UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Span design for entry-type excavations Lang, Brennan Davis Allan 1994

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

Item Metadata


831-ubc_1994-0557.pdf [ 8.11MB ]
JSON: 831-1.0081176.json
JSON-LD: 831-1.0081176-ld.json
RDF/XML (Pretty): 831-1.0081176-rdf.xml
RDF/JSON: 831-1.0081176-rdf.json
Turtle: 831-1.0081176-turtle.txt
N-Triples: 831-1.0081176-rdf-ntriples.txt
Original Record: 831-1.0081176-source.json
Full Text

Full Text

SPAN DESIGN FOR ENTRY-TYPE EXCAVATIONSbyBRENNAN DAVIS ALLAN LANGB.A.Sc., The University of British Columbia, 1987A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF APPLIED SCIENCEinTHE FACULTY OF GRADUATE STUDIESDepartment of Mining and Mineral Process EngineeringWe accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAOctober, 1994© Brennan Davis Allan Lang, 1994Signature(s) removed to protect privacyIn presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)Department of /{i4fl(j d )in rtc/ oi’sThe University of British ColumbiaVancouver, CanadaDate Jc/o/r /994-DE-6 (2/88)Signature(s) removed to protect privacyABSTRACTUnderground entry-type excavations require higher factors of safety than do non-entry excavationsyet not as high as those required for permanent underground structures. A review is made of undergroundexcavation span design techniques and the conditions under which they can be applied. Shortcomings ofthese existing methods, as they are applied to cut and fill stopes and other entry-type excavations, arehighlighted.A design procedure specific to conditions found in entry-type mining is proposed. At the centre ofthe procedure is an empirical span design chart, called the “Stability Graph for Entry-Type Excavations”,which provides a practical tool for mining engineers to design stable entry-type excavations. Thedevelopment of this chart and its use as a design tool is a result of the statistical analysis of 172 stopingcase histories collected at a large underground gold mine in northern Ontario,The influence of artificial support in maintaining stability and increasing span is investigated. Areport is given of a trial support program carried out at the same operation using a concentrated pattern ofcable bolts to replace a post pillar in order to increase span.11TABLE OF CONTENTSABSTRACT.UTABLE OF CONTENTS iiiLIST OF TABLESLIST OF FIGURESACKNOWLEDGMENTS x1. INTRODUCTION 11.1 BACKGROUND 11.2 RESEARCH METHODOLOGY 32. THE DETOUR LAKE MINE 52.1 INTRODUCTION 52.2 GEOLOGY 52.2.1 Regional Geology 52.2.2 Orogeny 52.2.3 MineGeology MamZone Quartz Zones TalcZones 72.3 MINiNG METHODS 72.3.1 Primary Development 72.3.2 MCF Stope Development 82.3.3 Drilling and Blasting 82.3.4 Mucking 82.3.5 Ground Support Mechanically Anchored Rockbolts Swellex Rockbolts Cable Bolts Wire Mesh Steel Straps 102.3.6 Backfilling 113. REVIEW OF DESIGN METHODOLOGIES 243.1 LITERATURE REVIEW 243.1.1 BeamandPlate Failure Beam Failure PlateFailure 273.1.2 Voussoir Block Failure VoussoirBeams VoussoirPlates 303.1.3 Structurally Controlled Failure 323.1.3.1 Stereonet Analysis Techniques 321113.1.3.2 Computational Techniques.333.1.4 Chimney Failure 343.1.5 Rock Mass Failure 343.1.5.1 NGI-Q Rock Mass Classification System 353.1.5.2 CSlRRockMassRating 363.1.5.3 ModifiedQ-Ratmg 383.1.5.4 Mining Rock Mass Rating 383.1.5.5 Golder Crown Piliar Study Database 403.1.6 Stress Induced Failure 423.1.6.1 Limitations of Numerical Modeling 423.1.6.2 The Boundary Element Method 433.1.6.3 The Finite Element Method 453.1.6.4 The Distinct Element Method 453.2 SURVEY OF OTHER CUT AND FILL OPERATORS 463.3 DESIGN PHILOSOPHY 474. DATABASE 874.1 iNTACT STRENGTH PROPERTIES 874.2 FABRIC ANALYSIS 884.2.1 Joint Orientation 884.2.2 Joint Roughness 894.2.3 Rock Strength 894.2.4 JointAperture 894.2.5 Joint Spacing 894.2.6 Joint Continuity 904.3 ROCK MASS CHARACTERIZATION 904.3.1 MainZone 904.3.2 Talc Zone 904.3.3 Hangingwall Rock 914.4 STRESS DETERMINATION 914.5 MINING HISTORY 915. NUMERICAL MODELING 1265.1 THREE-DIMENSIONAL BEAP 1265.2 NUMERICAL MODELS 1275.2.1 Post Pillar Modeling 1275.2.2 Sill Pillar Modeling 1295.2.3 OtherAreas 1316. SPAN DESIGN 1436.1 DATABASE 1436.1.1 Definitionof Span 1436.1.2 Definition of Stability 1436.1.2.1 Stable Excavations 1446.1.2.2 Potentially Unstable Excavations 1446.1.2.3 Unstable Excavations 1446.1.3 Detour Lake Mine Observations 1446.2 STATISTICAL ANALYSIS 1456.2.1 Objective 1456.2.2 Group Classification 1456.3 LIMITATIONS OF THE EMPIRICAL DESIGN METHOD 149iv6.3.1 Structure,1496.3.2 State ofStress.1496.4 COMPARISON WITH OTHER EMPIRICAL SPAN DESIGN METHODS 1506.4.1 Comparison to RIVIR Data 1506.4.2 Comparison to NGI Data 1506.4.3 Comparison to Golder Crown Pillar Data 1516.5 1NSTRUMENTATION AND OBSERVATION 1516.5.1 Instrumentation to Determine Instability 1516.5.2 Visual Monitoring of Ground Conditions 1526.6 CONCLUSIONS 1527. 1NFLUENCE OF GROUND SUPPORT ON SPAN DESIGN 1867.1 INTRODUCTION 1867.2 KEY-BLOCK SUPPORT 1867.3 CABLE BOLTING 1877.3.1 Cable Bolts 1877.3.2 Cable Bolt Modifications and Accessories 1887.3.3 Cable Bolt Support Design 1897.4 POST PILLARS 1897.5 CABLE BOLT PILLAR 1907.5.1 LocationofTest 1917.5.2 Geometry of Post Pillar 941 1917.5.3 Estimate of Pillar Strength 1917.5.4 Estimate of State of Stress in Pillar 1917.5.5 Description of the Rock Mass 1927.5.6 Cable Bolt Support Implementation 1927.5.6.1 Support Design 1927.5.6.2 Cable Bolt Installation 1937.5.7 Monitoring Instrumentation 1937.5.8 Results of Monitoring 1947.5.8.1 Visual Monitoring 1947.5.8.2 Instrumentation 1957.5.9 Summary of Cable Bolt Pillar Support Trial 1958. CONCLUSIONS AND RECOMMENDATIONS 2268.1 CONCLUSIONS 2268.2 RECOMMENDATIONS 228BIBLIOGRAPHY 230APPENDIX A: SURVEY OF CANADIAN CUT AND FILL MINES 235VLIST OF TABLESTable 2.1 Drilling and Loading Specifications for Cut and Fill Breasting 12Table 2.2 Technical Specifications of Mechanically Anchored Rockbolts 12Table 2.3 Technical Specifications of Swellex Bolts 13Table 2.4 Detour Lake Mine Backfill Properties 13Table 3.1 NGI- Q Classification System Ratings for Individual Parameters 50Table 3.2 Excavation Support Ratios 51Table 3.3 Geomecharncs Classification of Rock Masses 52Table 3.4 Geomechanics Classification Guide for Excavation and Support in Rock Tunnels 53Table 3.5 Mining Rock Mass Rating System Adjustments 54Table 3.6 Crown Pillar Study Data 55Table 3.7 Span Design Methodology at Canadian Cut and Fill Mines 56Table 3.8 Post Pillar Design Methodology at Canadian Cut and Fill Mines 56Table 3.9 Sill Pillar Design Methodology at Canadian Cut and Fill Mines 56Table 3.10 Cable Bolt Support Design Methodology at Canadian Cut and Fill Mines 56Table 4.1 Summary of Intact Rock Strength Characteristics 93Table 4.2 UBC Unconfined Compressive Strength Test Results - Talc Zone 94Table 4.3 Detour Lake Mine - Rock Properties 94Table 4.4 Approximate Classification of Rock Strength Based on Hardness 95Table 4.5 Joint Aperture Classification 95Table 4.6 Rock Mass Classification Detour Lake Mine Main Zone 96Table 4.7 Rock Mass Classification - Detour Lake Mine Talc Zone 97Table 4.8 Typical CSIR Rock Mass Rating at Detour Lake Mine 98Table 5.1 Major and Minor Principal Stresses in Crown and Sill Pillars 132Table 6.1 Raw Data - Detour Lake Mine 154Table 6.2 Corrected Data - Detour Lake Mine 158Table 6.3(a) Statistical Summary of Raw Data 162Table 6.3(b) Statistical Summary of Corrected Data 162Table 6.4 Mahalanobis Distance and Group Classification Probabilities 163Table 7.1 941 Post Pillar Geomechanics Rock Mass Rating 197Table 7.2 Main Zone Geomechanics Rock Mass Rating 197Table 7.3 Talc Zone Geomechanics Rock Mass Rating 197viLIST OF FIGURESFigure 2.1 Location Map - Detour Lake Mine 14Figure 2.2 Plan View of Lithologic Units - Detour Lake Mine 15Figure 2.3 Structural Geology Model - Detour Lake Mine 16Figure 2.4 Plan View of Ore Zones - Detour Lake Mine 17Figure 2.5 Longitudinal Section of Main Zone - Detour Lake Mine 18Figure 2.6 Main Zone Hangingwall Contact - Detour Lake Mine 19Figure 2.7 Primary Development - Detour Lake Mine 20Figure 2.8 Mechanized Cut and Fill Stope Development Schematic View 21Figure 2.9 Cut and Fill Breast Drilling and Loading - Detour Lake Mine 22Figure 2.10 Backf11 Grain Size Distribution - Detour Lake Mine 23Figure 3.1 Underground Roof Failure Mechanisms 57Figure 3.2 Beam Failure Mechanisms 58Figure 3.3 Plate Failure 59Figure 3.4 Voussoir Block Theory 60Figure 3.5 Voussoir Arch Failure Modes 61Figure 3.6 Voussoir Plate Failure Modes 62Figure 3.7 Wedge Failure 63Figure 3.8 Stereonet Analysis ofWedge Failures in Back 64Figure 3.9 Stereonet Analysis of Wedge Failure in Walls 65Figure 3.10 UNWEDGE Analysis of Wedge Failure 66Figure 3.11 Chimney Failure 67Figure 3.12 Progressive Rock Mass Failure 68Figure 3.13 Relationship Between the NGI-Q and Equivalent Dimension Dc 69Figure 3.14 Relationship Between the NGI-Q and Span for Temporary Mine Openings 70Figure 3.15 Relationship Between Span, RMR, and Stand-up Time 71Figure 3.16 Comparison of Span Versus Stand-up Time for NGI and RMR Data 72Figure 3.17 Variation of Rock Load as a Function of Span 73Figure 3.18 Variation of Rock Load Height as a Function of Span 74Figure 3.19 Modified Stability Graph 75Figure 3.20 Mining Rock Mass Rating Versus Hydraulic Radius 76Figure 3.21 Scaled Span Versus Rock Mass Rating 77Figure 3.22 Crown Pillar Study Data Analyzed in Terms of Span 78Figure 3.23 Stress Induced Failure in Pillars 79Figure 3.24 The Boundary Element Method - 2D 80Figure 3.25 EXAMINE-2D Output - Detour Lake Mine 81Figure 3.26 BEAP-3D Output - Detour Lake Mine 82Figure 3.27 The Finite Element Method - 2D 83Figure 3.28 The Distinct Element Method - 2D 84Figure 3.29 Span Versus Rock Mass Rating at Other Canadian Operations 85Figure 3.30 Span Design Methodology at Detour Lake Mine 86Figure 4.1 Detour Lake Mine UCS Testing - Main Zone and Talc Zone 99Figure 4.2 Detour Lake Mine - UCS Testing of Hangingwall Mafic Rock 100Figure 4.3 Detour Lake Mine - UCS Testing by UBC of Main Zone and Talc Zone 101Figure 4.4 Detour Lake Mine - UCS Testing by UBC of Hangingwall Mafic Rock 102Figure 4.5 Detour Lake Mine - UCS Testing Summary 103Figure 4.6 Geotechnical Mapping Form 104Figure 4.7 Pole Plots from Fabric Mapping 105viiFigure 4.8 Pole Plots from Fabric Mapping.106Figure 4.9 Contours of Pole Plots - DLM Mapping 107Figure 4.10 Contours of Pole Plots - DLM Mapping 108Figure 4.11 Contours of Pole Plots - DLM Mapping 109Figure 4.12 Contours of Pole Plots - DLM Mapping 110Figure 4.13 Contours of Pole Plots - All Structures in Main Zone and Talc Zone 111Figure 4.14 Typical Roughness Profiles for JRC Ranges 112Figure 4.15 Fabric Mapping - Joint Roughness Frequency and Rock Hardness Frequency 113Figure 4.16 Fabric Mapping - Joint Aperture and Spacing 114Figure 4.17 Fabric Mapping - Joint Continuity and Ends Visible 115Figure 4.18 Rock Mass Classification Form 116Figure 4.19 In-Situ Stresses in the Canadian Shield 117Figure 4.20 260 Sill Pillar Stress Monitoring 118Figure 4.21 360 and 460 Sill Pillar Stress Monitoring 119Figure 4.22 Surface Crown Pillar and 560 Sill Pillar Stress Monitoring 120Figure 4.23 Detour Lake Mine Production Schedule 121Figure 5.1 Post Pillar 941 BEAP-3D Model Geometry 133Figure 5.2 Post Pillar 941 - Major and Minor Principal Stresses at Mid-Height of Pillar 134Figure 5.3 Post Pillart 941 - 15 m Height Major and Minor Principal Stresses at Mid-Height 135Figure 5.4 Post Pillar 941 BEAP-3D Pillar Stress Versus Pillar Height 136Figure 5.5 Detour Lake Mine BEAP-3D Model Geometry 137Figure 5.6 Principal Stresses on Excavation Boundaries Detour Lake Mine 138Figure 5.7 Principal Stresses in Crown Pillar - Detour Lake Mine 139Figure 5.8 Principal Stresses in 260 Sill Pillar of Detour Lake Mine 140Figure 5.9 Principal Stresses in 360 Sill Pillar of Detour Lake Mine 141Figure 5.10 Principal Stresses Between Attack Ramps of 360 Stope 142Figure 6.1 Span Definition 166Figure 6.2 Raw Database - Detour Lake Mine 167Figure 6.3 Corrected Database - Detour Lake Mine 168Figure 6.4 Bivariate Ellipses on Corrected Data 169Figure 6.5 Data Density Contours by Group 170Figure 6.6 Data Density Contours of All Data 171Figure 6.7 Data Distribution Probability Plots by Group 172Figure 6.8 Data Distribution Dot Plots by Group 173Figure 6.9 Statistical Classification of Data Into Three Populations 174Figure 6.10 Original Group Classification Lines 175Figure 6.11 Modified Group Classification Lines 176Figure 6.12 Stability Graph for Entry-Type Excavations 177Figure 6.13 Stability Probability Graph 178Figure 6.14 Comparison of Bieniawski Data to Stability Graph for Entry-Type Excavations 179Figure 6.15 Comparison of Stable-Unstable Boundary Modified Bieniawski Curve 180Figure 6.16 Comparison ofNGI Span Design Curves to Stability Graph for Entry-Type Excavations.. 181Figure 6.17 Comparison of Golder Crown Pillar Data to Stability Graph for Entry-Type Excavations.. 182Figure 6.18 Ground Movement Monitor 183Figure 6.19 Typical GMM Response in Unstable Ground 184Figure 6.20 Mine Spider Ground Monitoring Device 185Figure 7.1 Bolt Factor Design Chart 198Figure 7.2 Effect of Support on the RMR Classification Value 199viiiFigure 7.3 Key Block Support Using Rockbolts.200Figure 7.4 Pre-Support of Cut and Fill Stopes Using Grouted Cable Bolts 201Figure 7.5 Water:Cement Ratio Versus Grout Strength and Pull-out Strength 202Figure 7.6 Typical Cable Bolt Grouting Arrangement for Up-Holes 203Figure 7.7 Strength Components of a Grouted Cable Bolt System 204Figure 7.8 A Summary of the Development of Cable Bolt Configurations 205Figure 7,9 Pull-out Strengths for Various Cable Bolt Configurations and Embedment Lengths 206Figure 7.10 Design Chart for Cable Bolt Density 207Figure 7.11 Typical Cut and Fill Stope Layout Using Post Pillars 208Figure 7.12 Progressive Rock Mass Failure 209Figure 7.13 Vertical Stress in a Post Pillar Versus Pillar Height 210Figure 7.14 Schematic Cross Section of a Post Pillar Cut and Fill Stope at Detour Lake Mine 211Figure 7.15 Location of Cable Bolt Support Trial 212Figure 7.16 Post Pillar 941 Plan Profiles 213Figure 7.17 Pillar Strength Versus Width:Height Ratio for Different Rock Masses 214Figure 7.18 Stereonet Projection of Structure Mapped Around Pillar 941 215Figure 7.19 Schematic Cross Section - Location of Instrumentation 216Figure 7.20 Photographs of Cable Bolts and Post Pillar 941 Prior to Pillar Blast 217Figure 7.21 Triple Anchor Wirefiex Extensometer Monitoring Heads 218Figure 7.22 Ground Movement Monitor (GMM) Measurement of Back Displacement 219Figure 7.23 Footwall Stress Change Monitoring 220Figure 7.24 Cable Bolt Load Monitoring - Long Term Monitoring 221Figure 7.25 Cable Bolt Load Monitoring - First 24 Hours 222Figure 7.26 Tensmeg Strain Gauge Load Calibration Curve 223Figure 7.27 Extensometer Monitoring of Back Displacement 224Figure 7.28 Effect of Cable Bolt and Post Pillar Support on Span 225ixACKNOWLEDGMENTSThe author would like to extend his sincere appreciation to Dr. Rimas Pakalnis for his valuableguidance and comments during the research and preparation of this thesis. Special thanks must be given tothe management of Detour Lake Mine for co-sponsoring this research and particularly to Mr. JackMacRory, Mr. Ron Moran, and Mr. Ken Cook for their valuable assistance. Appreciation is also extendedto Dr. Somchet Vongpaisal and CANMET for agreeing to co-sponsor this research. Finally, thank-you tomy wife Polly for her patience, understanding, and encouragement over the past several years.x1. INTRODUCTIONIn 1989, Placer Dome Inc.’s Detour Lake Mine undertook a major research focus in developing“Design Guidelines for Cut and Fill Stopes” in conjunction with CANMET. These guidelines makespecific reference to optimum stope dimensioning, ground support, mine sequencing, and pillar extraction.This thesis will focus on the span design portion of this research project and the role that support can havein increasing the allowable span of entry-type excavations in general.In designing spans for entry-type excavations, there are two limiting constraints which influencethe design. First, the nature of entry-type mimng is such that workers are exposed to freshly blastedground, unlike non-entry stopes. Therefore, higher safety factors are required for the design of entry-typestope spans. Secondly, profitable mining often demands the maximum extraction of the ore, which isachieved by maximizing the spans between pillars. In addition, stope excavations are required for only ashort duration and therefore the high safety factors which would be used for permanent underground civilengineering structures would be difficult to justify. This thesis will attempt to reconcile these conflictingdesign objectives by providing for the mining engineer a practical design tool developed specifically forspans in entry-type excavations.1.1 BACKGROUNDIn recent years, entry-type mining methods such as cut and fill, room and pillar, and shrinkagestoping have been replaced in many mining operations by lower cost, non-entry, bulk mining methods. Inmany mines, however, the nature of the orebody is such that more selective, entry-type mining methods arestill desirable. In 1989, cut and fill stoping and other entry-type mining methods still accounted for 37.1%of the total tonnes of ore extracted from underground metal mines in Canada (CMJ, 1990). Over the last50 years, Canadian mines have pioneered many innovations in cut and fill mining technology, includingrock fill, cemented fill, undercut and fill, and post pillar cut and fill mining (Singh et al., 1980). This needfor innovation will certainly continue as existing orebodies become depleted and mining reaches greaterdepths.Improved design procedures developed particularly for entry-type mining methods can result inthree major benefits for mining operations:• improved worker safety;• increased ore recovery; and• reduced dilution.1In its presentation to the Provincial Inquiry into Ground Control and Emergency Preparedness inOntario Mines in 1985, the Ontario Ministry of Labour provided statistics on mining related injuries overthe twenty-two year period ending in 1984 (Stevenson, 1986). The statistics indicate that fails of groundare the single highest cause of death in the mining industry in Ontario. Two-thirds of these fatalities occurwhile scaling, drilling, or from falling pieces of loose. These categories are predominantly associated withwork tasks at a freshly blasted face, as are encountered with-entry type mining methods. The trendtowards bulk mining techniques, as well as mechanized scaling, bolting, and drilling in entry-type stopes islikely to reduce the accident frequency in coming years.Improved design procedures and the use of alternative support measures can increase ore recoveryin entry-type stopes. In cut and fill stopes for example, post pillars are commonly left in the ore as a meansof support. This report will show how the use of cable bolts was successful in maintaining support in a cutand fill stope after the post pillar was mined out. In the future, the mining industry may face increasingsocial pressures to maximize extraction of the public resource they are licensed to exploit. Alternatesupport measures such as this may become more widespread if this is the case.Iniproved excavation design, mining techniques, and support methods can contribute to reduceddilution in entry-type stopes. While dilution in entry-type stopes is usually lower compared to open stopes,in the event of large failures, considerable waste may have to mined before the stope can be rehabilitated.There are no suitable methods for designing large open spans for entry-type stopes in jointed rock.Beam and plate theories, Voussoir block analysis, and numerical models which are described in Chapter 3,have been employed in the past. In general, however, they have been adopted from the field of civilengineering and are restricted by homogeneous, isotropic, and linear elastic assumptions about the rockmass. More recently, an empirical design method has been developed for the design of spans in non-entrystopes and has gained widespread acceptance in the mining industry in Canada. This method would not besuitable for entry-type mining methods, since the definition of stable in a cut and fill stope is much moreconservative then what is considered stable in a longhole stope. Other empirical methods have beenproposed as general purpose span design techniques for a range of excavations from temporary mineopenings to nuclear power stations. In general, however, they have been derived from a database consistingprimarily of civil engineering case histories which require long term stability and higher safety factors thanthose required for entry-type stoping.21.2 RESEARCH METHODOLOGYThe first phase of this research involved a questionnaire sent to underground cut and fill operatorsin Canada to determine what type of span and pillar design methods were being practiced in Canadianmines. From the returned questionnaires, it was evident that there is not a commonly accepted method usedby mining engineers to design stable excavation spans. Undocumented rule-of-thumb approaches andpast-practice plays a large part in the design procedure at most mines. The problem with these proceduresis that the experience of mining one orebody is not readily transferable to other orebodies. The results ofthe questionnaire did suggest that an empirical design method which would quantifr these rule-of-thumbapproaches, would be the best design method for predicting conditions of stability under varying conditionsof rock quality and stope geometry. Empirically based design methods are gaining increasingly widespreadacceptance in the mining industry. Procedures have been developed for such areas as:• open stope dimensioning;• cable bolt support design;• prediction of dilution in open stopes;• prediction of stand-up time; and• support requirements.The second phase of the project involved collection of span, rock quality, and stability data from alarge number of cut and fill stopes to establish the empirical database. Placer Dome Inc.’s Detour LakeMine, as a co-sponsor of this research project, provided access to its operation for the purpose of gatheringthese measurements over the period from December, 1989 to March, 1992. In addition, the mine madeavailable a large database of stope span, rock quality, and stability data gathered at the mine before theproject began. This information has been compiled on a Stability Graph for Entry-Type Excavationswhich plots the design span versus rock mass rating. The data was analyzed statistically to define regionson the graph as stable, potentially unstable, or unstable. This graph provides for the mining engineer apractical means of designing stable spans for entry-type stopes. The design procedure recognizes the needfor a comparatively low safety factor that is required for short-term underground excavations.The role of support systems such as post pillars and cable bolts is assessed and their affect on spanis studied. Post pillars have been used successfully at Detour Lake Mine and elsewhere to increase theoverall span which can be mined before instability occurs. To achieve greater ore recovery and miningefficiency, the Detour Lake Mine sought to replace the support provided by the post pillar with cable boltsupport. From a research perspective this work would provide a means of estimating the increase in spanwhich can be made possible with artificial support. A trial support project undertaken by Detour LakeMine and described in Chapter 7 demonstrated the effect of replacing a post pillar with a concentratedcable bolt pattern.3It is intended that the empirical span design procedure proposed in this thesis be used as part of anintegrated design philosophy which also combines analytical procedures, numerical modeling, andengineering judgment. This design approach as it is now practiced at Detour Lake Mine will be describedin further detail in Chapter 3.42. THE DETOUR LAKE MINE2.1 INTRODUCTIONPlacer Dome Inc.’s Detour Lake Mine (DLM) began production in 1984 as an 1800 tonne per dayopen pit gold mine (Figure 2.1). The pit reached an ultimate depth of 130 metres in 1987, at which timeproduction commenced from underground operations. Mining is carried out using mechanized cut and fill,longhole, and captive cut and fill techniques. Approximately 80% of production at the time of this studycame from mechanized cut and fill stopes, 10% from longhole, and 10% from captive cut and fill stopes.Productivity improvements and increased milling capacity have boosted the production rate to 2200 tonnesper day. Current ore reserves stand at approximately 6.0 million tonnes, sufficient for another 6.5 years ofproduction. The orebody has been proven to a depth of 660 metres below surface and is open along strikeat depth. The mine is serviced by an all-weather road from Cochrane, Ontario, and a gravel air strip at thesite. Most employees commute by bus to the mine and work a schedule of seven days in and seven daysout.2.2 GEOLOGY2.2.1 Regional GeologyThe Detour Lake Mine is located on the northwest rim of the Abitibi greenstone belt, which hosts aseries of Archean felsic, mafic, and ultramafic tuffs, flows, and intrusions, as well as volcaniclastic andchemical sediments (Miller, 1988). The deposit lies on the north limb of an east-west striking anticline.The lithologies strike on azimuth 070° to 080° and dip 60° to 80° north. Most of the ore discovered to dateis located at or adjacent to the contact between mafic and ultramafic rocks.The hangingwall rocks located to the north of a so-called chert horizon are iron rich maficvolcanics with increasing potassic alteration closer to the chert. The rocks to the south of the chert aremagnesium rich mafic and ultramafics which have been identified as chioritic greenstone and talc-chloriteschist. Figure 2.2 is a plan view of the lithologic units associated with the Detour Lake Mine orebody.2.2.2 OrogenyThe chert may have been formed as a mylonite zone, a chemical sediment, or a deformed felsicintrusive. Irrespective of its origin, the chert marker horizon occurs along a break in the stratigraphy thatestablishes the boundary between plastically deformed rocks to the south and brittly deformed rocks to thenorth. Overall, the mineralization appears to have developed in a wide fault zone that displays sinistralmovement. Vertical movement is reverse, with the north block moving up relative to the south. This faultzone subsequently served as a conduit for mineralizing fluids.5The faulting was probably a response to the regional stress field in the vicinity of the Detour LakeMine. The faulting strikes approximately 232° and dips 45°-60° north. This coincides with the majorprincipal stress direction, which at DLM has a measured azimuth of 257° and an inclination of 32° north.In the brittle hangingwall rocks, the faulting is manifested as small-scale folds and flexures in the chert andquartz veins. In the plastically deformed footwall rocks, it results in small-scale prolate boudins. Figure2.3 is a simplified geological model of the rock in the vicinity of Detour Lake Mine illustrating the directionof shearing relative to the major principal stress direction.2.2.3 Mine GeologyThere are three interrelated gold bearing zones, namely the Main Zone, Quartz Zone, and the TalcChlorite Zone. Figure 2.4 illustrates the relationship of these three zones on the 360 metre Level. MainZoneThe Main Zone contains 72% of proven reserves. It strikes east-west and has an average dip of60° north. The Main Zone is lens shaped, with widths of up to 45 metres in the centre and pinching downto under 5 metres at each end. Above the 560 metre Level horizon, the orebody plunges at 45° west andhas an average strike length of 200 metres. Below this elevation, the plunge gradually flattens to horizontaland the strike length increases (Figure 2.5). Present drilling indicates that the ore bottoms out atapproximately 760 metres below surface. The orebody is open along strike below this depth.The most persistent ore bearing feature is the chert horizon which dips at 60° north. Verticallydipping quartz-sulphide veins splay off this chert into the hangingwall. Immediately adjacent to the chert,the quartz-sulphide veining is quite dense. The veining pinches out, and in some cases the grade decreases,with increasing distance from the chert. Dropping these veins from the mining limit as they become too farfrom the chert accounts for the stepped hangingwall of stopes as illustrated in Figure 2.6. The footwallcontact is more regular on strike than the hangingwall contact but it undulates locally, with dips varyingfrom 30° to vertical. Ouartz ZonesThe ore in the Quartz Zones, which comprises 5% of proven reserves, exists as gold bearing veinswhich diverge north from the chert but continue to carry grade for long distances. The ore generallyconsists of three to five 10 cm thick veins with a combined average width of 3 metres. The Quartz Zoneshave an east-west strike, a dip of sub-vertical to vertical, and a plunge of 45° west. Mining of the QuartzZone is carried up as part of the Main Zone using mechanized cut and fill for up to 50 metres along strike,depending on scheduling constraints. Whatever ore remains further along strike is extracted by longholemethods. Talc ZonesThe Talc Zones are located to the south of the Main Zone and comprise 23% of proven reserves.They are discontinuous in plan and section with dips undulating locally from 20° to 700 north.Mineralization occurs in a relatively weak talc-chlorite schist which requires more ground support than theMain Zone. Strong fault structures containing several feet of gouge material control the distribution ofmineralization. Mafic intrusive dykes ranging in width from 0.3 to 5 metres commonly cut through themineralization. These features combine to make geological control and mining of the Talc Zones difficult,however, this can be offset by the generally higher grade.The Talc Zones are mined simultaneously with the Main Zone using mechanized cut and filltechniques if they are close enough and if it is economical to do so. If the Talc Zones are too far from theMain Zone or if they are too narrow, they may be mined using captive cut and fill methods.2.3 MINING METHODS2.3.1 Primary DeveiopmentThe mine is accessed from the hangingwall side of the ore body by a three compartment shaft sunkto a depth of 615 metres. Main levels have been driven from the shaft at 100 metre intervals. Five mainmining levels have been established, namely:• 260 Level;• 360 Level;• 460 Level; and• 560 Level.In addition, a connecting ramp system extends from surface to the 660 Level. Between each mainlevel, two sublevels are driven in the hangingwall at 30 metre elevation intervals to access the orebody(Figure 2.7). The Main Zone is accessed by means of two crosscuts, referred to as attack ramps, drivenfrom the hangingwall drift. Broken muck is hauled up the attack ramps to a central orepass system whichpasses ore to the 430 Level, where it feeds a crusher located on the 460 Level. Below the 460 Level, a fineore bin feeds a loading pocket from which ore is skipped to surface.Mechanized cut and fill stoping accounts for approximately 85 percent of the mine’s output.Longhole stoping of the Quartz Zones contributes 10 percent to the mine’s output and captive cut and fillmining of some of the Talc Zones provides the remaining 5 percent.72.3.2 MCF Stope DevelopmentMechanized cut and fill stopes were started on the 260, 360, 460, and 560 Levels. Attack rampswere driven from these levels at +3% to facilitate drainage. Typically, a stope is mined in two halves, eachaccessed by a separate attack ramp. As one half is mined, the other half is filled. In this way, a constantmining rate in each stope can be assured. When a lift has been completed, the attack is backslashed toprovide access to the next lift. This continues until the attack finally reaches an inclination of 20%, atwhich point another attack is driven at minus 20% from the next level, 30 metres above (Figure 2.8).Sill pillars are left beneath each of the stopes on the 260, 360, 460, and 560 Levels, and a crownpillar is maintained between the 260 stope and the pit. The ultimate pillar thickness is variable, dependingon stope span, rock quality, and stress conditions.2.3.3 Drilling and BlastingTwo boom hydraulic jumbos drill horizontal breasts 5 metres high and 3.6 metres deep. The facecan be up to 35 m wide, depending upon ground conditions and ore limits. 45 mm diameter holes aredrilled on a 1 metre square pattern. The holes are loaded with pneumatically placed ANFO initiated byNonel detonators.A number of ground problems experienced at Detour Lake Mine have been attributed to poordrilling and blasting practices. For this reason, particular emphasis is placed on drilling flat, parallel backholes for ground control purposes. Back holes are spaced 0.6 metres apart and loaded with 25 mmTrimrite cartridges to produce a decoupled charge. Figure 2.9 shows the location, loading, and sequencingof holes in a typical breast face at DLM. Technical details of cut and fill blasting at DLM are provided inTable MuckingFive-yard LI-ID’s and 26-tonne trucks are used to muck out stopes. The ore is trammed to orepasses which intersect the hangingwall ramp and pass the ore to the 430 Level, where it is transferred to thecoarse ore bin for crushing. Each lift is mucked out to within 0.4 to 0.6 metres of the sand fill in order tomaintain a good mucking floor for the equipment to work on. Before a stope is filled, the remaining muckis removed down to the fill level.82.3.5 Ground SupportIn accordance with the Ontario Ministry of Labour’s Policy on Ground Support, the freshly blastedarea is scaled and ground support is rnstalled immediately after the area is mucked out. Methods ofartificial ground support in use at DLM include:• mechanically anchored rockbolts;• Swellex bolts;• cable bolts;• wire mesh; and• steel straps. Mechanically Anchored RockboltsMechanically anchored rockbolts are used almost exclusively in development headings, QuartzZones, and the Main Zone. Rock in these areas has a compressive strength of approximately 165 MPamaking bail and wedge anchors effective. Spalling of rock around the collar of the hole is not a problem inthese zones. For these reasons, the mechanically anchored rockbolts, when used with a steel plate, provideexcellent active support. The purposes of the mechanically anchored rock bolt are:• To provide immediate support to potentially unstable blocks which cannotbe removed by scaling; and,• To support key-blocks at the immediate surface of the excavation which inturn provide geometric support to the overlying rock.The technical specifications of the mechanically anchored bolts employed at DLM are provided inTable 2.2.Rockbolting is carried out either with stopers operated from scissor lift vehicles or from arockbolting jumbo. A standard 1.2 metre rock bolt spacing is used which was found through experience tobe suitable for most of the DLM rock mass. In areas where joint spacing is as little as 0.3 metres, a 1.0 msquare pattern has been applied.Quality control on the installation of these bolts is maintained through routine torque testing bybolting crews and supervisors and pull testing carried out by the engineering department. Swellex RockboltsSwellex rockbolts are often preferred in the Talc Zones because the low rock strength does notpermit proper anchorage with a mechanically anchored bolt. Spalling of the drill hole collar often occurs,which also renders mechanical rockbolts ineffective. The Swellex bolt is a friction stabilizer, providinganchorage along the entire length of the bolt. This holds the rock together, reduces joint separation, andultimately helps the rock mass to support itself geometrically. The standard 1.2 metres pattern is also used9for Swellex bolts. Quality control on the installation of Swellex rockbolts is maintained by pull-out testsconducted by the engineering department.Super swellex bolts, similar to the standard swellex, are installed in areas where a potential wedgehas been identified and which cannot be supported by standard 1.8 metre bolts. Super swellex aremanufactured from a thicker steel and have a larger diameter than standard swellex. The installationpattern is designed for the specific failure geometry they are being used to stabilize.Technical specifications for standard and super Swellex bolts are provided in Table Cable BoltsCable bolting is an effective means of stabilizing and supporting rock masses too large forconventional bolts and for pre-supporting cut and fill stopes. Cable bolts can be cut to any length andinclude end-holding devices that allow for ease of installation in up-holes. A typical cable bolt is madefrom 16 mm diameter. 7 strand, stress relieved degreased steel cable, having an ultimate strength of 25tonnes. Cable bolts are grouted with a 0.4 water:cement mixture by weight.Detour Lake Mine has carried out a number of support trials involving regular steel cables, steelbirdcaged cables, and fibreglass birdcage cables. The mine has also demonstrated the use of concentratedcable bolt support as a means of replacing post pillars for support of wide spans in cut and fill stopes. Thiswork will be discussed in Section 7 of this report. Wire MeshWire mesh, commonly 5 cm x 5 cm galvanized chain link mesh or 10 cm x 10 cm weld mesh, isused in high traffic areas, shops, refuge stations, and areas of stopes with excessive small pieces of loose.The main purpose of the chain link mesh is to prevent injury to personnel or damage to equipment bycontaining small pieces of loose. In high traffic and work station areas where personnel and equipment areoften present, the mesh is installed as a long term safety measure to contain loose which may fall over anextended time period. The mesh is generally used in conjunction with mechanical rockbolts. A woodenplate is inserted between the mesh and the steel rock bolt plate to prevent cutting of the screen. Steel StrapsSteel straps are used to prevent joints or cracks from opening and to reinforce pillar corners. Dueto the higher costs of this type of support, they are used on a limited basis and only where local conditionsmake them necessary. Locations of installation are generally a front-line supervisor’s decision. Steel strapsused at DLM are made of 6mm thick steel, 100mm wide and vary in length from 1.2 m to 2.4 m. Thesteel straps are pinned with mechanically anchored or resin/rebar bolts.102.3.6 BackfiuingWhen a lift has been mined out to the ore limits, the remaining 0.4 to 0.6 metre layer of muck isscraped off the floor down to the backfill. Any waste muck mined on or near the level is then placed in thestope prior to placing the hydraulic fill. The hydraulically placed backfill is contained by building up abulkhead with muck to a level slightly higher than the planned fill level and covering it with fabrene. Thefill is normally placed to within one metre of the back.The sand fill used by Detour Lake Mine is obtained from a nearby esker sand borrow pit onsurface. Percolation tests and sieve analysis are performed on the sand in the pit to ensure an adequatepercolation rate before it is excavated. Sufficient sand is stockpiled during summer near the backfill plantfor use throughout the year. The backfill is mixed to 65% solids and delivered at a rate of 100 tonnes perhour. It is transported to the stopes from surface through a combination of drill holes and 10 cm diameterSclairpipe. The backfill mixture drains quickly, having an average percolation rate of 80 cm/hr. No othermechanisms to assist drainage are required. The drain water is collected in small sumps where it iscollected and drained through drill holes to a large sump on the 460 Level. It is pumped in a single stageinto the open pit on surface. Table 2.4 summarizes the characteristics of the backfill sand used at DLMand Figure 2.10 shows a typical sieve analysis.The first lift of each mechanized cut and fill stope was filled with a 10:1 sand to cement mixture tofacilitate sill pillar recovery. Rebar and screen were placed in the fill for added strength.11Table 2.1 Drillin2 and Loadin2 Specifications for Cut and Fill Breastin2BoreholesDiameter 45 mmLength (Drilled) 3.6 mL.ength (Loaded) 3.4 mBurden l.OmSpacing 1.OmExplosives1NFO (Nilite) for production holesl5 mm emulsified ANFO (Trimrite) in perimeter holesiIaximum Number of Holes per Delay = 201aximum Weight of Explosives per Delay = 105 kgTable 2.2 Technical Specifications of Mechanically Anchored Rockboltsrype LH Thread, Bail Type shell, forged head or threaded botimds, ASTM-F43Z-83 Standardteel Diameter 16 mmtield Load 12.5 tonnes.Jltimate Loads 16.3 tonnes3olt Lengths 1.8 m, 2.4 m, 3.0 m12(000H-.(D0 0—(JD0liiiliiiliii-IIII0001111iiIIIIb1IIL Iliii1cIIIIIIII-1IICD -t7’LOCATION MAP - DETOUR LAKE MINE FIGURE 2.1HUDSONBAYTOUR LAKE MINEQUEBEC0 200 400 km14SCALE: 1:2000ILEGENDMF = IRON RICH MAFIC VOLCANIC FLOWSKMF = POTASSIUM ALTERED MAFIC VOLCANIC FLOWSCHERTKB = KOMATITIC BASALTCG = CHLORITIC GREENSTONETC = TALC CHLORITEFANCF-IERTPLAN VIEW OF LITHOLOGIC UNITSDETOUR LAKE MINEFIGURE 2.2HANQUARTZ ZONESMFTCTALC ZONES} THOELIITIC BASALTSMAGNESIUM RICH MAFIC ANDULTRAMAFIC VOLCANICS15c=O.O29 MPa/mStrike of3 *FIGURE 2.3Dip of Qrebody=2.6*;Dip of Stra75igraphySTRUCTURAL GEOLOGY MODEL AND ORIENTATION OF PRINCIPALSTRESSES AT DETOUR LAKE MINE16MAIN ZONE0 5 50PLAN VIEW OF ORE ZONESDETOUR LAKE MINEFIGURE 2.4qO°QUARTZSTRINGERSTALC ZONESScale17LONGITUDINAL SECTION OF MAIN ZONE FIGURE 2.5DETOUR LAKE MINELOOKING NORTH18Plan ViewgingwalI4-MAIN ZONEFootwallSTEPPED HANGINGWALL CONTACT OF MAIN ZONEDETOUR LAKE MINEThese veins are ot mined outon Lift 3 as there is too muchwaste rock between them andthe Main Zone.Hangingwall Contact\MainZone\Section A-A5 \\,4 1Quartz ZonestringersFIGURE 2.619tn H> H C rnLA a’ r m r‘1b 7‘10 •1 C •1‘I‘I I ‘I ‘1 ‘1•15, I., C r I” rI) C, 3 II rdPi2-’-m 6’C I—2n>r’17 I,, z P1DETOUR LAKE MINEBACKSLASHED ATTACK RAMP (ROCKFILL)— PILLAR— ORE REMAININGACTIVE STOPE— DEVELOPMENT HEADINGSc:ii FILl.MECHANIZED CUT AND FILL STOPE DEVELOPMENTSCHEMATIC VIEWDETOUR LAKE MINEFIGURE2.8SCHEMATIC CROSS SECTIONOPEN PITCROWN PILLAR260 SILL PILLAR360 SILL PILLAR21Perimeter Hole Spacing = 0.65 m45 mm holes loaded with 25 mm Trimnte0000000000000000000000000000000000000000000000000000000000000000 00000000 000000000000000000000 000000000000000000000000000000000000000000000000000000000000000CUT AND FILL BREAST DRILLING AND LOADINGDETOUR LAKE MINEFIGURE 2.922CC HParticleSizeDistributionC C CParticleSize(mm)3. REVIEW OF DESIGN METHODOLOGIES3.1 LITERATURE REVIEWThe design of underground structures is a relatively recent practice compared with the time manhas been mining underground (Obert, 1973). The first types of design were simple rule-of-thumbapproaches which are practiced even to this day. Requirements for greater mining efficiency and highersafety standards have made necessary a more reliable and effective approach to the design of undergroundstructures.Many design methods have been developed over the years and they can be classified into threecategories:• Analytical Approximations;• Numerical Simulations; and• Empirical Methods.Analytical approximations include closed form solutions, limit equilibrium techniques, photoelasticmodeling, and physical modeling. Analytical methods usually involve gross simplifications of theexcavation geometry and rock properties. These simplifications can place severe restrictions on theirapplication to real mining problems.Numerical simulations, based on finite difference, finite element, boundary element, or distinctelement methods, are gaining increased usage thanks to the availability of software and the low cost forpersonal computers. Numerical simulations are becoming increasingly sophisticated with 3-D modelingpackages now commercially available at reasonable cost.Empirical design methods, which involve the application of knowledge based on documentedexperience with similar mining conditions, are gaining increased acceptance in the mining industry. Thisrequires a database of observations relating the stability of the underground structures to mine geometry,the rock mass characteristics, and other factors which influence stability. Empirical methods have beenmade possible in part by widespread acceptance of rock mass classification systems.These design methods will be discussed in detail as they are applied to the six failure modes whichaccount for instability of underground openings namely:• beam or plate failure;• Voussoir block failure;• wedge failure;• chimney failure;24• rock mass failure; and• stress induced failure.These failure mechanisms are illustrated in Figure Beam and Plate FailureBeam and plate failure analyses assume the rock mass behaves as an elastic beam or plate. Theanalysis methods have been adapted from civil engineering solutions for bending of homogeneous,isotropic, and linear elastic materials such as concrete. Obert et al. (1967) provide a good treatment ofbeam and plate failure analysis. In applying this type of analysis to the stability of underground structures,the following simplifying assumptions must be made:• In the case of beam failure, the strike length of the opening must betwice the width (beam span); and in the case of plate failure analysis,the strike length of the opening must be between 0.5 and 2 times thewidth;• The rock must be hard, massive, and free ofjointing to a degree that itis reasonable to consider it as homogeneous, isotropic and linearelastic;• The beam must be continuous with the stope walls so the beam endsare considered to be fixed;• In the case of beam bending, no loads are applied along the strike(plane strain condition); and,• The beam is considered to have a uniform thickness. BeamFailureThe two potential failure modes for beams are tensile (flexural) failure and shear failure, asindicated in Figure 3.2.(a) Tensile FailureFor a horizontally layered roof, subjected to gravity loading, the maximum tensile stress isgiven by:-Ya52 pS2— 2t + 2t (3.1)where,amaxmaximum tensile stress in beamS = span of roof layerp = any uniformly distributed load (i.e. a filled stope above)= adjusted unit weight of the lowest strata25t = thickness of roof layerThe adjusted unit weight (Ya), of the lowest strata used in the formula above to account for theweight of overlying strata, is calculated as follows:—E1t(y+y2t+y3t3...+yt)Ya 3 3E1t+E2t+E...+Et (3.2)where,= Young’s Modulus of nth layer= unit weight of nth layert = thickness of nth layerSetting 0max = R0, where R0 is the Modulus of Rupture (outer fibre tensile strength), andrearranging, the following span design equation for shear failure of horizontal strata can be derived:I 2t(R —a0)S=Ps(ya+p/t) (3.3)where,R0 = modulus of ruptureF = factor of safety(b) Shear FailureWhen the ratio of strata thickness to span exceeds approximately 0.2, shear failure begins todominate over flexural failure (Obert et al., 1967). For a horizontally layered roof subjected to gravityloading, the maximum shear stress is given by:tmax (34)The shear strength t of the beam is given by:t=c +a(tanp) (35)where,an = horizontal compressive stressc’ = cohesion on the plane of shear acting over the compressed zone= friction angle on the plane of shear26By rearranging these equations, and defining the factor of safety F5 as the ratio of shear strength toshear stress, the maximum allowable span to resist shear failure of the back in horizontally bedded strata isgiven by:4(c+a tan’)2 V (3.6)-‘Ya’s3.1.1.2 Plate FailureIn cases where the orebody has a strike length to width ratio of between 0.5:1 to 2:1, the backshould be treated as a slab spanning two directions with fixed support on all four sides (Figure 3.3). Themaximum tensile stress for such a plate occurs at the centre of the long edge. The maximum tensile stressfor bending of such a plate is given by:I3qSbumax—(3.7)where,f3 = coefficient which varies with the span ratio (See Figure 3 .3b)S = short span of plateb = long span of platet = thickness of plateq = loading on plate per unit areaSetting amaxRo, the maximum stable span is given by:2s — amaxt— I3qbF5 (3.8)Shear failure of the plate is analogous to chimney failure, which will be discussed later.3.1.2 Voussoir Block Failure3.1.2.1 Voussoir Beams(Evans, 1941) was the first to consider analyzing stope backs as discrete blocks as in a masonry orVoussoir arch. It has been recognized since Roman times that arching can greatly increase the load bearingcapacity of a beam. The Voussoir beam model was modified by Beer and Meek (1982) and is illustrated inFigure 3.4. The concept conveyed in this figure is that the line of lateral thrust within such an arch, whentraced on the beam span, approximates a parabolic arch.Voussoir beam theory makes the following assumptions about the rock mass being analyzed:• The rock mass is assumed to be cut by linear discontinuities trendingalong strike, such that the back can be assumed to be composed ofdiscrete blocks;27• It is assumed that there is no horizontal compressive stress in the backtransferred from the surrounding rock; and,• No tensile strength develops between individual blocks (c=0). Analysis ProcedureBecause the solution to the problem is indeterminate, two assumptions are required for theanalysis. First, the line of thrust is assumed to be parabolic, as mentioned; and secondly, the loaddistribution at the centre of the beam and the abutment contact is assumed to be triangular (Figure 3.4(b)).The triangular end load operates over a length nt where,1.5(1—.’l (3.9)tJwhere,n = lateral load to depth ratioz = arch heightt = beam thicknessApplying moment equilibrium around the centroid of the half beam yields:Y 2 fntz ‘yS2—tS = or,=— 3108 2 4nzwhere,= the horizontal compressive stress at the centre of the beam= unit weight of beamS horizontal span of beamAssuming that the shape of the thrust arch acting in the beam is parabolic, the arc length L, can beexpressed by:L=S+3s (3.11)where,L= arc length of parabolic thrust profilez = height of archS = horizontal spanThe resultant force acts through the centre of each force distribution, so the initial moment arm for1’c is given by2 ntz0=t——----- (3.12)where,z0 = initial moment arm of c28t = beam thicknessAs the beam deflects, the arch goes into compression and shortens by a length AL If the archheight z0 is shortened in compression due to AL, the new moment arm (z) can be computed by:v/(3s)[16z _AL] (3.13)The incremental elastic shortening of the arch, AL is given by:ALfL (3.14)where,av = the longitudinal stress in the beamE = Young’s Modulus of the beamThe average longitudinal stress in the beam is now estimated by considering the stresses in only aquarter of the beam, as shown in Figure 3.4(c). At a distance S/4 from the abutment, the stress distributionis uniform over the arch depth. The average longitudinal stressf for this quarter of the beam, and hencefor the entire beam, is given by:1 (2 n‘av=’cLj+) (3.15)An explicit solution for the loading in the beam and beam deformation is not possible. An iterativeprocedure is required, which begins with assuming a value for the initial load to depth ratio, n. An initialvalue ofn0.5 will normally produce a stable solution. The procedure involves calculating sequentiallyf,f, L, AL, z, and n. The process is repeated with the load to depth ratio n, used to calculatef. Iterationscontinue until stable load to depth ratios are obtained.(b) Failure ModesBeer and Meek (1982), identified three possible failure modes for Voussoir arches (Figure 3.5):• Crushing at the hinges formed in the upper portion of the centre of thebeam and at the lower abutment contacts;• Shear at the abutment when the limiting shear resistance T (tan 4) is lessthan the required abutment vertical reaction force V. (W/2); and,29• Buckling of the roof beam with increasing eccentricity of lateral thrustgiving rise to a snap through mechanism.Crushing or compressive failure is analyzed by comparing the maximum longitudinal compressivestressf to the uniaxial compressive strength of the beam. The factor of safety against compressive failureof the beam is then:UCSF= (3.16)The factor of safety against shear failure is defined by the frictional resistance to shearing dividedby the shear stress caused by the weight of the beam. The resistance to shearing is given by:F = Ttan4 = fctit(tati4) (3.17)The abutment shear force (V) is:v==X (3.18)2 2The factor of safety (F5) against shear failure at the abutments is given by:F5=.2.tancp (3.19)ySBuckling failure will occur when the moment arm z becomes negative; that is, when the centroid ofthe centre force distribution is lower than the abutment lateral force distribution. A check must should bemade in the iteration procedure described above to determine if z is negative and, therefore, if bucklingfailure occurs. Voussoir PlatesIn cases where the span to length ratio is greater than about 0.5, plane strain conditions do notapply. It is necessary to consider the roof of the stope as a plate supported on four edges. Tensile crackswill develop on the lower surface of the plate along the lines of maximum tensile stress. Beer and Meek,(1982) suggest this pattern of cracking results in two triangular segments on each end of a rectangularplate, and two trapezoidal segments on the sides (Figure 3.6). For a square plate, four equally sizedtriangular segments would be formed. Referring to Figure 3.6, the shape of the segments is given by:30= + 3 — k] (3.20)where,S short side of plateb = long side of platey = height of triangular segmentk = width to height ratioClearly, since the weight and moment arms of the trapezoidal segments are greater, their behaviourwill control overall roof stability.(a) Analysis ProcedureThe weight of the trapezoidal segment in Figure 3.6 is given by:ySt1 ,W=—b—yj (3.21)The centroid is located at a distance x from the plate edge where,= ( Sb Y! — (3.22)1b—y)4 3SJApplying moment equilibrium about the centroid yields:2(1 yk2 (1 yk fcI1t1ZyS3SyS bt4j,)=2or, c (3.23)The average longitudinal stress acting in the x direction as indicated in Figure 3.6 is given byEquation 3.14. In they direction, f is approximated by:7kffY = C (3.24)av 1231Using equations 3.14 and 3.23, the elastic shortening of the arch caused by deflection of the plateis given by:= favL(1 icy) (3.25)where,v = Poisson’s RatioIn order to solve for the state of stress in the plate, an iterative procedure is followed siniilar to theprocedure for Voussoir beams, except that Equations 3.9 and 3.13 are replaced by Equations 3.22 and 3.24respectively. The rest of the procedure is identical.(b) Failure ModesThe failure modes for plates are similar to those for a beam. Compressive or crushing failure willoccur at the top of the plate at the centre of the span or at the lower side of the abutment contacts. Thefactor of safety is given by Equation 3.1 8.The factor of safety against shear failure is defined by the frictional shear resistance due to themaximum longitudinal compressive stress.f divided by the shear stress due to the weight of the plate. It iscalculated by:F — fnbtan4S— yS(b—y) (3.26)As with Voussoir beams, buckling will occur if the moment arm z becomes negative. A checkshould be made during the iterative process to determine if buckling occurs.3.1.3 Structurally Controlled FailureStructurally controlled failure (wedge failure) is a relatively common occurrence in undergroundmetal mines. Wedges are delineated by intersecting discontinuity planes and the back or wall of anexcavation (Figure 3.7). Failure can occur by sliding along one of the planes in the case of wedge on a wallor by fall-out from the back. The frequency, condition, and orientation of the jointing combined with thesize of the excavation determine the size of potential wedges. The stress level around the excavation canalso influence the stability of wedges; however, most design procedures assume the immediate back to be ina relaxed state.323.1.3.1 Stereonet Analysis TechniquesPotential failure mechanisms can be analyzed quickly using a stereonet. A good introduction to theuse of stereonets for this purpose is provided by Hock and Brown, (1980). The first step is to determine theorientation of the dominant joints sets which are prevalent in the area of concern. In Figure 3.8(a) threejoint sets are plotted on a lower hemisphere, equal angle stereonet. These joints form three release planes,and with the roof of the excavation, form a tetrahedral wedge. It can be seen that the vertical line throughthe centre of the stereonet lies inside the triangle created by the three great circles defining the joint sets.This condition indicates that a vertical free fall of the wedge is kinematically possible.If the three great circles representing the joint sets intersect to form a wedge and the vertical line atthe centre of the stereonet lies outside the triangle formed by the great circles, instability can only bepossible by sliding along one of the discontinuities. In order for sliding to occur, the plane on which slidingtakes place must be steeper than the angle of ffiction which is represented by a continuous circle, as shownin 3.8(b). If the entire triangle falls outside of this circle, the wedge will be stable, In the example shown,the friction circle intersects Joint Set A, so sliding will occur along this plane.Sidewall failure can also be analyzed using stereonet projection. Figure 3.9(a) shows two joint setsand a 70° dipping wall. Since this line represents a wall on each side of the excavation, the failure modeson each side of the line must be assessed. Where two planes intersect in the wall of an excavation, slidingfailure is possible if the plunge of the intersection is less than the dip of the wall and greater than the angleof friction. This condition is illustrated in Figure 3.9(a). On the northeast wall of the excavation, slidingwill occur in the direction of the plunge of the intersection of Joints A and B.Another useful procedure for determining if sliding will occur on a plane or on the line ofintersection of the planes is discussed by Hocking, (1976). This procedure is illustrated in Figure 3.9(b).If the plunge of the line of intersection of two planes falls between the dip of the wall and the internalfriction angle circle, and if the dip direction of either of the planes falls between the dip direction of the walland the trend of the line of intersection, sliding will occur on that plane. Computational TechniquesA vector analysis technique for determining the stability of underground wedges, published in Hoekand Brown, (1980) has been adapted into an underground wedge stability computer program, UNWEDGE,produced by the University of Toronto. This program enables the user to graphically input the excavationgeometry, joint patterns, and joint strength properties for use in the analysis. The program can analyzewedges created by three intersecting joints in the roof or side walls of an excavation. The size and shape ofthe wedges can be displayed around the perimeter of the excavation. For a defined wedge, the programdetermines if the wedge is stable, falls out under gravity, or slides along one or two of the planes andcomputes factors of safety. An example of an UNWEDGE analysis is given in Figure 3.10. Such33programs are useful for determining whether specific wedges, which have been identified in a stope, will bestable or unstable. They can also assist the ground control engineer in the design of artificial groundsupport systems.Where the potential for wedge type failures is recognized, the span should be limited to control themaximum size of the wedge to one that can be supported with artificial ground support.3.1.4 Chimney FailureChimney failure occurs when the entire sill pillar or crown pillar above a stope slides as a blockinto the stope (Figure 3.11). This type of failure is not common but has been responsible for some largescale failures in the past. Chimney failures occur in very schistose rock masses or in orebodies where thefootwall and hangingwall are defined by weak discontinuities.Hoek, (1989) has developed an equation for determining the factor of safety against shear on thesides of the failure block. The factor of safety against downward sliding is given by:F=(i-+--- (3.27)Yr}LY x}where,Yr = unit weight of rockx = length of end wally = length of sidewallz = depth of block= shear strength along end walltyz = shear strength along side walland,= c+(a—u)tanp (3.28)where,c = cohesion along shear planea = horizontal stressu = groundwater pressureUsing this approach, chimney failure is defined to occur when the weight of the pillar exceeds thetotal shear strength developed on the sides of the block. The shear strength used in this equation is theshear strength along the sliding plane, which is a function of the cohesion, friction angle, horizontal stress,and water pressure along the shear plane.343.1.5 Rock Mass FailureGeneral rock mass failure or caving is characterized by a gradual failure of loose rock into thestope. Given sufficient time, the failure could continue to cave until the void is filled or it could stop whena stable shape has been created by the caving. Figure 3.12 illustrates these two conditions. Clearly, thesusceptibility of a stope to a rock mass failure is dependent upon many factors, the most important ofwhich are:• joint spacing;• joint orientation;• joint condition;• groundwater conditions;• stress conditions;• excavation geometry; and• rock hardness.Empirical design techniques are the only methods available for analyzing the susceptibility of arock mass to caving. All empirical methods rely on rock mass classification systems which attempt toquantify the rock mass parameters which contribute to weakness, Classification systems have been used toa limited extent in the past to predict stable spans, stand-up times, and support requirements inunderground openings. Unfortunately. however, much of the data has been compiled from civil engineeringcase histories which require higher safety factors than mining operations. Other empirical design methodshave been developed largely from open stoping (Potvin, 1988) or block caving databases (Laubscher,1981). The two most common systems which have been used to develop span design graphs and whichhave gained broad acceptance in the mining industry are the Norwegian Geotechnical Institute (NGI)- QRating System and the South African Council for Scientific and Industrial Research (CSIR) GeomechanicsRock Mass Rating system. NGI-Q Rock Mass Classification SystemThe NGI Tunneling Quality Index (Q) proposed by Barton, Lunde, and Lien (1974) is based on200 tunneling case histories in Scandinavia. The tunneling index value Q is defined by:(RQD”I (r’1Q—i ixi—ixi————n ) LaJ SRFwhere,RQD Deere’s Rock Quality DesignationJ = Joint Set Number= Joint Roughness NumberJoint Alteration Number= Joint Water Reduction Factor35SRF = Stress Reduction FactorThe first quotient, RQD/J is a rough measure of the relative block size. The second quotient 3i1arepresents the interbiock shear strength. Rough, tight, unaltered joints will have higher shear strengths thansmooth, open, and altered joints. The third quotient, is a measure of the active stress in the rockmass. SRF can be a measure of the loosening load in the case of shear zones, a measure of the inducedstress around the opening in the case of competent rock, or a measure of the squeezing or swelling load inplastic, incompetent rock. The factor w is a measure of the water pressure which reduces the effectiveshear strength of the joints. The ratings applied to individual parameters for the NGI Q system areprovided in Table 3.1 of this report.Barton et a!. have also defined two other factors for relating the Tunneling Index (Q) to the spanwhich can be supported.D— Excavation Span or Height (m)e—ESR (3.30)where,De = Equivalent DimensionESR= Excavation Support RatioThe excavation support ratio (ESR) is analogous to a safety factor and is dependent on the purposeof the excavation. The ESR ranges from 0.8 for underground nuclear power stations to 3-5 for temporarymine openings. A complete explanation of excavation support ratios is provided in Table 3.2. From thetable, it can be seen that the ESR for temporary mine openings is based on only 2 observations. Therefore,great care must be exercised if this procedure is to be used for design of mine openings. Figure 3.13 showsthe relationship between the Tunneling Index, Q and the equivalent dimension, De. This figure shows asharp line dividing the zone requiring support from the zone requiring no support. The equation of this lineis given by:Maximum Unsupported Span (m) = 2(ESR)Q°4 (3.31)In practice, there is a zone of potential instability which is not easily defined. Figure 3.12 has beenmodified in Figure 3.14 to show span versus Q for the excavation support ratios 3 and 5 defined to be theupper and lower bounds for temporary mine openings.363.1.5.2 CSIR Rock Mass RatingThe geomechanics classification system developed by Bieniawski (1976) is a general purpose rockmass classification system which has been used to predict stable spans, stand-up time, and supportrequirements. The rock mass rating (RMR) is defined as the sum of six parameters which can be obtainedin the field or estimated from borehole data.RIVER=A+B+C+D+E+F (3.32)where,A = unconfmed compressive strength of intact rockB = Deere’s Rock Quality DesignationC = spacing of discontinuitiesD= condition of discontinuitiesE = groundwater conditionsF = orientation of discontinuitiesThe range of values for each parameter is given in Table 3.3. Since the original publication of theclassification system, Bieniawski has made several updates to the ratings; however, the 1976 paper byBieniawski is considered the basic reference for this work and is the basis for other empirical studies (Hoeket al., 1994). The ratings for each parameter are summed up to obtain a value between 0 and 100. Basedon this rating, the rock is categorized into five classes ranging from very poor rock to very good rock(Table 3.3).Bieniawski has related the span to stand-up time and rock mass rating for tunneling and miningcase histories in Figure 3.15. This graph illustrates the wide band defining the unstable and potentiallyunstable zone.Bieniawski (1976) has proposed the following relationship between the NGI Q rating and the RMRbased on 117 case histories analyzed:RMR=9lnQ+44 (3.33)Based on this relationship, Bieniawski has compared the maximum unsupported span as predictedby the NGI and CSIR rock mass classification systems. In Figure 3.16 it can be observed that the RMR ismore conservative than the NGI system, which is probably a reflection of the different tunneling practice inScandinavia and the considerable experience they have in the particular rock conditions found there.The geomechanics classification provides guidelines for selecting the support for an opening.These support requirements are also dependent upon the size and shape of the excavation, the constructionmethod, and the stress around the opening. Support classifications for the geomechanics classification areprovided in Table 3.4.37Unal (1983) has proposed an equation for determining the support load using the rock mass rating.(100 — RMR “100 )YS=Yht (3.34)where,P = support loadht = rock load height (m)S = tunnel width (m)y = unit weight of rock (N/rn3)The variation of rock load per unit length of tunnel with span and rock mass rating is presented inFigure 3.17. This is analogous to determining the height of the relaxed zone of rock in the back of a stopewhich must be supported. The rock load height is plotted against tunnel width in Figure Modified 0 -RatingPotvin (1988) has developed an empirical design method for predicting stable spans in open stopesbased on a modified NGI- Q rating. The work is based upon an earlier study by Mathews et al. (1980)which looked at 26 ope stope case histories taken from three mines and 29 case histories taken from theliterature. The current Modified Stability Graph design technique is supported by 175 case historiescollected in more than forty Canadian underground mines. The chart known as the Modified StabilityGraph is constructed by plotting the rnodified stability number, N’ versus the hydraulic radius of the designsurface (Figure 3.19). The modified stability number N’ is given by the following equation:N’ =Q’ xA XB XC (335)where,Q’ = modified NGI tunneling indexA = rock stress factorB= rock defect orientation factorC = orientation of design surface factorThe modified NGI rating is taken to be the first two quotients of the NGI rating given in Equation3.29; that is, the stress reduction factor (SRF) and water pressure () terms have been ignored. The stresscondition is accounted for in Factor A. The values for factors A, B, and C can be obtained from the graphsin Figure 3.19. Each surface plotted on the graph was classified as stable, unstable, and caved; and fromthis data, three zones were defined: stable zone, supportable zone, and caving zone.38By calculating the parameters Q’, A, B, and C for a given stope surface, the stability number canbe plotted on the Modified Stability Graph to determine the potential for instability. It is important torecognize that the database was developed from open stoping case histories and that the design methodwould be unconservative for entry-type mining methods such as cut and fill. Mining Rock Mass RatingAnother rock mass classification system which has been developed is the Mining Rock MassRating (Laubscher, 1990). The MRMR takes account of the changes that a rock undergoes in a miningenvironment by taking the original CSIR rock mass rating and then adjusting it for weathering, mininginduced stress, joint orientation, and blasting effects. Laubscher has proposed a slightly different way ofcalculating the RMR, as follows:RMR=IRS++40{DxExFxG]m (3.36)where,IRS = unconfined compressive strength ratingFF/m = fracture frequency per metreD = large scale joint expression (i.e. wavy, planar, stepped)E = small scale joint expression (i.e. rough, smooth)F = joint wall alterationG joint infillingThe Mining Rock Mass Rating (MRMR) is defined as:MRMR=RMRx[WxJxBxT] (3.37)where,W = rating adjustment for weatheringJ = rating adjustment for joint wall orientationB = rating adjustment for blasting practiceT = rating adjustment for induced mining stressThe ratings for each RMR parameter and the adjustments are given in Table 3.5. The total stressparameter is the most difficult parameter to establish and may require an adjustment of 60% to 120%.Laubscher recommends that the mining induced stress be determined from published stress distributiondiagrams in the case of simple excavations, or from numerical modeling studies in the case of complexgeometries.Laubscher has also attempted to define the strength of the rock mass (RMS) in terms of the IRSand the RMR. Noting that large scale rock specimens give IRS values which are 80% of the small scaleIRS values, the rock mass is assumed to have a strength of 0.8 IRS if it had no joints at all. The RMS iscalculated by subtracting the IRS rating, A, from the full RMR rating, to give a rating out of 80 such that:39RMS = JRs1R1441 — A(’80 = IRS(RIvIR — A) (3.38)k 80 )‘l00) 100The Design Rock Mass Strength (DRMS) is defined as the strength of an unconfined rock mass ina specific mining environment. It could, for example, be compared to the mining induced stress in a pillarto compute a safety factor. In general, the immediate surface of an opening can be considered asunconfined. The depth of this zone depends on the size and shape of the opening. The same adjustmentsused to obtain the MRMR are used to obtain the DRMS.DRMS=RMSx[WxJxBxT] (339)This MRMR classification system has been successfully applied to assessing the suitability of arock mass for block caving. The objective in block caving is to open up a span which will remain unstableand continue to cave while the caved rock is gradually drawn out of the stope. The lower limit of what isconsidered a caveable rock mass could be considered to be the upper limit of the stable span of a stope.Laubscher (1990), has constructed a span design chart plotting MRMR versus the hydraulic radius (Figure3.20). Case histories were categorized as being stable (requiring only key block support), caving, or in atransition zone between the two where more intensive support was required to maintain stability. Thiscurve is not directly comparable to the RMR and Q span design curves presented earlier, since thehydraulic radius and not the span is presented. Golder Crown Pillar Study DatabaseCarter et a!. (1990) have undertaken an empirical evaluation of crown pillar stability involving 237individual pillar case histories. In most cases, sufficient data was present in the records for the authors toassign a CSIR Rock Mass Rating and an NGI Q value. A method was developed for relating the geometricfactors and rock mass parameters controlling pillar stability to the observed stability of the pillar, On thebasis of their study, the authors considered the most important geometric factors controlling crown pillarstability to be:• span(S);• thickness (T);• strike length (L);• foliation/ore dip (e); and• unit weight of rock (y).40These factors were appropriately combined to obtain a Crown Geometry Number (Cg). defined as:Cg=ft (3.40)where,F5t = span to thickness ratio = SITFe = stope inclination factor = (1 - 0.4cose)Fsr = span ratio factor = SI( 1 +S/L)F = specific gravity (y)The Crown Geometry Number is inversely proportional to stability. As the weight factor (F)increases, the weight of the pillar increases and stability decreases. The span to thickness ratio (Fi) is ahistorical rule-of-thumb approach for assessing pillar stability. As this factor increases, the pillar stabilitydecreases. The span ratio factor (Fsr), which is equal to twice the hydraulic radius, recognizes that whenthe length of the stope is greater than four times the span, failure is controlled by the short span. Forshorter strike lengths, stability is controlled by two-way spanning. The stope inclination factor isequivalent to Factor C in the Modified Stability Method described above. It recognizes that a vertical orsteeply dipping stope is more unstable than a shallow dipping one.The square root of Cg was taken to obtain a final empirical expression, C, tenned the ScaledCrown Span.C =S I (3.41)T(1+Sr)(10.4COS6)}The scaled span has been plotted against the rock mass rating for each case history in Figure 3.22.An empirical fit line proposed by Barton (1974) provides a good dividing line between the stable andunstable cases when superimposed on the crown pillar data. Carter et at. have added a hyperbolic sineterm to account for the non-linear trend to increasing stability in very good rock masses. The followingexpression was developed to describe this line:CriticalC = 3.3Q°43 [sinh 00016 (Q)] (3.42)Therefore, knowing the RMR or Q value for an area, the Critical Span can be obtained fromFigure 3.21. Knowing the pillar thickness, stope length, stope dip, and unit weight of the rock, the41minimum pillar thickness can be calculated by solving for T in Equation 3.41. The maximum allowablespan can be calculated by rearranging Equation 3.41 and solving the binomial equation for S.S2 —A2+ A2SLwhere,(3.43)A=C J(1—O.4cos9)As part of this research project, the raw data used Carter’s crown pillar study was reanalysedsolely on the basis of span and is shown in Table 3.6 and Figure 3.22. The individual points on the graphare defined as stable or unstable. A value is found adjacent to an observation on the graph if the crownthickness above the stope is less than 4 metres. These points have a very low thickness to span ratio wherefailure could result from beam failure rather than a rock mass failure.3.1.6 Stress Induced FailureStress induced failure is the result of mining induced stresses which exceed the strength of the rockmass. In competent, massive, elastic rock, this type of failure can take the form of rockbursting. In ajointed rock mass, a gradual yielding failure may take place. In cut and fill stopes, failure caused by highstress is most likely to occur in the back of the stope as the sill pillar width becomes smaller with each lift(Figure 3.23). It may also occur in very stiff post pillars. An analysis of the potential for this type offailure must take the form of analyzing the induced mining stresses at the boundaries of an excavation andcomparing it to the rock mass strength.Except in the early stages of development, most mines have complex excavation geometries inwhich stresses at the excavation boundaries cannot be estimated using closed-form solutions such as thosewhich have been compiled by Poulos and Davis (1974). In recent years a wide variety of computerprograms have become available which are capable of modeling excavations in two or three dimensions andcarrying out a stress analysis. The three main types of numerical analyses in use today are:• The Boundary Element Method;• The Finite Element Method; and• The Distinct Element Method.42These types of analyses are also useful for detennining the stresses around excavations and in turnserve as input for other design methods as described previously. Numerical modeling has many otherpractical applications for rock mechanics engineers including:• pillar design;• open stope span design/dilution studies;• shaft and service tunnel layouts;• analyses of complex excavation geometries;• stope sequencing studies; and• parametric design studies. Linütations of Numerical ModelingWith any numerical modeling procedure for determining stresses, the accuracy of the analysisdepends on the accuracy of three main input parameters:• The stress-strain relation(s) of the material(s);• The pre-mining stress conditions; and• The model geometry.Many numerical models (BEAP3D, EXAMiNE) assume the materials to be linear elastic, andisotropic. This assumption is considered accurate for intact drill core material but would not normallyrepresent the stress-strain relationship of the rock mass. Inhomogeneity of the rock mass can usually bemodeled by using different material parameters for different groups of elements in the model.The pre-mining stress conditions must be determined as input for the analysis. Both the magnitudeand direction of the principal stresses are required. These values are normally determined from in-situstress measurements. The Boundary Element MethodAn overview of two-dimensional boundary element stress analysis is provided by Hoek and Brown(1980). In general terms, the problem is to determine the stresses around an excavation given a twodimensional stress field as shown in Figure 3.24(a). This procedure describes the method of solving a two-dimensional problem; however, the 3-D problem is solved in a similar fashion. Prior to excavation, therock provides support for the area outside the excavation. This can be represented by normal andtangential tractions, as shown in Figure 3.24(b). The magnitudes of the tractions will vary along thesurface, depending upon the shape. When the opening is excavated, the stress at the boundary is reduced tozero. This is equivalent to introducing negative tractions at the boundary, as shown in Figure 3.24(c). Thefinal stress at the boundary can be considered to be the superposition of the original stress state and thestresses induced by the negative surface tractions.43The section shown in Figure 3.24(c) can be considered the true or actual situation. In order tomodel the problem, the boundary must be discretized into segments or elements. Each element is subjectedto a fictitious force acting in the plane of the section and with components Fn and Ft as shown in Figure3.24(d). These forces are assumed to act uniformly over the length of the element. An iterative procedureis used to adjust each of the fictitious forces in such a way that the normal and shear stresses at the centreof each element are equal to the normal and shear tractions. The stress at any point away from theboundary can be computed from standard expressions which sum the effects of the fictitious forces. Thesestresses are then added to the stresses from the original stress field to obtain the final stress. The elasticdisplacements are computed from standard solutions for displacements in an infinite medium due to pointor distributed loads.(a) 2-D Boundary Element ModelingTwo-dimensional boundary element programs can be used to model a cross-section of an openingwhere the dimension of the opening normal to the section is very long relative to the section dimensions(plane strain conditions). Most commercially available programs assume the medium to be linearly elasticand isotropic. The program FXAIVIJNE-2D (Curran et al., 1989) was used in the course of this study formodeling of cut and fill stopes and sill pillars. The program utilizes a graphical interface for input of theexcavation geometry and viewing of stresses and displacements. The program can use Mohr-Coulomb orHock-Brown failure criteria to compute factors of safety. Figure 3.25 is an example of the output createdby EXAIvIJNE-2D showing contours for the major principal stress around a vertical cross section of theDetour Lake Mine.(b) 3-D Boundary Element ModelingThe shape of many underground excavations and the influence of neighbouring excavations make2-D plane strain analysis inappropriate in many circumstances. In such cases a three-dimensional analysismay be required. The program BEAP-3D (CANMET, 1993) was used in the course of this researchproject to analyze the complex excavation geometry at Detour Lake Mine.BEAP-3D, or Boundary Element Applications Package, is a powerful numerical modeling packagedesigned specifically for modeling three-dimensional underground openings. The version used was capableof modeling up to 1000 elements. The program is designed to run from a Sun Sparc Workstation operatingunder Open Windows 3.0 or from a PC operating under Windows 3.1. The program utilizes a graphicalpreprocessor called MINE DESIGNER for creating a model geometry file as well as a graphical postprocessor called VIEWBEAP for viewing stresses.BEAP-3D has excellent graphics for viewing excavations and stress distributions; however input ofthe excavation geometry is still very time consuming, particularly when a high degree of detail is required.For this reason, three-dimensional modeling is still not widely used at mine sites. Further improvements to44these types of programs including linkages to existing mine design software will no doubt increase the useof this software in the near future. Figure 3.26 is an example of the output from BEAP-3D showing theprincipal stress contours around the stopes and the pit of the Detour Lake Mine. The results of themodeling of DLM will be discussed further in Chapter 5 of this report.(c) 2D versus 3D ModelingPakalnis (1991) has made a study of the relative differences in results between 2D and 3Dboundary element models. In general it was found that in the hangingwall of tabular stopes, the zone ofrelaxation predicted by 2D analysis is much larger than 3D for various stope geometries. The magnitudesof the tensile stresses in 2D are greater than in 3D. In the backs of stopes, the compressive stressespredicted by 2D modeling are greater than those evaluated by 3D modeling for various geometries. Since2-D results have been shown to be conservative under many conditions, it can remain a useful tool as a firstcheck. If stresses are found to be acceptably low using the 2-D model, 3-D modeling will not be required. The Finite Element MethodAn introduction to the finite element method applied to the field of rock mechanics can be found inBrady et al. (1985). Briefly, the finite element method involves discretizing the domain to be studied intoelements. The domain is subjected to initial stresses Pxx’ p, and Pzz shown in Figure 3.27(a).Appropriate boundary conditions are applied at the boundary of the domain to render the problem staticallydeterminate. For each element, appropriate functions are chosen which define the displacement of anypoint within the element in terms of the nodal displacements. The solution procedure results in a stiffnessmatrix, [k], for each element, based on the shape of the element, and an element load vector based on nodaltractions and boundary pressures such that,[k] [o]=[f] (3.44)The element stiffness matrices are assembled into a global stiffness matrix in such a way thatcompatibility of displacements is maintained. For the global system,[A] = [Kf’ [F] (3.45)The system is solved for the global nodal displacements; and, because the strain is the derivative ofthe displacements, the nodal strains can be computed. Stresses are then computed using a two dimensionalelasticity matrix involving Young’s Modulus and Poisson’s Ratio.45The finite element method does offer more flexibility in terms of defining the stress-strain relationsthan does the boundary element method. In theory, each element could have a different stress strainrelation but in practice, groups of elements are the same. Although most codes which have been developedfor rock mechanics applications rely on a simple, linear-elastic, stress-strain relationship, more complicatedcodes can be developed to model post peak strength behaviour. The disadvantage of the finite element isthat it is very time consuming to set up the model for all but the simplest problems because the entiredomain must be discretized rather than just the boundary as is the case in boundary element models. Thenumber of elements required also makes it very costly in terms of computer time and data storagerequirements. The Distinct Element MethodThe distinct element method treats the domain being modeled as an assemblage of blocks (Figure3.28). This may certainly be appropriate for many excavations where stability is controlled by structuraldiscontinuities. Where the stifffiess along these discontinuities is much less than the stiffness of the block,the block can be considered rigid and displacement only occurs along the discontinuities. Therefore, theblock shape does not change during the analysis; rather, the system adjusts to the prescribed boundaryconditions by movement parallel and normal to the joint surface. Difficulties with the distinct elementmethod arise in defining the stress-strain relationships parallel and perpendicular to the discontinuity. Theproblem is compounded if each discontinuity has a different strength. Furthermore, setting up the modelgeometry for actual mine problems is cumbersome and time consuming. These factors may account forwhy the procedure is not widely used in the mining industry except as a research tool.3.2 SURVEY OF OTHER CUT AND FILL OPERATORSA survey questionnaire was distributed in 1991 to cut and fill operations in Canada. The survey’sobjectives were to:• identify how operations were designing cut and fill stope spans;• obtain an operation’s typical stable stope span and rock mass quality;• obtain span and rock quality data from stopes which had experienced afall of ground;• identify how operations were designing post pillars and sill pillars;• determine types of instrumentation mines were using to predict instability;and• determine how cable bolts are being utilized in cut and fill stopes.46The following five companies representing 9 operations responded to the questionnaire:• Westmin Resources - Myra Falls Operations, Campbell River, B.C.;• Placer Dome Inc., - Dome Mine, Timmins, Ontario;• Hudson Bay Mining & Smelting -Trout Lake Mine, Trout LakeManitoba;• Inco Ltd. - Sudbury Operations, Sudbury, Ontario; and,• Falconbridge Inc. - Sudbury Operations, Sudbury, Ontario.The returned questionnaires are provided in Appendix A. Table 3.7 summarizes the span designdata obtained from this questionnaire. It is noteworthy that empirical design is used at least in part in threeof the four operations which require span design. Figure 3.29 is a plot of the span versus RMR for thesefive operations. The points on the graph plot as critical or design. A critical point indicates that basedupon past practice it was demonstrated that exceeding the span for the given rock quality results ininstability. A design point indicates the design mining span for an operation. This would have beendetermined through numerical modeling, past experience, or orebody constraints. It does not necessarilyimply that larger spans would be unstable.A number of pillar design methods are used by these mines and are identified in Table 3.8 andTable 3.9. Empirical design based on past experience at individual operations is the most common method,however two of the mines reported using the Hedley pillar formula for design (Hedley, 1972).Surprisingly, only one of the operations reported using numerical modeling in the pillar design process.Cable bolts were found to be used by all of the operations in specific situations where bad structureor a low quality rock mass has been identified. No firm rules were identified for where cable bolts wouldbe installed. Table 3.10 gives some detail of the bolting practice at these operations. One mine reported anexperimental support project involving replacing post pillar support by cable bolt support.3.3 DESIGN PHILOSOPHYIn the author’s experience, the best approach to span design is an integrated one which combineselements of analytical solutions, empirical design, and numerical modeling. This approach has beensuccessfully applied at Detour Lake Mine for designing safe yet practical spans in wide cut and fill stopes.The different analysis techniques are considered necessary given the different types of failure mechanismsdescribed earlier. Solving a solution from two or more approaches also provides confidence in the design ifthe two solutions can be made to closely agree. Of course, any final design must also comply with existingregulatory statutes such as those covering the minimum size of barrier pillars.47Before any design can be implemented, a detailed fabric analysis of the rock mass must be can-ledout and the intact rock strength parameters of the rock must be determined. This is required in order toobtain the fundamental parameters which are required regardless of the design procedure or failuremechanism. The intact rock strength parameters which should be determined prior to carrying out a designare:• Uniaxial Compressive Strength (UCS);• Young’s Modulus (E);• Poisson’s Ratio (t); and• Unit Weight (N/rn3).The fabric analysis must provide sufficient information on the structural characteristics of the rockmass for it to be classified using either the NGI or CSIR classification systems. At Detour Lake Mine, thisinformation is obtained from diamond drilling and geotechnical mapping of each lift. Structural data iscompiled on stereonets to determine the dominant joint sets. Faults or continuous joints are mapped on adaily basis by the geology staffAs the second step of the design process, the rock mechanics engineer must decide which is thecontrolling failure mechanism. At Detour Lake Mine for example, the presence of faults or continuousjoints in the back which dip at less than 300, can lead to wedge failure irrespective of span or the overallrock quality. Structural failure, therefore is the first failure mechanism which must be accounted for in thedesign. Prediction of structural failure requires a good database of geotechnical observations for the areaunder design as well as ongoing visual inspection. A stereonet analysis of the structures defining the wedgeis carried out to determine whether failure is kinematically possible. The computer program UNWEDGEis used to determine the size of the potential groundfall and to assess the support requirements. Ifstructural failure is predicted, the span could be reduced to prevent formation of the wedge. Alternatively,more intensive ground support could be specified to support the wedge. At this stage, numerical modelingwould be useful for detennining the extent of the relaxed zone in the immediate back.The next step is to decide which other failure mechanisms and design methods are relevant. AtDetour Lake Mine and most underground metal mines, beam and plate failure can be ruled out as a failuremechanism given that discontinuities are generally present and these theories assume horizontally stratifiedintact rock. Voussoir block theory does not apply at DLM since the jointing pattern does not meet the strictcriteria set out above. Chimney failure is unlikely at DLM because there is not any unfavourable structureon the hangingwall or footwall or strong foliation parallel to the orebody. At the ultimate stope heightwhere the pillar width is low, a quick assessment for chimney failure should be made using Equation 3.27.Numerical modeling may also be required at this stage in order to determine the horizontal stresses on thepillar such that the shear stress in Equation 3.27 may be calculated.48After structural failure, rock mass failure is the next most likely mode of failure. A rock massfailure assessment is made using an empirical approach utilizing a database of observations from stopingcase histories at Detour Lake Mine. These observations have been compiled and plotted on a span versusRMR graph to enable future prediction of stable spans given the RMR of the stope. This approach hasproven successful at predicting stable spans at DLM. The DLM database and empirical design methodwill be described in greater detail in Chapter 6.Finally, the potential for stress induced failure is assessed using 2-D and 3-D boundary elementmodeling. Stope design at DLM is an ongoing process. New structures and changes in the overall rockmass due to stress redistribution can develop from lift to lift, which may warrant design changes. TheDLM design procedure for cut and fill stopes is outlined in the diagram shown in Figure 3.30. Thedevelopment of an empirically based stability graph specifically designed for entry-type stopes is the focusof this study.49Table 3.1 NGI- Q Classification System Rating for Individual Parametersat., 1974)_________________________________________(Adapted from Barton etParameter Item and Description ValueRQD Rock Quality DesignationThe total length ofcore pieces over four inches in length in an interval 0-100divided by the length ofthe interval and expressed as a percent.Number of Sets of DiscontinuitiesMassive 0.5One Set 2.0One Set Plus Random 3.0Two Sets 4.0J1 Two Sets Plus Random 6.0Three Sets 9.0Three Sets Plus Random 12.0Four or More Sets 15.0Crushed Rock 20.0Roughness of DiscontinuitiesNon-continuous joints 4.0Rough and wavy 3.0Smooth and wavy 2.0r Rough and planar 1.5Smooth and planar 1.0Slickensided and planar 0.5Filled_discontinuities 1.0Filling and Wall Rock Alteration, EssentiallyUnfilledHealed Joints 0.75Staining only, no alteration 1.0Slightly altered joint walls 2.0Silty or sandy coatings 3.0Clay coatings 4.0J Filling and Wall Rock Alteration, Filled JointSand or crushed rock filling 4.0Stiff clay filling less than 5 tnmthick 6.0Soft clay filling less than 5 mm thick 8.0Swelling clay filling less than 5 mm thick 12.0Stiff clay filling more than 5 mm thick 10.0Soft clay filling more than 5 mm thick 15.0Swelling clay filling more than 5_mm thick 20.0Water ConditionsDry, or inflow < 5 litres/minute locally 1.0Medium water inflow 0.66Jw Large inflow, unfilled joints 0.5Large inflow, filled joints with washout 0.33Large inflow, filled joints, high transient flow 0.210 0.1Large inflow, filled joints, high continuous inflow 0.1 to 0.05Stress Condition ClassLoose rock with clay filled discontinuities 10.0SRF Loose rock with open discontinuities 5.0Shallow depth (50 m or less) rock with clay filled discontinuities 2.5Rock with tight unfilled discontinuities, medium stress 1.050Table 3.2 Excavation Support RatiosExcavation Category ESR No. of CasesA Temporary Mine Openings 3-5 2B. Vertical Shafts:Circular Section 2.5 -Rectangular or Square Section 2.0 -C. Permanent Mine Openings, water tunnels for hydropower 1.6 83(excluding high pressure penstocks), pilot tunnels, driftsand headings for large excavations.D. Storage rooms, water treatment plants, minor highway or 1.3 25railway tunnels, surge chambers, access tunnels.E. Power stations, major highway or railway tunnels, civil 1.0 73defense chambers, portals, intersections.F. Underground nuclear power stations, railroad stations and 0.8 2factories.51Table 3.3 Geomechanics Classification of Rock Masses (after Bieniawski, 1976)A. Classification Parameters and their RatingsPARAMETER RANGE OF VALUESStrength of Point Load For this low rangeIntact Rock Strength Index >8 MPa 4-8 MPa 2-4 MPa 1-2 MPa uniaxial tests are1 Material preferredJniaxial >200 MPa 100-200 MPa 50-100 MPa 25-50 MPa 5-25 1-5 <1 Mi’sompressive MPa MPatrengthating 15 12 7 4 2 1 02 Drill Core Quality, RQD 90%-100% 75%-90% 50%-75% 25%-50% <25%Rating 20 17 13 8 3SpacingofDiscontinuities >3m 1 .0-3.0m 0.3-1.Om 50-300mm <50mm3 Rating 30 25 20 10 5Condition of Discontinuities Very rough surfaces Slightly rough Slightly rough Slickensided Soft gouge > 5mmNot continuous surfaces] surfaces surfaces OR thick4 No separation Separation < Separation < 1mm Gouge <5 mm ORUnweathered wall lmm,slightly Highly weathered thick OR Separation lessrock weathered walls walls Separation 1-5 than 5mmmm continuous ContinuousRating 25 20 12 6 0Inflowperlont None <25 25-125 >125tunnel length litres/mm litres/mm litres/mm5 Groundwater Ratio 0 0.0-0.2 0.2-0.5 >0.5Joint Waer PreosureMajor Princip1 StressGeneral Completely Dry Moist only Water under Severe waterConditions (Interstitial Water) moderate problemspressureRating 10 7 4 0B. Rating Adjustment for Discontinuity OrientationVtrike and Dip Orientations of Joints Very Favourable Favourable Fair Unfavourable Very UnfavourableTunnels 0 -2 -5 -10 -12Ratings Foundations 0 -2 -7 -15 -25Slopes 0 -5 -25 -50 -60C. Rock Mass Classes Determined From Total RatingsRating 81-100 61-80 41-60 21-40 <20ClassNo. I II Ill IV VDescription Very Good Rock Good Rock Fair Rock Poor Rock Very Poor RockD. Meaning of Rock Mass ClassesClass Number I II III IV VAverage Stand-up Time 10 years for 15 m 6 months for 8 1 week for 5 m 10 hours for 2.5 m 30 minutes for 1 mspan m span span span spanCohesion of the Rock Mass >400 kPa 300-400 kPa 200-300 kPa 100-200 kPa <100 kPaFriction Angle ofthe Rock Mass <45 35°45° 25°-35° 15°-25° <15°52Table 3.4 Geomechanics Classification Guide for Excavation and Support in Rock Tunnels (afterBieniawski, 1984)Rock Mass Excavation SupportClassRockbolts (20mm Shotcrete Steel Setsdiam., fully bonded)Very Good Rock Full Face: Generally no support required except for occasional spotRMR: 81-100 3m advance boltingGood Rock Full Face: Locally bolts in crown 50 mm in NoneRMR: 61-80 1.0-1.5 madvance; 3 mlong, spaced 2.5 m crown whereComplete support 20 with occasional wire requiredm from face meshFair Rock Top Heading and Systematic bolts 4m 50-100mm in NoneRMR: 41-60 Bench: long,spaced 1.5-2.0 m crown and 301.5-3.0 m advance in in crown and walls with mmin sidestop heading; wire meshCommence support in crownafter each blastComplete support lOmfrom facePoor Rock Top Heading and Systematic bolts 4-5 m 100-150 mmin Light ribsRMR:21-40 Bench:1.0-1.5 m long, spaced 1-1.5 m in crown and 100 spaced 1.5 madvance in top heading; crown and walls with mm in sides where requiredInstall support wire meshconcurrently withexcavation - lOm fromfaceVery Poor Rock Multiple Drifts: Systematic bolts 5-6 m 150-200 mm in Medium toRMR: <20 0.5-1.5 m advance in long crown heavy ribstop heading; Spaced 1-1.5 m in 150 mm in spaced 0.75 mInstall support crown and walls with sides and with steelconcurrently with wire mesh Bolt invert 50mm on face lagging andexcavation; shotcrete forepoling ifas soon as possible required. Closeafter blasting invert53C x -4-4C>mZ><vrQCmC>0>-‘rzr-c,,rrmz--2pII&LSS00S——-----—-‘JA3=E-39-200000--0g,O C——-——-0’J0000‘0-0’0’-.040000‘0‘00000‘0—0-40’0’0’0’0’0’‘0’0’0’0’‘00’5 2c0’-00000-.00’0’-j-0000‘0-00000‘0—0’C’‘00.00’V0’0’-0C’0’0’0’0’0’0’0’—J--000’(0c00’0’-,--000000’-.C--a00‘00‘0‘fl0’0’0’0’‘00’‘000’0’0’0’U“0-0U0—0’-0-J0’V.0’0’-I--C000’0’I0000oO,,co’.oco©’0OO’0C’c0O0C’”0OOcO0’O0’C0JI 0 000’00z P,0-.00—,.z:::‘—00kCt-00N)_—EEEH—I’.C“00’000’k.0’000’Table 3.6 Crown Pillar Study Data (Carter, 1990)U = UnstableS = StableA number following an unstable condition (U2) refers to an unstable case where the thickness of the pillar was verysmall. The number refers to the pillar thickness in metres. For these cases, beam failure rather than rock massfailure may have been the failure mechanism.Case No. RMR Q Span Condition(%) (m)Case No. RMR Q Span Condition(%) (m)12A12B142122A22B25A25B262728293536383941465052535457596264A64B1234567891011128080453550502525505515752558505055202048548555708045457570075654560606085358054.654.’UUUUUUU3U3U4UUiU’SSSSSSSSSSSS131415161718192021222325303132333435363740424344454647484951555657586061626365666768697050454585708080704550603050605075502558607575756045506070454145506565604580603685655656701. 3.7 Span Design Methodology at Canadian Cut and Fill MinesTable 3.8 Post Pillar Desi n Methodology at Canadian Cut and Fill MinesDesign Method Falconbridge Ltd Placer Dome HBM&S Inco Limited WestminSudbwy Mines Inc. Trout Lake Mine Sudbuiy Mines ResourcesDome Mine HW MineMethods Based on Hedley .FormulaPast ExperienceTributary TheoryTable 3.9 Sill Pillar Desi n Methodology at Canadian Cut and Fill MinesDesign Method Falconbridge Placer Dome HBM&S Inco Limited WestminLtd. Inc. TroutLakeMine Sudburylvlines ResourcesSudbuty Mines Dome Mine 11W MineBased on Production . .RequirementsPast Experience .Designed to Support Loadof Fill Above PillarTable 3.10 Cable Bolting Support Design Methodology at Canadian Cut and Fill MinesDesign Method Falconbridge Placer Dome HBM&S Inco Limited WestminLtd. Inc. TroutLakeMine SudbuzyMines ResourcesSudbssy Mines Dome Mine HW MineNorandalPotvin Method .Numerical MethodsPast Experience at .MinesiteAnalytical MethodsExperience at Other MinesDesign Method Falconbridge Placer Dome HBM&S Inco Limited WestminLtd. Inc. Trout Lake Mine Sudbwy Mines ResourcesSudbury Mines Dome Mine 11W MineMathews Stability Graph .MethodNumerical MethodsPast Experience . .Not Designed (FullExtraction)56UNDERGROUND ROOF FAILURE MECHANISMS FIGURE 3.1—7-c--(a) Plate or Beam Failure (b) Voussoir Block Failure(c) Chimney Failure (d) Wedge Failure(e) Stress Induced Failure (f) Rock Mass Failure57I ItBeam Shear FailuretBeam Tensile FailureBEAM FAILURE MECHANISMS FIGURE 3.258MexluunTensileStress(l9a)0 I 3 3 ii I I Co •1 0;00 2I I CII Co •0 I 0I I I 0 I 0 3 000000Ma,dnrnTensileStress(lea)Ir13 C 3 0 CD 0 CD 0 0x(c) Line of Thrust in Beam to Abutments(after Brady and Brown, 1985)VOUSSOIR BLOCK THEORY FIGURE 3.4voussoir controflingcentral crack(a) Voussoir Block Excavation RoofI SI’t.1—1W/2TV(b) Parabolic Arch Developed In Roofline$/4 4’ tc—ijTTj60crushing zones(failure governed by UCS)(C) Buckling Failure with Increasing Eccentricity of Lateral ThrustVOUSSOIR ARCH FAILURE MODES FIGURE 3.5(a) Crushing at the Hinges(b) Shear Failure at the AbutmentsFailure controlled by outer fibretensile strength and horizontal stress61baN(after Beer & Meek, 1982)VOUSSOLR PLATE FAILURE MODES FIGURE 3.662\\ I‘ ,(b) Sliding Wedge Failure in WallWEDGE FAILURE FIGURE 3.7(a) Gravity Wedge Failure in RoofV63NACJoint SetCJoint Set AJoint Set B(a) Free Fall Wedge FailureNA ( + II\ A——Bc(b) Sliding Wedge FailureSTEREONET ANALYSIS OF WEDGE FAILURES iN BACK FIGURE 3.8—I64(a) Sliding on Line of Intersection of JointsDip Direction of WallTrend of Line of Intersectionof JointsDip Direction of Joint Set AJoint Set BDip Direction of Line of IntersectionDip Direction of Joint Set ADip Direction of Wall(b) Sliding on Joint ASTEREONET ANALYSIS OF WEDGE FAILURES IN WALLS FIGURE 3.9ATunnelDip Direction of Joint Set BTunnel Joint Set AWall65‘.4UNWEDGE ANALYSIS OF WEDGE FAILURE FIGURE 3.10rr2,C- Ii It)C’). wWi ‘I? W)-IIti W I oIR 4I WZt•-4I I 1’ I‘•41 ,— .. ——— —uJCIiiIICCwIU.CClUI Ui!I‘I.004’I-’I66NCHIMNEY FAILURE OFCROWN PILLARCHIMNEY FAILURE FIGURE 3.1167PROGRESSIVE ROCK MASS FAILURE FIGURE 3.12ROCK MASS FAILURE MAY STOP WHEN A STABLE SHAPEHAS BEEN ATTAINED1CAVING MAY CONTINUE UNTIL VOID IS FILLED68,o. d rri CD II100 10 I 0.1EXCEFROt’LAU.VEXWEkIELVTVERYEXT.EXC.FOORFOORVERYPOORPOORFAIRG000GOODGOODGOOD).2-—--JF[SUPPORTREQUIRED1HL:: ——I-I-f——————-——,,NOSUPPORTREQUIRED—,-—-++4-I-H-I-———1--1+—.‘.__h—.001.010.11101001000RQDxxNGITunnellingQualityQjfljaSRFE (U 0) a, t 0 0.D E E (UC H100 10 I TunnellingQualityQ=xxJaSRF1000zLI)LL00(after Bieniawski, 1984)FIGURE 3.152015106420.1 1 10 102 i05STAND-UP TIME (hr)D tunneling roof falls• =casehistonesofrooffallsinmrningcontour lines = limits of applicabilityRELATIONSHIP BETWEEN SPAN, RIvIR, AND STAND-UP TIME71U)UII-LUzCU)LUI00aDCl)zDCOMPARISON OF SPAN VERSUS STAND-UP TIME FOR NGI AND FIGURE 3.16RMR DATA$4b50403020151086543210.51 10 102 1O 1O 10STAND-UP TIME (HOURS)(after Bleniawakl, 1974)7220000SPAN (m)(after Unal, 1983)VARIATION OF ROCK LOAD AS A FUNCTION OF SPAN FIGURE 3.17“Iz-jLUzzILL.0xIC,zLU—IIzLU00-JC.)0150001000050000/IA///--___0 5 10 15 20 25 307325SPAN (m)(after Unal, 1983)FIGURE 3.1820151050/////;:2z—904____0 5 10 15 20 25 30VARIATION OF ROCK LOAD HEIGHT AS A FUNCTION OF SPAN74JO)NTORIENTATIONADJUSTMENTFACTORSROCKSTRESSFACTORACpbA0 Oz 88 ES cbz IMODIFIEDSTABILI1YNUMBER(N) 0000 C),0 0 0CD -I,0 CD 0 CD CD (0GRAYITYADJUSTMENTFACTORC(SUDING)00 > z ID U0 z 0 0 0 z0iGRAVITYADJUSTMENTFACTORC(GRAVITYFALL)z 0 O0 -Il m Cl) w I -1 -< G) -U I0 F’.)C),100U,C,z(0(I,C.,0wD-09080706050403020o STABLEA TRANSITIONAL• CAVING100 6010 20 30 30 40STABILITY INDEX = HYDRAULIC RADIUS = AREAPERIMETER(after loubscher. 1990)MINING ROCK MASS RATING VERSUS HYDRAULIC RADIUS FIGURE 3.2076NORWEGIANGEOTECHNICALINSTITUTETUNNELLINGQUALITYINDEX,00.0010.010.1EXTREM.XCER’EXCEPTIONALLY’EXTREMELYVERYPOORPOORFAIRGOODGOODIG00POORPOOR1.0410401004001000(I)6 0 C) z 0 -J010203040606070809D100VERYPOORPOORFAIRGOODVERYGOODSOUTHAFRICANCOUNCILFORSCIENTIFICANDINDUSTRIALRESEARCHGEOMECHANICSCLASSIFICATIONCrown Pillar Data (Carter, 1990)100101... ,...II.SUAUU.A’AUI 30 10 20 30 40 50 60 70 80 90 100CSIR Rock Mass Rating• UnstableA Unstable (possibly not rock mass failure)• StableNote:Numbers refer to crown pillarthickness. Failure mechanismmay have been beam bending forcases indicatedCROWN PILLAR STUDY DATA ANALYZED TN TERMS OF SPAN FIGURE 3.2278I.I.IT‘-‘h‘4IIl(a) High Horizontal Stress Developing in Sill Pillar0-v0-v(b) High Vertical Stress Develops in Post PillarsSTRESS INDUCED FAILURE IN PILLARS FIGURE 3.2379avH??avtJ_ /y(s—I-/1F,,3 \/(b) Tractions on potential boundarybefore excavation of hole in aninfinite plane(c) Negative tractions representingeffects of excavation(d) Fictitious forces and stresses onelements of imaginary surfaceon infinite planeTHE BOUNDARY ELEMENT METHOD - 2D FIGURE 3.24(after Hoek and Brown, 1980)(a) The problem to be solvedallah- 4-t?t?avNH-/çiTallalla ) 4-80EXAMINE-2D OUTPUT - DETOUR LAKE MINE81FIGURE 3.25xjy7 bc,BEAP-3D OUTPUT - DETOUR LAKE MINE FIGURE 3.26•1• ••82pyyttt t(a) Excavation to bemodelled with fieldstresses(b) Discretization of modeldomain with approriateboundary conditionsyu q(alter Brady el al., 1985)(C) Simple triangular elementwith six degrees offreedomTHE FIMTE ELEMENT METHOD - 2D FIGURE 3.274-4-4-4--.pyypyyft t t_pxxpxx--pxx--pxx4-4- -4---C)u,,u83a)Schematic representation of an opening in which thestability Is controlled by discrete rock blocks and thereforeamenable to modelling by the distinct element methodF4’,F1l1rb) Normal and shear modes ofinteraction between distinct elements(after Brady and Brown, 1985),i.w po.ld old 90.1dm,THE DISTINCT ELEMENT METHOD -2D FIGURE 3.28—\F,\ AF.84IAt’SPAN VERSUS ROCK MASS RATINGCANADIAN CUT & FILL MINES10 20 30 40 50 60 70 80 90 100CSIR Rock Mass Rating (%)• Westmin A Dome ?4( Trout Lake • Inco-Sudbury )< Falconbndge-SudburyFIGURE 3.29“U30?IE20U10 A <UUnSPAN VERSUS ROCK MASS RATiNG AT OTHER CANADIANOPERATIONS85DECISION TO MINE14FABRIC ANALYSISAND ROCK MASSCHARACTERIZATION(from core or from nearby workings)IS THERE A PROMINENT STRUCTURE CONTROLLING STABILITY?ANALYTICAL DESIGNSOLUTION(Wedge Analysis)I LIMIT SPAN I II TO PREVENT I I SUPPORTI FORMATION OF I IWEDGEWEDGEENGINEERINGJUDGEMENTEXCAVATIONANDMONITORINGFIGURE 3.30NONUMERICALMODELLING(stress)-FEM-BEM-DEMEMPIRICAL DESIGNSOLUTION(Stability Graph for Entiy Type Stopes)--,FJr ISPAN DESIGN METHODOLOGY AT DETOUR LAKE MThE864. DATABASE4.1 INTACT STRENGTH PROPERTIESAn estimate of the intact rock strength parameters is required for rock mass classification and forinput into stress modeling programs. Prior to this study, direct unconfined compressive strength test dataof Detour Lake Mine rock had been conducted by:• CANMET (March, 1985);• Terraprobe (July, 1984); and,• Smith Engineering (October, 1984).The results of these tests are summarized in Table 4.1. It must be recognized that the above testingwas conducted prior to the commencement of underground mining. Once underground access wasavailable, indirect UCS measurements were obtained from point load strength tests on representativesamples from 120 Level, 360 Level, and 560 Level. These results are presented in Figure 4.1 and 4.2.There was a large variation between all of these strengths so further UCS testing was carried out atUniversity of British Columbia.Representative AQ drill core (26 mm diameter) was obtained from the mine for testing. Allsamples failed through intact rock. The results are given in Figure 4.3 and 4.4. Unconfined compressivestrength test results from UBC for each of the three zones are compared to results of other researchers inFigure 4.5. Testing by UBC indicated a UCS for the mafic hangingwall rock of 165 MPa. This compareswell to a weighted average of 162 MPa for all other direct tests. The UCS of the MZ rock was very similarto that of the hangingwall at 166 MPa. This is to be expected since the lithology of the HW and the MZ issimilar except for the presence of veining in the MZ. The UBC result is similar to the value obtained byCANMET (169 MPa) but significantly higher than the Terraprobe value (91 MPa).UBC testing of the Talc Zone rock indicated a significantly lower average UCS of 28 MPa ascompared to the average of other researchers (93 MPa). Previous test work had been carried out on nearsurface rock which may account for the difference. The UBC results are supported by recent tests carriedout by Voest-Alpine (1991) in their recent assessment of the suitability of a roadheader for mining of theTalc Zones. Strengths were obtained between 22 and 38 MPa with an average of 29 MPa based on foursamples taken from the 460 Talc Zones. Another series of tests were carried out by UBC to confirm theseresults. Rock samples were obtained from attack drifts in the 360 T20 #3 and 460 T40 #6 stopes. Therock samples were subsequently cored to 52 mm diameter and tested. The results are given in Table 4.2.The unconfined compressive strength was higher at 48 MPa. The samples from these tests appeared to be87more siliceous than previous samples tested which may have contributed to the higher strength. Theseresults can be considered to represent an upper bound on the unconfined compressive strength of the TalcZone rock.Given that the UBC values were averaged from 10 tests and were taken from the most recentlymined areas, the values were used as design strengths for the purposes of this study. These and other rockproperties which have been measured at Detour Lake Mine are summarized in Table FABRIC ANALYSISAs part of this study, detailed fabric mapping was carried out in the following areas of the MainZone and Talc Zones:260MZ#l2Lift 360T20#4260T40#1 560MZ#1360 MZ #12 Lift 460 T40#6360 T40 #3, #4 460 T40 #5360 MZ#5 LiftDetailed line mapping was carried out and features were recorded on a standard geotechnicalmapping sheet (Figure 4.6). The following information was recorded: location, distance, rock type,structure type, number of features having similar orientation, spacing, rock and joint infihl hardness, strike,dip, aperture width, planarity, roughness, continuity, and water. Approximately 350 structures weremapped in the Main Zone and 419 in the Talc Zones.4.2.1 Joint OrientationThe orientations of the structures mapped during this study have been plotted on lower hemisphereequal area stereonets (Figures 4.7 to 4.8). In addition to the structures mapped as part of the detailed linemapping, a compilation was made of structural mapping recorded by the DLM geology department onpreviously mined lifts. Figures 4.9 to 4.12 show stereonets which record structural data for each liftstudied. These stereonets suggest a north-south joint set and an east-west joint set both dipping vertical tosubvertical. Random jointing is also present. Figure 4.13 shows stereonets for the Main Zone and for theTalc Zone with structures from all lifts plotted. Two joint sets are predominant:Joint Set A: Mean Orientation: Strike: 096° Dip: 90°Joint Set B: Mean Orientation: Strike: 353° Dip: 83°88The same two joint sets are predominant in the Talc Zones with only slightly different orientations:Joint Set A: Mean Orientation: Strike: 093° Dip: 90°Joint Set B: Mean Orientation: Strike: 352° Dip: 86°4.2.2 Joint RoughnessJoint Roughness has been measured by comparing the profile of each joint to 10 standard profilesshown in Figure 4.14. The joint roughness for the MZ and TZ have been plotted on a frequency histogram(Figure 4.15(a)). In the case of the mafics, the highest percentage of joints have a Joint RoughnessCoefficient (JRC) of 2-4 indicating at smooth, planar surface. Most other joints have a JRC of 4 to 10. Inthe Talc Zones, the highest percentage ofjoints have a JRC of 6-8 which is a rough stepped surface.4.2.3 Rock StrengthThe rock strength was established by field index testing described in Table 4.4. The rock hardnessis placed in 5 categories, Ri being the weakest to R5 being the hardest. The frequency histogram for rockhardness is given in Figure 4.15(b). The majority of Main Zone rocks tested have a hardness of R4indicating a UCS of 100 - 200 MPa. All other rocks in the Main Zone had an R5 hardness.Approximately 75% of the Talc Zone rocks had an R3 hardness (50 - 100 MPa) with the remainder beingR2 in hardness.4.2.4 Joint ApertureJoint aperture was classified into 5 categories ranging from very tight to moderately wide (Table4.5). Figure 4.16(a) is a frequency histogram indicating the joint aperture for talc and maflc rocks. Mostmaflcs and talc rocks have joints which are tight to very tight.4.2.5 Joint SpacingThe joint spacing is classified into the three categories used by the CSIR rock mass ratingclassification system (Figure 4.16(b)). Most main zone joints have a spacing between 5 cm and 1 metre. Asmaller percentage ofjoints have a spacing of 1-3 metres.4.2.6 Joint ContinuityJoint continuity has been classified into three categories, 0 - 3 m, 3 - 5 m, and 5 - 10 m. This datahas been plotted on a frequency histogram shown in Figure 4.17(a). Talc Zone joints occur in all89categories. Most joints in the mafics occur in the 3 - 5 m and 5 - 10 metre categories. In some cases, itmay not be possible to establish the full length of the joint if it continues beyond the wall of the excavation.Figure 4.17(b) indicates the number of joint ends which are visible. The Talc zone joints are generally lesscontinuous with 1 or 2 ends visible. Most mafics have 0 of 1 end visible. Zero ends visible indicates thatthe joint is longer than the dimensions of the excavation. The openings in the mafics are wider so thiswould indicate that joints are generally longer in the MZ than the TZ.4.3 ROCK MASS CHARACTERIZATIONAll work areas including those mined prior to the study were assessed a CSIR rock mass rating.The rating was performed using a standard rock classification form shown in Figure 4.18. The CSIR rockmass classification system was chosen for use at DLM for the following reasons.• Consistency of results among those performing the rating;• It relatively quick to use and understand; and,• The percentage scale is easier to get a “feel” for than the logarithmicNGI - Q rating.The RMR can be converted to the NGI- Q using Equation 3.33. This relationship was checkedperiodically and proven to be valid for the accuracy required.4.3.1 Main ZoneThe rock mass ratings for the MZ are compiled in Table 4.6. The average RMR in the Main Zoneis 73 with a standard deviation of 7.8. The lower ratings generally occur on the footwall side of ore bodywhere the joint spacing is closer and the orientation of the joints is more random. A typical CSIR rockmass rating, based on the fabric analysis presented previously, is provided in Table Talc ZoneCSIR rock mass ratings for all of the Talc Zone workings analyzed are provided in Table 4.7. Themean RMR for talc is 49 with a standard deviation of 11. The rock quality generally decreases withincreasing distance from the chert horizon. From the fabric analysis presented above, a typical CSIR -RMR for the TZ is given in Table Hangingwall RockRock mass ratings were not routinely performed in the hangingwall mafic rock since the rockquality is consistently high, (averaging approximately 80%) and the development headings are generally not90wider than 5 m. Rock mass ratings were sometimes carried out in larger excavations such as maintenanceshops and lunchrooms.4.4 STRESS DETERMINATIONThe principal stresses at Detour Lake Mine were determined in 1985 by CANMET usingoverconng techniques (Arjang et al. 1985). The measured stresses are comparable in direction andmagnitude to other measurements made at shallow depth in northeastern Ontario and Quebec (Figure 4.19).The vertical stress (03) has a magnitude of 0.029 MPalmetre depth. The major principal stress acts in aENE-WSW direction. For modeling purposes, the major principal stress (01) is assumed to acthorizontally, parallel to the strike of the orebody and with a magnitude of 2,6 times the vertical stress. Theintermediate principal stress (02) is assumed to act perpendicular to the strike of the orebody with amagnitude of 1.3 times the vertical stress.The major principal stress is favourably oriented in approximately the same direction as the strikeof the orebody (Figure 2.3). The stress conditions in the sill pillars will therefore be influenced mainly bythe intermediate stress, 02.Prior to filling the undercut of the three active stopes, five vibrating wire stress meters wereinstalled in each sill pillar at approximately 30 metre intervals along strike. The purpose of thisinstrumentation is to determine actual stress changes in the sill pillar and to use the information to calibratenumerical modeling estimates of the stress changes.The stress meters are normally read on a monthly basis but are read on a weekly basis when the sillpillar thickness becomes 30 metres or less. Figures 4.20 to 4.22 show the results of the stress monitoringfor the 260, 360, 460, 560 sill pillars as well as the surface crown pillar. All but one of the original stressmetres in the 260 pillar are no longer functional. It is believed that the cable has been severed within thefill. Three additional stress meters were installed in January, 1991.4.5 MINING HISTORYA historical record of the mining at Detour Lake Mine has been compiled such that recorded stresschanges can be related to mining activity Figure 4.23 indicates the time span during which mining of eachlift took place.Mining began on the 260 and 360 stopes in 1987 and proceeded at a rate of approximately 2months per lift. Due to the greater ore widths, mining on the 460 Level proceeded at a rate of91approximately 4 months per lift. Mechanized cut and fill stopmg recently began on the 560 Level and isprogressing at a rate of 2 months per lift. Mechanized cut and fill mining of the 260 stope was completedin October, 1991, leaving a 26 metre thick crown pillar between the pit bottom and the back of the 260stope. There are no plans to recover this pillar until the end of the mine life. Mechanized cut and fillmining of the 360 stope was completed in January, 1992, leaving a 16 metre thick sill pillar.92Table 4.1 Summary of Intact Rock Strength Characteristics** reported value, however this is higher than the possible range for rock()number of samples testedRock Type Specific UCS (MPa) Tensile Modulus of Poisson’s Friction CohesionGravity Strength Elasticity Ratio Angle (IviPa)(MPa) (GPa)MAFIC- Hangingwall and FootwallCANMET 2.9±0.02 270±42 (5) 20±1 (8) 88±4 (6) 0.25± 50 (4) 60 (4)(12) 0.01 (6)TERRAPROBE * 9 1±17 (5) 26±5 (5) 0.29±0.07 (5)J.SMITH 147±31 19±(12) 57±13(12) 0.1±0,02(12) (11)MAIN ZONECANMETSilicified 3,0±0,02 (9) 191±1 (3) 20±1 (6) 108±8 (3) 0.24± 51 (4) 44 (4)0.03 (3)Main 2,9±0,01 169±26 19±1 (9) 89±6 (10) 0.26± 43 (4) 40 (4)(19) (10) 0.02 (10)Average 3.0±0.06 174±69 20±2 93±18 (13) 0.26± 47 (8) 42 (8)(28) (13) (15) 0.05 (13)TERRAPROBE * 93±24 (5) 27±4 (4) 0.43±0.12 (4)TALC ZONECANMET 3,0±0.004 92±2 (3) 10±1 (7) 51±5 (3) 0.34± 42 (4) 20 (4)(9) 0.02 (3)TERRAPROBE 93±14 (5) 31±4 (5) 0.53**±0.11 (5)* failed along existing discontinuities93Table 4.2 UBC Unconfined Compressive Strength Test ResultsTalc Zone Number of TestsUnconfined Compressive Strength 48 ± 5 MPa (9 tests) 9Modulus of Deformation 32 ± 2 GPa (5 tests) 5Poisson’s Ratio 0.27 ± 0.06 (5 tests) 5Specific Gravity 3.0Table 4.3 Detour Lake Mine - Rock PropertiesMain ZoneUnconfined Compressive Strength 165 MPa (avg), 190 MPa(upper)Modulus of Deformation 93 GPaPoisson’s Ratio 0.26Specific Gravity 3.0Talc ZoneUnconfined Compressive Strength 28 MPa (lower), 48 MPa (upper)Modulus of Deformation 32 GPaPoisson’s Ratio 0.22Specific Gravity 3.094Table 4.4 Approximate Classification of Rock Based on Strength(after Brownq 1981)No. Description Unconfined Compressive ExamplesStrengthlb/in2 kg/cm2 MPaRi VERY WEAK ROCK 150-3500 10-250 1-25 chalk, rock salt-Crumbles under sharp blows with geologicalpick point, can be cut with a pocket knife.R2 MODERATELY WEAK ROCK 3500-7500 250-500 25-5 0 coal, schist-Shallow cuts or scraping with pocket knife siltstonewith difficulty, pick point indents deeply withfirm blow.R3 MODERATELY STRONG ROCK 7500-15000 500-1000 50-100 sandstone, slate-Knife cannot be used to scrape or peelsurface, shallow indentations under firm blowfrom pick.R4 STRONG ROCK 15000- 1000-2000 100- marble, granite30000 200-hand held sample breaks with one firm blow gneissfrom hammer end of geological picks.R5 VERY STRONG ROCK >30000 >2000 >200 quartzite-requires many blows from geological pick to dolerite,gabbro,break intact sample. basaltTable 4.5 Joint Aperture ClassificationVery Tight <0.1 mmTight 0.1-0.5 mmModerately Open 0.5-2.5 mmOpen 2.5-10 mmVery Wide 10-25 mm95Mean: 73,0Standard Deviation: 7.8Note: Q is calculated using RMR=9lnQ+44Table 4.6 Rock Mass Classification - Detour Lake Mine Main ZoneCASE DATE STOPE RMR QNO. RECORDED LOCATION (%) —CASE DATE STOPE RMR QNO. RECORDED LOCATION (%)35A5B9101 IAI1B12A12B141517A17B18222325A25B27A27B333435363738A38B3940414246A46B51545556575859606262B636466A66B6768697072A72B737576777980FEB/1990FEB/1990FEB/ 1990FEB/ 1990FEB/1990FEB/1990MAR/1990MAR/1990MAR/1990MARJ1990MARJ199OMAR/1990APR/1990APRJ199OAPRJ199OAPR/1990APR/I990MAY/1990MAY/1990MAY/1990MAY11990MAY/1990MAY/1990NOV/1989NOV/1989NOV/1989NOV/1989NOV/1989NOV/1989NOV/1989JANI2I9OJANI2I9OLAN12/90JANI2/90FEB13/89FEBI3/89DECI2/87OCT3/89SEP 17/88OCr/87SEP/87JUNE/90JUNE/90JUNE/90JUNE/90JUNE/90JULY/90JULY/90JULY/90JULY/90AUG/90AUG/90AUG/90AUG/90AUG/90AUG/90AUG/90AUG/90SEP/90SEP/90SEP/90SEP/90LUMlff I I230M3#1 1330M4#10330M3#1 1430M3#5430M3#5230M4#1 1230M3#1 1330M4#10330M4#10430M3#5430M3#5200m5#12230M4#1 1330M4#1 1330M4#1 1300M5#12200M5#12200M6#12300M5#12W300M5#12E430M6#6430M6#6230M3#10230M3#10230M4#10330M3#10330M4#10E430M3#5430M3#5230M3#1 1240M4#10330M4#10330M3#1 1460M1#3460M1#3360M2#2460M2#4360M2#4360M1#1260M1#1200M5#13200M6#12300M5#12430M3#5430M3#5200M6#13300M5#13430M3#6430M3#6200M5#12200M6#13300M5#13300M6#13430M3#7430M3#7W430M4#746OLONGHOLE200M5#14300M6#13430M4#7430M3#7858767777878858773637878738573637773777767787877878577637878878563777253606655696964707778788066646480806679796477678079797795.2118.812.939.143.743.795.2118.825.18.343.743.725. ISEP/91SEP/91OCT/91OCT/91OCT/9 IOCT/91OCT/91OCT/91OCT/91NOV/91NOV/91NOV/91NOV/91NOV/91DEC/91DEC/91DEC/91DEC/91DEC/91DEC/91JA2’1/92MARCH/92MARCH/92MARCH/92MARCH/92MARCH/92MARCH/9246OLONGHO200M5#14200M6#14300M5#14430M4#8430M3#7200M6#15300M6#14430M4#846OLONGHO200M6#16200M5#16300M5#I5430M4#8430M4#8430M3#8200M6#I6300M5#15300M6#16430M4#9430M3#8200M6#17200M5#17300M5#I6300M6#16430M4#9200M7#18200M8#19300M5#19300M6#16430M4#10200M7#19200M8#19300M5#17300M5#17430M4#10200M7#19200M8#20300M5#18300M5#10300M6#17430M3#I0430M4#I 1200M7/M8#20300M5M6 #18300M5M6 #18400M5#13430M4#1 I200M7/M8#20300M5M6 #18300M5M6 #18400M5#135605429266054292300M5M6 #18300M5M6 #18560M 192560M2#256054292575SLR59OSLR678080737777797877677082687878687068767868706768767879776964697703636369787965657580SI78636375797863637554756363815454707212.954.654.625. 4.7 Rock Mass Classification - Detour Lake Mine Talc ZoneCASE DATE STOPE RMR QNO. RECORDED LOCATION (%)6781316192021242628293031A31B3243444547484950525361657174788183848894100106113114114120A120B126127128129136137FEB/1990FEB/1990FEB/1990MARJ199OAPR11990APRJI99OAPRJI990APRJI99OMAY/1990MAY/1990MAY/1990MAY/1990MAY/1990FEB/1990FEBII99OFEBII99OJANI2I9OJANI2/90JANI2/90FEB 13/89NOVI/88FEBI3/90JANI9/89FEB28/89FEB28/89JUNE/90JULY/90AUG/90AUG/90SEP/90SEP/90JULY/90JULY/90OCT/90NOV/90FEB/91MARCH/91APRIL/91APRIL/91APRIL/91JULY/91JULY/91SEP/91SEP/91SEP/91SEP/91OCT/91OCT/91430T40#5430T5#5430T60#5430T60#526OT7OACCESS430T5#6430T40#6430T60#626OT7OACCESS360T20#4430T40#6430T20#6430T60#6360T40#3/4360T40#3/4360T20#3430T5#5430T40#5430T60#5230T40#5230T4064460T60#3360T40#3360T60#2360T60#1360T20#40360T40#10360T40#I0430T5#7360T40#10430T60#7560T15#156OTACCESS360T40#1 11360T60#1 1360T60#13360T60#13360T60#14360T60#14360T20#l0360T40#17360T40#17360T40#18360T40#18360T20#14360T20#14360T40#18360T40#185167505042624952425849614248485867555048482548282858484067406043255538383838386545454343565645452. 4&9Standard Deviation: 11.2Note: Q is calculated using RMR91nQ+4497Table 4.8 Typical CSIR Rock Mass Rating at Detour Lake MineCategory MAIN ZONE TALC ZONEDescription Rating Description RatingStrength 160-18OMPa 13 35-5OMPa 4RQD 90% 17 80% 16Joint Spacing 0.4 m 16 0.3 m 9Joint Condition smooth, hard, tight 17 smooth surfaces, soft 10Groundwater none 10 none 10Joint Orientation 0 0TOTAL 73 4998ri)UCa)0U)U-Detour Lake Mine - Main Zone UCS(Includes data from surface, 360 mL and 560 mL)Uniaxial Compressive Strength (MPa)Detour Lake Mine - Talc Zone UCS(includes data from surface, 360 mL and S6OmL)Uniaxial Compressive Strength (MPa)DETOUR LAKE MINE UCS TESTiNG FIGURE 4.1MAiN ZONE AND TALC ZONE2624222018161412100 25 50 75 100 125 150 175 200 225 250 275 300 325 350 375 400 425 450 475 5005010HHH I0 25 50 75 100 125 150 175 200 225 250 275 300 325 350 375 400 425 450 475 5009926242220Detour Lake Mine - Hangingwall UCS(includes data from surface, 36OmL, and 56OmL)C6)zU-I I I I I I I I I225 250 275 300 326 350 375 400 426 450 475 500Uniaxial Compressive Strength (MPa)Average UCS = 235.4 MPaDETOUR LAKE MiNE - UCS TESTING FIGURE 4.2HANGINGWALL MAFIC ROCK1818141210B6420 111111ih0I I I I I I I26 60 75 100 125 150 175 200100UBC UCS Testing - Talc Zone353025B20w3151:0250! 2004-Bj 150100C.)‘I0UC0UBC UCS Testing - Main ZoneSample NumberDETOUR LAKE MINE - UCS TESTING BY UBC OFTALC ZONE AND MAIN ZONEFIGURE4.31 2 3 4 5 6 7 8 9 10Sample Number1 2 3 4 5 6 7 8 9 101010.4-IUBC UCS Testing Hangingwall2502001501005001 2 3 4 5 6 7 8 9 10 11Sample NumberDETOUR LAKE MiNE - UCS TESTING BY UBC OF HANG1NGWALLMAFIC ROCKFIGURE 4.4102UCS Testing SummaryDETOUR LAKE MINEUCS TESTING SUMMARYFIGURE 4.5250200150100500HW MAIN ZONE TALC ZONEILIUBC ResultsCANMET ResultsTerraprobe ResultsVoest Alpine ResultsJ. Smith Results: Zone Mean UCS Standard Deviation No. of SamplesHW 165 27 11Main 166 22 10Talc (lower) 28 3 10Ta’c (upper) 48 9103ProjectBench.Drift,Ramp:Traverse:Ret.PointCoord.SurveyDirectionLength:I/I///,/1--—---.-..---.-..-.-.—.-------:—---------:14.NGICIossIflcotlonj-j.CSIRCIassIfTcatIonPMRA,B.C55.—-----——_________________J.2—---————--—-C)DelcititA-Rotg(•SQO.*itn,J,.——-AsengmO.xmvCordIloreliI-———-JB.500tvstrn-S4iij.sso.o00sDo,Co00e-.CtrB5DosB’QC) en C - c-i C) > C) 0UII DETOUR LAKE MINE eot.Es IPOLE PLOT I2600 I‘DETOUR LAKE MINEPoLEs[oLE PLOT j360T20*4L1••. NDETOUR LAKEPOLE PLDTMINE 27 POLES360140 #3/4NAPPING PROGRAM JULY/90POLE PLOTS FROM FABRIC MAPPING FIGURE 4.8NI.DETOUR LAKE MINEPOLE PLOT106CONTOURS OF POLE PLOTS - DLM MAPPING FIGURE 4.9107C-Sr3 ja-cU,I—jDz—JzUiIiijMD<cL-Ja::DDDIt.i 0_J‘0IUiUi—iMD<0-JLXDDDII-zWDmc-)• .11 .ii—• - .‘s.i ii if..._i+iiI i/I- .33 I •‘lI + I ‘%.‘% I I I:5:11+1+.1+1-1.+-..i ++/ i:zI— wU ..J—0.0- CuI—z-rLiiUJJMDzD-JLXDI—I-zbiD.0 U0 e,(i.000000000s6 gcO’00000%1111100000qqqqqU) O 1’. 00’00000-+ -‘ d0000qqq11111U.iZ.iji/ iiI I-iflitA +1/i+iIii +1.fill IiiI I i+d.xl •..i Si UI I ..i 1+‘I I .\ . . .11\\ + . i .\ iIII.i. .11 . . I .3 .1. . .‘..+ d+•%.d C4U>‘S.0‘0-Cuz-lUiMD-jLXDI—I—zbiDc-)1++dU\I Iii+I•...I .i++I• — ..i i%I I• . ..3 . . .I’%i/ii.it../I I I .•-I I U+ I1QUc4!21.3W-Ja-c0-CuI—biDLiibJiMD<0-jLXDDDII-ZbiD6ZJJDJa-c0-03I—LiiLIJMD-JLXDII-zbiDTJfl!N57Y1DETOURLAKEMINE97PULESCONTOURPOLEPLOT46OmL#3,(-J C C,’_____________________________________________C 11 C -e C C,,______________________DETOURLAKEMINEPOLESCONTOURPOLEPLOT260r,LU-NC/V,’Ic.E1+,• ——————.I—..—/c+,——,•—-/—-+..----+---•••—/+—-+1+—.--.--LEGEND-—1/—__________________________________SDETOURLAKEMINE102POLESDETOURLAKEMINE[°POLESCONTOURPOLEPLOTSj6OnLfl2CONTOURPOLEPLOT[60J.iLE112N•,•.—/++,,•I..*1/.++.11e,,.,,,,,,‘+,IIII•I__,•.•\—a—.—+—•,——ib+••.,**,•,....+,I•—I——..SNZ+/b—a—\/+,+.,I.1.,,_,,II0.q+b/—.—\+b++I*II—IIS00Nz——I+,-/+/-7,--...,/++*,*__,,,+II••I’II•-,/0.—.Ib,,++—+—I——.+..,.vIrb,,.aE‘11,/a,+—....b\++_.__...,I—.-I.,,/-+÷- —•—/baII*1*——I+—,+/—SDETOURLAKEMINEJ64POLESCONTOURPOLEPLOTt6Qi’L‘d#4DETOURLAKEMINEfrzpoxsCONTOURPOLEPLOTJLEft4Tj C 0kk’ffiLLLLTNV7.,,—,,++,+//\,•*——1——,•,Il_.I— I—+IVIIcii/f——Il,—++i•————•\.++——.IIII..—CONTOURS-Jio.oo—oi.co•os.oo—06.00b01.00—02.00—06.00—07.00c02.00—03.00+07.00—08.00ci03.00—04.00/08.00—09.00e04.00—05.00a09.00—99.90f-3 C Ct) C - C ru C -3 Ct)DTDURLAKEMINECONTOURPOLEPLOT36OmL#3Ic:::.zIDETOURLAKEMINECONTOURPOLEPLOTI36#5 /.————.—.—.—‘\.4•,—,,,,\:+—:——+—S____________________DETOUNLAIcEMINE56POLESCONTOURPOLEPLOT360ML#3TN—...—/b—.N——,-/1—.\7.,,———.../b../\1,11.•__...III’/•,••I_i..*11Ii—\‘iI.,.•.I61114.IEI——.——.—--/b—..I,,._.__,—,..1————.—b/../1.,.d,1’•sbb——....—I_IIIIII———.b/.SDETOURLAKEMINE[POLESCONTOURPOLEPLOTI3GomL#6&.ilIHTTiWhE:::‘CONTOURUiliS00.00—01.0001.00—02.00—oaoo—03.00+03.00—04.00/04.00—05.00u•DETOURLAKEMINEPOLESCONTOURPOLEPLOT6OtL#4DETOURLAKEMINE113POLESCONTOURPOLEPLOT360mL#705.00—06.00b06.00—07.00c07.00—08.00d08.00—09,00e09.00—99.90f’DETOUR LAKE MINE i pai.zs DETOUR L.AKE MINECONTOUR POLE PLOT 36 CONTOUR POLE PLOT JN N-Vas.. -—.......+— +I— +— . I —+Io + I++. . . —— II . . !E -.d--—_ II,.— — •—I fo.— I—+ I \+. —\_+_+/__..——ii I ++. •\. 1/. —/+— . . — / \—+. —/1. —\.——o./.—....——./ \.‘— —).•/o.o._. ..—.—.... ‘..——..+o.b/.———../ “%..+ ——S SCONTOUR LEGEND00.00—01.00. 05.00—0&DOb01.00—02.00— 06.00—07.OOc02.00—03.00+ 07.00—08.OOd03.00 — 04.00/ 08.00 — 09.00 e04.00 — 05.00 c. 09.00 — 99.90 fCONTOURS OF POLE PLOTS - DLM MAPPING FIGURE 4.12110MAIN ZONECONTOURS OF POLE PLOTS - ALL STRUCTURES IN MAIN ZONE AND FIGURE 4.13TALC ZONEa+)\C1*ItOT POLEcONcENTRAT I OHSof total par1.0 .t area<0< 10.5 xEQLLIII. II1ISPI675 POLES675 ENTRIESHO BIASOORCT I OH:CIIII Dl POLECOHCEHTRAT I OHSX of total par1.0 X araa<0 x<7 XLIII. IEHI SPIEJ634 POLES634 ENTRIESHO BIASOORIECT I 0t4TALC ZONE111TYPICAL ZOUGHNU POFILIS far JZC ta•gI 1— 0-22-43 H- - .4 -4 F———-——--—--——-—-----—--—--—-f - - ‘- s-”-..5. __—--——---——--——--j 1 -to6,—- —._7 12-148 14 -169 16 -1110 •11 -200 S 10I.I — SCALS. TYPICAL ROUGHNESS PROFILES FOR JRC RANGES FIGURE 4.14(after Barton et al., 1977)112JOINT ROUGHNESS300280 -260 -w 240-C-)220-200-C-)C) 180-0160140-z1200 -w 100‘- 8060 -40 -20 -0-130120110100C)o 90 I0>.. 80C)zw 70 1D Iow 60LI50•iiIHL..i,ir.u10JOINT ROUGHNESS NUMBEREZ] MAFICS TALCSROCK HARDNESSR5LiR4ROCK HARDNESS1:;[i 1FICS TPLCSRi. DETOUR LAKE MINE FABRIC.MAPPING FIGURE 4.15JOINT ROUGHNESS AND ROCK HARDNESS FREQUENCY113WIDTH OF JOINT OPENINGAPPARENT WIDTH[:::.: MAFICS — TALCSb4FICS — TALCSDETOUR LAKE MENE FABRIC MAPPENGJOINT APERTURE AND SPACINGFIGURE 4.16:1VT T170160150140w130w120110()o 100U-o 90>-80wD 70060U-403020100140 —130 —120 —w 1100100-900C-)o 807060D50- 4c30 -20 -10 -P0 0JOINT SPACING>3 1-3mL0.3-1 m <5cm5-30 cmJOINT SPACING114‘UC)z‘UC)C.)0U00zuJ0‘UU-160 -150 -140 -130z -w 120110-8 100-90U80C) 70z600Ui 50- 4030JOINT CONTINUITY1AFIC3-5mLength— TALCJOINT ENDS VISIBLEt14FICS — TALCDETOUR LAKE MINE FABRIC MAPPINGJOINT CONTINUITY AND ENDS VISIBLEFIGURE 4.175-lOm220200180160140 -1201008060402000I :z:jENDS VISIBLE2115ROCK MASS CLASSIFICATIONDATE:,45I nCATIrM•None020Completely dry10NAME:______________2120<25%3<50mm5Soft gouge>5 mm thickorJoints open >5 mmContinuous Joints0>125 Itireshiiin>0.5Severewater problems0DESCRIPTION VALUESfrengthRQD aointsi1)Joint SpacingCondition of JointsGroundwaterJont OrientationTOTALJOINT SETSStrike Dip SpacingJointSetlJoint Sot 2Joint Set 3ROCK MASS CLASSIFICATION FORM FIGURE 4.18RatingDfl Core Quality (RQD)15Rating90%—100%12Spadng of Joints2075%-90%7Rating>3m1750%-75%43014mCondition of Joints1325%-5o4250.3-ImVery rough surfacesNot continuousNo separationHard joint wall rockaRatingPARAMETER RANGE OF VAIUESPoint load Forthls low rangestrength Index >8 MPa 4-8 MPa 2-4 MPa 1-2 MPa uflla)dal compressive.p I teat Is preferredynIsxIeI Imaterial Compressive >200 NRa 100-200 MPa 50-100 MPa 25-50 UPs 10-25 I 3-10 I 1-3UPa UPs_I_MPa2050-300mmSlightly rough surfacesSeparetion< 1mmI-lard joint wall rock2510SlIghtly rough surfacesSeparatlon<lmmSoft Joint Wad RockGreundwaterInflow per lamtunnel length$—-0-.Gouge <5mm thIckSllckenald SurfacesJoints open 1-5 mmContinuous iolnta12 6General conditions<25 libee/minRatingor25-125 litreshrnnor0.0-0.02 02.5.5MoIst only(Interstitial Water7Water under moderatepressure4116C’, H C,) H C,,Cd)C,, H I?190 2?Okm260 SILL PILLAR STRESS MONITORING FIGURE 4.2015260 SILL PILLAR STRESS MONITORINGStress Meters installed Dec/901050(Uaa)CCU-c0U)Cl)a)U)(801)0)C(UC)U)U)‘I)C/)11/23/90 03/03/91 06/11/91 8/19/91 12/28191 04/06/92 07/15/92 1023I92 01131/93K 260-1 260-2 0 260-3260 SILL PILLAR STRESS MONITORINGOriginal Stress Meter-5-10- K-1.00 --2.00Aug-87 Dec-88 May-SO Sap-SI Jan-93v 5-104118360 SILL PILLAR STRESS MONITORING3-a,0,C(C-)U,U,4’004’0’C-CC.)0)U)Cl)360 AND 460 SILL PILLAR STRESS MONITORING FIGURE 4.2110.005.000.00-5.00-10.00Aug-87 Dec-88 May-90 Sep-91 Jan-93X 7-102 7-104 0 7-107 7-108 • 7-109460 SILL PILLAR STRESS MONITORING20. Dec-88 May-90 Sep-91Z 9-101 9-103 0 9-110 9-111 • 9-112Jan-93119560 SILL PILLAR STRESS MONITORING7.006.00 -500 -4.00 -3.00 -a)0)2.00 -C-)1.00 -0.00-1.00 --2.00 -___________-300Jun-91 Sep-91 Dec-91 Apr-92 Jul-92 Oct-92—4— 560-1 —h— 560-2 € 560-3CROWN PILLAR STRESS MONITORING7.006.00 -5.004.00 -0.3.00 -a,2.00 -C.)U) 1.00 -U)0.00-1.00 --2.00 --3.00’Dec-91 Feb-92 Apr-92 May-92 Jul-92 Sep-92—— 200-1 • 200-2 —e— 200-3 : 200-4SURFACE CROWN PILLAR AND 560 SILL PILLAR STRESS FIGURE 4.22MONITORING120mmUII Imx0zmmIIm0zUmII I-48Cdz<CdCdd%%‘-I-lu-u-u888CdCdCi-J8CdI‘I--J8CdIU-ICdIU.-48CdU.8CdU.-J-.48Cd-JCd8CdI’-j8IU.-4Cd2 8qm%%I-I_I-IUU.U.U.CII4.8-J8CdIU.8qCd-l8IU--48qI—U-I8DETOUR LAKE MINE PRODUCTION SCHEDULE FIGURE 4.23121UIIw>0zm ImxI.CD)--zIIII IIII IIa.xa. II I-J=II’=CII=C.,-J0*z*=C*Iz1%*==C.,=C,*I:0*=0*=§-JC.,*=UI*=UJa, I0*=C.,DETOUR LAKE MINE PRODUCTION SCHEDULE FIGURE 4.23. (continued)1220II Im I I00IIm -xI-a.‘,I0zzI III IIIzj::IIII II-IaxUaz I* *a,*U,C.,*C*U,*CU)*tzU)*C*U,-J0*=*D*C41*c.1*=C.)a, LiiItz t;:===== IDU)CU)CU)§gU)*C.,gp.*p..*DC.,DETOUR LAKE MINE PRODUCTION SCHEDULE FIGURE 4.23. (continued)123C)0z>-jzU-.mD.i-.-. —IIII ID4z•11CgoII Igo1:U,III0IC‘1,IC=Ilb=CIU)lb=goIgolb=U,IClb=goI1%lb=U,I§lbtz‘0IClbIz§Clb=C.)IClbI0lbDC.)I*IDETOUR LAKE MINE PRODUCTION SCHEDULE FIGURE 4.23. (continued)124C C 7: I —r1(D-4fi0LEVEL400M6UFT#12t’.)400115LJF#12400M5UFT#13400M6UFT#14400115L1Fr14—.19923ANUAJYFEBRUARYMARCHAPRILMAYJUNB_JULYAUOUSTSEPTEMBEROCTOBERNOVEMBSRDECEMBER360LEVEL300UFTI8FINAIU—.--I560LEVEL560112liFT2560111IJFTAI256OM2UFT#3560111LIFT#35. NUMERICAL MODELING5.1 THREE-DIMENSIONAL BEAPThe shape of many underground excavations and the influence of neighbouring excavations make2-D plane strain analysis inappropriate for many practical mining problems. In such cases, a three-dimensional analysis may be required to achieve a reasonable degree of accuracy.Three-dimensional boundary element modeling has many practical applications for rock mechanicsengineers including:• pillar design;• open stope span design/dilution studies;• shaft and service tunnel layouts;• analyses of complex excavation geometries;• stope sequencing studies; and• parametric design studiesBEAP-3D or Boundary Element Applications Package has been developed by CANMET and wasused in the course of this project to carry out three dimensional stress analysis. BEAP-3D is a powerfulnumerical modeling package designed specifically for modeling three-dimensional underground openings.The version used was capable of modeling up to 1000 elements. The program utilizes a graphical preprocessor called Mine Designer (CANMET, 1991) for creating a model geometry file as well as agraphical post-processor called ViewBeap (CANMET, 1991) for viewing stress contours.As with any numerical modeling procedure for determining stresses, the accuracy of the BEAP-3Danalysis depends on the accuracy of three main input parameters:1. the stress-strain relation(s) of the material(s);2. the pre-mining stress conditions; and3. the model geometry.The BEAP-3D analysis assumes the materials to be linear elastic, homogeneous, and isotropic.This assumption is considered accurate for laboratory scale specimens of intact rock but usually does notrepresent the stress-strain relationship of the rock mass since the strength is controlled by thediscontinuities. Inhomogeneity of the rock mass can be modeled by using different material parameters fordifferent groups of elements in the model.126The pre-mining stress conditions must be determined as input for the analysis. Both the magnitudeand direction of the principal stresses are required. These values are normally determined from in-situstress measurements.Three dimensional modeling permits better accuracy in modeling the actual mine geometry. Thedetail of the modeling is limited by the number of elements the program can handle. Large modelscontaining 800-1000 elements require approximately 15 hours to run on a Sparc workstation. As with anynumerical modeling program, the results obtained should be carefully scrutinized and applied with a gooddegree of engineering judgment, recognizing the simp1if’ing assumptions contained in the model.5.2 NUMERICAL MODELSThere were three main objectives to be achieved through the 3D numerical modeling:• Post Pillar - to estimate the stress conditions within a post pillar;• Sill Pillar - To estimate the stress conditions in the sill pillar and todetenmne whether horizontal stress was contributing to observed cases ofinstability; and,• Overall Mine - To locate areas of high stress concentration not alreadypredicted and which may contribute to instability.5.2.1 Post Pillar ModelingPost pillars are used at Detour Lake Mine as a means of reducing the exposed span in wide cut andfill stopes. The use of post pillars as a means of support is discussed in greater detail in Section 7 of thisreport. Post pillars are started on the footwall side of cut and fill stopes. With each lift mined, the stopeboundaries shifts southeast due the plunge and dip of the ore. The vertical post pillar migrates towards thecentre of the span. These pillars are typically 5 metres square and up to 25 metres in height. It is intendedthat the pillars will yield as the width to height ratio decreases, however experience at Detour Lake Mineand other operations demonstrate that the post-yield strength of the pillar is still capable of providingsupport to the immediate back.A 25 metre high post pillar has been modeled with BEAP-3D employing 489 elements. The modelassumes an opening 40 by 50 metre excavation, 25 metres high, plunging 45° to the west and dipping at56° to the north (Figure 5.1). The post pillar extends from the top centre of the modeled stope to the eastside of the stope which is commonly the case at Detour Lake Mine. The excavation and post pillar areplaced at a depth of between 420 and 460 metres below surface to simulate conditions on Pillar 941 of the460 Stope at DLM. This pillar was the subject of a support trial whereby the back was pre-supported with127cable bolts and the pillar removed. This support trial is described in further detail in Section 7 of thisreport.In the computer model, the uncemented backfill has been assumed to be incapable of carrying loadfrom the pillar since the elastic modulus of the fill is much lower than that of the pillar. In addition, theactive confining stress provided by the fill is assumed to be zero. In this sense, the model can be consideredconservative since the fill does indeed provide confinement to loose blocks which would normally spall offthe pillar. This confinement allows strength to be maintained in the pillar core.The major and minor principal stresses through the pillar are shown in Figures 5.2(a) and 5.2(b).The stresses are shown on a horizontal plane through the centre of the pillar. The higher stresses are shownto be on the east side of the pillar due to the pillar being “grown” from the east wall. The weighted averagepillar stress across the east-west centre line of the pillar has been calculated to be 19.8 MPa. Themaximum a1 of about 40 MPa occurs on the northeast and southeast corners of the pillar. Contours ofthe horizontal confining stress, a3, are shown in Figure 5.2(b) on the same horizontal plane. The confiningstress is shown to be less than zero across the entire cross section. The average pillar stress computed byBEAP-3D is much lower than what would be predicted using other estimation methods such as tributarytheory:rock column areau, =yghpillar area— (2700)(9. 81)(435)(475)Up— 25 (5.1)=219 MPaDespite the low average pillar strength relative to the intact uniaxial compressive strength of therock (165 MPa) the pillar can be expected to yield due to the existing structure and the lack of confiningstress, Using the modified Hock and Brown failure criterion for jointed rock masses the maximumprincipal stress at failure is given by (Hock et al., 1992):(O’ +ocmb—J (5.2)where,m and a are parameters which depend on the quality of the rock massThe failure constants m and a have been estimated to be 3.4 and 0.45 respectively, consistentwith a good quality, very blocky, fine grained basalt. Clearly, the use of this failure criterion predicts zerostrength when there is no confining stress. This confirms the expectation that the pillar has yielded.128Despite having yielded, post pillars at Detour Lake Mine do have residual, post-yield strength andcontinue to provide support to the immediate back. This will be documented in Section 6 of this reportwhere it will be shown that unstable back conditions develop where spans exceed approximately 20 metresat Detour Lake Mine. Where post pillars are used with a span of 20 metres between pillars, overall spancan be increased to the full width of the orebody (35-40 metres). It can therefore be concluded that a postpillar provides support to the stope back making use of its post-yield strength.The confinement provided by the fill makes a significant contribution to the post-yield strength ofthe pillar. As the pillar yields, it dilates and compresses the fill so the fill approaches the passive Rankinestate. Assuming a passive Rankine earth pressure coefficient, Kp3 .5, and a depth of fill of 7.5 m to thecentre of the 25 m tall pillar, the horizontal confining stress is calculated to be 0.5 MPa. From equation5.2, the post-yield pillar strength is then be estimated to be 22 MPa.Additional BEAP-3D runs were carried out for pillar heights of 20, 15, 10, and 5 metres. Themajor and minor principal stresses were found to increase at the mid-height of the pillar as the pillar heightdecreased. Figure 5,3 shows the increased ai and a3 stresses for a pillar height of 15 metres. Therelationship between pillar height and pillar stress as determined by the modeling is given in Figure 5.4.This relationship cannot be used for general design purposes since pillar stress is also a function of theexcavation geometry and the pre-mining stress field.5.2.2 Sill Pillar ModelingThe sill pillar modeling was carried out to assess whether high horizontal stresses were developingin the stope back as the sill pillar thickness decreased and whether these stresses contributed to an increasedfrequency of instability in the stope back. Previous boundary element modeling using the 2D BEMprogram EX4MJNE-2D, was believed to be overestimating the stress level since a composite verticalsection had to be used which modeled all stopes on the same vertical section. In reality this does not occurbecause the stope is plunging to the west at 45°. In addition, the influence of the backsiashed attack rampson the induced stress around the stopes could not be handled with a 2D model. The backsiashed attackramps were driven approximately perpendicular to the maximum principal stress so high stresses could beexpected to develop between attacks driven from different levels.The entire Detour Lake Mine was modeled from the 460 Level to surface. Major components ofthe model include the open pit, 260 Stope, 360 Stope, 460 Stope, Quartz Zone longhole stopes, and the 360Talc stopes. The components of the model are shown in Figure 5.5.129The major principal stress, a was approximated as acting horizontally in an east-west directionwith a magnitude of 0.075 MPa per metre depth. The intermediate principal stress was approximated asacting north-south with a magnitude of 0.038 MPa per metre depth. The minor principal stress. 03 wasassumed to be vertical and having a magnitude of 0.029 MPa per metre. The rock mass was assumed to behomogeneous, isotropic, and linear elastic with a Young’s Modulus of 30 GPa and Poisson’s Ratio of 0.26.The main model was run to simulate the mine geometry as it existed in February, 1992 with pillarshaving the following thickness:Thickness (m)Crown Pillar 24.1260 Sill Pillar 13.8360 Sill Pillar 53.0Figure 5.6 is a view of the entire mine showing the major and minor principal stresses at thesurfaces of the stopes as well as in the sill pillars. Figure 5.7 is a close up view showing stresses in thecrown pillar on a subvertical plane through the pillar. The magnitudes for maximum and average stressesin the pillar are provided in Table 5.1. The highest stresses occur in the lower centre part of the pillarbetween the top attack ramps of the 260 stope. The average pillar stress along the midline of the pillar is14.9 MPa with an average confining stress of 1.2 MPa. Applying the modified Hoek-Brown failurecriterion, the factor of safety for the pillar (defined as the average pillar strength divided by the averagepillar stress) is 2.17. A factor of safety of 1.5 is generally considered necessary for permanent support inunderground mines (Hoek and Brown. 1980). A maximum horizontal stress of 17.2 MPa occurs in theimmediate back of the 260 stope which is only one tenth the umaxial compressive strength. It can thereforebe concluded that high stresses were not a contributing factor for cases of instability recorded in the 260stope. Rather, the lack of confining stress in the immediate back results in relaxation of the rock into theopening which can lead to structural failure and general rock mass failure.Figure 5.8 shows maximum and minimum principal stresses in the 260 sill pillar. The highestpillar stress occurs in the centre of the pillar as an elongated band between the two attack ramps. Highstresses can also be observed at the top of the attack ramps indicating that they are likely contributing tohigh stresses in the pillar. The average a 1 pillar stress measured on a vertical line through the middle ofthe pillar is 35.6 MPa. The average confining stress, 03, 15 5.4 MPa. Applying the modified Hoek-Brownfailure criterion, the factor of safety is estimated to be 1.91. In the immediate back of the 360 Stope thehorizontal stress is in the range of 22 to 26 MPa. Again, this suggests that high stresses did not contributeto cases of instability recorded in this stope.130Figure 5.9 shows the major and minor principal stresses for the 360 sill pillar. Horizontal stress isbeing concentrated above the east attack ramp of the 460 stope and below the west ramp of the 360 stope.As the pillar decreases in thickness as mining proceeds, these two attacks will further concentrate stress inthe pillar at this location. The average major principal stress in the pillar is 33.5 MPa and the averageminor principal stress, (33 is 7.0 MPa. This yields a factor of safety of 2.3 for the pillar.5.2.3 Other AreasAnother area of stress concentration is between attack ramps driven approximately one above theother. Since the in-situ major principal stress acts roughly perpendicular to these attack ramps, highstresses can be expected to be easily generated. An example is shown in Figure 5.10 is the area betweenthe #3 and #5 attack ramps of the 360 Stope. High al stress in the order of 50 MPa is observed in thisarea. Stress is particularly concentrated at the intersection of the attacks and the stope. These high stressareas do not generally pose a threat to stability since they develop in the floor of the backslashed attack andaway from active workings.131Cj‘aCi)Ci)==— 1Y9UI-————c,JcrQ.‘D-,I-1)è—zzf::DzzzEIIz)‘J—————:‘,00II—L)—t’3a-tLt’JJiE-‘————O—00C’t’J00L..)—POST PILLAR 941 FIGURE 5.1BEAP-3D MODEL GEOMETRY133r.:——Stress Contoursin MPaFIGURE 5.2(a) a1 Stress (h) a3 StressPOST PILLAR 941 - MAJOR AND MINOR PRINCIPAL STRESSES AT MIDHEIGHT OF PILLAR134(a) a1 StressS1irii.c,oo-p.440[ool7.Oc11 • £0• 0001prarsL. OO’2.000. 00C•1. UtL’:— i.—2.000—. to-000(h) a3 StressPOST PILLAR 941 - 15 M HEIGHT FIGURE 5.3MAJOR AND MINOR PRINCIPAL STRESSES AT MID-HEIGHT—— S---& p -V-——— a..-.-•-- aStressContoursin MPa135Pillar Height versus BEAP-3D Pillar Stressigigma1Sigma 30• I I-100 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30Pillar Height (m)POST PILLAR 941 BEAP-3D PILLAR STRESS yERsus PILLAR HEIGHT FIGURE 5.4136460 Longhole Stopex or Y •- CDETOUR LAKE MINEBEAP-3D MODEL GEOMETRYFIGURE 5.5Open260 Stope360 Talc Stopes360 Stope460 Stope137U\\ .A\I/ /(a) r1 Stress on Excavation Boundaries(b) u Stress on Excavation Boundaries•? ):-..r.e‘ -.L.4 ?XPRINCIPAL STRESSES ON EXCAVATION BOUNDARIES DETOUR FIGURE 5.6LAKE MINE138•.i :PRINCIPAL STRESSES IN CROWN PILLARDETOUR LAKE MINEFIGURE 5.7Crown Pillar . Li 1: .. F(a) o StressCrown Pillar ‘260 Stop(b) o Stress139—--flt— --—r-t -‘Z--’t — —.L_ — ., -,— —.N Wi,je—-——h, .---‘-_.- ;c.—-‘zL, .:zs...k..... n.-- -Z•aesaaaaaar—— 1LN .aSa(a) a1 Stress-e—N 4- 4 - - -—‘4w aINrrr>.+ Z.-”--.—.—-____..___-,.--c.Lr——L--L2-zL._.—-—— ‘1..-.t4.(4N__N(b) 2 StressPRINCIPAL STRESSES IN 260 SILL PILLAR OF FIGURE 5.8DETOUR LAKE MINE140.,0.0035.5031.0026.5022.0017.50¶3.003.5004.000-O.5U0-5.000S qta3.• 20.0017.50.*5.00:12.50!to.oó:?.5005.0002.500i*bQ00çVLr2500:000PRINCIPAL STRESSES IN 360 SELL PILLAR OFDETOUR LAKE MINESI gmi.XM.. c,.c2O.tC• is. :•c• Ic.c.o-D. YJ(a) G StressI7OIE.OOi2so. fl$O• . :co(b)°‘2 StressFIGURE 5.9141(a) o Stress(b) 02 StressSiqL50.0044.n3.0O33.5028.0022,.5011.50• 6.0000.5(10-5.000Si q35000•44.5033.5028.0017 00ti-SO6.0000.5(10.—5.000.PRINCIPAL STRESSES BETWEEN ATTACK RAMPS OF 360 STOPE FIGURE 5.101426. SPAN DESIGN6.1 DATABASEThe Detour Lake Mine database consists of 172 records of observations made from thecommencement of underground mining in 1987 until March, 1992. Each record consists of the location,Geomechanics Rock Mass Rating (RMR), stability, span, support, and a note of the existence of majorstructures or flat jointing. The Geomechanics Rock Mass Rating has been described in Chapter 3. Theterms span, stability, and support require further definition.6.1.1 Definition of SpanFor the purposes of this study, the term span refers to the diameter of the largest circle which canbe drawn within the boundaries of the excavation as viewed in plan (Figure 6. la). This definition has beenadopted for two reasons. First, entry-type stopes can commonly have width (FW to HW) to length (alongstrike) ratios greater than 0.5, making plane strain analogies invalid. (Other analytical methods define spanin a way which assumes plane strain conditions.) Secondly, post pillars and irregular stope boundariesmake the calculation of hydraulic radius (used by other researchers) difficult and misleading. In cut and fillstopes, the span should also include the overhang material on the hangingwall, which is not contained bythe fill (Figure 6. lb).The term unsupported span refers to spans with no support or spans with pattern rock bolting(commonly 1.8 m long bolts on 1.2 m x 1.2 m pattern). It does not include spans which are supported bymore intensive ground support such as timber sets, shotcrete, cable bolts, or post pillars. The pattern rockbolting is designed to control loose close to the surface of the excavation, which may develop after scalingas a result of nearby blasting vibrations or stress redistribution caused by subsequent mining activity.6.1.2 Definition of StabilityBieniawski (1984) has shown that there is a relationship between span, rock mass rating, andstand-up time. Larger spans will result in lower stand-up times for a given rock mass rating. In this study,most of the data has been obtained from cut and fill stopes where the required stand-up time isapproximately three months. Therefore, a stable excavation is defined as one which has remained stablefor at least three months.An excavation’s stability is classified into three categories: stable, potentially unstable, andunstable.1436.1.2.1 Stable ExcavationsStable Excavations are characterized by the following:• There have been no uncontrolled falls of ground;• If instrumentation has been installed, there has not been any movement ofthe back which would warrant concern; and,• There were no extraordinary support measures implemented, Potentially Unstable ExcavationsPotentially unstable excavations are usually not difficult for experienced personnel, familiar withground conditions at their operation, to identif’. The openings may exhibit the following characteristics:• The opening may exhibit strong slips or faults with orientations formingpotential wedges in the back;• Extra ground support may have been installed to prevent potential falls ofground;• Instrumentation installed in the back has recorded continuing movement ofthe back; or,• There may be an increased frequency of popping and snapping indicatingworking of the ground. Unstable ExcavationsUnstable openings are simply those where an uncontrolled fall of ground has occurred. A fall ofground would usually involve failure through existing support; or, in the case of no support, the extent ofthe failure would be large enough to cause damage to pattern rock bolt support if it were installed.Uncontrolled falls of ground are distinguished from “loose incidents” which occur close to the face prior, toscaling and before rock bolts have been installed.6.1.3 Detour Lake Mine ObservationsThe raw data is provided in Table 6.1 and plotted in Figure 6.2. Figure 6.3 and Table 6.2 employthe same data except that the RMR has been reduced by 10% where unfavourable structural orientation isrecorded. At Detour Lake Mine, this correction is applied where:• faults or long, continuous, open joints are observed in the back (0-60degree dip); and,• flat jointing is observed in back (0-3 0 degree dip).It has been found that the combination of the flat structure combined with the dominant verticalEW and NS joint sets are the greatest source of instability at Detour Lake Mine. It can be seen in Table6.3 that the use of this correction results in a lower standard deviation of the corrected unstable data andvirtually unchanged standard deviations for the stable and potentially unstable data sets. The 10%144correction has been suggested by Biemawski (1984) as a correction factor applied for unfavourable jointorientation.6.2 STATISTICAL ANALYSIS6.2.1 ObjectiveThe objective of the following statistical analysis is to define three zones on the span versus RIvIRgraph; stable, unstable and potentially unstable, which can be used as a guide for the design of cut and fillstopes. Table 6.3 summarizes the main statistics for each of these three groups. The large variance in eachof the three groups results in the boundaries of the groups being ill-defined, Figure 6.4 is a plot of the datawith bivariate ellipses for each of the three groups. These ellipses illustrate the overlapping of data whichoccurs between groups. All three groups show parallel positive correlations and the group centroids areshown to lie approximately along the same line. This feature will prove useful in the statistical analysiswhich follows.Figure 6.5 is a density plot of the data for each of the three groups. The contours indicate thatmost of the data has been collected from a range of rock qualities from 40 to 85 and spans from 5 to 30metres. The choice of data collection sites was random in the sense that one or two observations weremade from each lift of each stope in the mine. It is not of interest, however what the probability is offinding a certain span at a certain rock mass rating. Rather, it is necessary to determine the probability of astope being stable, potentially unstable, or unstable given the span and RMR. For this reason, the idealsample would have a uniform density over the span-RMR domain. This was not possible however, givenactual stoping conditions. The actual density plot for all data points is given in Figure 6.6. Theconcentration of data in some areas and the lack of data in others, is an unavoidable source of error in theanalysis.6.2.2 Group ClassificationIn order to define the boundaries of the stable, potentially unstable, and unstable zones,conventional partitioned cluster analysis (Romesburg, 1990) was attempted but the technique tended tocreate compact clusters rather than linear ones, which could reasonably be expected given the paralleltrending bivariate ellipses in Figure 6.4. A form of discriminant analysis proved to be more successful indefining groups. The goal of discriminant analysis is to find the linear combination of variables thatmaximizes the variance between groups relative to the variance within groups. The technique employeduses the generalized Mahalanobis distance, D (Sneath and Sokal, 1973) to classify data points into thethree groups. The distances D1,D2, and D3 are computed for each case. The case can then be classifiedaccording to the minimum Mahalanobis distance.145There are two main criteria which must be satisfied in order for the method to be valid. The firstrequirement is for multivariate normality. This can be assessed using a normal distribution probability plotfor each of the three groups (Figure 6.7). The closer the points approximate a straight line, the morenormal the distribution. From the graphs, it can be seen that the expected values approximate a straightline. The second requirement is for homoscedasticity of the sample dispersion matrices. That is, thevariances between groups must be similar for each variable. This can be assessed using dot plots for eachvariable (Figure 6.8). The vertical length of the band of data in each column should be similar to achievehomoscedasticity. This is approximately the case for this data despite a low number of samples in thepotentially unstable range. Although the data does not show near perfect bivariate normality andhomoscedasticity, the Mahalanobis classification technique has shown considerable robustness to violationsof these requirements (Sneath, Sokal, 1973).The Mahalanobis distance classification technique is carried out as follows:E15Let x1 ... all pairs of stable data (span, RMR) i.e. x= [85r13Let y1 ...y, = all pairs of potentially stable data (span, RMR) i.e. y1= [61E 20Let z1 ...z = all pairs of unstable data (span, RMR) i.e. z1= [77then the mean sample vectors for each group are given by (Johnson and Wichern, 1988)= Inii1= Z.,n3 i=IThe sample covariance matrices can be evaluated for each group using the following relationships(Johnson and Wichern, 1988):146flIStable Group: = (x1— )(x1—T111—11=1“2Potentially Unstable Group: S2 =_____——n2Unstable Group: S3 = 1—— _)T— 11=1If the groups can be classified as multivariate normal distributions with the same variance, then apooled sample covariance matrix Spooled (Johnson and Wichem, 1988) can be formed where:(n1 —i)s +(n2 —i)s2 +(n3 —i)s3Spooled —nl +n2 +n3 —3Figure 6.9a contains three fictional groups (say stable, unstable, and potentially unstable) in a twodimensional domain. For any new point 1(I 1,12) it is possible to classify it into one of the groups using theMahalanobis Distance (Johnson and Wichem, 1988). The minimum Mahalanobis Distance defines thegroup to which the new case belongs.The Mahalanobis Distances, D3 are given by:/ _\T 1/—D1=ix— j [S] ix1—x/ _‘T 1/ —D2=y1— j [S] yi—y/ T —lr —D3 = z1 —zj [5] z1 —zThere will be points on the graph where the distance D1=D2,D2=D3,and D1=D3. These pointswill define boundaries of the stable-potentially unstable, unstable-potentially unstable, and stable-unstablezones respectively (Figure 6.9b). The equations of these linear boundaries can be derived as follows:147For any new point , , in the SPAN- RIVER domain , where £ = F SPAN[IRSetD1 =D2( -jTs’( - = ( - )T s1( -— —Ts’+Ts-1 Ts-’—yTs-’—Eliminating terms to obtain:(yT— yr)S1 +£TS1[y— ]+Ts-’—yTs-1= 0This is the equation for a straight line0where,a1/?+a2tc=0=aSPAN+a2RUR +c[: r [a] 2[S] [•i —c=[x] [S] [x]—[y] [S] {y]Similarly for D = D2 and D = D3.Using the corrected database , the variance covariance matrix has been computedr37.31 44.281 E 0.0438 —0.0143[S]=I I then [5] =1[44.28 135.7] [—0.0143 0.0120SettingD =D2 = Span =0.8858(RMR)—43.15SettingD1 =D3 = Span =0.7556(RMR)—31.35SettingD2 =D3 => Span =0.6496(RIVIR)—21.77The three lines corresponding to D1=D2,D2=D3,and D1=D3 have been plotted in Figure 6.10.The band width of the zone is wide at low RMR and converges to a point at RMR = 90. This feature is afunction of the statistical technique used and the fact that all three group centroids do not lie in a perfectlystraight line. The band width at the centre of the data is taken to be the most accurate since the highestconcentration of data is located there. If the centroids of the three groups formed a perfect line, all threelines would be parallel. For these reasons, the band width at the point where the centroid trend line (Figure1486.4) crosses the D1 = D3 line has been applied along the length of line D1 = D3 (Figure 6.11). The centreline defining points equidistant from the stable and unstable zones centroids is redundant and has beenremoved since we wish to define the boundaries of the potentially unstable zone. Finally, the graph shouldonly be applied to design of stopes which fall within the limits of the database used to create the graph.The potentially unstable zone has been shaded and cutoffs applied at the ends where insufficient data exists(Figure 6.12). This graph will be referred to as the Stability Graph for Entry-Type Stopes and isrecommended for design of spans in cut and fill stopes, shrinkage stopes, undercuts of longhole stopes andother temporary entry-type mine workings.The Mahalanobis distances have been computed for each data point and are provided in Table 6.4along with the probabilities that a point belongs to either of the three groups. The probabilities are simplya ratio of the Mahalanobis distances. A contour plot of these probabilities is given in Figure 6.13. It givesthe user of the Stability Graph for Entry-Type Excavations a measure of the confidence that a given designwill fall within each of the groups. For example, the probability of a design with a span of 20 metres andan RMR of 75, is stable is 0.6. The probability it is potentially unstable is 0.3. The probability it isunstable is LIMITATIONS OF THE EMPIRICAL DESIGN METHOD6.3.1 StructureIt is important to emphasize that the lower span design line shown in Figure 6.12 represents themaximum span for a given rock quality above which the potential for instability is high. The design spanshould always be below the shaded region. The most important factors controlling the degree to which thedesign span falls beneath the potentially unstable zone is how effectively the ground conditions are beingmonitored and whether more intensive support can be installed if required.6.3.2 State of StressThe empirical span design method presented above does not directly account for factors such as thestate of stress in the back and the influence of fill. One mine reported in the industry survey that thesefactors also influence their span design. In most rock types, changes in stress may be recognized by acorresponding change in rock mass quality. In such cases, stress conditions are being accounted forindirectly using rock mass ratings. Laubscher (1990) has suggested a rating factor which increases therock mass rating for increasing confinement since frictional resistance to sliding along joints will beincreased. This is the case for material away from the surface of the excavation however lack ofconfinement at excavation surface will allow for movement along joint surfaces. In non-bursting ground149this will be manifested as a lower rock mass rating. Therefore, proposed design curves remain valid sincethe Rock Mass Rating is a dynamic variable.The reduction of RMR due to bursting is one aspect of another CANMET sponsored researchstudy currently underway at Dickenson Mines in Ontario. Following a rockburst, the opening jointssurrounding an excavation could cause the rock mass rating to decrease by as much as 20%. Modeling ofstresses around the cut and fill stopes carried out in Chapter 5 indicated low stresses immediatelysurrounding the excavation so the influence of high stresses could not be assessed in any greater detail forthis study.6.4 COMPARISON WITH OTHER EMPIRICAL SPAN DESIGN METHODSOther empirical span design methods have been described in Chapter 3 of this report. Directcomparisons of the data and design curves are not possible due to the differences in each researcher’sdefinitions of terms such as unsupported span and stability. Other researchers such as Potvin (1988) andLaubscher have used hydraulic radius rather than span. Some initial comparisons can be made to the workof Bieniawski (1974) and Barton et al. (1974).6.4.1 Comparison to RMR DataBieniawski’s original failure case histories presented in Figure 3.15 have been superimposed on thethree stability zones established above (Figure 6.14). Note that the data does not include any stable casehistories. The lower bound of the failure band represents the point below which failure will never occur.This is roughly comparable to the lower bound of the potentially unstable zone of the cut and fill stabilitygraph. The upper bound of Bieniawski’s failure band is defined as the line above which instability willoccur immediately. Below the line are unstable cases where instability occurs after a period of time. Usingthe previous definition of stable (stable for at least 3 months), all points below 67% RMR would be classedas unstable since their stand-up times are less than 3 months. The upper bound on the band is valid forRMR greater than 73% (stand-up time greater than 3 months). A new curve based on this short-termstability requirement is given in Figure 6.15. In general, the Bieniawski curve is not a good predictor ofshort term stability in entry-type excavations. The method predicts 37% of the stable DLM cases in thestable zone compared with 77% using the Stability Graph for Entry-Type Excavations6.4.2 Comparison to NGI DataFigure 3.12 presents the relationship between the equivalent dimension, De and the NGI TunnelingIndex, Q. Assuming the recommended excavation support ratios of 3.0 and 5.0 for mining excavations,upper and lower bounds to the potentially unstable zone can be established. These bounds have been150plotted using the rock mass rating and span on linear scales for comparison purposes (Figure 6.16). TheNGI index Q was converted to RMR using Equation 3.33.Despite that only two case histories were used to determine the ESR’s of the curves in Figure 6.16,the results are remarkably close to the potentially unstable zone of the Stability Graph for Entry-TypeExcavations over the range 40-85 RMR. Although Barton’s curve is shown to extend below 40% RMRthere are too few case histories to justify it.6.4.3 Comparison to Golder Crown Pillar DataThe Golder crown pillar case histories (Carter, 1990), described in Section 3, are plotted on theStability Graph for Entry-Type Excavations in Figure 6.17. For this database, the boundaries of thepotentially unstable zone are somewhat conservative. Only one unstable point plots in the stable zone while19 stable points plot in the unstable zone. This result reflects a more conservative definition of “unstable”used for entry-type stopes in the Detour Lake Mine database.6.5 INSTRUMENTATION AND OBSERVATION6.5.1 Instrumentation to Determine InstabilityGround Movement Monitors (GMM’s) have proven to be the most effective type of instrumentationfor regular monitoring of stability in stope backs at Detour Lake Mine. The GMM consists of a slidinglinear potentiometer which is attached to a threaded-both-ends rock bolt. The rock bolt can be anchored atany depth depending on where the discontinuity is expected (Figure 6.18). At DLM they are installed at adepth of 3.8 metres since this is the limit of the drilling equipment and groundfalls greater than three metresdeep have never occurred. A plate on the GMM is glued to the collar of the hole and as the back movesrelative to the anchor, a resistance change can be measured, which is then converted to distance. The costof the GMM is approximately CDN $450.00 however they are durable and recoverable. GMM’s are usedat DLM wherever unfavourable structure is encountered or where the rock mass and span are such that thestope plots in the potential unstable zone of the Stability Graph for Non-Entry Excavations.Figure 6.19 shows the typical GMM response that could be expected in a stope which is beingprogressively enlarged, creating a wider span. For the first day, or until after the first nearby blast, slightfluctuations in the readings can be expected due to adjustment of the anchor glue setting and other factors.The readings should then stabilize with a slow steady movement rate or no movement at all. At DetourLake Mine, an area is considered potentially unstable if the rate of movement accelerates to 1 mm or morein a 24 hour period. This rate has been established only through experience. The normal monitoringinterval is once per day and this increases to once or twice per shift if the rate of movement is high. When151the movement exceeds 1 mm per day, production from the area is halted while monitoring continues. If themovement subsides, the stope will be reactivated and additional support will usually be installed.Other movement detectors such as Mine Spiders (Figure 6.20) have been used at Detour LakeMine but were not successful because blasting would set them off prematurely. These instruments consistof spring loaded reflective canisters which are attached to a rock bolt. Four arms extend from the canisterto the back and if the back moves, a fluorescent indicator will pop down indicating movement has takenplace. Another disadvantage of the Mine Spiders is that it is not possible to record the magnitude of themovement which has taken place.Multi-wire extensometers can provide the same information as a GMM however they are expensiveand non-recoverable. Readings must be taken by technicians whereas GMM readings can be taken bysupervisors and stope leaders. For these reasons, extensometers are only used in special circumstances.6.5.2 Visual Monitoring of Ground ConditionsAt DLM, design spans very closely approach the lower boundary of the potentially unstable zone.This is possible because ground conditions are closely monitored by geological personnel trained ingeotechnical mapping. Underground supervisors and crew leaders are also trained in recognizing andresponding to changes in ground conditions. An effective reporting system must be in place to ensure thatchanges in ground conditions are promptly communicated to the engineering department so design changescan be made. The more confidence a mine has in its ability to quickly recognize and adjust to changes inrock mass quality, the closer its design spans can safely approach the potentially unstable zone.6.6 CONCLUSIONSThe Stability Graph for Entry-Type Excavations derived above is a significant improvement overexisting methods of predicting stable spans in cut and fill stopes. The graph recognizes the real worlduncertainty which exists between stable and unstable excavations.Users of the graph must always be aware of the limits of the database which control itsapplicability. These are:• Span is determined by the diameter of the largest circle drawn betweenpillars and stope boundaries;• The term span refers to spans with keyblock support only;• The term stable refers to short term stability (approximately 3 months);152• The graph is considered applicable over the Geomechanics rock massrating of 40 to 85;• High horizontal stresses are not assumed to be a factor controllingstability; and• The graph applies to horizontal design surfaces.153Table 6.1 Raw Data - Detour Lake MineCase Date Stope Location RMR Q Span (m) Cond- STABILITYNo, Recorded (%) ition 5” = STABLE,”?” POTENTIALLY UNSTABLE“U” =UNSTABLE, “ “STABLE WITH SUPPORTI FEB/1990 230M4#11 85 95.2 15 S STABLE2 FEB/1990 230M3#11 87 118.8 25 S STABLE3 FEB/1990 330M4#10 77 39.1 20 ‘? FLAT OPEN JOINTS(0-20 DEG) - SMALL GROUND FALL 25T4 FEBII99O 330M3#11 77 39.1 16 S STABLE5A FEB/1990 430M3#5 78 43.7 19 S POST PILLAR OF Sm5B FEB/1990 430M3#5 78 43.7 28 P 28m INCLUDINGPOST PILLAR6 FEB/1990 430T40#5 61 6 13 U WEDGE 55 DEGREE+DYKE- COLLAPSE7 FEB/1990 430T5#5 67 12.9 9 58 FEB/1990 430T60#5 50 1.9 9 S BACK STABLE, WALL UNSTABLE9 MAR/1990 230M4#I1 85 95.2 15 SIOMARJI99O 230M3#11 87 118.8 15 5hA MAR/1990 330M4#10 73 25.1 20 5IIB MARJ199O 330M4#10 73 25.1 20 U BREAST FLAT + RELEASE12A MARJI99O 430M3#5 78 43.7 19 5 POST PILLAR OF Sm12B MARJI99O 430M3#5 78 43.7 28 P 28m INCLUDING POST PILLAR13 MAR/1990 430T60#5 50 1.9 9 S BACK STABLE,WALL UNSTABLE14 APR/1990 200m5#12 73 25.1 20 515 APRJI99O 30M4#11 85 95.2 15 516 APRJI99O 26OT7OACCESS 42 0.8 5 ShA APRII99O 330M4#hI 73 25.1 20 5 BACK17B APPJI99O 330M4#11 73 25.1 20 U BREASTFLAT+ RELEASE18 APRJ199O 300M5#12 77 39.1 18 S19 APRI1990 430T5#6 62 7.4 12 S STEEP STRUCTURE - 4m SUPER SWELLEX20 APRJI99O 430T40#6 49 1.7 8 ? 721 APRJI99O 430T60#6 52 2.4 6 S s22 MAY/1990 200M5#12 73 25.1 20 S23 MAY/1990 200M6#12 77 39.1 25 S24 MAY/1990 26OT7OACCESS 42 0.8 6 S s25A MAY/1990 300M5#12W 77 39.1 20 S25B MAY/1990 300M5#12E 77 39.1 23 U WEDGE26 MAY/1990 360T20#4 58 4.7 10 5 s27A MAY/1990 430M6#6 78 43.7 20 S POST PILLAR27B MAY/1990 430M6#6 78 433 39 P 30m INCLUDING POST PILLAR28 MAY/1990 430T40#6 49 13 12 7 Sm ADVANCE + SLASH RETREAT29 MAY/1990 430T20#6 61 6.6 10 S s30 MAY/1990 430T60#6 52 2A 6 7 STRUCTURE31A FEB/i 990 360T40#3/4 58 47 15 U WEDGE STRUCTURE - NO CABLE BOLTS31B FEB/1990 360T40#3/4 58 47 15 S WEDGE STRUCTURE + CABLE BOLTS32 FEB/1990 360T20#3 58 4.7 13 5 s33 NOV/1989 230M3#10 87 1188 22 S FLAT JOINT34N0V/1989 230M3#10 87 118.8 19 S35 NOV/1989 230M4#i0 85 95.2 14 536 NOV/1989 330M3#i0 77 39.1 20 S37 NOV/1989 330M4#1OE 73 25A 17 S FLAT JOINT38A NOV/I 989 430M3#5 78 433 18 S POST PILLAR BETWEEN PILLARS38B NOV/1989 430M3#5 78 43.7 28 S 28m INCLUDING POST PILLAR39 TAN12/1990 230M3#hI 87 118.8 15 S40 JANI2/1990 240M4#i0 85 95.2 16 741 JAN12/1990 330M4#h0 73 25A 17 7 FAULT/STRUCTURE154Table 6.1 Raw Data - Detour Lake Mine (continued)--Case Date Stope Location RMR Q Span (m) COND- STABILITYNo. Recorded (%) mo S’ = STABLE,”?” = POTENTIALLY UNSTABLE“U” =UNSTABLE, “*“5 WITH SUPPORT47 Tp.3sfl/99 330M3#l1 77 39.1 12 S43 JANI2/1990 430T5#5 67 12.9 2 S44 JANI2/1990 430t40#5 55 3.4 5 S DRILLING CAUSE UNSTABLE45 JAN12/1990 430T60#5 50 1.9 5 S46A FEBI3/1989 460M1#3 72 22.4 16 5 POST PILLAR SPANS IN MAFICS46B FEBI3/1989 460M1#3 53 2.7 10 S POST PILLAR SPANS IN TALC47 FEBI3/1989 230T40#5 48 1.6 8 548 NOV 1/1988 230T40#4 48 1.6 20 U COLLAPSE RMR FROM LIFT #549 FEBI3/1990 460T60#3 35 0.4 13 U FLAT STRUCTURE+WEDGE-COLLAPSE50 JAN19/1989 360T40#3 58 4.7 15 U WEDGE+FLAT-COLLAPSE51 DEC12/1987 360M2#2 70 18.0 22 U WEDGE+FLAT-COLLAPSE52 FEB28/1989 360T60#2 38 0.5 14 U AULT+FLAT-MONITORE-SEQUENcE CHANGED53 FEB28/1989 360T60#1 38 0.5 15 U54 OCT3/1989 460M2#4 76 35.0 22 U55 SEPTI7/1988 360M2#4 65 10.3 1256 OCT/1987 360M1#1 69 16.1 25 557 SEPT/1987 260M1#1 69 16.1 20 S58 J3JNE/1990 00M5#I3 64 9.2 16 559 JUNE/1990 200M6#12 70 18.0 20 560 JUNE/1990 300M5#12 77 39.1 23 561 YUNE/1990 360T20#4 58 4.7 10 562 TUNE/1990 430M3#5 78 43.7 20 5 STABLE WITHPOSTPILLAR62B IUNE/1990 430M3#5 78 43.7 39 * STABLE INCLUDES POSTPILLAR63 JULY/1990 200M6#13 80 54.6 12 S64 JULY/1990 300M5#13 76 35.0 17 7 WEDGE FLAT FAULT -RMR65 JULY/1990 360T40#10 48 1.6 7 S66A JULY/1990 430M3#6 74 28.0 35 5 INCLUDES POST PILLAR STRUCT 45-60 DEG66B JULY/1990 430M3#6 74 28.0 15 5 BETWEEN PILLAR STRUCTURE 45-60 DEG67 AUG1990 200M5#12 80 54.6 6 S68 AUG/1990 200M6#13 80 54.6 12 569 AUG/1990 300M5#13 76 35.0 17 7 FLAT STRUCTURE MODERATELY STABLE 4M SWELLEX70 AUG/1990 300M6#13 79 48.9 18 S71 AUG1990 360T40#10 40 0.6 7 S72A AUG/1990 430M3#7 79 48.9 25 S72B AUG/1990 430M3#7W 64 9.2 25 S73 AUG/1990 430M4#7 77 39.1 35 * INCLUDES POST PILLAR74 AUG/1990 430T5#7 67 12.9 9 575 AUG/1990 460 LONGHOLE 67 12.9 25 U76 SEPT/1990 200M5#14 80 54.6 19 577 SEPT/1990 300M6#13 79 48.9 18 S78 SEPT/1990 360T40#10 40 0.6 7 S79 SEPT/1990 430M4#7 79 48.9 25 S80 SEPT/I 990 430M3#7 77 39.1 20 581 SEP/90 430T60#7 60 5.9 5 S82 SEP/90 46OLONGHO 77 39.1 25 U *****I0DIJETOFLATSTRUCTTJRE(DLM)83 JULY/90 560T15#1 53 2.7 6 7 *******4ODTJETOFLATSTRUCTURE84 JULY/90 56OTACCESS 35 0.4 6 U HEADING STOPPED-FAULT85 OCT/90 200M5#14 80 54.6 18 586 OCT/90 200M6#14 80 54.6 20 S155Table 6.1 Raw Data- Detour Lake Mine (continued)Case Date Stope Location RMR Q Span (m) COND- STABILITYNo. Recorded (%) ON S” = STABLE,”?” = POTENTIALLY UNSTABLE“U” =UNSTABLE, “'“STABLE wITH SUPPORT87 OCT/90 300M5#14 73 25.1 25 S88 OCT/90 360T40#11] 55 3.4 7 S89 OCT/90 430M4#8 77 39.1 25 S90 OCT/90 430M3#7 77 39.1 35 S91 OCT/90 46OLONGHO 77 39.1 25 U92 NOV/90 200M6#15 79 48.9 14 S93 NOV/90 300M6#14 78 43.7 26 S94 NOV/90 360T60#11 38 0.5 5 S95 NOV/90 430M4#8 77 39.1 20 S96 NOV/90 46OLONGHO 77 39.1 25 U FLAT JOINTS97 FEB/91 200M6#16 70 18.0 11 S98 FEB/91 200M5#16 82 68.2 14 S99 FEB/91 300M5#15 78 43.7 21 U FLAT JOINTSIOQ FEB/91 360T60#13 38 C5 5 U WEDGE>45lOlA FEB/91 430M4#8 78 433 25 S POST PILLARIOIB FEB/91 430M4#8 78 437 35* SUPPORT NO POST102 FEB/91 430M3#8 78 43.7 20 ? DYKE 40-60103 MARCH/91 200M6#16 70 18.0 11 S104 MARCH/91 300M5#15 78 437 21 U FLAT JOINTS105 MARCH/91 300M6#16 76 35.0 24 5IOE MARCH/91 360T60#13 38 0.5 5* I .8m SWELLEX ON lmXlm PATERN107 MARCH/91 430M’#9 78 43.7 35* SUPPORT CABLE108 MARCH/91 430M3#8 78 43.7 20 ? DYKE 40-60109 APRIL/91 200M6#17 70 18.0 11 S110 APRIL/91 200M5#17 77 39.1 18 5 FLAT JOINTS111 APRIL/91 300M5#16 78 43.7 16 U FLAT JONTS(STABLE ONLY WITH BIRDCAGE)117 4PRIL/91 300M6#16 76 35.0 24 S BREASTFAILINGNOTBACK113 APRIL/91 60T60#14 38 0.5 5 U STABLE ONLY IF SUPPORTED114 APRIL/91 60T20#10 65 10.3 5 S115 APRIL/91 430M4#9 78 43.7 35 * CABLE SUPPORT PILLAR116 JULY/91 200M7#18 79 48.9 20 5117 JULY/91 200M8#19 77 39.1 16 5118 JULY/91 300M5#19 79 48.9 15 U FLAT JOINTS/STABLE WITH SUPPORT119 JULY/91 300M6#16 74 28.0 17 5 FLAT JOINTS120A JULY/91 360T40#17 45 1.1 7 U STABLE WITH SUPPORT ONLY120B JULY/91 360T40#17 45 1.1 7 * STABLE WITH SUPPORT ONLY121 JULY/91 430M4#1O 69 16.1 25 * STABLE WITH SUPPORT ONLY122 SEP/91 200M7#19 77 39.1 12 S123 SEP/91 200M8#19 83 76.2 15 S124 SEP/91 300M5#17 73 25.1 24 U FLAT JOINTS125 SEP/91 300M5#17 73 25.1 24 * FLAT JOINTS/STABLE WITH SUPPORT12f SEP/91 360T40#18 43 0.9 7 U127 SEP/91 360T40#18 43 0.9 7 * STABLE WITH SUPPORT12? SEP/91 60T20#14 56 3.8 5 U STABLE WITH SUPPORT129 SEP/91 360T20#14 56 3.8 5 * STABLE WITHSUPPORT130 SEP/91 430M4#10 69 16.1 25 * STABLE WITHSUPPORT131 OCT/91 200M7#19 78 43.7 18 5132 OCT/91 200M8#20 79 48.9 17 5133 OCT/91 300M5#18 75 31.3 18 U FLAT JOINTS156Table 6.1 Raw Data - Detour Lake Mine (continued)Case Date Stope Location RMR Q Span (m) CON])- STABILITYNo. Recorded (%) moN S” = STABLE,”?” = POTENTIALLY UNSTABLE“U” =UNSTABLE, **5TABLE WITH SUPPORT134 OCT/91 300M5#18 75 31.3 18 * STABLE WITH SUPPORT/FLAT JOINTS135 OCT/91 300M6#17 75 31.3 21 * STABLE WITHBIRDCAGE136 OCT/91 360T40#18 45 1.1 7 U137 OCT/91 360T40#18 45 1.1 7 * STABLE WITH SWELLEX138 OCT/91 430M3#10 80 54.6 20 S CAVED IN FEB/92139 OCT/91 430M4#1 1 81 61.0 23 * STABLE WITH BIRDCAGE140 NOV/91 200M7/M8#20 78 43.7 15 S141 NOV/91 300M5M6#18 73 25.1 24 U FLAT JOINTS142 NOV/91 300M5M6 #18 73 25.1 24 * FLAT JOINTS BIRCAGE IS STABLE143 NOV/91 400M5#13 75 31.3 26 * STRUCTURE STABLE WITH CABLE144 NOV/91 430M4#11 79 48.9 24 * CABLE PILLAR145 DEC/91 200M7/M8#20 78 43.7 15 S146 DEC/91 300M5M6#18 73 25.1 24 U FLAT JOINTS147 DEC/91 300M5M6#18 73 25.1 24 * FLAT JOINTS148 DEC/91 400M5#13 75 31.3 26 ? CAVED JAN/92 WAS MOVING149 DEC/91 560M2#2 54 3.0 10 STABLE WITH SWELLEX/TALC150 DEC/91 660M2#2 75 31.3 5 S151 JAN/92 300M5M6 #18 73 251 24 * FLAT JOINTS/STABLE WITH CABLES152 MARCH/92 300M5M6 #18 73 25.1 24 * FLAT JOINTS/STABLE WITH CABLES153 MARCH/92 560M1#2 81 61.0 9 5154 MARCH/92 560M2#2 54 3.0 10 U155 MARCH/92 560M2#2 54 10 10 * STABLE WITH SWELLEX15f MARCH/92 575SLR 70 18.0 5 S157 MARCH/92 59OSLR 72 22.4 5 5157Table 6.2 Corrected Data Detour Lake MineCase No. Date Stope Location RMR (%) Q Span (m) COND- STABILITYRecorded mo 5” = STABLE,”?” POTENTIALLY UNSTABLE“U” =UNSTABLE, *STAI8LE WITH SUPPORTI FEB/1990 230M4#l1 85 95.2 15 S STABLE2 FEB/1990 230M3#11 87 118.8 25 S STABLE3 FEB/1990 330M4#10 67 12.9 20 ? FLAT OPEN JOINTS(0-20 DEG) - SMALL GROUND FALL 25T4 FEB/1990 330M3#11 77 39.1 16 S STABLE5A FEB/1990 430M3#5 78 43.7 19 S POST PILLAR OF Sm5B FEBII99O 430M3#5 78 43.7 28 P 28mINCLUDINGPOSTPI[.LAR6 FEB/1990 430T40#5 51 2.2 13 U WEDGE 55 DEGREE+DYKE-COLLA.PSE7 FEB/1990 430T5#5 67 12.9 9 58 FEB/1990 430T60#5 50 1.9 9 S 3ACK STABLE, WALL UNSTABLE9 MAR/1990 230M4#11 85 95.2 15 S10 MAR/1990 230M3#11 87 118.8 15 5hi’ MARJI99O 330M4#10 73 25.1 20 5I1B MARJ199O 330M4#10 63 8.3 20 U BREASTFLATi-RELEASE12A MARJ199O 430M3#5 78 43.7 19 S POST PILLAR OF 5m12B MARJI99O 430M3#5 78 43.7 28 P 28mINCLUDINGPOSTPILLAR13 MARJI99O 430T60#5 50 1.9 9 5 BACK STABLE.WALL UNSTABLE14 APR/1990 200m5#12 73 25.1 20 515 APR/1990 230M4#11 85 95.2 15 516 APRJ199O 26OT7OACCESS 42 0.8 5 S17A APRJ199O 330M4#11 73 25.1 20 S BSCK17B APRI1990 330M4#1 1 63 25.1 20 U BREAST FLAT + RELEASE18 APRJI99O 300M5#12 77 39.1 18 519 APRJI99O 430T5#6 62 7.4 12 S STEEP STRUCTURE -4m SUPERSWELLEX20 APR/1990 430T40#6 49 1.7 8 ? ?21 APR/1990 430T60#6 52 2.4 6 522 MAY/1990 200M5#12 73 25.1 20 523 MAY/1990 200M6#12 77 39.1 25 524 MAY/1990 26OT7OACCESS 42 0.8 6 S25A MAY/1990 300M5#12W 77 39.1 20 S25P MAY/1990 300M5#12E 67 12.9 23 U WEDGE26 MAY/1990 360T20#4 58 4.7 10 527P MAY/1990 430M6#6 78 43.7 20 527B MAYII99O 430M6#6 78 43.7 39 P 3OmINCLUDINGPOSTPILLAR28 MAY/I 990 430T40#6 49 1.7 12 2 Sm ADVANCE + SLASH RETREAT29 MAY/1990 430T20#6 61 6.6 10 S30 MAY/1990 430T60#6 42 0.8 6 2 STRUCTURE31A FEB/1990 360T40#3/4 48 1.6 15 U WEDGE STRUCTURE - NO CABLE BOLTS31B FEB/1990 360T40#3/4 48 1.6 15 S WEDGE STRUCTURE + CABLE BOLTS32 FEB/1990 360T20#3 58 4.7 13 533 N0V11989 30M3#10 77 39.1 22 5 FLAT JOINT34 NOV/1989 230M3#10 87 118.8 19 535 NOV/1989 230M4#10 85 95.2 14 536 NOV/1989 330M3#10 77 39.1 20 537 NOV/1989 330M4#1OE 63 8.3 17 S FLAT JOINT38A NOV/1989 430M3#5 78 43.7 18 S38B NOV/1989 430M3#5 78 43.7 28 S 28mINCLUDINGPOSTPILLAR39 JANI2/90 230M3#11 87 118.8 15 S40 JANI2/90 240M4410 85 95.2 16 2 STRESS158Table 62 (cont.) Corrected Data - Detour Lake MineCase No. Date Stope Location R1WR (%) Q Span (m) COND- STABILITYRecorded mo 5” STABLE,”?” = POTENTIALLY UNSTABLE“U” =UNSTABLE, “ **‘53ft WITH SUPPORT41 JAN12/90 330M4#10 63 8.3 17 ? FAULT/STRUCTURE42 JAN12/90 330M3#11 77 39.1 12 S43 JANI2/90 430T5#5 67 12.9 2 S44 JANI2/90 430T40#5 55 3.4 5 S DRILLING CAUSE UNSTABLE45 JANI2/90 430T60#5 50 1.9 5 S46A FEBI3/89 460M1#3 72 22.4 16 S468 FEBI3/89 460M1#3 53 2.7 10 S4’ EB13/89 230T40#5 48 1.6 8 S48 NOVI/88 230T40#4 48 1.6 20 U COLLAPSEFROMLIFT#549 FEB13/90 460T60#3 25 0.1 13 U FLAT STRUCTURE+WEDGE-COLLAPSE50 JANI9/89 360T40#3 48 1.6 15 U WEDGE+FLAT-COLLAPSE51 DECI2/87 360M2#2 60 5.9 22 U WEDGE+FLAT-COLLAPSE52 FEB28/89 360T60#2 28 0.2 14 U FLAT STRUCTURE-COLLAPSE53 ‘EB28/89 360T60#1 28 0.2 15 U WEDGE+FLAT-COLLAPSE54 OCT3/89 460M2#4 66 11.5 22 U WEDGE+FLAT-COLLAPSE55 SEPI7/88 360M2#4 55 3.4 12 7 FAULT+FLAT-MONITORED56 OCT/87 360M1#1 69 16.1 25 557 SEP/87 260M1#1 69 16.1 20 S58 JUNE/90 200M5#13 64 9.2 16 559 JUNE/90 200M6#12 70 18.0 20 S60 EJNE/90 300M5#12 77 39.1 23 S61 JUNE/90 360T20#40 58 4.7 10 562 JUNE/90 430M3#5 78 43.7 20 5 STABLE WITHPOSTPILLAR628 JUNE/90 430M3#5 78 43.7 39 * STABLE INCLUDES POST PILLAR63 JULY/90 200M6#13 80 54.6 12 564 TULY/90 300M5#13 66 11.5 17 7 WEDGE FLAT FAULT RMR65 JULY/90 360T40#10 48 1.6 7 566A JULY/90 430M3#6 64 9.2 35 * INCLUDES POST PILLAR STRUCTURE 45-60 DEG668 JULY/90 430M3#6 64 9.2 15 5 BETWEEN PILLARS STRUCTURE 45-60 DEG67 AUG/90 200M5#12 80 54.6 6 568 AUG/90 200M6#13 80 54.6 12 569 AUG/90 300M5#13 66 11.5 17 ? FLAT STRUCTURE MOD.STABLE70 AUG/90 300M6#13 79 48.9 18 57’ AUG/90 360T40#10 40 0.6 7 5724 kUG/90 430M3#7 79 48.9 25 572P AUG/90 430M3#7W 64 9.2 25 S73 AUG/90 430M4#7 77 39.1 35 * INCLUDES POST PILLAR74 AUG/90 430T5#7 67 12.9 9 S75 AUG/90 46OLONGHOLE 67 12.9 25 U76 SEP/90 200M5#14 80 54.6 19 577 SEP/90 300M6#13 79 48.9 18 578 SEP/90 360T40#10 40 0.6 7 579 SEP/90 430M4#7 79 48.9 25 S80 SEP/90 430M3#7 77 39.1 20 58’ ‘EP/90 430T60#7 60 5.9 5 582 SEP/90 46OLONGHO 67 12.9 25 U *****10 DUE TO FLAT STRUCTURE(DLM)83 JULY/90 560T15#1 43 0.9 6 7 *******4() DUE TO FLAT STRUCTURE84 JULY/90 56OTACCESS 25 0.1 6 U HEADING STOPPED-FAULT85 OCT/90 200M5#14 80 54.6 18 5159Table 6.2 (cont.I Corrected Data- Detour Lake MineCase No. Date Stope Location RMR (%) Q Span (in) CON])- STABIUTYRecorded ON S” =STABLE,”” =POTENTIALLYUNSTABLE“U” tJNSTABLE, “ “STABLE WITH SUPPORT86 OCT/90 200M6#14 80 54.6 20 S87 OCT/90 300M5#14 73 25.1 25 S88 OCT/90 360T40#111 55 3.4 7 S89 OCT/90 430M4#8 77 39.1 25 S90 OCT/90 430M3#7 77 391 35 S91 OCT/90 46OLONGHO 77 391 25 U92 NOV/90 200M6#15 79 48.9 14 S93 NOV/90 300M6#14 78 43.7 26 S94 NOV/90 360T60#11 38 0.5 5 S95 NOV/90 430M4#8 77 39.1 20 S96 NOV/90 46OLONGHO 67 12.9 25 U FLAT JOINTS97 FEB/91 200M6#16 70 18.0 11 S98 FEB/91 200M5#16 82 68.2 14 S99 FEB/91 300M5#15 68 14.4-21 U FLAT JOINTS100 FEB/91 360T60#13 38 0.5 5 U WEDGE>45lOlA FEB/91 430M4#8 78 43.7 25 S POST PILLAR1O1B FEB/91 430M4#8 78 43.7 35 * SUPPORT NO POST102 FEB/91 430M3#8 68 14.4 20 ? DYKE4O-60103 MARCH/91 200M6#16 70 18.0 11 S104 MARCH/91 300M5#15 68 14.4 21 U FLAT JOINTS105 MARCH/91 300M6#16 76 35.0 24 S106 MARCHJ9I 360T60#13 38 0.5 5 * 1.8mSWELLEXON 1mX1mPATERN107 MARCH/91 430M4#9 78 43.7 35 * SUPPORT CABLE108 MARCH/91 430M3#8 68 14.4 20 ? DYKE 40-60109 APRIL/91 200M6#17 70 18.0 11 S110 APRIL/91 200M5#17 67 12.9 18 S FLAT JOINTS111 APRIL/91 300M5#16 68 14.4 16 U FLATJOINTS(STABLEONLYWITHBIRDCAGE)112 APRIL/91 300M6#16 76 35.0 24 5 BREAST FAILINGNOT BACK113 APRIL/91 360T60#14 38 0.5 5 U STABLE ONLY IF SUPPORTED114 APRIL/91 360T60#14 38 0.5 5 * STABLE ONLY IF SUPPORTED114 APRIL/91 360T20#10 65 10.3 5 S115 APRIL/91 430M4#9 78 43.7 35 * CABLE SUPPORT PILLAR116 JULY/91 200M7#18 79 48.9 20 S117 JULY/91 200M8#19 77 39.1 16 5118 JULY/91 300M5#19 69 16.1 15 U FLAT JOINTS/STABLE WITH SUPPORT119 JULY/91 300M6#16 64 9.2 17 S FLAT JOINTS120A JULY/91 360T40#17 45 1.1 7 U STABLE WITH SUPPORT ONLY120B JULY/91 360T40#17 45 1.1 7 * STABLE WITH SUPPORT ONLY121 JULY/91 430M4#l0 69 16.1 25 * STABLE WITH SUPPORT ONLY122 SEP/91 200M7#19 77 39.1 12 5123 SEP/91 200M8#19 83 76.2 15 S124 SEP/91 300M5#17 63 8.3 24 U FLAT JOINTS125 SEP/91 300M5#17 63 8.3 24 * FLAT JOINTS/STABLE WITH SUPPORT126 SEP/91 360T4O#18 43 0.9 7 U127 SEP/91 360T40#18 43 0.9 7 * STABLE WITH SUPPORT128 SEP/91 360T20#14 56 3.8 5 U STABLE WITH SUPPORT129 SEP/91 360T20#14 56 3.8 5 * STABLE WITH SUPPORT130 SEP/91 430M4#10 69 16.1 25 * STABLE WITH SUPPORT131 OCT/91 200M7#19 78 43.7 18 S160Table 6.2 (cont.) Corrected Data- Detour Lake MineCase No. Date Stope Location RMR (%) Q Span (m) COND- STABILITYRecorded mo S” = STABLE, “?“ = POTENTIALLY UNSTABLE“U” =IJNSTABLE,_*STABLE WITH SUPPORT132 OCT/91 200M8#20 79 48.9 17 S133 OCT/91 300M5#18 65 10.3 18 U FLAT JOINTS134 OCT/91 300M5#18 65 10.3 18 * STABLE WITH SUPPORT/FLAT JOINTS135 OCT/91 300M6#17 75 31.3 21 * STABLE WITH BIRECAGE136 OCT/91 360T40#18 45 1.1 7 U137 OCT/91 360T40#18 45 1.1 7 * STABLE WITH SWELLEX138 OCT/91 430M3#10 80 54.6 20 S CAVED IN FEB/92139 OCT/91 430M4#1 1 81 61.0 23 * STABLE WITH BIRDCAGE140 NOV/91 200M7/M8#20 78 43.7 15 S141 NOV/91 300M5M6#18 63 8.3 24 U FLATJOINTS142 NOV/91 300M5M6 #18 63 8.3 24 * FLAT JOINTS BIRCAGE IS STABLE143 NOV/91 400M5#13 75 31.3 26 * STRUCTURE STABLE WITH CABLE144 NOV/91 430M4#1 1 79 48.9 24 * CABLE PILLAR145 DEC/91 200M7/M8#20 78 43.7 15 S146 DEC/91 300M5M6 #18 63 8.3 24 U FLAT JOINTS147 DEC/91 300M5M6 #18 63 8.3 24 * FLAT JOINTS148 DEC/91 400M5#13 75 31.3 26 ? CAVED JAN/92 WAS MOVING149 DEC/91 560M2#2 54 3.0 10 * STABLE WITH SWELLEX/TALC150 DEC/91 660M2#2 75 31.3 5 S151 JAN/92 300M5M6 #18 63 8.3 24 * FLAT JOINTS/STABLE WITH CABLES152 MARCHJ92 300M5M6 #18 63 8.3 24 * FLAT JOINTS/STABLE WITH CABLES153 MARCHI92 560M1#2 81 61.0 9 S154 MARCHJ92 560M2#2 54 3.0 10 U155 MARCHJ92 560M2#2 54 3.0 10 * STABLE WITH SWELLEX156 MARCH/92 575SLR 70 18.0 5 S157 MARCH/92 59OSLR 72 22.4 5 5161Table 6.3(aI Statistical Summary of Raw DataStable Cases Potentially Unstable Unstable CasesCasesSPAN RMR SPAN RMR SPAN RMRNo. of Cases 98 98 13 13 32 32Minimum 2 38 6 49 5 35Maximum 35 87 26 85 25 79Range 33 49 20 36 20 44Mean 15.388 71.296 15.154 68.154 16.469 61.375Variance 45.003 144.664 36.474 165.308 48.386 259.145Std. Dev. 6.708 12.028 6.039 12.857 6.956 16.098Std. Error 0.678 1.2 15 1.675 3.566 1.23 2.846C.V. 0.436 0.169 0.399 0.189 0.422 0.262Median 15.5 77 17 75 17 68.5Table 6.3(b) Statistical Summary of Corrected DataStable Case Potentially Unstable Unstable Cases__________CasesSPAN RMR SPAN RMR SPAN RMRNo. of Cases 98 98 13 13 32 32Minimum 2 38 6 42 5 25Maximum 35 87 26 85 25 77Range 33 49 20 43 20 52Mean 15.388 70.684 15.154 61.231 16.469 54.5Variance 45.003 147.332 36.474 164.026 48.386 215.032Std. Dev. 6.708 12.138 6.039 12.807 6.956 14.664Std. Error 0.678 1.226 1.675 3.552 1.23 2.592C.V. 0.436 0.172 0.399 0.209 0.422 0.269Median 15.5 77 17 66 17 61.5162Table 6.4 Mahalanobis Distance and Group Classification Probabilities________________CASE RMR SPAN DISTANCE DISTANCE DISTANCE PROD. PROD. PROIL ORIGINAL PREDICTED MINIMUMNo (%) (m) 1 2 3 1 2 3 GROUP GROUP PROBABILIT85 15 1.464 2.423 3.271 0.856 0.133 0.012 1 1 0.8562 87 25 1.528 2.041 2.580 0.661 0.265 0.074 1 I 0.6613 67 20 1.198 0.726 0.989 0.261 0.411 0.328 2 2 0.4114 77 16 0.544 1.498 2,334 0.688 0.260 0.052 I I 0.6885A 78 19 0.622 1.357 2.104 0.619 0.299 0.082 1 I 0.6196 50 5 1.763 1.525 1.959 0.315 0.466 0.219 1 2 0.4666 51 13 1.765 0.838 0524 0.118 0394 0.488 3 3 0.4887 67 9 1.010 1.620 2,444 0.653 0.293 0.055 I I 0.6538 50 9 I 642 1.001 1.208 0.193 0.450 0.358 1 2 0.4509 85 15 1.464 2.423 3.271 0.856 0.133 0.012 1 1 0.85610 87 15 1.666 2.626 3.473 0.879 0.112 0.008 I 1 0.879hA 73 20 0.796 0.943 1536 0435 0.382 0.183 1 I 0.435I lB 63 20 1.543 0.834 0682 0169 0392 0 440 3 3 0.440I2A 78 19 0.622 1.357 2.104 0.619 0.299 0.082 1 1 0.61913 50 9 1.642 1.001 1.208 0.193 0.450 0.358 I 2 0.45014 73 20 0.796 0.943 1.536 0,435 0.382 0.183 1 I 0.43585 15 1.464 2.423 3,271 0.856 0.133 0.012 1 1 0.85616 42 5 2.271 1.680 1.724 0.139 0.447 0.414 1 2 0.44717A 73 20 0,796 0,943 1.536 0.435 0.382 0183 I 1 0.43517B 63 20 1.543 0.834 0.682 0.169 0.392 0.440 3 3 0.44018 77 18 0.503 1.320 2.103 0.625 0.297 0078 1 I 0.62519 62 12 0.698 0.660 1464 0.406 0.417 0.177 1 2 0.41720 49 8 1.716 1.134 1.355 0.199 0.455 0.346 2 2 0.45521 52 6 1.592 1.384 1.871 0335 0.458 0.207 I 2 0.45822 73 20 0.796 0.943 1.536 0.435 0.382 0.183 1 1 0.43523 77 25 1.564 1.529 1.781 0.363 0.384 0.253 1 2 0,38424 42 6 2267 1.605 1.569 0.119 0.428 0.453 1 3 0.45325A 77 20 0,712 1.241 1.921 0.556 0.331 0113 1 1 0.55625B 67 23 1,765 1.233 1.086 0.171 0379 0.450 3 3 0.45026 58 10 1.028 0.829 1.497 0.363 0,437 0.201 1 2 0.43727A 78 20 0.726 1.324 2.019 0.584 0.316 0.099 I 1 0,58428 49 12 1 878 0.976 0.670 0.108 0.390 0.502 2 3 0.50229 61 10 0.865 0.980 1.739 0.450 0,405 0,144 I I 0.45030 42 6 2.267 1.605 1.569 0.119 0.428 0,453 2 3 0.45331A 48 15 2.280 1.320 0,527 0.055 0.307 0.638 3 3 0.63832 58 13 1.079 0.329 0.934 0.260 0.440 0.300 1 2 0,44033 77 22 1.029 1.279 1.805 0.480 0.360 0,160 1 I 0.48034 87 19 1.306 2.216 3004 0.815 0.164 0.021 I I 0.81535 85 14 1.596 2.551 3.406 0.870 0.120 0.009 1 I 0.87036 77 20 0.712 1241 1.921 0.556 0.331 0.113 1 I 0.55637 63 17 1.038 0.280 0.799 0.257 0.423 0.320 1 2 0.42338A 78 18 0.567 1.416 2.203 0.652 0.281 0068 1 1 0.65239 87 15 1.666 2.626 3473 0.879 0.112 0.008 1 1 0.87940 85 16 1.347 2.305 3.143 0.839 0.146 0.015 2 1 0.83941 63 17 1.038 0.280 0.799 0.257 0.423 0.320 2 2 0.42342 77 12 1.130 2.034 2.898 0.789 0.189 0.022 1 1 0.78943 67 2 2.334 2.941 3.721 0.822 0.166 0,012 I 1 0.82244 55 5 1.579 1.637 2.246 0.457 0.416 0.128 I I 0.45746A 72 16 0.114 0.995 1.829 0.555 0.340 0,105 I 1 0.55546B 53 10 1.407 0.800 1.159 0.231 0.451 0.318 I 2 0.45147 48 8 1.796 1.170 1323 0.178 0.450 0,372 I 2 0.45048 48 20 2.973 2.060 1.209 0.020 0.195 0.785 3 3 0.78549 25 13 4.376 3.419 2.615 0.002 0.081 0.917 3 3 0,91750 48 15 2.280 1.320 0.527 0.055 0.307 0.638 3 3 0.638163Table 6.4 (cont.) Mahalanobis Distance and Grouçssiflcation_Probabilities_________CASE RMR SPAN DISTANCE DISTANCL DISTANCE PROB. PROB. PROB. ORIGINAL PREDICTED MINIMUMNe (%) (m) I 2 3 1 2 3 GROUP GROUP PROBABILITY51 60 22 2.166 1.403 0.837 0.082 0.318 0.600 3 3 0.60052 28 14 4.185 3.226 2.409 0.003 0.091 0.907 3 3 0.90753 28 15 4.303 3.343 2.512 0.002 0.081 0.917 3 3 0.91754 66 22 1.654 1.084 0.964 0.177 0.386 0.437 3 3 0.43755 55 12 1.302 0.533 0.895 0.218 0.441 0.341 2 2 0.44156 69 25 2.002 1.532 1.349 0.159 0.365 0.476 1 3 0.47657 69 20 1.042 0.753 1.165 0.315 0.409 0.276 1 2 0.40958 64 16 0.790 0.218 1.020 0.318 0.424 0.258 1 2 0.42459 70 20 0.971 0.785 1.256 0,344 0.405 0.251 I 2 0.40560 77 23 1.203 1.340 1.776 0.441 0.371 0.188 I I 0.44161 58 10 1.028 0.829 1.497 0.363 0.437 0.201 I 2 0.43762 78 20 0.726 1.324 2.019 0.584 0.316 0.099 1 1 0.58463 80 12 1.410 2.331 3.194 0.837 0.150 0.014 1 1 0.83764 66 17 0.748 0.378 1.102 0.339 0.417 0.244 2 2 0.41765 48 7 1.800 1.271 1.503 0.205 0.461 0.334 1 2 0.46166B 64 15 0.662 0.300 1.161 0.354 0.421 0.225 1 2 0.42167 80 6 2.477 3.309 4.160 0.914 0.082 0.003 1 1 0.91468 80 12 1.410 2.331 3.194 0.837 0.150 0.014 1 1 0.83769 66 17 0.748 0.378 1.102 0.339 0.417 0.244 2 2 0,41770 79 18 0,641 1.513 2.304 0.677 0.265 0.059 I I 0.67771 40 7 2.452 1.683 1.454 0.077 0.379 0.543 1 3 0.54372A 79 25 1.503 1.593 1.927 0.425 0.370 0.206 1 1 0.42572B 64 25 2.377 1.737 1.281 0.082 0,307 0.611 1 3 0.61173 77 35 3.451 3.085 2.778 0.080 0.266 0.654 1 3 0.65474 67 9 1.010 1.620 2.444 0.653 0.293 0,055 1 1 0.65375 67 25 2.146 1.598 1.299 0.124 0.345 0.532 3 3 0.53276 80 19 0.733 1.543 2.303 0.671 0.267 0.062 I 1 0,67177 79 18 0.641 1.513 2.304 0.677 0.265 0.059 1 1 0.67771 40 7 2.452 1,683 1.454 0,077 0,379 0.543 1 3 0.54379 79 25 1,503 1.593 1,927 0.425 0.370 0.206 1 1 0.42580 77 20 0.712 1.241 1.921 0.556 0.331 0.113 1 1 0.55681 60 5 1.545 1.884 2.600 0.598 0.335 0.067 1 1 0.59882 67 25 2.146 1,598 1.299 0.124 0.345 0.532 3 3 0.53283 43 6 2.188 1.555 1.573 0.134 0.439 0.427 2 2 0.43984 25 6 3.776 2.898 2.322 0.010 0.180 0.810 3 3 0.81085 80 18 0.721 1.610 2.404 0.701 0.249 0.050 I 1 0.70186 80 20 0.792 1.498 2.2 15 0.639 0.285 0.075 1 1 0.63987 73 25 1.750 1,476 1.527 0.250 0.389 0.361 1 2 0.38988 55 7 1.375 1.274 1.860 0385 0.440 0.176 I 2 0.44089 77 25 1.564 1.529 1.781 0.363 0.384 0.253 I 2 0.38491 77 25 1.564 1.529 1.781 0363 0,384 0.253 3 2 0.38492 79 14 0.996 1.946 2.803 0.781 0.193 0.025 1 I 0,78193 78 26 1.712 1.669 1.878 0.355 0,381 0.263 1 2 0.38194 38 5 2.583 1.887 1.741 0.084 0.398 0.518 1 3 0.51895 77 20 0.712 1.241 1.921 0.556 0.331 0,113 I 1 0.55696 67 25 2.146 1.598 1.299 0.124 0.345 0.532 3 3 0.53297 70 11 0.771 1.525 2378 0.666 0.281 0.053 1 1 0.66698 82 14 1.296 2.248 3.104 0.831 0.154 0.016 1 1 0.83199 68 21 1308 0.877 1.061 0.254 0.406 0.340 3 2 0.406100 38 5 2.583 1.887 1.741 0.084 0.398 0.518 3 3 0.518lOlA 78 25 1.531 1.558 1.852 0394 0.378 0.229 I I 0.394102 68 20 1.118 0,733 1,076 0.288 0.411 0.301 2 2 0.411103 70 11 0,771 1.525 2.378 0.666 0.281 0.053 I I 0.666104 68 21 1.308 0.877 1.061 0,254 0.406 0.340 3 2 0.406105 76 24 1.417 1.391 1.691 0.372 0.386 0.243 1 2 0,386108 68 20 1.118 0.733 1.076 0.288 0.411 0.301 2 2 0.411164Table 64 (cont.)_Mahalanobis Distance and Grp Classification Probabilities_________CASE RMR SPAN DISTANCE DISTANCE DISTANCE PROB. PROB. PROB, ORIGINAL PREDICTED MINIMUMNo (%) (m) 1 2 3 1 2 3 GROUP GROUP PROBABILITY109 70 11 0.771 1.525 2.378 0.666 0.281 0.053 1 1 0.666110 67 18 0.830 0.488 1.103 0.331 0.415 0.254 I 2 0.415Ill 68 16 0.394 0.596 1.424 0.435 0.394 0.171 3 1 0.435112 76 24 1.417 1.391 1.691 0.372 0.386 0.243 I 2 0.386113 38 5 2.583 1.887 1.741 0.084 0.398 0.518 3 3 0.518114 65 5 1.672 2.220 2.998 0.720 0.248 0.033 1 1 0.720116 79 20 0.753 1.410 2.116 0.612 0.301 0.017 1 1 0.612117 77 16 0.544 1.498 2.334 0.688 0.260 0.052 1 I 0.688118 69 15 0.160 0.805 1.660 0.503 0.369 0.128 3 I 0.503119 64 17 0.941 0.282 0.900 0.283 0.423 0.294 1 2 0.423120A 45 7 2.031 1.385 1.432 0.146 0.441 0.413 3 2 0.441122 77 12 1.130 2.034 2.198 0.789 0.189 0.022 I 1 0.789123 83 15 1.262 2.221 3.070 0.828 0.156 0.017 1 1 0.828124 63 24 2.272 1.601 1.130 0.086 0.315 0.599 3 3 0.599126 43 7 2.195 1.490 1.419 0.115 0.420 0.466 3 3 0.466128 56 5 1.559 1.677 2.312 0.486 0.401 0.113 3 I 0.486131 78 18 0.567 1.416 2.203 0.652 0.281 0.068 1 I 0.652132 79 17 0.667 1.597 2.415 0.706 0.246 0.048 1 1 0.706133 65 11 1.011 0.428 0.906 0.276 0.419 0.305 3 2 0.419136 45 7 2.031 1.385 1.432 0.146 0.441 0.413 3 2 0.441138 80 20 0.792 1.498 2.215 0.639 0.285 0.075 I 1 0.639140 78 15 0.757 1.715 2.565 0.738 0.226 0.037 1 1 0.738141 63 24 2.272 1.601 1.130 0.086 0.315 0.599 3 3 0.599145 78 15 0.757 1.715 2.565 0.738 0.226 0.037 1 1 0.738146 63 24 2.272 1.601 1.130 0.086 0.315 0.599 3 3 0.599148 75 26 1.837 1.627 1.697 0.269 0.387 0.345 2 2 0.387150 75 5 2.269 3.036 3.871 0.879 0.115 0.006 1 I 0.879153 81 9 2.020 2.899 3.759 0.891 0.103 0.006 1 I 0.891154 54 10 1.325 0.780 1.217 0.255 0.453 0.292 3 2 0.453156 70 5 1.928 2.611 3.425 0.813 0.172 0.015 1 I 0.813157 72 5 2.056 2.778 3.601 0.842 0.147 0.011 1 I 0.842165POST PILLARO SPAN = DIPMETER OF LARGEST CIRCLE WHICH CAN BE DRAWNBETWEEN PILLARS AND WALLS IN PLAN VIEW(a) Span Definition in Plan ViewSPAN INCLUDES PORTION OF HANGINGWALL OVERHANG NOT TIGHT FILLED(b) Span Definition in Section ViewSPAN DEFiNITION FIGURE 6.1SPAN166RAW DATA- DETOUR LAKE MINE400300CCl) 0z00) 0020 Ø©0 Occ98o0 00 0010o8 0C) 000 ocooooo00I I I0 20 40 60 80 100Rock Mass RatingL UNSTABLEEJ POTENTIALLY UNSTABLEo STABLERAW DATABASE - DETOUR LAKE MINE FIGURE 6.2167CORRECTED DATA- DETOUR LAKE MINE40 I030 -020 O9 0DD 0 0 0010 0aoo 0 00000 000I000 20 40 60 80 100Rock Mass RatingA UNSTABLE‘ POTENTIALLY UNSTABLEo STABLECORRECTED DATABASE - DETOUR LAKE MINE FIGURE 6.316840‘3°D 10 -020BIVARIATE ELLIPSES ON CORRECTED DATAUNSTABLERock Mass RatingD POTENTIALLY UNSTABLE0 STABLEBIVARIATE ELLIPSES ON CORRECTED DATA FIGURE 6.4000I I20040 60 80 100169Stable Data Density Contoiis0 20 40 60RUnstabie Data DensIty ContousDATA DENSITY CONTOURS BY GROUP FIGURE 6.5III I I0 20 40 60Potentially Unstable Data Density Contoxs60 100403004030I0403°I080 1000 20 40 60 60Pødc be100170DATA DENSITY CONTOURS - ALL DATA40 I3cCl)20t00.0.U,10I I I00 20 40 60 80 100Rock Mass RatingDATA DENSITY CONTOURS OFALL DATA FIGURE 6.617132—1—2-32I10—1-2Stable DataI.i/..Lkt.d eob v030DATA DISTRIBUTION PROBABILITY PLOTS BY GROUP FIGURE 6.717210 20 30 40L1ej 8Potentially Unstable Data10 15 20 25Lk1d 8CIUnstable Data320—1—2-30 10 20 30RMR Distribution by Group100080_______________o60 oo0__040200 IPotentIally Stable UnstableSpan Distribution by Group5040-030-00 0820-cxx7 0o +cx10- 0o_c8z00 I IPotentialy Stable UnstableDATA DISTRIBUTION DOT PLOTS BY GROUP FIGURE 6.8173FIGURE 6.9(a)Group I2(b)STATISTICAL CLASSIFICATION OF DATA iNTO THREE POPULATIONS’174ORIGINAL GROUP CLASSIFICATION LINES40 I I I00a.Co20 Q0o cia.0 ci D 00010 Q 0 00 0o 0 0 0 000 00I I I00 20 40 60 80 100Rock Mass RatingUNSTABLEC POTENTIALLY UNSTABLE0 STABLEORIGINAL GROUP CLASSIFICATION LINES FIGURE 6.10175Cl)=D POTENTIALLY UNSTABLE0 STABLEMODIFIED GROUP CLASSIFICATION L1NES FIGURE 6.11MODIFIED GROUP CLASSIFICATION LINES4030201000 20 40 60 80Rock Mass RatingL UNSTABLE10017630 -I20-Cl)G)STABILITY GRAPH FOR ENTRY-TYPE EXCAVATIONS FIGURE 6.1240STABILITY GRAPH FOR ENTRY-TYPE EXCAVATIONSI I I / // /,///,///UNSTABLE10 -0STABLE‘I / I0 20 40 60 80 100Rock Mass Rating177STABILITY PROBABILITY GRAPH40II/ /•// / //1/ d/ %‘II, o•30/ / ‘IIi// /I/i// /SO” ,, /Cl)/,/// / )20 / ‘/1 / /o I,, / // 1/, / /z A///,/ /U) /C /1 / /Z 10 / /1 / /II , / /I/I /0./ ,/ / /00 20 40 60 80 100Rock Mass RatingUNSTABLE PROBABILITY— — —- STABLE PROBABILITYPOTENTIALLY UNSTABLE PROBABIUTYSTABILITY PROBABILITY GRAPH FIGURE 6.13178I20Coa)4030 /CSIR DATA FAILURE CASE HISTORIESI I I //////UNSTABLE000 00 //00000/ /0 0/ I10 -0STABLE/—0 20 40 60 80 100Rock Mass Ratingo Tunneling Failure Case Histories (Bieniawski, 1974)0 Mining Failure Case Histories (Bieniawski, 1974)COMPARISON OF BIENIAWSKI DATA TO STABILITY GRAPH FORENTRY-TYPE EXCAVATIONSFIGURE 6.14179I4030201000COMPARISON OF STABLE-UNSTABLE BOUNDARYOF STABILITY GRAPH AND BIENIAWSKI CURVERock Mass RatingCentreline of Potentially UnstableZone on Cut and Fill Stability GraphModified Bieniawski CurveApplying 3 Month Stability CriterionFIGURE 6.1520 40 60 80 100COMPARISON OF STABLE-UNSTABLE BOUNDARY OF STABILITYGRAPH TO MODIFIED BIENIAWSKI CURVE180I4030201000COMPARISON OF NGI SPAN DESIGN CURVESTO STABILITY GRAPH FOR ENTRY-TYPE EXCAVATIONSRock Mass RatingFIGURE 6.1620 40 60 80 100COMPARISON OF NGI SPAN DESIGN CURVES TO STABILITY GRAPHFOR ENTRY-TYPE EXCAVATIONS18130 -I20-U)a)GOLDER CROWN PILLAR STUDY DATAUnstable Cases0 Stable CasesFIGURE 6.17400////////00000UNSTABLE0000000020010 -00000STABLE0 0000000/00 20 40 60 80 100Rock Mass RatingCOMPARISON OF GOLDER CROWN PILLAR DATA TO STABILITYGRAPH FOR ENTRY-TYPE EXCAVATIONS182\—GMM must be anchoredbeyond expect failure plane\-GROUND MOVEMENT MONITORROCK FACE— GLUING PLATE3 CONDUCTOR CABLEGROUND MOVEMENT MONITOR FIGURE 6.184 GMM plate glued to rock NDRILL HOLEROCKBOLT ANCHOR5/8’ THREADED BOTH ENDS ROCKBOLT183E0E00C000>C20TYPICAL GMM RESPONSE IN UNSTABLE GROUNDMonth/DayTYPICAL GMM RESPONSE IN UNSTABLE GROUND FIGURE 6.197/1 7/11 7/21 7/31 8/10 8/20 8/30184MINE SPIDER MONITORING DEVICEMINE SPIDER GROUND MONITORING DEVICE FIGURE 6.201T TEXISTING BOLT[1857. INFLUENCE OF GROUND SUPPORT ON SPAN DESIGN7.1 INTRODUCTIONSupport of cut and fill stopes is commonly provided by key-block support (rockbolts and frictionstabilizers), cable bolting, post pillars and backfill. The Stability Graph for Entry-Type Excavationsdemonstrates the increased stable span which is achieved with improved rock quality. If artificial supportis viewed as acting to reinforce the rock mass, support will have the effect of increasing the stable span.Bawden et ai.( 1989) have estimated the effect of support on rock quality by studying theimprovement in the Modified NGI-Q Rating that is achieved for varying concentrations of cable boltsupport. Figure 7.1 shows the relationship between the Q’-supported and Q’-unsupported for Bolt Factorsranging from 1 to 8. The Bolt Factor is defined as the length of cable per square metre of face supported.Another relationship which attempts to quantify the improvement in rock quality achieved with support wasdeveloped using the Rock Mass Rating System (Milne et al, 1987). Figure 7.2 shows the expected increasein RMR due to a Bolt Factor of 1 for cable bolts. This work is an initial attempt to quantify the effect ofsupport on rock quality. Although the approach appears promising, a more extensive database over a widerange of rock conditions and support concentrations is required before the validity of these relations can beestablished.7.2 KEY-BLOCK SUPPORTThe objective of key-block support is to reinforce the rock mass and make it self-supporting byholding in place key blocks at the immediate surface of the excavation. These blocks in turn providegeometric support to the surrounding rock (Figure 7.3). Key block support is normally provided by shortrock bolts, grouted dowels, or friction stabilizers. In hard, blocky ground, mechanical bolts will usuallysuffice whereas in softer, weaker ground, friction stabilizers such as Swellex or Split Sets may be used.Most key-block support design procedures rely on empirical relationships between bolt length, jointspacing and bolt spacing (Lang, 1961) and (U.S. Army Corps of Engineers, 1980). Other methods arebased on the reinforced arch concept whereby tensioned rockbolts are used to create a reinforced archwhich supports the loose rock above (Stillborg, 1986). Both methods have been used for civil engineeringexcavations but the author is not aware of any mining operations which routinely employ them. InCanadian mines, a minimum pattern is usually established from past experience and miners are ofteninstructed to decrease the spacing or increase the length of bolts to accommodate locally poor conditions.A typical pattern in Canadian mines is 1.8 m bolts on a 1.2 m square pattern. In Canada, it is becomingstandard practice to install key-block support for all excavations immediately after the face is mucked out.186Consequently, the case histories contained in the database presented in Chapter 6 all had key-block supportinstalled. Therefore, the ability of key-block support to increase the stable span cannot be assessed usingthe Detour Lake Mine database.7.3 CABLE BOLTINGCommencing approximately twenty years ago, a number of cut and fill mines developed systems topre-reinforce the rock mass prior to excavation (Fuller, 1980). The main purpose of pre-reinforcement isto control the amount of dilation into the opening immediately following excavation, thereby maintainingthe integrity of the original rock mass. Fuller, (1980) has suggested that, as with key block support, therole of cable bolts is to reinforce the rock mass rather than to directly support it. The most common type ofpre-reinforcement is a cement grouted steel cable bolt. The technique involves drilling holes of sufficientlength to pre-support at least two lifts (Figure 7.4). The holes can be angled to follow the dip and plunge ofthe ore or they can be angled to intersect the hangingwall. Cable bolts are grouted along the full length ofthe hole such that support to the back is immediately available when subsequent lifts are mined. Anycables left hanging after a blast can be cut with grinding saws or explosive cable cutting charges. In cutand fill mining, pre-reinforcement has the added benefit of supporting the breast face. Many cut and filloperations experience problems with the breast face collapsing onto the fill making drilling and blastingdifficult and slow (Ng, 1990). Pre-support of the breast face with cable bolts can reduce these difficulties,thereby improving productivity.7.3.1 Cable BoltsThe most common type of cable used for cable bolting is 16 mm diameter, seven strand cable withan ultimate tensile strength of 255 kN (25 tonnes). The cable is flexible allowing it to be installed inexcavations with low head room. The grout normally consists of Portland cement and water mixed at 0.30to 0.45 water to cement ratios by weight. Lowering the water content contributes significantly to highergrout strength (UCS) as shown in Figure 7.5a. The higher grout UCS in turn contributes to higher cablepull-out strengths (Figure 7.5b).Installation of cable bolts in up-holes involves first drilling the hole (usually 45-57 mm diameter) tothe desired length. Installations of 20 metres or more are common. A 6 mm ID plastic breather tube istaped to the top of the cable which is then pushed up the hole. An expansion shell or spring steel clips areattached to the top of the cable to prevent it from sliding down the hole before the cable is grouted. A 19mm ID PVC grout tube is inserted about 0.5 metres into the bottom of the hole which is then sealed withrags, wedges, or resin (Figure 7.6). The grout is then pumped up the hole through the grout tube. The holefills from the bottom up and air is allowed to escape through the breather tube at the top. When air stops187flowing from the breather tube, the hole is full. The ends of the grout tube and breather tube are tied off toprevent leakage and these tubes remain part of the grouted bolt system.A preferred but less common method involves inserting the grout tube to the top of the hole andpumping a thick grout (<0.35 W:C), withdrawing the grout tube as the hole fills. This may not be possiblewith some grout pumps or if the hole is too long.7.3.2 Cable Bolt Modifications and AccessoriesThe objective of the cable bolt support system is to mobilize as much of the cable strength aspossible by transferring the rock load through the grout to the cable. Therefore, the capacity of the systemis governed by three components (Figure 7.7):• Rock to grout bond;• Grout to cable bond; and• Strength of the cable.It has been demonstrated in laboratory tests and from failure case histories that failure of thesystem normally occurs at the cable-grout interface. A number of accessories have been added to cablebolts in an effort to increase the cable grout bond strength by providing a perpendicular load bearingsurface on the cable. These include ferrules or buttons hydraulically pressed on to the cable at regularintervals (Figure 7.8). The buttons are normally 25 mm to 32 mm in diameter and 37 mm to 44 mm long.Barrel and wedge type cable grips have also been used for this purpose. The spacing of the buttons on thecable should be less than the average joint spacing.Birdcage cables are a modification of the conventional steel cable. The birdcage cable bolt ismanufactured by destranding the cable at specified intervals. A common node spacing is about 20 cm.Birdcaging can be done on all or part of a cable. The destranded parts of the cable form anchors along thebolt where failure must occur by crushing and pulling through the grout. Laboratory pull tests conductedon birdcaged cables show increases in pull out strengths of between 36 and 79 percent. Figure 7.9 showsthe effect on pull-out strength of birdcaging, buttons, and using two cables per hole.The practice of post tensioning and plating of cable bolts is becoming more wide spread partly dueto the availability of lightweight, simple to use tensioning jacks. Plates are used to prevent unraveling ofthe rock around the cable at the collar of the hole. Steel plates or wooden head blocks can be used for thispurpose.1887.3.3 Cable Bolt Support DesignIn cases where structural features have delineated a potential wedge failure, the cable bolt patternmust be designed to support the expected load. The computer program UNWEDGE program (Hoek,1991) described in Chapter 3 is useful for calculating the size and weight of such a wedge knowing theorientation of the structures. Design charts such as Figure 7.5 can be used to obtain the pull-out strength.Dividing the rock load by the cable pull-out strength yields the number of cables required to support thedead load of the wedge.Where a specific hazard has not been identified, the support design should consider the bolts asacting to reinforce the rock mass. The most tried and proven technique for this type of cable bolt supportdesign is the empirical design method developed by Potvin (1988).Potvin has developed a cable bolt density chart based on 96 case histories of stopes using cablebolt support (Figure 7.10). On the x-axis of the chart is plotted the average block volume represented bythe RQD/Jn divided by the surface hydraulic radius. On the y-axis is plotted the cable bolt densityexpressed in bolts per square metre. As expected, an increase in the block volume or a decrease in thehydraulic radius will decrease the cable bolt density. The graph is divided into four zones for purposes ofcable bolt support design. The shaded area represents conditions where block size is so small that cablebolting would not be effective or where the cable bolt density is insufficient. The zone delineated betweenlines 1 and 2 is the least conservative design zone and is suitable for non-entry mining methods. The zonebetween lines 2 and 3 is considered to be a conservative design zone for non-entry mining methods as wellas being suitable for entry type mining methods such as cut and fill. The zone to the right of line 3 isconsidered to be the most conservative. Excavations requiring long term support such as civil engineeringprojects should have support which plots in this zone.7.4 POST PILLARSA variation of conventional overhand cut and fill for wider orebodies is post pillar cut and filldeveloped at Falconbridge Ltd.’s Strathcona Mine in the early 1970’s (Cleland et a!., 1973) A typical postpillar stope layout is given in Figure 7.11. Progressive mining of several lifts creates a pattern of tallpillars with height to width ratios exceeding 2:1. Since lifts are filled before the next one is started above,only the tops of the pillars are visible. Cases histories reported in the literature indicate this mining methodhas been used to depths of 600 metres and no limit to the overall stope span has been encountered.Post pillars are designed to gradually yield below the level of the fill. Support is provided by thepost-yield strength of the pillar. Figure 7.12 shows vertical stress measurements from a post pillar at KingIsland Scheelite Mine in Australia. The graph shows the stress to be decreasing as pillar height increases.189Figure 7.13 shows the increase in vertical movement within a pillar as the pillar height increases. Thiscombination of increasing strain with decreased stress is indicative of post-yield behaviour.Post pillars are employed at Detour Lake Mine as a means of reducing the unsupported span inmechanized cut and fill stopes. Normally, 5 metre square post pillars are left with the maximum spanbetween pillars or walls determined from the Stability Graph for Entry-Type Excavations (normally about20 metres). Current practice is to begin growing post pillars on the footwall side of the stope (Figure 7.14).As the stope shifts south and east with each lift due to the plunge and dip of the orebody, the vertical pillarmigrates from footwall to hangingwall.The post pillars do not provide significant benefit when they are located close to the sides of thestope since the walls are also providing support. They would provide maximum benefit when located at thecentre of the exposed span yet by the time they reach the centre of the span their width to height ratio is 4:1to 5:1. At this height:width ratio, the pillar has minimal load bearing capacity. Hedley, (1975) hassuggested that the effect of a post pillar is to provide support to the immediate back. With theseconsiderations in mind, a cable bolt support trial was initiated at DLM to effectively eliminate the need forpost pillars.7.5 CABLE BOLT PILLARCable bolting has a number of advantages over post pillars in cut and fill stopes:• It is possible to calculate and monitor the strength of the cable boltsupport whereas the strength of the post pillar is much harder to predict.• The cables can be installed in the centre of the stope and angled to followthe plunge and dip of the stope thereby remaining in the centre of the stopeproviding maximum benefit;• Greater ore extraction ratios can be achieved by mining the pillars; and,• Greater mining efficiency achieved by mining full face and not having tomine around a pillar.In order to compare the effectiveness of post pillars and cable bolting, a post pillar at Detour LakeMine was replaced with an artificial cable bolt pillar in the back. Cable bolts were installed in the backadjacent to the pillar. Instrumentation was installed to measure the displacement of the back as well as theload taken up by the cables. Finally the post pillar was removed and the loads and displacements weremonitored. This cable bolt support trial and the results of monitoring will be discussed below.1907.5.1 Location of TestPost pillars are only employed in the 460 Stope of DLM where the orebody is up to 45 metreswide. It was determined that Pillar 941 on the 8th lift of the 430 M4 stope (Figure 7.15) would be the bestlocation for the test for the following reasons:• The pillar had a width:height ratio of 0.2:1 and was approaching thecentre of the stope;• There was an access 20 metres above the stope necessary for locatingmonitoring instrumentation;• The cable bolting to be carried out adjacent to the pillar would be in aposition which would not interfere with regular stoping operations; and• Mine scheduling permitted time for cable bolt installation and pillarremoval before the stope had to be filled.7.5.2 Geometry of Post Pillar 941Pillar 941 was initiated on the 4th lift of the 430 Stope and was 25 metres in height prior toremoval. The cross sectional shape and area varies from lift to lift (Figure 7.16). The inconsistency in thesize and shape of the pillar on each lift is due in part to the angle that the pillar was approached by theadvancing breast face. The pillars are not trimmed to the design size after mining past them which alsocontributes to their larger than design size.7.5.3 Estimate of Pillar StrengthAn estimate of the post pillar strength was made in Chapter 5 using the modified Hock-Brownfailure criterion. The failure constants m and a have been estimated to be 3.4 and 0.45 respectively. Theconfining stress and therefore the pillar strength at the mid-height of the pillar was shown in Section 5 to benegligible using 3-D boundary element modeling. The above is based upon a rock mass rating of 80 for theundisturbed mafics (Table 7.2).7.5.4 Estimate of State of Stress in PillarThe stress on the post pillar can be initially estimated using tributary theory (Hock et al., 1980)where:(rock column areaa=yzI I (7.1)pillar area jwhere,= the unit weight of the rock191z = the depth of rock above pillarThe rock column area refers to the area supported by the post pillar which is assumed to be halfthe distance to the adjacent pillar. The minimum pillar area is 47.5 square metres which is the crosssectional area of the pillar on Lift #8. The rock column area is estimated from Figure 7.15 to be 420square metres. Therefore, the average vertical pillar stress is estimated to be:430m*0.O29MPaI m*(420147.5)= 1 1OMPa (7.2)Tributary theory severely overestimates the stress level in the pillar as compared to the results of 3-D boundary element modeling as shown in Chapter 5. This modeling has shown the average majorprincipal stress at the mid-height of the pillar is approximately 20 MPa, sufficient to cause yielding of thepillar. The above evaluation corresponds well to visual observations whereby the rock mass rating wasreduced from 80 to 58 at the fmal stages of extraction. It was concluded that the pillar had yielded but wasmaintaining support to the immediate back through its post-yield strength. The post-yield strength developsas the pillar dilates and compresses the confining fill.7.5.5 Description of the Rock MassVisual inspection of Pillar 941 identified several vertical joints on the east side of the pillar whichwere open 2-5 centimetres. Several blocks were close to slabbing off the wall of the pillar. The pillar islocated in an area between the Talc Zone and the Main Zone. A CSIR rock mass rating conducted on thenorth wall of the pillar yielded an RIVIR of 58 (Table 7.1). Rock to the north the pillar was typical MainZone rock which had a measured RIVIR of 80 (Table 7.2). Rock to the south of the pillar was typical Talcore which had a measured RMR of 65 (Table 7.3).Figure 7.18 is a lower hemisphere equal angle stereonet plot of the structural data collected aroundthe pillar during mining of lifts 4 to 8. The stereonet shows one major joint set and one minor one with thefollowing orientations:Joint Set A: Mean Orientation: Strike 297°, Dip 86°Joint Set B: Mean Orientation: Strike 160°, Dip 90°7.5.6 Cable Bolt Support Implementation7.5.6.1 Support DesignTo determine the number of cable bolts which would replace the post pillar, the maximum loadbeing supported by the pillar was assumed to be a wedge 20 m square at the base and 3 metres high. Threemetres is the maximum height of ground falls experienced at DLM to date. Using a specific gravity of 3.0192for DLM ore, the weight of this wedge would be 1800 tonnes. Thirty-five double, 5/8” diameter sevenstrand steel cable bolts were installed in a 1.5 m X 1.5 m square pattern adjacent to the pillar as shown inFigure 7.15. Each double cable bolt was assumed to have a strength of 54 tonnes. The bolts were installedin a concentrated pattern in the centre of the stope for three reasons:• The cable bolts were being used to simulate the effect of the post pillar inreducing the unsupported span;• To minimize the interference with normal mining activities; and• To allow for monitoring from the 400 MS Attack drift located above thestope. Cable Bolt InstallationA 2 metre high pad was built up from waste rock below the area to be cable bolted. This wasnecessary for the longhole drill to reach the back. A Boart BCI-2 pneumatic longhole drill mounted on arubber tired carrier was used to drill 52 mm diameter holes. The holes were all vertical and 19 metres long.Hole deviation was not measured but experience with drilling up-holes elsewhere in the mine with this rigindicated it to be 3 percent. Figure 7.19 is a section through Ring 1 showing the expected height of the drillholes and the future lift elevations. The bolts are intended to pre-support lifts 10 and 11.The cable bolts were inserted up the holes using a cable bolt inserter mounted on a scissor liftvehicle. The end of each cable bolt had a hydraulically “pressed-on” end holding device. Two 7.5 cm longspring steel clips were fastened to the end holding device to prevent it from slipping down the hole duringinsertion. A 6 mm diameter (I.D,) breather tube was taped to the top of the bolt prior to insertion. Afterpushing the bolts up the holes, a 19 mm diameter (I.D.) grout tube was inserted roughly 0.5 m into the hole.The hole was then sealed with a combination of cloth, wedges, and resin grout.Grouting was carried out with a Spedel Series 6000 grout pump and mixer. Type 30 high earlystrength cement was used in a 0.45 water cement ratio. A lower water cement ratio was desirable but notpractical with the Spedel pump. The end of the breather tube was placed in a bucket of water and pumpingcontinued until the air stopped coming out of the tube. Wooden squeeze blocks were attached to the cablebolts at the collar to contain any spalling around the hole collars and to contain small blocks with less thanthe required embedment length, in order to mobilize the full cable bolt strength (Figure 7.20(a)). Thecement had 12 days to cure before the adjacent pillar was removed.7.5.7 Monitoring InstrumentationFour types of instrumentation were used to monitor the extraction of the post pillar. These includetwo multi-point extensometers, a ground movement monitor, 1 cable bolt mounted with three Tensmegstrain gauges, and one vibrating wire stress meter.193The two Wireflex extensometers were installed in holes drilled from the 400 M5 Attack drift(Figure 7.15). This was necessary in order to permit monitoring to continue until lift #10. Eachextensometer had three anchors which were spaced to be 2 metres above the back elevations of the 8th, 9th,and 10th lifts (Figure 7.19). The extensometers were connected to an eight channel RST LE8200 DataLogger, which can take readings at programmable intervals and store the data for later retrieval.A ground movement monitor (GMM) was installed in the back of the 430 M4 #8 Stope near thecable bolts as shown in Figure 7.15. The 0MM was anchored at 3.6 metres into the back. There were noprominent joints in the vicinity which the 0MM was intended to cross. It was deep enough however torecord movement of a 3 metre high wedge which was the maximum height expected and for which the cablebolt pattern was designed. 0MM readings were taken manually.One cable bolt hole was drilled from the 400 M5 Attack. A single cable bolt was installed whichhad three Tensmeg strain gauges mounted on it to determine if the cables were taking any load after thepillar was removed. Like the extensometers, the strain gauges were mounted on the cable at intervals suchthat they would be 2 metres above the back of the 8th, 9th, and 10th lifts.A Geokon vibrating wire stress meter was installed in a hole on the south wall of the stope whichhad not been undercut. The hole was drilled 3 metres deep with a percussion drill. The stress meter wasoriented to measure the vertical stress change which would be expected to occur when the pillar wasremoved if the pillar was transmitting vertical stress. The stress meter was read manually before and afterpillar removal.7.5.8 Results of Monitoring7.5.8.1 Visual MonitoringAs a safety precaution, mining of the 8th lift was completed before the pillar was removed soworkers did not have to enter the area. Immediately prior to removing the pillar, a visual inspection of theback was made for subsequent comparison. The back was in good condition and had rock mass rating of80, typical of the Main Zone ore. The jointing was observed to be tight in the back. Rock bolt plates werenot showing any signs of taking load. The wooden squeeze blocks on the ends of the cable bolts were alsonot showing any squeezing.The pillar was drilled off on January 30, 1991. There were no problems with ‘jammed steel” orother signs of a highly jointed rock reported by the driller. The holes were clean except for the first 0.6metres where there were obvious signs of open joints. The pillar is shown in Figure 7.20(b) prior to beingblasted. The pillar was blasted on January 31, 1991 at 4:08 a.m.. A im x 3m pillar remained standingafter the blast but it was highly fractured and held up by the muckpile. It was recovered without additional194blasting. The open span was increased to 33 metres with the removal of the post pillar. Another visualinspection was made after the blast and no changes were observed. InstrumentationThe locations of the monitoring instmments are shown in Figure 7.15 and Figure 7.19. Figure 7.22refers to the GMM in the centre of the span showing no significant movement (under 1 mm) before andafter the pillar blast. Similar recordings were made on the 9th and 10th lifts indicating the backs of the liftsdid not move This is verified visually as the area was classified as “stable”.Figure 7.23 shows the stress change versus time plot for the stress meter located in the footwall ofthe 8th lift. No significant load change was observed with the removal of the pillar. A pillar that was aload bearing element would be expected to show some load transfer to the adjacent pillar. This reinforcesthe original assumption that the post pillar was a minimal support member.The Tensmeg monitoring data is shown in Figures 7.24. Figure 7.25 shows the same data over thefirst 24 hours after the blast. Strains on Anchor 1 of the Tensmeg located in the back of the 8th lift,indicated that load transfer onto the cables had occurred shortly after the removal of the post pillar.Loading on Anchor 2 above the 9th lift was minimal. The Tensmeg was calibrated to record maximumstrains of 10,000 to -2,500 microstrain. 8000 microstrain corresponds to a load on the cable of 25 tons(Figure 7.26). Loading on Anchor 1 climbed to over 25 tons within 24 hours of the pillar blast which issufficient load to cause failure.Visual observations and ground movement monitor readings did not support the Tensmegobservations. When the 9th lift was mined through in March, the Tensmeg in the back did not show thesame type of load increase, In fact the load decreased during this period. These results cannot beadequately explained unless damage had occurred to the instrumentation.Movement within the extensometers is erratic as shown in Figure 7.27. These gauges are affectedby the blast vibration and at this stage the displacement data is not considered to be reliable. Thepotentiometers are also believed to have been in contact with water which contributed to the poor quality ofthe data. The ground movement monitoring data is considered to be more reliable.7.5.9 Summary of Cable Bolt Pillar Support TrialCable bolts were used to pre-reinforce 2 lifts. Upon extraction of the 9th and 10th lifts of the 430M4 Stope, no movement of the ground movement monitors was observed. This was verified by visualobservations as the back was characterized as being “stable”. Extraction of the post pillar on the 8th liftdid not result in deteriorating conditions, however, loading on the cables did increase. Difficulties were195experienced with the instrumentation which prevented load magnitudes from being determined. Thedisplacement which caused the load increase was not measured by the GMM’s so it is expected that themagnitudes were small. It is concluded that the original post pillar was only providing support to theimmediate back.In Main Zone rock at Detour Lake Mine the maximum span which can be opened up beforeinstability begins to occur is about 20 metres based on the Stability Graph for Entry-Type Excavations.Post pillars are used to reduce the span where the width of the ore exceeds the design span. Observationsand analysis does indicate that the replacement of the post pillar with cables enabled spans on subsequentlifts to be increased up to 35 metres. The experience and confidence gained by this test has led to furthercable bolting of stopes in this manner at DLM wherever post pillars are not practical.It has been demonstrated that both cable bolts and post pillars can be used to increase the stablespan. The supported span case histories obtained at Detour Lake Mine have been plotted on the StabilityGraph for Entry-Type Excavations shown in Figure 7.28. It is difficult to assess the full potential benefitof the various support systems since there are no unstable supported case histories. A larger database ofsupported span case histories is required before such a graph could be used for design purposes however,this graph can be used to get an initial sense for the size of spans which can be designed for a given level ofsupport.196Table 7.1 941 Post Pillar Geomechanics Rock Mass RatingCategory Details Rating (%)Strength 170 MPa (R4) 14RQD 65% 15Joint Spacing 50-300 mm 10Joint Condition open joints 9Groundwater dry 10TOTAL 58Table 7.2 Main Zone Geomechanics Rock Mass RatingCategory Details Rating (%)Strength 170 MPa (R4) 14RQD 90% 18Joint Spacing 0.3-lm 20Joint Condition joints tight, slightly rough 18Groundwater dry 10TOTAL 80Table 7.3 Talc Zone Geomechanics Rock Mass RatingCategory Details Rating (%)Strength 40 MPa (R2) 5RQD 90% 18Joint Spacing 0.2-imnim 16Joint Condition joints tight, soft wall rock 16Groundwater dry 10TOTAL 651970)0)cuCa)1.a)aDt-D c) (‘4II U II II U II NILLLLLILLL. IL ILa2 a000(c11Hoddns) ID0BOLT FACTOR DESIGN CHART (After Bawden et al., 1989) FIGURE 7.1U0IF—uJ0FC)Uj0co00LUF0000C’)z0‘‘\\\ \ \ I‘D-0Ia)--------IL0U-- ---IHzEzLi.:::001981004.’0C.0.Cd)80604020040 100FIGURE 7.2cable boltingIncreased RMR due toRMR’’ 50- .5RMR(for RMR > 40)60 80Unsupported RMREFFECT OF SUPPORT ON THE RMR CLASSIFICATION VALUE199SUPPORT OF KEY BLOCKS PROVIDESGEOMETRIC SUPPORT TO BLOCKS ABOVEKEY BLOCK SUPPORT USING ROCKBOLTS FIGURE 7.3200I Cable bolts installed- to Pre-reinforce 2 lifts2__6_____—— 5 7 - Lift 3 filled4- Lift 4 Excavated32IFr - Hanging cables cut- More cables installed alongfootwallPRE-SUPPORT OF CUT AND FILL STOPES USING GROUTED FIGURE 7.4CABLE BOLTS201(a)100—I II I:. I0.25 0.35 0.45 0.55 0.65 0.75WATER:CEMENT RATIO(after Hyett et al., 1992)(b)250I I2001: z:zz;rz0 200 400 600 800Cable Embedment Length (mm)WATER:CEMENT RATIO VERSUS GROUT STRENGTH AND PULL FIGURE 7.5OUT STRENGTH202End Holding Device6 mm ID breather tubeElectrical Tapecement groutelectrical tape19 mm PVC grout tube (75 psi)FIGURE 7.657 mm diameter holeburlap plugTYPICAL CABLE BOLT GROUT1NG ARRANGEMENT FOR UP-HOLES203til, lit litft I I Itlii f ItStrength Components Affecting Pull Out Strengthof a Cable Bolt1. Rock Grout Bond Strength2. Cable Grout Bond Strength3. Tensile Strength of CableSTRENGTH COMPONENTS FOR A GROUTED CABLE BOLT FIGURE 7.7SYSTEM204C(1 C 00 Tj C-) rn 00.5 m Embedment500combination (slipped and snapped)400 -twin strand (slipp£apped)0(U0-J300- rdgedsn200 -10000 20 40 60Displacement (mm)I m Embedmentuuu.500 combination (snapped)twin strand (slipped)400.£300-I200-,_1/e.(snaPPe)........._____inIestrand (slipped)100-0 I I0 10 20 30 40 50 60Displacement (mm)(after Villeasusa, 1992)PULL OUT STRENGTHS FOR VARIOUS CABLE BOLT FIGURE 7.9CONFIGURATIONS AI%il) EMBEDMENT LENGTHS (after Villeasusa,1992)2060.Ci)zUi00DESIGN CHART FOR CABLE BOLT DENSITY0.400.350.300. 1 2 3 4 5 6 7 8(RQDIJ) I HYDRAULIC RADIUS(after Potvin, 1992)DESIGN CHART FOR CABLE BOLT DENSITY (after Potvin, 1992) FIGURE 7.10207drillingventilationpost pillarsdecline accessTYPICAL CUT AND FILL STOPE LAYOUT USiNG POST PILLARS FIGURE 7.11(after Barrett et al., 1981)rock boltingmuckingfilling208-JU-JC‘C-J‘pCU1C=LL - -(edIIlI) e6uetj sse;gMEASUREMENT OF VERTICAL STRESS IN A POST PILLAR FIGURE 7.12VERSUS PILLAR HEIGHT (after Barrett et al., 1981)C00-(0C,zzg— U,•0)C,(0/I/(0U0 0 0 01.209EUiUi>FIGURE 7.13VERTICAL MOVEMENT iN A POST PILLAR VERSUS PILLARHEIGHT (after Barrett et al., 1981)2104 20mSchematic Cross Section of Post PillarsFIGURE 7.145mPOST PILLAR5m1SCHEMATIC SECTION OF A POST PThLAR CUT AND HLL STOPEAT DETOUR LAKE MINE211LOCATION OF CABLE BOLT SUPPORT TRIAL0 10 20mT-60 ZoneFIGURE 7.15460 #3Attack APost Pillar— ExtensometerX GMMStress Meter, Cable BoltPillarT-20 Zone460 ml Stope - 8th Lift21220 000 N20 000 NLift 8LUCCCD0)0 5 1GmScalePOST PILLAR 941 PLAN PROFILES FIGURE 7.16LU00CD0)Lift 4r ft6Lift 5(N<ft72133.0Cl)C)Cl)I0. Samples offine grainedigneous crystallinerock.m=17, s=1Very goodqualityrock massm=8.5, s=0.1Good quality rock massm=1.7, s=0.004Fair quality rock massm=0.34, 8=0.0001Poor quality rock massm=0.09, s=0.000011 2 3 4Pillar Width I Pillar Height - WpIhPILLAR STRENGTH VERSUS WIDTH:HEIGHT RATIO FORDIFFERENT ROCK MASSES (after Hoek and Brown, 1980)FIGURE 7.17214_Jw wwZC.Q (4DC i O- Cw.‘ z —UI L)-oxUJ vC______________• ca Z4 C.)I r I Jii U 3 -l..4!:ifrLOWER HEMISPHERE EQUAL AREA STEREONET PROJECTION FIGURE 7.18OF STRUCTURE MAPPED AROUND PILLAR 941215DataloggerSCHEMATIC SECTION THROUGH BOLTING RING #1INDICATING LOCATION OF INSTRUMENTATIONFIGURE 7.19Cable BoltsLift 11Lift 10GroundMovementMonitorLift 9216t4-I:;L(a) Plated Double Steel Cable BoltsPHOTOGRAPHS OF CABLE BOLT A1’D POST PILLAR 941 PRIOR FIGURE 7.20TO PILLAR BLAST(b) Post Pillar 941 (Looking South)217TRIPLE ANCHOR WIREFLEX EXTENSOMETER MONITORINGHEADS (after Carter, 1990)218FIGURE 7.21GROUND MOVEME1Tf MONFIOR430 Stope 8th LIft1.00.5E0 —— 4— Pillar Blast-0.5-10 I I I I I I I I I I I I I I I I I I I1114 1118 1122 1126 1130 213MonthlDayGROUND MOVEMENT MONITOR (GMM) MEASUREMENT OF FIGURE 7.22BACK DISPLACEMENT219Footwall Stress Monitoring430 M4 Stope, 8th Lift4.03.002.0C.) 1.00.0-1.0 ——2.0 I I I I I I I I I I I1112 1114 1116 1118 1120 1122 1I24 1I26 1I28 1130 211 213 215 217Month DayFOOTWALL STRESS CHANGE MONITORING FIGURE 7.23220Cable Bolt LoadLong Term Monitoring300__________________________________________________________________200 N! 10: - - - - - - -Upper Limit of Measurement-100 —:+— Pillar BlastIi I I I I31101191 02)10191 02120/91 03102191 03/12/91 03/22)91 04/01191— Anchor I - --. Anchor 2Post Pillar 941 ExtractIonTensmeg Monitoring DataCABLE BOLT LOAD MONITORING USiNG TENSMEG STRAIN FIGURE 7.24GAUGES - LONG TERM MONITORiNG221z.0-JTensmeg Monitoring DataCable Bolt LoadShort Term MonitoringFIGURE 7.25250200150100500-50-100-1500200 04:00 06:00 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00Anchor I --- Anchor2Post Pillar 941 ExtractionCABLE BOLT LOAD MONITORING USING TENSMEG STRAINGAUGES - FIRST 24 HOURS222CEzI—Cl)0C.,TENSMEG STRAIN GAUGE LOAD CALIBRATION CURVE FIGURE 7.26TENSMEG CALIBRATION CURVE9,0008,0007,0006,0005,0004,0003,0002,0001,0000 50 100 150 200 250TENSION IN CABLE BOLT (kN)CALiBRATION CURVE FOR ThE READINGS ON A V1S)4AYSTRAININDICATOR OF A TENSMEG-70 GAUGE MOUNTED ON A B CABLEBOLT(GAUGE FACTOR - 2.0)22315Extensometer 2-5-10II‘. ‘“IF/1A A0131!91 02/10/91 02120191 03/02/91 03112191 03/22/91 04101191Anchor I — Anchor 2 Anchor3Post Pillar 941 ExtractionExtensometer Monitoring DataEXTENSOMETER MONITORING OF BACK DISPLACEMENT FIGURE 7.271050CL..a,a,E0U,CLU224STABLE SUPPORTED SPAN CASE HISTORIES40 —30 -2010 -00 20 40Support Methodo Cable Boltso Post PillarRock Mass Rating* Super SwellexFIGURE 7.280p iI’ // // // /K / //UNSTABLE///I! / I0STABLE/60 80 100EFFECT OF CABLE BOLT AND POST PILLAR SUPPORT ON SPAN2258. CONCLUSIONS AND RECOMMENDATIONS8.1 CONCLUSIONSBased on the results of a survey of cut and fill operations in Canada, it has been shown that there isnot a consistent or well established method of designing spans for underground entry-type excavations.Most operators are relying on past experience with ground conditions at their operations as a guide todesigning future stopes; however, this experience is not being systematically documented. Existingmethods of span design, including beam theory, Voussoir block theory, structural failure analysis,empirical design methods, and numerical modeling methods have been reviewed in Chapter 3. Beam theorycannot be applied because it assumes an unjointed rock mass, which does not normally occur inunderground metal mines. Voussoir block theory assumes that the back contains regularly spaced verticaljoints which would be uncommon in underground metal mines but may be applicable in stratified deposits.General purpose empirical design methods have been proposed; however they have been developed fromlargely civil engineering case histories. Other empirical span design methods such as the Modified StabilityGraph Method, have been developed from open-stope case histories and should not be used for the designof entry-type excavations.Any design procedure for entry-type stopes must attempt to reconcile two conflicting goals. First,a high enough safety factors is required that recognizes the higher risks which accompany mining entry-type stopes. Secondly, the design must be balanced with the requirement for a relatively low factor ofsafety as compared to civil engineering excavations, recognizing the short term nature of the stopes and thecosts associated with support. In this study, a database was established to develop an empirical spandesign technique specifically for entry-type excavations. 172 case histories were collected in whichstability, Rock Mass Rating, span, and major structure was recorded. The Detour Lake Mine, a largeunderground gold mine in northern Ontario, where the observations were collected, is well suited as a datacollection site because there is a good range of stope spans and rock qualities from which to collect thedata.An empirical span design chart, called the Stability Graph for Entry-Type Excavations wasdeveloped by plotting the span against Rock Mass Rating for the observed case histories. A statisticalanalysis of the data was carried out to define stable, potentially unstable, and unstable groups. A form ofdiscriminant analysis which employs the generalized Mahalanobis Distance was used to create boundariesbetween the three groups.The Stability Graph for Entry-Type Excavations is an easy to use method for designing spans inentry-type stopes but it is one which must be used with a reasonable degree of engineering judgment. To226use the stability graph, the engineer must first establish the expected Rock Mass Rating for the stope beingdesigned. Ideally, stope design should be carried out on a lift-by-lift basis so the RMR is measured onprevious lifts of a stope. Alternatively, the RMR can be measured in nearby workings or estimated fromgeotechnical core logging. It is recommended that the design span should fall below the lower boundary ofthe potentially unstable zone. Sound engineering judgment must be applied to determine the degree towhich the design span approaches the lower boundary of the potentially unstable zone. At Detour LakeMine, it is possible to design spans close to this lower bound because geologists, front-line supervisor’s,and stope leaders are given special training in recognizing and responding to hazardous ground conditions.Where a potentially hazardous condition arises, instrumentation is installed to detect and monitorinstability. An effective reporting system is also important so that changes in rock quality can be quicklyconmiunicated to the engineering department and design changes can be made. The more confidence amine has in its ability to quickly recognize and respond to changes in rock mass quality, the closer thedesign span can approach the potentially unstable zone boundary.The Stability Graph for Entry-Type Excavations that has been proposed is a significantimprovement over existing methods of predicting stable spans in cut and fill stopes. The potentiallyunstable zone on the graph recognizes the real-world uncertainty which exists between stable and unstableexcavations. It is important to emphasize the limitations of the database which control its applicability.These are:• The term span refers to spans with key block support only;• The term stable refers to short term stability (approximately 3 months);• The graph is considered applicable over the RMR range 40 to 85;• High horizontal stresses are not assumed to be a factor controllingstability; and• The graph applies to horizontal design surfaces.Three-dimensional boundary element modeling and vibrating wire stress meters were used to assessthe state of stress in the back of the cut and fill stopes. The modeling indicated the confining stress in thepillar to be near zero in the immediate back. Average a 1 stresses were well below the unconfinedcompressive strength of the rock, indicating that high stresses were not a contributing factor to cases ofinstability recorded at Detour Lake Mine. Rather, the lack of confining stress in the immediate back resultsin key-block failure, wedge failure, or rock mass failure.The role of support in increasing the allowable span before instability occurs has been brieflyexamined in this study. Yielding post pillars have been successful at Detour Lake Mine and elsewhere forincreasing the span which can be mined using cut and fill methods. A trial support program was carriedout at Detour Lake Mine to determine whether a concentrated pattern of cable bolts installed at the centreof the span would provide the same benefit as a post pillar. Based on the results of instrumentation227readings and visual observations, the cable bolts were successful in maintaining stability in the stope afterthe post pillar was removed and the span increased to 35 metres. Previous case histories without supportindicated that instability occurs when the span exceeds approximately 22 metres in the rock where the trialwas located. Two additional cable bolt pillars have since been installed at Detour Lake Mine, and bothhave been successful in increasing the stable span beyond 30 metres.It is suggested that the Stability Graph for Entry-Type Excavations be used as part of an integrateddesign approach which also uses analytical and numerical modeling techniques as described in Section 3.3and illustrated on the flow chart in Figure 3.30. After obtaining intact rock strength parameters fromlaboratory testing, and rock mass characteristics from geotechnical mapping, the design engineer shoulddetennine whether there are any major geological structures which will control stability. If so, a wedgestability analysis should be done as described in Section 3.1.3, and if necessary, the span should bedesigned to limit the formation of a wedge. Alternatively, a means of supporting the wedge can bespecified. If there is foliation parallel to the dip of the deposit, a quick assessment of chimney failurepotential should be made using Equation 3.29. Numerical modeling would be required at this stage todetermine the horizontal stress in the sill pillar for Equation 3.29, and it would also indicate if pillarcrushing was becoming a potential failure mechanism. After structural failure along recognized structureshas been accounted for in the design, the most likely mode of failure is rock mass failure which can beassessed using the Stability Graph for Entry-Type Excavations as described above.The design of a stope should be an on-going activity. During and after excavation of the stope, therock mass rating should be monitored on an on-going basis so the design can be adjusted accordingly.Visual monitoring, supplemented by instrumentation readings, should be used to assess stability ofindividual stopes on a regular basis.The above span design procedure has been incorporated into the mine design approach at DetourLake Mine, enabling the mine to maximize extraction of the orebody and improve profitability, whilemaintaining safe working conditions. In 1992, this approach contributed to the Detour Lake Mine beingawarded the “Award of Excellence” by the Mines Accident Prevention Association of Ontario as the safestmine in Ontario. During this time, the mine was also successful in reducing the cost of gold productionfrom $382/ounce to $340/ounce (US).8.2 RECOMMENDATIONSThe ideal database for the empirical study described above would contain a uniform density ofobservations on the Stability Graph for Entry-Type Excavations; however, this was not possible under228actual mining conditions. The concentration of data in some areas and the lack of it in others is a source oferror in the statistical analysis presented in Chapter 6. Additional unstable and potentially unstable casehistories would be desirable particularly at higher rock mass ratings. Potential sources for this data are theprovincial mines inspection authorities who may keep records of major groundfall incidents.The effect of post pillar support and cable bolt support has been examined briefly in this thesis. Ithas been shown how cable bolts have been used to increase the stable span at DLM from 22 to 35 metres.Additional case histories will be required before empirical guidelines can be established for confidentlypredicting the span which can be achieved for a given support. The author envisages that the StabilityGraph for Entry-Type Excavations can eventually be combined with supported entry-type excavation casehistories which would contain several design bands reflecting various support systems and intensities ofsupport. This would be a worthwhile and interesting subject for future researchers.229BIBLIOGRAPHYAmes, D.W., Bulk Mining and Worker Safety - A Preliminary Study, Ontario Ministry of Labour,Occupational Health and Safety Division, 1987.Arjang, B. and Herget, G., Stress Determinations at Detour Lake Mine, Northern Ontario, CANMET Report#MRP/MRL 85-(TR), 1985.Barrett, J.R. and Chester, G., Post Pillar Cut and Fill Mmin . Comparison of Theory and Practice, Proceedingsof the Conference on Applications of Rock Mechanics to Cut and Fill Mining, Lulea, Sweden, 1981.Barton, N., Lien, R. and Lunde, J., Analysis of Rock Mass Quality and Support Practice in Tunneling and aGuide for Estimating Support Requirements, Internal Report - Norwegian Geotechnical Institute, Oslo,1974.Barton, N. and Choubey, V., The Shear Strength of Rock Joints in Theory and Practice Rock Mechanics, Vol.10. 1977.Bawden, W.F., Sauriol, G., Milne, D., and Germain. P. Practical Rock Engineering Stope Design CaseHistories from Noranda Minerals Inc. CIM Bulletin 82 July, 1989.Beer, G. and Meek, J.L. Design Curves for Roofs and Hangingwalls Based on Voussoir Beam and PlateSolutions Trans. Instn. of Mining and Metallurgy, Vol. 91 London, 1992.Bernstein, I. H., Applied Multivariate Analysis, Springer-Verlag, New York, 1987.Biemawski, Z.T., Engineering Classification of Jointed Rock Masses, Transactions, South African Institute ofCivil Engineers, Vol. 15, No. 12, 1973.Biemawski, Z.T. Geomechamcs Classification of Rock Masses and its Application in Tunneling, Proc. 3rd.Intl. Congress on Rock Mechanics, International Society of Rock Mechanics, Denver, Colorado, Vol. hA,1974.Biemawski, Z.T., Rock Mass Classifications in Rock Engineering, Proceedings Symposium on Exploration forRock Engineering, Ed. Z.T. Biemawski, A.A. Balkema, Rotterdam, 1976.Biemawski, Z.T., Engineering Rock Mass Classifications, J. Wiley & Sons New York, 1989.Biemawski, Z.T., Rock Mechanics Design in Mining and Tunneling, A.A. Balkema, Rotterdam, 1984.Brady, B.H.G. and Brown, E.T., Rock Mechanics for Underground Mining, George, Allen, & Unwin, London,1985.Brawner, C .0. and Haugen, M., Pre-Reinforcement to Improve Underground Stability, Society of MiningEngineers, 112th AGM Littleton, Colorado, 1983.Brown, E.T. Rock Characterization. Testing, and Monitoring - ISRM Suggested Methods, Pergamon Press,1981.230Canadian Mining Journal, 1990 Mining Sourcebook 95th Edition, Southam Business Publications, Don Mills,Ontario, 1990.CANMET- Strength Determination of Detour Lake Mine Rock Division Report MRP/MRL 85-45(INT).,March, 1985.CANMET - BEAP User Manual Mining Research Laboratories and Gemcom Services Inc. MRL 93-070(TR)August, 1993.CANMET - VIEWBEAP - An Advanced 3-D Graphical Representation Program for BEAP3D,CANMET Project No. 23440-9-9 139, March 1991.CANMET- MINE DESIGNER - An Automated Graphical Data Generation Program IntegratingNumerical Modeling With the Mine Design Process, MRL 91-31 (TR), July, 1991.Carter, T.G., Wong, J.S., and Yuen, C,M,K,, Crown Pillar Stability Back-Analysis, CANMET Department ofEnergy, Mines, and Resources, Canada, DSS File No. 098Q23440-8-9074, 1990.Cleland, R. S. and Singh, K.H., Develonment of Post Pillar Mining at Falconbridge Nickel Mines Ltd., CIMBulletin, April, 1973.Clifford, R.L., Long Rockbolt Support at New Broken Hill Consolidated Ltd. Proceedings of the AustralianInstitute of Mining and Metallurgy Vol 251, 1974.Craig, R.F., Soil Mechanics, Chapman and Hall, London, 1987.Curran, J.H. and Corkum, B.T. Examine 2D Users Manual Rock Engineering Group,University of Toronto,1990.Diering, J.A.C., Numerical Modeling - Beap User Manual, CANMET Project No: 4-9147-4, 1990.Evans, W.H. The Strength of Undermined Strata. Transactions of the Institute of Mining and Metallurgy Vol.L, 1940-41 London, 1941.Fuller, P .G. Pre-reinforcement of Cut and Fill Stopes, Proceedings of the Conference on Applications of RockMechanics to Cut and Fill Mining, Lulea, Sweden, 1980.Griffel, W., Handbook of Formulas for Stress and Strain, F. Ungar Publishing Company, New York, 1966.Hassanni, F.P., Mitri, H.S., Khan, U.H., and Rajaie, H., Experimental and Numerical Studies of the Cable BoltSupport Systems Proc. International Symposium on Rock Support, Sudbury, Canada, 1992.Hedley, D.G.F. and Wilson, J.C. Rock Mechanics Applications in Canadian Underground Mines CIMBulletin, November, 1975.Hedley, D.G.F., and Grant, F., Stope and Pillar Design for the Elliot Lake Uranium Mines, CIMM Bulletin,Vol. 65, 1972.Herget,G., and Arjang,B., Update on Ground Stresses in the Canadian Shield, Proceedings - Stresses inUnderground Structures, CANMET, Ottawa, 1990.231Hocking, G.A. A Method for Distinguishing Between Single and Double Plane Sliding of Tetrahedral Wedgeshitni. Journal Rock Mechanics and Mining Sciences Vol. 13, 1976.Hoek, E., Kaiser. P.K., and Bawden, W.F. Design of Support for Underground Hardrock Mines, MiningResearch Directorate, Sudbury, Ont.,1994.Hoek, E., Wood, D. and Shah, S. A Modified Hoek-Brown Criterion for Jointed Rock Masses. Eurock’92London, 1992.Hoek, E. Unwedge Users Manual University of Toronto, 1991.Hoek, E., A Limit Equilibrium Analysis of Surface Crown Pillar Stability Proc. Intnl. Conf Surface CrownPillar Evaluation for Active and Abandoned Metal Mines, Timmins, Ont. 1989.Hoek, E., and Brown, E.T., The Hoek-Brown Failure Criterion - A 1988 Update, 15th Canadian RockMechanics Symposium, Toronto, Canada, 1988.Hoek, E., and Brown, E.T., Underground Excavations in Rock, Institution of Mining and Metallurgy, London,1980.Hyett, A.J., Bawden, W.F., and Coulen, A.L. Physical and Mechanical Properties of Normal Portland CementPertaining to Fully Grouted Cable Bolts. Proc. International Symposium on Rock Support, Sudbury,Canada, 1992.Johnson, R.A. and Wichern, D.W. Applied Multivariate Statistical Analysis, 2nd. Ed., Englewood Cliffs, N.J.,Prentice-Hall Inc. 1988.Knutsson, Sven, Stresses in the Hydraulic Backfill from Analytical Calculations and In-situ MeasurementsInstitute of Mining and Metallurgy, London, 1981.Lang, B., Pakalnis, R., and Vongpaisal, S., Cable Bolt Replacement of a Post Pillar at Detour Lake MineProc. International Symposium on Rock Support, Sudbury, Canada, 1992.Lang, T.A., Theory and Practice of Rockbolting, Tran. American Institute of Mining Engineering, Vol. 220,1961.Laubscher, D.H., A Geomechamcs Classification System for the Rating of Rock Mass in Mine Design, Journalof the South African Institute of Mining and Metallurgy. October, 1990.Laubscher, D.H. Selection of Mass Underground Mining Methods, Design and Operation of Caving andSublevel Stoping Mines, Society of Mining Engineers, AIME, 1981.Mathews, K.E. Hoek, E., Wylie, D.C. and Stewart, S., Prediction of Stable Excavation Spans for Mining atDepths Below 1000 m in Hard Rock, CANMET Report, DSS Serial No. OSQ8O-OO81DSS, File No.17SQ 23440-0-9020, 1980.Miller, P. Geology of the Detour Lake Mine Internal Report for Detour Lake Mine, Placer Dome Inc.Timmins, Ont. 1988.Milne, D., Unpublished Noranda Training Manual, Noranda Research, Pointe Claire, Quebec, 1987.232Mitchell, R.J., Sand Backfill Testing and Analysis for Cut and Fill Mining, Report Prepared for Detour LakeMine, John D. Smith Engineering Associates Ltd., March, 1987.Ng, Larry, Falconbridge Ltd., Personal Conversation, July, 1990.Obert, L.A., The Philosophy of Desinn U.S. Bureau of Mines Information Circular, IC 8585 1973.Obert, L.A. and Duvall, W.I., Rock Mechanics and the Design of Structures in Rock, John Wiley & Sons, NewYork, 1967.Pakalnis, R., Three Dimensional Modeling - An Applied Approach, 2nd Canadian Conference on ComputerApplications in the Mineral Industry, Vancouver, B.C. 1991.Pakalnis, R. Miller, H., Vongpaisal, S., and Madill, T. An Empirical Approach to Open Stone Design ISRMProceedings, Montreal, 1987.Potvin, Y., Empirical Open Stope Design in Canada - PhD Thesis, University of British Columbia, Vancouver,B.C. 1988,Potvin, Y., and Mime, D., Empirical Cable Bolt Design, Proceedings of the International Symposium on RockSupport, Sudbury, Canada, 1992.Poulos, H.G., and Davis, E.H., Elastic Solutions for Soil and Rock Mechanics, Wiley, New York, 1974.Romesburg, H.C., Cluster Analysis for Researchers Robert E. Krieger Publishing, Malabar Florida, 1990.Singh, K.H. and Hedley, D.G.F. Review of Fill Mining in Canada, Proc. Conference on Application of RockMechanics to Cut and Fill Mining, Institute of Mining and Metallurgy, London, 1981.Smith, J.D. Physical Property Tests - Basalt Samples from Shaft Pilot Hole Report from J.D. Smith andAssociates to Detour Lake Mine, October, 1984.Smith, J.D., Mitchell, R.J., Design and Control of Large Hydraulic Backfill Pours, CIM Bulletin, February,1982.Sneath, P., Sokal, R., Numerical Taxonomy: The Principles and Practice of Numerical Classification SanFrancisco, 1973.Stephansson and Jones, J., Application of Rock Mechanics to Cut and Fill Mining, Institute of Mining andMetallurgy, London, 1981.Stevenson, T., Improving Ground Stability and Mine Rescue - The Report of the Provincial Inquiry intoGround Control and Emergency Preparedness, Ontario Queen’s Printer, 1986.Stillborg, B. Professional Users Handbook for Rock Bolting Trans-Tech Publications, Clausthal-Zellerfeld,1986.Terraprobe - Laboratory Determination of Elastic Moduli of Intact Rock Cores, Report to Detour Lake MinesLtd. File 84137, July 1984.233Unal, E.,, Design Guidelines and Roof Control Standards for Coal Mine Roofs, PhD Thesis, Pennsylvania StateUniversity, 1983.U.S. Army Corps of Engineers, Engineering and Design: Rock Reinforcement. Engineer Manual EM 1110-1-2907, Office of the Chief Engineer, Washington, 1980.Villaesusa, E., Sandy, M.P., and Bywater, S., Ground Support Investigations and Practices at Mount Isa. Proc.International Symposium on Rock Support, Sudbury, Canada, 1992.Voest-Alpine Ltd. Assessment of Talc Zone Rock Strength, Report Prepared for Detour Lake Mines Ltd.,1991.Windsor, C.R,, Cable Bolting for Underground and Surface Excavations. Proc. International Symposium onRock Support, Sudbury, Canada, 1992.Wiles, T. MAP3D Mining Analysis Program in 3D Manual. Mine Modeling Ltd., Sudbury, Canada, 1991.234APPENDIX ASURVEY OF CANADIAN CUT AND FILL MINES235QUESTIONNAIREPLACER DOME INC.- PAKALMS AND ASSOCIATES- CANMETNAME OF OPERATION:CONTACT/POSiTION:PRODUCTION RATE FROMUNDERGROUND:DEPTH OF MINING:DIP OF OREBODY:TYPE OF ORE:PALCONBRJDGE LTD. (Overview ofC&F operations)DOUG HANSOA SENIOR GEOM’CIL4NfCS ENGilVEERFrom 2000 to 4000feet20’lo 60’Massive Sulphide and stringer ore1.Mining Method Percent of Total Production Backfill (YINIOverhand Cut and Fill 20 YPost Pillar Cut and Fill 50 yVCR Bench 15 yVCR Crater 15 y2. How do you determine the spans (FW-HW) within your cut and fill system?Currently evaluating back spans utilizingModfiedMathews stability graph method and recommending support spacing andlength. Will be incorporating the 3DEC-GC package whichfunctionsfrom the rock mass Q database.3. Have there been instances where full extraction of the ore was not possible by CIF because the spanwould be too wide? If so, what was the span? The rock mass rating? Percentage of ore left behind?Early 1980e StrathconaMine had groundfalls with dimensions of m x m between 5 mx 5 m postpillarx Q’ 20-404-11% ofore left behind4. Do you use post pillars? If so, how were they designed?Original concept based on 1973 design by K Singh which essentially uses D. Ffedleysformulas. Currently a postpillar projectis underway involving a new empirical approach combined with instrumentation and FL4C UDEC, and 3DEC sensitivityF’,’ ““‘“‘““Y’5. Do you use cable bolts? If so, what is the pattern and type? How was the pattern designed?Currently some mines use the MRA cable bolt design manual but we are working with Noranda to improve on the design.General patterns are 1.5m x 1.5m single strand 6.0 m long (length vanes). Again, 3DEC-GC program now being introducedwill analyze necessityfor support and reanalysis will determine fsupport is adequate.6. What is your spacing between sill pillars? How thick are your sill pillars? How were they designed?Spacing: 60-70 m. Thickness: 15 mTruthfully, a certain number ofhorizons are identified to achieve a certain production rate. However this results in highstressed burst prone sills as the mine matures.7. Have you experienced any cases of pillar failure?Horizontal sill pillars commonly burst at thicknesses 15 in. Vertical rib pillars experience stressfractures and verticalslabbing and hourglassing morefrom the seconda,y panel extraction side.8. Have you experienced any cases of back failure?Some backfailure in overcuts(gravitationa key block, low stress) where prelimina,y rock supportplaced Generally nofailure where cable bolt supportplaced9. What kind of monitoring instruments do you use in the back?Ground movement monitors (GMMs) standard and variable lengths. Sometimes use Geokon stress meters to examine stressarchine across back.In the pillars?Horizontal GI4U’x DISTOFOR (TelemaxExtensometer) Geokon Stress meters Rock Sm’s10. Have you experienced bursting of ground?Yesii. Would you describe failures to date as being structural or stress controlled?Bursting activity usually occurs first followed by debonding ofkey block wedges and gravitational failurn.236ROCK MECHANICS DATABASEIndicate with a V which of the following parameters have been estimated at your mine.Rock Unit Weight, y VStrength Elastic Modulus, VParameters Poisson’s Ratio, v VIn-Situ Measurement VStress Photo-Elastic ModuliInvestigations Computer Modeling VCompressive Strength, aLaboratory Tensile Strength, atTesting Triaxial Strength, a,i atShear Strength, tFailure Criterion VRQD VRock NGI Rating VMass CSIR Rating VClassification Laubscher MRMR RatingStructural Mapping VMulti-wire Extensometer VBoroscope ObservationMonitoring Compression PadClosure Station VLeveling Survey StationPiezometer237u-iAVArV0I:,1H iii0a911HpV0A0I1U aa a‘c,aaaaC.,,:zz*LIiC.Crrc‘.‘LLL‘ aaa——————‘—.‘.%‘aa½½33333333ppaaaaaaaa———————UpQUESTIONNAIRENAME OF OPERATION:C0NTACT/Posmow:PRODUCTION RATE FROMUNDERGROUNDDEPTH OF MINING:DIP OF OREBODY:TYPE OF ORE:PLACER DOME INC. - PAKALNIS AND ASSOCIATES- CANMETIMining Method Percent of Total Production Backfill (Y/N)Narrow Vein 2.5Panel Cut & Fill 28Longhole 26Other 20.5Open Pit 232. How do you determine the spans (FW-HW) within your cut and fill system?Experience / Orebody Type3. Have there been instances where full extraction of the ore was not possible by CIF because the spanwould be too wide? If so, what was the span? The rock mass rating? Percentage of ore left behind?Yes.. 30feet, Q=8-Ii Recovery8o%, Panel Stopes4. Do you use post pillars? If so, how were they designed?Yes Study by McGill University (July 1988) gives design criteria - not generally used to date.5. Do you use cable bolts? If so, what is the pattern and type? How was the pattern designed?Some cases - Normally 4’ to 8’ pattern20’ to 40’ long, designfrom experience and analysis ofstrength/stress expectations6. What is your spacing between sill pillars? How thick are your sill pillars? How were they designed?130feet between sill pillars. 20feet thick.7. Have you experienced any cases of pillar failure?Yes, 5 tons to 500.000 tons. Largefailures occur over long time periods in cave stopes.8. Have you experienced any cases of back failure?Yes, massive in cave stopes.9. What kind of monitoring instruments do you use in the back?GroundMovement Monitors, stress meters, slough metersIn the pillars?Stress meters, extensometers10. Have you experienced bursting of ground?Yes rareDOME NE& SELDON, ROCKMECHANICS ENGiNEER3300 Tons milled per dayFrom Surface toVaries 0-90’Gold associated with different host rocks and structuralfeatures11. Would you describe failures to date as being structural or stress controlled?ct,.,,, fM1,, 1 f’,mfr,,II,,d ,,, trn’I f.,ilrn’,, .,,,d ,,,,‘i,, Rjñ,.,.,tW,,r rfr,,,h,,’.,I f,,,,h,r,,239ROCK MECHANICS DATABASEIndicate with a V which of the following parameters have been estimated at your mine.Rock Unit Weight, y VStrength Elastic Modulus, VParameters Poisson’s Ratio, v VIn-Situ Measurement VStress Photo-Elastic ModuliInvestigations Computer Modeling VCompressive Strength, a VLaboratory Tensile Strength, a1- VTesting Triaxial Strength, Gd G1- VShear Strength, tFailure Criterion VRQD VRock NGI Rating VMass CSIR Rating VClassification Laubscher MRMR RatingStructural Mapping VMulti-wire Extensometer VBoroscope ObservationMonitoring Compression Pad VClosure Station VLeveling Survey StationPiezometer240r)ILI—a.rrrcC)rcci1b.-.•‘‘;‘-‘1.y3*-,.I1I1A————cJIQUESTIONNAIREPLACER DOME INC. - PAKALNIS AND ASSOCIATES - CANMETNAME OF OPERATION:CONTACT/POSITION:PRODUCTION RATE FROMUNDERGROUND:DEPTH OF MINING:DIP OF OREBODY:TYPE OF ORE:TROUT L4KEIvBNEJ. ROM4IVOWSKI, MINE ENGINEER2500 Tons per dayFrom 50 m to 400 m potentially to 800 m50’-70’Solid to disseminated suiphides within the altered zone1.Mining Method Percent of Total Production Backfill (Y/N)Cut and Fill 80-100 YLonghole Open Stoping 0-20 N2. How do you determine the spans (FW-HW) within your cut and fill system’It is determined by mineralization -full extraction3. Have there been instances where hill extraction of the ore was not possible by C/F because the spanwould be too wide? If so, what was the span? The rock mass rating? Percentage of ore left behind?It has not havened so far. you use post pillars? If so, how were they designed?No.Do you use cable bolts? If so, what is the pattern and type? How was the pattern designed?Yes. 7 strand 270K steel co.bles installed in vertical holes drilled into the back (65feet) and inclined (60 ‘-80 ) holes drilledinto the MW (30’-559. Basic pattern: 1. 8m x 1.8m, length and dip ofinclinedHWcables depends on local conditions Basicpattern was proposed by GolderAssoc. and was based on case histories rather than strictly designed It is sometimes modifIedWhat is your spacing between sill pillars? How thick are your sill pillars? How were they designed?Spacing: 60-65 metresThickness: approximately 10 metresThey were designed to contain the load ofbackfill over the maximum stope sparkHave you experienced any cases of pillar failure?NoHave you experienced any cases of back failure?Yes, Failure started always close to HWcontact andfreguently was controlled by natural jointing. It occurred withinfracture4desfressed zone in the back below the ‘pressure arch’What kind of monitoring instruments do you use in the back?NoneIn the pillars?NoneHave you experienced bursting of ground?Not YetWould you describe failures to date as being structural or stress controlled?BotA Most ofthem start in spots where we can expect stress concentration - so they are stress related Structure ofthe backand particularly MW controls extent and volume offaiture.242ROCK MECHANICS DATABASEIndicate with a V which of the following parameters have been estimated at your mine.Rock Unit Weight, y VStrength Elastic Modulus, € VParameters Poissonts Ratio, v VIn-Situ MeasurementStress Photo-Elastic ModuliInvestigations Computer Modeling VCompressive Strength, a VLaboratory Tensile Strength, a VTesting Triaxial Strength, a/ atShear Strength, tFailure CriterionRQDRock NGI Rating VMass CSIR RatingClassification Laubscher MRMR Rating VStructural Mapping VMulti-wire ExtensometerBoroscope ObservationMonitoring Compression PadClosure StationLeveling Survey StationPiezometer243RIAAVI0C00-AIcihJ0II-i-00-:i‘IQ.•.‘I?jIu‘:ts!0.€i.-0ic—————rl•.‘•rQUESTIONNAIREPLACER DOME INC.- PAKALNIS AND ASSOCIATES - CANMETNAME OF OPERATION:CONTACT/POSITION:PRODUCTION RATE FROM UNDERGROUNDDEPTH OF MINING: From Surface to 7200feetDIP OF OREBODY:TYPE OF ORE:Mining Method Percent of Total Production BacidlU (YIN)RAd 463 (1989) YBlasthole 21.0 (1989) NSLC 29.0 (1989) NCut & Fill 3.7 (1989) 1’2. How do you determine the spans (FW-H’i) within your cut and fill system?For true ore widths less than 35feet - minefidI width longitudinaL For widths>35feet mine with tran,werse stopes and ribsor with post pillars. Stoping spans rangefrom25 to 40feet depending on rock quality, mining depth, fill quality and stressstate.3. Have there been instances where full extraction of the ore was not possible by C/F because the spanwould be too wide? If so, what was the span? The rock mass rating? Percentage of ore left behind?No. The effective ore widths have never exceeded the critical span whichfor our stress state and yielding pillar approachto cut andfill mining is believed to be in excess of600feet at depths in excess of600feet4. Do you use post pillars? If so, how were they designed?Yes, Empirically. Yielding pillarsyield and therefore have safetyfactors less than 1.0. The sizing ofpillars and stope spansis afunction ofthe rock quality the horizontal stress state, and thefill quality. Overly stiffpillars can lead to high backsstresses. Overly softpillars can permit tensile conditions to develop in the stope bachs Their stiffness also influences thetiming and the nature ofthefailure ofthe sill pillar in wide orebodies.5. Do you use cable bolts? If so, what lathe pattern and type? How was the pattern designed?Yes, for specific situations. Forfall ofground or wedge situations, cable strength xNo. ofcables> weight and cable spacingnot to exceed point where the grout bond strength to cables exceeds 500 psi. (Grout w/c ratio <0.43) For pre-pinning, spacinggoverned by 500 psi bond strength 7ftx 7ftfor single cables. I Oftx loftfor double cables6. What is your spacing between sill pillars? How thick are your sill pillars? How were they designed?200foot spacing. Nil thickness, they are mined out They were mined throughfailure between 30feet and 40feet thickThey were the recovered by VC&F7. Have you experienced any cases of pillar failure?Yes, All ribs, post and sill pillarsfaiL Preferably by yield Rib and postpillars in the 30ft to SOft mining height region.Sill pillars anywherefrom 11 Ofeet to 40feet thick8. Have you expeiienced any cases of back failure?Who has not? There have been localized structurally bound groundfalls and material displaced by rockbursts. There have beenno massive collapses such as would occur ifthe critical span was exceeded9. What ldnd of monitoring instruments do you use in the back?No routine instrumentation other than visual obsereations. The adequacy ofthe design becomes apparent once the rib or postpillars yield through obsen’ation ofthe back conditions. There is normally little that can be done if the pillars are too softIn the pillars?for the stope spans and the back goes tensile10. Have you experienced bursting of ground?Yes11. Would you describe failures to date as being structural or stress controlled?RothINCO LIMITED - ONTARIO OPERATIONSPH. OLIVER/SENIOR SPECIALIST - ROCKMECHAI’llCS11,450,000 IONS- 1989Flat to vertical - Normal case isfor dips >55Massive to Disseminated sulphidesComment: There seems to be too little emvhasis vlaced on the role ofhorizontal stress state as it influences stable spans.245ROCK MECHANICS DATABASEIndicate with a V which of the following parameters have been estimated at your mine.*Not necessarily routinelyfor all applicationsRock Unit Weight, y VStrength Elastic Modulus, VParameters Poisson’s Ratio, v VIn-Situ Measurement VStress Photo-Elastic ModuliInvestigations Computer Modeling VCompressive Strength, crLaboratory Tensile Strength, a VTesting Triaxial Strength, aj VShear Strength, tFailure CriterionRQD VRock NGI Rating VMass CSIR Rating VClassification Laubscher MRMR RatingStructural Mapping VMulti-wire Extensometer VBoroscope ObservationMonitoring Compression PadClosure Station VLeveling Survey StationPiezometer246_%,-‘ItcL‘i.i?r),.I———ziI‘;‘‘UiI1I—————C)I0wIQUESTIONNAIREPLACER DOME INC. - PAKALNIS AND ASSOCIATES - CANMETNAME OF OPERATION:CONTACT/POSITION:PRODUCTION RATE FROMUNDERGROUND:DEPTH OF MINING:DIP OF OREBODY:TYPE OF ORE:WESThIINMINEMICHAEL CULLEN GEOTECHNICAL ENGINEER3650 TONS PER DAYFROM400 m To 600 m60’ to SubhorizontalMassive SulphidesI.Mining Method Percent of Total Production Backfill (Y/N)Mechanized Cut and Fill 75 yLonghole 25 y2. How do you determine the spans (FW-HW) within your cut and fill system?Determined by structural stability i.e. beneath thefaulted hangingwall instability occurs at approximately 6-8 metres,In massive sz4flde instability occurs at approximately 10 -15m.3. Have there been instances where full extraction of the ore was not possible by C/P because the spanwould be too wide? If so, what was the span? The rock mass rating? Percentage of ore left behind?Stopes greater than 10 m span (most) mined by post pillar. In the past 10 m stopes running strike length beneaththe faulted hangingwall utilized a two pass system 86% extraction in post pillar stopes. 100% extraction in double passQ<lforHWfaultQ>lofor massive sufides4. Do you use poet pillars? If so, how were they designed?Yes, Tributary areal Medley pillar strength/Exposed Span5. Do you use cable bolts? If so, what is the pattern and type? How was the pattern designed?Cable Time: 0.6” diameter. 7 strand 2 cablesPattern: 2mx2 m in waste, 2mxl.5m in ore- Design: Gravity Loa4 Experience, and practice at other mines6. What is your spacing between sill pillars? How thick are your sill pillars? How were they designed?N/A7. Have you experienced any cases of pillar failure?N/A8. Have you experienced any cases of back failure?Backfailures common in areas offaulting and geological contacts.9. What kind of monitoring instruments do you use in the back?Multi-point extensometersIn the pillars?Vibrating wire strain gauges10. Have you experienced bursting of ground?NoII. Would you describe failures to date as being structural or stress controlled?Predominantly structurally controllea however stress relatedfailures are increasing as extraction ratio increases248ROCK MECHANICS DATABASEIndicate with a V which of the following parameters have been estimated at your mine.Rock Unit Weight, y VStrength Elastic Modulus, s VParameters Poisson’s Ratio, v VIn-Situ Measurement VStress Photo-Elastic ModuliInvestigations Computer ModelingCompressive Strength, cr, VLaboratory Tensile Strength, atTesting Triaxial Strength, ad tShear Strength, ‘r VFailure CriterionRQD VRock NGI Rating VMass CSIR Rating VClassification Laubscher MRMR Rating VStructural Mapping VMulti-wire Extensometer VBoroscope ObservationMonitoring Compression PadClosure StationLeveling Survey StationPiezometer249i.iI.’:IAVrnn---IIC.,CC.(c‘O000\!SSSSSS4H.0——————(a


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items