable deformation patterns resulting from different failure mechanisms to be accommodated. The important features of the new Model are; 1) The provision of two different stages of event induced deformation behaviour prior to the onset- of-failure. The first stage encompasses relatively uniform deformation rate patterns and the second stage encompasses changing deformation rate patterns as a result of changing deformation rate decay functions. 2) A description of deformation behaviour which is largely independent of the resulting slope failure mechanism. 3) Description of the inter-relationship between vertical and horizontal deformation behaviour 4) Description of post collapse and/or post mining/recovery behaviour. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 11 : Time and Event Dependent Deformation Model Page 11-19 The application of the Model to both the detailed case studies and to case studies from the literature is considered in the following Section. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-1 12 APPLICATION OF THE TIME AND EVENT DEPENDENT DEFORMATION MODEL USING CASE STUDY DATA EXAMPLES 12.1 Introduction In the previous section a Generalised Time and Event Dependent Deformation Model was developed and presented. The most important feature of the Model is the way in which deformation behaviour changes as the slopes approach the onset of failure point ie the importance of the deformation rate decay curve. Therefore, contrary to opinions in most literature, it is the deformation behaviour patterns or deformation behaviour characteristics, and not instantaneous values of deformation or deformation rates, which are significant and meaningful. The instantaneous values vary continuously in any given slope according to the deformation rate decay functions. It is therefore in the view of the Author not meaningful in the context of this approach to calculate and compare these values between different slope case studies, although similarities in the instantaneous values of deformation within similar slopes, within the same mining environment or geological terrain, may be anticipated. The approach in demonstrating the application of the model is therefore to; ? Illustrate, in further detail, the typical horizontal displacement and displacement rate behaviour for each deformation pathway in the pre-collapse stages. ? Where possible, provide an example of each of the deformation pathways, using deformation data collected from the case studies. 12.2 Discussion on Stage 1 : Primary Rock Mass Creep Mode The stage 1 primary rock mass creep mode is a behaviour that can be anticipated in the early stages of mining activities, when slope heights remain relatively low and/or when overall slope angles remain relatively flat. The discussion and examples below specifically relate to Category 1, pre-onset-of- failure deformation behaviour, when slope deformation behaviour is being periodically influenced by new mining cuts or activities. The characteristics of the deformation behaviour of Stage 1 are; ? Relatively uniform deformation rate decay functions. ? Bi-linear deformation rate recovery functions. ? Relatively constant steady state rock mass creep rates. ? Overall regressive behaviour of the rock mass. Stage 1, S1?type 2 idealised mode of deformation behaviour is illustrated in Figure 12.1. One example of S1?type 1 deformation behaviour and two examples of S1?type 2 deformation behaviour were selected from prisms situated in the Navachab open pit case study, as shown in Figures 12.2a to 12.2c PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-2 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-3 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-4 12.3 Discussion on Stage 2 : Secondary Rock Mass Creep Mode 12.3.1 Category 1 Pre-onset-of-failure Deformation Behaviour As was discussed in Section 11, the most significant characteristic of Category 1 pre-onset-of-failure deformation behaviour is that it generally reflects a continuous period of specific macro event (mining) influenced deformation behaviour prior to the onset of failure, and the rate at which the slope progresses to the onset-of-failure is therefore largely influenced by the mining rate. The characteristic behaviour patterns of Category 1 Stage 2 deformation behaviour are; ? Changing recovery rate curves (decay rate functions) after each mining event, ? Changing steady state rock mass creep rates after each mining event, ? Overall regressive behaviour of the rock mass is maintained up until the onset-of-failure. In order to further illustrate the three types of associated deformation pathways for this Stage, each has been drawn individually showing the corresponding deformation rate behaviour in the following Figures; ? Figure 12.3 - Stage 2, S2?type 1 idealised mode of deformation behaviour. ? Figure 12.4 - Stage 2, S2?type 2 idealised mode of deformation behaviour. ? Figure 12.5 - Stage 2, S2?type 3 idealised mode of deformation behaviour. Illustrations of Stage 2 deformation behaviour from the literature review case studies, presented in Section 3, are shown in the following Figures; ? Figure 12.6 - Deformation behaviour reported by Broadbent and Ko (1980) using the Inspiration Mines open pits in Nevada. ? Figure 12.7 - Jinchuan Mine West Slope failure. ? Figure 12.8 - Nchanga Mine July 2004 failure. Selected examples from the detailed case studies, presented in Section 8 and 9, are shown in the following Figures; ? Figure 12.9 - 2 event Harmony open pit East Wall failure. ? Figure 12.10 - 3 event Venetia open pit July 2003 failure. The most important aspect of Stage 2 behaviour confirmed by Figures 12.6 to 12.10 is the changing deformations rate behaviour as each of the slope examples approaches collapse. No specific mining event related information was presented for the literature review case studies shown in Figures 12.6 and 12.7. However, for the July 2004 Nchanga failure Naismith and Wessels ( 2005) confirm that deformation rates ?show an increase in association with blasting in the area of concern followed by a decay to pre-blast levels within one or two days?. Records obtained from the mine, confirm that the accelerated deformation rates for both Harmony east wall failure (Figure 12.9) and Venetia?s July 2003 failure (Figure 12.10) were initiated by specific mining events (Leinster Nickel Operations Geotec Dept, 2006) and (Venetia Geotec Department, 2006). PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-5 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-6 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-7 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-8 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-9 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-10 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-11 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-12 12.3.2 Category 2 Pre-onset-of-failure Deformation Behaviour Once again, the most significant characteristic of Category 2 pre-onset-of-failure deformation behaviour is that they generally do not reflect a continuous period of specific macro event (mining) influenced deformation behaviour prior to the onset-of-failure. The onset-of-failure can occur with little or no prior warning in the way of changing deformation patterns. Although this behaviour is usually a characteristic of smaller hard rock failures, similar behaviour has also been found to be displayed in deeply weathered rocks (oxides or saprolites) which contain well defined relic structures. Selected examples from the detailed case studies, of Category 2 deformation, are shown in the following Figures; ? Figure 12.11 and 12.12 - Harmony pit saprolite failures (presented in Section 8). ? Figure 12.13 - Letlhakane July 2005 fa ilure (presented in Section 10). ? Figure 12.14 - Bibiani West Wall failure (presented in Section 6). 12.4 Discussion on Stage 3, Onset-of-failure to Collapse No specific examples are given here since Stage 3 behaviour has already been illustrated in Figures 12.8 to 12.14. The Stage 3 deformation in all these Figures reflects S3?type 1 deformation pathways. 12.5 Discussion on Stage 4, Post Collapse Once again it should be noted that post collapse behaviour is limited to failures in which the failed rock mass has largely maintained its integrity, and monitoring points are still able to be sighted. Generally, examples of post collapse behaviour are not easily obtained as monitoring of the prisms in failed areas is usually stopped, and/or the failed rock mass is usually mined out to reduce the risk of further collapses. In the examples used, failures have been left for removal by subsequent cuts, the pit has been closed (and monitoring continued), or immediate access to the failure has been prevented and monitoring has continued on prisms still able to be sighted. 12.5.1 Disintegration (S 4?type 1 ) Post collapse deformation demonstrating disintegration (S4?type 1 ) involves the rock mass losing its integrity, at which time no further monitoring of the failed material is possible and consequently no deformation data can be shown. Examples from the detailed case studies where disintegration has occurred include; ? Navachab Open Pit, March 2001 failure (presented in Section 5) ? Bibiani Open Pit, all failures (presented in Section 6) ? Mt Keith Open Pit, all failures except the December 2001 SE Wall failure (presented in Section 7) ? Letlhakane Open Pit, July 2005 failure (presented in Section 8) PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-13 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-14 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-15 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-16 12.5.2 Post Collapse Recovery (S 4?type 2 to S4?type 6 ) Post collapse deformation, demonstrating all other modes (S4?type 2 to S4?type 6 ), involves the rock mass maintaining its overall integrity. Identified controls on different post collapse deformation behaviour identified are numerous, and usually relate specifically to the mode of failure, residual strength characteristics of the rock mass, changing geometry, ground water response times and physical measures implemented by the Mine to control and limit further progression of the failure. Physical measures could include the construction of buttresses, dewatering, blasting and mining away failed material. Again examples are used to illustrate the deformation pathways. However, in this section comments are included with regard to the mode of failure and how this has influenced the behaviour. The examples used are as follows; ? S4?type 2 pathway : Partial recovery and gradual deceleration to creeping Figure 12.15 ? Mt Keith December 2001 SE Wall failure (presented in Section 7) This failure was a classic rotational /circular failure, which extended over a height of approximately 95m and was situated within the weathered saprolite zone. The deceleration was attributed to the gradual geometrical stabilisation of the failed material as it rotated along the failure plane. ? S4?type 3 pathway : Full recovery, velocity stops almost completely Figure 12.16 - Harmony pit March 2006 failure (presented in Section 8) This failure was a classic wedge failure where the resultant deformation vector movement of the failed material paralleled the line of intersection of the two controlling joint structures (J2 and J4) along which sliding took place. As the failed mass has not intersected any physical barriers it is likely that the deformation stopped when ground water pressures reduced as a result of internal slope deformation. ? S4?type 3 transitioning into S4?type 6 pathway Figure 12.17 ? Venetia September 2004 Extension to SE Wall Failure (presented in Section 9) This failure was a classic sliding failure, and the creeping movement was most likely a result of mining away of failed material at the toe. ? Complex S4 pathways: Sequential combinations of any S4 pathways. Figure 12.18 ? Mt Keith F Stage South-east Failure (presented in Section 7). This failure initially constituted a wedge mode of instability. The continuous reactivation of the failure which has resulted in the complex post collapse deformation behaviour reflected, has been caused by combinations of blasting and rainfall events, as well as by mining and removal of portions of the failed material. Examples of the following post collapse deformation pathways have been included from the literature case study database; PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-17 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-18 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-19 ? S4?type 4 : Partial recovery followed by a final collapse Luscar Mine, Pit 51-B-2 Northwall failure (Section 3). The northwall failure was a planar sliding failure along bedding, accompanied by horizontal failure through intact rock into the pit at the base of the slope. In this case study the initial collapse of the slope followed by a partial recovery and then final collapse, was attributed to a short lived relief of ground water pressures as a result of changing internal rock geometry during the initial collapse. As ground water pressures built up again, the slope accelerated to final collapse. ? S4?type 5 : Ratchet mechanism Cassiar Mine Case Study, North East Sector Failure (Section 3). Although specific deformation graphs are not given, a description of the ratchet mechanism was summarised from Martin (1993). ? S4?type 6 : Creeping mechanism Large scale creeping mechanisms of slopes are clearly described in Section 2.7 concerning the gravitational creeping behaviour of large scale rock mass. A number of examples of creeping instabilities in Alpine regions is given. 12.6 Discussion on Stage 5, Post Mining/Recovery Stage 5, S5?type 1 idealised mode of deformation behaviour is illustrated in Figure 12.19. The only example available to illustrate Stage 5 deformation behaviour is the Harmony pit Zone 5 prisms from April 2004 to March 2006, as shown in Figure 12.20. In the pit, monitoring was continued long after pit closure, in order to monitor the large east wall failure. The time frame selected includes approximately the last 8 months of mining plus approximately 15 months after mining was completed. The deformation behaviour of the prisms clearly shows the deceleration and stabilisation of the slope into a slow creep S5?type 1 mode of deformation. 12.7 Vertical Rebound As was discussed in Section 11, the vertical deformation behaviour for a consecutively mined slope shows that the magnitude of positive vertical elastic rebound increases with depth. However once the positive rebound has occurred, monitoring points tend to displace vertically downwards over the long term. The rate of mining plays a significant influence on the rebound behaviour. Rapid mining in hard rock masses can result in a stack of several benches all exhibiting positive vertical rebound simultaneously, but at differing rates, whereas slower mining in softer rock masses can result in upper prisms reflecting overall negative vertical velocities, whilst lower prisms exhibit positive vertical rebound velocities. Examples in the detailed case studies illustrating vertical rebound behaviour include; ? Figure 12.21 and 12.22 - Navachab open pit Figure 12.21 illustrates the long term behaviour of a selection of prisms situated in the middle of the east wall for a period of approximately seven years. These prisms were monitored manually. The significance of being situated in the centre of the wall is that these prisms can be expected PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-20 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-21 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-22 to experience the greatest magnitude of rebound. These prisms reflect a relatively rapid mining situation where, as discussed in Section 11.3.8 where in general the overall rebound velocity of the whole wall during the period of mining continues to be greater than the negative vertical deformation component occurring as a result of creep and rock mass relaxation. Figure 12.22 illustrates the vertical behaviour of three prisms on the East wall of the pit, using an automated theodolite and Geomos software, for a period of approximately 1 year (the final year of the seven year period illustrated in Figure 12.21). Prism C6 is situated towards the crest of the pit, C14 approximately half way down and C26 towards the final floor of the pit. In the light of Figure 12.21 it is clear that magnitude of deformation plotted in Figure 12.21 is not a true reflection of the total magnitude of vertical deformation experienced by prisms C6 and C14, as commencement of the automated prism monitoring occurred well into the mining operation. The graph is nevertheless significant as the prisms reflect a general long term trend in negative vertical movement with time as well as a time lag associated with sequential mining. Prism C26, having being installed after commencement of monitoring does reflect actual rebound behaviour. ? Figure 12.23 - Harmony open pit, West Wall In this example the selection of prisms illustrates the characteristic initial rebound behaviour followed by vertical relaxation as a result of mining, and the reduction in deformations as a result of pit closure and post mining Stage 5 recovery and stabilisation (see Figure 12.20 in conjunction with Figure 12.23). 12.8 Summary In this section measured deformation data obtained from the case studies has been applied to the Time and Event Dependent Deformation Model. This exercise has provided confirmation of the following attributes of the Model; 1) That the general deformation behaviour patterns which were described for different stages of the Model in Section 11 are indeed realistic. 2) Of the existence of two distinct stages of deformation (ie Stage 1 and 2) leading up to the onset- of-failure point (Stage 3) 3) That the first stage (Stage 1) is characterised by relatively constant steady state rock mass creep rates and the second stage (Stage 2) by constantly changing deformation behaviour patterns after each event which are associated with changing deformation rate decay functions. 4) Where mining records are available, that the accelerated deformation rates are initiated by specific events which are usually mining related. 5) Of the possibility of complex post collapse deformation behaviour provided the rock mass retains its integrity. Furthermore this exercise has demonstrated how both Category 1 and Category 2 pre-onset-of-failure deformation behaviour can be accommodated in the Model. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-23 The purpose of the Model is not just to characterise deformation behaviour patterns, but to use this knowledge for prediction and forecasting of instabilities and failure. The use of the deformation Model for forecasting is the topic for the following Section. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 12 : Application of the Model Usi ng Case Study Deformation Data Page 12-24 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-1 13 FORECASTING OF DEFORMATION BEHAVIOUR USING THE MODEL 13.1 Introduction One of the main objectives of the development of the Time and Event Dependent Deformation Model is to use the measured deformational characteristics of the model for predicting future behaviour. This section of the thesis presents a statistically based approach, which has been developed to enable successive forecasts of deformation rate and acceleration to be made on a real time basis. It can be anticipated that the principal reason for carrying out forecasting would be to predict the possible times at which onset-of-failure and slope collapse could occur. However, the method which will be described, is sufficiently general to enable it to be used to derive information regarding deformation rate recovery, or decay curves, of special importance in the post mining and recovery phase (Stage 5). It should be noted that a basic assumption regarding forecasting is that no further mining events, which can change deformation behaviour and hence prediction times, will occur. 13.2 The Problem The principal problem with using deformation rates de rived directly from measured deformations is that of dealing with very noisy data. This has alre ady been shown for example in the instantaneous horizontal velocities of the Nchanga Mine July 2004 failure (Figure 3.9d) as well as for the Navachab March 2001 Failure (Figure 5.4). The problem of noise is further illustrated using the relatively good data set from the February 2006 Harmony Pit East Wall failure. The instantaneous prism velocities for a prism situated on this failure are shown in Figure 13.1. The principal cause of the noise is that, in most cases, the change in the magnitude of the deformation between survey epochs is less than the achievable accuracy of the survey equipment, in the given conditions. For example, current survey instruments in use have laboratory calibrated accuracies of 1mm + 1 part per million (Thomson, 2006). This means that if prisms being su rveyed are exactly 1km away from the instrument, the error ellipse for the vertical and horizontal angles can be up to a maximum of 2mm (excluding atmospheric interference). Atmospheric interference, which is a function of temperature, pressure, humidity and air turbulence, can add another 4mm or more to the e rror ellipse (Thomson, 2006). Still further error can be introduced as a result of the survey instrument carrying out regular orientations onto poorly configured survey control networks (Thomson, 2006). A further complication is that this research has shown that evolution of instability and failure in a rock mass can result in a wide range of different deformation behaviour within the limits of the unstable rock mass. Deformation can range from relatively straight forward elastic displacement, to complex behaviour involving the progressive advance of multiple failure surfaces through rock mass, resulting in unequal deformation behaviour across the unstable rock mass. For example, inspection of the time dependent displacement magnitude data for selected prisms in the Harmony Pit East Wall failure (as shown in Figure 13.2) clearly demonstrates that; ? rock in different parts of the failure can experience different acceleration characteristics (or rates of acceleration). ? the onset-of-failure point can vary across the unstable rock mass. ? the magnitudes of deformation measured immediately preceding collapse can also vary across the failing rock mass. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-2 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-3 The difficulty of predicting time to collapse on a real time basis is therefore evident. Clearly, although the overall deformation rate trends may be evident, the magnitude of the noise is significantly amplified when calculating the displacement rate from measured displacements. As a result of this behaviour, general empirical relationships, which have been developed to predict time-to-failure (collapse) as discussed in Section 2.13, are no longer considered to be applicable. Intuitively, the parameter that can be considered to have the most uniform value in any particular failure is the rate of displacement. At the time of collapse, the rate of displacement can be considered to be approximately equal for all parts of a rock mass involved in the collapse. The basis for statistically forecasting collapse can therefore only be made using deformation rates. This is a confirmation, albeit for different reasons, of observations made as long ago as Zavodni and Broadbent (1980), in which literature the authors selected rate of movement as being the most sensitive indication of slope behaviour. 13.3 A Proposed Method of Forecasting Deformation Behaviour The following proposed method enables forecasting to be carried out at any time in which the rock mass is demonstrating Stages 1 to 3 deformation behaviour. The most important operational issue is that of forecasting the time of collapse (t c) when instability has been identified. This is especially important where no prior knowledge of characteristic deformation behaviour leading up to collapse is available. The method is set out in the following steps, ? Step 1 Pre-process the initial deformation data set. Th is would involve cleaning the data by removing outliers or misplaced data points from the data set. ? Step 2 Carry out curve fitting (conventional statistical cu rve fitting using linear or non-linear regression methods) on the displacement magnitude ( ?), or strain, data set and obtain the equation of the fitted curve. The equation that was used for curve fitting is a polynomial fit of the logarithm of the measured displacements, where n is referred to as the ?order? of the polynomial. Before carrying out the curve fitting procedure the deformation time is normalised. Normalisation is a process of scaling the numbers in a data set to improve the accuracy of the subsequent computations. The way in which normalisation is carried out is to centre the deformation time vector on a zero mean and scale it to a unit standard deviation as follows; )( ))(( date dateEdateNdate ? ?= (13.1) Where: ?Ndate? and ?date? are column vectors E(date) is the expected or mean value of the vector ? (date) is the standard deviation of the vector. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-4 ? Step 3 Evaluate the goodness-of-fit by calculating and pl otting the residuals ie the difference between the observed and predicted data within the time range of the fitted deformation data. ? Step 4 Confirm that the fitted curve is well behaved bey ond the fitted data, ie that erroneous polynomial behaviour does not take over past the fitted defor mation data, resulting in a deterioration of the approximation. ? Step 5 Carry out 1 st and 2nd order differentiation of the equation of the fitted curve to obtain the corresponding equations of the displacement rate and acceleration. ? Step 6 When required, solve the differential equation using a specified limiting displacement rate at collapse in order to establish the forecast time of collapse (t c). ? Steps 7 to 9 Repeat steps 1 to 6 whenever new deformation data becomes available, continuously updating forecasts. In a collapse prediction, the exer cise can be abandoned if the instability behaviour changes to regressive behaviour, or restarted if a mining event occurs which changes the deformation behaviour. It should be noted that all statistical methods of fo recasting inherently reflect increasing levels of uncertainty the further into the future the forecasts are made. Early forecasts, or significantly longer duration forecasts should therefore be expected to be relatively inaccurate. However, the accuracy can be expected to continuously improve with further da ta being added and/or as the time of the predicted event is approached. 13.4 Derivation of Equations of Displacement Rate and Velocity 13.4.1 Derivation Using Natural Logarithms )()ln( tf=? (13.2) Where 1 1 21 .....)( +? ++++= nnnn ptptptptf (13.3) therefore; ).....exp( 1121 +? ++++= nnnn ptptptp? (13.4) ))(exp().(' tftfdt d =? (13.5) With the first derivative of the polynomial being; PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-5 nn nn ptptpntpntf +++?+= ??? 12211 .....)1()()(' (13.6) Substituting equation 13.6; ).....exp()......)1()(( 112112211 +???? +++++++?+= nnnnnnnn ptptptpptptpntpndt d ? (13.7) ))(exp(.))('())(exp().('' 22 2 tftftftfdt d +=? (13.8) ))(exp(].))('()(''[ 22 2 tftftfdt d +=? (13.9) With the second derivative of the polynomial being; 1 3 2 2 1 .....)2)(1()1()('' ??? ++??+?= nnn ptpnntpnntf (13.10) Substituting equations 13.6 and 13.10; 1 3 2 2 12 2 .....)2)(1()1([ ??? ++??+?= nnn ptpnntpnndt d ? .....exp(].).....)1()(( 121212211 +++++?++ ???? nnnnnn tptpptptpntpn ).... 1+++ nn ptp (13.11) Substituting equation 13.4; 1 3 2 2 12 2 .....)2)(1()1([ ??? ++??+?= nnn ptpnntpnndt d ? ?].).....)1()(( 212211 nnnn ptptpntpn +++?++ ??? (13.12) 13.4.2 Derivation Using Logarithms to the Base 10 )()(log 10 tf=? (13.13) Where again 1 1 21 .....)( +? ++++= nnnn ptptptptf therefore; )(10 tf=? (13.14) )10exp(ln )( tf=? (13.15) PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-6 )10ln).(exp( tf=? (13.16) )10ln)......exp(( 1121 +? ++++= nnnn ptptptp? (13.17) )10ln).(exp(.10ln)(' tftfdt d =? (13.18) Substituting equation 13.6; .....exp((.10ln).....)1()(( 12112211 +++++?+= ???? nnnnnn tptpptptpntpndt d ? )10ln)...... 1+++ nn ptp (13.19) )10ln).(exp(.10ln)('.10ln)(')10ln).(exp(.10ln)(''2 2 tftftftftfdt d +=? (13.20) )10ln).(exp(].10ln))('()(''[10ln 22 2 tftftfdt d +=? (13.21) ?? ].10ln))('()(''[10ln 22 2 tftfdt d += (13.22) Substituting equations 13.6 and 13.10; +++??+?= ??? ).....)2)(1()1)([((10ln 132212 2 n nn ptpnntpnndt d ? ?].10ln).....)1()(( 212211 nnnn ptptpntpn +++?++ ??? (13.23) 13.5 Categories of Failures As has been discussed in Section 11, there are two categories of deformation leading to the onset-of- failure. It is worthwhile to once again review the deformational characteristics of these two categories; ? Category 1 Pre-onset-of-failure Deformation Behaviour The most significant deformation behavioural char acteristics of Category 1 pre-onset-of-failure deformation is that the deformation usually reflects a continuous period of specific, macro event (mining) driven, deformation behaviour prior to the onset-of-failure, ie they are event driven. This is because the stability of the slope is controlled or being maintained by the remaining material at the base of the slope which acts as a buttress. The rate at which the slope progresses to the onset-of-failure is therefore, to an extent, cont rolled by the rate at which mining removes the buttressing material or resisting forces. Catego ry 1 deformation behaviour evolution, leading to the onset-of-failure, generally occurs over a re latively long period of time compared with Category 2 deformation evolution. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-7 ? Category 2, Pre-onset-of-failure Deformation Behaviour The most significant characteristic of Category 2 pre-onset-of-failure def ormation is that the deformation does generally not reflect a continuous period of specific macro event (mining) influenced deformation behaviour prior to the onset-of-failure. The onset of failure can occur with little or no prior warning in the way of changing deformation patterns and can often be attributed to micro events, time dependent decay of the static coefficient of friction and stiffness of joints and/or crack propagation. Category 2 deformation behaviour typically occurs in the form of smaller, hard rocks failures, which develop relatively rapidly as ?hot spots? on the wall of the pit. The resulting failures usually occur after mining at the base of the instability has been completed and the location of the instability may therefore often be situated up on the pit wall above the floor of the pit. Category 2 pre-onset-of-failure deformation behaviour evolution appears to be largely ?automated? in the sense that once instability is in itiated it usually progresses to collapse without requiring any further ?assistance?. The resulting fa ilures generally occur over a relatively short span of time. The rate at which instability develops is largely a function of the characteristics of the rock mass. Stage 3 deformation behaviour is therefore relevant to both Category 1 and 2 pre-onset-of-failure deformation whilst Stage 2 is only relevant to Cat egory 2 pre-onset-of-failure deformation. The remainder of this section illustrates how the model can be applied to the prediction of time to collapse in Stage 3 and the onset-of-failure in Stage 2. 13.6 Forecasting of Collapse Times for Stage 3 Deformation Behaviour Consideration is now given as to how the Deformation Model can be used in a real time prediction of the time-to-collapse, t c. for Stage 3 deformational behaviour. This stage is characterised by continuous acceleration both in the magnitude of displacement and the displacement rate until the collapse point is reached. The defining point of collapse in the Model is where the acceleration of the slope of the deformation rate curve approaches a limiting deformation rate. Forecasting is therefore a question of predicting the remaining time left to collapse. There are several issues regardi ng the application of this method to Stage 3 deformation which are discussed in further detail below. 13.6.1 Identification of the Onset-of-failure Point A significant advantage of the proposed forecasting method is that the onset-of-failure point does not have to be specifically identified in order to successfully apply this method to forecast the time-to- collapse in Phase 3 deformation. T he reason for this is that knowl edge of the onset-of-failure point does not convey any added value or improvement in forecasting accuracy to the regression equations. The method can be applied to deformation data sets which include pre-onset-of-failure data or data sets which only contain deformation data obtained after the onset-of-failure point. Obviously forecasts regarding the latter are likely to be less accurate. The application of the forecasting method without knowledge of the onset-of-failure point will be demonstrated in the example to follow. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-8 13.6.2 Curve Fitting The power of modern computers and sophistication of available mathematical and statistical software such as MATLAB (Matlab, 2000) can facilitate curve fitting (regression) on a real time basis. Moreover the software can be programmed using script languages to evaluate the goodness-of-fit of a range of orders of polynomial equations, or alternative equation, forms if desired. In this instance the software finds the coefficients of a polynomial p(x) of degree n that fits the data, p(x(i)) to y(i), in a least squares sense. 13.6.3 Selection of Limiting Deformation Rate at the Point of Collapse A limiting deformation rate of 1m /day was rather ar bitrarily selected. The reas on for this assumption is that the exponential nature of the def ormation rate curves at failure makes the forecasting of collapse time relatively insensitive to the selected magnitude of the deformation rate. Other authors have determined that deformation rates at the time of collapse can vary widely. Examples are; ? 1.944m/day to 2.059 m/day at the Lusca r Mine(Cruden and Masoumzadeh,1987), ? 0.063m/day to 1.298m/day at the Li berty pit (Zavodni and Broadbent, 1980), ? 0.28m/day at the Chuquicamata pi t (Kennedy and Niermeyer, 1970), ? 0.8m/day at the Vaiont valley (Nonveiller, 1967) ? 0.1m/day at Jinchuan (Sijing, 1980) Perceptibly, a conservative approach to the selection of a limiting deformation rate is preferred and other influencing factors, such as local knowledge of previous failure velocities, quality of monitoring systems and mine management?s approach to risk management will all influence the selection of appropriate limiting rates. 13.6.4 Differentiation The application of numerical differentiation of the pol ynomial equations used for the regression curve fitting is considered trivial and can be easily programmed. Solving the differential equation for the selected limiting deformation rate is achieved using an algorithm that simply steps the equation forward in pre-selected time increments continuously checking to see when the value of the differential equation equals of exceeds the limiting deformation rate. The algorithm then converges on the solution using both forward and backward searches in smaller and smaller time increments. Once again this can be relatively easily programmed. 13.6.5 Selection of Appropriate Deformation Data Clearly forecasts of time-to-collapse are best carrie d out using deformation data associated with a rock mass experiencing continuous acceleration to failu re. One of the obvious problems using this direct method is that, on a real time basis, an element of judgement is still required to ensure that time-to- collapse forecasts are not carried out with inappropriate data, although this would to some extent be evident in the evaluation of residuals, which is a measure of the goodness-of-fit. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-9 13.7 An Example Illustrating the Use of this Forecasting Method for Stage 3 Deformation 13.7.1 Example Selection The Harmony open pit February 2006 East Wall Failure was selected as an example for illustrating the application of the method for forecasting the time of collapse (t c). Three prisms were selected for the analysis namely 422, 227B and 75. These prisms re present the middle and extremities of measured deformation behaviour leading up to the collapse. 13.7.2 Analysis and Results The data used for the forecasting commences on t he 27/10/2005, which is slig htly before the onset-of- failure point for each of the prisms. A new forecast was made at the end of every 10 new deformation measurements. A single failure example forecast us ing multiple orders of polynomial equation curve fitting for each prism can produce large quantities of data at each forecasting stage. For this reason the application of the method is best illustrated graphically using selected results. The results are shown graphically in Figures 13.3 to 13.11 and discussed in further detail below; ? Figure 13.3 Prism 422, Curve fitting for displacement magnitudes This Figure illustrates the application of the forecasting method to the available data for prism 422 using increments of 10 epochs (deformation survey measurements). It should be noted that the selected increment of 10 epochs does not ne cessarily correspond to a consistent time increment as it is possible that both the number of epochs and the time of the epochs may vary. It is therefore evident that it is better to use number of epoch increments rather than fixed time increments where extrapolation bet ween epochs would be necessary. The forecast method application in this exam ple has commenced at 90 epochs. This number has been arbitrarily selected in order to reduce the number of fitted curves shown in the Figure. As stated above, the method can however be applied continuously on and epoch by epoch basis if considered necessary. The Figure shows t he each fitted curve for each new set of epoch increments ending at the time of collapse which corresponds to 262 epochs. Each of the fitted curves is projected forward in time to the date corresponding to the deformation rate reaching its limiting value as discussed below. ? Figure 13.4 Prism 422, Displacement rate forecasts to 1m/day. This Figure illustrates the corresponding deformati on rate obtained by differentiating each of the fitted curves which are shown in Figure 13.3. T he deformation rate for each forecast stage is projected forward in pre-selected time increm ents continuously checking to see when the value of the differential equation equals of exceeds t he limiting deformation rate (1m/day). It is evident that the early forecasts are not conser vative in that they forecast collapse times up to 1 ? months later than the date at which the collapse eventually occurred. However, the forecast rapidly converges on the actual date of collapse. The forecasts from the last three sets of epochs are slightly conservative as they show slightly earlier forecast collapse date than actually occurred. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-10 ? Figure 13.5 Prism 75, Curve fitting for displacement magnitudes ? Figure 13.6 Prism 75, Displacement rate forecasts to 1m/day. Figures 13.5 and 13.6 show the results of t he forecast method applied to the deformation behaviour of prism 75. As was shown in Figur e 13.1, prism 75 represents the corresponding opposite limit in behaviour to prism 422 in that is displays the greatest delay for the onset-of- failure point, the highest rate of acceleration but lowest deformation magnitude at failure. The forecasting method for obtaining the fitted gr aphs was the same as that described for prism 422 in Figures 13.3 and 13.4 above. However, the only difference was that the method was applied commencing at 140 epochs. The reason for this was once again to reduce the number of fitted curves shown in the Figure and the fact that the onset-of-failure was delayed in this prism. The results once again illustrate a nd confirm the convergence of the forecast to the actual failure date. The forecasts of collapse times for this pa rticular prism appear to be more evenly spread between dates that are earlier or later then the actual collapse date although there does not appear to be any specifically identifiable reason for this. ? Figure 13.7 Prism 75, Time versus forecast time for date of collapse. This Figure simply plots the forecast times of collapse for each prism being analysed versus the actual time. By definition both axes of the gr aph are the same. The Fi gure clear illustrates the convergence of the forecast time for the three different prisms with the actual time of collapse. ? Figure 13.8 Prism 422, Prisms 422, Collapse forecasts on the 14/12/2005 illustrating curve fitting for different orders of polynomial. ? Figure 13.9 Prism 422, Prisms 422, Collapse fore casts on the 11/1/2005 illustrating curve fitting for different orders of polynomial. ? Figure 13.10 Prism 422, Prisms 422, Collapse forecasts on the 23/1/2006 illustrating curve fitting for different orders of polynomial. The purpose of the curves shown in Figures 13.8, 13.9 and 13.10 is to illustrate graphically the erroneous polynomial curves that can potentially be fitted to the measured deformation data and how these erroneous polynomial curves are gradually eliminated as more deformation data becomes available and the forecast progresses. Figure 13.8 shows polynomial curv es fitted to deformation data obtained up until the 14/12/2006. Only the 4 th order polynomial is a valid forecast with both the 5 th and 6th order polynomials and their derivatives being erroneous. However t he situation improves and after the 11/1/2006 (Figure 13.9) all polynomial curves fitted to def ormation data can be considered valid forecasts with only the question of accuracy still needing to be addressed. The accuracy is shown to continue to improve still further by the forecasts undertaken on the 23/1/2006 and once again no further erroneous behaviour is evident. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-11 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-12 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-13 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-14 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-15 ? Figure 13.11 Prism 422, Acceleration Forecast. Figure 13.11 illustrates an acceleration curve for a typical forecast carried out for prism 422. The acceleration characteristics were not shown in the previous Figures in order to maintain clarity. The acceleration profile is of the same form and shape as the fitted displacement magnitude curve as well as the displacement rate curve. 13.7.3 Discussion on Example The results in these Figures illustrate the potential of the developed approach to accurately predict the time of collapse. The accuracy of the prediction improves as more monitoring data becomes available, and as the time of failure is approached. In the early stages, the predicted time of collapse may be in error, and results indicate that the error is likely to be on the unconservative side. However, as more deformation data becomes available and more predictions are made with time, the results bracket and finally converge on the subsequent actual time of collapse. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-16 13.8 Forecasting of Deformation Behaviour for Stage 2 Consideration is now given as to how the Deformation Model can be used in the real time prediction of Stage 2 deformational behaviour. This stage is characterised by displacement rate decay curves in which the deformation rate of the rock mass continues to decrease until either; ? a new event sets off a new initial response and recovery deformation response, or ? the rock mass stabilises completely, at which time the creep velocities approach zero, or ? the onset-of-failure point is reached. The defining point for the onset-of-failure point in the Model is where the reducing deformation rate eventually converges on a non-zero value and the acceleration changes from negative to positive. Most of the issues discussed in Section 13.6 are applicable to using the forecasting method for Stage 2 deformation data, with the exception of the selection of the limiting deformation rate. In this instance there is clearly no way of pre-selecting non-zero deformation rates corresponding to the onset-of- failure point. 13.9 An Example Illustrating the Use of this Forecasting Method for Stage 2 Deformation In order to illustrate the prediction of the onset-of-failure point, the Harmony open pit February 2006 East Wall Failure data were again selected. However, this time an earlier (but still overlapping) data set for prism 422 was used, which included data from the 2nd June 2005 through to the 14 th December 2005. The application of the method follows the same procedure as set out in Section 13.3, and the results shown in Figure 13.12. This example clearly shows, for the deformation data analysed, how the deformation rate reduces, levels off and then starts to increase. This minimum corresponds with the onset-of -failure point as shown in the Figure. It is of interest to note that the minima of the acceleration curve and the deformation rate curve do not necessarily correspond. At the onset-of-failure point, the rate of acceleration changes over from negative to positive. It should finally be noted that the curve fitting was once again limited to a 4th order polynomial due to erroneous behaviour of the 5th and 6th order fitted polynomials. The prediction of the onset-of-failure point using this method is considered to be more accurate than the empirical approach using construction lines drawn on the deformation-time curves. This example has resulted in onset-of-failure time estimate which is approximately 1 month earlier than the value estimated empirically from the actual monitoring data (Figure 13.2). The reasons for this are that; ? the 4th order fitted curve appears to reflect fitted deformations increasing more rapidly than the actual values, which would result in slightly earlier onset-of-failure projections, ? the data set leading up to onset-of-failure extends over a long time line, resulting in a more accurate projection, ? the onset-of-failure points for the different prisms were estimated by using construction lines drawn on the deformation curves. This method although valid, is not considered as accurate as curve fitting. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-17 13.10 Further Validation Further validation work was carried out to demonstrate the applicability of the forecasting method, as set out in the previous sections, to other measurement techniques and geological terrains. The following additional examples were used; ? Mt Keith Failure #2, 2003-12-04 NE Corner 319RL (monitoring point 1) ? Letlhakane July 2005 Failure (prism S34) The Mt Keith failure was a short duration rapidly developing failure monitored over a period of approximately 3 days using radar in an ultramafic rock mass environment. The Letlhakane failure was a longer duration failure monitored over a period of a month using the Geomos Leica survey system in the Karoo Supergroup rock mass. Prisms and monitoring points selected for use in the forecasting method are considered to reflect the most representative behaviour of the failing rock mass. The individual collapse forecasts illustrating the process of curve fitting for different orders of polynomials at different time during the collapse, which were illustrated in detail for the first example, have not been shown again. The progressive failure forecasts for each of the two failures analysed above, are illustrated in the time versus forecast time graphs in Figures 13.13 and 13.14 respectively. Limiting deformation rates used were as follows; ? Mt Keith #2 Failure : 350mm/day ? Letlhakane July 2005 Failure : 1720mm/day PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-18 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-19 Discussion on Results ? Figure 13.13 : Mt Keith #2 Failure, Time versus forecast time for date of collapse. This Figure plots the forecast times of collapse for monitoring point 1 being analysed versus the actual time. Once again by definition both axes of the graph are the same. The Figure confirms the convergence of the forecast time for radar monitoring point 1 with the actual time of collapse. ? Figure 13.14 : Letlhakane July 2005 Failure, Time versus forecast time for date of collapse. This Figure plots the forecast times of collapse for prism S34 being analysed versus the actual time. Once again by definition both axes of the graph are the same. The Figure confirms the convergence of the forecast time of the prism with the actual time of collapse. 13.11 Discussion on Forecasting Method Forecasting of deformation behaviour using this new method offers several significant advantages over previous methods presented in literature. These are; 1) The potential exists for these forecasts to be carried out in a real time interactive environment, thereby enhancing its value and providing a practical tool for forecasting failures. 2) This forecasting method is relatively easily automated and programmed. The Author used an existing commercial mathematical software (MATLAB) and developed simple scripts which are automated series of commands stored in files, for processing input data as well as carrying out and storing results of the all the repetitive computations. 3) It is anticipated that this method can be relatively easily integrated into software running existing monitoring systems such as prism surveys and radar. 4) Unlike previous semi-empirical forecasti ng methods, no specific knowledge of the onset-of- failure point is required 5) No prior knowledge of the failure characteristics of a given rock mass is required. 6) The method is considered more accurate in forecasting the onset-of-failure point than empirical approaches which use construction lines drawn on deformation-time curves. 7) The results of the examples analysed illustrate and confirm the convergence of the forecasts to the actual failure date as the time of the predicted event is approached. 8) In predicting the time-to-collapse, the exponential nature of the deformation rate curves at collapse makes the forecasting of the collapse time relatively insensitive to the selected magnitude of the limiting deformation rate. This removes the onus on trying to accurately determine the deformation rate at collapse. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-20 9) Similarly, the defining point for the onset-of-failure point in the Model is where the reducing deformation rate (or deformation rate decay curve) eventually converges on a non-zero value and the acceleration changes from negative to positive. Once again this removes the onus on trying to accurately determine deformation rates when applying this forecasting method. There are nevertheless a few aspects which do affect the accuracy of the forecasts and that warrant further discussion; a) Erroneous polynomial behaviour Erroneous polynomial forecasting behaviour was shown to occur especially on early forecasts, where data sets are small, and there are insufficient deformation data to establish a suitable continuous acceleration profile (as was shown in Figure 13.8). As intuitively expected, erroneous polynomial forecasting behaviour eventually disappears on later forecasts which use progressively larger data sets, as shown in Figure 13.9 and 13.10 where all the higher order polynomials reflect suitable acceleration profiles. It is however, a relatively simple matter to identify and flag erroneous polynomial behaviour programmatically simply by testing forecasts for negative displacement rates. b) Goodness of fit and Error Bounds The forecasting method is essentially the application of a "best fit" logarithm of a polynomial (in the least squares sense) of a specified order for a given set of deformation data. This best fit log polynomial is then used to predict future failure based on a predetermined limiting deformation rate ie the derivative of the log polynomial function. In statistics the method of evaluating the "goodness of fit? of each curve (or prediction) is by way of evaluating the residuals, which is the simply the difference between the observed and predicted data. An evaluation of the suitability of a particular order of log polynomials can be made by simply comparing the residuals for the various fits. In the examples presented above this was carried out for each individual curve fit by simply inspecting the residual plots. This process is illustrated by comparing a 4th order and a 7th order log polynomial fit for a specific epoch using the Mt Keith #2 failure example as illustrated in Figures 13.15a and b and 13.16a and b. Clearly the 7th order log polynomial is a superior fit as compared with the 4th order fit. Error bounds are also useful for determining if the deformation data has been reasonably modeled by the curve fit. An optional second output parameter can be obtained from the polyfit function in MATLAB and passed as an input parameter to polyval function in MATLAB in order to obtain the error bounds. An example of the error bounds for the 4th order curve fit shown in Figure 13.15a is shown in Figure 13.15c and the error bounds for the 7th order curve fit shown in Figure 13.16a is shown in Figure 13.16c. This particular example uses an interval of ? ?2, corresponding to a 95% confidence interval. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-21 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-22 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-23 c) Convergence of forecasts to actual date It was found that only the lower order of polynomials are generally applicable for forecasts based on small data sets (as is shown in Figure 13.8) due to erroneous or invalid polynomial behaviour discussed above. These lower order polynomials do, as anticipated, tend to produce estimated times of collapse (tc) which vary unconservatively from the actual time of collapse. However, as progressively larger data sets become available for curve fitting, the accuracy of the forecasts increases rapidly, finally converging on the actual time of collapse. Generally speaking, in the example used, the higher order profiles tend to produce the more accurate tc forecasts, although this may not always be the case. d) Change in deformational behaviour of the instability itself. It should also be noted that the progressive deformational behaviour of the instability itself may change due to transient factors, which may also influence the eventual accuracy of the forecast. With reference to Figure 13.17, it is evident that midway through the progressive acceleration, on the 14/1/2006 there is a slight reduction in the rate of deformation. This has resulted in the curve fitting forecasts up until this date, converging on an earlier tc, which obviously influences the overall accuracy of the forecasts. e) Practioner?s skill. Clearly the practioner?s skill and knowledge will undoubtedly have an influence on the accuracy of the forecasts made. However the influence of individual practioner?s skill levels will diminish once the forecasting method has been fully automated (programmed). f) Influence of Transient Factors on Results. The forecasting method uses the deformation data as measured by either the radar or the conventional prism methods. At this stage no correction has been included in the forecasting method for the transient factors which affect the accuracy of the deformation survey measurements themselves. 13.12 Future Development This section of the thesis has demonstrated the application of the forecasting method developed to deformation data sets collected in the case study database. The method when applied using manual calculations is clearly very labour intensive and consequently of limited practical use in a real time mining environment. It is therefore recommended that the forecasting method be fully programmed and automated and fully tested using examples of failures in as many different geological terrains as practicable. Furthermore future application of the method will require a more detailed study of applicable limiting deformation rates. Finally, it is anticipated that implementation, ongoing testing and daily usage of the forecasting method in different real time mining environments will result in continual improvements to the forecasting method providing it is fully supported. Confidence in the forecasting method is therefore likely to evolve gradually and incrementally rather than be established at any particular point in time. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 13 : Forecasting of Deformation Behaviour using the Model Page 13-24 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 14 : Conclusions Page 14-1 14 CONCLUSIONS The achievements of this research are significant; Firstly, the collection of twelve international detailed mining case studies from Southern, Western and Eastern Africa together with Western Australia and South America, in itself, represents a very comprehensive, independently assembled large open pit mining database. A significant amount of the geological and structural data collected, has been modelled using 3D visualisation software (Vulcan) which has enabled a more accurate understanding of the orientations and interaction of larger structures within the rock. In addition, the deformation monitoring data which has been collected has included databases from the most up-to-date automated survey systems as well as slope stability radar and has even included video footage of failures. A considerable volume of further supporting data has also been collected, including regular internal mining reports, consultant?s reports, environmental data such as rainfall and groundwater monitoring, monthly pit surveys and geotechnical parameters. Secondly, a comprehensive new Time and Event Dependent Deformation Model has been developed which describes how deformation behaviour of excavated rock mass may be presented using deformation pathways. The model accommodates five principal stages of deformation ranging from primary and secondary rock mass creep modes through the onset-of-failure to collapse and post collapse or post mining recovery deformation behaviour. A very significant feature of the Model is the provision of changing deformation rate decay functions as a slope progresses towards failure. Thirdly, a new statistical method has been developed to use in conjunction with the Model to enable forecasting of deformation behaviour in order to make predictions such as the time to collapse. The new method offers many advantages over existing empirical and semi-empirical methods. These include, amongst others, the potential to undertake forecasting in a real time interactive environment, the method is easily automated and programmed and has the potential to be integrated into software running existing monitoring systems. Further advantages are, that no specific knowledge of the onset- of-failure point is required in order to make collapse predictions, the method is relatively insensitive to the selection of the limiting deformation rate and no prior knowledge of the failure characteristics of a given rock mass is required, in order to apply the forecasting method. This Section provides a concise summary of the most important findings for each of the principal sections of the research. 14.1 General review of Time Dependent Deformation The literature review showed that time dependent deformation of geological features has been widely recorded and documented in the past. Descriptions of deformation have ranged from large scale valley rebound and up warping of valley rims due to rapid erosion or glacier melting, deep seated continuous creep deformation of entire mountain sides, observed in the European Alps, rebound of excavations and foundations for large structures such as dams, to numerous small scale landslides and slope instabilities found all over the world. No two failure behaviours are the same and the characteristics of instabilities are highly site specific. The reason for this is that deformation and failure in fresh or hard rock slopes, are controlled PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 14 : Conclusions Page 14-2 principally by structural discontinuities such as bedding, joints, geological contacts, faults etc, which are themselves site specific. Failure seldom takes place through intact rock unless it is very soft or forms a rock bridge between discontinuities. Other important factors influencing deformation and failure are the rock lithologies and associated geomechanical properties, ground water levels, slope height, slope angle and blasting damage. It may be concluded that the majority of rock failures are, however, structurally controlled. Jointed rock masses consist of both time-dependent as well as time-independent resistance to deformation. It is important to consider the relevance of the ?time factor? in equating the potential reduction in the rock or joint strength within the time frame of the required active mining life of the slope. In this regard, the deformation of intact hard rock occurs over such a lengthy time scale in comparison to that of the limited required life of the slope that it can potentially be considered as insignificant. In contrast, the time dependent behaviour of discontinuities is considered significant where the potential creep deformation time leading to failure can be well within the required length of slope life. It has also been shown in studies on the time dependent deformation behaviour of discontinuities that the magnitude of the peak and residual shear strengths of the joints are themselves time related. The literature review identified one important previous research effort to characterise time dependent deformation behaviour, which was undertaken by Martin (1993). He was the first to illustrate the relationship between the accumulated displacement magnitude patterns and the associated displacement rate patterns. This work showed that, associated with the accumulating displacements, there was a corresponding pattern of deformation rate adjustments responding to changes in the global and local stress fields as a result of excavation. This pattern essentially consisted of a rapid increase (spike) in deformation followed by a period of gradual decrease in deformation rates which conformed to a negative exponential relationship (also referred to as strain hardening). However, a significant deficiency in Martin?s description of deformation was that he assumed that the decrease in deformation rates remained constant until failure. 14.2 Literature Review Case Study Findings The literature review case studies confirmed significant similarities in the characteristics of rock mass deformation in slopes. It was shown that specific events, especially mining (blasting) and high rainfall are responsible for triggering a deformational response of a slope. The initial part of the deformational response involves a sudden and rapid increase in the deformation rate which is also termed the initial response. Following this there is a period of relatively rapid reduction or decay in the deformation rate of the rock mass, which is in turn followed by a long period of slowly reducing steady state creep. This is termed the rock mass decay function and the overall behaviour is described as regressive. Importantly, it was identified that the closer to failure the slope becomes, the more the deformational behaviour of the decay functions change. The decay functions start to take an increasingly longer time to recover and eventually a point in time is reached where the deformation rates of the unstable rock mass starts to increase in an exponential manner, until collapse occurs. This is termed progressive behaviour. Although differing in rates and magnitudes, the pattern of behaviour as described above tends to be reflected in slopes displaying completely different failure mechanisms. Post collapse deformation can range from complete disintegration of the rock mass to complex alternating progressive, and/or regressive behaviour, provided that the collapsed rock mass has maintained its integrity. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 14 : Conclusions Page 14-3 14.3 Detailed Case Study Findings One of the biggest challenges in this research was to obtain good quality deformational data. In most instances, not much was known about either the quality of the monitoring systems or the nature and types of failures experienced until they were reviewed on the mine site. The data collection phase of the project showed the importance of ensuring that the increasingly sophisticated monitoring technology being deployed be administered by experienced staff. The six detailed case studies selected for inclusion in the thesis represent the best combination of accurate deformation monitoring data associated with relatively large failures that were able to be collected within the constraints of time and resources. Although most of the other detailed case studies not included in Section 5 to 10, had reasonably good deformation monitoring data, there were unfortunately no major failures (Potgietersrust Platinum, pre- underground Palabora, Geita, Kalgold and Kalgoorlie Superpit). Deformational data of the major failures from both Navachab and Bibiani were obtained at the time older manual survey systems were being used, although automated survey implementation has since been implemented at Navachab and was used to evaluate the high deformation rates experienced on the east wall. The Bibiani open pit has since been closed prematurely after the ramp failures. The deployment of radar systems in the Orapa, Letlhakane and Venetia open pits has occurred within the last 12 months. There have been no significant failures during this period, nevertheless information on smaller instabilities was reviewed. Data from the detailed case studies have shown that there exist two significantly different categories of pre-onset-of-failure deformational behaviour characteristics of rock slopes. The first category, (Category 1), can be described as large scale macro event driven deformation behaviour and the second, (Category 2), as deformational behaviour that is not specifically event driven or influenced. In Category 2 deformation, the onset of failure can occur with little or no prior warning and can often be attributed one or more factors such as micro events, time dependent decay of the static coefficient of friction and stiffness of joints and failure surface propagation. Large scale macro event driven deformation behaviour of rock slopes displaying Category 1 pre-onset- of-failure deformational behaviour, was closely studied using the Navachab east slope, Harmony pit east wall failure, Venetia July 2003 failure and the Orapa instability. These examples, in specific, confirmed how deformational response behaviour changes in response to mining events and as the slope approaches collapse. The Navachab 2001 failure, all the Bibiani west wall failures, Mt Keith fresh rock failures, Harmony west wall failures and Letlhakane July 2005 failure ar e examples of rapidly developing fresh rock failures displaying Category 2 pre-onset-of-failure deformational behaviour. In these failures the potential instability/failure mass is bounded by well defined failure planes or discontinuity boundaries (joints, shears or bedding). The excellent Mt Keith radar data enabled these types of failures to be studied in detail. The most important characteristic of these failures is that they generally do not reflect a continuous period of specific macro event (mining) influenced deformation behaviour prior to the onset of failure. In reality the onset of failure can occur with little or no prior warning. A further characteristic of these failures is that they are often characterised by easily detectable propagation of instability/failure through the rock mass. Despite occasional problems with monitoring data quality, the detailed case studies have provided additional confirmation of earlier observations, that there are many similarities in the deformation PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 14 : Conclusions Page 14-4 behaviour patterns of rock slopes, despite the considerable differences in the rock mass making up the slopes and the eventual failure mechanism. An important finding of this research is therefore, that the pre-collapse deformation rate behaviour is largely independent of the slope failure mechanism and mode. Therefore, the real potential for prediction is based on changing deformation rate patterns rather than identifying modes of instability or making comparisons of instantaneous deformation rates. 14.4 Development and Application of the Time and Event Dependent Deformation Model The finding of the independence of the deformation rate behaviour to slope failure mechanism facilitated the development of the Time and Event Dependent Deformation Model. The Model illustrates how differing deformation behaviour of the excavated rock mass may be described using deformation pathways. The deformation pathway for the point is plotted using a deformation time relationship where the deformation can be expressed in terms of horizontal or vertical displacements or displacement rates. The model describes five principal stages of deformation, ranging from primary and secondary rock mass creep modes through the onset-of-failure to collapse, and post collapse deformation behaviour. The model is sufficiently robust to enable deformation patterns resulting from different failure mechanisms to be accommodated. Key features of the new Model are; 1) The provision of two different stages of event induced deformation behaviour prior to the onset- of-failure. The first stage encompasses relatively uniform deformation rate patterns and the second stage encompasses changing deformation rate patterns as a result of changing deformation rate recovery functions. 2) A description of deformation behaviour which is largely independent of the resulting slope failure mechanism. 3) Description of the inter-relationship between vertical and horizontal deformation behaviour 4) Description of post collapse and/or post mining/recovery behaviour. 14.5 Forecasting of Deformation Behaviour A statistical based method has been developed, which enables forecasting of deformation behaviour to be carried out at any time in which the rock mass is demonstrating Stages 1 to 3 deformation behaviour. The new method offers many significant new advantages over existing empirical and semi- empirical methods found in literature. These include, 1) The potential for these forecasts to be carried out in a real time interactive environment, thereby enhancing its value and providing a practical tool for forecasting failures. 2) This forecasting method is relatively easily automated and programmed. The Author used an existing commercial mathematical software (MATLAB) and developed scripts which are automated series of commands stored in files, for processing input data as well as carrying out and storing results of the all the repetitive computations. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 14 : Conclusions Page 14-5 3) It is anticipated that this method can be relatively easily integrated into software running existing monitoring systems such as prism surveys and radar. 4) Unlike previous semi-empirical forecasti ng methods, no specific knowledge of the onset-of- failure point is required 5) No prior knowledge of the failure characteristics of a given rock mass is required. 6) The method is considered more accurate in forecasting the onset-of-failure point than empirical approaches which use construction lines drawn on deformation-time curves. 7) The results of the examples analysed illustrate and confirm the convergence of the forecasts to the actual failure date as the time of the predicted event is approached. 8) In predicting the time-to-collapse, the exponential nature of the deformation rate curves at collapse makes the forecasting of the collapse time relatively insensitive to the selected magnitude of the limiting deformation rate. This removes the onus on trying to accurately determine the deformation rate at collapse. 9) Similarly, the defining point for the onset-of-failure point in the Model is where the reducing deformation rate (or deformation rate decay curve) eventually converges on a non-zero value and the acceleration changes from negative to positive. Once again this removes the onus of trying to accurately determine deformation rates when applying this forecasting method. There are nevertheless a few minor aspects which do affect the accuracy of the forecasts which need to be considered. These include, erroneous polynomial behaviour, speed of convergence of forecasts to actual date, change in deformational behaviour of the instability itself and practioner?s skill. 14.6 Concluding Comments Deformational behaviour of failures can be extremely complicated and up to now there have been no methods or models that have adequately addressed the range of behaviour that is possible during excavation of rock slopes in different geological environments. This research project has significantly expanded the frontiers of the knowledge of time and event dependent deformational behaviour of rock slopes and provided a Model for both the interpretation and forecasting of deformational behaviour. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 15 : Recommendations Page 15-1 15 RECOMMENDATIONS FOR FURTHER RESEARCH In order to collect the data necessary to develop the model further, there are two different approaches suggested which a future research effort can adopt. The first is an extension of the approach adopted by the Author of the current research which was based around existing mining operations and the second is an implementation of a greenfields research site/s. These approaches are discussed further below; 1. The first is to select case study mining operations which have comprehensive monitoring systems and relatively frequent failures and visit and work with the mining operations over a period of possibly one to two years studying and observing deformation behaviour first hand. In this time the researcher should participate in overseeing the collection and interpretation of very precise data as well as continually reviewing further development of monitoring systems as the pit geometry and circumstances change. Deformation measuring instruments should include radar, automated prism survey networks and inclinometers and possibly laser scanning. Additionally, the researcher should record the location and time of all mining related events (such as blasting and excavation) as well as environmental events (seismic, rainfall, groundwater level, etc). Further information collected should include regular (monthly) pit surveys. The development of a geological model using 3D visualization software is considered a necessity in order to study the interaction of structures on deformation and failures. 2. The most ideal research project would be a greenfields site on a relatively small open pit with a limited life of 2 to 3 years where monitoring instrumentation devices such as prisms (initially around the open pit), slope inclinometers, piezometers and borehole extensometers can be installed in advance of the excavation of the planned open pit. The instrumentation can then be monitored before, during and for one to two years after mining operations are complete. Finally, a comparison of the deformation behaviour for several such research projects of this type would be ideal with each project having different geological environments and involving relatively uniform but different rock mass conditions (rock mass quality, structure and strength) in which the effects of groundwater could be controlled or were reasonably well defined Once this information has been obtained the comprehensive deformation data obtained could then be used to further refine the deformation pathways and recovery functions in the time dependent deformation model. In addition they can be used to calibrate numerical models and to assess the various constraints and controls on the time dependent deformation process. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 16 : References Page 1 of 7 16 REFERENCES Adams, R. (2006). ?Tactics versus Strategy : The Mount Keith Operations Experience 2003 to 2005?. Strategic versus Tactical Approaches in Mining 2006, Section 28, pg1-11. Amadei, B. and Curran, JH. (1980) ?Creep Behaviour of Rock Joints?. Proceedings Thirteenth Canadian Rock Mechanics Symposium, Toronto, Canadian Institute of Mining and Metallurgy. CIM Special Volume 22 pp. 146-150. Australian Mining Consultants (1997). ?Bibiani Mine Project, Feasibility Study, Volume 2 ? Geology and Mining?, Unpublished report. Barnett, W. P. (2003) ?Geological Control on Slope Failure Mechanisms in the Open Pit at Venetia Mine?. SA Journal of Geology, Vol. 106, pg 149-164. Basson, I.J. (2005) ?Country Rock Mapping of t he Venetia Open Pit?. TECT Geological Consutling. Unpublished report. Basson, I.J. (2006) ?3D Modelling of the Country Rocks Hosting the Orapa 2125A/K1 Kimberlite?. TECT Geological Consutling. Unpublished report. Basson, I.J. and Barnett, W.P. (2205) ?Structural Analysis Of Host Rocks To The Venetia Kimberlite Pipes, Central Zone, Limpopo Belt, South Africa? Tect Geological Consulting. Unpublished presentation. Bibiani Geotec Dept (2001 to 2005). Monthly pit inspection reports from 2001 to 2005. Unpublished reports. Bieniawski, Z.T. (1970) ?Time-dependent Behaviour of Fractured Rock?. Rock Mechanics, 2, pp. 123- 137. Blackwell, G., Pow, D. and Keast, M. (1975) ?Slope monitoring at Brenda Mine?. Proceedings, Tenth Canadian Rock Mechanics Symposium, Kingston, Canada, pp. 45-79. Brawner, C.O. and Stacey, P.F. (1979) ?Hogarth Pit Slope Failure, Ontario, Canada?. In Rockslides and Avalanches (Voight ed.), Elsevier, Amsterdam, pp. 691-707. Brawner, C.O., Stacey, P.F. and Stark, R. (1975) ?Monitoring of the Hogarth Pit Highwall, Steep Rock Mine, Atikokan, Ontario?. Canadian Rock Mechanics Symposium, Ottawa, Canada, Vol. 10/2 pp. 102- 138. Brawner, C.O., Stacey, P.F. and Stark, R. (1976) ?A successful application of mining with pitwall movement?. Open Pit Mining Technology CIM Bulletin (April), pp. 68-75. Broadbent, C. D. and Ko, K. C. (1971). "Rheological As pects of Rock Slope Failures." Cording, E.J. ed. Stability of Rock Slopes. Proceedings of 13th US Symposium on Rock Mechanics, Univ. of Illinois, Urbana, Ill. ASCE, pp. 573-593. Broadbent, C.D. and Zavodni, Z.M. (1981) ?Influence of Rock Structure on Stability?. Third International Conference on Stability in Surface Mining, (Brawner ed.) Vancouver, Canada, Society of Mining Engineers of AIME. Vol. 1 pp. 7-18. Buhmann, D., and M. Atanasova. (1997). ?Karoo Core Samples from Orapa and Letlhakane, Botswana: A mineralogical report?. Council of Geoscience, South African, Report no.:1997-0347. Unpublished. Bush, R.A., Hofmann, R.H., and van Rensburg, H. J. (1995). ?Dewatering Plan for Orapa Mine A/K1 Pit.? Unpublished Anglo American Corporation Civil Engineering Report. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 16 : References Page 2 of 7 Cahill, J. (2006). Personal communication. Cahill, J. and Lee, M. (2006). ?Ground Control at Leinster Nickel Operations?. SAIMM, International Symposium on Stability of Rock Slopes in Open Pit Mining and Civil Engineering Situations, S44, Cape Town, April 2006. pp 579 - 594. Cairncross, B. and Dixon, R, (1995). ?Minerals of South Africa?. Published by the Geological Society of South Africa. pp 45-46. Calder, P.N. and Blackwell, G. (1980) ?Investigation of a complex rock slope displacement at Brenda Mines?. CIM Bulletin, August, 1980 pp. 73-82. Call R. (1981) ?Monitoring Pit Slope Behaviour?, 3rd International Conference on Stability in Surface Mining, Vancouver, Canada. Cristescu, N.D. and Hunsche, U. (1998). ?Time Efffect s in Rock Mechanics?. Wiley Series in Materials, Modelling and Computation. John Wiley & Sons Colley, C. (2004) Leinster Nickel Operations- Geomechanics Section, Memorandum dated 17/8/2004, Unpublished. Cruden, D.M. and Masoumzadeh, S. (1987) ?Accelerating Creep of the Slopes of a Coal Mine?. Rock Mechanics and Rock Engineering, 20 pp. 123-135. Danckert, G.M. (1994) ?Managing the risk of failure in the Pit lA Highwall at Telfer Gold Mine?, Telfer, W A. Proceedings, Fourth Large Open Pit Mining Conference, Perth, Australia, AusIMM. Vol. 1 pp. 217-223. Deere, D.U., Hendron, AJ., Patton, F.D. and Cord ing, E.J. (1967) ?Design of Surface and Near Surface Construction in Rock?. In Failure and Breakage of Rock, Proceedings, Eighth US Symposium on Rock Mechanics, (Fairhurst ed.), New York, pp 237-302 Dieterich, J.H. (1972) ?Time-Dependent Friction in Rocks?. Journal of the Geophysical Research, Vol. 77, No. 20, pp 3771-3781. Dieterich, J.H. (1978) ?Time-Dependent Friction and the Mechanics of Stick-Slip?. Pageophy., Vol. 116, pp 790-805. Dight, P. (2006).?Pit Wall Failures on ?Unknown? Structures?. SAIMM, International Symposium on Stability of Rock Slopes in Open Pit Mining and Civil Engineering Situations, S44, Cape Town, April 2006. pp 11 to 29. Dirks, P.H.G.M. (2001) ?An Investigation Of The Main Failure Planes At Letlhakane Mine, Botswana? SRK Zimbabwe. Project No 506TF. Unpublished report. Dirks, P.H.G.M.(2003) ?An Investigation Of Major Structural Features With Relevance To Pit Stability At Orapa Mine, Botswana? SRK Consulting. Report No 125279. Unpublished report. Dodds, R.K. (1966). "Rock Movement Along Fractu res During Failure." Proceedings of 1st Congress of the International Society for Rock Mechanics. Lisbon. Paper 6.1, Vol. II, pp. 133-137. Feld. J. (1966). "Rock Movements from Load Re lease in Excavated Cuts." Proceedings of 1 st Int. Congress of the International Society for Rock Mechanics. Lisbon. Paper 6.2. Vol. II. pp. 139-140. Ferguson. H.F. (1967). "Valley Stress Relief in the Allegheny Plateau." Engineering Geology. Vol 4. No.1. pp. 63-71. Ferguson, H.F. and Hamel. J.V. (1981). "Valley Stre ss Relief in Flat Lying Sedimentary Rocks." Proceedings of the International Symposium on Weak Rock. Tokyo. September. pp. 1235-1240. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 16 : References Page 3 of 7 Franklin. J.A. and Dusseault, M.B. (1989)."Rock Engineering." McGraw- Hill. Toronto. 600p. Geology.com (2006) INTERNET. http://geology.com/dictionary Cited November 2006. Gibson, J. (1998) ?Brief History of Discovery, Sampling and Mining of Orapa and Letlhakane Mines?. Debswana internal note. Unpublished. Gillon. M.D. and Hancox. G.T. (1992). "Cromwell Gorge Landslides, A General Overview." Bell, D.H. ed. Proceedings of the 6th International Symposium on Landslides. Christchurch. New Zealand. February. Vol. I, pp. 83-102. Goel, S.C. (2003) ?Orientated Geotech Holes from 122RL for Kinematic Stability Analyses?. Ashanti Goldfields, 28 August 2003. Unpublished internal report. Goldberg, W.C. and Frizzell, E.M. (1989) ?Pit Slope Monitoring and Back Analysis of the Berkeley Pit? Journal of Mining Engineering. pp.1191-1195 Goodchild, D. (2006). Personal communication. Goodman. R.E. and Bray J.W. (1976) "Toppling of Rock Slopes." Rock Engineering for Foundations and Slopes, Proceedings of the ASCE Specialty Conference, Boulder, Colorado, August, 1975. Vol. II. pp. 201-234. Goodman, R.E. (1989). ?Introduction to Rock Mechanics?, Second Edition. John Wiley & Sons. ISBN 0-471-81200-5. Hanna A. and Little, T.E. 1992. "An Estimate of the Rebound Potential of the Shaftsbury Shales at a Damsite in British Columbia.". Canadaian Geotec hnical Journal. June, Vol. 29, No.3, pp. 375-392. Hannweg, L. (2006) Personal communication. Haywood,N. (2004). ?Structural Geology of the Mt Keith Area?. BFP Consultants. Unpublished report. Heim, A. (1932) ?Bergsturz und menschenleben.? (Landslides and Human Lives) Skermer, N.A. trans. 1989, Bitech Publications, Vancouver, British Columbia, 195p. Hewett, F.G.(1978) ?The Cassiar Story?. CIM Bulletin, 71 (792), pp. 56-63. Hoek, E. (1970). ?Estimating the Stability of Excavated Slopes in Opencast Mines?. Trans. Instn. Min. Metall. (Sect. A:Min. industry), 79, ppA109-A132. Hoek E. (1999). "Putting Numbers to Geology ? an engineer?s viewpoint?. Quarterly Journal of Engineering Geology, 32, 1-19. Hoek, E and Bray, JW (1981). ?Rock Slope Engineering? revised Third Edition?. Institution of Mining and Metallurgy, London. Hollingsworth, SE., Howard-Taylor, J and Kellaway, GA. (1943) ?Large Scale Superficial Structures in the Northhampton Ironstone Field.? Quarterly Journal of the Geological Society of London, Ser C, 397- 98, pp 1-44. Hutchinson, J. N. 1968. "General Report: Morphol ogical and Geotechnical Parameters of Landslides in Relation to Geology and Hydrogeology." Bonnar d, C. ed. Proceedings of 5th International Conference on Landslides. Luasanne, Switzerland. July, pp. 335. Itasca Consulting Group Inc. (2001). "FLAC: Fast Lagrangian Analysis of Continua." Users Manual, Version 4.0. Minneapolis, Minnesota. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 16 : References Page 4 of 7 Jakubec, J., Terbrugge, P.J., Guest, A.R. and Rams den, F. (2000). ?Pit Slope Design at Orapa Mine?. Proceedings, Slope Stability in Surface Mining, (Hustrulid, McCarter and vanZyl eds.) SME., pp. 227- 238. Janse van Rensburg, J.(2006) Personal communication. Jennings, J. E.(1968) ?'A preliminary theory for the stability of rock slopes based on wedge theory and using results of joint surveys,? Univ. of the Wi twatersrand, Internal Report (unpublished), 1968. Jennings, J. E. and Steffen, O. K. H. (1967) ?The analysis of the stability of slopes in deep opencast mines,? Trans. S. Afr. Inst. Civil Eng., Vol. 9, No.3, Mar. 1967, pp. 41-54. Jennings, J.E. and Steffen. O.K.H. (1979). "Monito ring of Deformations Associated with Excavation: Monitoring of Deformations with Special Reference to Open Excavations." Proceedings of 4th Int. Congress. of the International Society for Rock Mechanics. Montreux, Switzerland. Vol. III, pp. 383- 395. John, K. W. (1962) ?An Approach to Rock Mechani cs?, Journal Soil Mech. and Found. Div., ASCE, Vol. 88, No. SM4, Aug. 1962, pp. I-3D. Johnson, R.S. (1982) Slope stability monitoring. Proceedings, Fourth Canadian Symposium on Mine Surveying and Deformation Measurements., Banff, Canada, Vol. 1 pp.363-379. Kaiser, P.K. and Morgenstern, N.R. (1979). "T ime-dependent Deformation of Jointed Rock Near Failure." Proceedings of 4th Int. Congress of the In ternational Society for Rock Mechanics, Montreux, Switzerland. Vol. II, pp. 195-202. Kaiser, P.K. and Morgenstern, N.R. (1981) "P henomenological Model for Rock with Time-Dependent Strength." International Journal of Rock Mechanic s and Mining Science and Geomechanics Abstracts. Pergamon, Vol. 18, pp. 153-165. Kayesa, G. (2005) ?Slope Failure Analysis Report? Ref No LFR03/05, July 2005. Debswana - Orapa and Lelhakane Mines, Unpublished report. Kennedy, B.A. and Niermeyer, K.E. (1970) "Slope Mo nitoring Systems Used in Prediction of a Major Slope Failure at the Chuquicamata Mine, Chile." v an Rensburg, P.W.J. ed. Planning Open Pit Mines. Proceedings of the Symposium on the Theoretical Background to the Planning of Open Pit Mines with Special Reference to Slope Stability. Johannesb urg, South Africa. Balkema, pp. 215-225. Keyter, G. (2003) ?K1 South Face Failure 2nd July 2003? DeBeers Internal Memorandum. Unpublished. Lapworth, H. (1911). ?The Geology of Dam Trenches?. Transactions Institution of Water Engineers. London. Vol IIA, pp 336-341. Letlhakane Geotec Department (2006) Geomos prism survey database. Unpublished data. Leinster Nickel Operations Geotec Dept (2006) Weekly Mining Reports, Unpublished Marshall, N. (1997) ?Open Pit Geotechnical Investigation Report?. Bibiani Goldfields Limited. Unpublished report. Marshall, N. (2004) ?Bibiani Open Pit ? Portal Area Slope Failure ? Final Report?. SRK Consulting, Report No. U2618. SRK(UK) Ltd. September 2004. Unpublished report. Marshall, N. and Mercer, K. (2006) ?Bibiani Open Pit ? Cut Back Design. Geotechnical Review of Pit Slope Stability Report Prepared for: Anglogold Ashanti (Bibiani) Ltd.?, SRK Consulting, January 2006, Unpublished report. Martin,C.D. and Kaiser, P.K. (1984) Analysis of a rock slope with internal dilation. Canadian Geotechnical Journal, 21 pp. 605-620. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 16 : References Page 5 of 7 Martin, D.C. (1993) ?Time Dependent Deformation of Rock Slopes?. PhD, University of London. Martin, D.C. and Mehr, E.F. (1993) "Assessment of Slope Deformation and Deep Seated Instability in the Cassiar Open Pit." Canadian Institution of Mini ng and Metallurgy Bulletin. July, Vol. 86, No.972, pp 58-66. Matheson, D.S. and Thomson, S. (1973) "Geolog ical Implications of Valley Rebound." Canadian Journal of Earth Sciences Vol. 10, pp. 961-970. Matlab (2000) Matlab Version 6, Release 12, Copyright 1984-2000 The Mathworks, Inc Minproc Engineers Limited (1995) Bibiani Goldfields Limited, Bibiani Gold Project, Feasibility Study, CAN : 008992694, July 1995. Unpublished report. Mining One. (2003) ?Geotechnical Status Report, Mt Keith Nickel Operations (MKO) for BHPBilliton?. Unpublished report. MKO (Mt Keith Operations) Geotechnical Department (2006). Unpublished geological and monitoring data M?ller, L. (1968) "New Considerations on the Vaj ont Slide? Rock Mechanics and Engineering Geology 6, pp. 1-91. M?ller, L. (1974). ?Rock Mass Behaviour ? Determ ination and Application in Engineering Practice?. Proc. 3rd Int. Cong. International Society of Rock Mechanics, Denver, Vol 1, pp 205-215. Mulville, D. (2006) Personal communication Naidoo, R.(2004) ?K1 South Face Failure ? 11 September 2004? DeBeers Internal Memorandum. Unpublished. Naismith, W.A. and Wessels, S.D.N (2005). ?Management of a Major Slope Failure at Nchanga Open Pit, Chingola, Zambia?. The Journal of the South Af rican Institute of Mining and Metallurgy, Vol. 105, pp 619 to 626. Naismith, W.A (2006) Personal communication. Nemcok, A. (1972) "Gravitational Slope Deformation in High Mountains." 24th IGC, Montreal, Section 13, pp. 132-141. Nonveiller, E. (1967) ? Zur Frage der Felsrutschung im Vajont-Tal?, Rock Mechanics and Engineering Geology 5, pp 2-9. Orapa Geotec Department (2000) ?Update on t he A/K1 Failure? Unpublished presentation. Orapa Geotec Department (2005) ?Cut 2 Ramp Access Failure on 23rd December 2004 ? Presentation to General Manager? Unpublished presentation. Orapa Geotec Department (2006) Geomos prism survey database. Unpublished data. Patton, F.D. (1970) "Significant Geologic Factors in Rock Slope Stability?. Planning Open Pit Mines, van Rensburg, ed. Proceedings of the Symposium on the Theoretical Background to the Open Pit Mines with Special Reference to Slope Stability. J ohannesburg, South Africa. Balkema, pp. 143-151. Piteau, D. R. (1970b). "Geological Factors Significant to the Stability of Slopes Cut in Rock." Planning Open Pit Mines, van Rensburg, P.W.J. ed. Pr oceedings of the Symposium on the Theoretical Background to the Planning of Open Pit Mines with Special Reference to Slope Stability. Johannesburg, South Afri ca.Balkema, pp. 33-54. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 16 : References Page 6 of 7 Piteau, D.R. and Martin, D.C. (1977) "Slope Stab ility Analysis and Design Based on Probability Techniques at Cassiar Mine". CIM Bulletin. Vol. 70, No. 779 pp. 139-150, March. Piteau,D.R. and Martin, D.C. (1982) "Mechani cs of Slope Failure." Proceedings of the Third International Conference on Stability in Surface Mining, Brawner, C.O. ed. AlME. Vancouver. September, 1981, pp. 113-169. Radbruch-Hall, D.H. 1978. "Gravitational Creep of Rock Masses." Voight, B. ed. Rockslides and Avalanches, 1. Developments in Geotechnical Engineering, 14A. Elsevier, pp. 607-657. Ramsden, F. (2006). Personal communication Reeves, C.V.(1978) ?Reconnaissance Aeromagnetic Survey of Botswana 1975 ? 1977?. Final Interpretation Report, Geological Survey, Botswana and C.I.D.A. Read, H.H and Watson, J. (1970) ?Introdu ction to Geology? Vol. 1 Principles, 2nd Edition, MacMillian and Co., London Reid, G. and Stewart, D. (1986) "A Large Scale Toppli ng Failure at Afton." Singhal R. ed. International Symposium on Geotechnical Stability in Surface Mining, Calgary. November, pp. 215-223. Roux, R. (2005) Personal communication, mine monitoring and geological data. Roux, R. (2006) Personal communication Roux, R. and Terbrugge, P. (2006).? Slope Management at Navachab Gold Mine, Namibia?. SAIMM, International Symposium on Stability of Rock Slopes in Open Pit Mining and Civil Engineering Situations, S44, Cape Town, April 2006. Pp 579 to 594. Saito, M. (1969) ?Forecasting Time of Slope Failure by Tertiary Creep?. Proceedings, Seventh International Conference on Soil Mechanics and Foundation Engineering, Mexico, Vol. 2 pp. 677-683. Saito, M. (1980) ?Semi Logarithimic Representation for Forecasting Slope Failure?. Proceedings of the 3rd International Symposium on Landslides, New Dehli, 1, 321-324. Savely, J.P. and Kastner, V.L. (1981) ?Slope Instab ility at Inspiration's Mines.? Proceedings, Third International Conference on Stability in Surface Mining., (Brawner ed.) Vancouver, Canada, Society of Mining Engineers of AIME. pp. 609-634. Schneider, H. (1968). ?The Time-Dependence of Friction on Rock Joints?. Bull. Of Int. Ass. Of Eng. Geology, No16, pp 235-239. Sijing, W. (1980) ?On the Mechanism And Process of Slope Deformation in an Open Pit Mine?. Rock Mechanics, Vol 13, pp. 145-156. Skempton, A.W. and Hutchinson, J.N. (1969) ? Stability of Natural Slopes and Embankment Foundations, State of the Art Report? Proceedings, 7th Int. Conference on Soil Mechanics and Foundation Engineering, Mexico, Vol 1, pp. 291-340. Smith, R.A. (1984) ?The Lithostratigraphy of the Karoo Supergroup in Botswana?. Geological Survey Dept. Bulletin 26. Published by the Ministry of Mineral Resources and Water Affairs, Republic of Botswana. SRK Consulting (2001a). ?Erongo Mining & Exploration Company (Pty) Ltd, Report on the March 2001 Failure on the Footwall at Navachab Gold Mine?. Report No. 291783 March 2001. Unpublished report. SRK Consulting (2001b). ?Anglo Gold Namibia (Pty) Ltd, Report on the Footwall Slope Design Review for Navachab Gold Mine?. Report No. 291796. September 2001. Unpublished report. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 16 : References Page 7 of 7 SRK Consulting (2002). ?WMC Leinster Operation, Structural Geology and Geotechnical Review of the Harmony Open Pit?. SRK Project Code WM414. February 2002. Unpublished report. Stansfield, G. (1973) ?The Geology of the Area Around Dukwe and Tlalamabele, Central District, Botswana?. District Memoir 1, Geol. Surv., Botswana. Steffen, O. (2005) Personal communication. Stewart, D.H. and Reid, GJ. (1988) ?Afton - a Geotec hnical Pot-pourri?. CIM Bulletin, September, 1988 pp. 77-83. Summons, T. (2002). ?Retention of SE Felsic Failure Rill Material?. Western Mining Corporation, Geology Department Memorandum, Unpublished. Ter-Stepanian, G. (1966). "Types of Depth Creep of Slopes in Rock Masses." Proceedings of 1st Congress of the International Society for Rock Mechanics. Lisbon. Paper 6.6, Vol. II, pp 157-160. Ter-Stepanian, G. (1974). "Depth Creep of Slopes ." Bulletin of I.A.E.G.No.9, pp. 97-102. Terzaghi, K. (1950). ?Mechanism of landslides?, Application of Geology to Engineering Practice, Berkey vol., Geol. Soc. Am., 1950, pp 83-123. Terzaghi, K. (1962). ?Stability of Steep Slopes in Hard Unweathered Rock? Geotechnique, 12, No.4, Dec., pp 251-270. Thomson, C (2006). Personal communication. Varnes, D.J. (1982 ) Time-deformation Relations in Cr eep to Failure of Earth Materials. Proceedings, Seventh Southeast Asian Geotechnical Conference, (McFeat-Smith & Lumb eds.) Hong Kong, Vol 2, pp. 107-130. Venetia Geotec Department (2003) ?QSM Pres entation ? Slope Stability in the Venetia Pit?. Unpublished presentation. Venetia Geotec Department (2006) Unp ublished geological and monitoring data. Wawersik, W.R.(1974) ?Time-Dependent Behaviour of Rocks in Compression?. 3rd Int. Congress, ISRM, Denver, pp 357-363 Wessels, S.D.N. (2005) ?Evaluation of the Use of the Highwall Strain Slope Criteria at Nchanga Open Pit?. A course report submitted to the Univ ersity of the Witwatersrand. Unpublished. Western Mining Corporation Resources Limited (1999). ?Harmony Open Cut Feasibility Study?. Volume 1, Report. Unpublished internal report. Western Mining Corporation Resources Limited (2002). ?Leinster Nickel Operations, Harmony Review Study, Summary of Work January 2002 to Ju ne 2002?. Unpublished internal report. Zavodni, Z.M. and Broadbent, C.D. (1980). ?Slope Failure Kinematics." CIM Bulletin, April, pp. 69-74. Zay, K. (1807) ?Goldau und Seine Gegend, wie sie war und was sie Geworden.? Orell, Fuseli, Zurich, 390p. Zischinsky, U. (1966). "On the Deformation of High Rock Slopes." Proceedings of the First Int. Congress of the International Society for Rock Mechanics, Lisbon. Paper 6.11, Vol. II, pp. 179-185. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 17 :Bibliography Page 1 of 10 17 BIBLIOGRAPHY All the publications contained in this bibliography were reviewed as part of the literature review but were not specifically cited in the thesis text. This list also contains publications relating to the six detailed case studies which have not been presented. Anon. (1976) ?The Geology and the Economic Deposits of Copper, Iron and Vermiculite in the Palabora Igneous Complex: A Brief Review." PMC Staff, Econ. Geol. Vol. 71, No. 1. Jan-Feb 1976 pp 177-192 Bandis, S.C. (1990) ?Scale effects in the strength and deformability of rocks and rock joints?. Proceedings, First International Workshop on Scale Effects in Rock Masses, Loen, Norway, pp. 59-76. Barton, N. (1973) ?Review of a New Shear Strength Criterion for Rock Joints? Eng. Geol., 7, pp287- 332 Barton, N. (1983) ?Application of Q-System and Index Tests to Estimate Shear Strength and Deformability of Rock Masses?. Proceedings, International Symposium on Engineering Geology and Underground Construction, Boston, Balkema. pp. 51-7 Barton, N.R., Lien, R. and Lunde, J. (1974) ?Enginee ring Classification of Rock Masses for the Design of Tunnel Support?. Rock Mechanics, 6(4), pp 189-239. Baldelli, P. and Polloni, G. 1992. "A Landslide Relate d to Gravitational Phenomena in Maratea." Bell, D.H. ed. Proceedings of the 6th International Symposium on Landslides. Christchurch, New Zealand. February, Vol. 1, pp. 3-8. Bertuccioli, P., Esu, F., Federico, G. and Distefano, D. (1991) ?Initial Deformation of High Cuts in Overconsolidated Jointed Clays?. Proceedings, Si xth International Symposium on Landslides, (Bell ed.) Christchurch, New Zealand, Balkema. Vol. 2 pp 1265-1270. Bieniawski, Z.T. (1973) ?Engineering classification of jointed rock masses?. Trans. South African Institution of Civil Engineers, 15, pp. 335-344. Bieniawski, Z.T. (1974). "Geotec hnical Classification of Rock Masses and its Application to Tunnelling." Proceedings of the 3rd Congress of the In ternational Society for Rock Mechanics (ISRM) Denver. Vol. 2, Part A, pp. 27-32. Bieniawski, Z. (1976) ?Rock Mass Classification in Rock Engineering?. In Exploration for Rock Engineering (Bieniawski ed.), Johannesburg, Balkema, pp. 97-106. Bieniawski, Z.T. (1978) ?Determining Rock Mass Deformability - Experiences from Case Histories?. International Journal of Rock Mechanics and Mining Sciences and Geomechanics Abstracts, 15, pp. 237-247. Blackwell, G.H. and Calder, P.N. (1981) ?Practical Aspects of Wall Stability at Brenda Mines Ltd?, Peachland, B.C. Proceedings, Third International Conference on Stability in Surface Mining, (Brawner ed.) Vancouver, B.C., Society of Mining Engineers of AIME. pp. 573-608. Board, M, Chacon, E, Varona, P and Lorig, L (1996). ?Comparative Analysis of Toppling Behaviour at Chuquicamata Open-pit Mine?, Chile. Trans. Instit. Min. Metall. - Sect. A, 105: A11-A21. Boyd, J.M., Hinds, D.V., Moy, D. and Rogers, C. (1973) ?Two simple devices for monitoring movements in rock slopes?. Quarterly Journal of Engineering Geology, 6 pp 295-302. Bovis, M.J. (1990). "Rock Slope Deformation at A ffliction Creek, Southern Coast Mountains, British Columbia." Canadian Journal of Eart h Sciences. Vol 27, pp. 243-254. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 17 :Bibliography Page 2 of 10 Broadbent, C.D. and Rippere, K.H. (1970) ?Fracture Studies at the Kimbley Pit?. Symposium on the Theoretical Background to the Planning of Open Pit Mines with Special Reference to Slope Stability (van Rensburg ed.) Johannesburg, South Af rica, Balkema, Amsterdam. pp. 171-179. Brox, D. and Newcomen, W. (2003). ?Utilizing strain criteria to predict highwall stability performance?. ISRM 2003 ? Technology Roadmap for Rock Mechan ics, South Africa Institute of Mining and Metallurgy, 2003. pp 157-161. Caswell, G. (1992) ?Slope Stability Monitoring at Mt Whaleback?, Western Australian Conference on Mining Geomechanics, Kalgoorlie, Western Australia Coates, D.F., Yu, Y. and Gyenge, M. (1979) ?A Case History of Pit Slope Design?. Proceedings, Fourth International Congress of the International Society for Rock Mechanics., Montreaux, Switzerland, pp. 591-595. Conte, E., Dente, G. and Guerricchio, A. (1992). "Landslide Movements in Complex Geological Formations at Verbicaro, Southern Italy." Bell, D.H. ed. Proceedings of the 6th International Symposium on Landslides. Christchurch, New Zealand. February, Vol. 1, pp. 47-52. Coon, R.F. and Merritt, A.H. (1970) ?Predicting In Situ Modulus of Deformation Using Rock Quality Indexes?. In Determination of In Situ Modulus of Deformation of Rock ASTM, Philadelphia, pp. 154- 173. Crawford, A.M. and Curran, JH. (1982) ?The Influenc e of Rate- And Displacement Dependent Shear Resistance on the Response of Rock Slopes to Seismic Loads?. International Journal of Rock Mechanics and Mining Sciences and Geomechanics Abstracts, 19, pp. 9-14. Cundall, P.A. (1987). "Distinct Element Model s of Rock and Soil Structures." Analytical and Computational Methods in Engineering and Rock Mechanics Brown, E.T. ed., pp. 129-163. Davies, T.R.H. (1982). "Spreading of Rock Av alanche Debris by Mechanical Rock Mechanics Fluidization.", Vol. 15, pp. 9-24. D'Elia, B., Distefano, D., Esu, F. and Federico, G. (1986) ?Slope Movements in Structurally Complex Formations?. Proceedings of the International Symposium on Engineering in Complex Rock Formations, (Chengxiang & Ling eds.) Beijing, Pergamon Press. pp. 430-436. D'Elia, B., Distefano, D., Esu, F. and Federico, G. ( 1988) ?Deformations and Stability of High Cuts in a Structurally Complex Formation: Analysis and Prediction?. Proceedings, Fifth International Symposium on Landslides, (Bannard ed.) Lausanne, Switzerland, Bakema. Vol. 1, pp. 599-604. D'Elia, B., Distefano, D., Federico, G. and Oliva, S. (1984) ?Full-scale Study of a High Cut in a Structurally Complex Formation?. Proceedings, Fourth International Symposium on Landslides, Toronto, Vol. 2 pp. 57-62. Dunnicliff, I. (1988) ?Geotechnical Instrumentation for Monitoring Field Performance?, John Wiley, New York Du Plessis, L. and Martin, D.C. (1991) ?Numerical Modelling Studies for Design of High Rock Slopes at Palabora Copper Mine?. Proceedings, Seventh International Congress on Rock Mechanics, Aachen, Germany, ISRM. Vol. 1 pp. 799-804. Eberhardt, E, Stead D and Coggan JS.(2004). ?Numer ical Analysis of Initiation and Progressive Failure in Natural Rock Slopes ? the 1991 Randa Ro ckslide?, International Journal of Rock Mechanics & Mining Sciences 41 (2004) 69-87. Eisbacher, G.H. and Clague, J.J. (1984). "Destruc tive Mass Movements in High Mountains: Hazard and Management." Geological Survey of Canada, Paper 84-16, 230p. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 17 :Bibliography Page 3 of 10 Enegren, E.G. and Moore, D.P. "B ranham Ridge Slide - Twenty-two Years of Slope Monitoring." Bell, D.H. ed. 6th International Symposium on Landslides. Christchurch, New Zealand. Balkema. Vol 2. pp. 1129-1134. Farmer, I. (1983) ?Engineering Behaviour of Rocks?. Chapman & Hall, London, 208p. Feltus, W.R. (1993) ?Victorious East Failure at Ora Banda ? a Case Study?. Geotechnical Instrumentation and Monitoring in Open Pit and U nderground Mining, (Szwedzicki ed.) Kalgoorlie, Balkema, Rotterdam. pp. 155-159. Finn, W.D.L. and Emery, J.J. (1972) "Time-Dep endent Behaviour of Earth Slopes." Proceedings of 24th International Geological Congress. Montreal, Section 13, pp. 117-123. Giraud. A., Rochet, L. and Antoine. P. (1990). "P rocesses of Slope Failure in Crystallophyllian Formations." Engineering Geology, Else vier, Amsterdam, Vol. 29, pp. 241-253. Gillespie, P. (1999) ?Remote Slope Monitoring at Argyle Diamond Mine?. Australian Centre for Geomechnics Newsletter, 12, pp. 1-9. Gillette, D.R., Sture, S., Ko, R., Gould, M.C. and Scott, G.A. (1983) ?Dynamic Behaviour of Rock? Joints. 24th US Symposium on Rock Mechanics, Vol. 1 pp. 163-179. Glastonbury, J. and Fell, R. (2000) ?Report on the An alysis of ?Rapid? Natural Rock Slope Failures?. UNICIV Report No. R390, School of Civil and Enviro nmental Engineering, University of New South Wales. Glastonbury, J. and Fell, R. (2002) ?Report on the De formation Behaviour of Excavated Rock Slopes?. UNICIV Report No. R403, School of Civil and Enviro nmental Engineering, University of New South Wales. Glawe, U. and Lotter, M. (1996) ?Time Predict ion of Rock Slope Failures Based on Monitoring Results?. Proceedings, Seventh International Symposium on Landslides, (Senneset ed.) Trondheim, Norway, Balkema. Vol. 3, pp. 1551-1555. Goldberg, W.C. and Frizzell, E.M. (1989) ?Pit Slope Monitoring and Back-analysis of the Berkeley pit?. Mining Engineering, 41 (12), pp. 1191-1195. Gonano, L.P. (1980) ?An Integrated Report on Slope Failure Mechanisms at Goonyella, November, 1976?. Technical Report No. 114, CSIRO, Division of Applied Geomechanics. Gray, P.A. (1988)? The Problem of Estimating the Shear Strength of Unstable Rock Slopes?. Fifth Australia-New Zealand Conference on Geomechanics, Sydney, Australia, Vol. 1 pp. 375-380. Guerricchio, A., Melidorb. G. and Rizzo. V. (1988). "Instrumental Observations of the Slope Deformations and Deep Phenomena in Maratea Valley, Italy." Bannard, C. & Bell. D.H. ed. Proceedings of the 5th International Symposium on Landslides. Lausanne. Switzerland. pp. 415-422. Hajiabdolmajid, V. and Kaiser, P. ?Slope Stability in Strain Sensitive Rocks? Publication unknown Hajiabdolmajid, V. and Kaiser, P. ?Modeling Slopes in Brittle Rock? Publication unknown Haines, A and Terbrugge, P.J.(1991) ?Preliminary estimation of rock slope stability using rock mass classification systems?. Proceedings 7th Int. Cong. Soc.Rock Mech., Aachen, Balkema, Vol2, pp 887- 892. Hamel, J.V. (1971a) ?Kimbley Pit Slope Failure?. Fourth Panamerican Conference Soil Mechanics and Foundation Engineering, San Juan, Puerto Rico, ASCE. Vol. 2 pp.117-127. Hamel, J.V. (1971b) ?The Slide at Brilliant Cu t?. Proceedings, Thirteenth US Symposium on Rock Mechanics, (Cording ed.) Urbana, USA, ASCE. pp. 487-510. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 17 :Bibliography Page 4 of 10 Hamel, J.V. (1979) ?Pima Mine Slope Failure?, Arizo na, USA. In Rockslides Avalanches (Voight ed.), Elsevier, Amsterdam, pp. 633-649. Hanekom, H.J. et al, (1965) "The Geology of the Palabora Igneous Complex." South African Geological Survey, Mem. 54, 179 p. Hayashi, S. and Kawabe, H. (1996) ?A Model For Acceleration of Slide to Slope Failure?. Proceedings, Seventh International Symposium on Landslides, (Senneset ed. Trondheim, Norway, Balkema. Vol 2, pp. 1237-1242. Hencher, S.R., Liao, Q.H. and Monaghan, B.G. (1996) ?Modeling Slope Behavior in Open Pits?. Transactions of the Institution of Mining and Metallurgy (Sect A: Mining Industry), 105 pp. A37-A47. Hendron A.J and Patton, F.D. (1985). "The Vaiont Slide: A Geotecbnical Analysis Based on New Geologic Observations of the Failure Surface." Vols . I & II. Technical Report GL-85-5. Department of the Army, U.S. Army Corps of E ngineers, Washington, June, 324p. Hoek E. (1989). "Estimating Mohr-Coulomb Fric tion and Cohesion Values from the Hoek-Brown Failure Criterion." Technical Note, Dept.of Ci vil Engineering, University of Toronto, 6p. Hoek E. and Brown, E. T. (1987). ?Practical Estimates of Rock Mass Strength?. Int . J. Rock Mech. Min. Sci. Vol. 34,No 8 pp 1165-1186. Hoek E. and Brown, E. T. (1988). "The Hoek-Brown Failure Criterion , 1988 Update." Proceedings of the 15th Canadian Rock Mechanics Symposium, Toronto: University of Toronto, pp. 31-38. Hoek, E. and Brown, E.T. (1997) ?Practical Estimates of Rock Mass Strength?. International Journal of Rock Mechanics and Mining Sciences and Geomechanics Abstracts, 34 (8), pp. 1165-1186. Hoek, E. and Diederichs, M. (1989) ?Dips V.2.0?. User Manual, Rocscience, Toronto Hsu K.J. (1975). "Catastrophic D ebris Streams (Sturzstroms) Generated, by Rockfalls." Geological Society of America Bulletin. Vol 86, pp. 1.29-1.41. Hudson, J A., Brown, E.T. and Fairhurst, C. (1971) S hape of the complete stress strain curve for rock. Stability of Rock Slopes. Proceedings, 13th US Sympos ium on Rock Mechanics, University of Illinois., pp. 773-795. Hungr, O. and Evans, S.G. (in press) ?Prediction of the Failure Behaviour of Large Rockslides?. pp. 1- 40. Itasca Consulting Group Inc. (1989). "UDEC: Univ ersal Distinct Element Code." Users Manual, Version 1.64. Minneapolis, Minnesota. Jennings, J. E. (1970) ?'A Mathematical Theory for the Calculation of the Stability of Slopes in Open Cast Mines?, Planning Open Pit Mines, van Rensburg, P.W.J. ed. Proceeding s of the Symposium on the Theoretical Background to the Planning of Open Pit Mines with Special Reference to Slope Stability. Johannesburg, South Africa. Balkema, pp. 33-54. Jennings, J. E. and Black, R. A. L. (1963) ?Factors Affecting the Angle of Slope in Open Cast Mines?. Trans. Soc. Min. Eng., Mar., 1963, pp. 42-53. Jennings, J. E. and Robertson, A. Mac G. (1969) ?Procedures for the Prediction of the Stability of Slopes Cut into Natural Rock,' Proc. 7t h Int. Conf. Soil Mech., Mexico City. Jewell, R.J., Stone K. and Adhikary D., (1992) ?M odeling the Stability of Rock Slopes?, Western Australian Conference on Mining Geomechanics, Kalgoorlie, Western Australia. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 17 :Bibliography Page 5 of 10 Kalkani, E.C. and Piteau, D.R. (1976) ?Finite Element Analysis of Toppling Failure at Hells Gate Bluffs, British Columbia.? Bulletin of the Association of Engineering Geologists, 13 (4), pp. 315-327. Kovari, K. (1988) ?General report: Methods of Monitoring Landslides.? Proceedings, Fifth International Symposium on Landslides, (Bonnard ed.) Lausanne, Switzerland, Balkema, Rotterdam. Vol. 3 pp. 1421-1433. Lane, R.G.T and Knill, J.L. (1974) "Engineering Propert ies of Yielding Rocks." Proceeding of the 3rd Int. Cong. International Society for Rock Mechanics. Denver. Vol. IIA, pp. 336-341. Laubscher, D.H. (1990) "A Geomechanics Classification System for The Rating of Rock Mass in Mine Design." Journal of South African Institute of Mining and Metallurgy. October, Vol. 90, No. 10, pp. 257- 273. Leroueil, S:, Locat, J, Vaunat, J., Picarelli, L. and F aure, R. (1996) ?Geotechnical Characterisation of Slope Movements?. Proceedings, Seventh International Symposium on Landslides, (Senneset ed.) Trondheim, Norway, Balkema. Vol. 1, pp. 53-74. Leroy, AJ. and Lill, J W. (1990) ?Case Sstudy of Palabora Mining Company Ltd.? In Surface Mining (Kennedy ed.), Society for Mining, Metallurgy and Exploration Littleton, pp. 1098-1114. Li, L. and Villaescusa, E. (1999) ?Analysis of slope failure mechanism at the East Pit Daye Iron Ore Mine.? Proceedings, Eighth Australia New Zealand Conference on Geomechanics, (Vitharana & Colman eds.) Hobart, Australia, Vol. 1 pp. 121-125. Li, H. and Brummer, R. (2004). ?Analysis of Pit Wall Failure Mechanism and Assessment of Long-term Stability of Pit Walls, Palabora Mine?. Itasca Consulting Canada, Inc. Unpublished Report to Palabora Mining Company. Li, H., Boland, J. and Speight, H. (1997) ?Time Depe ndent Behaviour of Rock Masses?. CSIRO Report No. 342F. Liao, C. and Shi, Z. (1983) ?In-Situ Stress Measurements and Their Application to Engineering Design at the Jinchuan Mine?. Proceedings, Fifth Inte rnational Congress of Int. Soc. Rock Mechanics, Melbourne, Australia, ISRM. Vol. 2 pp. D87-D89. Liu, S.F., Champion, D.C. and Cassidy, K.F. (2002). Geology of the Sir Samuel 1:250,000 sheet area, Western Australia. Geoscience Australia Record 2002/14, Unpublished. Lajtai, E.Z. (1991) ?Time Dependent Behavior of the Rock Mass?. Geotechnical and Geological Engineering, Vol 9, pp. 109-124. Lajtai, E.Z. and Schmidtke, R.H. (1986) ?Delayed Failure in Rock Loaded in Uniaxial Compression?. Rock Mechanics & Rock Engineering, 19 pp. 11-25. Lombard.A.F., 1962 - "Geology and Exploration of the Copper deposit in Carbonatite at Loolekop, Palabora Complex." P.M.C. unpublished report, p46 Martin, D.C. (1990) "Deformation of Open Pit Slopes by Deep Seated Toppling." International Journal of Surface Mining and Reclamation, Vol. 4, No.4, pp. 153-164. Martin, D.C. (1997) ?The Effect of Cohesion Loss and Stress Path on Brittle Rock Strength?. Canadian Geotech. Journal 34, pp.698-725. Martin, D.C. and Carew, T.J. (1986) "Application of a Rock Buttress for Design of Slopes at Cassiar Mine." Canadian Institute of Mining Bulletin. December, Vol. 79, No. 896, pp. 60-66. Martin, D.C. and Franklin, J.A. 1990. "Deep Open Pi t Mines." Franklin J.A. ed. Mine Monitoring Manual, Canadian Institute of Mining and Metallurgy, Special Volume 42. Chapter 2, pp. 13-25. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 17 :Bibliography Page 6 of 10 Martin, D.C. and Newcomen, H.W. (1991) "Geote chnical Assessments and Numerical Modelling for the West Wall of the Palabora Open Pit." Technical Report prepared for Palabora Mining Company, July, 39p. Martin, D.C., Newcomen, H.W. and Gilmore, B".W. (1992) "Phase I Geotechnical Report on Geotechnical Assessments for the Design of Slopes in the Betze Pit." Unpublished report prepared for Barrick Goldstrike Mines Inc. March, 63p. Martin, D.C., Newcomen, H.W. and Sheehan, P. (1988) "Behaviour of a Steep Unbenched Footwall Slope at Smoky River Coal, Alberta." ClM Distri ct 6 Meeting, Fernie, B.C. September, 9p. Martin, D.C., Steenkamp, N.S.L. and Lill, lW. (1986) ?Application of a Statistical Analysis Technique for Design of High Rock Slopes at Palabora Mine, South Africa.? Mining Latin America, Santiago, Chile, Institution of Mining and Metallurgy. pp. 241-255. Martin, D.C. and Tod, J.D. (1990) "Engineering Geol ogy and Geotechnical Assessments of the Muddy Lake Landslides, Northwestern British Columbia." Symposium on Landslide Hazards in the Canadian Cordillera. Geological Association of Canada Annual Meeting, Vancouver. May, 14p. Mahr, T. (1977) "Deep-reaching Gravitational Defo rmations of High Mountain Slopes." Bulletin of LA.E.G. No. 16, pp. 121-127. Mahr, T. and Nemcok, A. (1977) "Deep Seated Creep Deformations in the Crystalline Cores of the Tatry Mountains." Bulletin of t he I.A.E.G. No. 16, pp. 104-106. Miller, VJ. (1981) ?The Northeast Tripp Slide - a 11.7 million Cubic Metre Wedge Failure at Kennecotts Nevada Mine Division?. Proceedings, Third International Conference on Stability in Surface Mining, (Brawner ed.) Vancouver, Canada, Society of Mining Engineers of AIME. pp. 725-743. Meyers, AG. and Priest SD. June (1992) ?The Predict ion of Rock Mass Strength and Deformability for Engineering Design ? Laboratory Practice?, Wester n Australian Conference on Mining Geomechanics, Kalgoorlie, Western Australia. Moore, D.P., Imrie, A.S. and Baker, D.G. (1991) "Rockslide Risk Reduction Using Monitoring." Proceedings of the Canadian Dam Safety Conference, Whistler, British Columbia. September, pp.245- 258. Moore, D.P., Ripley, B.D. and Groves, K.L. (1992) "Evaluation of Mountain Slope Movements at Wahleach." Geotechnique and Natural Hazards Symposium, Vancouver. May, pp. 99-107. Munn, FJ. (1985) ?Coping With Wall Failures?. ClM Bulletin, 78 (884), pp. 59-62. M?ller, L. (1964) "The Rock Slide in the Vajont Valley." Rock Mechanics and Engineering Geology, Journal of the International Society fo r Rock Mechanics. Vol. 2, pp. 148-212. Nichols, T.C. (1980) ?Rebound, its Nature and Effect on Engineering Works.? Quarterly Journal of Engineering Geology, 13 pp. 133-152. Newcomen, HW and Martin, DC (1988). ?Geotechnical Assessment of the Southeast Wall Slope Failure at Highmont Mine, British Columbia?. CIM Bulletin. September 1988. pp 71-76. Nieto, A.S. (1989) "Influence of Geologic Details on the Field Behaviour of Soils and Rocks: Some Case Histories." Cording et al eds. The Art and Scie nce of Geotechnical Engineering at the Dawn of the Twenty First Century. Prentice Hall, pp. 386-415. Pells, P.J.N. (1990) ?Stresses and Displaceme nts Around Deep Basements in the Sydney Area.? Proceedings, Seventh Australian Tunnelling Conference, Sydney, September 1990, The Institution of Engineers, Australia. pp. 241-249. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 17 :Bibliography Page 8 of 10 Ryan, T.M. and Call, R.D. (1992) ?Applications of Rock Mass Monitoring for Stability Assessment of Pit Slope Failure?. Proceedings of the 33rd US Rock Mechanics Symposium, (Tillerson & Wawersik ed.) Santa Fe, New Mexico, Balkema, Rotterdam. pp.221-229. Saito, M. (1965) ?Forecasting the Time of Occurrence of a Slope Failure?. Proceedings, Sixth International Conference on Soil Mechanics and Foundation Engineering, Montreal, Vol. 2 pp. 537- 541. Sharp, J. (1979) ?Drainage Used to Contro l Movements of a Large Rock Slide in Canada?. Proceedings, First International Mine Drainage Symposium, (Argall & Brawner eds.) Denver, Colorado, pp. 423-436. Sharp, J.C., Bergeron, M. and Ethier, R. (1987a) ?Excavation, Reinforcement and Monitoring of a 300m High Rock Face in Slate?. Proceedings, Sixth International Congress on Rock Mechanics, Montreal, Canada, Vol. 1 pp. 533-540. Sharp, J.C., Lemay, C. and Neville, B. (1987b) ?O bserved Behaviour of Major Open Pit Slopes in Weak Ultrabasic Rocks?. Proceedings, Sixth Inte rnational Congress on Rock Mechanics, Montreal, Canada, Vol. 1 pp. 541-550. Siddle, H.J. and Hutchinson, J.N. (1992) "A Slid ing and Multiple Toppling Failure in South Wales, U.K." Bell, D.H. ed. Proceedings of the 6th Inte rnational Symposium on Landslides. Christchurch, New Zealand. February, Vol. I, pp. 213-218. Sijing, W: and Xiaoning, W. (1990) ?A Study on Dy namics of Catastrophic High Speed Landslides in China?. Proceedings, Sixth International IAEG Congress, Amsterdam, Balkema, Rotterdam. Vol. 3, pp. 1747-1754. Sj?berg, J. (1999). ?Analysis of Large Scale Rock Slopes?, PhD Thesis, Lulea University of Technology. Small. C. A. and Morgenstern, N.R. (1992). "Per formance of a Highwall in Soft Rock, Highvale Mine, Alberta." Canadian Geotechnical Journal. Vol. 29, No.4, pp. 353-363. Smith, D. and Hungr, O. (1992) ?Failure Behaviour of Large Rockslides?, Report to the Geological Survey of Canada and BC Hydro and Power Authority. Thurber Engineering Ltd. Soregaroli, A.E. (1974) ?Geology of the Brenda copper-molybdenum Deposit in British Columbia?. CIM Bulletin, 67 (750), pp. 76-83. St John, CM (1972). ?Numerical and Observational Methods of Determining the Behaviour or Rock Slopes in Open Cast Mines.? PhD Thesis, University of London. Stacey, T.R, Yu, X., Amstrong, R. and Keyter GJ ( 2003) ?New Slope Stability Considerations for Deep Open Pit Mines.? Journal of the SA Institute of Mining and Metallurgy, July/August 2003. pp 373 to 389. Sullivan, T.D. (1986) ?An Approach to Coping with Unstable Pit Walls?. Proceedings, 13th Congress of the Council of Mining and Metallurgical Institutions, (Brady, Kay & Hore-Lacy eds.) Singapore, Vol. 3 pp. 193-199. Sullivan, T.D. (1993) ?Understanding Pit Sl ope Movements?. Proceedings of the Australian Conference on Geotechnical Instrumentation and Monitoring in Open Pit and Underground Mining, (Szwedzicki ed.) Kalgoorlie, Balkema. pp. 435-445. Suwa, H. (1991) ?Visually Observed Failure of a Rock Slope in Japan?. Landslide News, 5 pp.8-1O. Thakkar, B.O. (1992). ?An Approach to Dealing Safely with Slope Instability in Open Pits.? Western Australian Conference on Mining Geomechanics, Kalgoorlie, Western Australia PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 17 :Bibliography Page 7 of 10 Perry, V.D. (1952) ?Geology of the Chuquicamata Orebody.? Mining Engineering, 4(12), pp.1166- 1169. Peterson, R. (1958). "Rebound in the Bearpaw Shale, Western Canada." Bulletin of the Geological Society of America. September, Vol. 69, pp. 1113-1124. Piteau, D. R. (1970a). "Engineeri ng Geology Contribution to the Study of Stability of Slopes in Rock with Particular Reference to the DeBeers Mine?. Volume 1 and 2. Thesis submitted for the degree of Doctor of Philosophy in the Faculty of Science, University of the Witw atersrand, Johannesburg. Piteau and Associates.(1980). ?Palabora Mining Company, Phalaborwa, South Africa, Slope Stability Analysis and Design of the Open Pit Slopes?. May 1980. PMC unpublished report.p6 Piteau, D.R.. McLeod, B.C., Parkes, D.R. and Lou, J. K. (1979) "Rock Slope Failure at Hell's Gate, British Columbia, Canada." in Voight. B. ed. Ro ckslides and Avalanches, 2 Engineering Sites. Developments in Geotechnical Engineering l4B. Elsevier, pp.541-574. Piteau, D.R.. Stewart, A.F., Martin, D.C. and Trenholme, B.S. (1985). "A Combined Limit "Equilibrium and Statistical Approach for Designing Rock Slopes" . Specialty Conference on Rock Engineering. ASCE Spring Meeting, July. Pritchard, M.A. (1989) ?Numerical Modeling of Large Scale Toppling?. A Thesis Submitted in Partial Fulfillment of the Requirements for the Degree of Master of Applied Science in the Faculty of Graduate Studies, University of British Columbia. September 1989. Pritchard, M.A. and Savigny, K.W. (1990) ?Numerical Modeling of Toppling.? Canadian, Geotechnical Journal, Vol 27, pp. 823-834. Raisbeck, D. (1988) ?Instrumentation of Natural Slopes and Excavations?. Australian Geomechanics Society - Extension Course on Geotechnical Instrumentation Melbourne, Australia, pp. 10.1 - 10.14. Raphael, J.M. and Goodman, R.E. (1979) ?Strength and Deformability of Highly Fractured Rock?. ASCE Journal of Geotechnical Engineering, 105 pp. 1285-1300. Rapiman, M. (1993) ?Slope Stability and Rock Mechanics Analyses Chuquicamata mine, Codelco, Chile.? Innovative Mine Design for the 21st Century - International Congress on Mine Design, (Bawden & Archibald ed.) Kingston, Ontario, Balkema, Rotterdam.pp. 35-44. Rapiman, M. and Krstulovic, G. (1986) ?Geomechnics for slope design at the Chuquicamata Mine, Chile.? Proceedings, International Symposium on Engineering in Complex Rock Formations, (Tan ed.) Beijing, China, Science Press, Beijing. Vol. 1 pp. 399-407. Read, J.R.L. (1998) ?Deformation of High Rock Slopes in Open Pit Mines.? Proceedings, Eighth International Congress, International Association of Engineering Geology, Balkema, Rotterdam. Vol. 5 pp. 2983-2993. Rigby, R.J and Carr, R.J. (1987) ?Monitored Failure of an Excavation in an Ancient Landslide within the Newcastle Coal Measures.? In Soil Slope Instability and Stabilisation, (Walker & Fell ed.) Balkema, Rotterdam. pp. 397-402. Rosengren, K.J and Krehula, F.J. (1965) ?Earth Movements and Batter Stability in the Latrobe Valley Open Cuts.? Proceedings, Eighth Commonwealth Mining and Metallurgical Congress, Wellington, New Zealand, pp. 573-585. Rocscience (1998-2001). Rocscience Software Products ? Phase 2, Slide, DIPs, Swedge and RocPlane Rocscience Inc, Toronto, Ontario, Canada Ronaszeki, J. (1992). ?Structural Geology and its Application in Slope Stability Analysis and Pit Wall Design at the Mt Whaleback Mine?, Western Australian Conference on Mining Geomechanics, Kalgoorlie, Western Australia PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 17 :Bibliography Page 9 of 10 Thompson, P.W. (1993) ?Confirmation of a Failure Mechanism Using Open Pit Monitoring Methods?. Proceedings of the Australian Conference on Geotechnical Instrumentation and Monitoring in Open Pit and Underground Mining, (Szwedzicki ed.) Balkema. Vol. 1 pp. 483-489. Thompson, P.W. and Cierlitza, S. (1993) ?Identification of a Slope Failure Over a Year Before Final Collapse Using Multiple Monitoring Methods?. Proceedings of the Australian, Conference on Geotechnical Instrumentation and Monitoring in Open Pit and Underground Mining., (Szwedzicki ed.) Kalgoorlie, Balkema. Vol. 1 pp. 491-50 Tian, Y.S. and Zhang, W. (1995) ?Engineering Geologic al Characteristics and Rheological Property of Rock Mass In Jinchuan Nickel Mine?. Proceedings, 8 th Congress of Int. Soc. Rock Mechanics, (Fujii ed.) Tokyo, Japan, ISRM. Vol. I pp. 9-11. Tod, J.D. 1988. "A Preliminary Analysis and Engineering Geology of the Muddy Lake Rock Avalanche." Unpublished B.A.Sc. Thesis, University of British Columbia, 32p. Underwood, L.B., Thorfinnson, S.T. and Black, W. T. (1964) ?Rebound in the Redesign of Oahe Dam Hydraulic Structures?. ASCE Journal of Soil Mechan ics and Foundations Division, SM2 (3840), pp. 65- 86. US Army Corps of Engineers (1984) ?Libby Dam and Lake Koocanusa Project, Montana - Left Bank Slope?. US Army Corps of E ngineers - Seattle District. Van Dine, D. F. (1980) "Engineering Geology and Ge otechnical Study of Drynoch Landslide British Columbia." Geological Survey of Canada Paper 79-31. Energy Mines and Resources Canada. Ottawa, 34p. Van As, A. and Guest, A.R. ?Numerical Modeling as a Tool to Estimate Time-dependant Slope Failures?, Publication source unknown. Varnes, D.J. (1978) "Slope Movement Types a nd Processes." Landslides: Analysis and Control.? Transportation Research Board Special Publication 176. National Academy of Sciences, Washington D.C., pp. 11.-33. Voight, B. (1979) ?Wedge Rockslides, Libby Dam and Lake Koocanusa, Montana. In Rockslides and Avalanches? (Voight ed.), Elsevier, Amsterdam, pp. 281-315. Voight, B. and Kennedy, B.A. (1979) ?Slope failure of 1967-1969, Chuquicamata Mine, Chile?. In Rockslides and Avalanches (Voight ed.), Elsevier, Amsterdam, pp. 595-632. Volk, P.L. (1987) ?Monitoring of planar slope failure in rockfill quarry, Glenbawn Dam?. In Soil Slope Instability and Stabilisation (Walker & Fell eds.), Balkema, Rotterdam, pp.411-415. Wade, N.H. and Peterson, T.W.P. (1986) ?Highwall Monitoring, Instrumentation and Stability Analysis at Highvale Coal Mine, Alberta?. International Symposium on Geotechnical Stability in Surface Mining, Calgary, pp. 373-384. Warren-Newcomen, H. and Martin, D.C. (1988) ?Geotechnical Assessment of the Southeast Wall Slope Failure at Highmont Mine, British Columbia?. CIM Bulletin, September, 1988 pp. 71-76. Windsor, C.R. (1993) ?Measuring Stress and Deformation in Rock Masses?. Proceedings of the Australian Conference on Geotechnical Instrumentation and Monitoring in Open Pit and Underground Mining., (Szwedzicki ed.) Kalgoorlie, Balkema. Vol. 1 pp. 33-52. Wojciechowski, J. and Kasza, L. (1984) "Elastic Deformation of Open Pit Slopes." Geddes, J.D. ed. Proceedings of the 3rd International Conference on Ground Movements and Structures. Cardiff, July, pp. 751-766. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Section 17 :Bibliography Page 10 of 10 Wylie, D.C. and Munn, F.J. (1978) The Use Of Mo vement Monitoring to Minimise Production Losses Due to Pit Slope Failures. Proceedings, First International Symposium on Stability in Coal Mining, (Brawner &. Dorling ed.) Vancouver, Miller Freeman. p 75-94. Xu, D., Ge, x., Feng, S., Chen, C., Gu, X. and Xu, Y. (1996) ?Study of Stability of Rocky Slope Under High Displacement?. Proceedings, Korea-Japan Jo int Symposium on Rock Engineering, Vol. 1 pp. 445-449. Zavodni, Z.M. and McCarter, M.K. (1976).?Main Hill Slide Zone, Utah Copper Division?. Proceedings of the 17th U.S. Symposium on Rock Mechanics , Snowbird Resort, Alta, Utah. Zavodni, Z.M. (2000) ?Time Dependent Movements of Open Pit Slopes?. Proceedings, Slope Stability in Surface Mining, (Hustrulid, McCarter and vanZyl eds.) SME. pp. 81-87. Zvelebil, J. (1984) ?Time Predictio n of a Rockfall from a Sandstone Rock Slope?, Proceedings, Fourth International Symposium on Landslides, Balkema, Rotterdam Vol. 1, pp. 93-95. PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 1 APPENDIX 1 DETAILED CASE STUDY 1, NAVACHAB GOLD MINE, NAMIBIA PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 1 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 1 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 1 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 1 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 1 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 1 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 1 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 1 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 1 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 1 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 1 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 1 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 1 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 1 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 1 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 2 APPENDIX 2 DETAILED CASE STUDY 2, BIBIANI GOLD MINE, GHANA PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 2 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 2 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 2 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 2 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 2 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 2 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 2 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 2 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 2 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 2 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 2 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 2 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 3 APPENDIX 3 DETAILED CASE STUDY 3, MT KEITH, WESTERN AUSTRALIA PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 4 APPENDIX 4 DETAILED CASE STUDY 4, LEINSTER NI CKEL, HARMONY PIT, WESTERN AUSTRALIA PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 5 APPENDIX 5 DETAILED CASE STUDY 5, VENETIA, SOUTH AFRICA PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 5 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 5 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 5 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 5 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 5 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 5 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 5 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 5 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 5 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 5 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 5 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 5 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 5 Table A5.1 : Summary of Joint Orientations Joint Set Joint Strike/Name Mean Dip/Dipdir Domain 1 Domain 2 Domain 3 Domain 4 Domain 5 Domain 6 Domain 7 Domain 8 Domain 9 Domain 10 Domain 11 J0 Gneissic and schistocity banding (layering parallel) Varies across fold 49?/35? 82?/356? 58?/52? 49?/8? 37?/20? 27?/88? n/a 39?/348? 46?/13? 52?/2? J1 North to north-north east striking (West dippers) 85?/265? 77?/216? 82?/259? 79?/252? 85?/309? 66?/264? 86?/250? 83?/93? n/a 85?/263? J2 Northeast strike (South east dippers) 82?/140? 80?/134? 84?/114? 90?/355? 86?/164? 73?/145? 83?/317? 86?/144? n/a 7 8? /118? 89?/114? 76?/113? J4 West -northwest strike (South-southwest dippers) 81?/212? 7 5? /33? 87?/10? 63?/196? 81?/219? 81?/61? n/a 85?/14? 72?/204? 65?/221? J5 32?/295? 36?/321? n/a J6 Northwest striking (South east dippers) 47?/230? 59?/206? 3 8? /231? 50?/226? n/a J7 n/a 52?/126? J8 51?/66? 53?/7? n/a J9 8?/187? n/a PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 5 Table A5.2 : Average Intact Rock Properties from Laboratory Tests (Hannweg, 2006) Rock Type Rock Code UCS (MPa) Density (g/cm3) Poisson's ratio Young's Modulus (MPa) Base Friction Angle Tensile Strength (MPa) Diabase (Dolerite) DIA 296 3.00 0.27 89 33 25 Fuchsitic quartzite FQ 124 2.71 0.21 25 34 10 Marble MBL 153 2.79 0.29 73 20 11 Phylittes PHY 125 2.83 0.23 52 31 21 Biotite Gneiss BG 214 2.78 0.24 60 37 13 Biotite Schist BS 112 2.91 0.23 46 31 10 Amphibolite AM 191 3.07 0.25 68 36 15 Table A5.3 : Summary of Rock Mass (Hannweg, 2006) Rock Type Rock Code MRMR (Laubscher,1990) Q-Values (Barton et al, 1974) Diabase (Dolerite) DIA 68 5 Fuchsitic quartzite FQ 52 11 Marble MBL 64 6 Phylittes PHY 77 24 Biotite Gneiss BG 62 4 Biotite Schist BS 55 7 Amphibolite AM 66 7 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 6 APPENDIX 6 DETAILED CASE STUDY 6, ORAPA & LETLEKHANE, BOTSWANA PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 6 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 6 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 6 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 6 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 6 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 6 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 6 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 6 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 6 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 6 PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 6 Table A6.3: Summary of the Dominant Joint Sets in the Letlhakane Basalts (Dirks, 2001) Joint set Type Spacing Importance to large instabilities H (bedding); bedding +/- 5-15 m low 230/85 Shear zone (dextral) 50 m medium to high 250/85 Shear zone (sinistral) 50 m medium 240/85 (carbonate vein set) Laminated, open veins 50 m low 230/85 (master joints); J1: joints Individual joints 1-5m; concentrations of joints every 50 m. low to medium 300/90 J2: joints Irregularly spaced low 340/85 J3: joints Irregularly spaced low 210/20 joints <5 m low Table A6.4: Summary of the Dominant Joint Sets in the Letlhakane Sandstones and Mudstones (Dirks, 2001) Joint set Type Spacing Importance to large instabilities H (bedding); Bedding-parallel foliation domains Concentrated in mudstone beds Potentially very large 235/85 J1: joints (also shear joints) Individual joints 3-5 m; bundles of joints every 50 m. low 265/65 Shear joint (with normal component) 15 m medium 060/65 Shear joint (with normal component) 15 m medium Steep ENE trending jo joints Irregularly spaced low 190/20 joints <5 m low PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 6 Table A6.5: Average Values of Laboratory Test Results for Different Rock Types (Jakubec et al, 2000) Rock type / test Density Kg/t UCS MPa TS MPa PWV m/s SWV M/s Young?s modulus Gpa Poisson?s ratio Basalt 2600 81 9 4660 2630 42 0.20 Sandstone 2030 24 2 3330 1910 17 0.32 Mudstone 2380 48 4 2820 1540 6 0.14 L.sandstone 2410 58 - 3330 2250 16 0.24 C.mudstone 1960 23 7 2600 1580 7 0.22 Granite 2690 172 11 5480 3260 66 0.23 P.kimberlite 2300 22 3 2470 1320 66 0.23 TKB 2510 46 5 3410 1860 12 0.23 C = carbonaceous UCS = uniaxial Compressive Strength TS = tensile strength PWV = P wave velocity SWV = S wave velocity TKB = massive tuffisitic kimberlite breccia P = pyroclastic L = laminated Table A6.6 : Average Values of Geotechnical Parameters for Different Rock Types (Jakubec et al, 2000) Rock type / RQD JC IRS FF/m RMR MRMR DRMS Parameter MPa Basalt 14 25 12 25 62 57 58 Sandstone 14 20 4 26 50 46 16 Mudstone 14 14 3 22 39 28 4 L.sandstone 14 16 4 26 46 37 14 C.mudstone) 10 17 2 10 41 33 4 Granite 15 26 8 26 60 55 54 P. kimberlite 14 22 4 15 56 45 10 TKB 14 26 5 17 66 56 18 DRMS = Design rock-mass strength FF/m = Fracture frequency per metre IRS = Intact rock strength JC = Joint condition PhD Thesis : Time and Event Dependent Deformation Behaviour of Unsupported Rock Slopes Appendix 6 MRMR = Mining rock-mass rating RMR = Rock-mass rating RQD = Rock quality designation Table A6.7 : Intact Rock And Natural Joint Parameters Used in Stability Analyses (Jakubec et al, 2000) Intact rock Joints Rock type Density kg/t Youngs Modulus GPa Poisson?s Ratio Friction Degree Cohesion MPa Friction rating Cohesion (MPa) Basalt 2600 42 0.2 33.1 23.9 34.6 0 Sandstone 2100 17 0.3 43.2 10.9 35.7 0 Mudstone 2350 6 0.2 25.7 15.2 30.5 0 L.sandstone 2400 16 0.3 43.2 51.5 33.6 0 Granite 2650 66 0.2 37.5 44.4 31.0 0