libro Ernesto Villaseca Geotechnical desing for Sublevel Open Stoping

Ernesto
Villaescusa
Western Australian School of Mines
Geotechnical Design for
Sublevel Open Stoping
Geotechnical Design for
Sublevel Open Stoping
Ernesto
Villaescusa
Boca Raton London New York
CRC Press is an imprint of the
Taylor & Francis Group, an informa business
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2014 by Taylor & Francis Group, LLC
CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works
Version Date: 20130923
International Standard Book Number-13: 978-1-4822-1189-4 (eBook - PDF)
This book contains information obtained from authentic and highly regarded sources. Reasonable efforts
have been made to publish reliable data and information, but the author and publisher cannot assume
responsibility for the validity of all materials or the consequences of their use. The authors and publishers
have attempted to trace the copyright holders of all material reproduced in this publication and apologize to
copyright holders if permission to publish in this form has not been obtained. If any copyright material has
not been acknowledged please write and let us know so we may rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented,
including photocopying, microfilming, and recording, or in any information storage or retrieval system,
without written permission from the publishers.
For permission to photocopy or use material electronically from this work, please access www.copyright.
com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood
Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and
registration for a variety of users. For organizations that have been granted a photocopy license by the CCC,
a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used
only for identification and explanation without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
Contents
Foreword............................................................................................................... xiii
Preface................................................................................................................... xvii
Acknowledgments............................................................................................... xix
Author.................................................................................................................... xxi
1. Introduction......................................................................................................1
1.1 Mining Method Selection..................................................................... 1
1.2 Self-Supported Mining Methods......................................................... 1
1.3 Sublevel Open Stoping..........................................................................3
1.4 Factors Controlling Stope Wall Behavior...........................................7
1.4.1 Excavation Geometry............................................................... 7
1.4.2 Rock Mass Strength..................................................................9
1.4.3 Induced Stresses...................................................................... 12
1.4.4 Ground Support...................................................................... 13
1.4.5 Blast Damage........................................................................... 15
1.4.6 Drill Drive Layout................................................................... 16
1.5 Scope and Contents of This Book...................................................... 17
2. Sublevel Stoping Geometry........................................................................ 19
2.1 Introduction.......................................................................................... 19
2.2 Stoping Geometries............................................................................. 19
2.2.1 Cutoff Slot................................................................................ 19
2.2.2 Production Rings....................................................................22
2.2.3 Diaphragm Rings....................................................................22
2.2.4 Trough Undercut..................................................................... 23
2.2.5 Drawpoints.............................................................................. 26
2.3 Multiple-Lift Open Stoping................................................................ 26
2.3.1 Tabular Orebodies.................................................................. 28
2.3.2 Massive Orebodies................................................................. 29
2.4 Single-Lift Stoping............................................................................... 31
2.4.1 Conventional Vertical Crater Retreat Stoping....................34
2.4.2 Modified Vertical Retreat Stoping........................................ 36
2.5 Shallow Dipping Tabular Orebodies................................................ 37
2.6 Bench Stoping....................................................................................... 38
3. Planning and Design.................................................................................... 47
3.1 Introduction.......................................................................................... 47
3.2 Geological and Geotechnical Characterization............................... 49
3.3 Stress Analysis in Stope Design......................................................... 49
3.4 Design of Stoping Blocks.................................................................... 52
v
vi
Contents
3.4.1
3.4.2
3.5
Orebody Delineation.............................................................. 53
Global Extraction Sequences................................................. 53
3.4.2.1 Massive Orebodies..................................................54
3.4.2.2 Steeply Dipping Orebodies.................................... 58
3.4.3 Numerical Modeling.............................................................. 73
3.4.4 Regional Pillars....................................................................... 75
3.4.5 Block Development................................................................. 78
3.4.5.1 Shaft Stability........................................................... 78
3.4.5.2 Ramp Access............................................................ 81
3.4.5.3 Crown Pillar............................................................. 82
3.4.5.4 Sublevel Interval......................................................84
3.4.5.5 Access Crosscuts.....................................................84
3.4.5.6 Raises and Orepasses............................................. 86
3.4.5.7 Fill Infrastructure.................................................... 86
3.4.6 Stope Production Scheduling................................................ 89
3.4.6.1 Long-Term Production Scheduling.......................90
3.4.6.2 Medium-Term Activity Schedules........................ 90
3.4.6.3 Short-Term Activity Schedules.............................. 91
3.4.7 Ventilation................................................................................ 92
3.4.8 Global Economic Assessment............................................... 93
Detailed Stope Design......................................................................... 93
3.5.1 Geological Information.......................................................... 97
3.5.2 Development............................................................................ 98
3.5.3 Geotechnical Assessment.................................................... 100
3.5.4 Stope Design Philosophy..................................................... 102
3.5.4.1 Production Rings................................................... 102
3.5.4.2 Diaphragm Rings.................................................. 103
3.5.4.3 Cutoff Slot Design................................................. 104
3.5.4.4 Drawpoint Design................................................. 104
3.5.5 Stope Design Note................................................................ 106
3.5.6 Stope Firing Sequences........................................................ 107
3.5.7 Production Monitoring........................................................ 109
3.5.8 Ventilation.............................................................................. 110
3.5.9 Financial Analysis................................................................ 111
4. Rock Mass Characterization..................................................................... 113
4.1 Introduction........................................................................................ 113
4.2 Characterization from Exploration Core........................................ 115
4.2.1 Drilling Layout Design........................................................ 118
4.2.2 Underground Drilling.......................................................... 118
4.2.3 Core Transfer to Surface....................................................... 118
4.2.4 Drill Core Logging............................................................... 119
4.2.5 Geological Database............................................................. 120
4.2.6 Interpretation of the Orebody and Main
Geological Features.............................................................. 120
Contents
4.3
4.4
4.5
4.6
4.7
4.8
vii
4.2.7 Orebody Meshing in Three Dimensions........................... 121
4.2.8 Problems with Data Analysis.............................................. 121
Analysis of Logging Data................................................................. 122
4.3.1 Discontinuity Linear Frequency......................................... 122
4.3.2 Rock Quality Designation................................................... 125
4.3.3 Rock Mass Classifications from Core Logging................. 131
4.3.4 Advantages, Disadvantages, and Biases
in Core Logging.................................................................... 141
Geotechnical Mapping of Underground Exposures.................... 142
4.4.1 Cell Mapping......................................................................... 144
4.4.2 Line Mapping........................................................................ 145
4.4.3 Strip Mapping....................................................................... 146
4.4.4 Description of Mapping Parameters.................................. 147
4.4.5 Mapping Biases..................................................................... 151
4.4.6 Geological Strength Index................................................... 152
Analysis of Mapping Data................................................................ 153
4.5.1 Discontinuity Orientation................................................... 153
4.5.2 Number of Discontinuity Sets............................................ 155
4.5.3 Discontinuity Spacing.......................................................... 157
4.5.4 Discontinuity Trace Length................................................. 159
4.5.5 Rock Mass Classification Models....................................... 164
Intact Rock Strength.......................................................................... 165
4.6.1 Uniaxial Compressive Strength.......................................... 168
4.6.2 Point Load Strength.............................................................. 173
4.6.3 Confined Compressive Strength......................................... 175
Mechanical Properties of Rock Masses.......................................... 178
4.7.1 Hoek–Brown Empirical Strength Criterion...................... 179
4.7.2 Rock Mass Deformation Modulus...................................... 182
Rock Stress.......................................................................................... 183
4.8.1 Stress Tensor.......................................................................... 184
4.8.2 Stress Measurements Using Oriented Core...................... 185
5. Span and Pillar Design............................................................................... 191
5.1 Background......................................................................................... 191
5.2 Empirical Span Determination Using Rock Mass
Classification Methods...................................................................... 191
5.2.1 Span Determination Using Bieniawski’s
RMR System.......................................................................... 192
5.2.2 Span Determination Using the Tunnel
Quality Index (Q) System.................................................... 197
5.3 Stability Graph Method..................................................................... 197
5.3.1 Updated Determination of the Stability
Graph Parameters................................................................. 200
5.3.1.1 Factor A................................................................... 201
5.3.1.2 Factor B................................................................... 203
viii
Contents
5.4
5.5
5.3.1.3 Factor C................................................................... 206
5.3.1.4 Hydraulic Radius.................................................. 207
5.3.2 Prediction of Stope Stability................................................ 209
5.3.3 Use of the Stability Graph as a Design Tool...................... 213
5.3.4 Design Validation................................................................. 219
Numerical Modeling of Stope Wall Stability.................................222
5.4.1 Linear Elastic Numerical Modeling...................................223
5.4.2 Nonlinear Numerical Modeling.........................................225
Pillar Stability Analysis..................................................................... 231
5.5.1 Basic Concepts....................................................................... 231
5.5.2 Average Pillar Stress Using the Equivalent
Area Approach...................................................................... 232
5.5.3 Empirical Rib Pillar Stability Chart................................... 233
5.5.4 Confinement Pillar Stability Chart....................................234
5.5.5 Numerical Modeling for Pillar Design.............................. 240
6. Drilling and Blasting.................................................................................. 245
6.1 Introduction........................................................................................ 245
6.2 Longhole Drilling............................................................................... 245
6.2.1 Top-Hammer Drilling.......................................................... 247
6.2.2 In-the-Hole Drilling.............................................................. 247
6.2.3 Drilling Equipment Selection............................................. 248
6.2.4 Drilling Deviation................................................................. 249
6.2.4.1 Collar Positioning.................................................. 250
6.2.4.2 Drillhole Alignment............................................. 251
6.2.4.3 In-the-Hole Deviation........................................... 252
6.3 Blast Design Parameters................................................................... 258
6.3.1 Drilling Orientation.............................................................. 260
6.3.2 Blasthole Diameter................................................................ 262
6.3.3 Blasthole Length................................................................... 264
6.3.4 Burden.................................................................................... 265
6.3.5 Spacing................................................................................... 267
6.3.6 Stemming and Uncharged Length..................................... 268
6.4 Ring Design........................................................................................ 269
6.4.1 General Procedure................................................................ 270
6.4.2 Parallel Patterns.................................................................... 273
6.4.3 Radial Patterns...................................................................... 274
6.4.4 Vertical Crater Retreat Blasting.......................................... 277
6.5 Explosive Selection............................................................................ 280
6.5.1 Packaged versus Bulk Explosives....................................... 281
6.5.2 Ammonium Nitrate-Based Explosives.............................. 281
6.5.3 ANFO..................................................................................... 282
6.5.4 Watergels or Slurries............................................................ 282
6.5.5 Emulsions............................................................................... 283
6.5.6 Special ANFO and Emulsion Blends................................. 285
Contents
ix
6.6
Explosive Placement.......................................................................... 285
6.6.1 Powder Factor........................................................................ 287
6.6.2 Energy Distribution.............................................................. 288
6.7 Initiation Systems............................................................................... 289
6.7.1 Pyrotechnic Delay Element Detonators............................. 289
6.7.2 Available Timing and Sources of Timing Error for
Pyrotechnic Delay Elements............................................... 290
6.7.3 Electronic Delay Element Detonators................................ 292
6.7.4 Priming................................................................................... 294
6.7.5 Sequencing and Timing....................................................... 295
6.8 Raise and Cutoff Slot Blasting.......................................................... 298
6.8.1 Longhole Winzes.................................................................. 298
6.8.2 Cutoff Slots............................................................................. 302
6.9 Trough Undercut Blasting................................................................ 307
6.10 Rock Diaphragm Blasting.................................................................308
6.11 Mass Blasting......................................................................................309
6.11.1 Control of Ground Vibration............................................... 312
7. Rock Reinforcement and Support............................................................ 315
7.1 Introduction........................................................................................ 315
7.2 Terminology........................................................................................ 317
7.2.1 Continuous Mechanical Coupled....................................... 318
7.2.2 Continuous Friction Coupled.............................................. 318
7.2.3 Discrete Mechanical and Friction Coupled...................... 319
7.2.4 Load Transfer Concept......................................................... 319
7.2.5 Embedment Length Concept.............................................. 320
7.2.6 Reinforcement Performance Indicators............................. 321
7.3 Ground Support Design.................................................................... 322
7.3.1 Location of Failure due to Overstressing.......................... 324
7.3.2 Depth of Failure: Stress or Strain Controlled................... 324
7.3.3 Depth of Failure: Structurally Controlled......................... 326
7.3.4 Ground Reaction Curve Concept....................................... 328
7.3.5 Ground Support for Massive Rock and Low Stress......... 329
7.3.6 Ground Support for Massive Rock and
Moderate Stress..................................................................... 329
7.3.7 Ground Support for Massive Rock and High Stress....... 330
7.3.8 Ground Support for Layered Rock and Low Stress......... 332
7.3.9 Ground Support for Layered Rock and
Moderate Stress..................................................................... 333
7.3.10 Ground Support for Layered Rock and High Stress.......334
7.3.11 Ground Support for Jointed Rock and Low Stress..........334
7.3.12 Ground Support for Jointed Rock and
Moderate Stress.................................................................... 336
7.3.13 Ground Support for Jointed Rock and High Stress......... 337
7.3.14 Design by Precedent Rules.................................................. 338
x
Contents
7.4
7.5
7.6
7.7
7.8
7.3.15 Design by Rock Mass Classification...................................340
7.3.16 Reinforcement Layout..........................................................343
7.3.17 Energy Release......................................................................343
7.3.18 Rock Mass Demand..............................................................344
Rock Bolting of Open Stope Development Drives........................345
7.4.1 Continuous Mechanical Coupled Rock Bolts...................346
7.4.1.1 Cement-Encapsulated Threaded Bar..................346
7.4.1.2 Resin-Encapsulated Threaded Bar..................... 347
7.4.2 Continuous Friction Coupled Rock Bolts.......................... 352
7.4.2.1 Split-Tube Friction Rock Stabilizers.................... 352
7.4.3 Discrete Mechanical or Friction Coupled Rock Bolts......354
7.4.3.1 Expansion Shell Rock Bolts.................................. 355
7.4.4 Rock Bolts with Yielding Mechanisms............................... 357
Cable Bolting of Open Stope Walls................................................. 360
7.5.1 Cable Bolt Reinforcement Mechanisms............................. 363
7.5.2 Cable Bolt Types.................................................................... 366
7.5.2.1 Plain Strand Cable Bolts....................................... 366
7.5.2.2 Modified Strand Cable Bolts................................ 366
7.5.2.3 Debonded Plain Strand Cable Bolts................... 368
7.5.2.4 Cable Bolt Plates.................................................... 369
Cable Bolt Corrosion.......................................................................... 370
7.6.1 Corrosivity of Cable Bolt Strands....................................... 370
7.6.2 Corrosivity of Cable Bolt Anchors...................................... 374
Cement Grouting of Cable Bolts...................................................... 378
7.7.1 Collar to Toe Grouting......................................................... 378
7.7.2 Toe to Collar Grouting......................................................... 380
Support Systems................................................................................. 383
7.8.1 Plates....................................................................................... 383
7.8.2 Straps......................................................................................384
7.8.3 Mesh........................................................................................ 385
7.8.3.1 Mesh Testing.......................................................... 386
7.8.3.2 Mesh Force and Displacement............................ 389
7.8.4 Thin Spray on Liners............................................................ 395
7.8.5 Shotcrete Layers.................................................................... 395
7.8.5.1 Shotcrete Support Mechanisms.......................... 396
7.8.5.2 Shotcrete Reaction to Transverse Loading........ 397
7.8.5.3 Shotcrete Reaction in Tension.............................. 398
7.8.5.4 Shotcrete Reaction in Compression.................... 398
7.8.5.5 Shotcrete Toughness............................................. 399
8. Mine Fill........................................................................................................ 405
8.1 Introduction........................................................................................ 405
8.2 Unconsolidated Rock Fill.................................................................. 406
8.2.1 Rock Fill for Bench Stope Support...................................... 409
8.3 Cemented Rock Fill............................................................................ 412
Contents
8.4
8.5
8.6
8.7
xi
8.3.1 Cemented Aggregate Fill..................................................... 413
Hydraulic Fill...................................................................................... 418
Cemented Paste Fill...........................................................................422
Open Stope Fill Operations Systems............................................... 426
8.6.1 Material Preparation............................................................. 427
8.6.1.1 Chemistry and Mineralogy................................. 428
8.6.1.2 Particle Size Distribution..................................... 428
8.6.1.3 Binders.................................................................... 428
8.6.1.4 Admixtures............................................................ 431
8.6.1.5 Mixing Water......................................................... 431
8.6.1.6 Mix Design............................................................. 432
8.6.2 Stope Preparation..................................................................434
8.6.2.1 Design Criteria for Fill Barricades......................434
8.6.2.2 CHF Barricades......................................................434
8.6.2.3 CRF Barricades...................................................... 437
8.6.2.4 CPF Barricades....................................................... 438
8.6.3 Material Delivery.................................................................. 439
8.6.3.1 Rock Fill Passes......................................................440
8.6.3.2 Slurry Fill Passes...................................................440
8.6.4 Fill Placement........................................................................ 441
8.6.4.1 CHF Placement...................................................... 441
8.6.4.2 CRF Placement....................................................... 441
8.6.4.3 CPF Placement.......................................................442
Fill Monitoring and Quality Control..............................................443
8.7.1 Fill Supply..............................................................................443
8.7.2 Fill Plant.................................................................................443
8.7.3 Fill Reticulation.....................................................................444
8.7.4 Fill Placement........................................................................444
8.7.5 Barricade Performance.........................................................445
9. Dilution Control.......................................................................................... 447
9.1 Introduction........................................................................................ 447
9.2 Types of Dilution................................................................................ 449
9.2.1 Internal Dilution................................................................... 449
9.2.2 External Dilution................................................................... 450
9.2.3 Geological Dilution............................................................... 451
9.2.4 Ore Loss................................................................................. 451
9.3 Economic Impact of Dilution............................................................ 452
9.4 Parameters Influencing Dilution..................................................... 453
9.4.1 Dilution at the Orebody Delineation Stage....................... 455
9.4.2 Dilution at the Design and Sequencing Stages................. 456
9.4.3 Dilution at the Stope Development Stages........................ 458
9.4.4 Dilution at the Production Drilling
and Blasting Stages............................................................... 459
9.4.5 Dilution at the Production Stages....................................... 460
xii
Contents
9.5
9.6
9.7
9.4.6 Dilution Issues for Mine Management..............................463
Cavity Monitoring System................................................................464
Dilution Control Plan........................................................................ 466
9.6.1 Stope Performance Review.................................................. 468
Scale-Independent Measures of Stope Performance..................... 472
9.7.1 Conventional Measures....................................................... 474
9.7.2 Circularity Measures............................................................ 476
9.7.3 Extensivity Measures........................................................... 476
9.7.4 Hemisphericity Measures.................................................... 477
9.7.5 Cannington Mine Example................................................. 478
References............................................................................................................ 481
Foreword
Underground metalliferous mining in Australia began in the mid-1840s
at the copper and silver–lead mines in and around Kapunda and Burra in
South Australia. Mining in the Victorian goldfields following the discovery
of gold and the Gold Rush of 1851 was initially alluvial but soon evolved into
the underground mining of deep leads and then quartz veins. By 1895, the
180 Mine at Bendigo was, at 970 m deep, the deepest mine in the world. The
rich silver–lead–zinc orebodies of Broken Hill were discovered in 1883 and
gold in Western Australia in 1892. By that time, Australia’s mining industry
had already seen a number of boom-and-bust cycles. However, new discoveries have continued to be made and new mines developed up to the present
day, with mining remaining a mainstay of Australia’s export economy, particularly in recent decades.
In the 1950s, the dry fill formerly used was replaced by hydraulically placed
fill in a number of Australian underground metalliferous mines. Mechanized
cut-and-fill methods of mining were introduced for the lead orebodies at
Mount Isa in 1964 and were soon adopted by other mines. During the 1960s,
mining in Australia and elsewhere benefited greatly from the advances that
were then taking place in the emerging science of rock mechanics. By the
1970s, cut-and-fill was one of the major underground metalliferous mining methods used in Australia, and in Canada and Scandinavia as well, but
demand for higher productivity led to a transition to a range of sublevel and
longhole open stoping methods, usually with backfill, until these became the
most widely used methods in Australia. Although mass mining methods
using sublevel and block caving have been used increasingly since the 1990s
for mining some types of orebody, sublevel open stoping remains the primary method used for the underground mining of base and precious metals
in Australia.
The mining literature of recent decades includes conference proceedings
and specialist monographs on cut-and-fill and caving methods of mining
and on the mining of tabular orebodies such as the deep level gold-bearing
reefs of South Africa. Because of its continuing importance in most of the
world’s major metalliferous mining countries, including, but not limited to,
Australia, Canada, and the Scandinavian and South American countries, it
is entirely appropriate that a book should now appear synthesizing 40 years’
accumulated international experience with modern sublevel open stoping
methods. As will be argued in the following text, the author of this book,
Professor Ernesto Villaescusa, is supremely well qualified to undertake this
important task.
I first met Ernesto Villaescusa in early 1988 shortly after I had moved
to the University of Queensland, Brisbane, Australia, from Imperial
xiii
xiv
Foreword
College, London. Ernesto was introduced to me by Professor Alban Lynch,
AO, the distinguished foundation director of the University’s world famous
Julius Kruttschnitt Mineral Research Centre (JKMRC). Ernesto had just
joined the Centre as a research scholar in its then Mining Research Group. He
was interested in doing his PhD research in an area of mining rock mechanics and was looking for a supervisor. Previously, Professor Lynch had kindly
invited me to become associated with the JKMRC and to carry out my then
necessarily limited research-related activities through the Centre. I have to
admit that, initially, I was not at all enthusiastic about taking on a PhD student when I was trying to establish myself in a senior position in a new
university. However, Ernesto’s enthusiasm, persistence, and determination,
and Professor Lynch’s more gentle powers of persuasion, jointly won the day,
and I became Ernesto’s PhD supervisor for the next three years. That was
the beginning of a friendship and close professional relationship that has
continued now for 25 years.
After completing an excellent PhD thesis in 1991, Dr. Villaescusa joined
Mount Isa Mines as a rock mechanics engineer. In 1994–1995, he spent
some time at the Noranda Technology Centre in Canada, before returning
to Mount Isa Mines in 1995 as principal rock mechanics engineer. Then in
1997, at a very young age for a full professor in an Australian university, he
was appointed professor of mining geomechanics at the Western Australian
School of Mines (WASM), Kalgoorlie, a position that he continues to hold
in what since 2004 has been the industry-sponsored industry chair in mining rock mechanics. At WASM, Professor Villaescusa has built up a leading
applied mining rock mechanics research group, taught mining engineering
at undergraduate and postgraduate levels, carried out and/or supervised a
wide range of industry-sponsored mining rock mechanics research projects,
and acted as a consultant to the industry in Western Australia, elsewhere in
Australia, and in South America, mainly in the general area of underground
metalliferous mining.
Because of his directly relevant industry, applied research, teaching, and
consulting experience and his extensive list of publications in the area,
Professor Ernesto Villaescusa is eminently well-qualified to write this book,
Geotechnical Design for Sublevel Open Stoping. In particular, he has wide practical experience of sublevel open stoping and its variants at a large number of
mines that use these and other mining methods in Australia, Canada, Chile,
New Zealand, and in his native Mexico. He also has the great advantage of
having gained research training and practical mining experience in the basic
mining science of rock mechanics.
I have enjoyed the unusual privilege of having been asked by Professor
Villaescusa to offer comment and advice on the contents of his book as it has
developed through the various stages of its preparation. Although, as the
title suggests, the book has a geotechnical engineering orientation, it also
contains considerable practical detail on open stoping layouts, design, and
operations and includes chapters on drilling and blasting, rock support and
Foreword
xv
reinforcement, mine fill technology, and dilution control. Some of this material draws heavily on results obtained, and understandings developed, in
industrially sponsored research projects carried out by Professor Villaescusa,
his colleagues, and his students at WASM.
I believe that this book will serve multiple purposes. It will serve as a
specialist textbook for mining courses at the advanced undergraduate and
postgraduate levels. It will also provide an authoritative, practically oriented
reference work for those involved in the industry, both in mining operations
and as consulting engineers, particularly for those in the early stages of their
careers and those seeking to develop new understandings and skills. I congratulate Professor Villaescusa on this outstanding achievement and unhesitatingly recommend the book to those having an interest in the industrially
important sublevel open stoping methods of underground mining.
Edwin T. Brown, AC
Senior Consultant, Golder Associates Pty Ltd, Brisbane, Queensland, Australia
Emeritus Professor, University of Queensland, Brisbane, Queensland, Australia
President, International Society for Rock Mechanics, 1983–1987
Preface
Sublevel open stoping is one of the most widely used mining methods in
underground metalliferous mining. This method allows for low cost, high
recovery, and productivity while providing operational safety to personnel
and equipment. The success of the method relies on the stability of stope
walls and crowns, as well as any fill masses exposed. Although it is not a
selective method, the stope boundaries can be designed so that dilution and
ore loss can be minimized.
Over the last 30 years or so, increased understanding of the factors controlling stope spans and stability have been developed. In addition, improvements in drilling equipment, ventilation, cablebolt reinforcement, fill mass
strength, and routine implementation of stope void monitoring systems have
led to significant improvements in sublevel open stoping. In the future, the
method is likely to be used under more difficult geotechnical conditions,
and therefore, a better understanding of all technical and operating factors
influencing its success is required.
This book was written primarily for fourth year undergraduate students,
graduate students, and junior practitioners not yet entirely familiar with the
mining method. The book is divided into nine chapters that closely follow the
approach used by most mining houses in implementing the method worldwide. After the basic nomenclature is introduced, the method is reviewed
from orebody delineation, planning and design through key operations such
as drilling and blasting, ground support of access drives and stope walls,
as well as stope void filling. The book also includes a dilution control chapter given that documentation of stope performance is critical to improve the
design to optimize the method.
The material presented draws heavily on my experience at Mount Isa
Mines as well as from technical reviews of many mine sites worldwide. The
book also relies upon results of industry-sponsored research undertaken at
the Western Australian School of Mines (WASM) over the last 16 years or so.
Without the results of my postgraduate students, the book would not have
been possible. I wish to record my deep gratitude to Professor Ted Brown,
who has provided me with inspiration and advice throughout the entire process of book writing including technical content, layout, and numerous comments for improvement. My gratitude also goes to Professor Will F. Bawden
who early in the process provided me with unpublished material and comments to chapters. At WASM, I benefited from the friendship and technical support of the principal research fellows Dr. Alan Thompson and Chris
Windsor as well as the administrative and financial support from WASM
xvii
xviii
Preface
directors including Professors Peter Lilly, Eric Grimsey, Paul Dunn, and
Steve Hall. I also wish to thank the CRC Mining directors, Professors Mike
Hood and Paul Lever, for their financial support. I wish to thank Mount Isa
Mines for their permission to use previously unpublished material. Similarly,
I wish to thank other organizations and authors who freely gave me permission to reproduce published material.
Ernesto Villaescusa
Western Australian School of Mines
Acknowledgments
This book was written with university undergraduate students in mind. It
draws heavily on the knowledge and practical experience gained during my
years of employment at Mount Isa Mines from 1991 to 1997, my course notes
and interaction with students while teaching underground rock mechanics
at the Western Australian School of Mines (WASM) from 1997 to 2007, as well
as on outcomes from my WASM research students from 1997 to 2013. I wish
to acknowledge the following important contributions to this book:
• Professor Ted Brown, AC, who over the years has provided me with
many ideas and made invaluable suggestions about the book content and layout and has also carefully reviewed every chapter. His
friendship and technical advice started while doing my PhD studies
and continues to this day.
• Dr. Alan Thompson, who has been a great friend, for his technical
support and encouragement, which made writing the book a lot
easier. I also acknowledge his deep intellect and his contributions
to Chapter 7.
• Chris Windsor, who over the years has provided many technical
suggestions as to how to improve our research work at WASM. His
friendship has always been of great support. I would like to acknowledge his contributions to Chapters 4 and 7.
• Professor Will F. Bawden, who provided comments to some early
draft chapters and substantially wrote Sections 5.1 and 5.2. He also
gave me permission to use his contributions to write Sections 5.2.1
and 5.5.
• Dr. Peter Cepuritis, who, as part of his PhD studies, undertook many
of the calculations that are included in the book.
• Dr. Kelly Fleetwood, who carefully reviewed and made suggestions
to Chapter 6 and personally wrote Section 6.5.
• Dr. John Player, for his decade-long innovative work in ground support at WASM, some of which is presented in Chapter 7.
• Dr. Jianping Li, my first PhD student at WASM, who made many
contributions to rock testing and in situ stress measurements, the
results of which are reflected in Chapter 4.
• Dr. Rhett Hassell, for his contributions to the corrosion work
presented in Section 7.6.
xix
xx
Acknowledgments
• Nixon Saw, for his excellent work on fill testing presented in Chapter 8.
• Ellen Morton, for her work on mesh and shotcrete testing and for her
contributions to Chapter 7.
I am also grateful to a number of friends and colleagues that I have worked
with at a number of mine sites over the last 20 years or so. Their practical
approach and ideas have helped me write this book. In particular, I would
like to thank
• Leigh Neindorf, Mark Adams, and Mike Sandy, then of Mount Isa
Mines
• Dave Finn, then of WMC and later Placer Dome
• Peter Teasdale, then of WMC
• Cam Schubert, then of Mount Isa Mines and later McArthur River
Mining
I also wish to thank all those who supported me throughout this undertaking, especially
• Luis Machuca and Moises Cordova, for their great friendship and
constant, unwavering support
• Mike Westerman, for his permission to use previously unpublished
material from Mount Isa Mines; the support of Mount Isa Mines is
also gratefully acknowledged
• The WASM rock mechanics technical staff, including Brett Scott,
Lance Fraser, and Pat Hogan
• The administrative staff and a number of WASM undergraduate and
graduate students, including Ben Auld, Tom Parrott, Cesar Pardo,
Andrea Roth, Catherine Winder, Ayako Kusui, and Andres Brzovic,
among many others
• The sponsors of the WASM Rock Mechanics Chair who funded my
position at WASM, which include Goldfields, Barrick, Barminco,
Newcrest, and Curtin University of Technology; the financial support of CRC Mining is also gratefully acknowledged
• The authors and publishers who have given permission for the
reproduction of previously published figures and tables
• Finally, but most importantly, to Carolyn and Tiana, for their love,
patience, tolerance, and understanding of my dedication to exploration, mining, and rock mechanics
Author
Professor Ernesto Villaescusa received his
BEng in mining engineering (first class honors) from Universidad de Sonora, Mexico in
1984; his MSc in mining engineering from
Colorado School of Mines, Golden, Colorado
in 1987; and his PhD in mining engineering
from the University of Queensland, Brisbane,
Queensland, Australia in 1991.
He has over 25 years of applied research
experience having worked with a large number of mining houses such as MIM Holdings,
Noranda, WMC Resources, Peñoles, Minera Autlan, CODELCO, BHP Billiton,
Placer Dome Asia Pacific, and Normandy to develop guidelines for effective underground mining, leading to a safe, economical extraction of ore.
He has undertaken applied research in all aspects of mining methods for a
range of rock mass and geotechnical conditions ranging from shallow depth
cut-and-fill mines, room and pillar, to deep sublevel open stopes and block
cave mines (the picture below shows him inspecting stope hangingwalls at
Mount Isa Mines).
Over the last 16 years, he has worked at the
Western Australian School of Mines (WASM)
as a professor of mining geomechanics, where
he has secured over 21 million dollars of
­industry-funded mining research income. He
has supervised over 30 master’s and 10 PhD
student theses and has written over 100 technical papers. In 2004, he was appointed to an
industry chair in mining rock mechanics at
WASM. The chair is currently sponsored by
Barrick, Goldfields, Barminco, Newcrest, and
Curtin University.
xxi
1
Introduction
1.1 Mining Method Selection
The design and selection of a mining method requires a systematic approach,
with the dip, size, and shape of an orebody; the strengths of the ore and
the host rock mass; as well as economics being some of the fundamental
parameters influencing the planning and design process (Hamrin, 1982;
Brady and Brown, 2004). Distinctions can be made between orebodies having significant width, height, and length and those that are small in one
dimension and are either steeply or shallowly dipping. For example, orebodies with significant vertical dimensions can be accessed through drifts developed at successive depths. Gravity is used to advantage in ore-breaking and
ore-handling operations, as the broken material can be directed to the conveniently located draw (collection) points. When an orebody is thin, requiring
full entry for personnel and equipment, a critical consideration, as the mining face is advanced, is protection from rock falls (Figure 1.1). In most cases,
when an orebody is large in all dimensions, access is via small drifts that are
located outside the main production zones. The selected mining method will
exclude other options on a safety, productivity, recovery, and dilution control
basis. Brady and Brown (2004) have discussed the general relation between
the geotechnical properties of an orebody, the host rock mass, and the most
appropriate mining method.
1.2 Self-Supported Mining Methods
The stability of the rock mass greatly influences the choice of mining method.
Stable rock masses allow extensive exposures of the backs (roofs) and walls
of underground openings (Figure 1.2). Self-supported openings are those in
which the overlying load is redistributed through the rock mass and carried
by the side walls and pillars. The ore can be removed from an underground
1
2
Geotechnical Design for Sublevel Open Stoping
FIGURE 1.1
Stabilized stope access drift prior to sublevel open stoping extraction.
FIGURE 1.2
Very stable and large stope back in a good-quality rock mass.
opening without the use of materials for back and wall support. F
­ or safety,
ground support may still be required at individual locations or at regular
intervals. Examples of self-supporting mining methods (Brady and Brown,
2004) include open stope mining (the subject of this book) and room and pillar mining, which will not be discussed further here.
Introduction
3
Sublevel stoping is designed for the progressive extraction of specified
ore blocks between pillars of surrounding material. The objective is to mine
as much of a deposit as possible in the initial open stopes with low risk of
ground movement and without jeopardizing the recovery of adjacent pillar
ore. Therefore, open stoping represents an integrated and staged system of
total ore recovery. Primary open stoping is usually followed by secondary
and sometimes tertiary extraction phases to recover pillar ore. The stope walls
must be self-supporting to ensure that the excavation is stable to allow primary
stope mining without dilution. The ore should also be strong to ensure stable
secondary and tertiary pillars. Pillar recovery requires the use of consolidated
fill material that is placed into the primary stope voids to allow stable secondary and sometimes tertiary stope extraction. Although sublevel open stoping
is essentially a self-supported mining method, in this sense it can overlap with
artificially supported methods as identified by Brady and Brown (2004).
1.3 Sublevel Open Stoping
Sublevel open stoping methods are used to extract large massive or tabular,
often steeply dipping, competent orebodies surrounded by competent host
rocks, which in general have few constraints regarding the shape, size, and
continuity of the mineralization. The success of the method relies on the stability of large (mainly unreinforced) stope walls and crowns, as well as the
stability of any fill masses exposed. In good quality rock masses, open stopes
can be relatively large excavations (Figure 1.3), in which ring drilling and
blasting is the main method of rock breakage. Ore dilution consisting of lowgrade, waste rock or minefill materials may occur at the stope boundaries.
In addition, ore loss due to insufficient breakage can also occur within the
stope boundaries.
Two basic stope configurations are possible: longitudinal and transverse. In
both configurations, the ore is mined from sublevels by some form of benching
and flows by gravity to a drawpoint. Longitudinal sublevel stoping is used for
comparatively narrow, usually less than 15 m, steeply dipping orebodies with
the stoping running parallel to the strike of the orebody. For thick orebodies,
the stopes are oriented perpendicular (transverse) to the strike of the deposit
with pillars left between the primary stopes. Full recovery of stope and pillars requires the use of consolidated fill (Brady and Brown, 2004).
The method is widely applied worldwide and offers several advantages, including low cost and efficient nonentry production operations. It
utilizes highly mechanized, mobile drilling and loading production equipment to achieve high production rates with a minimum level of personnel.
Furthermore, the production operations of ring drilling, blasting, and drawpoint mucking are concentrated into few locations. The disadvantages
4
Geotechnical Design for Sublevel Open Stoping
Mount Isa Mines
Lead, silver, and zinc stopes
350 m
300 m
Typical
Tallest
250 m
200 m
150 m
100 m
50 m
0m
FIGURE 1.3
Large-scale stoping operations at Mount Isa Mines. (Courtesy of Mount Isa Mines, Mount Isa,
Queensland, Australia.)
include a requirement for a significant level of development infrastructure
before production starts, thus incurring a high initial capital investment.
However, a large part of the development occurs within the orebody. In
addition, the stopes must be designed with regular boundaries, and internal
waste pockets cannot be separated within the broken ore. Similarly, delineated ore cannot be recovered beyond a designed stope boundary.
Technical developments regarding the understanding of rock mass and fill
behavior, in conjunction with dilution measuring techniques, improved blasting, equipment, ventilation, and ground support practices, currently allow
for the successful application of this method in increasingly complex geological and mining situations, even at great depth. In particular, an increased
understanding of the method is required to facilitate improved stope access
configurations and optimized extraction sequences, leading to full orebody
recovery while achieving dilution control. The complexity of the method and
Introduction
5
the current depth of the orebodies being extracted worldwide suggest that
adequate planning and control of the operations are critical to the successful implementation of optimum stope sizes and sequences of extraction. The
method is commonly known throughout the world as open stoping, sublevel
stoping, and longhole or blasthole stoping. The following are the essential
common elements of sublevel stoping (Mathews, 1978; Bridges, 1983):
• The stopes are open and extracted without substantial wall collapse
or caving.
• The stopes extend from sublevel to sublevel, with operations carried
out only at these sublevels.
• The blasted rock moves by gravity alone to the stope drawpoints.
• The method uses long blastholes for rock breakage, achieving good
fragmentation (Figure 1.4).
• The blastholes are located within planes called rings.
• The holes can be drilled downward or upward.
• The initial expansion slot is located on the side, center, or bottom of
each stope.
• The method is nonentry, and personnel do not have access to the
open portion of a stope (Figures 1.5 and 1.6).
FIGURE 1.4
Typical rock fragmentation from sublevel open stoping blasting.
6
Geotechnical Design for Sublevel Open Stoping
FIGURE 1.5
A view inside an open stope.
Hangingwall
Drill
drives
Endwall
Production drill
rings
Access
crosscut
Footwall
access drives
Extraction
level
Trough undercut
Drawpoints
Tipple
FIGURE 1.6
Three-dimensional view of a multiple lift, transverse sublevel open stope. (From Villaescusa,
E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane,
Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne,
Victoria, Australia. With permission.)
Introduction
7
FIGURE 1.7
A large-scale longitudinal bench stope.
Within this context, the extraction of narrow, lenticular orebodies by longitudinal bench stoping (Villaescusa et al., 1994) is also included among
the sublevel stoping geometries and is considered in detail in this book
(Figure 1.7).
Over the last 20 years or so, technology has been developed to improve the
safety and economics of ore extraction by sublevel stoping and benching.
Experience indicates that geological discontinuities, stresses, blast damage,
excavation geometry, and ground support are the main factors controlling
stope wall behavior and stability. These factors will be introduced and discussed briefly in the following subsections. They will be discussed in more
detail in the subsequent chapters of this book.
1.4 Factors Controlling Stope Wall Behavior
1.4.1 Excavation Geometry
In sublevel stoping, drilling and blasting is undertaken from drilling
drives located on sublevels strategically placed over the height of a stope.
8
Geotechnical Design for Sublevel Open Stoping
Stope height
Transition zone
S
T
U
S = Stable
T = Transitional
U= Unstable
Unstable shapes
Critical dimension
Stope length/width
Unstable shapes
Critical dimension
Stope height
U
Transition zone
T
S
Stope length or width
FIGURE 1.8
Stable shapes for sublevel stoping.
Because of the limited cablebolt reinforcement that can be provided at the
exposed stope walls, the excavations must be designed to be inherently stable. In this regard, experience has shown that, in general, it is possible to
achieve stope wall stability with minimal dilution by creating openings having high vertical and short horizontal dimensions. An example would be a
stable, vertical raisebore that is extended laterally, until it becomes unstable.
Stability is also achieved by forming openings having long horizontal and
short vertical dimensions. An example would be a long, stable tunnel, whose
height is increased until it becomes unstable. Square-shaped stopes are the
most ineffective in terms of potentially stable volumes (Figure 1.8).
The shape of the conceptual transition curve in Figure 1.8 is hyperbolic
and indicates that for multiple lift sublevel open stopes (excavations with
walls that have high vertical and short horizontal dimensions) the critical
spans are either the exposed horizontal lengths or the stope widths. Length
and width, that is, dimensions in plan view, are the critical stope dimensions as they also control the dimensions of the stope crowns. Bench stopes
are excavations where the longest dimension is the strike length and the
critical spans are usually the exposed heights, as the orebody width is usually narrow. Figure 1.9 shows an example of hangingwall performance for
single- and double-lift stopes extracted in a similar geotechnical domain.
9
Introduction
60
Stope up-dip span (m)
50
40
30
Depth of failure (m)
0–1 m
1–2 m
20
2–3 m
10
0
3–4 m
>4 m
0
10
20
30
40
Stope strike length (m)
50
60
FIGURE 1.9
Stope performance—steeply dipping tabular rock mass, Mount Marion Mine.
The case study data show that for the single-lift stopes, stope performance
is not controlled by geometry, as the depth of failure is not correlated with
stope dimensions. However, as the stope height is increased, the depth of
failure increases with an increase in stope strike length. An immediate conclusion is that a reduction in stope size may not necessarily result in better
stope performance. Another case study is shown in Figure 1.10, in which
stope performance is clearly related to stope geometry.
1.4.2 Rock Mass Strength
It is generally accepted that the behavior of the stope walls is largely controlled by the strength of the rock mass surrounding the stope. This rock
mass strength depends upon the geometrical nature and strength of the geological discontinuities as well as the physical properties of the intact rock
bridges. Single or combinations of major discontinuities (usually continuous on the scale of a stoping block) such as faults, shears, and dykes usually
have very low shear strengths and, if oriented unfavorably, provide failure
surfaces when exposed by the stope walls (Figure 1.11). Such geological discontinuities largely control overbreak and stability around exposed stope
walls. This is particularly the case for those discontinuities having platy and
water-susceptible mineral infill such as talc, chlorite, and sericite.
10
Geotechnical Design for Sublevel Open Stoping
100
5.6
Stope up-dip span (m)
80
5.3
3.4
60
3.4
40
2.5
20
0
0
20
40
2.8
2.5
3.6
5.2
2.8
Stope depth of
failure (m)
60
80
100
Stope strike length (m)
FIGURE 1.10
Stope performance—shallowly dipping tabular rock mass, Davyhurst Mine. (From Parker, B.
2004. Geotechnical study of shallow dipping orebodies—Lights of Israel Underground Gold
Mine. BEng thesis, Mining Engineering Department, WA School of Mines, Curtin University
of Technology, Perth, Western Australia, Australia.)
FIGURE 1.11
Stope hangingwall stability controlled by large-scale faulting.
11
M
80
°
°
52
55
M
5500 N
N5
4
4500 N
5000 N
Introduction
44
J 54
T
45
S4
8
80
°
O
50
52
53
80
T
°
75°
1100 Cu orebody
70
°
°
75
°
FIGURE 1.12
Plan view of major structures affecting sublevel stoping—1100 Orebody, Mount Isa Mines.
(From Alexander, E.G. and Fabjanczyk, M.W., Extraction design using open stopes for pillar
recovery in the 1100 ore body at Mount Isa, in D.R. Stewart, ed., Design & Operation of Caving &
Sublevel Stoping Mines, SME of AIME, New York, 1981, pp. 437–458.)
In some cases, instability can be linked to activities in concurrent voids
along the strikes or dips of major geological features such as fault zones
(Logan et al., 1993). Ideally, the location of large-scale geological discontinuities is well defined and most open stoping mines have a threedimensional model of the local fault/shear network (Figure 1.12). These
features can also be seismically active, further increasing falloff at the excavation boundaries, especially in narrow orebodies. When large-scale structures are exposed, stope wall overbreak is usually very difficult to control,
regardless of the blasting practices used, and can only be minimized by
stope sequencing.
Stope wall behavior is also a function of the number, size, frequency, and
orientation of the minor-scale geological discontinuities. Such discontinuity
networks usually control the nature and amount of overbreak at the stope
boundaries. Rock mass characterization techniques can be used to estimate
the shapes and sizes of blocks likely to be exposed at the final stope walls.
The geometrical discontinuity set characteristics (size, frequency, orientation, persistence, surface strength, etc.) relative to the stope walls largely
control the amount of dilution experienced at those walls (Figure 1.13).
Individual joints have a limited size and they may either terminate in intact
rock, forming an intact rock bridge, or against another structure within a
discontinuity network. These intact rock bridges are significantly stronger
than the naturally occurring discontinuities and provide a higher resistance
to failure within a rock mass.
12
Geotechnical Design for Sublevel Open Stoping
FIGURE 1.13
Example of stope large-scale footwall and hangingwall falloff. (From Villaescusa, E., A review
of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland,
Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria,
Australia. With permission.)
1.4.3 Induced Stresses
Extraction within a stoping block can generate large concentrations of stress
around the excavation boundaries. If the local (induced) stresses increase
beyond the strength of a rock mass, then changes in the rock mass quality
around a stope will occur, and localized failures are likely to be experienced either following discontinuity surfaces or directly through intact rock.
Where movement through discontinuities occurs, stresses are relieved. This
may in turn lead to overbreak, dilution, or large failures (Figure 1.14).
Rock mass quality changes around the boundaries of a stope result from a
combination of stress redistributions, near field blast damage, and the effects
of the excavation itself. In cases where stope wall failures do not occur due to
stress concentration, vibration and gases from nearby blasting may damage
the intact rock bridges, which define and interlock the in situ rock blocks,
causing overbreak or dilution at the stope boundaries. Furthermore, the
dynamic behavior of an unsupported wall is directly related to the amount
of intact rock available within the rock mass. The less intact rock available,
the more cracking, slabbing, and visible stope wall displacement will result
from the blasting process.
13
Introduction
Roc
k fa
ll
FIGURE 1.14
Stress-related bench stope brow failure following ring blasting.
In addition, stope wall failures due to stress changes of a tensional nature
can also be experienced (Bywater et al., 1983). Stope extraction in a destressed
orebody may lead to normal stresses of very low magnitude across some of
the exposed walls. Buckling-type failures may occur, depending upon the frequency of discontinuities parallel to a stope wall, the size and frequency of any
cross discontinuities, and the size and shape of the exposed spans (Figure 1.15).
1.4.4 Ground Support
Reinforcement by cablebolting provided at selected locations, usually constrained by the distance between drilling sublevels, can be used to reduce
the deformations experienced at the stope boundaries (crowns, walls, and rib
pillars). Stope walls are pre-reinforced prior to any stope firings and, in most
cases, cablebolts are installed from rings drilled within the stope access drives.
Thus, stope wall reinforcement tends to be localized in continuous bands that
are separated by the distance between the sublevel intervals. The function of
such an arrangement is to divide the stope walls into a number of stable stope
wall spans as well as arresting up-dip hangingwall failures (Figure 1.16).
Support from fill can also be used to minimize the deformations experienced by the stope walls while providing a restraint to any adjacent rock
masses. In general, cemented fill is needed to recover ore from secondary
stopes where stable fill exposures are required to minimize dilution.
Cemented fill is essential in chequerboard extraction patterns within
14
Geotechnical Design for Sublevel Open Stoping
FIGURE 1.15
A large-scale, structurally controlled, stope hangingwall failure. (From Villaescusa, E., A
review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane,
Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne,
Victoria, Australia. With permission.)
Cablebolt reinforcement
Hangingwall
failure
Cablebolt reinforcement
Hangingwall
failure
FIGURE 1.16
A large stope hangingwall failure arrested by a row of cablebolts installed prior to stope firings. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the
MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590,
AusIMM, Melbourne, Victoria, Australia. With permission.)
15
Introduction
Filling
Bench limit
Production blasting
Critical strike length
Mucking
Ore
Fill support
Previous bench filled
FIGURE 1.17
Continuous extraction and filling operations in bench stoping.
massive orebodies (Bloss, 1992), while uncemented fill is normally used in
conjunction with bench stoping operations (Villaescusa and Kuganathan,
1998). An example of a bench stoping extraction strategy linked to fill is
shown in Figure 1.17. Here, the exposed wall length is usually limited to
a critical value, defined by the distance between the fill and the advancing
bench brow (Villaescusa et al., 1994).
1.4.5 Blast Damage
Blast damage to a blasted rock mass refers to any strength deterioration of
the remaining rock due to the presence of blast-induced cracks and to the
opening, shearing, and extension of a preexisting or newly generated planes
of weakness (Figure 1.18). It is generally accepted that the damage is caused
by expanding gases through the geological discontinuities and the vibrations experienced from the blasting process. However, it is not easy to establish the approximate contribution to damage caused by the expanding gases,
as it is difficult to measure their path within a rock mass discontinuity network. Nevertheless, significant backbreak may be regularly observed when
the explosive gases are well confined within a volume of rock, and in some
cases the gases can travel well beyond the location of the explosive charges.
Damage by the shock energy from an explosive charge close to a blast can
be related to the level of vibrations measured around the blasted volume.
Repetitive blastings also impose a dynamic loading to the exposed stope
walls away from a blasted volume, and may trigger structurally controlled
falloff and ultimately overbreak. Conventional blast monitoring and simple
geophysical techniques can be used to measure the effects of blasting in
the near field. Vibrations and frequency levels from the shock wave can be
measured reasonably accurately (Fleetwood, 2010). These data can be related
to damage provided the contribution (to the total damage) from the shock
energy can be estimated. Vibration and frequency levels at the mid-spans of
16
Burden
Open stope void
Hangingwall
Geotechnical Design for Sublevel Open Stoping
Stope brow
Blasthole
FIGURE 1.18
Structurally controlled damage around a hole in an open stope brow. (From Villaescusa, E.,
A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane,
Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne,
Victoria, Australia. With permission.)
instrumented stope walls can be used to characterize the dynamic response
to blasting at the stope boundaries (Villaescusa and Neindorf, 2000).
1.4.6 Drill Drive Layout
Additional factors such as poorly located or preexisting drives, which
undercut the stope walls, also contribute to dilution or falloff at the stope
boundaries. In general, the number and location of drilling drifts in open
stoping are usually functions of the width of the orebody. In wide orebodies,
hangingwall and footwall drill drives are used to provide cablebolt reinforcement and to minimize the impact of blasting at the stope boundaries
(Figure 1.19). In such cases, drilling and blasting can be carried out in a plane
parallel to the final stope walls or to any exposed backfill masses. Suitable
values of standoff distance for the perimeter holes parallel to a stope boundary can be determined depending upon the rock type and the hole size being
used (Villaescusa et al., 1994).
Excessive wall damage, dilution, and ore loss may be experienced in cases
where stoping requires drilling holes at an angle to a planned fill exposure
or a stope boundary. Furthermore, hole deviation at the toes may create
an uneven stope surface, thereby preventing effective rilling of the broken
Introduction
17
FIGURE 1.19
Twin drill drives allowing drilling parallel to the stope boundaries. (Courtesy of Mount Isa
Mines, Mount Isa, Queensland, Australia.)
material to the stope drawpoints. In addition, hole deviation may cause
excessive confinement at the hole toes, thus causing breakage beyond the
orebody boundaries.
1.5 Scope and Contents of This Book
Sublevel open stoping—including variants, such as bench stoping—is one of
the most widely used mining methods in underground mining. Improvements
in technology over the last 30 years or so have seen increases in sublevel
spacing due to advances in the drilling of longer and accurate production
holes, as well as advances in explosive types, charges, and initiation systems.
Improvements in slot rising either through vertical crater retreat, inverse drop
raise or raise boring have also been experienced. Increases in sublevel spacing
have meant larger unsupported stope walls that must stand without collapsing.
Consequently, an understanding of rock mass characterization is required to
minimize dilution and increase recovery. Methodologies to design optimum
open spans, pillars, rock reinforcement, and fill are required. Furthermore,
in the same period, a greater understanding has developed regarding the
sequencing of stoping blocks to minimize in situ stress concentrations. In
the future, sublevel stoping is likely to be practiced at ever-­increasing depths
(Thomson and Villaescusa, 2011) and a better understanding of all the variables required to optimize the method is required.
18
Geotechnical Design for Sublevel Open Stoping
This book will cover the topic in nine chapters, as follows:
1.
2.
3.
4.
5.
6.
7.
8.
9.
Introduction
Sublevel Stoping Geometry
Planning and Design
Rock Mass Characterization
Span and Pillar Design
Drilling and Blasting
Rock Reinforcement and Support
Mine Fill
Dilution Control
The chapter topics are presented according to the conventional sublevel
stoping process used by most mining houses, in which a sublevel stoping
geometry is chosen for a particular mining method, equipment availability,
and work force experience. Planning of access infrastructure and overall
extraction sequences takes into account rock mass characterization information, which is first collected from the orebody delineation process. Detailed
planning of stope span and pillars is followed by access development,
where production drilling and blasting take place. Ground support becomes
an important aspect to provide safe personnel and equipment access to a
limited number of areas where open stoping activities take place. Following
extraction, a number of strategies are available to fill resulting open stope
voids, in which a reconciliation of dilution control and ore loss is critical to
achieve the most economical extraction of ore.
The book has been written primarily for fourth-year undergraduate students who are not yet familiar with the mining method. The book presents
the state of the art and also results from the applied research at the Western
Australian School of Mines (WASM) and, hence, the book could also be used
for postgraduate student research. Furthermore, some mining practitioners
and junior consulting engineers may find the book useful.
2
Sublevel Stoping Geometry
2.1 Introduction
In sublevel stoping, ore is broken by drilling and blasting. Stope access is
achieved by mining drilling and extraction drives, which can be accessed
either transversally or longitudinally with respect to the orebody strike.
The first stage is to create a slot between the vertical horizons defining the
planned stope. This is achieved by enlarging a suitably located raise or longhole winze (LHW). The slot is created as an expansion void into which the
remainder of the stope is formed by the sequential blasting of production
holes. In most cases, the production holes are drilled in rings parallel to
the orebody dip between the drilling drives. Mining proceeds through
the sequential firing of production rings into the advancing void with the
broken ore being recovered from a specific extraction horizon (Figure 2.1).
The following section describes the stoping geometries required to achieve
production from sublevel open stoping.
2.2 Stoping Geometries
2.2.1 Cutoff Slot
Sublevel open stopes are created by the sequential blasting of production
rings into an initial expansion slot, called the cutoff slot. This initial opening is used to create sufficient void for the remaining portion of the stope to
break into (Figure 2.2). The cutoff slot is usually located on a side or in the
center of a stope either transversally (across) or longitudinally with respect
to the strike of the orebody. An important point relates to whether the cutoff blasting will expose a critical stope wall, such as a hangingwall or a
fill mass, very early in a stope blasting sequence. The cutoff slot raises are
blasted upward from sublevel to sublevel in order to expose the full stope
height. At each level, the expansion slots are formed by sequentially blasting
parallel holes into an LHW or a raise-bored hole. The slot must be expanded
19
20
Geotechnical Design for Sublevel Open Stoping
FIGURE 2.1
Remote-controlled production mucking in sublevel open stoping. (Courtesy of Mount Isa
Mines, Mount Isa, Queensland, Australia.)
Cutoff
slot
Drill
drives
tion
duc
Pro l rings
dril
Cutoff
slot
h
Troug
cut
under
Dr
Dr
aw
aw
poi
nts
po
int
s
FIGURE 2.2
A three-dimensional view of a cutoff slot. (Courtesy of WMC Resources, Kalgoorlie, Western
Australia, Australia.)
21
Sublevel Stoping Geometry
to the full width of the plane defined by the production holes that will be
subsequently blasted into this initial opening.
High powder factors are normally used during slot blasting in order to
ensure breakage and thus have a free face and a void available into which the
remainder of the stope is to be blasted. The choice of slot location depends
upon rock mass conditions, stope access, and the extraction sequence chosen. In a steeply dipping orebody, where the critical stope boundary is usually an inclined hangingwall, transversally oriented slots are used to ensure
sequential hangingwall exposure by the production rings. In large, massive
orebodies, the choice of slot orientation is also controlled by factors such as fill
exposures, stress regime, and preestablished access (Bloss and Morland, 1995).
In general, a slot must be designed so that failure within the main or production rings is minimized. In highly stressed pillars, a slot can be oriented
normally to the major principal stress to shadow the main production holes.
This is likely to minimize hole squeezing or dislocation due to stress-related
damage. In cases where a stope access can be redesigned, the slot should be
placed normal to any large-scale geological features likely to fail and damage the main ring geometries (Figure 2.3).
Damage to fill masses from cutoff slot blasting can be minimized by placing a cleaner ring between a cutoff and a fill boundary (Figure 2.4). The rock
mass adjacent to a fill mass is usually preconditioned by stress redistributions and is likely to fail following a cleaner ring blasting.
In order to minimize hangingwall failures, cutoff slots are oriented transversally to the orebody strike. This allows the hangingwall plane to be
sequentially exposed within a predetermined stable range. In secondary
stope extractions, where longitudinal cutoff slots may be located parallel
°
P4
1
5°
Fa
Potential falloff
within rings
ul
t
°
P4
1
5°
6
65
Fa
ul
t
Cutoff slot
/W
/W
1F
1F
P4
P4
t-off slotslot
Cutoff
6
65
a
spl
y
ay
spl
(a)
(b)
FIGURE 2.3
Exposure of weak geological features by a cutoff slot. (a) Poor (preliminary) design and
(b) improved (actual) design. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)
22
Geotechnical Design for Sublevel Open Stoping
Cleaner Production
ring
rings
Blasting sequence
1. Holes near raise
Extracted
(filled)
2. Mid cutoff
Cutoff
slot
3. Complete cutoff +
cleaner ring
Plan view
FIGURE 2.4
Cleaner ring geometry to minimize fill damage from blasting. (From Villaescusa, E., A review
of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland,
Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria,
Australia. With permission.)
(and adjacent) to a stope hangingwall, the expansion slot exposes a full hangingwall plane early in a blasting sequence. This usually limits the size of
exposures that can be safely excavated, as this critical wall of the stope may
fail when subjected to repetitive dynamic loading by the rest of the stope
firings as shown conceptually in Figure 2.5. In addition, when the stopes are
accessed centrally, drill design requires that the holes toe into any adjacent
fill masses, thereby increasing the likelihood of fill dilution.
2.2.2 Production Rings
A design stope shape is achieved by sequentially blasting rings of blastholes
into the opening created by the initial expansion or cutoff slot. Stopes are
usually sequentially sliced up, from sublevel to sublevel, firing rings toward
the open cutoff slot. The production rings are sequentially blasted, attempting to minimize undercutting of the internal solid portion of a stope. An
approximately straight face is kept along the entire stope height by firing
a similar number of rings at each sublevel. The firing sequence advances
upward as shown in Figure 2.6. Maintaining a straight retreating face minimizes the creation of large brows or corners, which can be highly stressed
or intersect large-scale structures, thereby contributing to stope falloff. This,
in turn, can severely affect productivity during the subsequent production
mucking operations.
2.2.3 Diaphragm Rings
Diaphragm rings consist of rings drilled parallel to a fill exposure. The purposes of a diaphragm ring are to prevent fill failure from a known weak
cemented fill mass, to contain uncemented fill in adjacent stopes, and to
prevent fill failure from exposures of greater dimension than is considered
23
Sublevel Stoping Geometry
Filled
Filled
Slot
Stope hangingwall plane
Repetitive production ring blasting
FIGURE 2.5
Dynamic loading of a fully exposed hangingwall plane. (From Villaescusa, E., A review of
sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland,
Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria,
Australia. With permission.)
stable. Experience has shown that although parts of a diaphragm against fill
do fall off, this rarely results in excessive fill dilution, as the fill mass remains
comparatively undisturbed, compared to when blasting takes place next to
the fill (Figure 2.7). A diaphragm is not capable of load-bearing capacity and
so is likely to deform considerably. However, when a large portion of the
diaphragm remains intact, this enables clean stope extraction until the diaphragm is either fired or the stope is completed.
2.2.4 Trough Undercut
The lower portion of a stope is shaped using trough undercut (TUC) rings
in order to facilitate the draw of fragmented ore to and from the stope drawpoints. A TUC ring consists of parallel upholes, drilled inclined toward the
cutoff slot. Usually the toes of the TUC ring interlock with the toes of the
main ring downholes from the sublevel above (see Figure 2.8). Drilling and
blasting of the TUCs is usually carried out using relatively small diameter
holes (70–89 mm) compared to production holes. An improved explosive
distribution likely to minimize rock mass damage around the stope drawpoints is achieved by using such small diameter holes. A disadvantage is
24
Geotechnical Design for Sublevel Open Stoping
Cutoff slot
Cutoff slot
Drill and blast
access
Drill and blast
access
Stope
undercut
fall-off
potential
Broken
ore
Mucking
Mucking
Mucking
FIGURE 2.6
A longitudinal section view showing two production blasting strategies. (From Villaescusa, E.,
A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane,
Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne,
Victoria, Australia. With permission.)
140 mm blasthole
Fill mass
·
· ·
· ·
·
·
· ·
· ·
Rock diaphragm
Remainder
of
stope
·
·
extracted
· ·
· ·
·
·
·
· ·
2m
from
fill
Fillmass
3 m burden on
diaphragm ring
FIGURE 2.7
Idealized sketch and photo showing a stope diaphragm ring. (From Villaescusa, E., A review
of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland,
Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria,
Australia. With permission.)
25
Sublevel Stoping Geometry
14
13
10
7
4
3
14
12
9
6
2
11
8
5
1
Longitudinal view
FIGURE 2.8
Firing sequence of a TUC with production rings in an open stope. (From Villaescusa, E.,
A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane,
Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne,
Victoria, Australia. With permission.)
the limited drilling length achieved, and the inability to match the burden
drilled for the production ring holes immediately above.
Because the TUC rings are drilled with a different burden to the production rings, the lower portion of a stope is usually blasted ahead of the main
rings. This leads to a moderate undercutting of the main rings, which can
lead to falloff, especially in cases where large geological discontinuities are
present or in regions of high stress redistribution.
26
Geotechnical Design for Sublevel Open Stoping
Drawpoints
S
P
Orebody
S
hangingw
P
all
FIGURE 2.9
A fixed, transverse drawpoint geometry in sublevel stoping. P, primary stope; S, secondary
stope. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the
MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590,
AusIMM, Melbourne, Victoria, Australia. With permission.)
2.2.5 Drawpoints
Production mucking can be carried out longitudinally or transversely across
the strike of an orebody. Transverse mucking requires the introduction of
fixed and specialized drawpoint geometries that may be located outside an
orebody boundary (Figure 2.9). The factors considered during drawpoint
design include size of equipment, tramming distance from access drives,
and gradient and orientation with respect to a stope boundary. The drawpoint dimensions must be sufficient to suit the equipment, but kept as small
as possible to minimize instability. Drawpoint access should be straight and
restricted to 15–20 m from a stope access drive to the stope brow. This will
ensure that auxiliary ventilation will not be required while mucking, and
also that the rear of the mucking unit is inside the drawpoint. Drawpoint
spacing is determined by ground conditions and stope geometry. In most
cases, the minimum spacing used is 10–15 m between center lines.
2.3 Multiple-Lift Open Stoping
Multiple-lift stopes extend vertically over a number of sublevel intervals,
in some cases exceeding hundreds of meters in vertical extension. The
method requires sequential blasting of the production rings into an initial
27
Sublevel Stoping Geometry
vertical opening formed by a cutoff slot. Ore breakage is achieved by rings
of parallel or fanned blastholes, depending upon the type of drilling access
used. TUCs are developed at the base of the stopes in order to direct the
broken ore into the drawpoints for extraction. Cablebolt reinforcement of
hangingwall and stope crowns can be provided from suitably located drilling drives.
The number of drawpoints is usually a function of the stope size, but in
most cases at least two drawpoints are designed. Because the drawpoint
location is fixed, permanent reinforcement can be achieved at minimum cost
per unit of ore extracted. Access to the stope on each of the other sublevel
locations is required for drilling, blasting, and filling purposes (Figure 2.10).
Usually, a single crosscut access is required on each sublevel, significantly
decreasing development in waste.
In general, multiple-lift stopes minimize back cablebolting within the
intermediate sublevels because a permanent back (full area) is only exposed
at the actual crown of a stope. Cablebolting coverage at a stope crown is
a function of the degree of development within the top sublevel. In addition, the requirements for permanent reinforcement within any intermediate
1. Development
2. Cablebolt drilling
3. Production drilling
4. Production blasting
5. Production mucking
FIGURE 2.10
Sequence of mining activities within a multiple-lift sublevel open stope at the Kanowna
Belle Mine.
28
Geotechnical Design for Sublevel Open Stoping
sublevel are minimized by the fact that all the back exposures within the
drill drives are consumed by the stoping process itself.
Conventional multiple sublevel stoping requires the sequential exposure of high vertical, short horizontal stope walls likely to remain stable
and provide undiluted ore. The strike lengths exposed during the initial
stope extraction are unlikely to exceed the critical stable stope spans. As
the excavations are enlarged and several rings are sequentially blasted into
the void formed by the cutoff and the initial production rings, confining
stresses are reduced, excess strain energy is induced, and displacement of
the stope walls is experienced. Depending on the structural nature of the
exposed walls, the rock may tend to displace following a sheetlike behavior,
in which a group of layers move together (in bedded rock), or the movement
may be isolated to individual blocks that partially rotate and slide against
each other.
2.3.1 Tabular Orebodies
The layout for multiple-lift sublevel stopes in tabular orebodies is usually
associated with the use of long blastholes drilled from drives parallel to the
orebody strike. Depending upon the orebody width, these drill drives may
be either of full orebody width or located at the boundaries of the orebodies. In such orebodies, the stope boundaries are usually well defined by the
orebody itself. Crown, hangingwall, footwall, endwalls, and a drawpoint
can be defined for each stope. The stability of stope crowns and hangingwalls is usually the most critical factor in the stope design and related extraction sequences. A conventional design usually consists of multiple drilling
sublevels with a single mucking horizon at the bottom of the stope as shown
in Figure 2.11.
One of the advantages of this design is that drilling and blasting can be
done in a plane parallel to the final stope walls. Hangingwall and footwall
drill drives are used to minimize the impact of blasting at the stope boundaries, greatly decreasing the likelihood of dilution due to blast damage. In
addition, the method reduces stope development in waste, given that, except
for the mucking horizon, a single stope drilling access is actually required at
each sublevel location.
In cases where sublevel stoping is used to extract large but tabular orebodies
having a moderately dipping hangingwall, the extraction can be divided into
a number of primary, secondary, and sometimes tertiary stopes, which can
be extracted in a checkerboard sequence. In order to optimize stope stability,
the stope walls are designed vertically, except for the hangingwall as shown
in Figure 2.12. Drilling drives parallel to the hangingwall can be used to
provide cablebolt reinforcement, and facilitate drilling and blasting parallel
to the hangingwall planes. Stope crown stability can be optimized with the
implementation of a floating sublevel to optimize cablebolt reinforcement.
The use of conventional drawpoint geometries increases productivity.
29
Sublevel Stoping Geometry
Production H/W drive Cleaner ring
rings
Extracted
(filled)
Trough undercut
Cablebolted
area
Cablebolted
area
Drawpoint 1
Cutoff slot
Cutoff slot
Hangingwall
Drawpoint 2
Cross cut
access
F/W drive
F/W access drive
F/W access drive
(a)
Extracted
(filled)
(b)
Crown reinforcement
Cutoff slot
H/W
rein
forc
eme
n
t
Crown reinforcement
Production
rings
Cutoff slot
F/W drive
Trough undercut
(c)
(d)
FIGURE 2.11
Sublevel stoping in a steeply dipping tabular orebody. (a) Plan view—mucking horizon, (b)
plan view—intermediate level, (c) cross section view—production rings, and (d) long section
view. (From Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the
MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590,
AusIMM, Melbourne, Victoria, Australia. With permission.)
2.3.2 Massive Orebodies
Open stoping in large, massive orebodies consists of a mining sequence
that requires several stages of stoping in conjunction with the application of
delayed fill methods to enable pillar recovery. Usually, a number of stopes
are designed between the orebody boundaries. In such cases, stoping comprises a number of stages that includes primary, secondary, and tertiary
stopes that are usually extracted using a checkerboard sequence (Alexander
and Fabjanczyk, 1981). The number of fill exposures ranges from none (in a
primary stope) up to three exposures in the late stages of stoping (Grant and
DeKruijff, 2000).
30
Geotechnical Design for Sublevel Open Stoping
Ha
ng
ing
wa
ll r
e
inf
o
rce
me
nt
Floating sublevel for crown reinforcement
Footwall access
drive
Drawpoint
FIGURE 2.12
Stope design for a large tabular orebody. (Courtesy of Mount Isa Mines, Mount Isa, Queensland,
Australia.)
Large vertical dimensions can be designed with the height of the stopes usually constrained by the orebody thickness or by the stability of any exposed
fill masses required for secondary and tertiary stope extraction. Stope dimensions in a plan view are usually constrained by stope crown instability. The
broken ore is extracted in the bottom part of the stope (Figure 2.13).
In cases where the ground conditions are favorable, stope dimensions can
be very large in plan, with full orebody height extraction being achieved in
a single stope (Bloss, 1996). Drilling and blasting is carried out from a series
of sublevel locations ranging from 40 to 60 m apart. Blastholes are mainly
drilled downward, with some short upholes drilled within the TUCs and
sometimes at the stope crown when a top access is not available.
Following pillar extraction (secondary and tertiary stopes), a number of
fill exposures are created depending upon the location of the stope in the
mining sequence. Early on in the life of a massive orebody, primary stopes
usually account for a significant part of the production. As orebody extraction increases, the shift to pillar mining as the primary method of extraction
31
Sublevel Stoping Geometry
7
1
Extraction sequence
5
4
2
8
3
1
7
6
(a)
(b)
FIGURE 2.13
Multiple-lift stoping in a massive orebody. (a) Plan view and (b) three-dimensional view. (From
Villaescusa, E., A review of sublevel stoping, in G. Chitombo, ed., Proceedings of the MassMin
2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 577–590, AusIMM,
Melbourne, Victoria, Australia. With permission.)
becomes evident. In such cases, the stability of the fill exposures is of primary importance in achieving the target production figures (Bloss and
Morland, 1995).
In cases where the upper orebody boundary does not coincide with the predetermined location of the upper sublevel interval, drilling into or through
the orebody crown may be required. If the top of the orebody is above the
highest sublevel interval location, upholes may be drilled into the stope
crown in order to define a designed stope shape. In cases where the highest
sublevel is located above the orebody boundary, downholes may be drilled
through the orebody crown, with the lowest portion of the holes blasted to
define a stope shape. In both cases, the stope crown remains unsupported,
and a preferred alternative is to develop a “floating” sublevel through the top
of the stopes to facilitate deep cablebolt reinforcement and drilling of holes
parallel to the designed stope crown (Figure 2.14).
2.4 Single-Lift Stoping
A single-lift design is the most basic arrangement for sublevel open stope
extraction. The stope shape and size is constrained by two sublevels:
the extraction or undercut horizon, and the drilling or overcut horizon.
32
Geotechnical Design for Sublevel Open Stoping
le
rp
Pu
fa
ul
t
Gr
f
ay
lt
au
24A
2200 RL
S613
lt
au
df
Re
25A
2150 RL
26D
26B
FIGURE 2.14
Drilling and blasting strategies for a stope crown. (Courtesy of Mount Isa Mines, Mount Isa,
Queensland, Australia.)
Access to the stopes is via crosscuts off a permanent access drive parallel
to the orebody. Effectively, this method requires a “moving” drawpoint
system, as the stoping extraction progresses upward. Following the filling
of a stope void, a previous drilling horizon becomes the next extraction
level (Figure 2.15).
Development
Cablebolting/drilling
Blasting
Filling
Drawpoints
Drawpoints
Drawpoints
FIGURE 2.15
Three-dimensional view of single-lift sublevel stoping. (From Potvin, Y. et al., CIM Bull.,
82(926), 53, 1989.)
33
Sublevel Stoping Geometry
West
East
Stoping block 3
Block 3–4 sill pillar
Cave zone
Stoping block 4
Cave
zone
FIGURE 2.16
Longitudinal section view of the Williams Mine B zone. (From Bawden, W. F. et al., Lessons
in control of mine costs from instrumented cable bolt support. In J. Girard, M. Leibman, C.
Breeds, and T. Doe (eds.), Proceedings of the Fourth North American Rock Mechanics Symposium,
Seattle, WA, 31 July–3 August, A.A. Balkema, Rotterdam, the Netherlands, 2000, pp. 633–642.)
In order to optimize mucking productivity, up to two access crosscuts
per stope may be required at each sublevel interval. This actually increases
the overall access development in waste to actual stoping ratio. The method
requires very good control of the stope back and brow stability, especially
in a highly stressed environment. Stress redistribution due to the stoping
sequence itself can create significant back failures, especially if shallow dipping discontinuities are present within a rock mass. Figure 2.16 shows a typical extraction configuration using single-lift stopes at the Williams Mine in
Canada, where several major rockfall occurrences within the sill pillar have
been reported by Bawden et al. (2000). The rockfalls delayed the mining of
approximately 1 million tons containing some 300,000 oz, seriously affecting
production from the mine.
Extended backs, pillars, and highly stressed brows are likely to be formed
somewhere within the stoping sequence, and full cablebolting coverage of
the stope backs is required to minimize the potential failures at each sublevel location. Full cablebolting coverage requires stripping the orebody
access to the full stope width, thereby minimizing the sizes of stopes that
can be developed safely. As a result, single-lift stopes tend to be relatively
small openings compared with multiple-lift stopes.
34
Geotechnical Design for Sublevel Open Stoping
Pendant pillar
Filled
Filled
Pendant pillar
FIGURE 2.17
Idealized stoping sequence for single stopes on a 1-4-7 extraction sequence. (From Potvin, Y.
et al., CIM Bull., 82(926), 53, 1989.)
Primary development requires the extension of the access crosscut to a
proposed hangingwall location, where both the drill and the extraction sublevels are completely silled out to allow the installation of cablebolt reinforcement. In addition, the drilling of parallel blastholes is also facilitated with
full stope undercut and overcut geometries. Drilling of parallel holes is the
preferred way in vertical retreat stoping, which is linked to single-lift stoping. The method requires a significant amount of remote mucking due to the
flat-bottom nature of the single-lift stope geometries, thereby increasing the
overall mining cost compared to a conventional TUC drawpoint geometry.
In wide orebodies, a number of stopes may be designed across the strike
in a given area, and in all cases, adjacent primary stopes are extracted to a
level above that of a secondary stope. This type of sequence creates what is
called a pendant pillar. A pendant pillar is a solid piece of ground that has
many degrees of freedom for movement, as most stopes around it have been
extracted (Figure 2.17). Large pillar failures may be experienced in such stoping geometries (Milne and Gendron, 1990).
2.4.1 Conventional Vertical Crater Retreat Stoping
Vertical crater retreat (VCR) is a single-lift stoping method where the
stope’s shape is defined by a lower (undercut) and upper (overcut) horizon
35
Sublevel Stoping Geometry
Full
overcut
Filled
FIGURE 2.18
VCR mining within a single-lift stope. (From Villaescusa, E., A review of sublevel stoping,
in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia,
October 29 to November 2, 2000, pp. 577–590, AusIMM, Melbourne, Victoria, Australia. With
permission.)
(Trotter, 1991). Large-diameter holes are drilled in order to minimize deviation, and the holes are charged from the overcut and blasted by means of
horizontal slices of ore progressing from the bottom level to the top level
(Figure 2.18). The separation between the undercut and overcut is a function
of stope wall stability, the nature of the orebody, and drilling accuracy.
Following blasting, only a slight amount of broken ore is mucked, so that
enough room is available for a subsequent blast to break into. This keeps the
stope full of broken rock, thereby providing passive support to the exposed
stope walls until blasting to the stope overcut is complete. Once blasting is
completed and all the ore within the stope is mucked, the undercut accesses
are closed off and the stope is filled. As mining progresses upward, the stope
overcut becomes the next mucking horizon in the sequence.
The method has a number of perceived advantages including the requirement for few large-diameter blastholes, likely to reduce the overall in-thestope drilling. Large holes enable a larger sublevel interval to be used, thus
reducing the overall sublevel development cost. The cost of raising and slashing to create a slot is eliminated, and all the drilling and loading operations
are carried out from the overcut, thereby increasing safety.
36
Geotechnical Design for Sublevel Open Stoping
The disadvantage of this method is the potential for blast damage from
crater blasting at the stope boundaries (Platford et al., 1989). Small-diameter
holes cannot be used due to hole closure caused by ground movement following the individual stope blasts (Hills and Gearing, 1993). In addition, this
method may be susceptible to poor fragmentation (falloff) from the unsupported areas defined by blasting, especially if an uneven back is formed and
high stresses are subsequently redistributed upward. Blast damage from cratering is even more detrimental when shallowly dipping geological discontinuities are present within a rock mass.
2.4.2 Modified Vertical Retreat Stoping
A modified vertical retreat method uses a winze or a raise-bored hole, which
is located near the middle of the stope, into which a radiating pattern of
blastholes is sequentially fired in horizontal lifts. The raise is used to overcome the limited free face available in a conventional vertical retreat stope.
In order to facilitate the initial blasting, the method requires close spacing of the holes near the raise (Figure 2.19). All the holes in a horizontal lift
are fired, and the possibility of collar damage exists when the inner holes
near the raise do not perform. In addition, hole damage (closure, requiring
Parallel rings of vertical blastholes
to be fired in a radiating pattern
from the raise
Plan view
Raise bored hole
1.1 m diameter
Position after one ring firing
Section view
FIGURE 2.19
Typical blast layout for a modified vertical retreat stope in the Porgera Mine. (From Hills,
P.B. and Gearing, W.G., Gold ore mining by the Porgera Joint Venture at Porgera, Papua New
Guinea, in J.T. Woodcock and J.K. Hamilton, eds., Processing Australasian Mining and Metallurgy,
AusIMM, Melbourne, Victoria, Australia, 1993, Chapter 12: Gold, pp. 897–902.)
37
Sublevel Stoping Geometry
redrilling) within the last lift in the stope may be continuously experienced
with this method (Hills and Gearing, 1993). On the other hand, the method
is considered to be relatively safe because no vertical opening is made within
the stope until the last firing.
2.5 Shallow Dipping Tabular Orebodies
Tabular orebodies in which the dip angle does not allow the flow of broken ore
utilizing gravity can be extracted using a type of sublevel stoping called uphole
retreat panel stoping (Kaesehagen and Boffey, 1998). Typically, an orebody
can be divided into panels, running parallel to the strike of the orebody and
defined down-dip as shown in Figure 2.20. The stopes are extracted by developing a footwall extraction drive from which drilling, blasting, and mucking
operations can be carried out. The stopes are accessed from a footwall drive,
with a slot established at the far end of the panels, and the stopes are progressively blasted retreating back to the access end of a panel (Figure 2.21).
Cablebolt reinforcement is provided from the hangingwall drives located
within the primary stopes. In addition, permanent pillars can be left within
the secondary stopes to provide additional hangingwall support.
Flat lying orebodies can also be extracted by individual stopes in conjunction with cablebolting drives and mine fill operations. The stopes are
extracted by developing a TUC horizon in waste to allow the flow of ore
Hangingwall cablebolting
Production
up-holes
Panel slot
(drilled
downhole)
Decline
Panel access drive
Longitudinal view
p
S
p
S
S
Section view
FIGURE 2.20
Typical panel stoping layout. P, primary stope; S, secondary stope. (From Kaesehagen, M.R. and
Boffey, R.H., Development of the Osborne Mine—with a focus on technical and operational
aspects, Proceedings of the Seventh Underground Operators’ Conference, Townsville, Queensland,
Australia, June 30 to July 3, 1998, pp. 29–37, AusIMM, Melbourne, Victoria, Australia. With
permission.)
38
Geotechnical Design for Sublevel Open Stoping
FIGURE 2.21
An unsupported uphole panel stope following extraction.
to the stope drawpoints. Downhole drilling is undertaken from a series of
hangingwall drives, from which cablebolt reinforcement is also provided
(Figure 2.22). This method results in an increased lead time in stope preparation as well as additional costs, as noneconomical material is developed.
The overall stope extraction retreats up-dip and toward the access end of
the drilling drives. Experience indicates that only half of the back of a previously extracted stope (down-dip) can be filled effectively. The methodology
consists of extracting stopes having either single or double drilling drives,
depending upon their location with respect to the orebody abutment and
with respect to each other in the extraction sequence. Alternating single and
double drilling drives is likely to optimize hangingwall reinforcement as the
extraction progresses up-dip.
2.6 Bench Stoping
Bench stoping is used to extract steeply dipping and relatively narrow (up
to 12–15 m wide) veins, lenses, lodes, or any stratiform deposit extending in
39
Sublevel Stoping Geometry
18
2
12
6
15
1
3
19
10
7
2
11
Plan view
20
17
8
13
Ca
16
4
9
5
11
ble
bo
ltin
g
7
3
14
1
Extraction sequence
Stope boundary
Filled stope
1
Section view
FIGURE 2.22
Overall extraction sequence and cross section showing cablebolt reinforcement. (From
Villaescusa, E., Extraction sequences in sublevel stoping, Proceedings of the 12th International
Symposium on Mine Planning & Equipment Selection, Kalgoorlie, Western Australia, Australia,
April 23–25, 2003, pp. 9–18, AusIMM, Melbourne, Victoria, Australia. With permission.)
two dimensions (along strike and down-dip). The method involves the initial mining of both a drilling and an extraction drive for the entire length
and width of the orebody (Figure 2.23). A slot is created between the two
horizons at one end of the orebody by enlarging a cutoff raise (or LHW)
located near the footwall of the orebody. The slot created is used as an
expansion void into which the remainder of the bench stope is formed by
the sequential blasting of production holes. In most cases, the production
holes are drilled in rings parallel to the orebody dip between the two drives
(Figure 2.24).
Stoping proceeds through the sequential firing of downhole (or uphole)
blasthole rings into the advancing void, and ore is then remotely mucked
along the orebody from the extraction horizon (Figure 2.25). The likelihood of ore dilution increases if the ore is left within the stope floor for
long periods of time and wall failures may cause ore loss or damage the
mucking units.
The success of bench stoping relies on the stability of the exposed unsupported spans, the ability to provide support with cablebolting and fill, tight
control on drilling and blasting, as well as the application of remote mucking
technology (Villaescusa et al., 1994). Downhole bench stoping geometries are
linked to up-dip overall extraction sequences in conjunction with fill. Uphole
bench stopes are often extracted without the use of fill, and retreating topdown in conjunction with permanent, nonrecoverable pillars.
In most mining operations, the bench heights are fixed during the early
stages of mine development, and the extraction strategy is the only variable
that can be used to optimize the economics of bench stoping. In downhole
benches, the extraction is followed by filling of the void with waste, hydraulic sand fill, or aggregate to the floor of the drilling drive, which becomes the
new extraction drive on the next lift up-dip. A number of extraction strategies
Ore
Auxiliary ventilation
Bench brow closed
Rockfill
Retreat
direction
Crosscut
(b)
Longitudinal
access
Access FW drive
Crosscut
Sto
pe
reat
Ret tion
c
dire
kfill
Roc
FIGURE 2.23
Details of bench stope extraction. (a) Longitudinal view and (b) 3D view. (Courtesy of Kanowna Belle Mines, Kalgoorlie, Western Australia, Australia.)
(a)
RAR to
primary
exhaust
40
Geotechnical Design for Sublevel Open Stoping
Sublevel Stoping Geometry
41
0.5 m
0.8 m
17D
6865 N
78°
83°
17B
0.5 m
72°
0.8 m
67°
17.5
m
17.5
m
17.6 m
17.9 m
2m
FIGURE 2.24
A typical cross-sectional view and the results of exceptional downhole bench stoping.
(Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)
FIGURE 2.25
Longitudinal remote mucking of broken ore in bench stoping. (Courtesy of Mount Isa Mines,
Mount Isa, Queensland, Australia.)
42
Geotechnical Design for Sublevel Open Stoping
FIGURE 2.26
Longitudinal ore extraction in conjunction with fill support. (Courtesy of Mount Isa Mines,
Mount Isa, Queensland, Australia.)
have been considered for downhole benching (Villaescusa and Kuganathan,
1998). The most common involves using a continuous dry fill mass (waste
rock having a rill angle between 38° and 42°) that follows an advancing
bench stope brow at a fixed distance (not exceeding a critical unsupported
strike length) along the entire bench length (Figure 2.26).
Benches can also be extracted using hydraulic fill, with the stopes
extended to a maximum stable unsupported strike length, followed by fill
in conjunction with brick bulkheads. Filling is followed by pillar recovery
and the process is repeated along the entire bench length (Figure 2.27).
Although this strategy is primarily linked to hydraulic fill, the use of
cemented fill would ensure that minimal fill dilution would be experienced
following pillar recovery. Cemented fill can only be justified during extraction of very high-grade orebodies. Recent applications of cemented paste
fill are replacing the use of hydraulic fill, thus minimizing the need for
brick bulkheads.
Another strategy is to leave (planned) permanent pillars between independent (unfilled) hangingwall spans along the entire bench length. Filling
is done on bench completion using either dry or hydraulic fill (Figure 2.28).
In this strategy, it is critical to establish the optimum distances between
the pillars in order to minimize the number of pillars required, especially
in high-grade orebodies. Pillar dimensions are a function of the ground
conditions, the expected stress levels, and the optimum extraction of the
adjacent LHWs. In weak rock masses, the stability of unfilled spans may
be affected by blasting in adjacent spans along the strike of the orebody,
as the individual spans may show time-dependent behavior with related
43
Sublevel Stoping Geometry
Temporary pillar (drilled)
New slot
Recovered pillar
Production
blasting
Hydraulic fill
Maximum
strike length
(void filled)
Mucking
Ore
Bench
limit
Barricades
Extracted and
filled
FIGURE 2.27
Hydraulic fill and pillar recovery. (From Villaescusa, E. and Kuganathan, K., Backfill for
bench stoping operations, in M.L. Bloss, ed., Minefill 98, Proceedings of the Sixth International
Symposium on Mining with Backfill, Brisbane, Queensland, Australia, April 14–16, 1998, pp. 179–
184, AusIMM, Melbourne, Victoria, Australia. With permission.)
Permanent pillar
New slot
Production
blasting
Permanent pillar
Maximum
unsupported
strike length
(void to be
filled at bench
completion)
Mucking
Ore
Extracted and
filled
Bench
limit
Permanent pillar
Extracted and
filled
FIGURE 2.28
Nonrecoverable permanent pillars in conjunction with fill. (From Villaescusa, E. and
Kuganathan, K., Backfill for bench stoping operations, in M.L. Bloss, ed., Minefill 98, Proceedings
of the Sixth International Symposium on Mining with Backfill, Brisbane, Queensland, Australia,
April 14–16, 1998, pp. 179–184, AusIMM, Melbourne, Victoria, Australia. With permission.)
44
Geotechnical Design for Sublevel Open Stoping
FIGURE 2.29
Unsupported spans and permanent pillars in shallow dipping bench stoping.
deformation. Figure 2.29 shows a top-down bench extraction strategy that
relies on a combination of unsupported spans and permanent pillars.
Bench stopes can be also extracted using a continuous and tight filling
technique called Avoca. Initially, the bench stope is extracted to a maximum
stable length, followed by tight filling to the brow. Any subsequent blasting
is then undertaken with no free face as shown in Figure 2.30. The success of
Filling
Filling
Production
blasting
(no free face)
Mucking
Continuous AVOCA fill
Bench
limit
Ore
Extracted
and filled
FIGURE 2.30
Full Avoca bench extraction method. (From Villaescusa, E. and Kuganathan, K., Backfill for
bench stoping operations, in M.L. Bloss, ed., Minefill 98, Proceedings of the Sixth International
Symposium on Mining with Backfill, Brisbane, Queensland, Australia, April 14–16, 1998, pp. 179–
184, AusIMM, Melbourne, Victoria, Australia. With permission.)
45
Sublevel Stoping Geometry
Slot
this method is a function of the fill stability following blasting. This is controlled by the orebody width and height and the moisture and particle size
distribution of the fill material used.
The option of extracting a bench beyond its stable limits and then leaving
a (unplanned) pillar to arrest a hangingwall failure has not been considered
because it does not represent good design or operational practice. The extraction option shown in Figure 2.27 is related to extracting the bench using
pillars that have been designed at the very early stages, and it is assumed
that the spans between the pillars are stable and independent (from a deformational point of view) of each other.
Uphole benches are often related to top-down sequences of extraction
where the orebodies are partitioned into blocks separated by horizontal
crown pillars. Individual uphole benches are defined within a block, and
retreated to a central or end access crosscut. Typical uphole drilling heights
range from 15 to 25 m, and the individual rings are inclined forward (70°) to
promote a safe brow for the blasthole charging crews. The design of forward
dumping rings also reduces muck throw, which in turn minimizes remote
mucking. Hangingwall reinforcement is provided from the drilling drives.
In addition, in good-quality rock masses, filling can be introduced following
the extraction of an entire stoping block (Figure 2.31).
Hydraulic fill
Crown pillar extraction
ble
bo
lti
ng
Crown pillar
in
gw
all
ca
985 L
Ha
ng
Hydraulic fill
Hydraulic fill
965 L
Rock fill
945 L
925 L
Crown pillar
Pillar extraction
Open void
890 L
Ramp access
Slot drilling
Broken ore
870 L
850 L
Pillar
Slot
Uphole drilling
Pillar extraction
Retreat to access
Open void
Broken ore
Uphole drilling
830 L
Crown pillar
785 L
Cross-sectional view
Central access
Level development
Longitudinal view
FIGURE 2.31
Schematic of uphole bench extraction sequences, Osborne Mine. (After Kaesehagen, M.R. and
Boffey, R.H., Development of the Osborne Mine—with a focus on technical and operational
aspects, Proceedings of the Seventh Underground Operators’ Conference, Townsville, Queensland,
Australia, June 30 to July 3, 1998, pp. 29–37, AusIMM, Melbourne, Victoria, Australia. With
permission.)
3
Planning and Design
3.1 Introduction
Mine planning is an engineering process that encompasses all of the major
technical functions undertaken in sublevel open stoping, with the key performance indicators being safety, dilution control, recovery, productivity,
and mining cost. Mine planning provides the means for the safe, efficient,
continuous, and economic recovery of ore while considering the life of mine
issues and their implications for short-term planning and design. It also
helps to maintain the long-term security of production, while ensuring satisfactory economic returns (Trout, 1997).
Mine planning prepares and evaluates all future stope design and
operating strategies. Parameters such as ore reserve estimation, overall
sequences of extraction, dimensioning of regional pillars and sublevel intervals, design of ore haulage systems, as well as fill and ventilation systems
are determined during the process (see Figure 3.1). Although it is beyond
the scope of this book to review such topics in detail, geotechnical aspects
of the process from orebody delineation to stope extraction are considered
within this chapter.
The approach suggested here requires interaction between geology,
mine planning, rock mechanics, and operating personnel throughout
the entire mine-planning process (Villaescusa, 1998). The overall rational
methodology for the stope planning process is shown in Table 3.1. The
orebody delineation and rock mass characterization stages constitute the
basic inputs. The requirements consist of an early determination of rock
mass properties on a block scale, followed by the selection of the mining
method and an estimate of the likely loading conditions from the stoping
sequences. The process requires both global and detailed design stages.
Global design issues are relevant and applicable within entire areas of a
mine, such as an extension of an existing orebody, while detailed design
issues are applicable to the extraction of individual stopes. Finally, a monitoring and back analysis strategy that allows a documented closure of the
design loop is required.
47
48
Geotechnical Design for Sublevel Open Stoping
Orebody
Orebodydelineation
delineation
Geology
Rockmass
characterization
Rockmass
characterization
Geology
and
Geology
& rock
rock
mechanics
Mining method selection
Mine planning
Access
and&infrastructure
Access
infrastructure
Rock
Rockmechanics
Mechanics
Global
Globalsequences
sequences
(stress
(Stressanalysis)
analysis)
Mine planning and rock mechanics
Global
Globaleconomics
Economics
Mine
Mine planning
planning
d
e
s
i
g
n
Acceptable
Acceptable
design
design
Yes
Infill
Infilldelineation
delineation drilling
drilling
D
e
t
a
i
l
e
d
Geology
Geology
Drill
blastdesign
design
Drilland
& blast
Mine
Mine planning
planning
Rock
Rockreinforcement
reinforcement
No
G
l
o
b
a
l
Mine
Mine planning
planning
Stope
and
pillarsizes
sizes
Stope
& pillar
No
Input
data
Rock
Rockmechanics
mechanics
Detailed
Detailedeconomics
economics
Mine
Mine planning
planning
Extraction monitoring
Operations, mine planning
geology and rock mechanics
Acceptable
Acceptable
design
design
Yes
Document
Document
results
results
End
d
e
s
i
g
n
Closure of
design loop
FIGURE 3.1
Flowchart of mine-planning process. (From Villaescusa, E., Geotechnical design for dilution
control in underground mining, in R.K. Singhal, ed., Proceedings of the Seventh International
Symposium on Mine Planning & Equipment Selection, Calgary, Alberta, Canada, October 5–9, 1998,
pp. 141–149, Balkema, Rotterdam, the Netherlands.)
TABLE 3.1
Key Stages within a Stope Planning and Design Process
Stope Design Process Stages
Basic Input
Orebody delineation
Rock mass characterization
Mining method selection
Control of Ground Behavior
Closure of the Design Loop
Stope block design
Detailed stope design
Monitoring
Back analysis
Documentation
Planning and Design
49
3.2 Geological and Geotechnical Characterization
The orebody delineation and rock mass characterization stages provide the
input for the entire stope design process. In most cases, however, the main
role of a mine geology department is limited to the definition and delineation
of the ore zones within a deposit, the geological interpretation for further
delineation and exploration strategies, and making ore reserve estimations.
Rock mass characterization is rarely undertaken by mine geology as a routine process, as the significant demands of a robust orebody delineation
leave no time for additional geotechnically related duties. Sometimes, a lack
of proper training and awareness of the relevant geotechnical issues by the
mine geologists also contributes to deficient data collection approaches.
The suggested approach is to obtain representative, mine-wide, rock mass
properties required during the subsequent global excavation design and stability analysis stages. This information is obtained from diamond drill holes
consisting of mainly core logging of nonoriented holes followed by direct
mapping of underground openings. Geophysical tools can also be used for
orebody delineation and rock mass characterization, but such techniques
have not been widely implemented to date.
The confidence in the geological information must be sufficient to establish
the nature and irregularities of the orebody, the nature and location of major
controlling geological structures, the general rock mass characteristics, as
well as allowing an economic evaluation to be carried out to determine
whether a particular stoping block should be mined. This type of information
requires that the sampling process extends beyond the orebody boundaries
in order to determine the likelihood of failure from orebody hangingwalls,
footwalls, or stope crowns.
The first step in any rock mass characterization process is a three-­dimensional
definition of rock-type contacts and alteration halos. In addition, large-scale
geological discontinuities such as faults and shears likely to play a major role
in the overall mechanical behavior of the entire deposit must be identified.
The second step in a rock mass characterization program is to determine the
rock mass behavior away from the main geological discontinuities by defining
structural domains for design. This can be achieved by core logging and direct
mapping of joint set characteristics such as number of joint sets, joint orientation, frequency, trace length, planarity, and surface strength (Brown, 1981).
3.3 Stress Analysis in Stope Design
As illustrated in Figure 3.1 and discussed in more detail subsequently
in this chapter and in Chapter 5, many of the steps in the overall stope
50
Geotechnical Design for Sublevel Open Stoping
block and detailed stope design processes require the use of some form
of stress analysis.
In modern mining practice, computational or numerical methods of stress
and deformation analysis are used to evaluate the stresses and deformations induced around excavation boundaries and within the surrounding
rock mass as a result of excavation. They find widespread use as aids to
decision making in establishing overall open stoping layouts and extraction sequences, and in the detailed design and dimensioning of components
of the overall mining structure, including items of infrastructure, accesses,
stopes, and pillars. The numerical methods that are most commonly used
in addressing mining rock mechanics problems may be classified as follows
(Jing, 2003):
Continuum methods
• Finite difference method (FDM)
• Finite element method (FEM)
• Boundary element method (BEM)
Discontinuum methods
• Discrete or distinct element method (DEM)
• Discrete fracture network (DFN) methods
Hybrid continuum/discontinuum models
•
•
•
•
Hybrid FEM/BEM
Hybrid BEM/DEM
Hybrid FEM/DEM
Other hybrid models
Figure 3.2 illustrates the two-dimensional discretization concepts used in the
FDM, FEM, BEM, and DEM for a fractured or discontinuous rock mass. As
shown in Figure 3.2, the modeling of faults in FDM, FEM, and BEM requires
the introduction of special joint or displacement discontinuity elements.
The discussion of the details of the numerical methods of stress and deformation analysis used in underground mining, including open stoping, is
beyond the scope of this book. Useful general introductions are given by
Brady and Brown (2004) and Jing (2003). The following overview of the main
numerical methods is based on that given by Brady and Brown (2004).
Computational methods of stress analysis may be divided into two
categories: differential methods and integral methods. In differential methods, the problem domain is divided or discretized into a set of subdomains
or elements as shown in Figure 3.2b. The solution procedure may be based
51
Planning and Design
Joints
Faults
Joint
element
(b)
(a)
Region 1
Region 4
Block
Region 2
Block
Region 3
(c)
Element of
displacement
discontinuity
(d)
Regularized
discontinuity
FIGURE 3.2
Two-dimensional representation of the fractured rock mass shown in (a) by (b) FDM or FEM,
(c) BEM, and (d) DEM. (Reprinted from Int. J. Rock Mech. Min. Sci., 40(3), Jing, L., A review of
techniques, advances and outstanding issues in numerical modelling for rock mechanics and
rock engineering, 283–353, Copyright 2003, with permission from Elsevier.)
on numerical approximations of the governing equations, that is, the differential equations of equilibrium, the strain–displacement relations, and the
stress–strain equations, as in FDM. Alternatively, the procedure may exploit
approximations to the connectivity of elements, and continuity of displacements and stresses between elements as in the FEM. The FEM can readily
accommodate nonlinear and heterogeneous material properties, but the
problem domain is defined arbitrarily, and discretization errors may occur
throughout the domain (Brady and Brown, 2004).
In integral methods, the problem is specified and solved in terms of surface values of the field variables of traction (surface stress components) and
displacement. As only the problem boundary is defined and discretized as
in Figure 3.2c, this BEM effectively provides a unit reduction in the dimensional order of the problem. This offers a significant advantage in terms of
computational efficiency, particularly in the solution of three-dimensional
problems. These methods are best suited to linear material behavior and
homogeneous material properties. However, they model far-field boundary
conditions correctly, restrict discretization errors to the problem boundary,
and ensure fully continuous variations of stress and displacement throughout the medium (Brady and Brown, 2004).
52
Geotechnical Design for Sublevel Open Stoping
The DEM represents the fractured rock mass as an assembly of blocks interacting through deformable discontinuities having definable stiffnesses. The
equations of motion of these blocks are solved through continuous detection
and treatment of the contacts between blocks. The blocks can be rigid or
made deformable using FDM or FEM discretizations. The method can model
large displacements caused by the rigid body motion of individual blocks
including block rotation, fracture opening, and complete block detachment (Jing, 2003). Detailed accounts of the fundamentals of discrete element
methods and of their application in rock engineering are given by Jing and
Stephansson (2007).
Examples of the use of numerical modeling in extraction sequencing,
assessing stope wall stability, pillar design, and ground support design are
given in Sections 3.4, 5.4, 5.5, and 7.3, respectively.
3.4 Design of Stoping Blocks
Stope block design issues are related to the global design and stability of
large sections of a mine, such as a new adjacent orebody, extensions at depth,
or in the abutment of an existing deposit (Chileshe and Kulkarni, 1995).
Global design issues are represented schematically in Figure 3.1 and listed
in detail in Table 3.2. The issues involved include global orebody delineation, mine access and infrastructure, dimensions of sublevel intervals, fill
requirements, equipment, and ventilation considerations. Stress analysis of
the global production schedules is critical to determine the loading conditions likely to result from any proposed mine-wide stoping sequences.
TABLE 3.2
Stope Block Design Issues
Exploration drilling requirements for orebody delineation for the designed area
Area-wide rock mass characterization from borehole data and direct access
Overall mining method selection
Quantity and grade of ore required with respect to scheduled metal targets
Access and infrastructure development requirements—ore-handling systems,
workshops, etc.
Production scheduling, details, and timing
Induced stresses from scheduled sequences, including extraction directions
Primary and secondary stope dimensions, including regional access pillars
Fill system requirements
Equipment requirements
Ventilation
Global economic assessment
Planning and Design
53
3.4.1 Orebody Delineation
The geological analysis on a block scale requires information on orebody
boundaries, grade, major geological structures, as well as the major rock
types within and around the orebodies. A grade distribution and a geotechnical model on a block scale are constructed from the geological interpretation of the data, which is initially collected from widely spaced surface
diamond drill holes. The preliminary design of a stoping block layout is
based on confirmatory exploration drilling, with holes drilled at 60–80 m
spacing. Additional geological information is required to provide the ore
limits and grade information suitable for a detailed stope design. This information can be collected as underground access becomes available and stope
delineation drilling at 20–40 m spacing can be carried out. In addition, geological and geotechnical mapping is carried out from the exposed rock mass
around a stope block development.
The geological and geotechnical models are used by a mine-planning engineer to develop a geometrical model of a stoping block in three dimensions.
The major geological structures likely to influence overall block stability are
determined and included in the analysis. The resulting three-dimensional
model is then used to calculate tones and grade for the global design block
(Thomas and Earl, 1999). Following mining method selection and an economic analysis for the block, the design of the development, ore- and wastehandling systems, services, and ventilation can be undertaken.
3.4.2 Global Extraction Sequences
One of the limiting factors affecting the design of an underground excavation is the maximum excavation size that a rock mass can sustain without
failure. This failure may take place either as a function of movement along
planes of weakness or through a combination of failures through intact rock
and on geological discontinuities. In most orebodies suitable for open stoping, the volume that may be excavated safely such that stope wall failures
are avoided, is many times smaller than the orebody itself. Consequently,
a series of individual stopes must be excavated to achieve full orebody
extraction.
One of the most important tools that a design and planning engineer has
for controlling the overall behavior of a rock mass is the extraction sequence
of the stopes contained within a given area of an orebody. Extraction
sequences are fundamental to achieving production targets safely and economically throughout a stope life. In most stoping mines, various stages of
development, production, and filling occur at any one time. The ore sources
are likely to be scheduled from a number of locations and extraction methods. In general, a stoping sequence is driven by ore grade requirements,
operational issues, and induced stress considerations (Potvin and Hudyma,
2000). A technically sound strategy is to avoid creating blocks of highly
54
Geotechnical Design for Sublevel Open Stoping
stressed rock within an orebody. This can be achieved by retreating stopes
to an orebody abutment instead of creating pillars located within central
orebody areas (Beck and Sandy, 2003). In general, the overall stope extraction
sequence is influenced by the nature of the orebody in question.
3.4.2.1 Massive Orebodies
Massive orebodies are extracted using multiple stopes (primary, secondary,
and, when required, tertiary) in conjunction with mass blasting techniques
and cemented fill. A number of sequencing options can be used including
temporary or permanent rib or transverse pillars, strike slots with continuous or discontinuous advance, and chequerboard sequences. Each overall
extraction sequence can be engineered to manage the induced stress redistributions on a global scale. Ideally, the initial stopes are extracted within a chosen area of an orebody and subsequent stopes are retreated systematically
toward orebody abutments taking into account the stress redistributions,
production tonnage requirements, and access constraints.
One extraction option used in extremely good-quality rock masses is to
mass blast secondary stopes into adjacent primary stopes to create very large,
but stable, openings (Mikula and Lee, 2000). In order to increase recovery and
achieve stability, the resulting voids can be filled using either consolidated or
unconsolidated fill with the individual stopes separated by rib (longitudinal)
and transverse pillars (Figures 3.3 and 3.4). The latter option leaves a high
proportion of ore tied up in the rib and transverse pillars. Methods such as
sublevel caving retreat have been used to achieve complete recovery of these
pillars (Alexander and Fabjanczyk, 1981).
The concept of a discontinuous strike slot for a 12-stope extraction sequence
is shown in Figure 3.5. Assuming the major principal stress to be normal to
the long axis of an orebody, the primary, secondary, and tertiary stopes are
designed with an overall stress management philosophy consisting of stress
shadowing and orebody abutment retreat. Once the strike slot has been completed (stopes 1–4), all the remaining stopes are effectively stress shadowed.
Stress shadowing occurs when two or more excavations are aligned along a
major principal stress trajectory. Stresses redistribute, and some areas may
be stress-relieved as the rock lies in the shadow cast by the excavations. In
addition, stress may be intensified in other areas, depending upon the distance between the excavations (Figure 3.6).
In very high stress environments, sequences using transverse pillars or
discontinuous transverse/strike slots may concentrate stress even in the early
stages of extraction. At the Creighton Mine in Canada, a series of central
stopes were extracted adjacent to each other to form a continuous slot within
an initial mining block in order to create a stress shadow for the remaining
stopes (Figure 3.7). In order to form a continuous strike slot, the fill from
the initial stope must be cured before extraction of the immediately adjacent
stopes can proceed. Production from the first three stopes is slowed by the
55
Planning and Design
Open cut
Open cut
100 m
5L
Block A
Cut and fill
(filled)
A
Cut and fill
7
Open stope
B3
9L
300 m
Crown pillar
C
2
C3
C1
D1
13L
F4
B1
B2
B1
3
1
2
6
C4
F3
Block B detailed
extraction sequence
E1
Blocks A,B,C, D, E, and F
extracted and filled with
unconsolidated fill
F1
F2
5
7
21L
G4
G3
G4
G1
G2
4
B3
D2
E2
E3
15L
500 m
B2
5
9
G5
G3
4
6
3
G2
1
2
8
H
Orebody outline
Rib pillar
Block G detailed
extraction sequence
5500 N
5000 N
4500 N
FIGURE 3.3
Temporary rib pillars within a top-down sequence at Mount Charlotte Mine. (From Ulla, Z.,
Applicability of the Mathews stability graph for evaluating stability of open stopes at the
Mount Charlotte Mine, MEngSc thesis, Curtin University of Technology, Kalgoorlie, Western
Australia, Australia, 1997.)
1500 E
Rib
pillar
S4
8f
(a)
Transverse
pillar
au
2000 E
lt
(b)
FIGURE 3.4
(a) Rib and transverse pillars at Mount Isa Mines and (b) transverse pillar at the Darlot Gold
Mine. (From Alexander, E.G. and Fabjanczyk, M.W., Extraction design using open stopes for
pillar recovery in the 1100 ore body at Mount Isa, in D.R. Stewart, ed., Design & Operation of
Caving & Sublevel Stoping Mines, SME of AIME, New York, 1981, pp. 437–458.)
56
Geotechnical Design for Sublevel Open Stoping
10
3
σ1
7
2SLOS 1SLOS 2SLOS
σ1
8
1
6
4
12
2
2SLOS
P
5
1SLOS
9
1SLOS 2SLOS 2SLOS
σ1
2SLOS
P
11
2SLOS
10
3
8
σ1
2SLOS
σ1
σ1 σ
1
7
2SLOS 1SLOS 2SLOS
6
12
σ1
11
7
2SLOS
6
σ1
σ1
2SLOS
8
1SLOS
9
10
2SLOS
5
4
1SLOS 2SLOS2SLOS
2SLOS
σ1
5
4
1SLOS
9
1SLOS 2SLOS2SLOS
12
2SLOS
2SLOS
σ1 σ
1
11
2SLOS
σ1
FIGURE 3.5
Plan view of discontinuous transverse slot extraction sequence for a massive orebody. (From
Villaescusa, E., Extraction sequences in sublevel stoping, Proceedings of the 12th International
Symposium on Mine Planning & Equipment Selection, Kalgoorlie, Western Australia, Australia,
April 23–25, 2003, pp. 9–18, AusIMM, Melbourne, Victoria, Australia. With permission.)
Low stress
Stress concentration
area
lines
Extracted stopes
Low stress
area
Highly stressed
area
Low stress
area
Low stress
area
Undisturbed stress field
FIGURE 3.6
Plan view showing stress shadowing across a series of stopes. (From Villaescusa, E., Extraction
sequences in sublevel stoping, Proceedings of the 12th International Symposium on Mine Planning
& Equipment Selection, Kalgoorlie, Western Australia, Australia, April 23–25, 2003, pp. 9–18,
AusIMM, Melbourne, Victoria, Australia. With permission.)
requirements to not expose the initial fill mass simultaneously on both sides.
This means that the third stope within the strike slot must wait until the fill
mass in the second stope has cured.
Another alternative is to adopt a chequerboard pattern of stope extraction.
The process starts with primary stopes filled with consolidated fill followed
by secondary and tertiary stope extraction of stope pillars having multiple
fill mass exposures (Grant and DeKruijff, 2000). The stoping front can either
move longitudinally or adopt a continuous retreat strategy depending upon
57
Planning and Design
12
9
9
10
6
10
7
3
7
4
1
5
8
2
8
11
6
11
9
9
12
FIGURE 3.7
Plan view showing a continuous, pillarless, stoping sequence. (After Min. Sci. Technol., 13,
Trotter, D.A., Vertical crater retreat mining in the Sudbury Basin, 131–143, Copyright 1991, with
permission from Elsevier.)
the level of in situ stress and the production tonnage requirements. Figure 3.8
shows the massive 1100 orebody at Mount Isa Mines, in which a north to
south global extraction sequence has continuously stepped out to access primary stoping blocks. The extraction was designed with large, 40 × 300–400 m
east–west transverse pillars for access, ventilation, and services (Grant and
DeKruijff, 2000).
1100 Orebody
1500 m
Hangingwall
lens
Lower
footwall
lens
Footwall lens
Northern
1100 orebody
1900 Orebody
Man and
supply shaft
Stopes to be mined
Stopes filled or empty
Mount ISA mines
Copper mine
FIGURE 3.8
Plan view of the 1100 orebody showing extracted stopes. (From Grant, D. and DeKruijff, S.,
Mount Isa Mines—1100 orebody, 35 years on, in G. Chitombo, ed., Proceedings of the MassMin
2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 591–600, AusIMM,
Melbourne, Victoria, Australia. With permission.)
58
Geotechnical Design for Sublevel Open Stoping
The advantages of a properly designed chequerboard extraction sequence
include stable primary stopes which must be tight-filled in time to provide support to the remaining stopes and crown pillar (Alexander and
Fabjanczyk, 1981). A disadvantage is the large amount of ore tied up within
the remaining tertiary stope pillars, where localized stope design can be
complex and a function of existing development and the number of fill exposures as mine life progresses. A chequerboard sequence is dependent upon
successful mass blasting practices and the development of stable fill masses
that provide support to adjacent rock masses with minimal dilution during
multiple-fill exposures (Bloss, 1992).
3.4.2.2 Steeply Dipping Orebodies
In the case of steeply dipping and relatively narrow orebodies, the most common orebody access is through crosscuts off access drives that are connected
to ramps located in the footwall of the orebodies. The crosscuts intersect the
orebodies from footwall to hangingwall and ore drives are developed from
the crosscuts along the strike of the intersected orebodies. In cases where
bench stoping is used as the preferred mining method, extraction can be
retreated toward the access crosscuts using either a top-down or a bottomup extraction sequence. A top-down bench stope extraction sequence usually requires permanent rib pillars to minimize dilution between individual
stopes along strike. In addition, a series of sill pillars may be required to
control overall stability and dilution and to isolate any unconsolidated fill
that may be introduced into the upper stopes as extraction progresses downward (Figure 3.9). A bottom-up sequence requires fill in order to provide a
working floor as the extraction proceeds upward. The need for crown pillars
is minimized by the use of rib pillars along the strike of the orebody and the
beneficial impact of the fill masses (Figure 3.10).
Flexibility and productivity can be greatly enhanced with the introduction
of two access crosscuts as shown in Figure 3.11. Although more costly, this
configuration increases tonnage and allows for a better stress redistribution
as the initial stopes can be located in the center of the mining block with
subsequent retreat toward the abutments.
In cases where multiple-lift sublevel stoping is used to extract narrow
tabular orebodies, a series of primary and secondary stopes can be
designed along the strike of the orebody. The pillar stopes are designed
large enough to enable safe recovery between primary stopes. Figure 3.12
shows the stoping sequence for the Kanowna Belle orebody, Stoping
Block A. The extraction sequence is based upon a primary stope extraction
and filling with consolidated fill, before the secondary pillars are extracted.
In other cases, stope extraction in conjunction with unconsolidated fill and
separated by permanent pillars can be used to extract low-grade orebodies
(Figure 3.13).
59
Planning and Design
Drill drives parallel to orebody
1
3
2
Permanent
rib
pillars
4
Central
access
crosscuts
5
2
3
Stope void unfilled
Permanent sill pillar
8
7
8
10
9
4
5
Permanent sill pillar
6
1
6
7
10
9
FIGURE 3.9
Longitudinal view of a top-down extraction sequence, permanent pillars, and retreat to a central crosscut, no fill.
Drill drives
9
8
6
Rib
pillars
optional
4
3
1
5
2
8
Filled stope
Central
access
crosscuts
7
9
10
10
6
7
5
3
4
2
1
FIGURE 3.10
Longitudinal view of a bottom-up extraction sequence, retreating to a central crosscut; pillars
are optional.
Figure 3.14 shows an example of primary and secondary stope extraction
from Mount Isa Mines, where some of the secondary stopes are mass blasted
into the void created by the primary stopes in what is called a “triplet” stope
extraction (Bywater et al., 1983).
Figure 3.15 shows a primary and secondary stope extraction sequence for
a shallowly plunging orebody at Kambalda, Western Australia. The cost of
cement binder in the fill is minimized by filling the secondary stopes with
unconsolidated waste rock.
60
Geotechnical Design for Sublevel Open Stoping
Retreat to abutment
access
Retreat to abutment
access
6
5
4
5
6
3
2
1
2
3
Optional pillars
Uncemented fill
FIGURE 3.11
Longitudinal view of a bottom-up extraction sequence with double access.
The advantage of primary and secondary stoping sequences lies in the initial high flexibility and productivity and low cost during primary stoping.
The overall cost is minimized by the use of unconsolidated fill within the
secondary stopes. A disadvantage is that stress redistributions may cause
rock mass damage late in the extraction sequence.
Figure 3.16 shows an example in which the induced stresses increase as
the stope extraction progresses within an abutment area. The results show
normal stress in excess of 70 MPa in the crown pillar region of the seven
orebody L692–L698 stopes at Mount Isa Mines. The induced stresses were
predicted using the program NFOLD, and confirmed with in situ stress
measurements. Previous studies indicated that 70 MPa compressive stress
was considered to be a critical value within the seven orebodies. Field observations identified severe spalling and cracking of the L690 pillar on the 13th
level (Bywater et al., 1983).
The effects of stress can be minimized by avoiding the undercutting of
individual stopes and by mass blasting those highly stressed regions within
a stoping block. Multiple-lift primary and secondary stopes have been used
very successfully to achieve complete extraction with minimal dilution
within the steeply dipping lead orebodies at Mount Isa Mines (Goddard,
1981; Bywater et al., 1983; Beck et al., 1997) and also at the Kanowna Belle
Mine (Magee, 2005; Cepuritis et al., 2007).
3.4.2.2.1 Pillarless, Center-Out Sequences
Pillarless, center-out mining sequences have been proposed to eliminate the
need for secondary stopes (Morrison, 1996). The perceived advantage from
such sequences is the slow rate of convergence of the host rocks as stoping
190 m Sub
13
Filled
9
2
7
8
1
14
Fill in progress
18
10
3
5
11 15
4
Current source
17
6
16
Decline
Scheduled
12
Vent raise
FIGURE 3.12
Longitudinal view of Kanowna Belle Mine—Stoping Block A. (From Bywater, S. and Fuller, P.G., Cable support of lead open stope hangingwalls
at Mount Isa Mines Limited, in O. Stephansson, ed., Rock Bolting: Theory and Application in Mining and Underground Construction, Proceedings of the
International Symposium on Rock Bolting, Abisko, Sweden, 1983, pp. 539–555, Balkema, Rotterdam, the Netherlands.)
200 m RL
160 m Sub
100 m RL
110 m Sub
135 m Sub
Vent raise
Planning and Design
61
62
10/7/01
3/7/01
23/6/01
19/6/01
13/6/01
6/6/01
14/5/01
9/5/01
11/5/01
Geotechnical Design for Sublevel Open Stoping
10/7/01
3/7/01
23/6/01
1
10/7/01
3/7/01
3
23/6/01
21/6/01
19/6/01 21/6/01
Permanent pillar
6/6/01
6/6/01
20/5/01
20/5/01
17/5/01
3/5/01
14/5/01
14/5/01
2
11/5/01
17/5/01
4/5/01
3/5/01
1
Permanent pillar
2/5/01
Overall
sequence
FIGURE 3.13
Permanent pillars left between primary stopes (filled with unconsolidated fill)—Mount
Marion Mine. (From Villaescusa, E., Extraction sequences in sublevel stoping, Proceedings of
the 12th International Symposium on Mine Planning & Equipment Selection, Kalgoorlie, Western
Australia, Australia, April 23–25, 2003, pp. 9–18, AusIMM, Melbourne, Victoria, Australia. With
permission.)
proceeds from the center toward the orebody abutments (Figure 3.17). It is
argued that the slow rate of convergence is likely to minimize the magnitude of the local seismic events. In addition, the small single-lift stopes
may reduce the amount of released seismic energy. Such a pillarless stoping sequence was used in Block 3 at the Golden Giant Mine in Canada and
named pyramid retreat, as mining progresses in a triangular shape (Potvin
and Hudyma, 2000). The Golden Giant Mine pyramid retreat sequence is
illustrated in Figure 3.18.
Although a continuous advancing stoping sequence is an attractive idea,
it is very difficult to implement in practice, especially when fill is introduced into the system (Grice, 1999). The overall productivity is severely
constrained by the individual stope cycle times as stopes must be mined,
filled, and cured before an adjacent stope can be extracted. With active
mining on a large number of sublevels, substantial development, scheduling, and logistic challenges are experienced throughout the stoping block
(Potvin and Hudyma, 2000). As an example, the extraction of stope No. 6
M 654
4
M 657
1 Extraction
sequence
CHF
6 7
M 660
5
0
in meters
1
M 667
Scale
25 50
3
M 662
2
11
100
M 676
9
M 674
8
M 683
M 678
N
3 10 4
Cut
and
fill
15 L
15 B
14 L
HF
J 668
CHF
7
J 672
3
1 Extraction
sequence
4
J 679
5
4
J 684
2
100
J 686
in meters
Scale
0 25 50
1
8 6
12
13
9 4 11 10
J 694
J 696
J 698
N
FIGURE 3.14
Longitudinal section view showing secondary stopes mass blasted into the void created by a primary stope, Mount Isa Mines. (From Bywater, S. and
Fuller, P.G., Cable support of lead open stope hangingwalls at Mount Isa Mines Limited, in O. Stephansson, ed., Rock Bolting: Theory and Application in
Mining and Underground Construction, Proceedings of the International Symposium on Rock Bolting, Abisko, Sweden, 1983, pp. 539–555, Balkema, Rotterdam,
the Netherlands.)
HF
15 B
15 L
M 651
6 500 N
14 C
2800 RL
14 L
M 665
6 600 N
M 671
6 700 N
M 680
6 800 N
13 L
14 C
6 700 N
J 670
J 675
13 L
6 000 N
J 682
6 800 N
J 701
7 000 N
J 689
J 691
6 900 N
J 704
Cut and fill (11L–13L)
Planning and Design
63
64
Geotechnical Design for Sublevel Open Stoping
P1
P1
S1
S1
P2
P2
S2
S2
P3
P3
S3
S3
P4
P4
S4
S4
FIGURE 3.15
Primary and secondary stope extraction sequence for a shallowly plunging orebody. P1, primary stope cemented fill; S1, secondary stope unconsolidated fill. The numbers show the
sequence of stope extraction.
(Figure 3.18), although very early in the sequence, requires seven operational sublevels.
Pillarless stoping sequences are more suited to paste fill as they require
rapidly curing cemented fill with minimal drainage delays in all the stopes,
thus potentially increasing the operating cost. In addition, tight backfill of
the stope crowns is rarely achieved, especially when cemented rock fill is
used (Figure 3.19). Introducing hydraulic fill to achieve tight fill is timeconsuming, expensive, and sometimes not practical. Consequently, large
crowns that require extensive rock reinforcement are exposed by this
method. In some cases, damage from stress concentrations (cracking through
intact rock or slip on geological structures) in the stope brows is also experienced. This may create difficulties during drilling and blasting and make
the reinforcement schemes inefficient, as very large slabs of rock parallel to
the stope edges may be released.
Figure 3.20 shows a pillarless stope extraction sequence where the stopes are
partially mined under cemented paste fill. This sequence was implemented to
allow the extraction of stopes under very high stress at the Junction Mine,
Kambalda, Western Australia. The pillarless sequence was facilitated by the
large initial extraction already under way within the center of the orebody at
the Junction Mine. The pillarless extraction sequence actually evolved around
the edges of the previous extraction. Figure 3.21 shows a typical view inside
one of the open stopes, with paste fill constituting the back or roof of the stope.
65
Planning and Design
Cut and fill (11–13/L)
L698
L695
L692
L690
L687
L685
L683
13/L
Planned stopes
14/C
Cut
and
fill
14/L
15/D
15/B
15/L
MPa
Extracted
stopes
Depth stress
Extracted
28
31
Extracted
stopes
Extracted
stopes
Induced
normal
stress
+70 MPa
50–59 MPa
60–69 MPa
40–49 MPa
0
25
50
100 m
FIGURE 3.16
Induced stress as stope extraction progresses. (From Bywater, S. et al., Stress measurements
and analysis for mine planning, Proceedings of the Fifth Congress of the International Society
for Rock Mechanics, Melbourne, Victoria, Australia, April 11–15, 1983, pp. D29–D37, Balkema,
Rotterdam, the Netherlands.)
In practice, continuous retreating sequences can only be applied to individual stoping blocks that are separated by crown or waste pillars. An
increased number of advancing fronts increases extraction flexibility, but
also increases the number of pillars that must be dealt with at a later stage.
Extraction of the pillars between the continuous fronts may be complicated
in areas where high induced stresses are experienced. Figure 3.22 shows a
proposed longitudinal view of two stoping blocks extracted using a continuously advancing front and single-lift stopes.
66
Geotechnical Design for Sublevel Open Stoping
4
4
3
3
2
3
2
1
2
3
4
FIGURE 3.17
A conceptual pillarless stoping sequence, center-out extraction. (From Morrison, D.M., CIM
Bull., 89, 46, 1996.)
3.4.2.2.2 Primary and Secondary 1-3-5 or 1-5-9 Stoping Sequences
A compromise to a pillarless sequence is to use a general triangular retreat
shape but with a short lift primary and secondary stope arrangement. This
system has been implemented at the Williams mine in Canada and is illustrated in Figure 3.23. This methodology allows for a number of primary
stopes to be mined simultaneously, hence increasing the productivity within
the stoping block. Because of the detrimental effects of the stress redistributions on the pendant pillars formed in the sequence, secondary pillar
stopes must be recovered as early as possible in the extraction sequence.
In general, no more than two sublevels are mined ahead of a pillar before
recovering it and both sides of a pillar cannot be mined simultaneously
(Potvin and Hudyma, 2000). In practice, however, stoping blocks are likely
to interact with one another, making extraction of sill pillars extremely difficult and costly using this method. Figure 3.24 shows a longitudinal section view of a sill pillar extraction at the Williams mine, where a single
seismic event required over $4 million expenditure on rehabilitation and
additional development in order to resume mining. In addition, significant
delays were incurred.
A variation of this method was proposed for the George Fisher Mine,
Queensland, Australia, where a 1-5-9 stoping sequence was selected for
extraction (Neindorf and Karunatillake, 2000). Stopes 1-5-9 are extracted as
two-lift primaries and filled with consolidated fill (Figure 3.25). This is followed by another set of primary two-lift stopes (3-7-11), also filled with consolidated fill. Following the fill cure within the primary stopes 1-3-5-9-11, a
set of single-lift stopes (2-6-10) is then extracted and filled with unconsolidated fill. This creates a pendant pillar, which has many degrees of freedom
and relies on the fill support from the primary stopes for stability. Finally,
the single-lift stopes 4-8-12 are extracted and filled with unconsolidated fill
67
Planning and Design
18
27
19
10
20
21
13
6
12
22
23
15
9
4
8
14
24
25 17
11
7
1
5
2
16
66 m
Stope height
4600 level
4533 level
4466 level
26
3
4400 level
1 Extraction sequence
FIGURE 3.18
Pyramid retreat at the Golden Giant Mine, Canada. (From Potvin, Y. and Hudyma, M., Open
stope mining in Canada, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane,
Queensland, Australia, October 29 to November 2, 2000, pp. 661–674, AusIMM, Melbourne,
Victoria, Australia. With permission.)
68
Geotechnical Design for Sublevel Open Stoping
Cutoff raise
used as fill pass
Filling
Drilling
Producing
FIGURE 3.19
Conceptual continuous advance for single-lift stopes. (From Grice, T., Mine backfill—course
notes for the masters of engineering science in mining geomechanics, MEngSc thesis, Western
Australian School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia,
Australia, 1999.)
5
9
13
4
11
3
8
14
7
3
Extracted
area
7
10
6
4
6
5
8
9
8
12
11
10
13
12
FIGURE 3.20
Top-down, pillarless stoping sequence, mining under paste fill at the Junction Mine, Kambalda,
Western Australia.
Planning and Design
69
FIGURE 3.21
Top-down stoping under paste fill at the Junction Mine, Kambalda.
before the entire sequence is repeated up-dip. The extraction of stopes 4-8-12
also creates pendant pillars.
A major disadvantage of 1-5-9 (or 1-4-7) extraction sequences using short
lift stopes is their inefficient production mucking characteristics. The method
effectively requires (an upward) moving drawpoint sequence (even in primary stopes), which necessarily follows the vertical retreat of the stopes, as
shown in Figure 2.15. This implies that production mucking is carried out in
areas that had previously been subjected to stress redistribution and stope
blasting at the stope crowns. Each stope access becomes a stope drawpoint
and a significant amount of reinforcement using cablebolting is required in
all the stope accesses and exposed backs to minimize large-scale back failures (Figure 3.26).
Reinforcement can be largely inefficient within the bottoms of pendant
secondary pillars where remote mucking is required for 100% of the tonnage. Furthermore, additional footwall development access in waste may be
required on each sublevel, as more than one access may be required for effective production mucking of each individual stope.
3.4.2.2.3 Multiple Steeply Dipping Orebodies
The extraction sequence for multiple, steeply dipping parallel orebodies,
which are accessed by a common crosscut off a footwall ramp, requires
70
Geotechnical Design for Sublevel Open Stoping
22
20
18
16
8
6
4
2
Stoping
block 1
12
14
10
Crown pillar between stoping blocks
21
19
17
5
3
15
Stoping
block 2
13
11
9
7
Waste
pillar
FIGURE 3.22
Bottom-up continuous extraction sequences on each stoping block.
1
71
Planning and Design
P
S
P
S
P
S
P
Production
blasting
1
3
Development
P
Slot
raise
Filling
uncemented
rockfill
Cablebolts
S
25 m
25 m
5
Production
blasting
25 m
Cemented rockfill
FIGURE 3.23
A conceptual longitudinal section view showing a 1-3-5 extraction sequence. (From Bronkhorst,
D. et al., Geotechnical principles governing mine design at the Williams Mine, in W.F. Bawden
and J.F. Archibald, eds., Proceedings of the International Congress on Innovative Mine Design for the
21st Century, Kingston, Ontario, Canada, August 23–26, Balkema, Rotterdam, the Netherlands,
1993, pp. 433–442.)
31
30
29
28
27
26
25
24
23
9475
20
19
17
16
15
14
13
12
9475
9450
9415
Major damage
Caved
zone
9380
Minor damage
9370
9345
9310
18
Moderate damage
December 17 2.6 Mn rockburst
March 29 rockburst
9415
9370
21
Filled
zone
9450
9380
22
9345
Filled
zone
9310
FIGURE 3.24
Longitudinal section view of crown pillar between stoping blocks 3 and 4 at the Williams mine.
(After Bawden, W.F. et al., Lessons in control of mine costs from instrumented cablebolt support, in J. Girard et al., eds., Proceedings of the Fourth North American Rock Mechanics Symposium,
Seattle, WA, July 31 to August 3, 2000, pp. 633–642, Balkema, Rotterdam, the Netherlands.)
additional consideration, as the stope extractions at particular locations can
be interrelated. In those cases where the thickness between orebodies is
less than half the stope height, the stopes are likely to interact and the stope
hangingwall deformations are minimized by extracting the orebodies from
footwall to hangingwall. The extraction sequence shown in Figure 3.27 aims
72
Geotechnical Design for Sublevel Open Stoping
D orebody 1, 5, 9 sequence
11L
12C
12L
13C
6
5
8
7
1
13L
1P
3
2S
6
5
8
7
1
2
3
3P 4S
4
3
4
8S
2 Fill stopes 1, 5, 9 and extract 3, 7, 11
3 Fill and cure stopes 1, 3, 5, 7, 9, 11 and extract 2, 6, 10
2
1
5P 6S 7P
1 Extract stopes 1, 5, 9
8
7
2
4
6
5
4 Fill stopes 2, 6, 10 and extract 4, 8, 12
North
9P 10S 11P 12S
3-22
3-12
3-06
3-11
3-21
3-31
3-14
3-08
3-03
3-07
3-13
3-23
3-02
3-10
FIGURE 3.25
Longitudinal section view of George Fisher conceptual orebody extraction. (From Neindorf,
L.B. and Karunatillake, G.S.B., George Fisher Mine—Feasibility and construction, in G.
Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29
to November 2, 2000, pp. 601–609, AusIMM, Melbourne, Victoria, Australia. With permission.)
3-04
3-01
3-05
3-09
Extracted
Longitudinal view
stope extraction
sequence
Potential failure
surface under
investigation
Open slot
Stope
3-11
Filled
Mucking level
Filled
Isometric view stope 3-11
FIGURE 3.26
Brow instability in highly stressed single-lift open stopes at Hemlo Gold Mine, Canada. (From
Milne, D. and Gendron, A., Borehole camera monitoring for safety and design, Presented at
92nd CIM Annual General Meeting, Ottawa, Ontario, Canada, May 6–10, 1990, 13pp.)
73
Planning and Design
σ1
Hig
hly
st
zon ressed
e
σ1
4
3
2
1
Lea
ext ding
rac
tion
5
Extraction
sequence
Hangingwall
orebody
Footwall
orebody
1
Filled
FIGURE 3.27
Footwall stope extracted ahead of other stopes in the same lift. (From Villaescusa, E., Trans.
Inst. Min. Metall., Sect. A Min. Ind., 105, A1–A10, 1996.)
to minimize the effects that stopes might have on each other. The stopes
interact as the block extraction sequence advances up-dip toward a region
of high induced stress below a mining block extracted earlier. Within this
sequence, the footwall stopes are always extracted one or two lifts ahead
of the hangingwall stopes, effectively creating a “leading” stope geometry.
The sequence is devised to “shield” the rest of the stopes in a particular lift
from excessive stress-induced damage, as well as to minimize the effects of
blasting, as most hangingwalls are mined in undisturbed ground. In some
cases, the leading orebodies may experience stress-related crown damage,
and adequate rock reinforcement must be provided to minimize failures.
Alternatively, the leading orebody must be selected following considerations
of rock mass strength, orebody width, and orebody grade. In such cases,
it may be advisable to select a narrow orebody (located anywhere in the
sequence) as the leading orebody.
3.4.3 Numerical Modeling
Induced stresses from a particular extraction sequence can be determined
using numerical modeling (Beck and Duplancic, 2005; Wiles, 2006). The
inputs required are an estimate of the stress field orientation and magnitude
with depth, the rock mass deformational properties, the initial excavation
74
Geotechnical Design for Sublevel Open Stoping
geometry, and the chosen overall stope extraction sequence. The limitations
of linear elastic modeling include the inability to predict movement, falloff,
or dilution from fault or shear zones. Consequently, the results must be used
in conjunction with structural information, for example, large fault behavior,
in order to interpret the different stoping sequences. Alternatively, nonlinear
modeling which is able to predict rock mass failure and any stress redistribution resulting from such failures can be used (Beck and Duplancic, 2005).
Progressive orebody extraction may induce several phases of post-peak
behavior in a rock mass, and small changes to the stress field induced by
distant stope extraction may cause significant rock mass damage around the
stope boundaries.
Typical outputs from numerical modeling include stresses and displacements, which in turn can be compared with empirical failure criteria established for the different domains within an orebody (Brady and Brown, 2004).
Any predictive models must be validated against field data and observations.
Modern numerical modeling tools allow realistic assessments to be made of
mine-wide extraction sequences (Figure 3.28). The model preprocessing is
usually linked to a three-dimensional model of the excavation geometries in
σ1
(MPa)
50.0
45.0
40.0
35.0
30.0
25.0
20.0
15.0
10.0
5.0
0.0
z
y
x
FIGURE 3.28
Major principal stress distribution in a stoping block using the program MAP3D. (From
Villaescusa, E. et al., Open stope design and sequences at great depth at Kanowna Belle,
Unpublished Research Report for Placer Dome Asia Pacific, 2003a, 217p.)
Planning and Design
75
order to reduce mesh generation times. A link to mine scheduling is required
in order to analyze the different extraction sequencing options.
3.4.4 Regional Pillars
The use of regional pillars is sometimes required to control the overall stability and to provide safe access to active stoping areas across an
existing orebody. In some cases, the pillars are required for permanent
access throughout the entire life of a stoping block. The use of transverse
pillars to control the overall stability of massive orebodies, such as the
1100 orebody at Mount Isa Mines is well documented (Alexander and
Fabyanczyk, 1981). Transverse pillars are an efficient way of controlling
overall crown subsidence, while ensuring safe access through the orebody
(Figure 3.29).
Regional pillars were also used to provide permanent access to multiple,
steeply dipping orebodies in the Lead Mine at Mount Isa Mines. Access
to each of the orebodies was provided through crosscuts centrally located
within 40 m wide pillars. Extraction of the orebodies retreated toward the
pillar edge as shown in Figure 3.30. Provision for the recovery of such permanent pillars can be designed for late in the extraction sequence of a mining block.
Stress redistributions from a global stoping sequence may cause damage
to transverse or regional pillars. This damage may require rehabilitation
or loss of access development through the pillar. Extension strain cracking
(Stacey, 1981) parallel to the direction of the major principal stress orientation
may be experienced, especially in rock masses exhibiting a high modulus.
Consequently, an eventual recovery of transverse pillars must be planned
carefully, ideally with the initial pillar stope located in the best-quality rock
mass area. Extraction of the initial stope may allow an overall stress reduction within the pillar, as a stress shadow is likely to be created for the adjacent transverse pillar stopes. In the example shown in Figure 3.31, extraction
of stope A, as the first stope in the transverse pillar, may actually activate the
fault causing shearing and failure into the stope. On the other hand, extraction of stope B as the first stope within the transverse pillar may cause a
reduction of stresses through the pillar, minimizing the potential for shearing along the fault.
Damage to permanent pillars is not entirely determined by stress-induced
behavior, as preexisting geological discontinuities can also influence the performance of a pillar. Monitoring has linked stoping activities and instability
in concurrent extraction areas along the strike lengths of large fault zones
(Logan et al., 1993). The resulting behavior can be linked to induced stress
relief along the structures with increased loading and degree of freedom.
Large stope blasts can transmit energy along continuous fault zones, and fill
drainage may introduce water into fault systems. As a result, production and
4000 mN
V405
S434
Y434
U434 U438
R454
T446 U450
T454
S447 T450 S454
S450
R450
Q451 Q455
Recently filled stope
U442
T438 T442
S442 S446
R442
Q442 Q446
Q450 Q454
P450 P454
P458
P465
Q465
Q465
Q461
P461
5
Scheduled stope
T4
Producing or empty stope
W426
V430
Q438
P438
R432 S438
R434
T430 T434
S430
Q431
R430
Q435
P442 P446
N462
O458
Q461
°
Filled stope
V409
T422 T426
R426
Q426 Q430
Q434
P426 P430 P434 O438 O442 O446
N454
P471
M465 M469
66°
J46
N461
N465
N458
80
V401
U418
M444
L473
L473
S4
5000 mN
65
°
8
FIGURE 3.29
A plan view of the Mount Isa Mines 1100 orebody showing transverse pillar access, large-scale discontinuities, and scheduled stopes. (From Villaescusa,
E., Extraction sequences in sublevel stoping, Proceedings of the 12th International Symposium on Mine Planning & Equipment Selection, Kalgoorlie, Western
Australia, Australia, April 23–25, 2003, pp. 9–18, AusIMM, Melbourne, Victoria, Australia. With permission.)
55°
P41
U409
N430 N434 M438
4500 mN
O426 O430 O434 N438 N442 O447
N426
°
70
44
M
U403
ary
rim
6 p llar M405 M409 M413 M418
39
i
p
N422
M
99
N3
01
N401 N405 N409 N413 N418
N3 2
9
97
N3
N3
O418 O422
5
39
O401 O405 O409
O
2
O413
9
O3
P418 P422
P397 P401 P405 P409
94
P413
3
P
Q397
Q418 Q422
Q401 Q405 Q409
83
Q413
Q3
Q398
R401 R405 R409
R418 R422
96
R3
7
9
S408 R413
S405
3
Q421
S
S400
95
S3
S413
S409 S409
S418 S422
S401
60°
J46
T405 T409 T413
M422
76
Geotechnical Design for Sublevel Open Stoping
77
Planning and Design
N
S
Drilling horizon
To additional orebodies
(Sill drive)
Ring blasting
Access
crosscut
Broken
ore
Mucking horizon
Permanent
pillar
Drill holes
Sill
drives
To access ramp
Crosscut
Previous stope filled
To access ramp
Previous mucking horizon
Longitudinal section
Cross section looking north
FIGURE 3.30
Longitudinal and cross-sectional view of a typical permanent pillar in the Lead Mine, Mount
Isa. (From Kropp, W. and Villaescusa, E., Development of mining practices in the Lead/Zinc
Mine, Mount Isa, Proceedings of the Ninth Australian Tunnelling Conference, Sydney, New South
Wales, Australia, August 27–29, 1996, pp. 461–466. With permission.)
High stress
Stope A
Fault zone
Stope B
Extracted
High stress
Pillar stope
FIGURE 3.31
Plan view of an initial extraction in a transverse pillar stope. (From Logan, A.S. et al.,
Geotechnical instrumentation and ground behavior at Mount Isa, in T. Szwedzicki, ed.,
Geotechnical Instrumentation and Monitoring in Open Pit and Underground Mining, Proceedings of
the Australian Conference, Kalgoorlie, Western Australia, Australia, June 21–23, 1993, pp. 321–329,
Balkema, Rotterdam, the Netherlands.)
78
Geotechnical Design for Sublevel Open Stoping
filling strategies must minimize stope interaction along common faults that
intersect permanent pillars (Logan et al., 1993).
3.4.5 Block Development
The purpose of a block development is to provide suitable access for stoping and ore handling, fill reticulation, ventilation, mine services, as well
as gaining further and more detailed information about the nature and
size of the orebody. The two main factors to be considered are the mode
of entry into the underground workings and the related lateral development required to extract the orebodies. The layout of the basic development
depends upon the orebody characteristics, the nature of the host rock, and
the stoping method chosen for extraction. Properly designed block development is critical to the ongoing success of a stoping operation. Figure 3.32
shows the ore-handling flow from sublevel stoping in the 1100 orebody at
Mount Isa Mines, with some key infrastructure being illustrated (Grant
and DeKruijff, 2000).
3.4.5.1 Shaft Stability
Vertical shafts are the most common type of access for deep underground
orebodies. Shaft sinking and equipping is a specialized, complex procedure
usually costing millions of dollars. Consequently, it is economically justifiable to spend a significant amount of time and money on shaft site selection and characterization. The rock mass investigations require geotechnical
drilling to assess the presence of large-scale geological discontinuities, the
hydrological regime, the nature and strength of jointing, and the physical
properties of the rock types intersected. This is likely to indicate any potential instability problems during shaft sinking and the subsequent access
maintenance.
A shaft is sunk to a depth that will ensure many years of production during the life of a mine. Shaft location is controlled by the mining method used
as well as the rock types present on a particular site. In sublevel stoping, the
location of the shaft is usually at the footwall of the orebodies, where it is
likely to be outside the influence of any ground disturbance caused by the
stoping operations. In cases where the shaft is located within an orebody, a
large amount of level development can be carried out within the orebody.
However, a large amount of ore around the shaft must be left unmined as a
shaft pillar (Figure 3.33). For example, the main and services supply shafts of
the 1100 orebody at Mount Isa Mines have a shaft pillar that exceeds 200 m
in diameter (Grant and DeKruijff, 2000).
The design and monitoring of shaft pillars usually include the prediction
of elastic/plastic strain profiles as a first-pass design, followed by physical
monitoring of rock mass response to mining in order to identify displacement
on preexisting geological discontinuities intersecting the shaft. A maximum
15 level conveyor Fill passes
to stopes
CHF and
HF lines
Ground support
Development + mucking
23E Sub
6.2
Ore bin
Ore bin
Truck
haulage
Production
mucking
S60
5.2 C
/V
Loading
flasks
No.3
Skips
Concentrator
22/L
21/L
20/L
19/L
Crude ore bin
U62 shaft
Surface bin
Copper mine process flow diagram
lt
bo
ble
Ca
Crushers
ipples
No.4 T
Production and
blasting
superintendent
Drilling and services
superintendent
Development
superintendent
Copper planning
team leader
Ore
pass
19L Hallage
21C Sub
Ore
pass
Note: Most stopes are
Filled
adjacent to at least
Stope
one filled stope
Production
drilling
24D Sub
P41
Crusher
22D Sub
Spyder
passes
Blasting
Ore
Exhaust
vent, shafts
V33, M37, L44,
K48, I54 V33, Y37
water to death
adder gully
Geological model
and planning
Diamond
drilling
Copper mine
accountability
FIGURE 3.32
Schematic of ore process flow throughout the stoping cycle in the 1100 orebody, Mount Isa Mines. (From Grant, D. and DeKruijff, S., Mount Isa
Mines—1100 orebody, 35 years on, in G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2,
2000, pp. 591–600, AusIMM, Melbourne, Victoria, Australia. With permission.)
Surface
fill passes
Wet fill plant X41 Shaft
Surface fill conveyor
(ex. K.S.O.C.)
Planning and Design
79
80
Geotechnical Design for Sublevel Open Stoping
60
F
61
62
63
64
G
H
I
J
K
N643
M
N
N645
Restricted mining area
L
O
P
S
R62 supply and ore shaft
No mining
6500 N
R
6000 N
Q
FIGURE 3.33
Plan view of no mining and restricted mining pillars around the R62 shaft complex at Mount
Isa Mines. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)
strain criterion of ±150 με in any direction was historically established for
shaft conveyance at the R62 shaft complex at Mount Isa Mines.
Numerical modeling can be used to predict movement on regional faults
intersecting a shaft complex. The change on the stoping geometries, the fault
locations, and the history of shaft problems must be considered in the analysis. The results and predictions of the numerical analysis must be supported
by the actual shaft inspection results. Regular direct shaft inspections coupled with kinematic shaft surveys can provide a baseline for monitoring
the actual shaft deformation with time. Reusch and Beck (2007) have used
results from nonlinear numerical modeling to compare plastic strain with
measured shaft deflections. Their results are shown in Figure 3.34, where the
simulated magnitude of the shaft deflection matches measured values with
an error of less than 10%. The main deviation between model results and
measurements occurs over a short section of the shaft, where some significant
perturbations exist. The local change in shear strain usually corresponds to
81
Planning and Design
West
500
Shaft deflection (mm)
400
300
200
100
0
East
–100 –200
250.0
WE
WE FE
150.0
50.0
–150.0
–250.0
–350.0
Shaft depth (m)
–50.0
–450.0
–550.0
–650.0
–750.0
–850.0
FIGURE 3.34
Modeled and measured shaft deflections (left) and modeled plastic strain (right) due to sublevel open stoping. (From Reusch, F. and Beck, D., Simulating shaft and crusher damage in
deep mines, in Y. Potvin, ed., Proceedings of the Fourth International Seminar on Deep and High
Stress Mining, November 7–8, 2007, Australian Centre for Geomechanics, Perth, Western
Australia, Australia, pp. 65–79.)
significant mechanical difficulties related to deflection of the rails or damage
to shaft lining. The largest amount of the modeled plastic strain corresponds
with an area of significant damage in the hangingwalls of the stopes in close
proximity to the shaft (Reusch and Beck, 2007).
3.4.5.2 Ramp Access
In some cases, major access to stoping blocks is provided by ramps, which
are usually located within the footwalls of the orebodies (Figure 3.35). Access
and trucking ramp systems are generally used, with major trucking ramps
usually graded and designed with enough radius of curvature to preserve
sight distance, enable maneuverability, and minimize tyre wear. Ideally,
ramps are designed anticlockwise downward in order to provide optimum
sight distance to left-hand drive (LHD) drivers, which must descend bucket
first. Ramps must not lead directly into accesses to major mining excavations
such as workshops, fueling bays, etc.
The ramp dimensions are determined by the size of the mining equipment utilized. In particular, the design of ramp intersections with other
roadways is important, as they must remain stable. Ramps may undergo
high stress redistributions as the stopes are usually retreated toward
82
Geotechnical Design for Sublevel Open Stoping
N
Portal
1340RL
1240 mL
L7 stope
1200 mL
O8 stope
1160 mL
1120 mL
L8 stope
1080 mL
N8 stope
1040 mL
1000 mL
960 mL
H7 stope
K7 stope
N7 stope
920 mL
FIGURE 3.35
Ramp access at the Mount Wright Mine, Queensland. (From De Vries, R., Sublevel shrinkage
at the Mount Wright underground gold mine, MEngSc thesis, Western Australian School of
Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 2013, 89pp.)
crosscuts off a ramping system. The locations and geometries of the ramps
must take into account factors such as the orebody geometry, the rock
mass strength, and the stress loading as a result of the overall extraction
sequence. Typical horizontal distances for ramp location from an orebody
range from 50 to 70 m.
3.4.5.3 Crown Pillar
In some cases, a major crown pillar is left in place to separate open pit and
underground excavations within the same orebody (Figure 3.36). Conse­
quently, crown pillar stability is then critical to ensure safe underground
extraction. The crown pillar dimensions and stability are a function of a
83
Planning and Design
Open pit extraction
Crown pillar under open pit
–150 m
–250 m
–350 m
100 m
–600 m
Planned delineation drillhole
FIGURE 3.36
Crown pillar at the Kundana Gold Mine, Western Australia.
number of parameters. The most important are the width of the orebody,
the stress regime, the blasting practices, the rock mass strength within the
pillar, the overall extraction sequence (top-down or bottom-up), and whether
backfill will be introduced into the system.
The actual crown pillar dimensions will depend upon the stress environment. Indications of high stress could include obvious signs of mininginduced stress fracturing or rock burst activity. High stresses may also be
induced in otherwise low stress environments near the surface, due to the
geometry of the orebody and the extraction ratio below and above the pillar. In addition, if a crown pillar is situated within a stress shadow environment, consideration must also be given to potential unraveling due to loss of
clamping across the pillar. As a general rule of thumb, for narrow orebodies
(<10 m), the crown pillar height or thickness is based on a width/height ratio
of 1:1 plus 5–10 m. For orebodies wider than 10 m, the crown pillar heights
are designed with a width/height ratio of 1:1 plus 20–25 m. However, numerical modeling is required to determine whether excessive stress concentrations are likely to occur within a pillar.
A strategy to minimize the effect of stress and potential seismicity within
crown pillars is to place cemented fill within the first stoping lift, thus
allowing the recovery of all the ore and minimizing the buildup of stress.
Alternatively, the crown pillar may be recovered early in the stoping life by
84
Geotechnical Design for Sublevel Open Stoping
incorporating the extraction of portions of the crown pillar above each individual stope extraction.
3.4.5.4 Sublevel Interval
The selection of a sublevel interval is controlled by a global economic decision that provides the lowest cost per tonne of ore for the mining method
chosen for a particular stoping block. The selection of the sublevel interval
is not always controlled by stope wall stability. In most cases, the sublevel
interval is based on factors such as development cost, down dip orebody
irregularity, the available drilling equipment, and considerations of rock
mass damage from explosives (Figure 3.37).
The underlying criteria should be the control of dilution and the reduction
of the ore loss, as increased sublevel intervals reduce the required sublevel
development, but may increase dilution. Consequently, an assessment is
required of the anticipated economic impact of ore loss and dilution for each
particular sublevel interval. Although this is not an issue that is well understood, an attempt must be made during the economic evaluation to cost the
additional development required to reduce the sublevel interval in order to
minimize dilution and ore loss.
3.4.5.5 Access Crosscuts
Crosscuts are designed to provide access to the orebodies at the selected
sublevel interval. In cases where the crosscuts are located within a regional
pillar, they are designed to be directly above one another (see Figure 3.30).
FIGURE 3.37
The effects of orebody geometry on the chosen sublevel interval.
Planning and Design
85
FIGURE 3.38
Footwall development access for a tabular orebody. (From Cepuritis, P.M., Three-dimensional
rock mass characterization for the design of excavations and estimation of ground support
requirements, in E. Villaescusa and Y. Potvin, eds., Ground Support in Mining & Underground
Construction, Proceedings of the Fifth International Symposium on Ground Support, Perth, Western
Australia, Australia, September 28–30, 2004, pp. 115–127, Balkema, Leiden, the Netherlands.)
In single steeply dipping orebodies that are extracted using a single ramp
access as shown in Figure 3.38, the access crosscuts are not fixed at a particular location along the strike of the orebody, but rather where the ramp
intersects the sublevel elevation. In both cases, crosscut development must
be maintained at minimum size and design shape in order to improve stability at the crosscut–orebody intersection. The best practice for construction
is to anticipate the crosscut design position near the orebody hangingwall
boundary, with the final mining cut taken under geological control. Probe
drilling using a mobile drilling machine can be used for orebody delineation
prior to the mining of the last development cut near the hangingwall of the
orebody. Such a step may be required to avoid undercutting the hangingwall
planes, and thus minimize any falloff during subsequent stoping operations
(Figure 3.39).
Crosscuts also play a key role in orebody delineation and rock mass characterization as the orebody boundaries are delineated within the crosscut
walls prior to the orebody drive breakoff. The geotechnical behavior of the
stope boundaries can be predicted from the results of crosscut mapping
(Landmark and Villaescusa, 1992).
86
Geotechnical Design for Sublevel Open Stoping
Crosscut
Probe drilling
Crosscut
Last development cut
FIGURE 3.39
A conceptual section view of a development strategy to minimize hangingwall undercutting.
3.4.5.6 Raises and Orepasses
Raises can be used to connect different vertical mine elevations with each
other (Figure 3.40). Raises are used for a number of purposes such as ventilation, ore passes, and travelways. Ore passes are usually designed with an
angle exceeding 55° to the vertical in order to allow the broken rock to flow
by gravitational means. The modern mechanical methods of raising in open
stoping include raising by longhole drilling and raise boring.
Raising by longhole drilling consists of drilling holes in a suitable pattern,
through the full depth of the ground, up to 60 m long in some cases. Drilling
is usually carried out from the top level using conventional longhole drilling
equipment. In some cases, such as in a top-down bench extraction, uphole
drilling is utilized. Downhole raises are blasted in sections of approximately
5–10 m, while uphole raises are blasted over their entire length, usually less
than 25 m.
Raise-boring machines are capable of reaming raises with a sufficient
diameter and height to match any stoping or mine development requirements. Although the cost per cubic meter of rock removed is higher, this
type of development offers speed in advance, compared with conventional drill and blast methods. Raise boring is of particular importance for
mine ventilation. Decline development can be undertaken blind and with
increasing depths due to exhaust ventilation by raise-bored ventilation
shafts. A major advantage is their smooth-walled finish, which reduces air
resistance.
3.4.5.7 Fill Infrastructure
Fill masses are required to provide large-scale ground support, as well
as localized stability for pillar recovery. The key stages of a fill operation
for sublevel stoping are material and stope preparation, fill delivery or
P 49 SHAFT
JA 51 MP
JG 51 RAR
HD 50 RAR
J 53 Service shaft
J 52 Exhaust shaft
JF 5260 RAR
JC 53 RAR
J53 Sink
Hilton mine
orebody passes
Longitudinal section looking west
12 Level
12 B Sub
12 D Sub
11 A Sub
11 C Sub
10 Level
10 C Sub
8 Level
Legend
Developed Designed
pass
pass
Return
Air raises
IJ 52 HICAF JA 51 MP
IB 52 HICAF
JD 53 MP
OP/RAR
OP/RAR
JC 52 OP
IC 52 OP
HF 50 MP HD 50 MP
IG 50 OP/RAR
IG 50 N OP/RAR
JC 52 OP
5000 N
FIGURE 3.40
Hilton Mine vertical opening system. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)
2750 RL
IG 50 S OP/RAR
3000 RL
Planning and Design
87
88
Geotechnical Design for Sublevel Open Stoping
N52
Fill pass
S50
Fill pass
Screening Crushing
Conveyor
2468 m
KSOC
384 m
Fill passes
(2–4 m diam)
530-560
538
545
530
522
515
500
507
492
484
476
469
461
446
454
13C Sub
Multiple lift sublevel open stopes
15 level
19 level
FIGURE 3.41
Schematic of fill distribution system at Mount Isa Mines. (From Mathews, K.E. and
Kaesehagen, F., The development and design of a cemented rock filling system at the Mount Isa
Mine Australia, in E.G. Thomas, ed., Proceedings of the Jubilee Symposium on Mining with Backfill,
Mount Isa, Queensland, Australia, 1973, pp. 13–23. With permission.)
reticulation, placement, and drainage. Development for fill delivery and
reticulation is usually addressed during a stope block design. The options
may include fill delivery from a surface material station using raise holes or
boreholes, trucked to stopes via ramp access or from underground sources.
Underground fill reticulation is achieved by means of gravity feed or pumping to stoped-out areas. Conveyor belts, pipeline distributions, or standard
or ejection tray trucks can be used.
Fill reticulation for massive orebodies usually requires long-term development within the crown of an orebody (Figure 3.41). In such cases, crown
subsidence may threaten the stability of the development associated with
a fill system above an orebody. To minimize this likelihood, progressive
tight filling of stope voids is required, as the combined effect of unfilled
stope crowns can result in regional subsidence. Geological and operational
factors such as delaying filling can influence the rate of subsidence (Logan
et al., 1993).
Large unfilled voids as well as progressive stoping may cause dilation of
geological discontinuities, which in turn can be linked to rotation and sliding of large blocks within the crown of a deposit (Logan et al., 1993). This
localized block behavior may produce significant changes in the relative elevations along the strike of an orebody. Continued monitoring using precise
level-surveying techniques can be used to obtain an understanding and to
manage subsidence.
Planning and Design
89
3.4.6 Stope Production Scheduling
Scheduling is an essential component of the stope planning process, as it adds
a time dimension to all the functions within the process. Schedules specify
the sequence, timing, and allocation of resources to events that extend from
daily operations to the life of mine scenarios (Trout, 1997). Scheduling has
time frames that may vary from mine to mine. Mines with a shorter life
will have a different scheduling perspective to larger mines. In general, the
objective of production scheduling is to provide direction to the mine production personnel ensuring that established metal targets are accomplished.
Planning personnel must have an understanding of the overall production
targets and issues required in order to achieve desired outcomes within a
business plan.
In practice, mine scheduling is usually carried out either at a broad level
or, conversely, at a more detailed level. Production schedules are used to
establish the long-term strategic issues in conjunction with their economic
implications. Activity schedules are used to set out the details of how the
production schedules will be achieved (Trout, 1997). While the production schedules deal with a broad picture, the targets set within the activity
schedules must be compatible with the long-term scheduling goals.
The details included in a schedule change with the size of the source.
Small stopes generally require a schedule with shorter time intervals and
more detail (Trout, 1997). Usually, the geological information, and hence the
degree of confidence, increases for the activity schedules. The entire mineplanning team including geologists, surveyors, mining engineers, and production supervisors must be familiar with the targets set during all levels of
production scheduling.
In general, scheduling identifies critical production activities while providing the baseline means for monitoring progress and whether any deviation from the overall objectives is occurring. The schedule must identify key
or critical events to guide the mine management decision-making process.
The design status of each production source should be identified within a
production schedule.
Typical scheduled items include development and production targets,
capital and operating expenditures, equipment replacement, maintenance,
and diamond drilling. A document is required to document the critical issues
and assumptions of a particular production schedule. The stopes scheduled
to be extracted may also be represented in plan views and longitudinal sections on which the scheduled targets (development, stope extraction, and
filling) are clearly identified (Figure 3.42).
A number of generic processes are available for undertaking mine scheduling. These include manual techniques using computer spreadsheets, project management (critical path) approaches, and other methods available
in computerized mine-planning software in which all the interdependencies and constraints are taken into account (Trout, 1997). The procedures,
90
200 m
7000 N
6800 N
Om surface
6600 N
6400 N
Geotechnical Design for Sublevel Open Stoping
Scheduled
400 m
Extracted
600 m
Extracted
800 m
1000 m
(a)
Normal induced
stress (MPa)
(b)
>90
80–90
70–80
60–70
50–60
40–60
30–40
FIGURE 3.42
(a) Extracted and scheduled stopes and (b) induced normal stress following extraction, Mount
Isa Mines. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)
frequency of preparation, time periods, level of detail, format information,
and communication process for mine scheduling may differ between mine
sites (Trout, 1997).
3.4.6.1 Long-Term Production Scheduling
Production scheduling is the highest level of scheduling and provides a longterm view of the mining process by focusing on issues such as ore grade,
extraction sequences, and production quantities. Production schedules
typically extend over a number of years and are expressed in terms of ore
sources relating to stoping blocks (Trout, 1997). These schedules can extend
through to the life of a mine, depending upon which event comes first. The
items included in a scheduling exercise are long-term production targets,
fill, development, raising, and diamond-drilling requirements. Annual estimates for equipment replacement, capital, and operating expenditure may
also be determined. The most common restrictions imposed on scheduling
may include capital availability, expected life of the mine, infrastructure,
and equipment life.
3.4.6.2 Medium-Term Activity Schedules
The second level of scheduling undertaken in underground mining is called
medium-term activity scheduling. This schedule usually consists of a 2-year
91
Planning and Design
(or similar length of time) production period. Similarly to a long-term schedule, production targets, backfill, development, raising, and diamond drilling
requirements are considered within this schedule. However, the activities
are updated (using a rolling format) and issued every 3 months. Usually, a
1-year budget schedule is developed and adopted within a medium-term
activity schedule. This full-year forecast is a critical document that sets the
formal budget for the subsequent production year. The forecast is based on
preliminary stope designs, in order to ensure that the budget metal, capital,
and operating expenditure can be effectively achieved. Depending on the
size of the mine and the number of ore sources, mine size and number of
sources, the full-year forecast may be reviewed and updated each month.
Priorities are then determined to ensure that the budget targets are met.
3.4.6.3 Short-Term Activity Schedules
Short-term activity scheduling plays a tactical role while providing a detailed
schedule over a short time horizon. The activity schedule contains sufficient
details to allow underground personnel to plan and perform their work
(Trout, 1997). Usually, this schedule considers the production activities within
a 3-month period. It is updated and issued each month, primarily to assist
production personnel in identifying the short-term activities (day-to-day
mine operation) required to fulfill yearly budget targets. The short-term activity schedules are usually presented during a meeting between the planning
and production personnel, where stope preparation (stope access development, ground support, services installations, stope drilling) and production
issues (blasting, material handling, and filling) are discussed (Figure 3.43).
Stope production
phase
Preparation
phase
Extraction
phase
Filling
phase
FIGURE 3.43
A time-based representation of stope mining phases. (From Trout, P.L., Formulation and
application of new underground mine scheduling models, PhD thesis, The University of
Queensland, Brisbane, Queensland, Australia, 1997, 344pp.)
92
Geotechnical Design for Sublevel Open Stoping
3.4.7 Ventilation
A mine ventilation system is related to the magnitude and direction of air
movement through the various working places in the mine. The supply of air
is referred to as air distribution, and it is accomplished by adopting a ventilation circuit suitable for the particular mining method used for extraction. In
sublevel stoping, primary development openings such as shafts and ramps
are used for main airways for ventilation, while the individual levels can be
used as intakes and outlets using unidirectional air distribution.
Sublevel stoping mines are likely to have extensive workings on each level,
as well as between levels, and therefore require ventilation from combined
vertical and horizontal circuits. The stopes are designed to allow flowthrough ventilation between the sublevels connected by the stopes. The
overall objective is to supply fresh air to each level from a downcast pressure source, radiating outward and upward through the working places to
exhaust airways leading to upcast shafts (Figure 3.44).
In general, the airflow should be in an opposite direction to the stope
retreat direction, so that dust and fumes are kept away from the operators. Consequently, the ventilation design for a stoping block will consist
of access to fresh air, either from fresh air raises or a decline, as well as a
return air exhaust system. The preferred approach is to ventilate each stope
with a separate split of air, with the air introduced to the working places
from the lowest level. Separate exhaust openings may be required to prevent
contaminated air from entering other stopes in a stoping block. Ventilation
shafts and airways must be located and maintained in ground which will
not be caved and lost during the lifetime of the operation. In addition, short
Exhaust fan
Escape way
Decline
VR
Development
Stope
Intake
Return
Bulkhead
Security door
VR Vent raise
Production
Stope
Stope
Development
FIGURE 3.44
Schematic of primary ventilation, Konkola deep mining project, Zambia. (From Calizaya, F.,
Schematic of a primary ventilation network, pers. commun., 2013.)
Planning and Design
93
circuitry and dust hazard created by air leakage up or down partially filled
orepasses must be prevented.
3.4.8 Global Economic Assessment
A number of global design considerations must be analyzed and economically evaluated to arrive at the optimum design for a stoping block.
The outline of the orebody is determined by cutoff grade evaluations that
account for the cost of block development, mining cost, haulage, surface cartage, mineral processing, and general overhead costs. A financial model is
used to determine the viability by comparing the unit cost of all the steps
involved in mining and processing with the estimated revenue. This could
be an iterative process as, once the cost of development is included, some
stoping blocks may prove not to be economical. However, they may become
economic if development is carried out through those blocks to access other
more economic areas.
Thomas and Earl (1999) have described a computerized stope optimization
tool that can be applied in the strategic planning of underground stopes.
The technique can be used to generate an extraction sequence in conjunction
with an optimum stope configuration that maximizes the net present value
of an operation. The tool is used to generate inventories for a series of cutoff
grades, and the results are scheduled to produce net present value (NPV)
versus tonnage relationships.
3.5 Detailed Stope Design
Detailed stope design relates to the extraction of individual stopes within a
stoping block or global area. Detailed design is the process of establishing
an optimum extraction method for an individual stope, subject to a number of variables and constraints. Blasthole geometry, firing sequence, ground
support, ventilation, and economics are some of the key variables considered. The constraints include the orebody boundaries, the geological structures, any existing development and, in some cases, any adjacent fill masses
(Figures 3.45 and 3.46).
Figure 3.47 shows a typical process for taking an open stope from conceptual design through to production at the then WMC Resources, Australia
(Teasdale, 2001). The detailed design process begins when the geological
team undertakes detailed orebody delineation for a particular stope extraction. In-fill delineation drilling, mapping, sampling, and geological interpretations on a stope scale are then completed. The mine planning engineer
uses geological sections from a mine design package to do a preliminary
stope design, while the rock mechanics engineer completes a rock mass
94
Geotechnical Design for Sublevel Open Stoping
16B
16B
17D
18E
18E
18B
19C
19C
19A
19/L
19/L
FIGURE 3.45
Isometric view of the P446 stope in the 1100 orebody, Mount Isa Mines. (From Grant, D. and De
Kruijff, S., Mount Isa Mines—1100 orebody, 35 years on, in G. Chitombo, ed., Proceedings of the
MassMin 2000, Brisbane, Queensland, Australia, October 29 to November 2, 2000, pp. 591–600,
AusIMM, Melbourne, Victoria, Australia. With permission.)
characterization program, providing guidelines for stope stability, dilution
control, reinforcement, and blast sequencing.
At this stage, extraction factors that account for dilution as well as back
analysis of performance from any adjacent stopes are taken into account.
Drill and blast design is undertaken considering the equipment capabilities
to ensure that the designed stope shape is achievable. This is then followed
by an economic analysis that determines stope viability by considering the
break-even revenue cutoff figures including a calculation of net revenue versus total mining, concentrating, and overhead cost. Finally, a stope design
document that includes detail of the overall extraction philosophy, plans
of sublevel development, sections showing blasthole design concepts and
drilling and blasting parameters, ore- and waste-handling systems, ventilation, geology, rock mechanics, and overall firing sequence is issued to the
operating personnel.
All the topics included in a stope design document are interrelated. The
extraction philosophy provides a general overview of the design, safety,
95
Planning and Design
4500 N
1800 E
Q450
filled
4480 XC
Production
rings
4457 XC
Cutoff slot
N
Western
cutoff
slot
1800 E
DP
T
2
4500 N
4450 N
4450 N
Eastern
cutoff
slot
P442
filled
(a)
(b)
4500 N
Q450
filled
Q450
filled
Cutoff slot
Production
rings
1800 E
1800 E
4500 N
Production
rings
4500 N
4500 N
P442
filled
(c)
Cutoff slot
P442
filled
(d)
FIGURE 3.46
Plan view of several sublevels through stope P446 showing drilling layouts and adjacent fill
masses. (a) Extraction level, (b) mid height sublevel 18B, (c) mid height sublevel 18E, and (d) top
sublevel 17D. (From Grant, D. and De Kruijff, S., Mount Isa Mines—1100 orebody, 35 years on, in
G. Chitombo, ed., Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, October 29
to November 2, 2000, pp. 591–600, AusIMM, Melbourne, Victoria, Australia. With permission.)
96
Geotechnical Design for Sublevel Open Stoping
Drilling and sampling
Kriging and wireframe
Preliminary design
Final design
Survey pickup
Development and ground support
Ring design
Face mapping, geological mark-up
Geological wireframe
Production drilling
Blasting, mucking
CMS survey
Filling
Reconciliation
FIGURE 3.47
Typical process for open stope design, WMC Resources. (From Teasdale, P., Open stoping mining method of mining at WMC Resources Gold Business Unit operations, design process and
operating practices, MEngSc thesis, Western Australian School of Mines, Curtin University of
Technology, Kalgoorlie, Western Australia, Australia, 2001, 68pp.)
and production issues for a particular section of an orebody. Properly reinforced stope development is required to allow access for drilling, blasting,
and mucking. Development size is a function of the stoping method and
the equipment utilized. Development allows for the drill geometries to
be designed, as well as subsequent ring firers’ access to charge the rings.
Knowledge of the nature and stability of the adjacent fill masses is needed to
design cleaner rings or to avoid toeing of blastholes into the fill.
Geological considerations such as the presence of major geological discontinuities often influence the blasting sequences. Other factors considered are
the stress redistributions within and around a stope that are likely to control
falloff behavior on the exposed walls. In addition, the retreat direction of the
blasthole rings must take into account the stope ventilation network, with a
retreat direction into fresh air. Progress through a detailed design process
can be tracked using a stope control sheet that can be used to track progress
with preliminary design, production, and filled stopes (Figure 3.48).
97
Planning and Design
Stope Control Sheet
Stope name: _______ Orebody: _______ Upper level: _______ Lower level: _______
Task
Responsible
Development completed
Development superintendent
Rock mass characterization completed
Rock mechanics engineer
Survey pick-up completed
Survey department
Wireframe geology
Geology department
Ring design completed
Mine planning department
Ring grade
Geology department
Update database
Mine planning department
Drilling completed
Production superintendent
Production completed
Cavity monitoring completed
Production superintendent
Survey department
Filling completed
Fill superintendent
Stope finished
Update database
Mine planning department
Stope reconciliation note
Mine planning department
Update ore reserve long section
Geology department
Initial
Date
Comments:
FIGURE 3.48
A stope control sheet developed at Mount Isa Mines. (Courtesy of Mount Isa Mines, Mount Isa,
Queensland, Australia.)
3.5.1 Geological Information
The typical geological information required for stope design consists of
grade and tonnage, dilution factors, and delineation of the main geological features intersecting a stoping area. The initial information is usually
collected from the diamond drillholes intersecting the area of interest. This
information is used to create a conceptual three-dimensional orebody delineation design with ore tonnages and grades. Empirical extraction and dilution factors that account for the expected tonnage and dilution are issued
prior to the preliminary economic analysis. A 110/96 factor indicates that
up to 10% additional tonnes are expected from the stope. In addition, a 4%
reduction on the grade is also expected.
Major geological structures provide the greatest potential for large falloff in a stope void. Thus, information on the major geological structures
anticipated can be used to delineate potentially unstable zones adjacent to
an exposed stope wall. Unraveling and block release is possible along major
structures, resulting in a zone of disturbance. In some cases, failure can
98
Geotechnical Design for Sublevel Open Stoping
17D
Overbreak
80 m
S48 fault
zone
O434
stope
18B
Stope design
outline
FIGURE 3.49
O434 stope hangingwall failure, Mount Isa Mines. (From Logan, A.S. et al., Geotechnical
instrumentation and ground behavior at Mount Isa, in T. Szwedzicki, ed., Geotechnical
Instrumentation and Monitoring in Open Pit and Underground Mining, Proceedings of the Australian
Conference, Kalgoorlie, Western Australia, Australia, June 21–23, 1993, pp. 321–329, Balkema,
Rotterdam, the Netherlands.)
progress beyond the weak zone itself. If a stope design requires blasting to
the top of a fault or potential failure zone (by considering that such material has a high probability of failure), the design dilution factors are actually
increased. However, the problems related to poor fragmentation from fault
falloff may actually be minimized.
In other cases, stope designs attempt to leave weak faults in place by leaving a beam of good-quality material against a fault or potential failure zone
in order to improve the stability. An accurate assessment of fault location and
knowledge of the likely behavior and deformational characteristics of the
rock beam are required. The reduced fragmentation problems due to minimal
falloff must be balanced against the ore loss occurring within the rock beam.
A successful outcome during stope extraction involving a large weak zone is
not always guaranteed, even by leaving ore beams, as shown in Figure 3.49.
3.5.2 Development
The orebody characteristics and the type of equipment used are likely to
influence the locations as well as the final sizes and shapes of the stope
Planning and Design
99
FIGURE 3.50
Development access prior to stope drilling at the Mount Marion Mine, Kalgoorlie, Western
Australia.
development accesses. Geological control during development of the ore
drives is required to minimize undercut and blast damage at the orebody
boundaries. Geological mapping and orebody contact markup are undertaken at every development cut through a stope. This information is entered
into a computerized database that can be used for orebody delineation
purposes. Stope production drilling is facilitated when the development
drill drives exhibit straight walls and good floor profiles (Figure 3.50).
Development inside the stope, namely, cutoff slot and production blasting
access, does not require such tight control as does the development located
on the ore/waste contact, and so can be mined under survey control.
The length of the development rounds must be compatible with orebody
boundary variations along strike. Long round development may not be compatible with orebodies that pinch and swell along strike, as the chances of
hangingwall and footwall undercutting may actually increase. In addition,
the excavation size and shape must suit the equipment used during each task
within the stoping cycle. Strike drives and crosscuts must take into account
the dimensions and capabilities of development jumbos, longhole drilling,
and production mucking equipment. When possible, stope drill drives are
developed along the stope boundaries to limit the subsequent maximum
drill hole length during production drilling within the stope. This may also
100
Geotechnical Design for Sublevel Open Stoping
prevent the blastholes toeing into stope footwalls or hangingwalls. However,
depending on the orebody width, sometimes it is not always possible to
locate twin drill drives at the boundary of a stope.
3.5.3 Geotechnical Assessment
Geotechnical assessment for a stope design is carried out following completion of the strike and crosscut development within the stope limits.
Geotechnical data can then be collected from direct mapping of the exposed
walls within the stope development. These data are used to complement the
initial data collected from core logging of the exploration diamond holes
intersecting the rock mass within and around the stope. The data from
mapping, logging, and the information on major geological discontinuities and rock type variations provided by the geologists are fundamental
to the assessment of structurally controlled stope wall behavior. Figure 3.51
shows three major shear zones intersecting a planned stope design boundary. An initial interpretation from diamond drilling was confirmed with
ear
25A
he
ar
W6
0 sh
Recrystallized shale
Ur
qu
ha
rt s
25B
W
63
she
ar
Interpretation from
mapping and
diamond drilling
27C
Interpretation from
diamond drilling
28D
FIGURE 3.51
Interpretation of the main geological discontinuities on a stope scale, Mount Isa Mines.
(Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)
101
Planning and Design
direct access mapping within the upper portion of the stope (25A and 26B
sublevels). Experience with similar structures was used to predict a potential
failure zone in the stope crown (Logan et al., 1993).
The determination of stable stope wall dimensions is a critical aspect of the
geotechnical assessment for a particular stoping area. Experience has shown
that localized dilution as well as large block failures can be experienced in
poorly dimensioned (very large) stope walls. On the other hand, designing for
a worst-case geological scenario (small stopes) means that stope productivity
may be unnecessarily affected throughout the operation. In most mines, the
maximum stable stope wall length (or width) dimension is influenced by the
height of the sublevel interval chosen. As the dimensions for the sublevel interval are systematically applied throughout a design block, considerations of
stope wall stability are used to calculate the maximum permissible length or
stope width for a particular stoping scenario. Stope wall dimensions become
a very important economic parameter within individual stope design as they
also control the size of the exposed spans at the stope crowns.
The effect of external dilution (due to failures) or any unrecovered ore that
must remain in pillars required to stabilize large spans is a key factor that
requires consideration during stope size determination. The dimensions of a
maximum stable length or width for a particular stope area are usually determined using local experience or an empirical rock mass classification system
(Potvin et al., 1989). A geotechnical model of the maximum permissible stope
lengths (or widths) for a fixed sublevel interval can be established for each
particular stoping area (Figure 3.52). The model is based on geotechnical
Hangingwall stability (55° dip)
100
90
Floor to
floor
Permissible length (m)
80
70
50
12
15
22.5
30
45
40
Meters
60
30
20
10
0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
Hydraulic radius (m)
FIGURE 3.52
Maximum permissible stope length for different (floor-to-floor) sublevel intervals.
102
Geotechnical Design for Sublevel Open Stoping
mapping of stope development exposures and the localized stope delineation drilling. An iterative process is required to determine the optimum
stope dimensions, as different degrees of stability may be predicted along
and across the strike of an orebody (see Chapter 5).
Geotechnical assessment also requires numerical modeling of the blasting
sequences within the stope itself. This will determine the stress redistribution
in and around the stope area. The likely damage due to stress changes, either
compressive or tensile, in the stope brows, exposed hangingwalls, and adjacent pillars can then be predicted from back analysis in other similar areas.
3.5.4 Stope Design Philosophy
Stope wall falloff and its subsequent influence on production efficiency
is largely controlled by geology. Consequently, the stope design philosophy must consider the influence of any large geological feature intersecting a stoping area. The stopes must be designed to minimize falloff, rather
than to maximize direct short-term cost savings by utilizing existing
development that may force unfavorable positioning of blasthole rings,
cutoff slots, retreat directions, and sequencing of the blast within the stope
itself. Priority must be given to the analysis of the geological structures and
their influence on production blasting and stope wall stability. Short-term
savings on development may subsequently lead to poor fragmentation,
falloff, and production losses several orders of magnitude higher than the
savings on development.
3.5.4.1 Production Rings
Ring blasting establishes the location of blastholes in relation to the drill
drives, the orebody, and, most importantly, the planned stope outline. The
designed volume and shape of the stope to be blasted as well as the positions and shapes of the drill drives and production mucking horizons are
established for each ring section. Each individual ring design layout consists
of a section through the orebody. The information presented in a ring design
consists of the collar positions and the lengths and angles of the holes to be
drilled and blasted. The hole size, the amount of explosives used on each
hole, and the tonnes fired in each ring can also be indicated. In addition, the
lengths of any uncharged collars on the holes are also provided. A plan view
of the drill drives is used to determine the position of the cutoff slot in relation to the rings as well as the burden on the rings (Figure 3.53).
A flexible blast design is one that allows the engineer a choice of single
or multiple ring firings avoiding significant undercutting of stope areas.
Blasting of the initial rings around the cutoff slot creates enough room for
the remainder of the stope to be blasted. Considerations such as the level of
the induced stresses and production and access constraint requirements are
taken into account to determine the number of rings to be blasted together.
103
Planning and Design
Mount Isa Mines Limited, Mount Isa
16A sublevel
0.5m
Cross section view taken at ring 31
looking north at 6684.9 m
Machine
Hole diameter
Explosive density kg/m ANFO
Explosive density kg/m LD450
Explosive density kg/m LD425
Tonnes broken
Task code
Scale 1:250
Designed
Drawn
Checked
Approved
Number of holes: 5/4
Footwall
contact
Hangingwall
contact
1.5
m
1.5
m
m
m
12.
0°
3m
64.
1
5°
2.8
m
65.
13.
0°
2m
65.
13.
0°
7m
65.
14.
0°
2m
64.
0.8 m
1.5
1.5
0.8 m
1.5
16B sublevel
Average orebody width: 9.3 m
m
Rings 29–33
R39
0.5 m
R38
2.0 m
0.8 m
L
4
L
3
R30
L
4
R29
L
0.5 m
Plan view drilling layout
guide
SIMBA
70 mm
4.30
2.15
1.58
820
2005
Notes:
Dashed line represents the orebody outline
All collar and breakthrough positions are relative to
Orebody
All holes are blow loaded
Explosive densities shown are for blow loading
Hangingwall hole to be loaded with LD450
12 Orebody
Bench stope 12 C8
FIGURE 3.53
Typical section and plan view of drilling layout for bench stoping at Mount Isa Mines. (From
Tucker, G. et al., Bench stoping at Mount Isa Mine, Mount Isa, Queensland, Proceedings of the
7th Underground Operators Conference, Townsville, Queensland, Australia, 30 June–3 July, 1998,
pp. 135–147, AusIMM, Melbourne, Victoria, Australia. With permission.)
Important information such as the actual firing sequence, blasting results
(fragmentation, freezing of holes, misfires, etc.), and any stope wall failures
related to blasting must be recorded during ring blasting.
3.5.4.2 Diaphragm Rings
Diaphragm rings are used where there is a moderate to high probability of
fill exposure failure. Diaphragm ring design is complicated by issues such
as different drilling and blasting techniques, different exposure sequences,
varying stress regimes, and containment of anything from cemented to
104
Geotechnical Design for Sublevel Open Stoping
uncemented fill. Diaphragms are potentially unstable where undercutting
of the diaphragm by the main rings is experienced. This may occur due to
poor drilling resulting in hole deviation. In addition, failure may occur when
a weak geological structure intersects a diaphragm in an unfavorable orientation or when extraction from previous stoping has damaged the rock mass
within the diaphragm sufficiently to reduce stability.
Another factor that assists diaphragm design is accurate knowledge of the
backfill–rock interface. This knowledge would allow a proper determination
to be made of the diaphragm thickness, in cases of uneven and sometimes
overhanging fill masses. Stope surveys using the cavity monitoring system
must be conducted following stope completion. However, stope wall falloff
may still occur after the final stope survey, and probe drilling may be needed
to accurately determine the actual rock–fill interface.
3.5.4.3 Cutoff Slot Design
A cutoff slot is a very important element in a stope extraction sequence as
it provides a free face and the required void for the rest of the stope to be
blasted. The cutoff is created by the sequential enlargement of a long hole
winze (LHW) geometry or a raise-bored opening. The decision to use one
or the other is controlled by equipment availability, the height of the stope,
the position of the existing development, and the desire to minimize damage from blasting. In multiple-lift sublevel stopes, existing development
may be offset on alternative sublevels, making one straight raise-bored
hole impossible to accommodate (Rosengren and Jones, 1992). LHWs are
more flexible, but they limit the speed with which a stope can be brought
into production.
Stope ventilation requirements must also be considered as raise-bored
holes improve the initial stope ventilation circuits. In some cases, a combination of raise-bored and LHW can be used within a single stope design.
Figure 3.54 shows a stope design that incorporates an LHW at the western
lower boundary (27C–28D) with cutoff holes retreating east. The top section
of the stope was designed using a 1.8 m diameter raise-bored hole. Cutoff
holes retreat west from 25A to 27C.
3.5.4.4 Drawpoint Design
Production mucking can be carried out either longitudinally or transversely,
across the strike of an orebody. Longitudinal mucking requires exposure
of the loader under a retreating bench stope brow, while transverse mucking requires the use of fixed specialized drawpoint geometries that may be
located outside an orebody boundary.
In longitudinal mucking, the stability of a retreating brow is a function
of the orebody width, the nature and strength of the geological discontinuities, the blasting practices, and the induced stresses. Mucking is carried out
105
Planning and Design
East
West
Ha
ng
ing
wa
ll
25A level
Raisebore
26B level
Raisebore
27C level
LHW
28D level
FIGURE 3.54
Cross section view showing a cutoff slot design parallel to a hangingwall to prevent blast damage. (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)
by remote control because of the dangers of rock falls near the brow. The
operator is positioned in a safe location just outside the stope and in constant
visual contact with the loader. The loader is driven to the muck, loaded, and
returned to the safe area under remote control. The operator then boards the
loader and completes the mucking cycle under manual control.
Longitudinal mucking of highly stressed, narrow, and long orebodies
(where the muck is typically thrown up to 20 m away from the brow) is well
suited to teleoperated mucking. Mucking is carried out from transportable,
air-conditioned control stations that may be located up to 300 m from the
stopes (Figure 3.55). The operator is seated on a comfortable chair that has
joy stick control in each arm rest. Color video cameras are mounted at the
front and rear of each loader and the images are continually transmitted back
to the operator via a radio link. Teleoperated mucking provides improved
106
Geotechnical Design for Sublevel Open Stoping
FIGURE 3.55
A modern cabin for teleoperated mucking. (From McHugh, C., Introduction of autonomous
loaders to Olympic dam operations, Australia, Proceedings of the Ninth Underground Operators
Conference, Perth, Western Australia, Australia, 7–9 March, 2005, pp. 127–132, AusIMM,
Melbourne, Victoria, Australia. With permission.)
occupational health for the operators, especially in highly stressed, seismic
conditions. In addition, increased production rates, greater tolerance of hot
and dusty conditions, and a reduced loader fleet have been accomplished
with this method (Villaescusa et al., 1994).
For transverse mucking, a number of factors must be considered during
drawpoint design, including the size of equipment, tramming distances
from access drives, as well as gradient and orientation with respect to a stope
boundary. The drawpoint dimensions must be sufficient to suit the equipment, but kept as small as possible to minimize instability.
3.5.5 Stope Design Note
A stope design note covers many factors involved in the development of, and
production from, a stope (Table 3.3). Technical presentations are required to
encourage technical input from all the members of the design team (geology, rock mechanics, planning, operations, and management). They usually occur twice within the design process, at the conceptual design stage
and prior to the issue of the final drill and blast design. Feedback from both
meetings should be incorporated into the final stope design.
Once a final stope design status has been achieved, the blasthole design is
undertaken by considering the production rigs that will be used, the ore limits, the survey pickup of the access development, the extent and sublevels of
the stope, as well as the ring burden and toe spacing. The ore limits are usually updated in accordance with the completed stope development. A scaled
floor plan showing details of the latest survey information including any
vertical openings and status of surrounding stopes is provided to assist
drilling. Locations of hangingwall, footwalls, cutoff slot detail, and locations
of the production rings are also included (Figure 3.56). A long section that
Planning and Design
107
TABLE 3.3
Stope Design Presentation Issues
Geological structures
Stope access and development requirements
Ore passes, loading bays, etc.
Stope cutoff location
Selection of drill rig and hole size
Selection of explosive type
Blasting sequence
Stability issues, ground support requirements
Stress redistributions assessment
Fill requirements or permanent pillar demands
Production schedule
Ventilation requirements
Detailed economic analysis
includes a schematic view of the stope cutoff raise, the cutoff slot, the production rings, and the trough undercuts is also completed. This section helps to
explain the stope design philosophy, and becomes a useful tool during drilling and blasting of the stope. Table 3.4 lists a number of issues that should be
considered during stope design.
3.5.6 Stope Firing Sequences
The actual firing sequence used to extract individual stopes is likely to influence the stress redistribution as well as blast-induced damage within a stope.
Stress and blast-induced falloff within a stope boundary may lead to poor
mucking performance during extraction. Although falloff resulting from
stope firing is not the only source of poor fragmentation, it can be minimized
by avoiding excessive undercutting of the stope walls. Stope undercutting is
usually linked to single-lift stopes (Figure 3.57). As a guideline, undercutting should not be undertaken when the stope is well advanced, and should
never be attempted in poor ground where large-scale structures are present. Unfortunately, in single-lift stopes, where in most cases a cutoff slot is
not available, undercutting is required by the method, regardless of the rock
mass conditions.
A number of design options can be used to reduce stope undercutting
including firing the cutoff slot to the full height of the stope before blasting of the main rings commences. This can be followed by the sequential
blasting of the main rings to the full stope height (Figure 3.58). The objective is to reduce the number of stope faces exposed, thereby reducing the
potential for time-related structurally controlled falloff. Undercutting of
the main rings can be avoided by designing the troughs to be blasted with
coinciding faces.
108
Geotechnical Design for Sublevel Open Stoping
6750 XC
16B
16A
Bench limit 6730 N
Bench limit 6730 N
13C8 Sill drive
N
12C8 Sill drive
13C9 Sill drive
6700 N
12C9 Sill drive
11C9 Sill drive
6700 N
6650 N
6650 N
Bench limit 6620 N
6600 N
Bench limit 6620 N
6601 XC
6600 XC
Note:
Bottom sill is shown
to the left
Revision
6600 N
Mine design
12C8 bench stope
Floor plan
16B-16A
Scale 1:500
FIGURE 3.56
Floor plan of 12CB bench stope showing cutoff slot position and main rings, Mount Isa Mines.
(From Tucker, G. et al., Bench stoping at Mount Isa Mine, Mount Isa, Queensland, Proceedings
of the 7th Underground Operators Conference, Townsville, Queensland, Australia, 30 June–3 July,
1998, pp. 135–147, AusIMM, Melbourne, Victoria, Australia. With permission.)
Planning and Design
109
TABLE 3.4
Stope Design Checklist
Location, orientation, and strength properties of large-scale geological structures
Size of existing development and suitability for available drilling rig
Additional development requirements, size, shape, and gradient
Ground support requirements for development and stope walls
Equipment needs for development including drilling, mucking, charging, and
ground support
Water drainage
Tramming distances and alternate ore and waste passes
Emergency escape routes during development and production
Drill drive layout, blasthole design, and firing sequence
Ring firers’ access to stope
Drawpoint brow location and ground support requirements
Ventilation requirements during development and stope production
Bomb bays for storage of oversized rocks and secondary blasting
Explosive types for development and production blasting
Location, size, and orientation of pillars
Overall rock mass (and fill mass) stability of the area prior to, during, and after
stope extraction
Detailed scheduling of stope development, production blasting, and filling
Cost comparison of alternative designs
Fill requirements including fill passes, reticulation, and delivery to stope
Continuing stope performance monitoring during extraction
Undertaking stope performance review after stope extraction
A stope firing sequence also determines the rate of exposure of the main
geological discontinuities intersecting a stope. Rapid exposure of a large
fault may occur after mass blasting or after progressive firing to a fault. Such
exposures may not allow sufficient time for gradual stress relief. If the orientation of the stress field is unfavorable, large shear stresses may result inducing local and regional fault movements leading to stope falloff.
3.5.7 Production Monitoring
Regular inspections of a producing stope are required, especially after each
firing, in order to monitor wall, crown, and drawpoint conditions. Any significant rock noise, falloff, or underbreak should be documented. In addition,
dilution exceeding more than 10% should be reported so that the actual stope
grade can be adjusted accordingly. Geologists should conduct drawpoint
investigations to estimate the grade of the ore being produced. Secondary
blasting of oversized rocks and hung-up drawpoints may be required. In
some cases, a bomb bay may be available for stockpiling oversized rocks and
undertaking secondary blasting.
110
Geotechnical Design for Sublevel Open Stoping
33
2
1
2
FIGURE 3.57
Stope wall undercutting within a stope-firing sequence. 1–3 indicate blasting sequence for a
single stope.
Broken ore is mucked conventionally when the drawpoints are full, but it
is sometimes required to remote muck the last ore remaining on the floor of
a stope, especially in large flat-bottomed stopes with retreating drawpoints.
Significant disruptions to mucking productivity can occur when excessive delays are experienced during a stope extraction. Stopes left open over
long periods of time may be influenced by time-dependent regional fault
behavior. Stress redistribution, production blasting, and backfill drainage
from adjacent stopes are likely to influence stope stability over a period of
time. Blast damage and the effects of water from backfill can be transmitted
along common fault structures intersecting a number of stopes. Instability
may create difficult remote mucking conditions due to large-sized material falling off into the stope. These delays (stope production tails) actually
extend the stope life, which in turn may contribute to more overbreak and
more mucking delays.
3.5.8 Ventilation
Stope ventilation is required during stope development and during stope
production. Ventilation during development requires auxiliary fans that
are used to force ventilate before a circuit is established. In steeply dipping
orebodies with a single ramp access, the fresh air usually flows through
the access ramp, where it can be force ventilated to the crosscut and ore
drives using auxiliary fans. The fans are equipped with flexible ventilation
111
Planning and Design
Ring
blast
ing
Stop
e vo
id
17 level
18B sublevel
Ring
Blasted
4
3-1-83
5 and 6 12-1-83
6 and 7 19-1-83
19C sublevel
FIGURE 3.58
Full stope height blasting with matching trough undercut geometries to minimize undercut.
(Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)
bags running along the crosscuts and the ore drives during development
and production drilling. Once a stope raise is blasted and a ventilation circuit is established, the air is exhausted by means of specialized drives connected to return air raises. Stopes with interlevel connection are usually
ventilated with air introduced in the lowest level and exhausted through
the top level.
3.5.9 Financial Analysis
The estimated cash per tonne of extraction reserves is calculated using the
delineated mining reserve (tonnes and grade), the metal prices, and the
extraction and dilution factors expected. The total cash profit (or loss) is
determined using a proper ore value model suited to the particular economics of a mine site. The input factors may include tonnes mined, grades
112
Geotechnical Design for Sublevel Open Stoping
and metal prices, mining, milling, smelting, overheads and royalties, and
exchanges rates.
In periods of excess mining, hoisting, and milling capacity, the total net
cash revenue can be increased by mining marginal stopes or marginal ore
within stope boundaries. Marginal ore can be included within a stope design
provided that little or no extra cost (no excessive extra development or additional reinforcement, etc.) will be incurred. An individual stope should be
extracted if it can return a positive total net cash revenue after covering the
costs of the remaining work required for extraction. Specific stopes may not
break even but may be sufficiently advanced in terms of development and
ground support to warrant a reduction in the break even value.
4
Rock Mass Characterization
4.1 Introduction
A rock mass is a three-dimensional discontinuous medium that can be
thought of as an assembly of potential blocks that can be disaggregated by
the excavation process. The size distribution, shape, and degree of interlock
of the blocks are functions of the distribution and nature of usually at least
three main discontinuity sets. Rock masses are rarely uniform or isotropic;
even within the confines of a design area, there are likely to be major geological structures, significant changes of lithology, and a prevailing anisotropy.
The nature and degree of this anisotropy and the heterogeneity of the rock
mass properties are likely to exert considerable influence on the extent of
damage to, and dilution from, the final stope walls.
During the last 35 years or so, a great deal of effort has been devoted to the
characterization of discontinuity networks and to modeling them quantitatively (Call et al., 1976; Hudson and Priest, 1979; Villaescusa, 1991; Brzovic,
2010; Cepuritis, 2011a). Systematic collection of geotechnical information in
conjunction with an appreciation of rock mechanics and geological factors
are essential in planning and designing stable stopes. The structural data
are initially utilized in the design of ground support configurations for the
stope infrastructure, including access drives, crosscuts, and drawpoints.
These data are used to develop an understanding of the various structural
domains within an orebody, which can be used to predict the likely wall
behavior during stope extraction. An optimized stope extraction sequence
can be determined from this information.
Some of the most important geological factors influencing a rock mass
are shown schematically in Figure 4.1. The main features include the
following:
1. Intact rock: This is the solid material between the discontinuities.
Failure modes may involve failure of intact rock bridges.
2. Rock stress: The vertical stress caused by the weight of overlying
strata and the horizontal stress caused by tectonic forces within the
earth’s crust.
113
114
Geotechnical Design for Sublevel Open Stoping
Termination
Wall
strength
Large
discontinuity
Infill
Waviness
or
planarity
g
in
ac
Sp Discontinuity
In situ stress
ce
en
ist
rs
Pe
Block size
(intact rock)
σ1
σ3
σ2
Dip and
dip
direction
set
Water seepage
FIGURE 4.1
Some of the major geological factors influencing the engineering behavior of a rock mass.
3. Number of discontinuity sets: A discontinuity is a mechanical break (of
geological origin) within the rock mass. Because of geological process, discontinuities are formed in sets. In addition, a rock mass may
be divided by single, large-scale geological discontinuity.
4. Discontinuity orientation: The three-dimensional attitude of a discontinuity in space is measured using dip direction (azimuth from north
to the steepest line on the plane measured in a horizontal plane) and
dip angle (the angle that the steepest line makes with the horizontal
plane).
5. Discontinuity frequency and spacing: The frequency is the number of
discontinuities per unit distance in space. It is the reciprocal of the
spacing and can be defined globally for all discontinuity sets or by
individual sets.
6. Discontinuity persistence and termination: Persistence is the observed
trace length of a discontinuity within a rock mass. It provides
a measure of areal extent or penetration for each discontinuity.
Termination of a discontinuity can be either in intact rock or against
another discontinuity.
7. Block shape and size: The shape and size of an intact rock block within
a rock mass. The block size is a function of the number of sets, frequency, orientation, size, and termination of the geological discontinuities present within the rock mass.
8. Discontinuity roughness and planarity: Inherent surface roughness and
planarity (or waviness) with respect to the naturally occurring mean
Rock Mass Characterization
9.
10.
11.
12.
115
plane defining a discontinuity. Both roughness and planarity contribute to shear strength.
Aperture: Perpendicular distance across adjacent walls of a
discontinuity.
Wall strength: Compressive strength of adjacent walls of a discontinuity. Usually lower than the rock block strength due to alteration
of the walls (by migrating fluids). Constitutes a key component of
discontinuity shear strength if the walls are in contact.
Infill: Material that separates adjacent rock surfaces of a discontinuity. The material may be weaker (usually) or stronger than the adjacent rock walls.
Water seepage: Moisture or water flow within individual discontinuities or through intact rock.
Some aspects of rock mass structure, strength, and stress can be measured
by the logging of drill cores, directly by the structural mapping of exposed
faces, or can be deduced from indirect measurements made using geophysical techniques. At most mining sites, conventional geological mapping is
completed for all horizontally developed excavations, while geotechnical mapping is restricted to areas of specific concern where greater characterization of the rock mass is required. However, the largest amount of
information in terms of areal coverage across an orebody is collected from
diamond drilling during the several stages of the orebody delineation process. This process includes data collection from widely spaced surface drilling programs and any subsequent underground drilling for detailed stope
design purposes.
4.2 Characterization from Exploration Core
Diamond drilling, with geological core logging, is the most commonly used
method for orebody delineation. Information obtained from drill intersections is extrapolated hole-to-hole using geological assumptions to provide
estimates of lithological boundaries, alteration, weathering, hydrogeology, orebody size, shape, grades, continuity, tonnage, and some geotechnical characteristics (Figure 4.2). The advantages are the depth to which
the information can be obtained, and a relatively routine data analysis and
interpretation. Holes near the center of the mineralization provide critical
information for stope design, while holes near the periphery are critical
to the design of mine infrastructure such as shafts, access declines, and
crusher chambers.
116
Geotechnical Design for Sublevel Open Stoping
FIGURE 4.2
Core details showing shear zones and faults intersecting orebodies at depth.
2200 RL
9000 E
Orebody
boundary
8800 E
8600 E
2000 RL
HW intercept
FW intercept
FIGURE 4.3
Longitudinal section view showing typical exploration drillholes.
Another advantage of geological logging is that characterization encompasses every drilled hole through a geological deposit (Figure 4.3). If some
relevant geotechnical parameters are collected within this program, an
extensive and representative database within and across the immediate
boundaries of an orebody can be established. Parameters such as discontinuity linear frequency and rock mass classification data can be used to
determine spatial variations in rock quality across an orebody. A perceived
disadvantage is that a large number of individuals may perform the geological and geotechnical logging, introducing the chance of bias arising from
different practices and interpretations. In addition, some of the drilling data
117
Rock Mass Characterization
may be collected from small-sized unoriented core that is not ideally suited
for geotechnical logging.
The approach suggested here is to carry out geotechnical logging on a
number of selected holes within each exploration ring as part of the orebody
delineation drilling program. The approach does not require oriented core to
carry out the geotechnical logging, with the level of detail required during
geotechnical investigations usually depending upon the stages of a particular project (mine prefeasibility, feasibility, etc.). Estimates of the likely stable
stope sizes and shapes, dimensions of regional pillars, the best locations for
underground infrastructure, and reinforcement schemes can be provided by
such investigations. Figure 4.4 shows a cross section of a typical exploration ring where horizontal, steeply inclined, and steeply declined holes were
logged systematically across the orebodies.
Experience has shown that the choice of data format is important to
facilitate the subsequent stages of the stope design process. In some cases,
the computerized geological and geotechnical data are meshed as a threedimensional model. In some mines, such a geological/geotechnical model
is not available, and the information is presented on paper plans/sections
from which it can be digitized for printing purposes only. It is important
that the initial geological model built is not only for grade control purposes
but also intended for use in predicting the likely engineering performance
of the excavations. The following sections describe a procedure that can be
followed to carry out a rock mass characterization program from routine
underground orebody delineation drilling.
Surface holes
Underground
drilling
Core logged for geotechnical
purposes
FIGURE 4.4
Cross-sectional view showing typical underground exploration ring.
118
Geotechnical Design for Sublevel Open Stoping
4.2.1 Drilling Layout Design
The drilling layout is based on information obtained from the initial surface
delineation and a subsequent geological evaluation program. Section spacing
and the distance between intersections down dip are determined based on
local orebody complexity and the experience of the site geologists. An understanding of the associated risks incurred by the inability to interpret the
orebody geometry and grade must be developed. Table 4.1 shows three stages
of diamond drilling and the required confidence levels associated with each
stage. Detailed stope design usually requires a 20 m average spacing between
sections. With such a drilling spacing, the number of holes per stope is likely
to be sufficient for an effective rock mass characterization process.
Development access must be maintained ahead of production in order to
have sufficient time and locations to undertake the proper orebody delineation. In most cases, development is maintained at least a year ahead of production, providing enough time to complete the task.
4.2.2 Underground Drilling
Following the completion of the drill layout design, the holes are drilled from
a suitable underground access (footwall access or ramps, see Figure 4.5). In
most cases, the borehole diameter used during the underground delineation stages ranges from AQ to BQ (27–40.7 mm core, see Table 4.2). These
hole sizes may not always be appropriate for the collection of geotechnical
parameters without proper correction for the mechanical effects of drilling
upon the core.
4.2.3 Core Transfer to Surface
Following the completion of drilling, the recovered core must be transferred
to surface to a core shed or similar, for logging. Significant damage to the
TABLE 4.1
Typical Drill Spacing during Orebody Delineation
Orebody Nature
Tabular
Stope Design
Stage
Feasibility
Block design
Detailed design
Structurally Complex
Drill Spacing
(m × m)
Confidence (%)
Drill Spacing
(m × m)
Confidence (%)
80 × 80
20 × 40
20 × 20
50
70
90
80 × 80
20 × 20
10 × (10 or 20)
50
80
90
Source: Teasdale, P., Open stoping mining method of mining at WMC Resources Gold
Business Unit operations, design process and operating practices, MEngSc thesis,
Western Australian School of Mines, Curtin University of Technology, Kalgoorlie,
Western Australia, Australia, 2001, 68pp.
119
Rock Mass Characterization
FIGURE 4.5
Delineation drilling prior to detailed stope design.
TABLE 4.2
Nominal Hole and Core Diameters from
Wireline Drilling
Drill Size
AQ
BQ3
BQ
LTK60
NQ3
NQ
HQ3
HQ
PQ3
Nominal Core
Diameter (mm)
Nominal Hole
Diameter (mm)
27
33.5
36.5
43
45.1
47.6
61.1
63.5
83.1
48
59.9
60
60
75.7
75.8
96.1
96.1
122.6
recovered core can occur at this stage due to mishandling of the core trays.
This damage must be minimized so that an accurate estimation of the geotechnical parameters can be facilitated.
4.2.4 Drill Core Logging
The geologists, geological technicians, or geotechnical engineers log the
recovered drill core. The logged data are (manually or computerized) entered
into a geological database. Mineralized zones are identified and prepared for
assaying. The logged core is then photographed, split, bagged, and sent to
120
Geotechnical Design for Sublevel Open Stoping
FIGURE 4.6
Core splitting within an ore zone and immediate boundaries.
be assayed (Figure 4.6). If, at this stage, geotechnical logging has not been
undertaken, critical information regarding the mechanical behavior of the
rock mass will be lost permanently. Therefore, it is strongly recommended
that laboratory assay data are obtained after all geological and geotechnical
logging are completed.
Geotechnical logging must be carried out over lengths of 1 m for at least
the first 5–10 m immediately outside an orebody boundary. Logging intervals within an orebody are dependent upon the geological split defined by
the general geological interpretations (Figure 4.7). Geotechnical logging
must include interpretation and identification of major structures likely to
form a discrete failure surface.
4.2.5 Geological Database
Most mines have some type of database system, with data entry efficiency
ranging from manual to highly computerized digital drillhole logging systems. Following the completion of data entry into a geological database,
two-dimensional sections showing the geological logs and assays can be
displayed on computer screens or paper plots.
4.2.6 Interpretation of the Orebody and Main Geological Features
Based on the geologist’s experience and interpretative skills, orebody contacts and the main geological features such as faults, dikes, and shear zones
are established between the boreholes on each section. This step is the
121
Rock Mass Characterization
DDH
Orebody
Logging each meter
(for 5–10 m)
DDH
Logging by
geological
split
Logging each meter
(for 5–10 m)
Orebody
Underground
access
FIGURE 4.7
Recommended core logging intervals across an orebody.
most important in terms of predicting the mechanical behavior of the stope
walls and controlling the geological dilution (see Chapter 8). The quality
of the information is important as additional data are not easily collected.
The geologist must decide whether or not enough geological and assay
information is available for an adequate interpretation of the mineralized
zones and the main geological features. If a clear geological interpretation
is not possible, then additional underground drilling may be undertaken
(Figure 4.8). A second phase of underground delineation drilling (spacing
between sections and down dip as close as 10 m) may be needed to facilitate the interpretations. Information from locations near the center of the
orebody can be used for stope design, while information from the orebody
periphery can be used to design infrastructure, such as shafts, ramps, and
other related vertical infrastructure.
4.2.7 Orebody Meshing in Three Dimensions
Once the orebody and the main geological features have been sufficiently
sampled, the resulting shapes can be digitized in the two-dimensional paper
sections. After that, the final three-dimensional orebody shape and volume
can be established using computerized meshing tools. Computer manipulation and visualization of the meshed ore zones and controlling geological
features can be used to establish geological and geotechnical models of a
stoping block in three dimensions (Cepuritis, 2011a).
4.2.8 Problems with Data Analysis
An orebody delineation process usually produces information that flows in a
linear and sequential fashion. As the logging is undertaken, new information
122
Geotechnical Design for Sublevel Open Stoping
FIGURE 4.8
Two stages of drilling (global and detailed) for orebody delineation and characterization.
is being added to a geological database. However, very rarely a geological
or geotechnical model, provided one actually exists, is centrally updated as
soon as new information is obtained through data exchange between geology and mine planning. Relevant data required for long-term planning and
detailed design may not be made available on time to the mine planning
engineer. Data manipulation and visualization systems to update, access,
retrieve, and display geological and geotechnical information with minimal
effort are required (Cepuritis, 2011a).
4.3 Analysis of Logging Data
4.3.1 Discontinuity Linear Frequency
Back analysis of unsupported hangingwall performance in open stoping
carried out by Baczynski (1974) indicates that the number of discontinuities per meter within the first 3–5 m of a stope wall usually has a major
control on the behavior of an exposed opening, including dilution control. The number of geological discontinuities is recorded for every split
123
Rock Mass Characterization
of core logged and then manipulated to determine the linear frequency
per meter. In doing this, attention is required to identify and discount
fractures caused by the drilling process or core handling. This process
can be subjective, but most natural discontinuities have distinguishing
characteristics such as mineral coating, while artificially broken core often
has a rough, jagged appearance. Figure 4.9 shows a core logging sheet in
which a common rock quality designation (RQD) data collection sheet has
been modified to include information on discontinuity linear frequency.
The rock mass class ranges have been established by back analysis of
unsupported stope spans at Mount Isa Mines (Baczynski, 1974; Villaescusa
et al., 1992).
Discontinuity linear frequency is defined as the number of geological discontinuities per meter of a borehole through the rock mass. In three dimensions, the linear value depends on the orientation of the line with respect to
the structural discontinuity network. The linear frequency can be calculated
for a single joint set or a number of combined sets since the total number of
joints encountered along a line is additive. It is calculated from
lL =
nT
LT
(4.1)
where nT is the total number of discontinuities intersected by a borehole of
total length LT.
Hudson and Priest (1983) have established that variation in the discontinuity linear frequency value, λL, when calculated in different directions in
space, is a function of the existence of any anisotropies or preferred discontinuity orientations.
The discontinuity frequency within the first 5–10 m adjacent to a stope
hangingwall or footwall is calculated for each logged hole. The data for all
the holes intersecting an orebody can then be interpolated and represented
as a contour plot on a longitudinal section view (Figure 4.10).
Interpolation techniques such as kriging can also be used to display discontinuity frequency data, which can be used to predict ground behavior
following a core logging program. A kriged model of discontinuity frequency for an orebody can be produced using equivalent kriging weights.
Figure 4.11 shows variograms of bedding plane frequency calculated from
the closely spaced drilling fans in some of the orebodies at the George Fisher
Mine in Mount Isa. A strong anisotropy ratio (across versus along bedding)
was found for these orebodies, and the equivalent kriging weights used were
based on a strong 9:1 anisotropy ratio. The advantage of this type of analysis
is that estimates of ground behavior for an entire deposit can be made using
a number of commercially available software packages. The estimated conditions can be predicted using geostatistical block model data and displayed
on cross sections or plan views (Figure 4.12).
m
m
m
Fair
12
Poor
17
Discontinuity linear frequency
Very
Exc
Good
1.5good 4
7
FIGURE 4.9
Discontinuity frequency and RQD logging sheet.
Depth (m)
Lithologic log
Lithologic, Linear Frequency and RQD Logs– Project Name
Very
poor
Rock quality designation
Date
Page
Very poor
Poor
Fair
Good Exc
10% 20% 30% 40% 50% 60% 70% 80% 90%
By
124
Geotechnical Design for Sublevel Open Stoping
125
Rock Mass Characterization
2
17
1
15.3
13.6
11.9
10.2
8.5
6.8
5.1
3.4
1.7
0
FF/m
FIGURE 4.10
Longitudinal section view showing contours of discontinuity linear frequency and large scale
geological discontinuities for the first 5 m into an orebody hangingwall.
Variogram/(mean + 4.70)2
0.28
Cross bedding
0.24
0.20
Along bedding vertical
0.16
Along bedding horizontal
0.12
0.08
0.04
0.00
0
10
20
30
40
50
60
70
80
90
100
Distance (m)
FIGURE 4.11
Relative variograms of bedding plane frequency.
4.3.2 Rock Quality Designation
The concept of RQD as described by Deere (1964) is a quantitative index
based on core recovery in which the measure of the quality of the core is
determined incorporating only those pieces of intact sound core greater than
a threshold value tc in length. This value is generally twice the core diameter
126
Geotechnical Design for Sublevel Open Stoping
Color
Condition
Very good
Fair
Very poor
Disc/meter
1.5–4
7–12
>17
Color
Condition Disc/meter
Good
4–7
12–17
Poor
FIGURE 4.12
Geostatistical display of bedding plane frequency for a number of orebodies (grid is
100 m × 100 m).
dimensions, and shorter lengths of core are usually ignored. RQD can be
formally defined as
RQD = 100
n
ÂL
i =1
xi
(4.2)
T
where
xi is the length of ith length greater than the threshold value tc
n is the number of such intact lengths greater than tc
LT is the length of the borehole along which the RQD is calculated
The original concept of RQD was based on data from NQ size core (Table 4.2)
with tc = 100 mm. However, core from underground exploration drilling is
typically smaller than NQ, and such core is likely to be more sensitive to
drilling and handling conditions than larger diameter core. Consequently,
the threshold value used in the evaluation of RQD should reflect the
sensitivity to core diameter.
Although the global statistics and distributional nature of the RQD values per stope surface (hangingwall, orebody, or footwall) can be determined
(Figure 4.13), the mean values may not be relevant for individual stope
designs. The spatial variability of the RQD values needs to be taken into
127
Rock Mass Characterization
N = 719
X = 83
SD = 19.5
Max = 100
75% = 97
50% = 91
25% = 79
Min = 0
25
Observed frequency
20
15
10
5
0
0
15
30
45
60
75
90
100
Rock quality designation (%)
FIGURE 4.13
Histogram of RQD values for a hangingwall surface.
account when designing individual stopes. As shown in Figure 4.14, stope
outlines can be superimposed on RQD contours, allowing local RQD values
to be determined and used for design. The data for Figures 4.13 and 4.14 were
collected for the first 5 m surface of the hangingwall at the Kanowna Belle
Gold Mine, Western Australia.
Priest and Hudson (1976) formulated the theoretical RQD as an integration
of the probability density function of discontinuity spacing. If the spacings
are negative exponentially distributed along a borehole axis, the RQD values
can be approximated by
RQD t = 100 e(
-lL t c ){lL t c +1}
(4.3)
where
λL is the linear frequency along a borehole of total length LT
tc is the threshold value
Equation 4.3 provides a theoretical link between RQD and the linear
frequency, and provided the global spacing is negative exponentially
distributed, it can be used to give a reasonable estimation of the actual
RQD values.
A practical alternative is to use the empirical correlation between linear
frequency and RQD initially suggested by Baczynski (1980). Following a
similar approach and based on the data collected using the logging sheet
in Figure 4.9, the linear frequency and the RQD values can be calculated for
128
Geotechnical Design for Sublevel Open Stoping
100
90
Block A
10000 N
80
Block B
70
60
9800 N
50
Block C
40
9600 N
30
20
20400 E
RQD
20200 E
9400 N
19800 E
19000 E
0
20000 E
Block D
10
FIGURE 4.14
Longitudinal section view of the Kanowna Belle Mine showing contours of RQD values and
individual stope outlines (grid is 200 m × 200 m).
each meter split of core logged along the axis of a borehole. Significant relationships can be found between the data sets using linear, polynomial, and
logarithmic fits. Figure 4.15 shows an empirical best fit using a polynomial fit
to a typical set of data as follows:
RQD = 100 - 6 {lL }+ 0.08 {lL }
2
(4.4)
where
RQD is the calculated rock quality designation
λL is the observed linear frequency along the borehole axis
As noted earlier, the initial guidelines for RQD calculation developed by
Deere (1964) were based on core logging of NQ diameter core. Experience
indicates that the larger the diameter of the core, the less likely the influence
of drilling damage on core fractures. RQD values calculated from smalldiameter core typically used underground may be affected by the mechanical disturbance from drilling and effectively underestimate the actual
ground conditions. Therefore, careful consideration must be given when
using small-diameter core to calculate RQD values. An example of logged
data from two closely spaced holes on the same area (1.5 m apart) at Mount
129
Rock Mass Characterization
100
90
RQD = 100 – 6LF + 0.08 (LF)2
80
RQD
70
60
50
40
30
20
10
0
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Discontinuity linear frequency (J/m)
FIGURE 4.15
An empirical relationship between discontinuity linear frequency and RQD.
Isa Mines indicated that on average, up to two additional breaks per meter
can be expected when BQTK rather than LTK60 drilling is used as shown in
Figure 4.16. The effect of the core size is evident as an increased discontinuity frequency along the smaller-diameter core axis. Figure 4.17 shows two
empirical relationships between RQD and linear frequency calculated from
different core sizes and orientations at the New Celebration Mine, Western
Australia (Cepuritis, 1987).
Bedding plane breaks/meter
25
LTK60 (45 mm)
Very
poor
BQTK (38 mm)
17
Poor
12
Fair
Good
7
Very 4
good
1.5
Exc 0
0
10
20
30
40
50
60
Drillhole depth (m)
FIGURE 4.16
Comparison of observed linear frequency with different hole sizes.
70
80
130
Geotechnical Design for Sublevel Open Stoping
HQ3 Core (63.5 mm) — Hole HBG_1559_1 (50/64)
100
RQD = 100 –3.78LF + 0.011LF2
R2 = 0.7603
RQD (%)
80
60
40
20
0
0
2
4
6
8
10
12
14
16
18
20
Discontinuity linear frequency (J/m)
LTK 56 Core (45.6 mm) — Hole UHS_3302_1 (–30/077)
100
RQD = 100 –3.081LF + 0.009LF2
R2 = 0.684
RQD (%)
80
60
40
20
0
0
2
4
6
8
10
12
14
16
18
20
Discontinuity linear frequency (J/m)
FIGURE 4.17
Empirical relationship between RQD and linear frequency for two core sizes. (From
Cepuritis, P.M., Hampton boulder haulage shaft geotechnical study. MEngSc thesis, Western
Australian School of Mines, Curtin University of Technology, Perth, Western Australia,
Australia, 1987.)
Rock Mass Characterization
131
Figure 4.18 shows a comparison of RQD and linear frequency for the hangingwall of the Kanowna Belle Mine, where both parameters were calculated
over identical 1 m intervals down the hole. The parameters were calculated
using a 10 m × 10 m grid model overlain on a solid model defined by the first
5 m into the orebody hangingwall, where the parameters were calculated.
The inverse distance square method was used for interpolation. The data
suggest that both methods predict similar variability of the rock mass conditions at the Kanowna Belle open stoping mine.
4.3.3 Rock Mass Classifications from Core Logging
Rock mass classification systems are used in mining and civil engineering applications to characterize the rock mass and to determine maximum
unsupported spans, support, and reinforcement requirements and estimated
rock mass strengths. They are empirical methods that have been developed
from the back analysis of excavation performance. Where multiple joint sets
with differing discontinuity characteristics are present in a rock mass, a
decision must be made as to the most important discontinuity set that is
likely to control rock mass behavior and potential failure.
A program of data collection for rock mass classification purposes can also
take place when the geologists are logging the core for orebody delineation
purposes. A number of conventional classification systems can be considered for rock mass characterization, the most widely used in the Australian
mining industry being the Q (Barton et al., 1974) and the RMR (Bieniawski,
1976, 1989) systems. Basically all the holes can be logged in a manner consistent with obtaining information on discontinuity linear frequency and
the information required to carry out a rock mass classification. For each
split, the data collected may include the number of discontinuities per split,
joint condition, joint set number, joint roughness, joint alteration, point load
strength, and the position of faults. The rock mass classification systems
(Tables 4.3 and 4.4) have been extensively described elsewhere (e.g., Hoek
et al., 1995) and so are not described in detail within this book. However, a
number of guidelines are presented to enable logging of rock mass classification parameters using standard exploration core.
The individual parameters for RMR and Q using conventional exploration
core can be estimated as follows:
1. UCS can be determined from standard uniaxial compression or
point load testing of selected core samples.
2. RQD can be determined using the total number of discontinuities
per meter or split to calculate the discontinuity linear frequency.
The actual estimate of RQD can be obtained by evaluating Equation
4.3 or a locally established empirical relationship using core data as
shown in Figure 4.15.
RQD
9400 N
9600 N
9800 N
10000 N
Block D
Block C
Block B
Block A
0
FF/m
1.7
3.4
5.1
6.8
8.5
10.2
11.9
13.6
15.3
17
9400 N
9600 N
9800 N
10000 N
19800 E
20400 E
20200 E
20000 E
19800 E
FIGURE 4.18
A comparison of RQD and discontinuity linear frequency contours for the same orebody (grid is 200 m × 200 m).
0
10
20
30
40
50
60
70
80
90
19000 E
100
19000 E
Block D
Block C
Block B
Block A
132
Geotechnical Design for Sublevel Open Stoping
20400 E
20200 E
20000 E
Strength of
intact rock
material
Drill core
quality RQD
Spacing of
discontinuities
Condition of
discontinuities
1
2
3
4
Parameter
Rating
Rating
30
25
12
75–90
17
0.6–2 m
15
Slightly rough
surfaces,
separation <1
mm, slightly
weathered wall
rock
15
90–100
20
>2 m
20
Very rough
surfaces not
continuous,
no separation,
unweathered
wall rock
Rating
100–250
>250
Uniaxial
compressive
strength (UCS)
(MPa)
Rating
4–10
>10
Point load
strength index
(MPa)
Rock Mass Rating (RMR) System
TABLE 4.3
20
7
50–75
13
200–600 mm
10
Slightly rough
surfaces,
separation <1 mm,
highly weathered
wall rock
50–100
2–4
Range of Values
4
25–50
8
60–200 mm
8
Slickensided
surfaces or
gouge <5 mm
thick or
separation
1–5 mm,
continuous
10
25–50
1–2
0
(continued )
1
0
<25
3
<60
5
Soft gouge
>5 mm thick
or separation
>5 mm
continuous
2
For this low
range,
conventional
UCS testing is
preferred
5–25
1–5
<1
Rock Mass Characterization
133
Groundwater
Inflow per 10 m
tunnel length
(L/min)
Ratio (joint
water
pressure/
major principal
stress)
General
conditions
Rating
Weathering
6
6
Unweathered
1–3 m
4
<0.1 mm
5
Rough
5
Hard filling
<5 mm
4
Slightly
weathered
5
10
Damp
<0.1
0
Completely
dry
15
<10
None
Guidelines for classification of discontinuity conditions
Parameter
Discontinuity length (persistence)
<1 m
6
Separation (aperture)
None
6
Roughness
Very rough
6
Infilling (gouge)
None
5
Parameter
Rock Mass Rating (RMR) System
TABLE 4.3 (continued)
7
>5 mm
2
Moderately
weathered
3
Ratings
3–10 m
2
0.1–1.0 mm
4
Slightly rough
3
Wet
0.1–0.2
10–25
Range of Values
10–20 m
1
1–5 mm
1
Smooth
1
Soft filling
<5 mm
2
Highly
weathered
1
4
Dripping
0.2–0.5
25–125
0
>5 mm
0
Decomposed
>20 m
0
>5 mm
0
Slickensided
0
0
Flowing
>0.5
>125
134
Geotechnical Design for Sublevel Open Stoping
Dip 20–45
Favorable
II
1 year for 10 m
span
300–400
35–45
200–300
25–35
III
1 week for 5 m span
Source: Bieniawski, Z.T., Engineering Rock Mass Classification, John Wiley, New York, 1989, 251pp.
Cohesion of rock mass (kPa)
Friction angle of rock mass (deg)
Meaning of rock masses
Class number
Average stand-up time
I
20 years for
15 m span
>400
>45
IV
10 h for 2.5 m
span
100–200
15–25
V
30 min for 1 m
span
<100
<15
<20
V
Very poor rock
Orientation of Discontinuities
Ratings
Tunnels and mines
Foundations
Slopes
Rock mass ratings determined from total ratings: RMR = ∑ (classification parameters) + discontinuity orientation adjustment
Rating
81–100
61–80
41–60
21–40
Class number
I
II
III
IV
Description
Very good rock
Good rock
Fair rock
Poor rock
Dip 20–45
Unfavorable
Irrespective of strike
Dip 0–20
Unfavorable
Drive against Dip
Very
Unfavorable
−12
−25
−60
Fair
−5
−7
−25
Dip 45–90
Fair
Unfavorable
−10
−15
−50
Very
Favorable
0
0
Favorable
−2
−2
−5
Strike perpendicular to tunnel axis
Dip 45–90
Very unfavorable
Strike parallel to tunnel axis
Rating adjustment for discontinuity orientations
Dip 20–45
Fair
Dip 45–90
Very favorable
Drive with Dip
Effect of discontinuity orientations in tunneling
Rock Mass Characterization
135
136
Geotechnical Design for Sublevel Open Stoping
TABLE 4.4
Tunneling Quality Index Q
Description
Value
Notes
Rock quality designation
A. Very poor
B. Poor
C. Fair
D. Good
E. Excellent
RQD
0–25
25–50
50–75
75–90
90–100
1. Where RQD is reported
or measured as ≤10
(including 0), a nominal
value of 10 is used to
evaluate Q.
2. RQD intervals of 5, that
is, 100, 95, 90, etc., are
sufficiently accurate.
Jn
0.5–1
2
3
4
6
9
12
15
1. For intersections, use
(3.0 × Jn).
2. For portals, use (2.0 × Jn).
Joint set number
A. Massive, none to few joints
B. One joint set
C. One joint set plus random
D. Two joint sets
E. Two joint sets plus random
F. Three joint sets
G. Three joint sets plus random
H. Four or more joint sets, random,
heavily jointed sugar cube, etc.
I. Crushed rock, earthlike
Joint roughness number
a. Rock wall contact and
b. Rock wall contact before 10 cm shear
A. Discontinuous joints
B. Rough or irregular, undulating
C. Smooth, undulating
D. Slickensided, undulating
E. Rough or irregular, planar
F. Smooth, planar
G. Slickensided, planar
c. No rock wall contact when sheared
A. Zone containing clay minerals
thick enough to prevent rock wall
contact
B. Sandy, gravely, or crushed zone
thick enough to prevent rock wall
contact
Joint alteration number
a. Rock wall contact
A. Tightly healed, hard, nonsoftening,
impermeable filling.
B. Unaltered joint walls, surface
staining only.
20
Jr
4
3
2
1.5
1.5
1.0
0.5
1. Add 1.0 if the mean
spacing of the relevant
joint set is greater
than 3 m.
2. Jr = 0.5 can be used for
planar, slickensided
joints having lineations,
provided the lineations
are oriented for
minimum strength.
1.0 (nominal)
1.0 (nominal)
Ja φr (approx.)
0.75
1.0 (25°–35°)
1. Values of φr, the residual
friction angle, are
intended as an
approximate guide to the
mineralogical properties
of the alteration
products, if present.
137
Rock Mass Characterization
TABLE 4.4 (continued)
Tunneling Quality Index Q
Description
C. Slightly altered joint walls
nonsoftening mineral coatings,
sandy particles, clay-free
disintegrated rock.
D. Silty or sandy clay coatings, small
clay fraction (nonsoftening).
E. Softening or low-friction clay
mineral coatings, that is, kaolinite,
mica. Also chlorite, talc, gypsum,
graphite, etc., a small quantity of
swelling clays (discontinuous
coatings, 1–2 mm or less in
thickness).
b. Rock wall contact before 10 cm shear
F. Sandy particles, clay-free
disintegrated rock, etc.
G. Strongly overconsolidated,
nonsoftening clay mineral filings
(continuous, <5 mm thick).
H. Medium or low overconsolidation,
softening, clay mineral fillings
(continuous, <5 mm thick).
J. Swelling clay fillings, that is,
montmorillonite (continuous,
<5 mm thick). Values of Ja depend
on percentage of swelling clay-size
particles and access to water.
c. No rock wall contact when sheared
K–M. Zones or bands of disintegrated or
crushed rock and clay (see G, H,
and J for description of clay
conditions).
N. Zones or bands of silty or sandy
clay, small clay fraction
(nonsoftening).
O–R. Thick, continuous zones or bands
of clay (see G, H, and J for clay
conditions).
Joint water reduction
A. Dry excavations or minor inflow,
that is, <5 L/min locally
B. Medium inflow or pressure,
occasional outwash of joint fillings
Value
Notes
2.0 (25°–30°)
3.0 (20°–25°)
4.0 (8°–16°)
4.0 (25°–30°)
6.0 (16°–24°)
8.0 (12°–16°)
8.0–12.0
(6°–12°)
6.0, 8.0
or 8.0–12.0
(6°–24°)
5.0
10.0–13.0
13.0–20.0
(6°–24°)
Jw
1.0
0.66
Approx. water pressure
(kgf/cm2)
<1.0
1.0–2.5
(continued )
138
Geotechnical Design for Sublevel Open Stoping
TABLE 4.4 (continued)
Tunneling Quality Index Q
Description
Value
Notes
C. Large inflow or high pressure in
0.50
2.5–10.0
competent rock with unfilled joints
D. Large inflow or high pressure,
0.33
2.5–10.0
considerable outwash of joint
fillings
E. Exceptionally high inflow or water
0.2–0.1
>10.0
pressure at blasting, decaying with
time
F. Exceptionally high inflow or water
0.1–0.05
>10.0
pressure continuing without
noticeable decay
Factors C–F are crude estimates; increase Jw if drainage installed. Special problems caused
by ice are not considered.
Stress reduction factor
SRF
a. Weakness zones intersecting excavation, which may cause
loosening of rock mass when tunnel is excavated
A. Multiple occurrences of weakness
10.0
zones containing clay or chemically
disintegrated rock, very loose
surrounding rock (any depth).
B. Single weakness zones containing
5.0
clay or chemically disintegrated
rock (depth of excavation <50 m).
C. Single weakness zones containing
2.5
clay or chemically disintegrated
rock (depth of excavation >50 m).
D. Multiple shear zones in competent
7.5
rock (clay-free), loose surrounding
rock (any depth).
E. Single shear zones in competent
5.0
rock (clay-free), (depth of
excavation <50 m).
F. Single shear zones in competent
2.5
rock (clay-free), (depth of
excavation >50 m).
G. Loose, open joints, heavily jointed
5.0
or sugar cube, etc. (any depth).
Stress reduction factor
σc/σ1
b. Competent rock, rock stress problems
H. Low stress, near surface, open
>200
joints.
J. Medium stress, favorable stress
10–200
condition.
K. High stress, very tight structure.
5–10
Usually favorable to stability, may
be unfavorable for wall stability.
1. Reduce these values of
SRF by 25%–50% if the
relevant shear zones
only influence but do
not intersect the
excavation.
σt/σ1
SRF
>13
2.5
0.66–13
1.0
0.33–0.66
0.5–2.0
139
Rock Mass Characterization
TABLE 4.4 (continued)
Tunneling Quality Index Q
Description
Value
Notes
L. Mild rockburst (massive rock).
2.5–5.0
0.16–0.33
5–10
M. Heavy rockburst (massive rock).
<2.5
<0.16
10–20
1.For a strongly anisotropic virgin stress field (if measured): when 5 ≤ σ1/σ3 ≤ 10,
reduce σc to 0.8σc and σt to 0.8σt. When σ1/σ3 > 10, reduce σc and σt to 0.6σc and
0.6σt where σc is the unconfined compressive strength, σt is the tensile strength
(point load), and σ1 and σ3 are the major and minor principal stresses.
2.Few case records available where depth of crown below surface is less than span
width. Suggest SRF increase from 2.5 to such cases (See H).
c. Squeezing rock: plastic flow of incompetent rock under the influence of high
SRF
rock pressure
Mild squeezing rock pressure.
5–10
Heavy squeezing rock pressure.
10–20
d. Swelling rock: chemical swelling activity depending on the presence of water
SRF
Mild swelling rock pressure.
5–10
Heavy swelling rock pressure.
10–15
Additional notes on the use of these tables:
When making estimates of the rock mass quality (Q), the following guidelines should be
followed in addition to the notes listed on the tables:
1. When borehole core is unavailable, RQD can be estimated from the number of joints per
unit volume, in which the number of joints per meter for each joint set is added.
A simple relationship can be used to convert this number to RQD for the case of clay-free
rock masses: RQD = 115–3.3 Jv (approx.), where Jv = total number of joints per m3
(0 < RQD < 100 for 35 > Jv > 4.5).
2. The parameter Jn representing the number of joint sets will often be affected by foliation,
schistosity, slaty cleavage or bedding, etc. If strongly developed, these parallel joints
should obviously be counted as a complete joint set. However, if there are few joints
visible, or if only occasional breaks in the core are due to these features, then it will be
more appropriate to count them as random joints when evaluating Jn.
3. The parameters Jr and Ja (representing shear strength) should be relevant to the weakest
significant joint set or clay-filled discontinuity in the given zone. However, if the joint set
or discontinuity with the minimum value of Jr/Ja is favorably oriented for stability, then
a second, less favorably oriented joint set or discontinuity may sometimes be more
significant, and its higher value of Jr/Ja should be used when evaluating Q. The value of
Jr/Ja should in fact relate to the surface most likely to allow failure to initiate.
4. When a rock mass contains clay, the factor SRF appropriate to loosening loads should be
evaluated. In such cases, the strength of the intact rock is of little interest. However,
when jointing is minimal and clay is completely absent, the strength of the intact rock
may become the weakest link, and the stability will depend on the ratio rock stress/rock
strength. A strongly anisotropic stress field is unfavorable for stability and is roughly
accounted for as in note 2 in the table for SRF evaluation.
5. The compressive and tensile strength (σc and σt) of the intact rock should be evaluated in
the saturated condition if this is appropriate to the present and future in situ conditions.
A very conservative estimate of the strength should be made for those rocks that
deteriorate when exposed to moist or saturated conditions.
Source: Barton, N. et al., Rock Mech., 6(4), 189, 1974.
140
Geotechnical Design for Sublevel Open Stoping
3. The number of joint sets forming potentially unstable blocks
requires engineering judgment when oriented core is not available,
as is normally the case during conventional exploration drilling.
A limited amount of oriented core or mapping of underground
exposures may be required to identify the exact number of discontinuity sets.
4. Joint spacing of the most significant (potentially unstable) discontinuity set can be determined by simply inverting the discontinuity
linear frequency for the critical set. Oriented core may be needed to
identify the individual sets.
5. Condition of the discontinuities from core logging can be determined on the basis of the infill material. The guidelines provided in
Table 4.5 are recommended to establish a relationship between the
values proposed by Bieniawski (1989) and the actual infill logged.
6. Groundwater conditions can be determined based on engineering
judgment of the site under consideration.
7. Estimation of joint roughness (Jr) from core logging is difficult
because the undulation (large scale) cannot be readily determined.
The guidelines provided in Table 4.6 are recommended for the
determination of the small-scale roughness on the basis of the infill
material.
8. Determination of joint alteration (Ja) from core logging can be based
on a simple empirical scratch test of the joint surfaces (Milne et al.,
1991). If the joint surface can be scratched with a knife blade, then
Ja ranges from 1 to 1.5. If the joint surface can be scratched with a
fingernail, or feel slippery, then Ja is equal to 2. Similarly, where an
altered joint surface can be dented with a fingernail, or the joint is
infilled with gouge, then Ja is set to 4. The guidelines provided in
Table 4.7 are recommended for the determination of the joint alteration on the basis of the infill material.
TABLE 4.5
Estimates of Joint Condition from Logged
Infill Materials
Logged Roughness (Average
of Interval)
Fault gouge
1. Slickensided/polished
2. Smooth
3. Defined ridges
4. Small steps, quartz infill
5. Very rough, quartz infill
RMR Joint
Condition Rating
0
10
15
20
25
30
141
Rock Mass Characterization
TABLE 4.6
Estimates of Joint Roughness (Small Scale)
from Logged Infill Materials
Logged Roughness
(Average of Interval)
Fault gouge
1. Slickensided/polished
2. Smooth
3. Defined ridges
4. Small steps, quartz infill
5. Very rough, quartz infill
Equivalent Q Rating
Gouge
Slickensided
Smooth
Rough
Rough
Rough
TABLE 4.7
Estimates of Joint Alteration
from Logged Infill Materials
Type of Infill (Average
of Interval)
Quartz
Limonite
Clay
Chlorite
Faults and shears
Equivalent
Ja Rating
1
2
4
4
8
9. The stress reduction factor (SRF) can be estimated using the UCS
information and by determining the in situ stress using oriented
core (Villaescusa et al., 2002, 2003b, 2012).
4.3.4 Advantages, Disadvantages, and Biases in Core Logging
Core logging allows for the effective three-dimensional establishment of
lithological boundaries. If most of the holes drilled through a deposit are
logged, an investigation into the spatial variations in rock mass quality
across an orebody can be facilitated. Identification of severely fractured
zones (Figure 4.19) and alteration boundaries around an orebody can be
accomplished. Logging of diamond-drilled core can be used to define
the discontinuity linear frequency, rock quality data (RQD), and rock
mass classification data. This provides an extensive and representative
database of the frequency of geological discontinuities within the immediate boundary of an orebody. Mine-wide estimation of laboratory testing
strength indices such as UCS and the point load test index, Is(50), can be
determined. Joint condition and strength can also be determined from
logging.
142
Geotechnical Design for Sublevel Open Stoping
FIGURE 4.19
Main shear zones close to an orebody boundary.
When a large number of individuals perform logging, a chance to some
bias arises because of potentially different geological and geotechnical interpretations. Logging data collected from small-diameter core (27–40.7 mm)
may not be ideally suited for geotechnical logging. Mechanical disturbance
of the core from the drilling process itself can be experienced. Core loss may
occur in heavily fractured rock masses. Infill material may be washed out
during the drilling process. The size or persistence of discontinuities cannot be determined. Core orientation and drill deviation must be taken into
account.
Orientation bias occurs due to the preferential sampling of joints oriented
normal to the drillhole axis (Terzaghi, 1965). In tabular orebodies, information for stope hangingwall design is more readily collected than information
for back design. Size bias occurs due to large discontinuities having a greater
chance of being intersected by drillholes. The weakest, and perhaps the most
relevant, (infill) materials may be lost during drilling.
4.4 Geotechnical Mapping of Underground Exposures
Mapping of exposed rock faces allows for the direct assessment of several
rock mass characterization parameters that cannot be established by routine
drillhole logging for orebody delineation purposes. Geotechnical mapping
of direct accesses such as stope drill drives, and access crosscuts can be used
143
Rock Mass Characterization
to determine the orientation, linear frequency, size, and surface strength of
the geological discontinuities. In addition, observation, interpretation and
description of faults and shears, precise determination of the number discontinuity sets, trace lengths, and observation of the large-scale planarity
and joint roughness characteristics can all be achieved by direct geotechnical
mapping. Importantly, the need for oriented core is minimized if unbiased
data from direct mapping can be used to establish reliable joint set orientation boundaries leading to the geotechnical description of each geological
discontinuity set.
Several methods are available to determine the geological discontinuity set characteristics including line sampling (Call, 1972; Priest, 1985), cell
sampling techniques (Mathis, 1988), and strip mapping (Landmark and
Villaescusa, 1992). The data collected can be divided into two classes (Call
et al., 1976): major structures and minor geological features. Major structures,
such as faults, dikes, contacts, and related features, usually have a size of
the same order of magnitude as that of the site to be characterized. They are
usually continuous, have low shear strength, and sometimes can be seismically active. The position in space, physical properties, and geometrical characteristics are usually established deterministically for each of those main
discontinuities (Figure 4.20).
Major structures are characterized by routine geological mapping carried out by the geologists, who usually gather data on orebody boundaries,
rock types, alteration, and location of the main structural features using at
a 1:500 or 1:1000 scale. Following the completion of mapping along several
drives and elevations, the mine geologists undertake data interpolation to
determine which structures are continuous across several drives and levels,
thus forming a large-scale structure (Cepuritis et al., 2007; Cepuritis, 2011a).
D3 Stope
16750E
D4 Stope
16700E
16650E
D2 Stope
FIGURE 4.20
A major geological structure intersecting a number of stopes and associated access
development.
144
(a)
Geotechnical Design for Sublevel Open Stoping
(b)
FIGURE 4.21
Interpretation of location and orientation of large-scale geological features on a stoping block
scale. (a) Location of faults from geotechnical mapping and (b) interpretation of faults.
The interpretation is usually based on structure type, orientation, alteration, and infill type and thickness. Figure 4.21 shows an interpretive longitudinal section featuring the position of the major discontinuities with
respect to an entire stope block area.
For practical purposes, minor features represent and infinite population in
the area of a stope design. As a result, their geometrical characteristics and
physical properties must be estimated by measurements of a representative
sampled (smaller) population using the methods described later.
4.4.1 Cell Mapping
This is a form of areal sampling or two-dimensional mapping in which
an area interception criterion is established in order to collect the field
data. Rectangular or square windows, which are called cells are defined
along excavation walls (Figure 4.22). A statistical value based on the properties of the geological discontinuities found within the boundaries is assigned
to each cell (Pahl, 1981; Laslett, 1982; Kulatilake and Wu, 1984, Mathis, 1988).
In this method, the individual discontinuity sets are defined visually
within the cell boundaries. This process requires the grouping by eye of a
family of discontinuities with similar orientational properties in order to
form a geological design set. For each discontinuity set, the orientation, location, and end points of all the discontinuities within the cell boundaries are
recorded. A sampling line can be used to calculate the average apparent discontinuity spacing.
Mathis (1988) developed a quick areal sampling method, in which the discontinuity properties are sampled from a reduced number of observations that
appear to represent the mean values for each set. Nevertheless, cell mapping
methods are time consuming compared with routine geological mapping or
the more conventional geotechnical mapping based on line sampling.
145
Rock Mass Characterization
Roof or back
Only joint set shown
NT— No. of disc. sampled = 18
N2 — both ends exposed = 11
N1 — one end exposed = 3
N0 — no end exposed = 3
Line used for spacing
calculations
Sampling window
Floor
FIGURE 4.22
Longitudinal section view showing typical cell mapping procedures.
4.4.2 Line Mapping
This is a systematic, one-dimensional spot sampling technique, which can be
extended to two dimensions if the line is located inside a sampling window.
The method consists of stretching a measuring tape along an exposed face
and recording the measurements and features of interest of every discontinuity that intersects the tape (Figure 4.23). Ideally, the sampling sites should
be randomly selected in three equal-length, mutually orthogonal directions.
In this way, any discontinuity ignored by one line, because of its orientation, will be sampled preferentially by one or two lines. In practice, however, the sites are determined by the availability and accessibility of the rock
exposures. For example, vertical sampling lines are very important in determining the properties of any flat-lying discontinuity sets. However, vertical lines are difficult to obtain due to the absence of vertical development
within an area of interest. A recommended compromise is to use several
Roof or back
Sampling
line
Upper limit for
trace length
measurements
Only joint set shown
NT — No. of disc. sampled = 5
N2 — both ends exposed = 3
N1 — one end exposed = 1
N0 — no ends exposed = 1
Floor
FIGURE 4.23
Line sampling of rock mass discontinuities.
Lower limit for
trace length
measurements
146
Geotechnical Design for Sublevel Open Stoping
randomly located, short ladder-based lines within the drives or face walls
where heights approaching 3–4 m are available.
Experience has shown that approximately 2 days are required to choose
an appropriate mapping site, establish the line, and record the data required.
Mapping should be undertaken on clean (washed) or newly exposed rock
surfaces, which allow for a better exposure of the discontinuity characteristics (Figure 4.24). The length of the sampling line is usually extended until
a prerequisite number of observations are obtained. Savely (1972) determined that at least 60 observations are required to stereographically define
the discontinuity sets found along a particular sampling line. Villaescusa
(1991) found that at least 40 observations per set are required in order to
construct experimental histograms of spacing, trace length, and discontinuity orientation. In practice, however, depending upon the complexity of the
discontinuity network (a rock mass generally contains between three and six
discontinuity sets), and the number of sampling lines used, between 200 and
300 observations are required to establish a structural domain for design.
4.4.3 Strip Mapping
Strip mapping was developed as an alternative to conventional mapping
techniques that were found to be too slow for routine use in a mining environment. The strip mapping technique was developed to record relevant
data to be used by rock mechanics, geology, and planning personnel. It was
envisaged that the technique should not be time consuming in both the collection and the manipulation of the data. The initial results correlated well
FIGURE 4.24
Washing of rock walls prior to geotechnical mapping at the Golden Grove Mine, Western
Australia.
147
Rock Mass Characterization
with those obtained from conventional techniques such as line and cell
mapping (Landmark and Villaescusa, 1992).
Strip mapping is a two-dimensional mapping method that incorporates
features used in both line and cell mapping. The excavation walls are divided
into 10 m intervals and the midpoint of each interval marked on the wall and
located at least 1.5 m above the floor. Geological discontinuities occurring
within the 10–20 m interval are visually grouped into sets based on their
orientations. A 3 m by 1 m window (the strip), centered on the interval midpoint and oriented with the long axis parallel to the projection of each set,
is marked on the excavation wall. A 1 m long sampling line is then located
through the interval midpoint and aligned normal to the long axis of the
strip (Figure 4.25).
Every discontinuity crossing this line is then noted. Discontinuity set
characteristics determined by the strip mapping method include orientation, linear frequency, and mean trace length values. The number of end
points for the discontinuity occurring within the window is recorded. N0
Long section view of development
Backor roof
7160 N Floor
7170 N
Set 2
Set 1
3m
3m
1 m long
sampling
line
Date: 18/12/91 O/B: 7 Sill: X700 Level: 9A
Wall orientation: 75/276
Northing: 7165
Joint
Set
No.
1
2
Orientation
DIP DIPDIR
45
37
356
151
Linear
Frequency
(J/m)
5
5
Trace Length
Endpoints
N0
1
1
N1
2
1
N2
2
3
FIGURE 4.25
Strip mapping method. (From Landmark, J. and Villaescusa, E., Geotechnical mapping at
Mount Isa Mines, in T. Szwedzicki et al., eds., Proceedings of the Western Australian Conference
on Mining Geomechanics, Kalgoorlie, Western Australia, Australia, 8–10 June, 1992, pp. 329–333,
Western Australian School of Mines, Kalgoorlie, Western Australia, Australia.)
148
Geotechnical Design for Sublevel Open Stoping
signifies the number of transecting joints. They have no end points exposed
within the 3 m by 1 m strip, implying that the discontinuity is at least 3 m
long. N1 is used to denote the number of discontinuities with one end point
exposed within the strip. N2 indicates the number of intersecting discontinuities that are contained within the strip and are smaller than 3 m in
length. A frequency count for each discontinuity set is obtained from summing the N0, N1, and N2 values. A mean orientation is recorded for each
set of any convenient surface occurring within the 10 m interval. A strip
is then defined for the next joint set and the sampling procedure repeated.
This operation is carried out at each midpoint for all the suitable intervals
within an exposed wall. The strip mapping method allows an estimate of
the mean trace length (Lmean) to be made using the method proposed by
Pahl (1981):
L mean =
h (2N 0 + N1 )
2N 2 + N1
(4.5)
where h is the height of the observation cell, that is, 3 m.
Pahl (1981) considered this a special case where the observation window is
parallel to the discontinuity set orientation, otherwise an angular correction
factor has to be applied. However, in strip mapping, the observation window is always parallel to each discontinuity set, as the window is rotated
for each set.
4.4.4 Description of Mapping Parameters
Geotechnical mapping requires the determination of the number of families, location (frequency and spacing), orientation and size, intact rock
strength, hydrological condition, and surface properties of critical discontinuities likely to influence stope stability. The information is usually
recorded on a tabular data sheet suitable for subsequent computer analysis. The dimensions of the observation window, regardless of the mapping
technique used, should be kept constant at each site and across sites, since
data from different locations are usually grouped together. Basic information at each site should include the number, location, elevation, bearing,
and plunge of the reference line used to collect the frequency, the dip and
dip direction of the rock exposure, and the censoring levels of the convex
sampling window (up and down from the observation line; see Figures
4.22 and 4.23). A mnemonic system compatible with the notation used by
the local geologists should be implemented to identify different rock types
and discontinuity infill material at each site (Call et al., 1976). The following
characteristics and geological factors influencing a rock mass should be
recorded (for cell and line mapping techniques):
Rock Mass Characterization
149
1. Distance along the tape where the discontinuity intersects the sampling line. Discontinuity spacing is calculated from intercept distances,
the mean set orientation, and the orientation of the sampling line.
Numerically, the individual apparent spacing values are defined by
sorting the discontinuities by individual sets down the line and subtracting the distances between adjacent discontinuities of the same set.
2. End points of the discontinuities intersecting the tape. When the
discontinuities are observed through a convex window, three set of
observations are obtained: joints totally contained (two trace length
end points observable), joints intersecting only one of the window
boundaries (one trace length end point observable), and finally, joints
transecting the window (no trace length end points observable).
3. Type of structures, naturally occurring features such as faults, shears,
bedding, veins, joints, contacts, etc.
4. Orientation: dip and dip direction of features intersecting the tape.
5. Rock type recorded using a mnemonic system code.
6. Roughness: A qualitative measure of the small-scale (2 cm or less)
asperities on the discontinuity surface. Rough, smooth, and slickensided categories are used.
7. Planarity: A qualitative indication of the geometrical nature of the
discontinuities on a large scale. Planar, wavy, and irregular categories are used.
The rock mass classification parameter joint roughness (Jr) is determined based on small-scale roughness and large-scale planarity.
However, the guidelines for the classification of discontinuity joint
roughness can be subjective due to the qualitative nature of their
description. Milne et al. (1991) have developed a method that can be
used to measure joint roughness quantitatively.
Following Milne et al. (1991), small-scale roughness (Jrr) can be calculated by determining the maximum joint amplitude over a 10 cm
profile, and the large-scale planarity (Jrw) can be calculated by establishing the maximum joint amplitude over a 1 m profile. These two
parameters can be measured using a 10 cm steel rule and a 1 m folding rule, where each ruler is placed along the discontinuity surface,
and the maximum amplitude is recorded for both 10 cm and 1 m
profiles as shown in Figure 4.26. Small-scale amplitudes of less than
2.5 mm are assigned a Jrr of 1.0, while amplitudes greater than 2.5 mm
are assigned a Jrr of 1.5. Similarly, the large-scale planarity Jrw is
assigned values of 1.0, when the amplitudes measured are less than
10 mm, a value of 1.5 when the amplitudes measured fall between
10 and 20 mm, and finally a Jrw value of 2.0 is assigned to measured
amplitudes greater than 20 mm. The resulting joint roughness (Jr)
150
Geotechnical Design for Sublevel Open Stoping
1000
Amplitude
20
16
12
8
Length
1m
Profile
Amplitude of asperities (mm)
100
Wavy
Jr/w ~ 2.0
50
10 cm
Profile
10
10 mm
Planar to
wavy
Jr/w ~ 1.5
1
Planar
Jr/w ~ 1.0
Rough
Jr/r ~ 1.5
5
20 mm
4
2.5 mm
Smooth
Jr/r ~ 1.0
1
Joint roughness coefficient (JRC)
500
(Jr/r) (Jr/w) = Jr
0.1
0
0.1
1.0
10
Length of profile (m)
FIGURE 4.26
The band’s joint amplitude versus trace length showing the Milne et al. proposed method.
(From Milne, D.P. et al., Systematic rock mass characterization for underground mine design,
in W. Wittke, ed., Proceedings of the Seventh International Congress on Rock Mechanics, Aachen,
Germany, September 16–20, vol. 1, 1991, pp. 293–298, A.A. Balkema, Rotterdam, the Netherlands.)
8.
9.
10.
11.
12.
for classification purposes is determined by multiplying the Jrr and
Jrw terms (Milne et al., 1991).
Infill material: In some discontinuities such as faults, this may be
gouge and slickensides, and in others, it may be a quantitative
mineralogical assemblage.
Thickness: The measured width across a discontinuity wall.
Alteration of the discontinuity walls, determined using scratching
tests. Guidelines were suggested in Section 4.3.3.
Trace length measured as seen in the rock face. The trace length is the
maximum measurable length of the resulting intersection between a
discontinuity and a planar excavation in rock.
Termination as observed in the top and bottom of a discontinuity in the dip direction, but only if the discontinuity is contained;
151
Rock Mass Characterization
O
O
J-H
J-L
IR
IR
J-H
J-H
J-H
J-H
Top of mapping
window
IR
IR
IR
Measuring
tape
J-H
IR
O
O
O
Bottom of mapping
window
Obscured by rubble
O, obscured; IR, intact rock; J-L and J-H, against a discontinuity at low angle and
high angle; , mapped structure.
FIGURE 4.27
Types of discontinuity termination, line mapping.
otherwise, the termination is artificially obscured by the observation
window or the excavation geometry. Discontinuities can terminate
either against another joint or within intact rock. Call et al. (1976)
introduced the concept of high- (>20°) and low (<20°)-angle termination against another discontinuity; when this occurs, observations
suggest that one of the discontinuities is likely to propagate further
along a common direction (Figure 4.27).
13. Remarks are used to describe other characteristics such as alteration,
observed hydrological conditions, etc.
The format of the geotechnical data collected must allow flexibility such that
a geotechnical engineer can utilize any number of rock mass characterization
or classification techniques as required. Therefore, the parameters collected
should preferably describe engineering geology data, rather than interpreted
rock mass classification parameters (Cepuritis, 2004). As an example, it is recommended to collect planarity and roughness data instead of an interpreted
Jr from the Q system (Barton et al., 1974). It is very important to note that collecting fundamental engineering geology characteristics in the form of rock mass
classification parameters may introduce bias, as these parameters are interpretations and simplifications of the actual rock mass characteristics. Appendix A
shows the proposed data collection sheets for core logging and geotechnical
mapping taking into account the 12 fundamental characteristics described later.
4.4.5 Mapping Biases
Geological mapping of geological discontinuities introduces a number
of biases into the sampled data, namely, orientation, size truncation, and
152
Geotechnical Design for Sublevel Open Stoping
censoring. Carefully defined sampling or correction in procedures is essential to eliminate or minimize these effects. At each site, at least three different
directions of mapping should be chosen to reduce the orientation bias introduced when discontinuities striking parallel to a surveying line are sampled
to a lesser degree than discontinuities striking normal to the sampling direction (Terzaghi, 1965). Furthermore, a quantitative correction of this bias as
described by Priest (1983) can be implemented during data analysis.
Size bias in discontinuity sampling occurs at two levels. At the first level,
larger discontinuities are more likely to intersect an outcrop or excavation
wall than smaller discontinuities. At the second level, the likelihood that a
sampling line intersects a discontinuity trace is directly proportional to the
length of the trace (Baecher and Lanney, 1978). A method based on mathematical stereology (Warburton, 1980) can be used to correct the first-level
bias, and the second-level bias can be corrected using a method developed
by Laslett (1982).
In data collection, a decision is made to disregard any discontinuity with
a trace length smaller than an arbitrary cut-off. Also, the dimension of the
artificial window imposed on the geological discontinuities limits the size of
the observed structures. As a result, the sampled trace length distributions
are both truncated and censored. Truncation occurs when trace length values below a certain threshold are not recorded. Censoring occurs when the
observed length of a trace is shortened due to the edge effect of the observation (artificial or not) window (i.e., when one or both ends of the trace are not
visible). Censored traces provide only lower-bound estimates of their lengths.
The walls of an underground excavation are rarely smooth or planar
especially when the openings are created by traditional drill-blast-scale
techniques. Even in cases in which overbreak is negligible, hole deviation
alone will control the conditions or geometry of the final excavated walls.
As pointed out by Mathis (1988), the excavation process will expose discontinuities that would normally be hidden behind the plane of mapping
(Figure 4.28) leading to overestimation of parameters such as observed trace
length, discontinuity frequency, and also a reduction in the discontinuity
orientation bias.
4.4.6 Geological Strength Index
As part of the continuing development and practical application of the Hoek–
Brown empirical rock mass strength criterion to be discussed in Section 4.7,
Hoek (1994) and Hoek et al. (1995) introduced a new rock mass classification
scheme known as the Geological Strength Index (GSI). The GSI was developed to overcome some of the deficiencies that had been identified in more
than a decade of experience in using Bieniawski’s rock mass rating (RMR)
with the rock mass strength criterion.
The GSI is an index developed specifically as a method of accounting for
those properties of a discontinuous or jointed rock mass, which influence
Rock Mass Characterization
153
FIGURE 4.28
Actual mapping surfaces after blasting and scaling. (From Mathis, J.I., Development and verification of a three dimensional rock joint model, PhD thesis, Lulea University, Lulea, Sweden,
1988.)
its strength and deformability. Accordingly, the GSI seeks to account for
two features of the rock mass—its structure as represented by its blockiness
and degree of interlocking, and the condition of the discontinuity surfaces.
Using Figure 4.29, the GSI may be estimated directly from visual examination or mapping of exposures of the rock mass, and more indirectly from
borehole core.
The GSI does not include an evaluation of the uniaxial compressive
strength (UCS; σc) of the intact pieces of rock because this factor is taken
into account when rock mass strength estimates are made using the
Hoek–Brown criterion (see Section 4.7.2). It should be noted that massive
or sparsely jointed rock masses do not satisfy the assumed conditions of
isotropy and homogeneity and that tectonically presheared rock masses do
not satisfy the assumed conditions of interlocking and peak shear strength
development on joints. Although they are shown in Figure 4.29, GSI estimates for these two types of structure should be made with great care or
not at all. Further discussions of the development of the GSI and of its use
in the context of underground excavation design are given by Brown (2007)
and Marinos et al. (2007).
4.5 Analysis of Mapping Data
4.5.1 Discontinuity Orientation
Discontinuity orientation is defined by two field measurements that can
be expressed as either strike and dip, or most commonly, dip and dip
154
Intact or massive—Intact
rock specimens or massive in situ
rock with few widely spaced
discontinuities
Blocky/disturbed/seamy—
Folded with angular blocks
formed by many intersecting
discontinuity sets. Persistence
of bedding planes or schistocity
Disintegrated—Poorly interlocked, heavily broken rock mass
with mixture of angular and
rounded rock pieces
Laminated/sheared—Lack
of blockiness due to close spacing
of weak schistocity or shear planes
90
N/A
N/A
80
70
Decreasing interlocking of rock pieces
Very blocky—interlocked
Partially disturbed mass with
multifaceted angular blocks
formed by four or more joint sets
Very poor
Slickensided, highly weathered surfaces with soft clay
Coatings or fillings
Decreasing surface quality
Structure
Blocky—Well interlocked
undisturbed rock mass consisting
of cubic blocks formed by three
intersecting discontinuity sets
Fair
Smooth, moderately weathered and altered surfaces
Good
Rough, slightly unweathered, nonstained surfaces
Very good
Very rough, fresh unweathered surfaces
From the lithology, structure, and surface
conditions of the discontinuities, estimate
the average value of GSI. Do not try to be
too precise.Quoting a range from 33 to 37
is more realistic than stating that GSI = 35.
Note that the table does not apply to
structurally controlled failures.
Where weak planar structural planes are
present in an unfavorable orientation with
respect to the excavation face, these will
dominate the rock mass behavior.
The shear strength of surfaces in rocks that
are prone to deterioration as a result of
changes in moisture content will be reduced
if water is present. When working with
rocks in the fair to very poor categories, a
shift to the right may be made for wet
conditions. Water pressure is dealt with by
effective stress analysis.
Surface conditions
Geological Strength Index
Jointed Rocks (Hoek and Marinos, 2000)
Poor
Slickensided, highly weathered surfaces with compact
Coatings or fillings or angular fragments
Geotechnical Design for Sublevel Open Stoping
60
50
40
30
20
N/A
N/A
10
FIGURE 4.29
Geological strength index (GSI) chart for jointed rock masses. (From Hoek, E., Rock mass
properties, in Practical Rock Engineering, 2007. Available online at http://www.rocscience.com/
hoek/corner/11_Rock_mass_properties.pdf.)
155
Rock Mass Characterization
Z (R/L)
Z
Normal to
discontinuity
plane
D
ip
Plunge (φ)
di
re
ct
Normal to
discontinuity
plane
(θ)
io
n
Y (North)
Trend (β)
X
0 < β < 2π
–π/2 < φ < π/2
θ = π/2–φ
φ=β
Y
Dip (θ)
di Dip
re
c
(φ tio
) n
X (East)
FIGURE 4.30
Polar coordinates defining orientation of a unit normal to a discontinuity plane.
direction. Both forms of measurements can be used to describe the attitude in space of planar structural features contained within a rock mass.
For data analysis in three dimensions, discontinuity orientation can also
be represented by unit vectors normal to the discontinuity planar surfaces
(Priest, 1985). The orientation of the unit normals is recorded unambiguously by polar coordinates using two angles, the plunge φ and the trend β
(Figure 4.30).
The plunge φ is the acute angle, measured in a vertical plane, between the
horizontal and the unit normal. It can vary between −π/2 and π/2. However,
it suffices to consider only one orientation hemisphere (upper or lower),
since in the other hemisphere each pole is duplicated. The trend β can vary
around the full circle and is defined as the angle between north and the
vertical plane containing the unit normal to the plunge φ. By convention,
the trend is measured in a clockwise rotation in the direction of the plunge
(Priest, 1985).
4.5.2 Number of Discontinuity Sets
The orientation of the structural data from each sampling location is usually
displayed as conventional lower-hemisphere equal-angle projections where
the statistical calculation of the pole densities can be used to accurately define
families of discontinuities. The accuracy of the discontinuity set boundary
determination depends on the ability to detect changes in the stereographic
projection patterns, as it is sometimes difficult to group overlapping clusters
into design sets.
156
Geotechnical Design for Sublevel Open Stoping
7%
Percent of total points
6
5
P = 0.02
(98%)
P = 0.20
(80%)
4
P = 0.001
(99.9%)
3
2
1
0
10
20
30
50
100
200 300
500
1,000
2,000
5,000
10,000
Total number of points
FIGURE 4.31
Poisson exponential binomial limit for a large number of samples and small probability of
occurrence. (From Villaescusa, E., Slope stability analysis at La Caridad Mine, Nacozari,
Sonora Mexico, Masters thesis, Colorado School of Mines, Golden, CO, 1987, 120pp.)
A method of defining a discontinuity family consists of treating their orientations on a stereographic projection in a purely statistical manner. This
allows a determination to be made of whether a sampled distribution deviates significantly from one which is randomly oriented. The probability of
obtaining concentrations on a point diagram that deviate from a random distribution can be calculated by the Poisson exponential binomial limit (Abel,
1983). Values for the function, given the number of poles in an equal-area
stereographic projection and the confidence limit desired, can be calculated
referring to tables of the Poisson function (Figure 4.31). The 80% confidence
level represents the number of discontinuity orientations lying within 1%
of the total area of the plot that would normally occur only once in five
plots of that total number of randomly generated dips and dip directions.
Conversely, if the 80% confidence level contour is reached, one is 80% confident that the discontinuity is real and not the result of chance or error (Abel,
1983; Villaescusa, 1987). The 98% confidence level reduces the chance of error
to 1 in 50.
In addition to the statistical techniques, it must be emphasized that from
a geological point of view, hemispherical contouring of the data provides a
good indication of the number and orientation of the discontinuity sets present in a rock mass. Figure 4.32 shows a plan view of drawpoint development
at the Bronzewing Gold Mine, Western Australia, where cell, line, and strip
mapping techniques were independently set up to compare the collected
discontinuity data.
The underlying scope of the data interpretation is to determine if more
than one structural domain for design is present across an orebody.
A structural domain for design is an area of the mine in which the rock
Rock Mass Characterization
157
Horizontal line
Vertical line
line
Cell mapping
Cell mappin g
Strip mapping
Strip mappin g
FIGURE 4.32
Plan view of geotechnical mapping locations. (From Baldwin, N., The implementation of geotechnical mapping at Bronzewing Gold Mine, WA, BEng (Mining Geology) thesis, Western
Australian School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia,
Australia, 1998, 96pp.)
mass geometry and the related mechanical strength and deformational
behavior are likely to be similar for all engineering purposes. The results
from Figures 4.33 through 4.35 indicate that the same structural domain
was determined using cell, line, and strip mapping techniques. Strip mapping is fast and accurate, and can be used for systematic collection of discontinuity data during routine development mapping by mine geologists.
Once the discontinuity set boundaries are established, subsequent analysis of the data can be undertaken for each design set to establish the number
of observations per set, mean dip and dip direction, spacing and trace length
calculations (Table 4.8).
4.5.3 Discontinuity Spacing
Discontinuity spacing largely controls rock mass parameters such as in situ
block size, rock mass permeability, and seepage characteristics. Spacing is
defined as the distance between successive discontinuities intersecting a
sampling line or borehole in space. Spacing can be calculated for an individual discontinuity set or for any superposition of sets along a sampling
line. Spacing calculations for an individual set assume that all the discontinuities in a particular set are subparallel and have the same orientation
as the mean orientation of the set. The spacing between two consecutive
158
Geotechnical Design for Sublevel Open Stoping
Bronzewing mine
Contour plot
Fisher pole
concentrations % of
total per1.0 % area
N
W
E
< 0
%
< 1
%
< 2
%
< 3
%
< 4
%
< 5
%
< 6
%
< More
Equal
Angle
Lwr. hemisphere
384
Poles
384
Entries
No bias correction
S
Joint sets from cell mapping
FIGURE 4.33
Discontinuity set definition from cell mapping techniques.
discontinuities of a same set separated by a distance Sa along a sampling
line or borehole is calculated as
Ts = S a cos dM
(4.6)
where
Sa is the measured distance or apparent spacing
δM is the angle between the orientation of the sampling line and the vector
normal to the mean set
Discontinuity observations striking parallel to a sampling line (δM = π/2) are
sampled to a lesser degree than discontinuities striking normal to it (δM = 0).
Analysis of joint spacing data from line sampling supports either negative exponential or logarithmic spacing distributions, depending upon the
degree of periodicity of the spacing data. It is commonly observed that joint
spacings in individual joint sets tend to be clustered, with short spacings
between joints of the same cluster and large spacings between joints belonging to two different clusters (Villaescusa and Brown, 1990). To verify the
distributional nature of the experimental joint spacings, it is necessary to
prepare histograms of joint spacing by set (Figure 4.36). Goodness-of-fit tests
159
Rock Mass Characterization
Bronzewing mine
Contour plot
Fisher pole
concentrations % of
total per 1.0% area
N
E
W
< 0
%
< 1
%
< 2
%
< 3
%
< 4
%
< 5
%
< 6
%
< More
Equal
Angle
LWR. hemisphere
169
Poles
169
Entries
No bias correction
S
Joint sets from line mapping
FIGURE 4.34
Discontinuity set definition from line mapping techniques.
such as the chi-square test can then be used to determine the distributional
function (Villaescusa and Brown, 1990).
In cases where the intersection of a sampling line with a geological discontinuity network is a purely random event, successive spacings are independent with a negative exponential distribution defined by
F (s )= lL e(
-lL s )
(4.7)
where s = 1/λL is the average joint spacing.
As suggested by Priest (1985), a negative exponential distribution of discontinuity spacing might suggest, but does not confirm, that joint intersection along a sampling line or borehole is a purely random event. Histograms
of discontinuity frequency often show repetitive clustering behavior along a
sampling line (Figure 4.37). This suggests that the spacing of individual sets
can be spatially correlated.
4.5.4 Discontinuity Trace Length
Trace length, persistence, or continuity refers to the areal extent or the surface area over which a geological discontinuity extends. In practice, this is
160
Geotechnical Design for Sublevel Open Stoping
Bronzewing mine
Contour plot
Fisher pole
concentrations % of
total per 1.0% area
N
W
< 0
%
< 1
%
< 2
%
< 3
%
< 4
%
< 5
%
< 6
%
< More
Equal
Angle
LWR. hemisphere
166
Poles
166
Entries
No bias correction
E
S
Joint sets from strip mapping
FIGURE 4.35
Discontinuity set definition from strip mapping techniques.
TABLE 4.8
Comparison of Discontinuity Set Characteristics
Geotechnical
Parameter and
Method
Mean orientation
Cell mapping
Line sampling
Strip mapping
Mean spacing
Cell mapping
Line sampling
Strip mapping
Mean trace length
Cell mapping
Line sampling
Strip mapping
Set 1
81/190
85/190
83/187
0.74
0.72
0.43
1.54
5.90
2.43
Set 2
Set 3
Dip/dip direction
55/094
76/052
47/097
73/059
52/096
77/046
Meters
0.42
0.45
0.64
0.42
0.30
0.42
Meters
1.03
1.23
0.97
1.11
1.04
1.59
Set 4
16/329
15/018
17/294
0.31
0.35
0.14
0.78
2.77
1.08
161
Rock Mass Characterization
Observed frequency
80
X = 0.214
SD = 0.203
N = 212
60
40
20
0
0.0
0.2
0.4
0.6
0.8
1.0
1.2
> 1.3
Spacing (m)
FIGURE 4.36
A typical joint spacing distribution.
Discontinuity linear frequency
20
10
0
0
10
20
30
40
Distance (m)
FIGURE 4.37
Clustering of discontinuity linear frequency.
measured as a trace length, which is the minimum measurable length of
the resulting intersection between a geological discontinuity and a planar
excavation in rock.
Quantifying persistence is not easy because of the difficulty of establishing
the areal extent of a discontinuity plane unless the rock mass is dismantled
block by block. Nevertheless, discontinuities that are persistent over the scale
of the exposed spans are more likely to be part of a failure geometry.
162
Geotechnical Design for Sublevel Open Stoping
N = 212
X = 1.97
SD = 1.78
Max = 10.50
Min = 0.27
40
30
20
10
0
0.5
1.5
2.5
3.5
40
Observed frequency
Observed frequency
50
N = 140
X = 0.60
SD = 0.69
Max = 4.00
Min = 0.08
30
20
10
0
4.5 5.5 > 6.5
0.0
Trace length (m)
0.30 0.60 0.90 1.20 1.50 > 1.80
Trace length (m)
FIGURE 4.38
Typical trace length distributions per set for two mine sites.
Sampled trace length distributions are both truncated and censored.
Truncation occurs when observations below or above a certain value are not
recorded. In practice, trace lengths shorter than 200 mm are usually disregarded. Censoring occurs when the observed trace is shortened due to the
edge effects of the observation window, that is, when one or both ends of the
joint trace are not visible. Truncation can be corrected using a method suggested by Warburton (1980), whereas censoring can be corrected using an
algorithm developed by Laslett (1982). Both methodologies are beyond the
scope of this book. The data shown in Figure 4.38 suggest that trace length
distributions appear to be lognormally distributed (Villaescusa and Brown,
1990). The figures also show a quantitative description of the mean, minimum, and maximum trace lengths that can be used to describe the trace
lengths using the schemes suggested by the IRSM (Brown, 1981).
The continuity of the trace lengths is likely to affect the potential for planes
of failure to develop as shown in Figure 4.39, where the existence of broken
and intact rock bridges is conceptualized. For failure to occur, the thrust
must exceed the resistance due to friction and cohesion along a potential failure surface. The maximum and minimum shear strengths are those of the
intact rock and of a smooth planar surface, respectively. For failure to occur
along a discontinuous structure, some intact rock (solid) rock bridges must
be broken between individual discontinuities. This results in an effective
intact rock cohesion defined by Terzaghi (1962) as
C = Ci
As
A
where
C is the effective intact rock cohesion
Ci is the cohesion of intact rock established from triaxial testing
As is the total solid area
A is the total area
(4.8)
163
Rock Mass Characterization
Intact
rock
bridge
Sub-persistent
set
Intact rock
Non-persistent
set
(Maximum trace length)2
(Maximum trace length +
minimum discontinuity
spacing)2
Intact rock
Broken rock
Persistent
set
(a)
(b)
FIGURE 4.39
Trace length continuity and failure plane development. (a) Three-dimensional reality and
(b) two-dimensional estimate. (From Abel, J.F., MN 321 rock mechanics handouts: Unpublished
lecture notes. Mining Engineering Department, Colorado School of Mines, Golden, CO, 1983.)
The factor As/A represents the proportion of intact rock in the potential plane
of failure (Marek, 1975). An estimate of this ratio is a critical input to stability
evaluation. However, it is not possible to actually measure the proportion of
intact rock present along the potential failure surface(s) because the removal
of the block above a fracture will guarantee that no intact rock will remain.
Therefore, it is recommended to conservatively estimate the ratio for each
significant discontinuity set as follows (Abel, 1983):
Minimum discontinuity spacing
Intact area
=
Total area Maximum trace length + Minimum discontinuity spacing
The termination index (Brown, 1981) provides a relative measure of the number of times the geological discontinuities terminate within intact rock. The
termination index TIR is given by
TIR =
100N IR
N IR +N J +N O
(4.9)
where
NIR is the number of trace lengths terminating within intact rock (IR)
NJ is the number of discontinuities terminating against other discontinuities (J–L or J–H)
NO represents the number of geological discontinuities whose trace
lengths are obscured (O) as illustrated in Figure 4.27
164
Geotechnical Design for Sublevel Open Stoping
1
2
3
1
2
3
Termination index
Face 1 = 79%
Face 2 = 89%
Face 3 = 86%
Termination index
Face 1 = 17%
Face 2 = 31%
Face 3 = 23%
1
3
Termination index
Face 1 = 45%
Face 2 = 58%
Face 3 = 59%
3
Termination index
Face 1 = 0%
Face 2 = 0%
Face 3 = 0%
2
1
2
FIGURE 4.40
Trace length continuity and the number of blocks formed. (From Windsor, C.R., Rock mass
characterization—A course on structural characterization and structural analysis, Course
notes for the Masters of Engineering Science in Mining Geomechanics, Western Australian
School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia,
1995, 462pp.)
The effect of trace length on block size and its formation within a rock mass
is shown in Figure 4.40. As the trace length increases, the number of fully
formed blocks increases as does the potential for instability within the
exposed faces of a rock mass.
4.5.5 Rock Mass Classification Models
An objective of rock mass characterization methodologies is to divide a rock
mass into domains of similar geotechnical characteristics. Rock mass variability and heterogeneity potentially leading to different stoping behaviors
can be determined using conventional rock mass classifications (Barton, et al.,
1974; Bieniawski, 1989). Cepuritis (2004) has described three-­dimensional
rock mass model construction methods and considerations to account for
local variations in rock mass conditions using conventional rock mass classification methods. The methodology is compatible with data collection during
orebody delineation and subsequent mapping, as long as there is a sufficient
quantity and quality of detailed geotechnical data available. Conventional
mining software can be used to display plans and sections of both stoping
layouts and structural data (Dempers et al., 2010).
Three-dimensional modeling of rock mass classification parameters is
somewhat similar to the process developed for geological and resource modeling, with the properties of a rock mass volume being simulated from a
Rock Mass Characterization
165
limited number of data points (see Figure 4.3). A typical process, as described
by Cepuritis (2004), is as follows:
1. Evaluation of input data sources, data accuracy, reliability, and data
distribution
2. Preliminary geotechnical domain definition
3. Determination of the most appropriate modeling types for each
domain
4. Compositing input parameters into regularly sized data intervals
5. Statistical analysis and sub/redomaining (if required)
6. Defining and applying interpolation techniques
7. Model verification
Important considerations include the process of geotechnical domain definition, the choice of interpolation model, and the display of the results in
three dimensions. The modeling results can be displayed using grid or block
models to show the variations in rock mass conditions throughout a stoping
block or even mine wide (Cepuritis, 2004; Dempers et al., 2010).
In particular, grid models can be very effective to show variations in rock
mass classification parameters within the hangingwall, footwall, and midore intersection of steeply dipping tabular orebodies extracted by open
stoping. Data from drillhole logging and underground mapping can be combined into a database and used as the data source for estimating the various
geotechnical parameters within a relevant grid model. The data source is
first composited into 1 m intervals with only data relevant to each grid model
surface used in the estimation process. For example, for an ore zone, only
data samples lying within the ore zone are used. Similarly, for the footwall
and hangingwall models, only data lying within a specified distance from
the ore contacts are used.
For each geotechnical domain, the rock mass classification data can be
examined to generate contoured models of estimated Jn, RQD, Jr, and Ja
parameters. The rock mass classification parameter values are calculated for
each point in the grid as per the conventional guidelines in Tables 4.3 and
4.4. The estimates can be subsequently contoured as shown in Figures 4.41
through 4.44.
4.6 Intact Rock Strength
Physical testing of suitable rock core specimens allows the determination
of the mechanical properties of intact rock required for sublevel open stope
166
Geotechnical Design for Sublevel Open Stoping
20
18.05
16.1
2215
2195
2175
2155
2135
2115
2095
2070
2200 RL
14.15
12.2
10.25
8.3
2045
2020
1995
1970
1945
1920
2000 RL
6.35
4.4
9000 E
8800 E
1800 RL
8600 E
0.5 Jn
8400 E
2.45
FIGURE 4.41
Contoured grid model of hangingwall joint set number Jn.
100
90
80
2215
2195
2175
2155
2135
2115
2095
2070
2200 RL
70
60
50
40
2045
2020
1995
1970
1945
1920
2000 RL
30
20
9000 E
1800 RL
8800 E
RQD
8600 E
0
8400 E
10
FIGURE 4.42
Contoured grid model of hangingwall RQD.
design using rock mass classification or numerical analysis methods. The
intact rock strength is commonly measured in uniaxial compression, point
load, indirect tensile, and triaxial compression tests (Brady and Brown,
2004). Usually, a limited (but representative) number of cylindrical specimens of each rock type should be tested for UCS in a suitable laboratory
167
Rock Mass Characterization
3
2.75
2.5
2215
2195
2175
2155
2135
2115
2095
2070
2200 RL
2.25
2
1.75
1.5
2000 RL
1.25
1920
1
2045
2020
1995
1970
1945
9000 E
1800 RL
8800 E
Jr
8600 E
.5
8400 E
.75
FIGURE 4.43
Contoured grid model of hangingwall joint roughness number Jr.
4.5
4.15
3.8
2215
2195
2175
2155
2135
2115
2095
2070
2200 RL
3.45
3.1
2.75
2.4
2000 RL
2.05
1920
1.7
2045
2020
1995
1970
1945
9000 E
1800 RL
8800 E
Ja
8600 E
1
8400 E
1.35
FIGURE 4.44
Contoured grid model of hangingwall joint alteration number Ja.
equipped with a stiff testing machine. A larger number of point load tests
can be carried out during the core logging process for orebody delineation.
A comprehensive set of suggested testing methods has been published by
the International Society for Rock Mechanics (ISRM) (Brown, 1981; Ulusay
and Hudson, 2007).
168
Geotechnical Design for Sublevel Open Stoping
4.6.1 Uniaxial Compressive Strength
The uniaxial (or unconfined) compressive strength (UCS) of the intact rock
is one of the most widely used parameters in rock engineering. The UCS
is determined by loading a cylindrical specimen between steel platens of
a similar diameter (Figure 4.45) and is calculated as the average peak axial
stress:
sc =
P
A
(4.10)
where
σc is the UCS
P is the peak axial load
A is the cylindrical specimen cross-sectional area
The UCS values for intact rock are usually reported in units of mega pascals
(MPa) where 1 MPa = 1 N/mm2.
The UCS can be readily determined from cylindrical specimens ranging from PQ, NX, and HQ core sizes (see Table 4.2). However, according
to the ISRM suggested methods (Brown, 1981; Ulusay and Hudson, 2007),
the diameter should be preferably not less than NX core size (54 mm).
In essence, the diameter of the specimen should be 10 times the largest
grain size.
FIGURE 4.45
Conventional uniaxial compression testing.
Rock Mass Characterization
169
For each rock type, a number of specimens are selected containing a
minimum number of discontinuities (preferably none) in an attempt to
cause failure to occur through the intact rock. In order to minimize the
end effect interaction between the specimen and testing platens, it is
recommended that the specimen shape be limited to cylinders having a
height to diameter ratio ranging from 2.5 to 3.0. Furthermore, the ISRM
suggested method provides for the test specimens to be loaded at either
end through a steel disk having a diameter of between D and D + 2 mm
and a thickness of 15 mm or D/3, where D is the diameter of the test
specimen.
The quality of sample preparation is critically important. The specimen
ends must be parallel to each other and aligned normal to the specimen axis.
In addition, the specimen ends must be lapped parallel to within 0.02 mm.
The recommended rate of loading during testing is 0.5–1 MPa/s, and the
specimens should be tested with water contents reflecting the field conditions.
Other factors that have been found to affect the recorded UCS values include
friction between the platens and the specimen ends, specimen geometry and
volume, testing system stiffness, rate of loading or strain rate (Brady and
Brown, 2004).
It is preferable to test some specimens with diametrically opposed vertical and horizontal strain gauges, or some other type of deformation gauge,
attached to permit axial and lateral deformations to be recorded during testing so that the deformation moduli, Young’s modulus and Poisson’s ratio,
can be determined. Monitoring deformations during loading is also useful
in checking for the effects of poor end preparation and nonuniform load
application.
Intact rock material contains mineral grains, cracks, and pores of different sizes and orientations. This can result in large variations in intact
rock strength being obtained from a suite of similar or apparently identical samples. Depending on the grain size distribution and the microstructure of the rock, the fracture or failure process begins with the initiation of
damage caused by small cracks growing in the direction of the maximum
applied load at an applied stress of approximately 0.25–0.5 times the UCS
of the intact rock. As the axial load increases, these stable cracks continue
to accumulate. Eventually, when the specimen contains a sufficient density
of cracks, they start to interact, and an unstable cracking process may be
initiated.
The way in which the macrofailure of the specimen develops as it
unloads after the peak axial load has been reached depends on the relative stiffnesses of the specimen and the loading system. This failure process can be controlled and studied using servo-controlled testing systems.
It must also be recognized that differences in the relative stiffnesses of the
intact rock and the testing system in the laboratory, and of the rock mass
and the surrounding rock loading it around an underground excavation,
170
(a)
Geotechnical Design for Sublevel Open Stoping
(b)
(c)
(d)
FIGURE 4.46
(a) Axial splitting—mode A, (b) shear through intact rock—mode Bi, (c) shear through
structure—mode Bs, and (d) multiple cracking failure—mode C.
may be quite different and so may produce different macrofailure modes.
A detailed discussion of this issue is beyond the scope of this book.
The interested reader is referred to the account provided by Brady and
Brown (2004).
The macrofailure modes usually observed in uniaxial compression tests
on isotropic specimens of rock can be classified as axial splitting (failure mode A), shearing through intact rock (failure mode Bi), and multiple cracking (failure mode C). If a specimen is anisotropic or contains a
plane of weakness or structural feature of some type, macrofailure may
occur along a preferred orientation (failure mode Bs). Figure 4.46 shows
examples of these four failure modes obtained in the WASM stiff testing
machine.
As shown in Figure 4.47, for a given rock type, a wide range of UCS values may be obtained for each failure mode. The UCS of intact rock used
for stope design should be determined by considering only the strengths
of specimens failing by axial splitting, multiple cracking, or shear through
intact rock. Failure through intact rock appears to be normally distributed,
while structurally controlled failure appears to be lognormally distributed
(Figure 4.48). The data shown in this figure are from BX core size, which is a
typical core used for deep drilling in the mining industry.
Up-to-date, conventional open stope design requiring UCS input largely
relies on average values per rock type rather than considering the spatial variability of the samples. It is recommended that UCS databases
take into account the spatial location of the samples in order to determine the local UCS values for different stope locations within an orebody
(Figure 4.49).
0%
5%
10%
15%
20%
25%
30%
0–20
20–40
40–60
FIGURE 4.47
UCS data dispersion per failure mode.
Observations
35%
60–80
Mode B — Shear
Mode C — Multiple cracking
UCS Range (MPa)
80–100 100–120 120–140 140–160 160–180 180–200 200–220 220–240 240–260 260–280 280–300
Mode A — Axial splitting
Rock Mass Characterization
171
172
Geotechnical Design for Sublevel Open Stoping
Frequency
Rock type: shale — Core size: BX
45
40
35
30
25
20
15
10
5
0
N = 140
X = 214
S2 = 1707
0
50
100
150
200
250
300
350
400
450
UCS (MPa)
(a)
35
N = 68
X = 99
S2 = 1999
Frequency
30
25
20
15
10
5
0
0
50
100
150
200
250
300
UCS (MPa)
Frequency
(b)
35
30
25
20
15
10
5
0
N = 68
X = 112
S2 = 2297
0
50
100
150
200
250
300
350
UCS (MPa)
(c)
FIGURE 4.48
Distributional nature and values of UCS per failure mode. (a) Through intact rock, (b) along
bedding, and (c) along other weakness.
173
Rock Mass Characterization
–140 UCS
(MPa)
2200 RL
–130
–120
–110
–100
–90
2000 RL
–80
9000 E
1800 RL
8800 E
–50
8600 E
–60
8400 E
–70
FIGURE 4.49
Modeled UCS variability across an orebody hangingwall boundary.
4.6.2 Point Load Strength
Field- or laboratory-based point load strength testing can be used to complement the conventional laboratory UCS testing. Point load testing is performed by a portable device with the specimen being a piece of core or an
irregular rock lump. In both cases, the specimens are loaded using a pair of
spherically truncated conical platens (Figure 4.50).
The point load strength (Is) is calculated as follows:
Is =
P
(De)2
(4.11)
where
P is the failure load
De is the equivalent diameter, given by
De =
4WD
p
(4.12)
where W and D are the dimensions of the specimen calculated as shown in
Figure 4.51.
Diametral and axial point load testing can be routinely carried in the field
as part of an orebody delineation process. The standard for diametrical
point load testing is for a 50 mm core (Is50). Smaller-diameter cores may yield
174
Geotechnical Design for Sublevel Open Stoping
FIGURE 4.50
Portable point load tester with electronic measurement of load.
P
P
L
De
D
D
W
L > D/2
(a)
L
P
W
L
De
D
P
W1
D
De
W2
L > D/2
(c)
0.3W < D < W
(b)
0.3W < D < W
W = (W1 + W2)/2
(d)
FIGURE 4.51
Specimen shape requirements for (a) diametral test, (b) axial test, (c) block test, and (d) irregular lump. (From Brown, E.T., ed., Rock Characterization, Testing and Monitoring—ISRM Suggested
Methods, Pergamon, Oxford, U.K., 1981, 211pp.; Int. J. Rock Mech. Min. Sci. Geomech., 22(2),
ISRM, Suggested method for determining point load strength, 51–60, Copyright 1985, with
permission from Elsevier.)
175
Rock Mass Characterization
500
UCS = 20.103Is(50)
R2 = 0.4203
UCS (MPa)
400
300
200
100
0
0
5
10
15
20
25
Point load index—Is(50) (MPa)
FIGURE 4.52
UCS versus point load index (Is(50)).
higher point load strengths, since those specimens are less likely to contain
preexisting flaws, and a correction factor is required as follows (Brown, 1981;
ISRM, 1985; Ulusay and Hudson, 2007):
0.45
ÊD ˆ
I s(50 ) = Á e ˜
Ë 50 ¯
(4.13)
Although theoretical considerations show that Is provides a measure of
tensile strength, the experimental results show that Is is also sufficiently
related to σc as shown in Figure 4.52. The data in the figure were determined by calculating pairs of σc and Is from adjacent pieces of core for a
large number of deposits and host rock masses in Australia. On average, σc
is about 20 times Is(50), which agrees with the multiplication factor of 20–25
suggested by Broch and Franklin (1972) and the ISRM (1985). In some rocks,
different Is(50) values are obtained when the core sample is loaded axially
or diametrally.
4.6.3 Confined Compressive Strength
As noted in the opening of Section 4.6, triaxial compression testing is one
of the commonly used mechanical property tests for intact rock. This test
is carried out on cylindrical samples subjected to a range of uniform allround confining pressures and loaded in axial compression (Figure 4.53).
Discussion of the test procedures, the factors influencing rock behavior in
176
Geotechnical Design for Sublevel Open Stoping
σ1
Platens
σn
σ3
α
e
lan
τn
p
re
ilu
Fa
σ3
Specimen
σ1
FIGURE 4.53
Schematic of triaxial compressive testing showing failure plane.
these tests, and the interpretation of the test results is beyond the scope of
this book. Discussions of these types are given by Brady and Brown (2004)
and Hoek (2007), for example.
Figure 4.54 shows the complete axial stress (σa)–axial strain (εa) curves
obtained by Wawersik and Fairhurst (1970) in a series of triaxial compression
tests carried out on a marble. These and similar data for other rocks show
that, with increasing confining pressure, the following results are obtained:
a. The peak strength increases.
b. A transition occurs from typically brittle (with a postpeak reduction in strength) to fully ductile (continued deformation at constant
differential stress) behavior with the introduction of plastic mechanisms of deformation including cataclastic flow and grain-sliding
effects.
c. The region incorporating the peak of the σa–εa curve flattens and
widens.
d. The postpeak drop in stress to the residual strength reduces and
disappears at high values of the confining pressure, σ3.
177
Rock Mass Characterization
Increasing confining
pressure (MPa)
300
48.3
Axial stress (MPa)
34.5
27.6
200
20.7
13.8
100
0
6.9
3.4
0
0
0.10
0.20
0.30
0.40
0.50
0.60
0.70
Axial strain, εa (%)
FIGURE 4.54
Complete axial stress–axial strain curves obtained in triaxial compression tests on Tennessee
marble at the confining pressures indicated by the numbers on the curves. (After Int. J. Rock
Mech. Min. Sci., 7(5), Wawersik, W.R. and Fairhurst, C., A study of brittle rock fracture in laboratory compression tests, 561–575, Copyright 1970, with permission from Elsevier.)
σ1
τn
ϕ
ψ
σc
c
2β
(a)
σ3
σ1
σn
(b)
σ3
FIGURE 4.55
Mohr–Coulomb peak strength envelopes in terms of (a) shear and normal stresses, and (b)
principal stresses. (From Brady, B.H.G. and Brown, E.T., Rock Mechanics for Underground Mining,
3rd edn., Kluwer, Dordrecht, the Netherlands, 2004, 628pp.)
As illustrated in Figure 4.55, the peak axial stress (σ1) reached at each value
of confining pressure (σ3) in a series of triaxial compression tests may be
plotted as Mohr’s circles of stress on shear stress (τn)–normal stress (σn)
axes (Figure 4.55a), or as plots of σ1 against σ3 (Figure 4.55b). The resulting
peak strength envelopes for intact rock are customarily curved and may
be described by the nonlinear Hoek–Brown empirical strength equation
to be introduced in Section 4.7.2 or by other empirical criteria (Brady and
178
Geotechnical Design for Sublevel Open Stoping
Brown, 2004). However, for many rocks, particularly over limited ranges of
the stresses, σn and σ3, they may be approximated closely by straight lines.
As shown in Figure 4.55a, the straight line peak strength envelope on τn–σn
axes is a representation of the classical Coulomb (often referred to as the
Mohr–Coulomb) shear strength criterion:
tn = c + s n tan j
(4.14)
where
c is the cohesion
φ is the angle of internal friction
Note that the intercept of the principal stress envelope on the σ1 axis
(Figure 4.55b) gives the UCS, σc. The slope of the σ1 versus σ3 envelope, ψ, is
a function of the angle of internal friction, φ, as follows:
tan y =
1 + sin j
1 - sin j
(4.15)
The UCS, σc, is related to cohesion, c, and the angle of internal friction, φ, as
follows (Brady and Brown, 2004):
sc =
2c cos j
1 - sin j
(4.16)
The Mohr–Coulomb criterion is also used as the basis of a range of expressions used to describe the shear strengths of smooth, rough, and infilled
discontinuities in rock (Brady and Brown, 2004).
4.7 Mechanical Properties of Rock Masses
In analyzing geotechnical problems encountered in the design of open
stoping layouts and sequences, often by using numerical modeling, it is
necessary to estimate the mechanical properties of the rock mass, usually represented by its stress–strain behavior. Important aspects of this
behavior are the constants relating stresses and strains in the elastic range,
the stress levels at which yield, fracturing, or slip occurs within the rock
mass, and the postpeak stress–strain behavior of the fractured or failed
Rock Mass Characterization
179
rock (Brady and Brown, 2004). The collection of data for use in estimating
some of these properties is part of the rock mass characterization process
discussed in this chapter.
In some problems, it is the behavior of the intact rock material discussed
in Section 4.6 that will be of concern. This will be the case when considering the excavation of rock by drilling and blasting (to be discussed in
Chapter 6) or when considering the stability of excavations in good quality brittle rock. In other cases, the behavior of single discontinuities or
of small numbers of discontinuities may be of paramount importance.
This class of problem includes the equilibrium of blocks of rock formed
by the intersection of three or more discontinuities with the roof or wall
of an excavation, and cases in which slip on a fault must be considered.
A different class of problem is that in which the rock mass must be analyzed as an assembly of discrete blocks as discussed in Section 3.3. In
this case, the normal and shear force–displacement relations at face-toface and corner-to-face block contacts are of importance in the analysis.
Finally, it is sometimes necessary to consider the overall response of a
jointed rock mass in which the discontinuity spacing is small on the scale
of the problem domain and the rock mass can be treated as an equivalent
continuum having isotropic material properties. The remainder of Section
4.7 will consider the strength and deformability of rock masses in these
circumstances.
4.7.1 Hoek–Brown Empirical Strength Criterion
In an attempt to provide a first-pass method of estimating the strength of
jointed rock masses for use in underground excavation design, Hoek and
Brown (1980) developed an empirical rock mass strength criterion based
on their earlier work on the brittle fracture of rock and the mechanical
behavior of discontinuous rock masses. The criterion took the strength
of the intact rock as the starting point and introduced factors to reduce
the strength on the basis of the spacing and characteristics of the joints
within the rock mass. Initially, Hoek and Brown (1980) used the 1976 version of Bieniawski’s RMR (see Section 4.2.3) as an index of the geological
characteristics considered likely to influence the mechanical properties of
the rock mass.
Because of a lack of suitable alternatives, the Hoek–Brown criterion was
soon adopted by rock mechanics practitioners and sometimes used for purposes for which it was not originally intended and which lay outside the
limits of the data and methods used in its derivation. Because of this, and
as experience was acquired with its practical application, a series of changes
were made and new elements were introduced into the criterion (e.g., Hoek
and Brown, 1997). The current version of the criterion is that given by Hoek
et al. (2002) and discussed by Hoek (2007).
180
Geotechnical Design for Sublevel Open Stoping
The generalized Hoek–Brown empirical strength criterion for jointed rock
masses is given by
a
È Ês ˆ ˘
s1 = s 3 + sci Ím b Á 3 ˜+ s ˙
Î Ë sc ¯ ˚
(4.17)
where
σ1 and σ3 are the major and minor principal stresses at peak strength
σci is the UCS of the intact rock
mb is a parameter that reflects the frictional strength of the rock mass
s is a parameter that reflects the cohesive strength of the rock mass and
depends on the rock mass quality as does the index a, which takes a
value of close to 0.5 for hard, fresh rock
For intact rock, s = 1.0.
The values of mb, s, and a are related to the GSI of the rock mass by the
following relations:
m b = m ieGSI-100 28-14D
(4.18)
s = eGSI-100 9-3D
(4.19)
and
a = 0.5 +
e-GSI 15 - e-20 3
6
(4.20)
where
mi is a strength parameter for the intact rock (Figure 4.56)
D is disturbance factor that varies with the degree of disturbance due to
blast damage and stress relaxation
D varies from 0 for undisturbed in situ rock masses to 1.0 for very disturbed
rock masses.
The rock material parameter, m i, is obtained by the statistical analysis of
a set of triaxial compression tests on carefully prepared 50 mm diameter
core samples of the intact rock. If it is not possible to carry out a set of
triaxial tests, m i may be estimated as σci/T where T is the uniaxial tensile
strength of the intact rock (Brown, 2007). Because of potential differences
in the failure mode, the value of the UCS estimated from the intercept
of the peak strength envelope with the σ1 axis as shown in Figure 4.55b
181
Rock Mass Characterization
Rock
Type
Class
Group
Texture
Coarse
a
Conglomerates
Brecciasa
Sedimentary
Clastic
Carbonates
Nonclastic
Crystalline
limestone
(12 ± 3)
Evaporates
Medium
Fine
Sandstones
(17 ± 4)
Siltstones
(4 ± 2)
Graywackes
(18 ± 3)
Sparitic
limestone
(10 ± 2)
Gypsum
(8 ± 2)
Micritic
limestone
(9 ± 2)
Anhydrate
(12 ± 2)
Metamorphic
Organic
Marble
(9 ± 3)
Nonfoliated
Slightly foliated
Foliatedb
Light
Igneous
Plutonic
Dark
Hypabyssal
Volcanic
Lava
Pyroclastic
Hornfels
Quartzites
(19 ± 4)
(20 ± 3)
Metasandstone
(19 ± 3)
Migmatite Amphibolites
Gneiss
(29 ± 3)
(26 ± 6)
(28 ± 5)
Phyllites
Schists
(7 ± 3)
(12 ± 3)
Granite
Diorite
(32 ± 3)
(25 ± 5)
Granodiorite
(29 ± 3)
Gabro
Dolerite
(27 ± 3)
(16 ± 5)
Norite
(20 ± 5)
Porphyries
Diabase
(20 ± 5)
(15 ± 5)
Dacite
Rhyolite
(25 ± 3)
(25 ± 5)
Andesite
Basalt
(25 ± 5)
(25 ± 5)
Agglomerate
Breccia
Tuff
(19 ± 3)
(19 ± 5)
(13 ± 5)
Very Fine
Claytones
(7 ± 2)
Shales
(6 ± 2)
Marls
(7 ± 2)
Dolomites
(9 ± 3)
Chalk
(7 ± 2)
Slates
(7 ± 4)
Peridotite
(25 ± 5)
FIGURE 4.56
Values of the constant m i for intact rock by rock group. (From Hoek, E. et al., Support of
Underground Excavations in Hard Rock, Balkema, Rotterdam, the Netherlands, 1995.)
a Conglomerates and breccias may present a wide range of m values, depending on the nature
i
of the cementing material and the degree of cementation, so they may range from values similar to sandstone to values used for fine grained sediments (even under 10).
b These values are for intact rock specimens tested normal to bedding or foliation. The values of
mi will be significantly different if failure occurs along a weakness plane.
182
Geotechnical Design for Sublevel Open Stoping
may differ from the mean value of UCS obtained from a series of uniaxial compression tests as discussed in Section 4.6.1. It is the value of UCS
obtained by extrapolating the peak strength envelope back to σ3 = 0, represented by the symbol, σci, that should be used in the Hoek–Brown criterion
(Equation 4.17).
4.7.2 Rock Mass Deformation Modulus
As noted in Section 4.7.1, stress and deformation analyses of the responses of
rock masses to the creation of mining excavations within them require the
input of a range of parameters describing their stress–strain behavior. In the
case being considered here in which a jointed rock mass may be represented
as an isotropic equivalent continuum, the main parameter required is the
deformation modulus, Em.
Over the years, a wide range of methods of estimating Em for different
purposes have been proposed in the literature. In the main, these methods
use some measure of rock mass quality such as joint spacing, RQD, RMR,
Q, or GSI to give empirical estimates of the rock mass modulus, Em, sometimes by reducing the modulus of the intact rock, Ei (e.g., Bieniawski, 1976;
Serafim and Periera, 1983; Barton, 2002; Zhang and Einstein, 2004; Hoek and
Diederichs, 2006). Figure 4.57 shows plots of a range of measured values of
Em fitted by the equations based on RMR proposed by Bieniawski (1978) and
by Serafim and Periera (1983), and an equation based on Qc = Q σc/100 proposed by Barton (2002).
As Figure 4.57 shows, these empirical equations generally give unrealistically high estimates of rock mass modulus at high values of RMR
or Q, in some cases being asymptotic to infinity as RMR approaches 100.
Hoek and Diederichs (2006) evaluated a wider range of field measurements of rock mass deformation modulus and fitted them by a sigmoidal
relation to overcome the problem of exponentially increasing values of Em
at high values of RMR, Q, or GSI. The expression developed by Hoek and
Diederichs (2006) is
È
1 - (D 2 ) ˘
Em = Ei Í0.02 +
60 +15 D-GSI ) 11 ˙
1+e (
Í
˙
˚
Î
where
Ei is the modulus of the intact rock
GSI is the Geological Strength Index introduced in Section 4.4.6
D is the disturbance factor introduced in Section 4.7.2
(4.21)
183
Rock Mass Characterization
Compromise RMR = 15 log Q + 50
Deformation modulus Emass (GPa)
90
80
Emass = 2 RMR – 100
70
60
Emass = 10 Q1/3
c
50
40
30
Emass = 10
20
Case histories
(RMR – 10)
40
Bieniawski (1978)
Serafim and
Pereira (1983)
10
0
0
10
0.001
30
40
50
60
70
80
20
Geomechanics rock mass rating (RMR)
0.01
0.1
1.0
Q rating
10
100
90
100
1000
FIGURE 4.57
Measured values of static rock mass modulus, Em, and some empirical relations. (After Int. J.
Rock Mech. Min. Sci., 39, Barton, N., Some new Q-value correlations to assist in site characterization and tunnel design, 185–221, Copyright 2002, with permission from Elsevier.)
If a laboratory-determined value of Ei is not available, a value may be estimated from Ei = (MR) σc, where MR is the modulus ratio for the rock type
concerned as given in a table provided by Hoek and Diederichs (2006) and
σc is the UCS of the intact rock.
4.8 Rock Stress
In sublevel open stoping, knowledge of the in situ stress field is critical in
order to achieve extraction sequences giving 100% recovery with minimal
dilution and ore loss. In particular, the stress field data are used as an
input to rock mass classification and numerical modeling, thus enabling
various sized, shaped, and oriented stopes to be arranged and extracted
within manageable expressions of rock mass failure. Clearly, formal engineering design of open stoping including pillars cannot be attempted
without a reasonable knowledge of the stress field. Figure 4.58 shows a
184
Geotechnical Design for Sublevel Open Stoping
FIGURE 4.58
Highly stressed stope and pillar damaged by rock bursting.
highly stressed stope and pillar where excessive stress caused significant
seismicity and related damage.
Another typical expression of high stress is commonly found around
the vertical development required in open stoping, such as raises for cutoff slots and ventilation shafts or even blastholes. Large concentrations
of stress at the boundaries of subvertical raises create rock mass failures
that can be used to estimate the orientation of the major principal stress
(Figure 4.59).
4.8.1 Stress Tensor
Stress is a mathematical concept used to represent stored strain energy
within a rock mass volume. However, it is beyond the scope of this book to
address with any detail the tensorial nature of stress in three dimensions.
For a complete description of the fundamental principles of stress, the reader
is advised to study the books by Brady and Brown (2004) and Hudson and
Harrison (1997). This book will focus on stress measurements using oriented
exploration core and their interpretation.
A reliable and representative estimation of in situ stress is a major
requirement for the optimized design of an extraction sequence of open
stopes, especially at depth. Stress tensor notation can be represented as
follows:
185
Rock Mass Characterization
FIGURE 4.59
Raise wall damage due to excessive horizontal stresses.
s11
sij = t21
t31
t12
s 22
t32
t13
t23
s 33
(4.22)
The tensor may be transformed to a unique orientation in which the normal
stresses are maximized and the shear stresses vanish. These maximized normal stresses are termed the principal stresses, denoted by σ1, σ2, and σ3 and
referred to as the major, intermediate, and minor principal stresses, respectively (Brady and Brown, 2004).
4.8.2 Stress Measurements Using Oriented Core
Stress measurements using oriented core are classified as destressing–
restressing techniques. These techniques involve completely decoupling
a volume of rock from the stress field, then reloading the rock volume
back to its original stressed condition (Villaescusa et al., 2003b). The intention is to return a rock core volume to its in situ state. The method discussed here has been called the WASM AE stress measurement technique
(Villaescusa et al., 2002). It is a technique that utilizes a completely decoupled volume of rock from exploration core that is reloaded to its original
stress state by reference to one indirect parameter, the acoustic emission
event count.
186
Geotechnical Design for Sublevel Open Stoping
Basically, the method involves a sequence of six steps:
1. An oriented sample volume, usually common oriented exploration core (termed here the main core), is isolated from a rock mass.
2. The main core is transported to a rock mechanics laboratory and
resampled by a number of smaller subcores that are taken at certain
orientations relative to the axis of the main core.
3. The oriented subcores are precision ground for rightness and flatness, then fitted with suitable acoustic emission sensors.
4. Each subcore is tested under monotonically increasing uniaxial
load (stress). The acoustic sensors measure the event count rate
attributed to the deformation, dislocation, and propagation of preexisting cracks and the initiation of new cracks, as the stress is
increased.
5. The applied stress versus the count rate is approximately bilinear with
the change of relationship indicated by a demonstrable increase in
noise count rate at a certain stress level (Figure 4.60). This transition point is taken to indicate the largest contemporary stress experienced by the subcore in the direction of the subcore axis.
6. The stress measurements for the oriented subcores are used in conjunction with their orientations relative to the original oriented core
to determine the largest contemporary stress field experienced by
the main core (Figure 4.61). Provided the rock specimen has been
selected from an area previously in equilibrium with gravitational
loading and tectonics (Windsor et al., 2006, 2007), this is the maximum previous stress to which a particular rock mass has been subjected by its environment.
Cumulative AE events
30
25
20
Previous maximum stress
15
10
5
0
0
5
10
15
Stress (MPa)
FIGURE 4.60
Typical AE cumulative events versus applied uniaxial stress.
20
25
187
Rock Mass Characterization
WASM AE stress measurements
Pole plot
N
σ1
σ2
E
W
σ3
S
WA School of Mines
90
σ1 = 0.0406 × Depth + 6.1
Stress magnitude (MPa)
80
σ2 = 0.0334 × Depth + 1.7
70
σ3 = 0.0270 × Depth
60
σv = 0.0278 × Depth
50
40
30
20
10
0
0
200
400
600
800 1000 1200 1400
Vertical depth from surface (m)
1600
1800
FIGURE 4.61
Principal stress orientations and magnitudes determined using oriented core.
2000
188
Geotechnical Design for Sublevel Open Stoping
This section presents the scalar characteristics (i.e., the stress magnitudes alone)
from approximately 240 WASM AE rock stress tensor determinations obtained
from different geological and geodynamic regimes from different continents and compares them to results compiled in an Earth Rock Stress Tensor
Database (ERSTD) (Windsor, 2009). The data comprise results from techniques
that attempt to measure, without a priori assumption, the complete rock stress
tensor (e.g., it does not include results obtained from the hydraulic fracturing
technique). The data are presented as reported, without prejudice or censorship.
The distributions of the vertical stress, the principal normal stresses, and
the maximum shear stress with depth in the upper 3 km of Earth’s crust
from the WASM AE data set and from the ERSTD are shown in Figures
4.61 through 4.63, respectively. Figure 4.62 indicates that both data sets are
distributed about a theoretical linear relationship for vertical stress given
Vertical stress (MPa)
0
0
20
40
60
80
100
120
ERSTD
WASM AE
500
Depth (m)
1000
1500
2000
2500
Theoretical vertical stress
(unit weight 27 kN/m3)
3000
FIGURE 4.62
Distribution of vertical stress with depth, measured by WASM AE and from the ERSTD. (From
Villaescusa, E. et al., Stress measurements from oriented core—A decade of results, Presented
at MassMin 2012, Sixth International Conference & Exhibition on Mass Mining, Sudbury, Ontario,
Canada, June 10–14, 2012a, Paper 6842, 9pp.)
189
Rock Mass Characterization
–20
0
500
0
Principal normal stresses σ1, σ2, σ3 (MPa)
20
40
60
80
100
120
140
160
S1 ERSTD
S2 ERSTD
S3 ERSTD
S1 WASM AE
S2 WASM AE
S3 WASM AE
Depth (m)
1000
1500
2000
2500
3000
FIGURE 4.63
Distributions of principal normal stresses with depth, measured by WASM AE and from
the ERSTD. (From Villaescusa, E. et al., Stress measurements from oriented core—A decade of
results, Presented at MassMin 2012, Sixth International Conference & Exhibition on Mass Mining,
Sudbury, Ontario, Canada, June 10–14, 2012a, Paper 6842, 9pp.)
by σv = zγr where z is the overburden depth and γr is the unit weight of
rock, which is set here at 27 kN/m3. The WASM AE data appear to fit better
with this relation than the ERSTD (Villaescusa et al., 2012).
The distribution of principal normal stresses (σ1, σ2, and σ3) with depth
given in Figure 4.63 shows a low frequency of tensor measurement below
1.5 km, with scatter increasing with depth. It indicates slight nonlinearity of
the WASM AE data set and greater nonlinearity of the ERSTD. Note that the
ERSTD is influenced at depth by a greater frequency of deeper- and lowerstress magnitudes measured around South African mine sites. Figure 4.64
shows the distribution of the maximum shear stress from WASM AE and from
the ERSTD. Both data sets show nonlinearity and considerable scatter with
depth, which is thought to be linked to the variability in the shear strength of
Earth’s crust and its ability to sustain shear stresses (Windsor, 2009).
190
Geotechnical Design for Sublevel Open Stoping
Maximum shear stress τmax (MPa)
0
0
10
20
30
40
50
ERSTD
500
WASM AE
Depth (m)
1000
1500
2000
2500
3000
FIGURE 4.64
Distribution of maximum shear stress with depth, measured by WASM AE and from
the ERSTD. (From Villaescusa, E. et al., Stress measurements from oriented core—A decade of
results, Presented at MassMin 2012, Sixth International Conference & Exhibition on Mass Mining,
Sudbury, Ontario, Canada, June 10–14, 2012a, Paper 6842, 9pp.)
5
Span and Pillar Design
5.1 Background
The development of sublevel open stope mining methods enhanced the
mechanization and increased productivity of underground bulk mining
operations. This in turn led to a need to optimize the size and shape of the
open stopes in order to maximize production. Unacceptable waste dilution plagued many bulk mining operations, and traditional trial-and-error
approaches to optimizing stope dimensions became economically unacceptable. Furthermore, inadequate design methodologies often resulted in failure of secondary stopes with resulting production delays, increased costs,
and, in some cases, loss of ore reserves. In this chapter, modern stope and
pillar design methodologies will be discussed.
5.2 Empirical Span Determination Using
Rock Mass Classification Methods
Rock masses represent extremely complex media in which to design and
construct engineered structures. During the early design stages of a project,
such as the scoping and prefeasibility stages, when little detailed information
on a rock mass and its stress and hydrologic characteristics are available, the
use of a rock mass classification scheme can be of benefit. At its simplest, this
may involve the use of a classification scheme as a checklist to ensure that
some geotechnical information has been considered. At the other extreme,
one or more classification schemes can be used to build up a picture of the
composition and characteristics of a rock mass to provide initial estimates of
allowable spans and support requirements, and to provide estimates of its
strength and deformation responses to the excavation process.
Classification and its application to underground support is primarily founded in civil engineering tunnel construction (e.g., rock quality
191
192
Geotechnical Design for Sublevel Open Stoping
designation (RQD)—Deere et al., 1967; rock mass rating (RMR)—Bieniawski,
1989; tunnel quality index (Q)—Barton et al., 1974). Due to the relatively
modest depth (0–500 m) of many of these case studies and the relatively
high-safety factors demanded in civil works, design recommendations from
these classification systems may be difficult to apply directly in an open
stoping context. They can, however, provide a first or conservative estimate of allowable span and support requirements. Laubscher and Taylor
(1976) and Laubscher (1993) modified RMR for use in the design of blockcaving mines. Caving operations are beyond the scope of this book, and
Laubscher’s method will therefore not be discussed further. Mathews et al.
(1980) and Potvin (1988) modified the Q system and applied it to open stope
design. Their methodology has been modified slightly and is presented in
this chapter.
A problem with rock mass classifications is that, in addition to being conservative, they are likely to be missing a key parameter, for example, joint
termination (see Chapter 4). Furthermore, the stress path is not really considered and this is a significant difference with respect to civil engineering,
where there is less interaction among excavations compared to the complex
extraction sequences utilized in the mining industry.
5.2.1 Span Determination Using Bieniawski’s RMR System
The rock mass rating (RMR) system was originally developed by Bieniawski
(1973). Over the years, it has been successively refined, as more case studies have been added to its database. The reader should be aware that, over
time, Bieniawski has made several changes to the ratings assigned to the
different parameters (Bieniawski, 1976, 1989). Figure 5.1 presents an additional modification to Bieniawski’s (1989) span versus stand-up time graph.
The changes have been made to account for the very large and stable open
spans that are being achieved in massive silicified skarns at medium confining stress (Figure 5.2). This is in part due to the silicification of the orebodies and host rocks, the relatively shallow depths being mined and also the
favorable condition of the geological discontinuities with respect to the
exposed spans. The concept of stand-up time was originally conceived by
Lauffer (1958, 1960) to indicate the time period within which an excavation
will remain serviceable and after which significant instability and caving
would be experienced. A stope span is defined as the minimum dimension
of an open stope wall.
Hutchinson and Diederichs (1996) have presented the maximum stable
unsupported span as a function of Bieniawski’s (1989) RMR (RMR89) value
as shown in Figure 5.3. In the absence of large-scale geological discontinuities, or very high induced stress, a temporary mine opening such as a
10 m-wide drill drive in downhole bench stoping can be analyzed. If the
required stand-up time is typically less than 5 years, then it can be seen
193
Span and Pillar Design
50
1
h
10
h
1
day
1
week
1
month
20
5
10
years years
10
90
80
70
R 89
RM
15
60
50
8
40
6
5
4
30
50
40
2
RMR 89
No support
required
30
1
70
60
3
1
6
1
months year
Immediate
collapse
30
Unsupported span (m)
Maximum unsupported stand-up time
10
100
1,000
10,000
100,000
Maximum unsupported stand-up time (h)
Tunnels
U/G mines
FIGURE 5.1
Unsupported tunnel limits. (Modified from Bieniawski, Z.T., Engineering Rock Mass Classification,
John Wiley, New York, 1989, 251pp. With permission.)
that for a rock mass having an RMR89 of greater than 80, the drill drive may
not need systematic cablebolt reinforcement, with the exception of bolts and
mesh for personnel safety.
The RMR89 data shown in Figure 5.4 indicate that few unsupported spans
are stable when their dimension exceeds 20 m. This is due to the majority of
the data being collected in cut-and-fill operations (Pakalnis, 2002), where full
operator access is required and potentially unstable spans cannot be effectively stabilized even with the implementation of cablebolting. However,
recent experience in open stoping in extremely hard rock mines, where the
orebody and host rocks have been altered by a strong silicification, unsupported stable spans ranging from 20 to 40 m can be safely achieved. The
open stoping data (spans exceeding 20 m) in Figure 5.4 show circles representing stable spans (depths of failure less than 2 m), square symbols representing transitional spans (depths of failure ranging from 2 to 4 m), and
194
Geotechnical Design for Sublevel Open Stoping
FIGURE 5.2
Very large and stable span exceeding 25 m, Sabinas mine, Mexico. (Photo courtesy of Peñoles,
Mexico City, Mexico.)
Fair
25
Good
Very good
s
ear
s
ear
5y
ear
1y
6m
ont
hs
ay
1w
10
1m
ont
eek
h
15
Hard rock
mine design zone
Unsupported stand-up time
Immediate collapse
10 y
20
1d
Maximum stable unsupported span (m)
Span
5
No support required
0
40
50
60
70
80
90
Rock mass rating (RMR89)
FIGURE 5.3
Alternate representation of RMR89 stand-up time guidelines. (From Hutchinson, D.J. and
Diederichs, M.S., Cablebolting in Underground Mines, Bitech Publishers, Richmond, British
Columbia, Canada, 1996, 406pp. With permission.)
195
Span and Pillar Design
nal
50
Tra
nsi
ti o
45
Unsupported span (m)
40
35
30
Stable
Unstable
25
20
15
10
5
0
10
20
30
40
50
60
70
80
90
100
Rock mass rating (RMR89)
FIGURE 5.4
Span design using the RMR89 method. (After Pakalnis, R., Empirical design methods—UBC
geomechanics an update, in R. Hammah, W. Bawden, J. Curran, and M. Telesnicki, eds., Mining
and Tunnelling Innovation and Opportunity, Proceedings of the 5th North American Rock Mech
Symp & 17th Tunnelling Association of Canada Conference, Toronto, July 7–10, 2002, pp. 203–210,
University of Toronto, Toronto, Ontario, Canada.)
triangles representing unstable spans (depths of failure exceeding 4 m). The
data can also be used as a guideline for design against immediate collapse,
large instabilities, or an indication where systematic cablebolting may be
required. A point to notice when using the RMR89 method for span design is
that the stress path effects, as well as the localized effect of large-scale structures likely to form wedges, must also be considered. Hence, for safe access,
ground support is always recommended for sublevel stope access infrastructure, even in very hard rock masses.
Modern sublevel open stoping mines use cavity-monitoring systems
(CMS) to continually collect data and develop databases that encompass
the final geometry of the stope voids. Stope performance is determined by
the depth of failure, which is defined as the distance from a design surface
to a resulting wall following complete stope extraction (Villaescusa, 2004).
Furthermore, rock mass classification databases from drill holes (Cepuritis,
2004; Dempers et al., 2010) can be used to establish contours of RMR89 values
for each stope wall (Figure 5.5).
The rock mass classification data coupled with the depths of failure from
the CMS and the design stope geometry can be used to establish relationships similar to that shown in Figure 5.6. The proposed limits for the stable
(depth of failure < 2 m), transitional (depth of failure 2–4 m), unstable (depth
of failure 4–6 m), and collapsed (depth of failure > 6 m) regions for stope
spans exceeding 20 m are usually based upon local mine economics.
196
Geotechnical Design for Sublevel Open Stoping
85
82
2200 RL
79
76
73
70
67
2000 RL
64
RMR
8600 E
8400 E
55
1800 RL
9000 E
58
8800 E
61
FIGURE 5.5
Contoured grid model of hangingwall RMR89 values.
50
45
40
6.8 m
6.2 m
30
20.0 m
5.5 m
25
7.0 m
20
4.8 m
4.0 m
6.0 m
6.0 m
5.0 m
2.5 m
2.0 m
0.5 m
2.0 m
0
10
20
30
40
50
1.0,2.5 m
1.5,2.0 m
0.5,3.0 m
0.5 m
4.8 m
0.5 m
6.0 m
1.0 m
m)
(<2
4.0 m
1.0 m
ble
Sta
Un
5
(4 –
6m
Tr
)
ans
itio
nal
(2 –
4m
)
m)
Ca
ble
ved
(> 6
10
0.5 m
1.0 m
1.0 m
15
sta
Stope span (m)
35
0
4.5 m
2.6 m
60
Stope, depth of failure
1.0 m
70
RMR89
FIGURE 5.6
Depths of failure for a number of stope spans and varying RMR89 values.
80
90
100
Span and Pillar Design
197
The data shown in Figure 5.6 are for stope designs in very hard, silicified rock masses extracted by conventional sublevel open stoping. The stope
data shown earlier relate to medium confining stress in mining epithermal
orebodies having depths of less than 500 m. A limitation is that induced
stresses cannot readily be considered when calculating the RMR89 values.
Hence, a designer trying to implement a similar strategy would need to
ensure that stress-driven instability is not a prominent failure mode prior to
implementing an approach similar to the one described here.
5.2.2 Span Determination Using the Tunnel Quality Index (Q) System
Barton et al. (1974) described the application of the Q system for rock mass
classification for the determination of no-support limits for various types of
excavations. Some 200 original case studies were used in the original calibration of the method. Over the next 18 years, more than 2000 new empirical
tunnel and large cavern designs were successfully carried out (Barton et al.,
1992). Figure 5.7 shows the updated plot for ground support recommendations. The solid lines bound the limits of practical support application, with
the lower line demarcating the stability limit for unsupported excavations of
a given equivalent span, ES = Span/ESR, where values for excavation support
ratio (ESR) are given in Table 5.1.
The ESR is a factor used by Barton to allow for varying degrees of instability based on excavation service life and use. The actual span of the excavation is divided by the ESR value to obtain the equivalent span for use in
Figures 5.7 and 5.8. Hutchinson and Diederichs (1996) note that the number
of mining case histories leading to the recommendation of ESR = 3–5 for
temporary mine openings is extremely limited and therefore recommend
using a maximum ESR of 3 for these openings unless local experience justifies the use of higher values. Certain mining excavations are more critical
than others from both operational and safety points of view. Figure 5.8 (after
Hutchinson and Diederichs, 1996) provides guidelines for no-support limits in order of decreasing reliability, relating them to Barton’s original ESR
values. Figure 5.8 is plotted against actual excavation span. Nevertheless,
the direct use of Q for open span design is not well documented within the
mining industry.
5.3 Stability Graph Method
Sublevel open stoping has become one of the most common underground
mining methods in the world due largely to its safety and efficiency.
Dimensioning of sublevel intervals, strike spans, pillars, and their location is
0.001
1
2
5
10
20
50
100
ESR
0.004
0.01
0.04
1.5 m
1.7 m
0.4
1
1.3 m
5c
m
Good
B
2.3 m 2.5 m
ng
2.0 m
10
Jw
SRF
i
1.6 m
pac
lt s
Bo
4
RQD Jr
Rock mass quality Q =
Jn
Ja
0.1
m
1m
9c
e
te
et
e
re
et
cr
tc
cr
ot lts
o
t
h
h
t
o
s
s bo
ls
s
sh olt
d bo
ed d
ce nd
ed d b
rc an
r
c
o
)
fo a
or an
nf
in )
ei cm
nf )
re cm
-r 5
ei m
r- –9
er 2–1
- r 12 c
e
r
b
b
Fi (1
be –
Fi (5
Fi (9
m)
5 c lts
tc
>1 d bo
(
Cas
e
ret e an
otc et
sh otcr
d
h
rce of s
nfo s
cm
cm
cm
rei d rib
12
25
15
r
e
e rc
b
i
F nfo
i
Re
ing
e lin
ret
onc
1m
a
re
eted a
shotcr
ing in
c
a
p
s
Bolt
1.3 m
1.2 m
Unsupportable
Fair
Poor
2.1 m
C
Rock classes
D
i
sh
n
nu
rea
da
ete
40
100
1.5
2.4
3
5
7
11
20
1000
Exc.
good
400
No support required
r
otc
3.0 m
A
Ext.
good
Spot bolting
4.0 m
Very
good
Bolt length in metres for ESR=1
FIGURE 5.7
Updated ground support recommendations. (After Grimstad, E. and Barton, N, Updating the Q-system for NMT, in R. Kompen, O.A. Opsahl, and
K.R. Berg, eds., Proceedings of the International Symposium on Sprayed Concrete—Modern Use of Wet Mix Sprayed Concrete for Underground Support, Fagernes,
Norway, October 17–21, 1993, pp. 46–66, Norwegian Concrete Association, Oslo, Norway.)
Excavation span (m)
E
Very
poor
ste
F
Extremely
poor
ing
G
Exceptionably
poor
nd
(a
)
cm
4
un Sy
re st
in em
4 c forc ati
ed c b
m
sh olt
o
tc ing
re
te
,>
olt
cb
ma
ti
Sy
198
Geotechnical Design for Sublevel Open Stoping
199
Span and Pillar Design
TABLE 5.1
Excavation Support Ratio
Type of Excavation
Number of
Cases
ESR
(Approx.)
2
83
3–5
1.6
25
1.3
79
1
Temporary mine openings
Permanent mine openings: low-pressure water tunnels;
pilot tunnels; drifts and headings for large openings
Storage caverns; water treatment plants; minor road and
railway tunnels; surge chambers; access tunnels, etc.
Power stations; major road and railway tunnels; civil
defense chambers; portals; intersections
Underground nuclear power stations; railway stations;
sports and public facilities; factories
2
0.8
Source: Barton, N.R., Rock mass classification and tunnel reinforcement selection
using the Q-system, in L. Kirkaldie (ed.), Rock Classification Systems for
Engineering Purposes: ASTM Special Technical Publication 984, ASTM
International, Philadelphia, PA, 1988, pp. 59–88.
Poor
Fair
Very
good
Good
Extremely
good
Maximum unsupported span (m)
200
100
50
try
n
ope
alls
—w
es
stop
ited
rifts
d
cess
ac
10
5
2
1
4
10
im
s—l
40
100
1.6
1.3
1.0
gs
enin
p
nt o
ane gs
ack
m
b
r
n
e
e
i
op
nd pal open s
n st
ls a
Ope
nne critic station
u
t
e
e
nd
ulag ns a efug
e ha statio and r
n
i
M sher ions
t
d
Cru aft sta
uire
req
Sh
t
r
po
sup
No
en
Non
20
1
3
e
ilur
fa
iate
ed
Imm
Exceptionally
good ESR
5
400
1000
Rock tunnelling quality index (Q)
FIGURE 5.8
Q system; no-support span limits for underground mine openings. (After Hutchinson,
D.J. and Diederichs, M.S., Cablebolting in Underground Mines, Bitech Publishers, Richmond,
British Columbia, Canada, 1996, 406pp. With permission.)
200
Geotechnical Design for Sublevel Open Stoping
very important to the success of the method. An empirical method of evaluating strike length span stability was developed in Canada by Mathews
et al. (1980). The method was further developed and applied by Potvin
(1988), Bawden et al. (1988, 1989), Nickson (1992), and Mawdesley et al. (2001),
among others. The original intent was to provide a practical design tool for
Canadian mine operators. The following five objectives were set for model
development (Bawden, 1993):
1. The model should be capable of predicting the overall stability of a
stope in terms of operating problems. Instead of focusing on precise
calculations and the identification of every single potential block
fall, the model should concentrate on defining conservative stope
dimensions, less conservative stope dimensions, and critical stope
dimensions above which open stoping becomes impractical.
2. The model must be reliable and hence sensitive to all key geotechnical parameters affecting underground stope design. It is
also important that the different conditions associated with open
stope mining such as stope geometry, mining sequence, blasting, and support from fill and cablebolts are directly or indirectly
accounted for.
3. The model must be easy to use by mining or geological engineers
on site. The input parameters should rely mainly on observational
methods rather than expensive testing, lengthy studies, and sophisticated equipment.
4. The model should be usable at any stage of mining (i.e., at the feasibility study and for short-term and long-term planning). Although
the precision of any model is largely a function of the quality of
the input parameters, which are better understood as mining progresses, the model should be capable of providing at least approximate answers at the feasibility study stage.
5. The model should be representative of rock mass behavior and be
capable of identifying underground modes of failure. This will provide a better understanding of the ground conditions and help in
selecting proper remedial solutions to ground control problems.
5.3.1 Updated Determination of the Stability Graph Parameters
The stability graph method is effectively a modification of the Q (1974) rock
mass classification method. The method relies on relating a stability number
(N′) to a stope wall hydraulic radius by way of a number of curves, each
depicting various levels of stability. For each stope wall, a stability number
is defined as follows:
N¢= Q¢ABC
(5.1)
201
Span and Pillar Design
where
A is a stress factor
B is a rock defect orientation factor
C is a design surface orientation factor (Potvin, 1988)
Q′ is defined following Barton et al. (1974) as
Q¢=
RQD Jr
Jn Ja
(5.2)
where RQD, Jn, Jr, and Ja are defined as per Table 4.4 and the different classification guidelines described in Section 4.3.3. Furthermore, Figures 4.40
through 4.43 illustrate the typical variability in the individual parameters
required to determine Q′. Following a similar procedure to that described
in Section 4.5.5, the value and variability of Q′ can be evaluated as shown
in Figure 5.9.
The parameters A, B, and C are defined individually as in the following
subsections.
5.3.1.1 Factor A
The rock stress factor A was initially designed to replace the stress reduction factor (SRF) in the original Q (Barton et al., 1974) system (Mathews et al.,
1980). Similarly to the SRF, it was defined as the ratio of uniaxial compressive
strength (UCS) of intact rock to the induced compressive stress parallel to
20
18
2200 RL
16
14
12
10
2000 RL
8
0
Q
1800 RL
FIGURE 5.9
Contoured grid model of hangingwall stability number Q′.
9000 E
2
8800 E
8400 E
4
8600 E
6
202
Geotechnical Design for Sublevel Open Stoping
20
σ1 = In situ main principal stress
σc = Uniaxial compressive strength
of intact rock
Stress reduction factor (SRF)
10
5
Near
surface
2.5–5.0
Heavy
rockburst
zone
2
1
High
stress
0.5
Low
confining
stress
Medium
confining
stress
0.2
0.1
1
2
5
10
σc
20
50
100
200
σ1
FIGURE 5.10
Stress reduction factor. (After Hutchinson, D.J. and Diederichs, M.S., Cablebolting in Underground
Mines, Bitech Publishers, Richmond, British Columbia, Canada, 1996, 406pp. With permission.)
the stope surface under consideration. However, the factor A as proposed
by Mathews et al. (1980) does not specifically account for the loss of confinement, as the SRF does (Figure 5.10). Stress relaxation may have a large effect
on jointed rock masses as it provides freedom of movement for individual
blocks. This is taken into account by the SRF within the low confining stress
zone. In addition, the original SRF factor accounts for improved stability
while mining under medium confining stress conditions. Experience has
shown that even a modest amount of confining pressure is likely to increase
the ultimate strength around stope walls.
Data from many years of numerical modeling and observations of open
stoping at Mount Isa Mines (Villaescusa, 1996), as well as the relative data
from the SRF, were used to review the stress factor A for input into stope
stability assessment. The results in Figure 5.11 show that the original factor A
(Potvin, 1988) is significantly more conservative than the SRF in the original
Q (1974) method.
Consideration of modern stope blasting practices and back analysis
of strength/stress ratio data from open stope walls at a large number of
Australian mines have been used to define a new A factor (Figure 5.12). As
with the original SRF, the benefits of medium confining stress are taken
into account, and it is suggested that no correction for compressive failure need be undertaken when the ratio of UCS/induced stress exceeds 5.5.
The resulting variability using the new factor A throughout four stoping
blocks and individual stope outlines is shown in Figure 5.13. The results
203
Span and Pillar Design
1.0
0.9
Stress factor A
0.8
Villaescusa (1996)
Potvin (1988)
Q-1974
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0
1
2
3
4
5
6
7
8
9
UCS/stress
10
11
12
13
14
15
FIGURE 5.11
A comparison of SRFs from a number of sources.
Rock stress factor A
1.0
0.9
Sto
0.6
0.5
σmax
0.0
Medium confining stress
High stress
0.4
0.3
0.2
0.1
all
pe w
0.8
0.7
Very high stress
0
1
2
3
4
5
6
7
8
9
10
σmax from 3D numerical modeling
σ
Ratio: σ c
max
FIGURE 5.12
Stress factor A and regions of stress considered.
are in accordance with observed conditions at the Kanowna Belle mine,
where very little stress-related failure was experienced within stoping
blocks A, B, and C.
5.3.1.2 Factor B
The rock defect orientation factor B is a weighting factor based on the orientation of the discontinuity set that is considered most likely to detract from
the stability of a particular stope surface (Potvin, 1988). The method requires
analysis of the discontinuity data to determine the most critical discontinuity
204
Geotechnical Design for Sublevel Open Stoping
1
Block A
.9
10000 N
.8
Block B
.7
.6
9800 N
Block C
.5
.4
.3
9600 N
20400 E
9400 N
Factor A
20200 E
0
19600 E
.1
20000 E
Block D
19800 E
.2
FIGURE 5.13
Contoured grid model using new factor A for stope hangingwalls, Kanowna Belle mine,
Western Australia.
likely to control stability. The determination of factor B requires the calculation of the true angle between a planar stope surface and the critical geological feature.
Considering that the most critical discontinuities are subparallel to a
stope surface, a few changes have been implemented to the original factor
B (Potvin, 1988). Based on many observations of actual stope wall failures,
it is suggested that no correction for discontinuity orientation should be
undertaken when the true angle with a stope surface exceeds 65° as shown
in Figure 5.14. In addition, a maximum penalty of 60% to the calculated Q′
is suggested for the effects of subparallel discontinuities. An example of the
variability of factor B throughout a number of stoping blocks is shown in
Figure 5.15.
The solid angle α (Figure 5.14) between the poles of a stope wall
(P) and a critical geological discontinuity (D) can be calculated from the
dot product
P ◊D = P D cos a
(5.3)
The vector P is the unit vector of the direction cosines of the normal to a
stope wall (P), which are defined by
205
N
Discontinuity pole
Stope wall
pole
α
0
10
20
30
40
50
60
70
80
ll
wa
W
pe
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
o
St
Joint orientation factor B
Span and Pillar Design
90
E
y
uit
tin
con
Dis
S
True angle between face and discontinuity
(angle α between poles)
FIGURE 5.14
Influence of joint orientation—factor B. (Modified after Potvin, Y., Empirical open stope design
in Canada, PhD thesis, University of British Columbia, Vancouver, British Columbia, Canada,
1988, 350pp.)
1
.9
10000 N
.8
Block B
.7
.6
9800 N
.55
Block C
.5
.45
9600 N
.4
20200 E
20000 E
9400 N
Factor B
19800 E
19600 E
0
20400 E
Block D
.3
FIGURE 5.15
Contoured grid model of new factor B for stope hangingwalls, Kanowna Belle mine.
206
Geotechnical Design for Sublevel Open Stoping
Px = cos jp sin bp
Py = cos jp cos bp
(5.4)
Pz = sin jp ,
where βp and φp are the trend and plunge of the normal to the stope wall
plane (see Figure 4.30). The vector D is the unit vector of the direction cosines
of the normal to a critical discontinuity, which are defined by
D x = cos jd sin bd
D y = cos jd cos bd
(5.5)
D z = sin jd ,
where βd and φd are the trend and plunge of the normal to a critical discontinuity, as defined in Figure 4.29.
The angle α is thus given by
cos d = Px D x + Py D y + Pz D z
(5.6)
5.3.1.3 Factor C
The design surface orientation factor C was proposed to account for the
influence of gravity on the stability of the stope surface (Potvin, 1988). The
factor is based on the assumption that under the effects of gravity, a vertical stope wall is more stable than a horizontal stope back. Surfaces where
sliding blocks can form or where significant overhangs occur (i.e., stope
backs and hangingwalls) will have the most detrimental influence on stability. Two adjustment factors were proposed by Potvin (1988) and have been
modified here to account for the back analysis of stope stability at a number
of Australian mines. The effects of gravity fall and slabbing are considered
in Figure 5.16. The adjustment factor has been made constant for flat stope
backs having a dip of less than 20° (Bieniawski, 1989).
The second adjustment factor proposed by Potvin (1988) to analyze sliding
modes of failure of stope walls is shown in Figure 5.17. Assuming that the
frictional resistance of a critical discontinuity exceeds the driving force, the
amount of adjustment has a maximum value of 8 when the dip of a critical
discontinuity is less than 30°. It is proposed here that as the dip of a critical discontinuity increases, the adjustment will decrease to a minimum value of 4.
According to Potvin (1988), the potential mode of failure can be determined with a simple diagram in which the excavation and the critical joint
are sketched. If a gravity vector represented by a vertical arrow drawn from
207
Span and Pillar Design
pe
Dip
n sto
10
9
8
7
6
5
4
3
2
1
0
Slabbing
Ope
Gravity adjustment factor C
Gravity fall
0
10
20
30
40
50
60
70
80
90
Dip of stope wall (degrees)
FIGURE 5.16
Determination of gravity effects—factor C. (Modified after Potvin, Y., Empirical open stope
design in Canada, PhD thesis, University of British Columbia, Vancouver, British Columbia,
Canada, 1988, 350pp.)
the approximate center of gravity of the block formed by the critical discontinuity falls directly inside the opening, the mode of failure will be gravity fall. In addition, if the gravity vector stays inside the medium without
intersecting the critical discontinuity, slabbing or buckling failure can occur.
Furthermore, when the gravity vector crosses the critical joint, the potential for sliding failure exists (Potvin, 1988). An example of the variability of
factor C throughout a number of stoping blocks is shown in Figure 5.18.
5.3.1.4 Hydraulic Radius
The hydraulic radius concept to account for the size and shape of a stope plane
under analysis was introduced by Laubscher and Taylor (1976). Hydraulic
radius is the quotient of the stope wall area and the stope wall perimeter,
and favors long and narrow shapes over square shapes (see Figure 1.5).
208
Geotechnical Design for Sublevel Open Stoping
Ope
n st
ope
Discontinuity
dip
Gravity adjustment factor C
Sliding
10
9
8
7
6
5
4
3
2
1
0
0
10
20
30
40
50
60
70
80
90
Dip of critical discontinuity (degrees)
FIGURE 5.17
Determination of sliding effect on critical joint—factor C. (Modified after Potvin, Y., Empirical
open stope design in Canada, PhD thesis, University of British Columbia, Vancouver, British
Columbia, Canada, 1988, 350pp.)
Hydraulic radius is easy to assess as most stope shapes are not very complex.
The methodology allows the analysis of stope surfaces wall by wall. The
relationship between hydraulic radius (i.e., area/perimeter) and excavation
length, given a fixed height, usually defined by the sublevel interval, is
given by
(H)(L)
2(H + L)
(5.7)
2(H)(HR )
H - 2(HR )
(5.8)
HR =
and
L=
209
Span and Pillar Design
10
Block A
9
10000 N
8
Block B
7
6
9800 N
Block C
5
4
3
9600 N
2
Block D
20400 E
20200 E
20000 E
9400 N
Factor C
19600 E
0
19800 E
1
FIGURE 5.18
Contoured grid model of new factor C for stope hangingwalls, Kanowna Belle mine.
where HR is the hydraulic radius and H and L are the height and length of
the stope wall, respectively.
In order to determine the maximum allowable unsupported lengths, the
height or width of the excavations needs to be first determined. For vertical
walls, this generally relates to floor-to-floor dimensions for the stope surface
under consideration. Consider, for example, Figure 5.19a, which shows that
for a footwall, the stope down-dip span is “fixed,” as it is determined by the
sublevel interval chosen. For the stope backs and end walls, the width is
generally controlled by the ore or stope width (as for narrow vein, generally,
stopes are purposely not mined wider than the ore width). For a hangingwall, because of the cablebolting reinforcement at every sublevel interval, the
“fixed” dimension is the down-dip span between the cablebolts (for steeply
dipping orebodies, this is approximately equal to the level interval spacing).
Figure 5.19b shows that the rock mass exposed between the cablebolts must
be inherently stable, as the cablebolts only minimize the deformation locally
near the stope drives. The cablebolts are also very effective in arresting failures up-dip (see Figure 1.13).
5.3.2 Prediction of Stope Stability
The calculation of the stability number (Equation 5.1) for a particular stope
wall is achieved by multiplying the variables accounting for the geotechnical
210
Geotechnical Design for Sublevel Open Stoping
HR =
HR back =
Area
Perimeter
L*W
L*H
HR hangingwall =
2 * (L + W)
2 * (L + H)
Hangin
gw
Down-d all
ip
span (H
)
Foo
Down-d twall
ip span
(H)
Maximum
2 * HR * H
=
allowable
Back
(H – 2 * HR)
length (Lmax)
) Stope width (W)
x
a
ll
wa gth (L m
ing
ng le len
a
H wab
llo
xa
a
M
x
Ma
(a)
)
ax
all th (L m
w
g
t
n
o
Fo ble le
wa
allo
(b)
FIGURE 5.19
(a) Fixed and allowable stope dimensions and (b) hangingwall failure. (b: Courtesy of Mount
Isa Mines, Mount Isa, Queensland, Australia.)
parameters previously described. The initial back analysis work in Canada
included a total of 175 case studies of unsupported open stope walls from
23 Canadian mines (Potvin, 1988). The initial stability graph shown in
Figure 5.20 is composed of stable and caving zones, separated by a transition zone. The stope walls were divided by Potvin (1988) into three groups.
Stable walls that experienced low dilution were represented by roundshaped points. Stope walls that experienced dilution and rock falls causing
operational problems were classified as unstable. They are shown on the
graph as square-shaped points. The triangular points represent stope walls
that experienced severe instability.
The solid black line shown in Figure 5.20 was calculated by Nickson (1992)
to statistically account for the difference between stable and caved points.
The relationship between the stability number N′ and the maximum allowable unsupported hydraulic radius (HRallowed) is given as a function of the
stability number by
HR allowed = 10[0.573 +0.338 log N¢]
(5.9)
Nickson’s boundary allows for larger stope dimensions than those predicted
by Potvin’s unsupported transitional zone. The statistical boundary developed by Nickson can be used to predict the maximum allowable stable open
211
Span and Pillar Design
1000
Stable zone
Stability number (N')
100
HR== 10
HR
10
(0.573 + 0.338 log N')
Caved zone
1.0
0.1
0
5
10
15
20
25
Hydraulic radius (m)
FIGURE 5.20
Initial stability graph calculated from 175 case histories of unsupported open stope walls.
(After Potvin, Y., Empirical open stope design in Canada, PhD thesis, University of British
Columbia, Vancouver, British Columbia, Canada, 1988, 350pp.)
stope surface relating to that particular stability number. For example, for a
stability number of 11, a corresponding hydraulic radius of 10 is allowed and
is recommended as a first estimate for stope span design.
Nickson (1992) also increased the initial database for the stability graph
method and eventually updated the stability graph to the form shown in
Figure 5.21. This figure can be used to evaluate maximum allowable stope
wall sizes for either unsupported or pattern (full coverage) cablebolted
stope walls. However, Nickson (1992) clearly stated that the graph cannot be used to design cablebolted hangingwall spans where the cables are
installed from localized drill drive locations (point anchored hangingwall
cablebolting or rib rock; see Rauert, 1995). Stability evaluations of cablebolted stope hangingwalls must ensure that any unsupported rock mass
exposed down-dip between finite cablebolting locations is inherently stable, as per Equation 5.9.
Figure 5.22 shows stope stability data for unsupported and completely
stable (zero depth of failure) open stope walls at the Cannington mine,
Queensland (Coles, 2007). The figure compares the calculations from Potvin
(1988) with the prediction using the updated parameters presented in this
chapter. The figure also shows the original relationship developed by
Nickson (1992).
212
Geotechnical Design for Sublevel Open Stoping
1000
500
200
HRallowed = 10[0.573 + 0.338 log N']
Stable zone
50
20
U
10
5
2
1
0.5
e
zon
Caving zone
Re
0.2
0.1
r
dt
rte
po
p
nsu
on
iti
ans
St
ca able
ble w
bo ith
inf
lt
orc
re full c
ed
in
tra
fo ove
rc ra
nsi
em ge
tio
en
nz
t
on
e
Stability number (N')
100
0
5
Full coverage cablebolting
10
Hydraulic radius (m)
15
20
Localized cablebolting
d
Unsu
pporte
Unsu
pporte
d
g
ltin
ebo
l
cab
Ponit an
ch
cablebo ored (Rib-roc)
lt reinfo
rcemen
t
age
ver
l co
Ful
Extended chart is only applicable for full coverage cablebolt reinforcement geometries
Limit for unsupported design given by HRallowed = 10[0.573 + 0.338 log N']
FIGURE 5.21
Stability graph showing zones of stable ground, caving ground, and ground requiring cablebolt reinforcement. (After Nickson, S.D., Cablebolt support guidelines for underground hard
rock mine operations. MASc thesis (unpublished), University of British Columbia, Vancouver,
British Columbia, Canada, 1992.)
213
Span and Pillar Design
1000
Potvin (1988)
Stable zone
Modified after Potvin (1988)
Stability number (N')
100
HR
10
= 10
.338
3+0
0.57
´
log N
1
Caved zone
0.1
0
5
10
15
20
Hydraulic radius (m)
FIGURE 5.22
Stability graph for unsupported, completely stable (zero depth of failure) open stope walls,
Cannington mine, Queensland. (Data from Coles, D., Performance of open stopes at BHPBilliton Cannington mine, BEng thesis, Western Australian School on Mines, Curtin University
of Technology, Kalgoorlie, Western Australia, Australia, 2007, 161pp.)
Regardless of the empirical methodology chosen, the final design of an
open stope must always consider the geotechnical issues described earlier
together with economic, scheduling, and mining constraints. Consequently,
engineering judgment is always required to establish the most efficient stope
wall design.
5.3.3 Use of the Stability Graph as a Design Tool
Relational geotechnical databases that include information on UCS, rock
mass classification data such as RQD, joint set number, joint orientation and
condition, and stope wall performance such as depth of failure can be used to
calibrate the stability graph for existing stoping blocks. The sample locations
(X,Y,Z) for each data point in the database can be plotted in three dimensions to obtain a visual appreciation of the spatial distribution and density
of the database with respect to the orebody and its immediate boundaries. It
is important to note that, although the total number of samples in the database is always significant, it is critical to ensure that the relevant samples
are actually located within the immediate hangingwall and footwall or the
orebody in question (see Figure 4.7). Figure 5.23 shows the modeled spatial
variability of the Q′ parameter at the Kanowna Belle Orebody, Kalgoorlie,
Western Australia. The model predicts a reduction on rock mass quality at
214
Geotechnical Design for Sublevel Open Stoping
20
18
Block A
10000 N
16
14
Block B
12
9800 N
10
Block C
8
9600 N
6
4
20200 E
20000 E
Q
19800 E
9400 N
19600 E
0
20400 E
Block D
2
FIGURE 5.23
Contoured grid model of Q′ for stope hangingwalls, Kanowna Belle mine.
depth for the Kanowna Belle hangingwall boundary and the stope design
must account for such variation in space.
In addition to rock mass classification data, the anticipated maximum and
minimum induced principal stresses tangential to the stope walls are also
required to more accurately determine the stress conditions required to calculate factor A in the stability graph method. The induced stresses can be estimated using three-dimensional numerical modeling. For each mining step
within the numerical model, the major and minor induced principal stresses
across each mining surface are located and recorded, along with the threedimensional coordinates of these points (Figure 5.24). It must be remembered
that the induced stresses depend on the stoping extraction sequence.
An example of a longitudinal view of the induced major principal stresses
on a hangingwall plane for four stoping blocks is shown in Figure 5.25.
A significant increase in induced stress with depth can be seen. Very high
stresses are expected in stoping block D where induced stresses up to three
times higher than those experienced in block A are predicted.
The stability number (N′) must be calculated independently for each stope
wall. Instability will occur in surfaces where sliding blocks can form, or
where significant overhangs occur. Flat joints are likely to have a significant
effect on stope backs (crowns) and the stability within vertical walls will be
X
X
0.0
0.0
34 MPa
2
3
1
20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 180.0 200.0
65 MPa
20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 180.0 200.0
3
2
1
CT98-62
CT02-62
CP03-62
DR03-59
Induced (σ1) model
CS00-62
CP80-62
• Major induced stress (σ1) at time of stope
extraction utilized (i.e., sequence dependent)
σ1 (MPa)
0
12
24
36
48
60
72
84
96
108
120
FIGURE 5.24
Estimating the induced major principal stress using the computer program MAP3D.
Y
σ1
(MPa)
Y
σ1
(MPa)
CT05-62
CP07-
Span and Pillar Design
215
216
Geotechnical Design for Sublevel Open Stoping
120
108
Block A
10000 N
96
84
72
Block B
9800 N
60
Block C
48
36
9600 N
24
20200 E
20000 E
19800 E
9400 N
19600 E
0
Signal (MPa)
20400 E
Block D
12
FIGURE 5.25
Contoured grid model of the induced major principal stress (σ1) for stope hangingwalls,
Kanowna Belle mine.
controlled by the presence of subvertical to moderately dipping geological
discontinuities having strikes oriented subparallel to a stope surface. The
mode of failure, however, is dependent on the dip direction of the critical
joint with respect to the particular stope wall. Figure 5.26 shows back analysis data from open stoping at the Olympic Dam mine, South Australia, showing different degrees of instability for the different stope walls forming the
stope shapes.
The stability number for a particular stope surface can be calculated for the
grid models by multiplying the component terms from each of the grid models to evaluate Equation 5.1. An example of the resulting stability number
is presented in Figure 5.27. The allowable hydraulic radius (HRallowed) for a
given N′ value is given by Equation 5.9 (Nickson, 1992). The HRallowed results
for the grid model are shown in Figure 5.28.
In order to determine the maximum allowable unsupported lengths
(Lmax), the height of the designed stopes needs to be first established.
A decision must be taken to determine if cablebolt reinforcement effectively reduces the down-dip span as shown in Figure 5.19a. An example
of a contoured grid model of interlevel down-dip height for the stope
hangingwall surfaces at the Kanowna Belle mine is given in Figure 5.29.
217
Span and Pillar Design
100
Stable
90
Unstable
Failed
Occurrence (%)
80
70
60
50
40
30
20
10
0
Crown
N
E
S
W
Stope stability by wall orientation
FIGURE 5.26
Stope stability by wall orientation at the Olympic Dam mine, South Australia. (From
Oddie, M.E. and Pascoe, M.J., Stope performance at Olympic Dam mine, Proceedings of the
9th Underground Operators’ Conference, Perth, Western Australia, Australia, March 7–9, 2005,
pp. 265–272, AusIMM, Melbourne, Victoria, Australia. With permission.)
20
Block A
18
10000 N
16
Block B
14
12
9800 N
10
Block C
8
9600 N
6
4
20200 E
20000 E
N
19800 E
9400 N
19600 E
0
20400 E
Block D
2
FIGURE 5.27
Contoured grid model of stability number N′ for stope hangingwalls, Kanowna Belle mine.
218
Geotechnical Design for Sublevel Open Stoping
15
13.5
Block A
10000 N
12
Block B
10.5
9
9800 N
7.5
Block C
6
4.5
9600 N
3
20200 E
20000 E
19800 E
9400 N
19600 E
0
Hydraulic radius
20400 E
Block D
1.5
FIGURE 5.28
Contoured grid model of stable unsupported hydraulic radius (HR allowed) for stope hangingwalls, Kanowna Belle mine.
40
36
Block A
10000 N
32
Block B
28
24
9800 N
Block C
20
16
12
9600 N
8
Block D
20400 E
20200 E
20000 E
9400 N
19600 E
0
Down dip height
19800 E
4
FIGURE 5.29
Contoured grid model of interlevel down-dip height for the hangingwall surfaces, Kanowna
Belle mine.
Span and Pillar Design
219
Given a fixed sublevel interval, the stable hydraulic radius contour plots
(HRallowed) enable a determination of the maximum allowable stable lengths
(Equation 5.8) as shown in Figure 5.30. The plot also shows actual (mined
or designed) dimensions for comparison. The close agreement for block A
suggests that the modifications to the factors A, B, and C presented here
are well established.
5.3.4 Design Validation
The stability graph method was originally developed as an initial assessment of stability at the prefeasibility stages of projects. Currently, the method
is being used worldwide as a design tool in all stages of stope dimensioning
and has become an established empirical tool for dimensioning open stope
walls. However, the system has a number of limitations that must be understood in order to assess its applicability in any particular geotechnical environment. Over the years, the applicability and limitations of the method for
open stope design has been reviewed by several authors (Pakalnis et al., 1995;
Stewart and Forsyth, 1995; Suorineni et al., 2001; Suorineni, 2012). In particular, the following observations are considered to be important:
1. The definitions of stable versus caving conditions are subjective since
the depth of failure is not reported. In addition, the method does not
incorporate complex failure mechanisms involving more than one
family of geological discontinuities. Specifically, the method does
not consider buckling in which the frequency of subparallel discontinuities may be critical.
2. Despite the use of quantifiable input values, the precise degree of
inherent conservatism is not known.
3. The method reflects mining practice, which may have been influenced by factors such as legislation, local practices, and particular
geological peculiarities. The method lacks sufficient precision for
stope dimensioning (excessive scatter).
The following factors are likely to have affected the stable/unstable boundaries identified during method development in Canada and may not necessarily be the same elsewhere:
•
•
•
•
•
•
Stoping style methodology
Volume of overbreak or dilution levels
Blasting practices
Stress regime (including destressing or tensile failures)
Determination of induced stress in complex stoping geometries
Mine-wide determination of intact rock parameters such as UCS
9400 N
9600 N
9800 N
10000 N
(a)
Block D
Block C
Block B
Block A
(b)
20200 E
20000 E
19800 E
20200 E
20000 E
19800 E
Block D
Block C
Block B
Block A
FIGURE 5.30
Contoured grid model of (a) maximum allowable unsupported length and (b) mined and designed strike length for the hangingwall, Kanowna
Belle mine.
Strike length
–0
–4
–8
–12
–16
–20
–24
–28
–32
19600
–36
20400 E
–40
20400 E
220
Geotechnical Design for Sublevel Open Stoping
221
Span and Pillar Design
• Quality control on reinforcement installation
• Type of reinforcement, use of plates, etc.
• Quality of rock mass characterization, detailed mapping including
biases
Therefore, the stability graph method may not necessarily constitute an optimum design methodology but, rather, a starting point for each particular
geotechnical environment. Empirical evidence and ongoing documentation
are therefore critical to the implementation of optimized stoping geometries
at any particular mine site. Consequently, design validation represents a critical component in the application of the stability graph. Validation is accomplished through the use of various instrumentation strategies ranging from
simple underground observations at the most basic, to minewide microseismic systems at the most complex. Geotechnical instrumentation is of critical
importance to the mine design approach discussed herein. Other than for
local safety considerations, instrumentation should be placed to help calibrate design models. It is essential that all instrumentation be very carefully
designed and located to ensure maximum benefit and interpretability.
In order to emphasize this applicability and validation point, Figure 5.31
from a published back analysis of open stopes at the Olympic Dam mine
1000
Stable
Unstable
Stable region
Failed
Stability number (N')
100
n
Tra
ion
l reg
na
sitio
10
Unstable region
1.0
0.1
0
5
10
15
20
Hydraulic radius (m)
FIGURE 5.31
Stability graph calculations for unsupported stope walls at the Olympic Dam mine. (From
Oddie, M.E. and Pascoe, M.J., Stope performance at Olympic Dam mine, Proceedings of the
9th Underground Operators’ Conference, Perth, Western Australia, Australia, March 7–9, 2005,
pp. 265–272, AusIMM, Melbourne, Victoria, Australia. With permission.)
222
Geotechnical Design for Sublevel Open Stoping
(Oddie and Pascoe, 2005) is presented. The resulting data show little or no
correlation with the stability graph, suggesting that a local parameter, perhaps not well accounted for by the stability graph methodology, controls the
stability of the Olympic Dam mine open stope walls.
5.4 Numerical Modeling of Stope Wall Stability
The main objective of numerical modeling is to quantify the effects of
induced stress on stope performance. This is achieved by relating different levels of induced stress to different levels of rock mass damage around
a stoping void. The underlying assumption is that stress-induced failure
occurs from induced stresses exceeding the local rock mass strength, thus
resulting in stope wall overbreak. Unfortunately, this assumption could lead
to variability in back analysis of open stope performance results because
the resulting stope void geometry may not necessarily define the excavation damage zone or yield zone of the rock mass around a stope (Cepuritis
et al., 2007, Figure 5.32). Material around a stope void could actually represent
“Unyielded”
“unremoved”
rock mass outside
planned void
?
Amount of “yielded”
rock mass unknown
with CMS data
“Yielded”
“unremoved”
rock mass inside
planned void
Unyielded/yielded boundary (EDZ)
Unremoved/removed boundary
Inside/outside planned void boundary
“Yielded” “removed”
rock mass outside
planned void
“Unyielded” “unremoved”
rock mass inside
planned void
“Yielded”
“removed”
rock mass inside
planned void
“Yielded”
“unremoved”
rock mass outside
planned void
FIGURE 5.32
Schematic showing resulting stope void with respect to possible yielded rock mass conditions and planned void geometry. (From Cepuritis, PM. et al., Back analysis and performance
of block A long hole open stopes—Kanowna Belle Gold mine, in E. Eberhardt, D. Stead, and
T. Morrison, eds., Rock Mechanics: Meeting Society’s Challenges & Demands, Proceedings of the First
Canada—US Rock Mechanics Symposium, Vancouver, British Columbia, Canada, May 27–31,
2007, pp. 1431–1439, Taylor & Francis, Leiden, the Netherlands.)
Span and Pillar Design
223
“yielded” yet “unremoved” rock mass, where the local shape and span could
have arched holding up yielded material. In addition, “yielding” of a rock
mass cannot always be solely attributed to stress-induced rock mass damage,
as other influences such as poor drill and blast practices may also contribute.
Nevertheless, numerical modeling techniques can be used to identify and
quantify the relative contributions of the various factors that influence stope
performance, including stope geometry, development location and undercutting, rock mass characteristics, in situ and induced stresses, and the influence
of large-scale geological structures. For open stoping, the choice of modeling
technique includes linear elastic numerical modeling, such as the program
Map3D (Mine Modelling, 2013), and nonlinear continuum or discontinuum
finite element analysis, such as Abaqus (Beck and Duplancic, 2005). In particular, Abaqus is used specifically for the analysis of stoping problems where
there is potential for significant plasticity and high levels of deformation
with large-scale structures explicitly incorporated in the model.
5.4.1 Linear Elastic Numerical Modeling
Wiles (2001) suggested that rock mass damage can be related to the relative level of linear elastic overstressing (Figure 5.33a). The critical stress
levels are dependent on mine site-specific parameters and can be correlated
using the observed rock mass response and the results from numerical
modeling. The assumption is that below a site-specific damage threshold,
the rock mass response is elastic and consequently very little damage is
observable. As the level of overstressing increases, the observed damage
(i.e., irrecoverable strain) should increase, leading to a zone of potential
overbreak around the excavation. Increased overstressing beyond this
level may cause stress-driven failures and eventually the rock mass may
become unsupportable. Wiles (2001) proposed that this methodology could
be incorporated into a comprehensive back analysis technique to assist in
quantitative stope design (Figure 5.33b). Furthermore, the damage model
assumes that the level of overstressing is a direct cause of an increase in σ1,
while confinement is kept constant. In practice, the stress path experienced
by a rock mass can vary (Figure 5.33c) with “excess stress” generated by
any of the following:
• A loss of confinement, for example, a stope wall or back (−∆σ3)
• An increase in load, for example, a pillar or stope wall (+∆σ1)
• A combination of both, typical of a stope block abutment failure
(+∆τmax)
Back analysis of stress-driven open stope damage is best undertaken for primary stopes, where a condition of minimal stress-induced damage prior to
stoping can be assumed. Thus, the stress path in the immediate vicinity of
224
Geotechnical Design for Sublevel Open Stoping
Unsupportable
σ-driven failure
POB
σ1
Damage
(a)
Unsupportable
σ1
POB
σ-driven
failure
ε1
σ3-Confinement
Collapse
σ1
Increasing
damage
(b)
Unstable
Stable
σ1
∆σ3
∆σ1
∆τmax
Undamaged
σ3
(c)
σ3
FIGURE 5.33
(a) Linear elastic stress damage model for monotonically increasing stresses, together
with related strain damage. (After Wiles, T.D., Map3D course notes. Masters of Mining
Geomechanics, Western Australian School of Mines, Mine Modelling Pty Ltd., Mount Eliza,
Victoria, Australia, 2001, 124pp.) (b) Generalized damage model. (After Wiles, T.D., Map3D
course notes. Masters of Mining Geomechanics, Western Australian School of Mines, Mine
Modelling Pvt Ltd, Leinster, Western Australia, Australia, 2001, 124pp.) (c) Stress path overstressing. (From Cepuritis, P.M. et al., Back analysis and performance of block A long hole
open stopes—Kanowna Belle Gold mine, in E. Eberhardt, D. Stead, and T. Morrison, eds.,
Rock Mechanics: Meeting Society’s Challenges & Demands, Proceedings of the 1st Canada—US Rock
Mechanics Symposium, Vancouver, British Columbia, Canada, May 27–31, 2007, pp. 1431–1439,
Taylor & Francis, Leiden, the Netherlands.)
the stopes may be attributed to the primary stope extraction sequence. The
number, location, and orientation of large-scale geological discontinuities
(Villaescusa and Cepuritis, 2005) must be also taken into account to facilitate
the interpretation of the numerical modeling results (Cepuritis et al., 2007).
Cepuritis et al. (2007) show example results of σ1 versus σ3 contoured
by the depth of stope wall overbreak or the calculated depth of failure.
The results were subdivided into regions based on the likely stress path
experienced (Figure 5.34). For moderately jointed to massive rock masses
(Figure 5.35), the onset of increased depth of failure shows good correlation
with an estimated Hoek–Brown strength envelope (Cepuritis et al., 2007).
225
Span and Pillar Design
σ1
Monotonic
Shear
–45°
–15°
–15°
Confined
∆σ1,∆σ3
Low
confinement
In situ stress
–90°
Unloading
–180°
σ3
FIGURE 5.34
A stress path classification used in back analysis of stope wall overbreak. (From Cepuritis,
P.M. et al., Back analysis and performance of block A long hole open stopes—Kanowna Belle
Gold mine, in E. Eberhardt, D. Stead, and T. Morrison, eds., Rock Mechanics: Meeting Society’s
Challenges & Demands, Proceedings of the 1st Canada—US Rock Mechanics Symposium, Vancouver,
British Columbia, Canada, May 27–31, 2007, pp. 1431–1439, Taylor & Francis, Leiden, the
Netherlands.)
More significantly, the depth of overbreak increases with overstressing, and
progressively increases as the stress path changes from monotonic loading
and shear, through to low confinement conditions. Increased falloff occurs
under unloading conditions, particularly close to the stope-scale rock mass
damage initiation criteria (Cepuritis et al., 2007). For highly fractured rock
masses influenced by large-scale geological discontinuities, the overbreak
generally occurs at lower stress levels, and the extent occurs over a wider
range of stress conditions (see Figure 5.36, Cepuritis et al., 2007).
5.4.2 Nonlinear Numerical Modeling
Nonlinear modeling of complex open stoping sequences can be undertaken
using a nonlinear, general purpose, three-dimensional finite element analysis program such as the Abaqus Explicit (Beck and Duplancic, 2005). Abaqus
is well suited to the analysis of mining problems where a potential exists
for significant plasticity, complex extraction sequences, high levels of deformation, and large numbers of material discontinuities. Models required to
represent global stoping sequences, large-scale geological discontinuities,
and stope-scale structures are routinely implemented (Beck and Duplancic,
2005). Large-scale global models are constructed incorporating all stoping
geometries including shafts, ramps, access development, and mine-scale geological discontinuities. Smaller, more detailed submodels are subsequently
226
Geotechnical Design for Sublevel Open Stoping
80
70
Estimated
rock mass
strength
Monotonic
loading
High
confinement
σ1 (MPa)
60
50
Shear
Rock mass
damage
initiation
Stope wall
Depth of failure (m)
40
30
Low
confinement
< 2.5 m
2.5–3.0 m
3.0 –3.5 m
20
3.5 – 4.0 m
4.0 –4.5 m
10
–20
4.5–5.0 m
Unloading
–10
0
10
20
30
σ3 (MPa)
FIGURE 5.35
Example of σ1 versus σ3 for moderately jointed to massive rock masses. (From Cepuritis,
P.M. et al., Back analysis and performance of block A long hole open stopes—Kanowna Belle
Gold mine, in E. Eberhardt, D. Stead, and T. Morrison, eds., Rock Mechanics: Meeting Society’s
Challenges & Demands, Proceedings of the 1st Canada—US Rock Mechanics Symposium, Vancouver,
British Columbia, Canada, May 27–31, 2007, pp. 1431–1439, Taylor & Francis, Leiden, the
Netherlands.)
constructed in key areas, with strain outputs and tractions from the global
models being used as the boundary conditions for the submodels. Modeling
is specifically targeted at understanding rock mass response and the influence
of stope-scale structures on stope wall performance (Cepuritis et al., 2010).
Extraction sequences in a global model are implemented in approximately
quarterly steps, while block-scale models are extracted in steps no larger than
one stope at a time. Selected stopes are extracted and then filled sequentially.
A large number of extraction steps are required to ensure that the stress path
throughout an entire area of interest is captured. For the submodels, each
stope can be extracted involving a number of intricate firings, usually consisting of (a) a full-height cutoff slot or approximately 10% of the final void,
227
Span and Pillar Design
Monotonic
loading
80
70
σ1 (MPa)
60
50
High
confinement
Subperpendicular
to wall surface
Sub-parallel to
wall surface
Shear
40
30
Low
confinement
σ1 – σ3 = 25 MPa
Stope wall
Depth of failure (m)
<2.5 m
20
2.5 – 3.0 m
3.0 – 3.5 m
10
–20
3.5 – 4.0 m
4.0 – 4.5 m
Unloading
–10
4.5 – 5.0 m
0
10
20
30
σ3 (MPa)
FIGURE 5.36
Example of σ1 versus σ3 for highly fractured rock masses. (From Cepuritis, P.M. et al., Back
analysis and performance of block A long hole open stopes—Kanowna Belle Gold mine, in
E. Eberhardt, D. Stead, and T. Morrison, eds., Rock Mechanics: Meeting Society’s Challenges &
Demands, Proceedings of the 1st Canada—US Rock Mechanics Symposium, Vancouver, British
Columbia, Canada, May 27–31, 2007, pp. 1431–1439, Taylor & Francis, Leiden, the Netherlands.)
(b) void creation of approximately 30%–40% of the final void, and (c) final
stope mass firing to create the final stope void.
The inclusion of detailed and extensive structural stope-scale discontinuity geometries is significant. It allows the model to be able to represent the
physics and interactions between stope-scale structures, excavations, and
the continuum rock mass components. It also allows for the efficient computation of displacements, damage, and deformation to the required level of
detail across large numbers of stopes, in a number of stoping blocks.
The model results are calculated using a grid of result points enabling
the calculation of various model parameters at each mining step at
228
Geotechnical Design for Sublevel Open Stoping
FIGURE 5.37
Arrangement and distribution of result points for a stoping block and a single stope. (From
Cepuritis, P.M. et al., Back analysis of over-break in a longhole open stope operation using nonlinear elasto-plastic numerical modelling, Proceedings of the 44th US Symposium Rock Mechanics
& 5th Canada—US Rock Mechanics Symposium, Salt Lake City, UT, June 27–30, 2010, Paper ARMA
10-124, 11pp.)
varying distances into a stope wall rock mass. The result points are generally located at approximately 1 m intervals into a stope wall, using an
approximate 5 m × 5 m pattern across a stope surface. Figure 5.37 shows a
general arrangement of result points described by Cepuritis et al. (2010). For
each result point and mining step, the output parameters are entered into
a purpose-built database. Additional information such as stope name, true
distance to a stope surface, distance to the nearest stope-scale structure, and
its final stability condition are also included. The final stability condition
is assigned by determining whether a point lies within the final surveyed
void and beyond the planned geometry (i.e., overbreak) and is therefore
assigned as unstable or whether it is located outside the surveyed volume,
within the stable rock mass.
Stope wall instability is generally defined by an unacceptable displacement of the rock mass into the stoping void. The criteria for instability are
generally defined by a certain critical limit of displacement or velocity, or in
the case of open stoping, a certain volume of rock mass. These criteria occur
within a certain time frame, typically prior to complete removal of ore and
stope filling. These criteria can be measured, albeit with various degrees of
accuracy and precision, using geotechnical instrumentation, such as extensometers, or laser cavity surveys (Miller et al., 1992).
Critical limits of stope instability can be accessed from nonlinear numerical modeling primarily using velocity and plastic strain values computed
229
Span and Pillar Design
0.10
0.10
Stope extraction
0.08
0.06
0.06
Plastic strain
Velocity
0.04
0.04
0.02
0.02
0.00
Velocity (m/step)
Plastic strain
0.08
20
40
60
80
100
120
140
150
0.00
180
Model step
FIGURE 5.38
Plastic strain and maximum velocity values versus mining step. (From Cepuritis, P.M. et al.,
Back analysis of over-break in a longhole open stope operation using non-linear elasto-plastic
numerical modelling, Proceedings of the 44th US Symposium Rock Mechanics and Fifth Canada—
US Rock Mechanics Symposium, Salt Lake City, UT, June 27–30, 2010, Paper ARMA 10-124, 11pp.)
during stope extraction. With regard to the modeling results, velocity
here refers to the magnitude of a computed resultant displacement vector
between mining steps, expressed as meter per step (i.e., m/step). An example of velocity and plastic strain output from Cepuritis et al. (2010) is shown
in Figure 5.38. Velocity can be considered as an upper-bound criterion for
instability, as all points with high velocity should theoretically be considered unstable. Hence, rock without damage that has a high velocity must
be unstable (e.g., a moving rock mass bounded by structure). Plastic strain
or damage can be considered a lower-bound criterion for stope wall instability, as material may be damaged, but may still be stable if the velocity
is low. An unstable point in the rock mass can therefore have a number of
combinations of velocity and plastic strain. In terms of prediction of rock
mass failure using these two variables, they are not mutually exclusive. In
addition, plotting plastic strain versus velocity indicates that these variables
are independent, with the covariance and correlation coefficient effectively
zero (Cepuritis et al., 2010).
The maximum levels of plastic strain and velocity during stope extraction can be compared to the frequency with which they correspond to stable
and unstable points within a stoping boundary. The percentage of unstable
points for a selected interval range can be considered an empirical “probability of instability” as it is calibrated on the actual mining geometry, sequence,
and performance. An example relationship between maximum velocity
during stope extraction (regardless of plastic strain) and the percentage of
230
Geotechnical Design for Sublevel Open Stoping
0.50
0.45
Probability of instability
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
0.00
Probability of instability = 0.031 e25.3 * Velocity
R2 = 0.83
0.05
0.10
0.15
Velocity (m/step)
FIGURE 5.39
Maximum velocity values versus mining step. (From Cepuritis, P.M. et al., Back analysis of
over-break in a longhole open stope operation using non-linear elasto-plastic numerical modelling, Proceedings of the 44th US Symposium Rock Mechanics and Fifth Canada—US Rock Mechanics
Symposium, Salt Lake City, UT, June 27–30, 2010, Paper ARMA 10-124, 11pp.)
“unstable” points is shown in Figure 5.39. The relationship indicates that
at velocities >100 mm per step in the model, a 0.50 correspondence with
observed falloff was determined.
An example of the relationship between plastic strain during stope extraction (regardless of velocity) and the percentage of “unstable” points is shown
in Figure 5.40. The data indicate that instability due solely to plastic strain
only accounts for a maximum of around 25%–30% of observed instabilities.
This highlights the importance of stope-scale structure, its role in instability, and its influence on the strain field itself. The criterion is a reasonable
predictor of overall instability, with a peak probability of falloff of 0.15–0.2
at more than 5% plastic strain, which corresponds to extremely comminuted
material, or crushed rock (Beck and Duplancic, 2005). Rock masses with
this corresponding level of plastic strain would almost certainly unravel if
unconfined and exposed on a stope wall.
The correlations of instability with velocity and plastic strain are encouraging in terms of predictors of stope wall instability, and hence appear
attractive as design tools (Cepuritis et al., 2010). Levels of instability can be
predicted for a variety of stope geometries, layouts, and sequences by forward numerical analysis. Simplistically, points in the forward analysis that
display large velocities are predicted to have a very high likelihood of being
associated with instability. Points showing high levels of plastic strain, low
231
Span and Pillar Design
0.30
Probability of instability
0.25
0.20
0.15
0.10
Probability of instability = 0.686 (Plastic strain)0.452
2
R = 0.98
0.05
0.00
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
0.18
0.20
Plastic strain
FIGURE 5.40
Plastic strain values versus mining step. (From Cepuritis, P.M. et al., Back analysis of overbreak in a longhole open stope operation using non-linear elasto-plastic numerical modelling, Proceedings of the 44th US Symposium Rock Mechanics and Fifth Canada—US Rock Mechanics
Symposium, Salt Lake City, UT, June 27–30, 2010, Paper ARMA 10-124, 11pp.)
levels of confinement, and that are exposed at a stope wall are expected to
have a moderate chance of reporting as falloff.
5.5 Pillar Stability Analysis
5.5.1 Basic Concepts
Pillar design and stability analysis is a critical component of the stope design
process. Although the fundamental concepts of factor of safety as the pillar strength/average pillar stress ratio and pillar stability have been understood for some time, it is only more recently that the tools have become
available to allow more quantitative analyses of pillar strength and stability
to be carried out.
In basic engineering mechanics terms, stability refers to the stability of
equilibrium, or the ability of the overall structure, or an element of that
structure (in the present case, a mine pillar), to undergo a small change
in the equilibrium state of loading without producing a state of unstable
232
Geotechnical Design for Sublevel Open Stoping
equilibrium involving a sudden release of stored strain energy or large
deformations (Salamon, 1970; Brady and Brown, 2004). This form of instability may lead to crushing and the total collapse of a pillar and, in some cases,
its surrounds. In other cases, the peak load-carrying capacity of a pillar may
be exceeded and it may show visible signs of having been overloaded, but
it may retain some load-carrying capacity and continue to provide support
to the mine structure without undergoing unacceptably large deformations. The analysis of pillar stability in these engineering mechanics terms
is beyond the scope of this book. Here, the emphasis will be on the relationship of the average pillar stress to the pillar strength. The terms stability
and instability will not always be used in the strict engineering mechanics
sense, but may be used simply to indicate that the stress imposed on the pillar exceeds its strength.
Early developments in empirical pillar design were dominated by contributions from room and pillar methods, particularly in coal mining. More
recently, reliable hard rock empirical and numerical pillar design tools have
become available and have been implemented in sublevel open stoping for
the design of secondary stope geometries. In general, pillar strength and
stability are controlled by a large number of factors that include structural
geology, compressive strength and deformability of the rock mass, the pillar
dimensions including the width/height ratio, the degree of confinement, the
percentage extraction, and the quality of mining such as drilling and blasting.
5.5.2 Average Pillar Stress Using the Equivalent Area Approach
The stress analysis approach to pillar design requires that the load acting on
the pillar be determined using either analytical or numerical techniques. The
average pillar strength must then be evaluated and the pillar strength/stress
ratio can then be used to estimate pillar stability. The simplest approach to
the evaluation of pillar stability uses the “equivalent pillar area” technique
to estimate pillar loads.
Figure 5.41 illustrates a typical square room and pillar layout used in
mining horizontally bedded deposits. Assuming that the pillars shown are
part of a large array of pillars and that the rock load is uniformly distributed over these pillars (Hoek and Brown, 1980), the average pillar stress, σp
is given by
2
2
Ê Wo ˆ
Ê Wo ˆ
sp = sz Á
Á1 + Wp ˜
˜ = gz Á
Á1 + Wp ˜
˜
Ë
¯
Ë
¯
where
γ is the unit weight of the rock
z is the depth below surface
Wo and Wp are the widths of the opening and the pillar, respectively.
(5.10)
233
Span and Pillar Design
Plan area of pillar on surface
Wp
Wo + Wp
Z
Wp
Wp
Wo
FIGURE 5.41
Load carried by a single pillar assuming total rock load to be uniformly distributed over all
pillars. (After Hoek, E. and Brown, E.T., Underground Excavations in Rock, IMM, London, U.K.,
1980, 527pp. With permission.)
The average pillar stresses for different pillar layouts are summarized in
Figure 5.42 and, in all cases, the value of σp is given by the ratio of the weight
of the rock column carried by an individual pillar to the plan area of the
pillar. The analysis incorporates several significant simplifications and in
practice its use is restricted to shallow flat-lying deposits of significant lateral extent. As such, it may be of limited use for most hard rock mine pillar
analyses. Hence, this method must be used with caution in sublevel open
stope design, as it can be very conservative.
5.5.3 Empirical Rib Pillar Stability Chart
Hudyma (1988) analyzed data from rib pillars in a number of Canadian
open stope mines and plotted this in terms of the Y-axis (normalized pillar load to material UCS) and X-axis (pillar width/height). The database
234
Geotechnical Design for Sublevel Open Stoping
Unit length
Wp
Wo
Rib pillars σp = γz (1 + Wo /Wp)
Rock column area
Wo + Wp
Pillar area
Irregular pillars σp = γz =
Rock column area
Pillar area
FIGURE 5.42
Average vertical pillar stress in typical pillar layouts using equivalent area method—plan
views. (After Hoek, E. and Brown, E.T., Underground Excavations in Rock, IMM, London, U.K.,
1980, 527pp. With permission.)
incorporated a wide variety of rock types and pillar loads that were derived
from three-dimensional linear elastic numerical modeling. The data showed
that squat pillars under low stress were stable (lower right quadrant, Figure
5.43). Pillars become less stable as they move toward the upper left region.
Hudyma divided the graph into three general zones: failed, transition, and
stable. The database also included 13 case studies in which pillars were originally stable and subsequently yielded. These cases were observed to move
correctly through the three zones on the graph. Hudyma also suggested that
the graph could be used to predict pillar yield in open stoping design.
5.5.4 Confinement Pillar Stability Chart
A pillar stability database was developed at Westmin Resources Myra Falls
operations and was combined with seven existing pillar databases, four consisting of detailed information and three with limited information. Detailed
databases included the Westmin Resources data, Hudyma’s database collected from 13 Canadian operations, a database from the Selbi-Phikwe mine
in Botswana (Von Kimmelmann et al., 1984), and the Hedley and Grant (1972)
database from the Elliot Lake district in Ontario. The three limited databases
were from the Black Angel mine in Greenland (Krauland and Soder, 1987),
from the Zinkgruvan mine in Sweden (Sjöberg, 1992), and from Brady (1977)
from Mount Isa Mines in Australia.
Each of the databases listed used some form of pillar stability classification. In order to bring these data to a common frame of reference, a simplified pillar stability classification scale was developed (Lunder, 1994; Lunder
and Pakalnis, 1997). Pillar stability was classified as being stable, unstable,
or failed. The classification methods used for the combined database ranged
from a six-level classification quantifying various levels of pillar instability
to a more limited classification identifying only stable, sloughing, or failed
235
Span and Pillar Design
Open stope rib pillar data
0.60
Stable
Sloughing
Failure
Pillar load / UCS
0.50
0.40
0.30
0.20
0.10
0.00
0.0
0.4
0.8
1.2
1.6
2.0
2.4
Pillar width/pillar height
FIGURE 5.43
Pillar stability graph—stable, transition, and failed zones. (After Hudyma, M., Development
of empirical rib pillar failure criterion for open stope mining, MASc thesis, Department of
Mining and Mineral Processing, University of British Columbia, Vancouver, British Columbia,
Canada, 1988.)
conditions. Figure 5.44 is a schematic illustration of the pillar stability classification method developed for use at the Myra Falls mine. Pillar classifications
of 2–4 represent an unstable pillar classification for the combined database.
Table 5.2 describes the criteria used at Myra Falls to make an assessment of
the pillar stability classification. Figure 5.45 shows the excellent rock mass
conditions for typical class 1 pillars.
The average pillar stresses considered in this analysis were predominantly calculated using linear elastic numerical modeling with the exception of Hedley and Grant (1972), who used tributary area theory. Pillar
strength was presented in a general form as shown in Equation 5.11. This
equation is divided into two general terms, the first representing the
strength of the intact pillar and the second representing the effect of pillar
shape on pillar strength:
Ps = Size ¥ shape
(5.11)
where Ps is the estimated pillar strength (MPa), size is a strength term that
incorporates the size effect and the strength of the intact pillar material (MPa),
and shape is a geometric term that incorporates the shape effect of the pillar.
236
Geotechnical Design for Sublevel Open Stoping
Opening
Opening
Class 1
Class 2
Opening
Opening
Class 3
Class 4
Opening
Class 5
FIGURE 5.44
Schematic illustration of the pillar stability classification method developed for use at Westmin
Resources Ltd. (After Lunder, P., Hard rock pillar strength estimation: An applied empirical
approach, MASc thesis, University of British Columbia, Vancouver, British Columbia, Canada,
1994, 166pp.)
TABLE 5.2
Visual Assessment of Pillar Stability
Pillar Stability
Classification
1
2
3
4
5
Observed Pillar Conditions
No sign of stress-induced fracturing
Corner breaking up only
Fracturing in pillar walls
Fractures < half pillar height in length
Fracture aperture < 5 mm
Fractures > half pillar height in length
Fracture aperture > 5 mm, <10 mm
Disintegration of pillar
Blocks falling out from pillar
Fracture aperture > 10 mm
Fractures through pillar core
Source: Lunder, P., Hard rock pillar strength estimation:
An applied empirical approach, MASc thesis,
University of British Columbia, Vancouver,
British Columbia, Canada, 1994, 166pp.
Span and Pillar Design
237
FIGURE 5.45
Excellent rock mass conditions: Example of Class 1 pillars (MRM, Northern Territory).
Two formulae that can be used for the estimation of pillar strength were
developed by Lunder (1994) including the “log-power shape effect formula”
and the “confinement formula.” Both formulae are virtually identical when
plotted on a stability graph. However, the difference is that the “log-power
formula” is a purely empirical formula, while the “confinement formula” is
a modified form of the Mohr–Coulomb failure criterion (Lunder, 1994). Both
formulae use the average pillar confinement term as subsequently described.
The combined database was analyzed in order to determine if any past
methods could be applied successfully to the combined database. It was determined that these historical methods could not adequately represent the combined database over the full range of pillar width/height ratios (Lunder, 1994).
Individual linear shape effect constants were derived for each of the databases described earlier. These values enabled the assignment of a strength
factor that is used to correct (i.e., scale) the unconfined compressive strength
of intact pillar material to the full-size unconfined compressive strength of the
pillar. This value is the “size” term in Equation 5.11, where the full-size unconfined compressive strength of a mine pillar can be represented by ≈44% of the
unconfined compressive strength of the intact pillar material (Lunder, 1994).
Pillar strength has been related to the pillar width/height ratio extensively
in the past. However, the strength of a rock mass is known to be a function of the applied and the confining stresses. Using two-dimensional elastic boundary element modeling, it was determined that a relationship exists
between the pillar width/height ratio and a term called the “average pillar
confinement” and represented by the symbol Cpav. The average pillar confinement is defined as the ratio of the average minor pillar stress (σ3) and the
238
Geotechnical Design for Sublevel Open Stoping
average major pillar stress (σ1). These values are measured at the mid-height
of the pillar. Equation 5.12 is the relationship that was determined to relate
the pillar width/height ratio and the “average pillar confinement.” The value
of “coeff” in Equation 5.12 is dependent on the extraction ratio in the vicinity
of the pillar. For typical extraction ratios in underground hard rock mines of
70–90%, a value of 0.46 for “coeff” has been determined to be acceptable with
less than 10% error (Lunder, 1994):
1.4
Cpav
È Êw
ˆ˘Êw ˆ
= coeff Ílog Á + 0.75 ˜˙ÁË h ˜¯
¯˚
Î Ëh
(5.12)
where
Cpav is the average pillar confinement
coeff is the coefficient of pillar confinement and set to 0.46
w is the pillar width (m)
h is the pillar height (m)
A modified strength formula, “the confinement formula,” that resembles the
Mohr–Coulomb strength criterion was determined by Lunder (1994) to represent the combined database with a prediction success that has a slightly
higher predictability rate (87% versus 85%) than the “log-power” formula.
The “confinement formula” is represented by Equation 5.13. Empirical constants representing rock mass properties have been determined for C1 and C2
to be 0.68 and 0.52, respectively. This method is presented graphically along
with all of the case histories in the combined database on Figure 5.46.
The fundamental difference between the “log-power formula” and the
“confinement formula” is that the latter is based on the theory of the strength
of a rock mass, while the “log-power formula” is a purely empirical formula
for which curve-fitting parameters have been determined. Pillar strength in
the “confinement formula” is driven by the mine pillar friction term “kappa,”
as defined in Equation 5.14, which is a function of the applied and confining
stresses on the pillar only:
Ps = (k sc ) (C1 + C2 kappa )
(5.13)
where
Ps is the pillar strength (MPa)
k is the pillar size factor = 0.44
σc is the unconfined compressive strength of the pillar material (in MPa for
a 50 mm diameter sample)
C1 and C2 are empirical rock mass constants (0.68 and 0.52, respectively)
kappa is a mine pillar friction term, calculated as follows:
239
Span and Pillar Design
0.7
F.S. = 1.0
Average pillar stress/UCS
0.6
0.5
F.S. = 1.4
0.4
0.3
0.2
Stable
Sloughing
Failure
0.1
0.0
0.0
0.4
0.8
1.2
1.6
2.0
2.4
2.8
3.2
Pillar width/pillar height
FIGURE 5.46
The confinement formula stability graph plotted with all case histories from the combined
databases. (From Lunder, P., Hard rock pillar strength estimation: An applied empirical
approach, MASc thesis, University of British Columbia, Vancouver, British Columbia, Canada,
1994, 166pp.)
È
Ê1 - Cpav
kappa = tan Ícos -1 Á
Á1 + Cpav
Í
Ë
Î
ˆ˘
˜
˜˙
¯˙
˚
(5.14)
where Cpav is the average pillar confinement and is defined by Equation 5.12.
The lines dividing each of the pillar stability classifications have been
assigned a factor of safety. This assignment is based upon the assumption
that the line dividing the unstable and failed pillars has a factor of safety
of 1.0. Using this as a baseline, it was determined that the transition from
unstable to stable pillar conditions would have a calculated factor of safety
of 1.4 (Lunder, 1994).
In order to use the design guidelines developed with confidence, the
method must be calibrated to existing conditions. Calibration is accomplished through the observation of existing pillar conditions and calculated
stress values. If the observed pillars do not fall in the correct region on the
pillar stability plots, modification to the input parameters is required. The
modification can either be to the values that are used as input for stress
determination (the in situ stress values) or to the unconfined compressive
strength of the intact pillar material such that the pillars used for calibration fall in the correct region on the pillar stability plots. Figure 5.47 shows
Lunder’s Canadian database and over 50 points from the McArthur River
240
Geotechnical Design for Sublevel Open Stoping
10.0
Lunder
Stable
Unstable
Failure
UCS/average pillar stress
9.0
8.0
7.0
MRM
Stable
Unstable
Failure
FS1.4
FS1.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
0
0.5
1
1.5
2
2.5
3
3.5
Pillar width/height ratio
FIGURE 5.47
Pillar stability graph—Lunder and MRM data. (From Schubert, C.J. and Villaescusa, E., An
approach to hard rock pillar design at the McArthur River mine, Proceedings of the AusIMM
Annual Conference—The Mining Cycle, Mount Isa, Queensland, Australia, April 19–23, 1998,
pp. 255–259, AusIMM, Melbourne, Victoria, Australia. With permission.)
Mine (MRM) in Australia (Schubert and Villaescusa, 1998). The MRM
results confirm the generality of the method. The data suggest that for a
ratio of σc/σp less than 2, the majority of the pillars are unstable, regardless
of the pillar W/H ratio. Furthermore, when the σc/σp ratio is greater than 5,
even slender pillars are stable. This supports the changes suggested earlier
to the factor A in Figure 5.12.
5.5.5 Numerical Modeling for Pillar Design
Both three-dimensional linear elastic and nonelastic numerical models can
be used for pillar design. For linear elastic analysis, the three-dimensional
stoping geometries can be represented in almost any required detail
incorporating sequencing. Elastic models are generally run as single
material models as incorporation of multiple geological materials generally has limited effect on the final stress outcome. Some models allow
inclusion of a limited number of major geological discontinuities. The
programs MAP3D (Wiles, 2006) and Examine3D (Rocscience Inc, 1990)
are typical of the three-dimensional elastic numerical analysis software
available. Output from such models is generally relatively straightforward
to interpret with contours of principal stress and factor of safety often
displayed (Figure 5.48).
241
Span and Pillar Design
While three-dimensional elastic models provide a reasonable representation of the stress redistribution resulting from a stoping process, many
pillars are subject to varying degrees of failure, particularly at the exposed
pillar faces, and resultant stress redistribution to the pillar core cannot
be simulated unless nonlinear models are used to analyze the responses
σ1
(MPa)
100.0
90.0
80.0
70.0
60.0
50.0
40.0
30.0
20.0
10.0
0.0
400 Level
380 Level
360 Level
340 Level
320 Level
300 Level
σ3
(MPa)
50.0
45.0
40.0
35.0
30.0
25.0
20.0
15.0
10.0
5.0
0.0
400 Level
380 Level
360 Level
340 Level
320 Level
300 Level
(a)
FIGURE 5.48
Major principal stress and strength factor for Eloise Deeps mine 30 m wide pillar using the
program MAP3D. (a) Longitudinal view of major and minor principal stresses.
242
Geotechnical Design for Sublevel Open Stoping
Strength
Factor-A
1.00
1.10
1.20
1.30
1.40
1.50
1.60
1.70
1.80
1.90
2.00
Average
pillar
SF-A= 1.7
400 Level
380 Level
2.8
m
UCS = 48
= 45°
6.5
m
360 Level
340 Level
320 Level
300 Level
Strength
Factor-A
1.00
1.10
1.20
1.30
1.40
1.50
1.60
1.70
1.80
1.90
2.00
UCS = 48
= 45°
N
(b)
FIGURE 5.48 (continued)
Major principal stress and strength factor for Eloise Deeps mine 30 m wide pillar using the
program MAP3D. (b) Longitudinal and plan view of Strength Factor-A.
of yielding pillars. Consequently, one of the most significant improvements in mine design has come from a move toward calibrated, multiscale,
nonlinear numerical modeling. Gross deformation simulated at a global
stope sequencing scale can be used to provide the boundary conditions for
a smaller, stope length scale model that incorporates more detailed material
properties incorporating discrete fracture networks (Beck et al., 2010).
Span and Pillar Design
243
Massive, strain softening, dilatant analysis is often used for multiscale
stope design and analysis. The greatest improvement has been the rationalization of the use of submodels, which have the ability to correctly replicate
observed displacements at all length scales. An immediate consequence is
the ability to use velocity and displacement as criteria for instability (see
Section 5.4.2). The mechanisms of damage and deformation that affect stability at each stoping sequence can then be successfully captured.
Currently, nonlinear three-dimensional modeling can be conducted
using various commercially available finite element codes such as the program Abaqus. Other specialized codes such as FLAC3D (three-­dimensional
finite difference) and 3DEC (three-dimensional distinct element) are also
available. Each of these techniques can be very useful depending on the
specific problem to be solved. For example, if the nature of the problem
involves slip on major geological structures intersecting the pillar, then a
distinct element program such as 3DEC may provide the appropriate analysis tool. In cases where general plasticity (crushing) failure is dominant,
a continuum three-dimensional code may provide the most appropriate
analysis tool.
In all cases, however, the key to successful prediction of rock mass behavior
is the ability to quantify rock mass failure and its behavior after failure. The
selection of a realistic nonlinear constitutive model to provide the relation
between the stresses and strains that can be sustained by a fractured hard
rock mass is required. However, a detailed development of such a topic is
beyond the scope of this book.
6
Drilling and Blasting
6.1 Introduction
Drilling and blasting in sublevel open stoping involves the interaction of
the rock mass, the drillhole patterns, the explosives types, and the initiation
sequences. The performance is measured in terms of safety, rock fragmentation, muckpile characteristics, stability of the exposed stope walls, and damage to nearby areas and equipment (Figure 6.1).
The objective of the blast design process is to determine the number,
position, and length of required blastholes with respect to the available development and the stope boundaries, while taking into account the orebody
shape, ground conditions, groundwater, available equipment, stope access
geometry, hole size, and the explosive types. In addition, the economic objective is to achieve the desired fragmentation (with minimum damage to the
exposed stope walls and stope accesses) by means of a minimal use of explosives, materials, and time. Damage to the surrounding areas such as dented
ventilation fans, ripped ventilation bags, dislodged and broken service pipes,
and electrical cables must be avoided. Furthermore, the consequences of
coarse fragmentation range from hung-up drawpoints and excessive secondary breakage to difficult mucking, loading, and tramming, causing increased
maintenance costs on trucks and loaders. The effects of undersized fragmentation are excessive fines, overloaded equipment, and milling problems.
6.2 Longhole Drilling
Sublevel open stoping requires the accurate and efficient drilling of relatively
long blastholes within a designed stope boundary. Depending upon the rock
mass conditions and stoping geometry, ring drilling may involve upholes,
downholes, one-sided rings, and full 360° rings in vertical, inclined, or horizontal planes. Drilling is achieved by percussion mechanisms and adequate
feed pressure, with bit penetration resulting from localized crushing and
245
246
Geotechnical Design for Sublevel Open Stoping
In situ block size distribution
Geological discontinuities
Intact rock bridges
+
Blast energy
Fragmentation
Gas expansion
Vibrations
Drilling accuracy
Explosive strength
Confinement
Stand-off distance
Muckpile shape, looseness,
and muckability
+
Damage
Instability, dilution,
airblast
FIGURE 6.1
The drilling and blasting process in sublevel stoping.
chipping at the rock–bit interface. In addition, rotation is required to change
the button position within the toe of the hole following each percussive
impact of the striker bar on the drill string or bit. Finally, flushing is required
to remove the rock cuttings and also to cool the drilling tools (Puhakka, 1997).
Drilling for longholes in sublevel stoping involves either top-hammer or
in-the-hole (ITH) drilling mechanisms. In a top-hammer configuration, the
rock drill or drifter remains on the top of the drill string, requiring transfer
of the impact energy from the drifter through the entire drill string to reach
the bit. In ITH drilling, the impact mechanism is located directly above the
bit and enters the hole as the first piece of the drill string (Hamrin, 1993).
Therefore, the impact energy is transferred over a shorter distance of the
drill string prior to reaching the rock–bit interface. The minimum drillhole
diameter and the required drilling accuracies determine which type of drilling configuration is suitable for each application.
Percussion drilling is restricted by the ability of a drill steel to transmit
energy. Drill stems are likely to deteriorate when subjected to excessive
energy during impact force transmission. Consequently, percussion pressure settings must be established considering penetration rates and drill
steel economy. Optimal feed pressures can be determined for a particular
rock type following observations of penetration rates, bit wear, and steel
threadwear (Puhakka, 1997). Excessively high feed forces do not necessarily
Drilling and Blasting
247
achieve increased penetration rates (Schunnesson and Holme, 1997). One
problem experienced with excessive feed forces during drilling is bending
of the drill steels resulting in increased drillhole deviation.
6.2.1 Top-Hammer Drilling
Top-hammer drilling relies on the transferal of percussive energy (torque and
impact) to the rock–drill bit interface via the drill stem. This energy is generated by a piston in the rock drill using pneumatic or electrohydraulic means.
The drill bit contains no moving parts and simply screws onto the drill rod
end. The rate of bit penetration is a function of the transferred impact force,
the blow frequency, rotation speed, and the flushing efficiency (Puhakka,
1997). Energy losses along the drill string increase with hole depth, thereby
reducing penetration rates.
The hole diameter for top-hammer production hole drilling ranges from
51 to 127 mm with the hole length limited to 50 m (using a 127 mm hole
diameter) due to the weight of the drill string and storage capacity of the
tube magazine (Hamrin, 1993). In most cases, however, the hole length is
usually restricted to less than 35 m due to limitations in hole drilling accuracy. Top-hammer rigs have drifters that are suitable for a small range of hole
diameters and a typical rig is only capable of covering a spread of 50 mm
between the minimum and maximum hole diameters. In order to drill a different-sized hole, a change of drifter as well as a change of drill string and
hydraulic pumps may be required.
6.2.2 In-the-Hole Drilling
In this drilling method, the percussive hammer is located inside the hole
directly above the bit. The drilling bit is a continuation of the shank on which
the drill piston impacts directly. Consequently, little energy is lost during
the drilling process and penetration rates are almost constant regardless of
hole depth. ITH drilling is typically only applicable to larger-diameter blastholes due to the space required to house the in-hole striker element and the
increased drill string diameter. Drilling directions are logistically limited to
subhorizontal to vertical downholes due to the inherent difficulties of charging explosives into large-diameter upholes. The main advantage of ITH
drilling of longholes is improved hole accuracy compared with top-hammer
drilling. This is very important in sublevel open stoping where the ability
to accurately drill long, large-diameter holes allows for greater distances
between sublevels, thereby reducing the costs of stope development access.
Commonly used hole sizes for ITH drilling range from 85 to 215 mm,
with holes extending up to 60 m in length. A disadvantage of ITH drilling
is that low penetration rates, compared with the top-hammer technique, are
likely to be achieved. In addition, the need for a large separate compressor
results in reduced equipment mobility. Nevertheless, ITH drilling is the
248
Geotechnical Design for Sublevel Open Stoping
only technique capable of drilling very longholes with satisfactory accuracy.
Another advantage with ITH drilling is that all the specified diameters can
be drilled using one drill rig as the ITH hammer can be exchanged for a
hammer of the required diameter and the existing drill string is retained.
6.2.3 Drilling Equipment Selection
Considerations of the general mine layout including any special drilling
needs are required during equipment selection. The equipment must be
mobile and versatile as it is likely that it will perform a number of tasks while
traveling to different locations in a reasonable amount of time. Typical tasks
may include drilling holes of varying lengths, multiple diameters, different
dip and dump angles, and upholes or downholes. In all cases, the selection
of the stiffest rod–bit combination within a drill steel is critical to minimize
hole deviation. Table 6.1 shows some suitable combinations of bit and rod
diameters for longhole drill strings for production drilling in sublevel open
stoping.
Additional capabilities that require consideration during rig selection
include crawler or wheel-mounted carriers and the selection of a boom capable of drilling a full 360° ring while tilting backward and forward. Other
considerations are the selection of a feed system that can provide an adequate and smooth feed force at all feed pressures (to ensure that straighter
holes are drilled), selection of a suitable rod changer, drill bit type, shape and
cutter configuration, drill rod types and couplers, and any additional drill
string stabilizing elements such as tubes or guides.
One important operational factor is the flushing velocity of the air/water/
oil required to remove rock fragments from the face of the drill bit and propel
them out of the drillhole. The ITH drilling system uses large-diameter tubes
TABLE 6.1
Selection of Drill String Combinations for
Longhole Drilling
Hole/Bit
Diameter (mm)
51
64
73
76
89
102
115
127
140
165
Rod Diameter
(mm)
Tube Diameter
(mm)
32
38
38
45
51
–
–
–
–
–
–
–
–
64
76
85
89
89
115
115
249
Drilling and Blasting
resulting in small apertures between the tubes and the wall of the hole. Given
that a constant volume of air is pumped down the string to operate the hammer, high air velocities with excellent flushing capabilities are achieved. On
the other hand, top-hammer drilling utilizes small-diameter strings resulting in low flushing velocities due to the large aperture between the drill steel
and the wall of a hole. However, if a drill string composed of drill tubes is
used, the flushing capabilities of a top-hammer system can be increased. In
addition, drillhole deviation can be minimized.
For similar diameter holes, the initial cost of a top-hammer drill rig is usually higher than that of an ITH drill rig. However, for short-length, smalldiameter holes, the high productivity achieved with the top-hammer drill
rig ensures that it remains competitive. In summary, the decision on which
method will be used depends on many factors, some of which may be sitespecific. Usually the depth of the holes and required accuracies are primary considerations, with ITH drilling preferred for holes exceeding 35 m
in length. Conversely, for short-length, small-diameter holes, top-hammer
drilling is well suited.
6.2.4 Drilling Deviation
Drill deviation
(equivalent blasthole diameters, m)
Blasthole deviation is defined as the difference between the designed path of
a drillhole and its actual trajectory. The total deviation from a planned drillhole location can be attributed to three factors. These are incorrect collar positioning, drill alignment error, and ITH deviation from a planned trajectory
(Figure 6.2). The extent of each of the three sources of error depends upon the
rock properties and geometry of the blast, type of drilling equipment, drill bit
25
20
l
ota
T
15
Bending
error
10
Setup error
5
0
or
err
Collaring error
0
50
100
150
200
250
Drilled depth (equivalent blasthole diameters, m)
300
FIGURE 6.2
Drill deviation types in longhole drilling. (After Heilig, J., Blast engineering—course notes
for the masters of engineering science in mining geomechanics, Western Australian School of
Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 1999.)
250
Geotechnical Design for Sublevel Open Stoping
and rod specifications, and drill operation parameters (Kleine et al., 1992). The
first two types of error are usually random in nature, and can be minimized
by adequate markup and drilling procedures. ITH deviation is the bending of
the holes as they are drilled and is a function of the forces acting on the drill
strings and the drill string flexibility. This third type of error can compound
the effect of either or both of the previous error types leading to an aggregate
total error greater than any of the three alone (Kleine et al., 1992).
6.2.4.1 Collar Positioning
2.0
1.5
1.0
North (m)
A collar position error arises from the inaccurate location of the drill rig prior
to drilling. Usually, the drive centerline and ring positions are marked on
the backs or walls of a drilling drive either by the mining survey department
or by the drillers. The collar positions of the holes within each individual
ring can be painted on the floor, walls, or back and spaced out using a tape
measure. The rig is then positioned between the ring markings and drilling
is undertaken over the marked collars. For modern computer-controlled drill
rigs, drill collar location control can be maximized by positioning the drill rig
at a designed pivot point within the drilling drive. The hole collar positions
are then identified by a hole dip and dump angle as drilled from the specified
pivot point.
Errors in hole collaring are independent of the hole diameter, length, and
drilling equipment used. The errors can be determined by comparing the
actual collar locations with the planned collar locations (Figure 6.3). In this
figure, the planned locations of the hole collars are represented by the intersection of the two axes, while the actual collar locations are represented by
each of the points in the plot. The data suggest a smaller error in the north–
south direction than in the east–west direction. In this particular case, errors
in a north–south direction are minimized by marking each ring position on
both of the drill drive walls. Errors in an east–west direction are incurred
by the poor location of each individual collar within the ring. Therefore,
0.5
–2.0
–1.5
–1.0
–0.5
–0.5
–1.0
–1.5
–2.0
FIGURE 6.3
Collaring error due to poor drill rig positioning.
East (m)
0.5
1.0
1.5
2.0
251
Drilling and Blasting
good drilling surfaces and the accurate marking of each individual collar is
required to minimize such errors.
Quality control during the drillhole design process and drill setup is the simplest way to reduce collar-positioning errors. However, this is potentially the
most difficult solution to implement consistently, as it depends upon the attitude and work procedures followed by the drillers. Setup errors are increased
by driller boredom and compounded when the drillers are paid large meterage
bonuses. Having a quality component as part of the wages has been known to
reduce this type of error, as drilling inaccuracies can be considered a symptom
of a “people problem,” usually caused by an underlying management problem.
Design issues when using computer-controlled drill patterns have also
been identified (Fleetwood, 2010). If the actual floor elevation is different
from that used in the design, hole collar location errors are increased. This is
typical of when the floor of the drive is loose from overblasting of the lifters
during development and the floors are cleaned up prior to drilling to make
for easier collaring. This change in floor elevation is not taken into account
in the drillhole design which used the initial drive laser surveys for collar
location designs.
6.2.4.2 Drillhole Alignment
Drill alignment error arises during the siting of the drill boom such that
the initial orientation of the drillholes does not match the design. A change
from design in either bearing or plunge will cause drill deviation that will
increase as the drill path progresses. This error can be detected either by
monitoring the initial drill setup angles or by calculations from down the
hole survey measurements within the initial 2.5 m from the hole collar. Little
or no in-hole deviation would be expected to occur in the first 2.5 m due to
bending of the drill string. The alignment error can be calculated as the solid
angle between the planned bearing and plunge and the surveyed bearing
and plunge (Figure 6.4).
Z (Elev)
Planned (P)
Drilled (D)
δ
Collar
p
d
βp
Px = cos
Py = cos
Pz = sin
p
p
sin βp
cos βp
p
Y (North)
βd
X (East)
FIGURE 6.4
Solid angle between a drilled path and a designed path.
Dx = cos
Dy = cos
Dz = sin
d
d
d
sin βd
cos βd
252
Geotechnical Design for Sublevel Open Stoping
The solid angle δ can be calculated from the dot product of the unit vectors
of the direction cosines of the planned and the drilled holes (see Equation 5.3).
Extensive surveying data from a number of typical bench stoping operations
in Australia indicate that average solid angles of about 2° are typical. The
estimated deviation due to ring misalignment from such a solid angle is a
very significant ±3.5%.
Experience suggests that in addition to incorrect drill positioning, uneven
drilling surfaces also contribute to this type of error. Errors in azimuth are related
to errors in burden, which can be minimized with the use of ring laser alignment. Plunge misalignment relates to deviations in toe spacing, where boom
kickback when collaring also contributes to the error. In general, boom stability
can be improved with the use of sufficiently long “stingers” capable of reaching
both the floor and the back of the drilling drives or by a “horseshoe” stabilizer
on the drill boom. The use of electronic pendulums or digital tilt-meters can
help to monitor the alignment of the boom while drilling (Hamrin, 1993).
Conventional ring drilling alignment consists of aligning the drill boom by
eye, to a pair of paint marks on each side of the drill drive defining the planes
of the rings. This technique is subject to markup, and setup errors and deviation from a design plane could be high as different drillers line up the drill
rig from slightly different positions. An alternative is to set up a longitudinal
alignment technique in which laser beams are used to locate the drill rig
parallel to the drill drive center line (Figure 6.5). A set of suitable targets are
accurately surveyed into position at each end of the drilling drives, allowing
rig alignment by means of the laser beams. Hole survey results indicate that
the bearing alignment error can be reduced by up to 5° using this technique.
Additional lasers can be added perpendicular to the spine of the rig for accurate alignment with the plane of the ring as specified by the wall markups.
6.2.4.3 In-the-Hole Deviation
ITH deviation is related to the bending of drillholes and occurs when the
bit deviates from a straight path as it drills through a rock mass. Bending of
Ring line
Laser beam
Filled stope
Drill rig
Targets
FIGURE 6.5
Longitudinal drill rig alignment using lasers. (Courtesy of Mount Isa Mines, Mount Isa,
Queensland, Australia.)
253
Drilling and Blasting
holes is a function of noncontrollable factors (rock properties and geological
features) as well as drill operation parameters such as thrust and torque and
rod and drill bit specifications (Kleine et al., 1992). Rod deviations are caused
primarily by the forces acting on the drill string and also due to the drill
string flexibility. The flexibility depends upon the rod stiffness, which is a
function of the physical makeup and the active length of the drill string to
diameter (l/d) ratio.
As drill string flexibility increases and/or the annulus (area difference
between rod and drill bit) increases, a greater possibility of ITH deviation
arises. In the early stages of drilling, drill string flexibility (related to the
ratio l/d) is low. As drilling progresses and the ratio increases, so too does
the flexibility, and increased bending is likely. In addition, more flexible
strings will offer less resistance to side-loading changes on the bit. This can
occur when the bit drills across rock types having different strengths or
stiffnesses. Each change on side loading causes the bit to drill off in a slightly
different direction, thus contributing to deviation.
Consequently, to minimize ITH deviation, it is important to use stiff rods
to prevent flexing as well as a suitable choice of rod and drill bit combination. For a given hole length and drill bit diameter, smaller diameter rods
have more space within which to flex in a drillhole. Hence, it is expected that
a T45 speed rod and a 76 mm bit combination would drill straighter holes
than a T45 rod using an 89 mm bit. In both cases, the flexibility is the same;
however, the T45 rod–76 mm bit combination has a smaller annulus, thereby
reducing in-hole deviation. One problem with a smaller annulus is that rod
couplings may become entrapped if the bit wears down or rocks fall behind
a coupler. Flexing of the drill string can also be reduced with the use of tube
drilling technology. Hence, a combination of 64 mm tube and 76 mm bit is
expected to deviate less than a T45 rod–76 mm bit combination due to the
increased stiffness provided by the tubes.
The ITH deviation magnitude and orientation can be calculated using a
three-dimensional vector analysis. ITH deviation is analyzed by considering
both the planned drillhole trajectory and the actual surveyed path as follows
(see Figure 6.4).
P is the unit vector in the direction of the planned hole whose direction
cosines are Px, Py, and Pz as defined by Figure 6.4. Let the coordinates of a
surveyed point S along a drilled hole be
S = (Xs , Ys , Zs )
(6.1)
A general vector on a planned hole is given by
T = (tPx , tPy , tPz )
where t is a real parameter.
(6.2)
254
Geotechnical Design for Sublevel Open Stoping
The distance of the surveyed point S to the planned hole is the length of
the vector
TS = (Xs - tPx , Ys - tPy , Zs - tPz )
(6.3)
precisely when the vector TS is perpendicular to the planned hole, that is,
when the dot product of the vector TS and the unit vector P is zero:
(Xs - Pl x , Ys - Pl y , Zs - Pl z ) i (Px , Py , Pz )= 0
(6.4)
This implies that
t=
XsPx + YsPy + ZsPz
Px2 + Py2 + Pz2
(6.5)
Therefore, the distance d from the surveyed point S to the planned hole,
which is the ITH deviation, is given by
d = ÷ (Xs - tPx )2 + (Ys - tPy )2 + (Zs - tPz )2
(6.6)
where t is given by Equation 6.5.
The average orientation of the deviation for each surveyed depth along a
drilled hole can be estimated by a method suggested by Priest (1985). First,
each surveyed point can be represented by a unit vector centered at the
origin of the system of coordinates shown in Figure 6.4. The X, Y, Z coordinates of the terminal point of the ith vector are given by
Nix = cos jd sin bd
Niy = cos jd cos bd
(6.7)
Niz = sin jd
where βd and φd are the trend and plunge of the drilled hole at the point of
measurement.
The X, Y, Z coordinates of the terminal point of the resultant or average
hole deviation are given by
rx =
N
 Nix
i =1
ry =
N
 Niy
i=1
rz =
N
 Niz
i =1
(6.8)
255
Drilling and Blasting
The trend (βave) and plunge (φave) of the average deviation are given by
Ê rx ˆ
bave = arctan Á ˜+ q
Ëry ¯
(6.9)
and
Ê
rz
jave = arctan Á
2
Á (rx) + (ry)2
Ë
ˆ
˜
˜
¯
(6.10)
where the term q is an angle that, depending on the sign of rx and ry, ensures
that βave lies in the proper quadrant. If (rx ≥ 0 and ry ≥ 0), then q = 0 and when
(rx < 0 and ry > 0), then q = 0, otherwise, q = π.
ITH deviation calculated from surveyed data from a number of drill
string and rod bit combinations can be analyzed for different hole lengths
in order to determine the critical depth for each rod–bit combination.
Figure 6.6 presents a comparison of average deviation with depth from
downhole bench stoping using a top hammer Atlas Copco Simba H221
fitted with T38 and T45 speed rods. Each hole was surveyed at depths of
0
T38–73 mm
T45–76 mm
T45–89 mm
Depth (m)
–5
–10
–15
–20
0.0
0.1
0.2
0.3
0.4
0.5
Drill deviation (m)
FIGURE 6.6
Example of average deviation for different drilling strings.
0.6
0.7
0.8
256
Geotechnical Design for Sublevel Open Stoping
2.5, 5, 10, and 15 m. The rod deviations from the collar to a 10 m depth
­differed only slightly for the three combinations, with an absolute deviation of about 0.2 m. However, a significant difference was found at 15 m,
where the straightest drilling string was T45 rods with a 76 mm bit. An
average deviation of 0.28 m was observed with a standard deviation of
0.09 m. The worst performing combination was the T45 rods in conjunction
with 89 mm bits that resulted in an absolute average deviation of 0.60 m
with a standard deviation of 0.26 m. In comparison, an average collar position error of 0.24 m with a standard deviation of 0.15 m was determined for
this particular drilling operation.
In addition to the average deviation values, individual hole deviation distribution from a design target must be considered. Figure 6.7 presents a comparison of deviation distribution at a hole depth of 10 m for each of the drill
string combinations. A definitive trend for the blastholes to deviate in an
east–west direction (toe spacing) was found at this site and the T45–89 mm
string combination was the worst performer.
0.8
0.8
∆Northing (m)
0.4
0.6
0.4
∆Northing (m)
0.6
T38–73 mm
at 10 m
0.2
0.0
–0.2
0.2
0.0
–0.2
–0.4
–0.4
–0.6
–0.6
–0.8
–0.8 –0.6 –0.4 –0.2 0.0 0.2
0.4 0.6 0.8
T45–76 mm
at 10 m
–0.8
–0.8 –0.6 –0.4 –0.2 0.0 0.2
∆Easting (m)
∆Easting (m)
0.8
0.6
∆Northing (m)
0.4
T45–89 mm
at 10 m
0.2
0.0
–0.2
–0.4
–0.6
–0.8
–0.8 –0.6 –0.4 –0.2 0.0 0.2
0.4 0.6 0.8
∆Easting (m)
FIGURE 6.7
Drill deviation at 10 m hole depth from different drill string combinations.
0.4 0.6 0.8
257
Drilling and Blasting
Deviation
North (m)
1.50
1.00
Deviation
East (m)
0.50
3.00
–2.00
–1.00
1.00
Average
–0.50
–1.00
–1.50
2.00
3.00
Deviation at 8.5 m depth
Deviation
North (m)
1.50
1.00
Deviation
East (m) 3.00
0.50
–2.00
–1.00
1.00
–0.50
–1.00
–1.50
2.00
3.00
Average
Deviation at 15 m depth
Deviation
North (m)
1.50
1.00
Deviation
East (m) 3.00
0.50
–2.00
–1.00
–0.50
–1.00
–1.50
1.00
2.00
3.00
Average
Deviation at 20 m depth
FIGURE 6.8
Drill deviation for different hole depths. (After Cameron, A. and Paley, N., Assessment of blasting to reduce damage in B704 bench stope at Mount Isa Mines, in T. Szwedzicki, G.R. Baird,
and T.N. Little, eds., Proceedings of the Western Australian Conference on Mining Geomechanics,
Kalgoorlie, Western Australia, Australia, June 8–10, 1992, pp. 375–383, Western Australian
School of Mines , Kalgoorlie, Western Australia, Australia.)
Figure 6.8 presents a comparison of average deviation with depth from
downhole bench stoping using a top hammer Atlas Copco Simba H221 fitted
with T38 speed rods. Each hole was surveyed at depths of 8.5, 15, and 20 m.
The results show that for the 20 m-long blastholes there was a high probability of both excessively small and large toe burdens. A definitive trend
for the blastholes to deviate in an east–west direction was also found at this
particular site.
258
Geotechnical Design for Sublevel Open Stoping
Studies of hole deviation have shown that greater accuracy can be achieved
by adding guide rods to a drill string or by using a tube string. Results from
a study entitled the “straight hole” project carried out by Atlas Copco and
the LKAB Kiruna iron ore mine attempted to quantify the effects of drillhole
diameter on the expected deviation. The study was based on uphole drilling
(drillhole length up to 50 m) using top-hammer drills in conjunction with tube
strings (Hamrin, 1993). The results shown in Figure 6.9 can be used to determine the maximum drillhole length for a given diameter, where the target
deviation for 95% of the holes does not exceed half the normal ring burden.
It is important to note that the “straight hole” project guidelines are only
applicable if the conditions associated with modern techniques of precision
drilling apply. The maximum hole depth in Figure 6.9 may be achieved by
a drill rig with appropriate angle instrumentation setup and by the use of
a rigid tube drill system. In addition, a minimum collaring error of ±0.10 m
and a rig alignment error of only ±1.0% have been assumed. As a comparison,
Table 6.2 shows deviation data collected from 89 mm-diameter holes drilled
with 64-mm diameter tube strings. Average bearing and plunge misalignments of 3.9° and 1.1°, respectively, were determined from the calculations.
Another factor causing deviation is the effect of gravity on the bit. A pendulum effect may be experienced in longholes when gravity forces acting in a
bit cause it to cut the bottom of the hole, gradually steepening the hole (Figure
6.10). Solutions such as increasing drilling thrust or placing stabilizing devices
near the bit (to rotate it into the desired orientation) have been suggested to
correct this problem. A negative offset in Figure 6.10a means that the hole
has deviated north. The shallow holes are closest to the design, but tend to
deviate to the north, perhaps due to drill setup error. The calculated burden
in Figure 6.10b shows that all the holes begin within the correct plane, but get
out of plane by the toe, especially the steeper holes.
6.3 Blast Design Parameters
The dimensions of a blasthole pattern must be selected to suit the rock mass
conditions, the geometry of the orebody, and the limitations of the drilling
equipment. Blast patterns can then be adjusted to determine an optimal
design for the different stope geometries such as production rings, cutoff slot
(COS) holes, fill diaphragm, and trough undercut (TUC) rings. This process
is based upon accumulated knowledge from previous experience in rock
masses having similar strength and jointing conditions. The factors considered are the drilling access, the blasthole diameter and length, the burden
and spacing, the explosive types, and the effects of timing and sequencing.
The benefits achieved when a blast design is optimized include increased
excavation stability, good fragmentation with reduced mucking (loader) unit
50
89
100
150
89
102 115 127 140 155 165
Hole diameter (mm)
57 64 76
ting
las
ole b
gh
Lon
bla
Top hammer range ITH drilling range
h
nc
Be
0.50
1.00
1.50
2.00
0
(1.3 m)
10
Example: hole diameter: 89 mm
Limit for
hole spread (m) Estimated in-the-hole deviation: 2.5%
Maximum hole depth: 35 m
30
Maximum hole length (m)
20
ed
1.0
%
40
50
Set-up and direction
1.0% of hole depth
hol
Collaring 0.1 m
the
In
n
atio
evi
%
2.0
%
3.0
FIGURE 6.9
Maximum hole length using precision drilling. (After Hamrin, H., Precision drilling extends the range of longhole blasting, in G. Almgren, U. Kumar
and N. Vagenas, eds., Proceedings of the 2nd International Symposium on Mine Mechanization & Automation, Luleå, Sweden, June 7–10, 1993, pp. 143–151,
Balkema, Rotterdam, the Netherlands.)
1.00
2.00
(2.6 m)
3.00
4.00
ng
sti
u
um
Ac
c
r
sp
d
lat
e
0%
4.
ea
d
Nominal
burden (m)
Drilling and Blasting
259
260
Geotechnical Design for Sublevel Open Stoping
TABLE 6.2
Hole Deviation for 89 mm Holes Drilled with 64 mm Tube Strings
Hole ID
Depth
(m)
Total
Deviation (m)
In-the-Hole
Deviation (m)
Bearing
Misalignment (°)
Dip
Misalignment (°)
R1-HW
R1-FW
R1-easer
R2-FW
R3-FW
R3-easer
Average
28.1
29.0
28.5
15.0
15.0
28.7
24.0
0.64
2.07
0.61
1.04
0.59
0.18
0.86
0.32
0.52
0.36
0.12
0.05
0.16
0.26
−3.17
10.06
1.74
7.81
−0.42
−0.34
3.92
0.78
−1.36
−0.15
1.97
−2.04
0.11
1.07
2606
2606
2596
Offset (m)
<–0.9
–0.9 to –0.6
–0.6 to –0.3
–0.3 to 0.0
0.0 to 0.3
2586
Mine RL (m)
Mine RL (m)
2596
2576
(a)
2566
1906
Burden (m)
5.4
5.1
4.8
4.5
4.2
3.9
3.6
2586
2576
1916
1926
1936
Mine easting (m)
(b)
1906
1916
1926
Mine easting (m)
1936
FIGURE 6.10
(a) Hole deviation and (b) related burden calculation.
wear and higher mucking productivity, reduced secondary blasting, fewer
orepass hang-ups and less dilution from stope wall failures.
6.3.1 Drilling Orientation
Sublevel open stoping features rings of either radial or parallel blastholes.
Radial holes toe into a designed stope boundary and are usually drilled
from stope development accesses that are narrower than the planned stope
boundary. In situations where a sill is excavated across the entire orebody
width, the blastholes can be drilled parallel to the stope boundary using a
regular (burden and spacing) blast pattern. The size and shape of the drilling
drives are controlled by issues of excavation stability and overall development cost, with rock mass conditions sometimes precluding the use of full
orebody undercuts.
Drilling parallel to a planned stope boundary provides greater control
of the breakage plane and aids in minimizing blast damage (Figure 6.11).
Explosive types, loading densities, and standoff distances of the perimeter
261
Drilling and Blasting
0.8 m
11
.8 m
2m
2m
2m
1m
11.
m
9.7 m
°
10.4
9.1m
73.5°
2m
2m
16A sublevel
60
.8°
°
63.
4°
66.3
69.7
0.8 m
16 level
0.5 m max.
FIGURE 6.11
Typical bench stope drilling and charging pattern. (Courtesy of Mount Isa Mines, Mount Isa,
Queensland, Australia.)
holes can be modified to minimize overbreak. Where full-width sills or ore
drives exist, drillhole deviation of breakthrough blastholes can be established by visual inspections. Redrilling can take place while the equipment
is still within the stope, saving significant time and drilling resources. A disadvantage is the increased development and ground support cost within a
stripped-out drilling drive. In cases where the stope access is in excess of the
stable span, instability problems may be encountered during stope blasting
even if additional deep reinforcement (cablebolting) is implemented.
In cases where the holes toe into a designed stope boundary, the line of
breakage is defined by the positions of the blasthole toes. Such radial patterns
of drillholes are usually drilled from relatively small excavation accesses.
This decreases the development and ground support costs while enhancing
stability of the drilling access during the stoping operations. A disadvantage is that drilling and blasting become more difficult, because a uniform
explosive distribution may not be achieved from a fanned blasthole pattern.
Depending upon the rock mass strength, holes that terminate at a stope wall
may create a “saw-toothed” profile that could prove to be unstable or lead
to ore loss due to restrictions on broken ore rill. Limited control of the toe
location due to hole deviation or hole length from overdrilling may lead to
confined charges causing damage. This is particularly true in a stope hangingwall where a single hole may be sufficient to cause failure. Holes toeing
into a footwall may cause an uneven surface affecting the rill of broken ore at
that stope boundary (Figure 6.12).
262
Geotechnical Design for Sublevel Open Stoping
1650 E
3200 mRL
1600 E
6/L
7/L
3250 mRL
8/L
FIGURE 6.12
Blasthole toeing into a stope footwall. (Courtesy of Mount Isa Mines, Mount Isa, Queensland,
Australia.)
In cases where wide orebodies are extracted, a compromise on safe and
economical access (while providing some flexibility for drill and blast control) can be achieved. Separate drilling drives parallel to a stope outline
can be established to drill breakthrough holes at the stope boundaries
(Figure 6.13). Furthermore, the explosive charging for a radial pattern can
be engineered to achieve a more or less even distribution of explosives for
each ring.
6.3.2 Blasthole Diameter
The nominal blasthole diameter (d) is defined as the diameter of a new bit
of the specified size. It is one of the most important factors in the design, since
the majority of the other blasting parameters are geometrically related to the
hole diameter. The blasthole diameters used in open stoping may range from
51 to 200 mm depending on the geometry of the stope (stope dimensions and
maximum hole length), the rock mass conditions, and the drilling equipment
263
1500 E
Drilling and Blasting
2850 mRL
13/L
2800 mRL
14/L
FIGURE 6.13
Parallel drilling to a stope boundary within a large tabular orebody. (Courtesy of Mount Isa
Mines, Mount Isa, Queensland, Australia.)
available. Large-diameter blastholes actually decrease the specific drilling cost
(dollars per cubic meter of rock blasted), while improving drilling accuracy.
A disadvantage of large-diameter blastholes is the potential to create
greater damage to the surrounding rock due to an increased explosive concentration. In addition, the uniformity of the resulting rock fragmentation
may be affected due to high powder factors which are poorly distributed
throughout the blasted volume. Another disadvantage is the inherent difficulty of the conventional charging of explosives into large-diameter upholes.
In general, similar fragmentation can be achieved with different blasthole
sizes in a homogeneous rock mass, provided the blasting pattern (burden,
spacing, uncharged length, etc.) is adjusted to suit the local geotechnical
conditions. The resulting powder factor (explosive distribution) in different
parts of a stope should be analyzed to avoid excessive or inadequate charge
concentrations, especially in high aspect ratio drilling regions. Nevertheless,
damage beyond a designed stope outline will necessarily increase with
larger blasthole sizes unless hole standoffs or blasthole lengths are suitably
adjusted. In addition, consideration of the actual rock mass conditions is
required since broken ground may preclude the use of smaller holes due to
blasthole closure or excessive shearing, causing hole losses.
264
Geotechnical Design for Sublevel Open Stoping
Sublevel intervals are also related to the blasthole diameter, since a compromise must be reached between the capabilities of the drilling equipment
to minimize drillhole deviation and the cost of sublevel access development. Excessive drillhole deviation may be experienced with a reduced hole
diameter in cases where the blasthole length exceeds drilling equipment
capabilities.
In theory, ring designs are based on nominal bit sizes and do not allow for
bit wear. In practice, however, new drill bits start at a slightly larger diameter
than the nominal hole size and are discarded when a minimum bit size is
reached. This means that the actual hole diameter in the critical toe area may
be considerably smaller than at the collar due to gauge loss from bit wear
while drilling a long blasthole. This can be significant in longholes drilled in
highly abrasive rock masses, such as those having high silica contents. As an
example, a 40 m-longhole drilled using a nominal 140 mm bit could actually
be collared with a 136 mm bit and finished at 133 mm, representing a reduction in powder factor of approximately 10% at the most critical area of the
holes (the toes). Similarly, nominal 70 mm-diameter button bits are discarded
(after several resharpenings) at 64 mm diameter.
As the toes of ring charges are critical to achieve good fragmentation
and final stope shape, the potential impact of hole size reduction on the
actual powder factor distribution needs consideration. The detrimental
effect on the blasthole toe powder factor could be compounded by drill
deviation leading to excessive ring burdens or toe spacing. However, in
some cases, the toes of wet downholes are loaded with water-resistant
cartridge explosives which are more powerful than ammonium nitrate/
fuel oil (ANFO), and actually compensate for a reduced hole diameter at
the blasthole toes. The placement of the booster at the toe of the blasthole
also increases the explosive energy.
6.3.3 Blasthole Length
A number of empirical rules are available to choose the most appropriate blasthole length as a function of the blasthole diameter. The recommendations are
usually based on studies of hole deviation, with the suggested lengths aimed
at minimizing the probability of overlap at the toes of the holes. The greatest
impact of blasthole deviation is at the toes, where problems such as “excessive
ground to pull” (due to large distances between holes) and out-of-sequence
detonation and sympathetic detonation (due to hole overlap) can occur.
Blasthole length (L) is a function of the hole size and the drilling technology used. Table 6.3 shows a typical range of hole lengths for different
drilling technologies, selected to minimize hole deviation. They represent a
starting point and the results should be evaluated against local experience.
In some cases, the width of the orebody also plays a role in determining
the hole diameter, as increased blast damage may be expected with blasting
large-diameter holes in heavily confined narrow orebodies.
265
Drilling and Blasting
TABLE 6.3
Suggested Drillhole Lengths for Downholes in Sublevel Open Stoping
Hole Diameter
(mm)
Burden
(m)
Stand-Off
Distance (m)
Drilling
Technology
Hole
Depth (m)
51
64
73
76
89
102
115
140
1.0–1.5
1.3–1.8
2.0–2.5
2.0–2.5
2.5–2.8
3.0
3.0–3.5
3.5–4.0
0.4
0.6
0.8
1.0
1.1
1.2
1.3
1.5
Rods
Rods
Rods + stabilizers
Rods + tubes
Tubes–top hammer
Tubes–top hammer
In-the-hole hammer
In-the-hole hammer
10–15
10–15
12–20
20–25
25–35
25–40
40–50
40–60
A limiting factor on hole length is cleaning of the blastholes prior to hole
charging with explosives. Operators must prepare blastholes for charging by
cleaning out the water and drill cuttings contained at the toe of nonbreakthrough holes. Consequently, some difficulties may be experienced in largediameter holes (say 140 mm) longer than 45 m. A solution is to charge the
bottom 15%–20% of long downholes with a more powerful explosive (such
as bulk emulsion) to ensure effective explosive density at the toes of the holes
where some explosive contamination may occur.
Upholes in open stoping are usually drilled using 70–115 mm-­diameter
blastholes. Experience shows that conventional charging restricts the
lengths and diameters of the blastholes that can be pneumatically loaded
with ANFO; a 25 m blasthole length is the maximum limit for efficient
charging.
6.3.4 Burden
Burden (V) is defined as the distance between an explosive charge and a
free face or the nominal distance between production rings. Ring burden
is typically determined from the blasthole diameter and is one the most
important blast design parameters in sublevel stoping. Burden and its
related toe spacing are critical to the resulting fragmentation, damage, and
drilling cost.
Empirical blasthole diameter–burden relationships are commonly used
for blasting operations in sublevel open stoping (Rustan, 1990; Heilig, 1999).
These relationships are based on fully coupled, high strength explosives
and define the burden V (in meters) as a function of the hole diameter d
(in meters), as in Figure 6.14. Values based on these rules of thumb can be
used as a first approximation and are consequently fine-tuned based on
actual observations of drill and blast performance. A methodology relying
on Langefors uniformity law can be used to determine burden dimensions
based on different explosive products for a given rock type.
266
Geotechnical Design for Sublevel Open Stoping
46
d
15
bu
rd
Burden, V (m)
en
=
10
eS
we
dis
hr
oc
k)
Rustan, surface mines
Burden (V) = 18.8d 0.689
La
ng
efo
rs
(av
era
g
5
0
0.0
Rustan, underground mines
Burden (V) = 11.8d 0.630
0.1
0.2
0.3
Blasthole diameter, d (m)
0.4
FIGURE 6.14
Burden as a function of blasthole diameter. (After Rustan, A., Burden, spacing and borehole
diameter at rock blasting, Proceedings of the Third International Symposium on Rock Fragmentation
by Blasting, Brisbane, Queensland, Australia, August 26–31, 1990, pp. 303–310, The AusIMM,
Melbourne, Victoria, Australia. With permission.)
3
q1 ÈV1 ˘
=
˙
q2 Í
ÎV2 ˚
(6.11)
where
q is the charging density (kg/m)
V is the burden (m)
Table 6.4 shows design burdens for pneumatically charged ANFO (density of
approximately 0.90 g/cm3) using a typical range of hole diameters for open
stope extraction.
In practice, the optimum burden will depend upon the rock mass properties and the requirements for fragmentation and control of overbreak.
Excessive burdens are likely to result in coarse fragmentation, tighter muck,
and large overbreak behind the final ring of holes. On the other hand, insufficient burdens may produce excessive muckpile throw, excessive underground air overpressure, and promote interaction between charges of
different rows. Rings designed at pillar edges can be designed at 60% of the
267
Drilling and Blasting
TABLE 6.4
Design Burden Data for a Range of Diameter Holes
Hole
Diameter,
d (mm)
57
64
70
115
140
165
Explosive
Density,
ρ (g/cm3)
Charging
Density,
q (kg/m)
Velocity of
Detonation
(km/s)
Design
Burden,
V (m)
Maximum
Burden,
Vmax (m)
0.90
0.90
0.90
0.80
0.80
0.80
2.31
2.86
3.45
8.21
12.26
17.13
3.2
3.3
3.3
3.5
3.7
3.8
2.10
2.25
2.40
3.20
3.70
4.10
2.60
2.90
3.20
5.20
6.30
7.50
design burdens in order to achieve clean walls and minimize backbreak. In
general, design burden may be adjusted by ±10% to suit stope dimensions
without negatively influencing blast performance.
6.3.5 Spacing
Hole (or toe) spacing (ε) is defined as the distance between blastholes in
the same ring. The toe spacing within a ring of blastholes is related to the
designed blasthole burden, the orebody geometry, and the capabilities of
the drilling equipment. Toe spacing values are typically larger than the burden values to ensure rock breakage toward a free face, rather than shearing
across adjacent holes. Toe spacing can be varied to suit stope dimensions
and to allow staggering of holes between adjacent rows. The typical range of
blasthole toe spacings (ε) as a function of the burden V is given by
1.15 V < e < 2.0 V
(6.12)
The nominal value for the toe spacing of parallel holes is usually 1.5 times the
burden. However, the actual value depends upon the type of drilling used.
Spacing values in the upper range are used where interlocking toes from radial
fans of blastholes are designed. Toe spacing in the lower range can be used for
rings of parallel blastholes (Heilig, 1999). Large toe spacings for radial drillholes
are likely to cause insufficient breakage and localized rock mass damage.
The maximum toe spacing should not exceed twice the design ring burden. The maximum toe spacing relationship is applicable at the toes of
long, subparallel holes such that a compromise is reached between excessive spacing at the toes and wasted drilling and unnecessary low spacing
between adjacent holes over much of their length. As shown in Figure 6.15,
a margin (M) is usually left between adjacent noninterlocking blastholes to
minimize intersecting holes due to hole deviation as well as sympathetic
detonation between adjacent holes.
268
Geotechnical Design for Sublevel Open Stoping
c
M
ε
ε
M
ε
ε
(a)
(b)
ε: Toe spacing
M: Margin for noninterlocking holes
C: Uncharged collar
FIGURE 6.15
Design layout for (a) radial and (b) parallel ring blasting in open stoping. (Courtesy of Mount
Isa Mines, Mount Isa, Queensland, Australia.)
As described earlier, a reduction in toe spacing is likely to increase the
amount of uncharged drillhole length within a single ring of blastholes. This
increased cost must be balanced with the production cost benefit achieved
through more uniform fragmentation from a better explosive distribution.
Table 6.5 lists some typical ring blast patterns for open stoping in hard rock.
6.3.6 Stemming and Uncharged Length
Stemming is an inert material that is placed in a blasthole to contain the
explosive gas energy. Stemming can also act as an inert decking material
between two charges in the same hole so that they can detonate independently. Stemming near the hole collar helps to contain the detonating gases
from the explosion, thereby achieving greater fragmentation. Stemming
lengths in holes that are used for a single firing are sometimes equal to the
burden, while in holes that are blasted repeatedly (such as in vertical crater
retreat [VCR] and raising), the stemming length is reduced to facilitate cleaning before blasting the next lift (Heilig, 1999).
A more common practice for the design of suitable stemming lengths is
to relate the stemming column height to the blasthole diameter, taking into
account experience from blast monitoring and postblast observations of overbreak at the top of the explosive charge. Heilig (1999) suggested that the stemming lengths for individual blastholes and for holes that are to be fired again
be set to 20 and 10 times the hole diameter, respectively. In addition, the length
of stemming material between decks within the same blasthole was suggested
269
Drilling and Blasting
TABLE 6.5
Open Stoping Blasthole Diameters, Ring Burdens, and Spacings
Stope Area
Hole Diameter,
d (mm)
Cutoff slot
Trough undercut
(TUC)
Primary stope
ring blasting
Tertiary stope
ring blasting
Rock
diaphragms
a
b
c
Burden,
V (m)
Spacing,
ε (m)
Explosive
Type
140
70
3
2.2
3.5
3
89
3.5
4
70
2.4
3.0–3.5
ANFO
ENERGANa/
ANFO
ENERGAN/
ANFO
ANFOb
89
115
140
140
2.7–3.0
3.1–3.5
3.3–3.8
3.7
3.4–4.7
5.5–6.0
6.0–7.0
7.0
ANFO
ANFO
ANFO
ANFO
140
3.0
4.5–5.0
Low strength
ANFO
Comments
Parallel holes
TUC at 70°
Shaped to 50°
Include short
upholes
Used with
discretion
ISANOL50c
ENERGAN: Blow-loaded = 1.08 g/cm3; pour-loaded = 0.93 g/cm3.
ANFO: Blow-loaded = 0.95 g/cm3; pour-loaded = 0.80 g/cm3.
ISANOL50: Blow-loaded = 0.63 g/cm3; pour-loaded = 0.45 g/cm3.
as 20 times the blasthole diameter to reduce the risk of sympathetic detonation and charge dislocation. In wet hole charging applications, the decking
between charges should be increased. A good-quality crushed/screened stone
stemming material is recommended with a particle size of approximately
1/10th the blasthole diameter to ensure adequate confinement.
In cases where stemming is not used, blasting practices require that a
portion of the hole remains uncharged around the collar where the spacing
between adjacent converging holes becomes less than 1.5 burdens. This is to
reduce overcharging around the collar region and to minimize overbreak
and possible loss of adjacent hole collars. Additionally, inadequate uncharged
collars can lead to excessive overpressure in underground workings and
damage to ground support above the blast. The minimum uncharged length
(C in meters) can also be related to the blasthole diameter (d) as follows:
C = (18 to 20 )d (m )
(6.13)
6.4 Ring Design
A ring design uses mathematical relationships and the stope dimensions
and blast design parameters to locate blastholes within a stope outline and
270
Geotechnical Design for Sublevel Open Stoping
ensure an adequate distribution of explosives (Onederra and Chitombo,
2007). Explosive placement is determined to avoid zones of excessive or low
energy concentrations to facilitate acceptable breakage and rock displacement. A ring design document is part of the stope design process and is
issued following the completion of a survey of the drilling access development and prior to the drilling equipment moving to the stope.
The general information required for a ring design includes the stope outline and extent, geological data, the orebody boundaries, survey pickup data
to position the drill rig in the ring design (drill pivot points), and information on any adjacent excavation or fill mass boundaries. Information on the
specific drilling rigs to be used and blasting parameters such as toe spacing
and burden are also needed.
Modern ring design algorithms are used within three-dimensional mineplanning packages to facilitate data input, storage, retrieval, calculations,
and analysis. The latest survey and geological wireframes must be updated
in accordance with the chosen stope development prior to starting the ring
design work.
6.4.1 General Procedure
The basic objective of a modern, computerized ring design layout is to provide scaled drawings of the drilling plans, showing the locations of the blastholes in relation to the drill drives, the interpreted orebody and the stope
outlines. Equipment constraints, such as the drillable dip and dump angles,
hole lengths, and drilling accuracy must also be considered. One constraint on
the ring designer, however, is the availability and location of drilling accesses.
Sometimes, it is not possible to drill parallel to critical walls such as the stope
hangingwalls due to a lack of necessary development. Designed blasthole
angles are also constrained by the geological boundaries, the dip and dump
capabilities of the drill rig, the required offset of the rig from a designed stope
wall due to rig dimensions and the position of the drill steel carousel, the
explosives used, and the method of charging. The angle of a stope footwall
should exceed 45°–50°, so that the broken ore is able to rill. To achieve this,
it may be required to drill outside an orebody boundary and dictate the final
stope perimeter with charging controls, which may lead to dilution or ore loss.
In general, a ring design layout consists of sections at each ring (usually
looking toward the COS, scale 1:250) through the orebody showing the following details (Figure 6.16):
•
•
•
•
•
The stope name, level, drilling horizon, and breakthrough name
The ring number and a reference mark (e.g., Easting and RL)
The designed volume and shape of the rock to be blasted
The position and shape of the drill and mucking drives
The hole diameter, length, and orientation of the holes to be drilled
271
4300 N
Drilling and Blasting
11. 9
8
3
0°
3
3
3
8
13.0 4°
2
9
3
18.0
0°
14
44
12
40
12
7
6.
14
1
°
6
10
9
27
°
20.
2
23
.3
.9
75°
28
59°
24
22
24
°
50
.8
59
°
18
23
2
75°
24
5
4
67°
6
83°
Meters drilled
Meters changed
Tonnes broken
Tonnes/m drilled
kgs ANFO/ring
kgs ANFO/tonne
kgs ANFO/m drilled
485.2
290.9
14037.3
28.9
3576
0.25
7.40
7
Detonation sequence
(double primed)
4
Security primer
.2
36
82°
4
5
3
67°
34.2
3
Q430 M/R 4 ANFO
140 mm holes
8
7
4
17D sublevel
18.5
14°
°
7
3
41.9
39.0
40.4
2
39.5
3
38.7
40.0
5
1
2
3
4
FIGURE 6.16
Typical ring design—section view. (Courtesy of Mount Isa Mines, Mount Isa, Queensland,
Australia.)
•
•
•
•
•
The collar positions or drill rig pivot positions in the drill drive
Expected depth of breakthrough (if any)
The amount of explosive used in each hole
The length of any uncharged collars
The angle of inclination if the ring section is not vertical
Additional information provided with the ring design includes
•
•
•
•
•
•
Tonnes in the ring
Meters drilled and meters charged
Tonnes/meter drilled
Weight (kg) of explosive/ring
Weight (kg) of explosive/tonne
Weight (kg) of explosive/meter drilled
A typical floor plan (usually at a scale of 1:250 or 1:500) shows the active stope
as well as the status (extracted, current, or scheduled) of any adjacent stopes.
All vertical and horizontal development falling within 20 m of the planned
272
Geotechnical Design for Sublevel Open Stoping
H718 filled
N
MR7
MR6
MR5
MR4
MR3
MR2
H713 HW DR
7150 N
H713 CO
36
37
35
33
32
34
30
31
27
24
1
25
2
26
29
28
4
3
5
7
6
8
10 13 15 19 22
9 12 16 18
MR1
21
11 14 17 20 22
MR8
MR9
7115 XC
H710 planned
1500 E
7100 N
FIGURE 6.17
Plan view of typical floor plan showing development required. (Courtesy of Mount Isa Mines,
Mount Isa, Queensland, Australia.)
stope outline is also indicated. A plan of the drill drive showing the following details is provided (Figure 6.17):
• The stope name, drilling horizon, or level name including a north
arrow
• The ring numbers and position of the blasthole rings along the drive
as a solid line
• The easer section lines as dotted lines
• The shape and contours of the drill drive
•
•
•
•
•
•
The ore outline at the collar elevation
The position of the raise with respect to the cutoff slot
The position of the COS with respect to the blasthole rings
The burden of the blasthole rings
The burden of the holes in the COS
The position of any vertical openings
Drilling and Blasting
273
In addition to the ring section and plan views, a cover note that includes the
following information must be issued with every ring design:
• The stope name and drilling horizon
• The hole diameter and burdens for main rings and easer offsets
from main rings
• Dimensions of the raise or longhole winze (LHW)
• A table of tonnes per ring showing the actual ring tonnes, the cumulative tonnes, and the tonnes remaining after each ring blasting
• The explosive types and where they are to be loaded
• A table showing meters to drill, explosive types, and quantities per
each ring
• The actual firing sequence within the stope
During ring drilling, an accurate record of hole lengths must be kept, indicating where redrilling has taken place. This is especially important in secondary stopes to minimize dilution from adjacent fill masses. If a fill mass is
intersected by a blasthole prior to the designed length, the contact distance
must be recorded in order to modify the firing sequence and actual charged
lengths. When scheduling drilling, it is best to start working at the top level
of the stope. Drilling should take place from the top of the stope and progress down to the lower sublevels to minimize water, cuttings, and sludge
from the upper sublevels disrupting the holes below. Drilling of nonbreakthrough holes requires that the collars be blocked to prevent sludge and drill
cuttings from other holes going down the holes.
6.4.2 Parallel Patterns
Charging of parallel blastholes is a relatively straightforward procedure in
which all the holes within a ring contain a specified column length to achieve
a required stope shape (Figure 6.18). The explosive column length in each hole
is not determined by the interaction of adjacent blastholes within the ring as
with fan pattern charging, but depends upon the orebody geometry and the
required void shape after blasting. All the blastholes are individually charged
to comply with specific charge weight per delay requirements or, alternatively, decking and stemming may be introduced accordingly (Heilig, 1999).
A significant amount of experience is available in the blasting of parallel holes, since numerous explosive research programs have been carried
out with parallel holes in open pits (Andrieux et al., 1994). An advantage of
parallel holes is that an even distribution of explosives throughout a ring
plane can be achieved. A limitation is the requirement for full orebody overcut and undercut, thereby limiting the sizes of stopes that can be blasted.
Usually parallel holes are typical of bench stope blasting operations where
the widths of the orebodies are limited to typically less than 12–15 m.
274
Geotechnical Design for Sublevel Open Stoping
–75°
BT
2
3
1 Detonation
sequence
(a)
Footwall
°
–71
m
18.2
BT
2
23.5 m
m
18.9
–66
°
1
Hangingwall
3
1
Stope
void
(b)
FIGURE 6.18
(a) Cross section and (b) plan view of narrow bench stope drilling.
In cases where parallel patterns and downhole drilling are used, a row
of parallel holes is drilled to breakthrough, thus allowing the blastholes to
drain prior to blasting. The burden and spacing can be large in wide orebodies, as the blastholes are not as constrained as in narrow vein stoping. The
extent of drill deviation increases as the stope height increases; nevertheless,
the ability to check breakthrough hole toe locations allows the hole positions
to be located and redrilling can occur as required.
6.4.3 Radial Patterns
Production blasting in sublevel stoping sometimes requires the drilling
of holes in a radial pattern. Radial drilling layouts and charging criteria
can be significantly more complex than those encountered in parallel drilling. For instance, radial drilling patterns make it very difficult to get a
completely even distribution within a drilled ring. In fact, the aim of ring
design is to ensure the most even distribution possible, while achieving
minimum drilled meters and maximum fragmentation. Alternate rings are
usually drilled on a staggered pattern to distribute the explosives more
effectively and to minimize the effects of misfired holes. A preferred,
275
Drilling and Blasting
Ring 1, 3, etc.
Ring 2, 4, etc.
(a)
(b)
FIGURE 6.19
Staggered radial ring layouts. (a) Good design and (b) poor design. (Courtesy of Mount Isa
Mines, Mount Isa, Queensland, Australia.)
simplified staggering design (Figure 6.19a) has an even distribution of
explosives allowing the firing more than one ring at a time.
Figure 6.20 shows that interlocking blasthole toes actually overlap by
2–3 m to minimize the adverse effect of holes that may be “short” due to
the adverse effect of drilling sludge or excessive toe burden due to hole
deviation. In addition, hole interlocking is important when firing a stope by
sublevel rather than the full stope height, as it allows the toes of downholes
to be cleared prior to mass blasting.
Individual blastholes are placed within a particular drill region using
simple geometrical rules and the distance between the ends of adjacent holes
within a ring is defined by toe spacing rules. A normal spacing (having a
length equal to the toe spacing, ε) to the median bisector at the drill region
boundary can be used to space the holes (Figure 6.21).
Having defined a drill region, a number of control (critical) holes are placed
on the boundaries or in the corners. The control holes are placed from the
drilling positions, and the remaining holes in the ring are placed between the
control holes using a specified toe spacing rule (Figure 6.22).
The procedure starts at a control hole and works clockwise placing holes
to satisfy a specific toe spacing rule. A number of holes may be placed before
another control hole is encountered. If there is not enough room for the
last hole on the sequence, the holes already placed are redistributed evenly
within the ring so that another blasthole can be inserted. Design rules are
required to avoid excessive increases or decreases in the toe spacing within a
ring to accommodate extra blastholes. If the calculated spacing for the blasthole is within 0.8–1.20 times the design toe spacing, the calculated spacing is
left unchanged (i.e., lower and upper limits of 0.8ε and 1.2ε, respectively, are
276
Geotechnical Design for Sublevel Open Stoping
Common
tonnage
region
2m
Common tonnage
region
Drill drive A
Drill drive B
Drill region
A
Drill region
B
2m
FIGURE 6.20
Overlap of hole toes in ring blasting. (After Rosengren, M. and Jones, S., How can we improve
fragmentation in the copper mine? Unpublished Mount Isa Mines Limited Internal Report, 1992.
With permission.)
accepted). Otherwise, an overall correction including an additional hole is
required as described previously.
In regularly shaped stope outlines, alternate rings of radial blastholes are usually staggered to allow the explosive to be distributed more uniformly throughout the stope volume. This also provides some insurance in the case that a single
hole is lost in a preceding ring. Staggering is achieved by adding an additional
hole to every other ring, and closing the toe spacing on that ring. Staggering
works well when there is a reasonable number of blastholes between the control holes. The general practice is that by the third hole in the ring, the stagger
should be on the median, with the stagger being reduced toward the control
holes on the boundary of the drilling region (Figure 6.23). Staggering may not
be effective in cases where the drilling region is small due to the increase in the
powder factor achieved by adding one hole in every other ring.
277
Drilling and Blasting
ε
Median
bisector
Drill region
FIGURE 6.21
Toe spacing rule to place holes within a ring. (After Rosengren, M. and Jones, S., How can we
improve fragmentation in the copper mine? Unpublished Mount Isa Mines Limited Internal
Report, 1992. With permission.)
ε/2
ε
ε/2
ε
ε
Control holes
minimum collar charge
FIGURE 6.22
Charging holes within a ring. (After Rosengren, M. and Jones, S. How can we improve fragmentation in the copper mine? Unpublished Mount Isa Mines Limited Internal Report, 1992.
With permission.)
6.4.4 Vertical Crater Retreat Blasting
VCR blasting is based upon the use of near-spherical charges, where the
explosive length to diameter ratio does not exceed 6:1 (Livingston, 1956). The
ideal explosives for hard rock blasting are watergels and emulsions. ANFO
has very limited applications and is used only in soft rocks (Lopez Jimeno
et al., 1995). The charges are placed in positions within the blastholes (usually drilled normal to a free face) so that they are located at an optimum
distance from an advancing stope back. Once a charge detonates, an inverted
crater is produced around the blasthole collars. In practice, a number of
278
Geotechnical Design for Sublevel Open Stoping
Fa
Drilling
region
ult
e
tur
uc
str
Alternate
ring
On median
FIGURE 6.23
Staggering of holes on the median bisector. (After Rosengren, M. and Jones, S., How can we
improve fragmentation in the copper mine? Unpublished Mount Isa Mines Limited Internal
Report, 1992. With permission.)
holes are detonated such that the blasted craters overlap at the free face.
Generally, a pattern of holes is blasted in sequence, resulting in horizontal
lifts being removed from the stope backs. The extraction progresses upward
until the full stope is extracted.
The correct depth of burial to achieve the greatest crater volume is usually determined from test shots on site, where the rock and explosive type
are the same as in the production blasts. The depth of burial is the measured distance from a free face to the explosive charge center of mass. The
optimum depth of burial is determined by drilling a series of test holes of
the proposed VCR diameter with the test depths increased sequentially.
The test blastholes should be drilled to various depths from 0.75 to 4 m in
steps of 0.25 m. A spherical charge is then fired in each test hole and the crater volumes are measured and compared to determine the optimum depth
of burial. A graph of charge depth versus crater volume, with the average
crater volume being calculated from a series of cross sections, is established
(Figure 6.24).
The depth of burial for VCR blasting can be established from crater blasting
theory (Lopez Jimeno et al., 1995). The critical depth Dc, at which the first
signs of external damage in the form of cracks and fractures are noted, can
be given by
D c = Et Q1/3
(6.14)
where
Et is the strain energy factor (a characteristic constant in each rock–
explosive combination)
Q is the weight of the explosive in kilograms
279
Drilling and Blasting
0.8
0.7
Crater volume (m3)
0.6
0.5
0.4
0.3
0.2
0.1
0
0
0.20
0.40
0.60
0.80 1.00 1.20
Charge depth (m)
1.40
1.60
1.80
2.00
FIGURE 6.24
Experimental determination of crater volumes. (Courtesy of Mount Isa Mines, Mount Isa,
Queensland, Australia.)
Equation 6.14 can be re­written as
D g = DEt Q1/3
(6.15)
where
Dg is the depth burial measured from the surface to the center of gravity of
the explosive charge
∆ is a dimensionless parameter equal to Dg/Dc
The optimum depth ratio ∆o at which the explosive maximizes the crater
volume is
Do =
Do
Dc
(6.16)
where Do is the optimum depth of burial. In practice, the crater radius can be
estimated as the depth of burial for the optimum depth case. The hole spacing is usually taken as 1.25 times the crater radius.
Blasthole diameters ranging from 150 to 165 mm are commonly used in
VCR blasting. The larger blasthole diameters increase the drilling accuracy while allowing for a large explosive mass. This in turn increases the
optimum depth of burial and the volume of displaced rock. The depth of
charge placement and proper stemming are very important to maintain an
even back during stope advance. The maximum length of top stemming
280
Geotechnical Design for Sublevel Open Stoping
should be equal to or greater than the bottom stemming. Usually, a top
stemming length equal to 12 times the diameter of the blasthole is used
(Lopez Jimeno et al., 1995). Interlocking angular crushed stone is recommended for bottom stemming, while sand is preferred for top stemming.
Sand does not lock or cement together above the blasted charge and it can
be washed or blown out to clear the hole for preparation of future firings.
Blasthole patterns must be drilled and charged to ensure that craters intersect and remove a horizontal slab of rock. In cases where the blastholes are
drilled too far apart, or where an excessive depth of burial is used, humps of
unbroken rock will remain, causing major problems as stoping proceeds. If
required, the depth of burial must be reduced in some blastholes to even up
the stope back. Each blasthole should be examined before charging to determine breakthrough or blocked depths.
In the case of blocked holes, the options are to redrill, excluding the hole
from the firing, or to charge above the blockage by estimating the location of
the free face using information from adjacent holes. The initiation sequence
to maximize fragmentation must consider the longest charges or those
lowest in the stope horizon, such that two free faces are provided for each
charge (Lopez Jimeno et al., 1995). Ideally, the charges near the stope walls
are blasted last, so that breakage is toward previously blasted craters, thus
minimizing stope wall damage.
A number of operational and safety problems unique to VCR may be experienced during production blasting. In some cases, irregular rather than
flat free faces are created by insufficient depth of pull or by the inability to
charge some of the holes. Blastholes may shift with ground movement due
to sequential blasting, thus preventing charging. Also, plugging of expanded
(or belled out) blastholes may not be possible at all. In addition, the depth of
pull is largely controlled by the presence of geological discontinuities and
large failures of the stope back are possible. Similarly, rock masses subjected
to excessive stresses may experience uncontrollable cave in at the free faces.
The procedures for examination and handling of misfires are complicated by
the depth of occurrence down the blasthole and recovery of the final crown
of the stope (immediately underneath the charging level) may be complicated if the crown gets too thin to operate safely.
6.5 Explosive Selection
In recent years, the number of explosive formulations available for underground use has increased significantly with the introduction of chemically
sensitized emulsions and customized explosive blends used in both development and stope production blasting. Selection of the explosive type used in
stope blasting can depend on many factors including water conditions, rock
Drilling and Blasting
281
mass properties, blasthole lengths, diameters and orientations, the desired
fragmentation, available blasthole charging equipment and personnel, and
limitations on blast-induced vibrations. These and other factors are considered in deciding on the most suitable explosive formulation, strength, and
delivery system for the particular application.
6.5.1 Packaged versus Bulk Explosives
For stope production blasting applications, two systems of explosives
packaging or delivery exist: packaged formulations and bulk in-hole
delivery. The term “packaged explosive” can refer to either individual
explosive cartridges of high-strength or high-sensitivity emulsion or
watergel formulations (ranging from approximately 100 g to 5 kg in size)
or reduced-quantity packaging of typical bulk explosives such as ANFO
(e.g., 20 kg bags). Bulk in-hole delivery of explosives generally refers to
mechanized explosive loading using pneumatic, electric, or hydraulically
operated pumps, augers, or air pressure.
Due to explosive storage concerns and constraints on transport and delivery to underground magazines, an intermediate system exists in which
large-quantity bulk-bags (for ANFO) or small bins (for emulsions) are used.
These containers can range in size from 100 kg to 1 tonne and allow for the
unassisted delivery of the product to the blastholes by charge-up personnel
with minimal transport and handling of a large quantity of individual bags
or cases.
The decision as to which delivery system is suitable for each application
depends on the type and quantity of explosive to be used, the operating
environment, and the training and experience of the charge-up personnel. In
modern open stope blasting with hole diameters ranging from 64 to 127 mm,
intermediate or bulk explosive delivery is preferred due to the high rate of
explosive delivery and ease of charging. Contract blasthole loading is sometimes offered as a service by explosives manufacturers using pumping technologies from mobile explosive-manufacturing units or bulk auger trucks.
6.5.2 Ammonium Nitrate-Based Explosives
A vast majority of modern commercial explosives used in mining applications are composed of an ammonium nitrate (AN) oxidizer, a fuel component, and a sensitizing agent. The oxidizer is generally comprised of either
a dry AN prill of specific size and characteristics or an AN solution. The
fuel component can consist of any organic carbon-based material, although
fuel oil or other organic oils are preferred due to material handling, explosive detonation efficiency, and ease of mixing. Sensitizing agents can include
physical voids in the explosive formulation through prill porosity or voids
in prill packing, small voids created by glass microballoons or gas bubbles,
or various chemicals. Additional chemicals or additives (such as propellants
282
Geotechnical Design for Sublevel Open Stoping
or aluminum powders) can further modify the detonation characteristics of
the explosive for customized applications. Three basic formulations of AN
explosives exist, each differing in the way the fuel, oxidizer, and sensitizing
components interact during manufacturing and detonation. The three main
categories of AN explosives are ammonium nitrate prill/fuel oil (ANFO),
watergels, and emulsions.
6.5.3 ANFO
The most widely used commercial explosive for surface or underground
blasting in a wide range of hole diameters (38 to >900 mm) is the standard dry AN prill and fuel oil mixture known as ANFO. ANFO consists
of 1.5–3 mm-diameter AN prill coated in fuel oil at an optimum mixture of
5.7% fuel by weight. The detonation efficiency of ANFO varies with the percentage of fuel, where underfueling results in a greater reduction in output
energy than overfueling. The typical loaded density of ANFO ranges from
0.82 to 0.95 g/cc based on pour loading or pneumatic loading, respectively.
Several different modified formulations of ANFO are available for use in
a wide range of applications. These include formulations having properties
of reduced density, low fume production, water resistance, buffering against
thermally active or chemically reactive ground, and high-strength breakageresistant prill for pneumatic loading. For mostly dry blasting conditions
requiring an even distribution of shock and heave energy in medium- to
large-diameter holes, ANFO is the preferred explosive choice. Due to the
popularity of ANFO as the preferred explosive choice over several decades,
the strength characteristics of other explosive formulations are regularly
listed with reference to the standard strength of ANFO. These properties
include relative bulk strength and relative weight strength.
Due to the susceptibility of ANFO to water-induced explosive degradation, blasting in wet holes or where long sleep times are required is not recommended. Additionally, pneumatic charging of ANFO in large-diameter
upholes (>89 mm) can result in excessive explosive loss due to fallout and
therefore is advised only for subhorizontal to vertical downholes or smallerdiameter upholes. The lack of water resistance and cohesion of ANFO were
two properties that prompted the development of other explosive formulations. These desired properties were drivers for the development of fluidbased bulk watergel and emulsion explosive formulations to aid in replacing
ANFO for certain applications. Of the two types of explosives, emulsions
have been developed more extensively for use in modern bulk delivery
application.
6.5.4 Watergels or Slurries
Ammonium nitrate watergels, commonly known as “slurries,” began development in the late 1950s (Du Pont, 1977), and consist of the same three
Drilling and Blasting
283
components as ANFO (oxidizer, fuel, and sensitizer). The phases and mixing
procedures of the three components differ from ANFO, leading to increased
water resistance and detonation characteristics. Slurry explosives contain
oxidizer salts, fuels, and sensitizers dispersed in a continuous liquid phase.
The addition of gelling agents or cross-linking agents retards the separation
of the three components, controls the density and viscosity of the product,
and adds water resistance to the mixture. The droplet size of the oxidizer in
a slurry explosive is in the order of 0.2 mm (Bampfield and Morrey, 1984).
The detonation characteristics of watergels are generally more efficient
than those of ANFO because of the decreased size of the particles and the
increased intimacy between the components. The method of oxidizer, fuel,
and sensitizer suspension in watergels causes poor gap sensitivity and high
sensitivity to changes in product and ground temperatures. For these reasons, watergel explosives are not used extensively in the modern mining
industry, having been largely replaced by emulsion explosives.
6.5.5 Emulsions
Emulsion explosives are similar to slurries in that the active components are
suspended in a continuous liquid phase and are therefore water-resistant
and highly pumpable. The differences between watergels and emulsions
become apparent when reviewing the mixing process of the separate phases
and the common sensitizing agents used in each type of explosive. The basic
formula of an AN emulsion explosive is the suspension of small droplets
of AN solution in a continuous oil (fuel) matrix. The droplet size of the AN
solution in the emulsion matrix is on the order of 0.001 mm or less (Bampfield
and Morrey, 1984).
Common sensitizing agents used in bulk emulsion explosives are glass
microballoons or gas bubbles formed by a chemical reaction within the
emulsion after it is delivered into the blasthole. The required charging equipment, loaded densities, desired detonation characteristics, and storage and
transportation requirements of each type of sensitized product generally
determine which is the most suitable for particular applications. Each of
these factors is closely linked to the method of product sensitization prior
to or during blasthole loading. Under current explosives regulations, unsensitized emulsion is considered to be a bulk oxidizer much like agricultural
fertilizer. Once the sensitizing agent is introduced, the emulsion becomes a
blasting agent and is therefore subjected to more stringent storage and transportation regulations.
The method of emulsion sensitization significantly influences the physical
properties and the detonation characteristics. Microballoons generally yield
an emulsion product that is more resistant to dynamic shock-induced desensitization and is better suited to close-in or highly confined blasting conditions. Microballoon-sensitized emulsions can also be sheared to change the
rheology for loading into larger-diameter upholes or where a lower-viscosity
284
Geotechnical Design for Sublevel Open Stoping
Unconfined velocity of detonation (m/s)
4500
4000
3500
“Gassed” emulsion
(0.85 g/cc)
3000
2500
2000
ANFO (0.85 g/cc)
1500
1000
500
50
60
70
80
90
100
Blasthole diameter (mm)
110
120
130
FIGURE 6.25
Comparison of ANFO and emulsion velocity of detonation.
product is required. In general, microballoon-sensitized emulsions have a
high loading density (1.2–1.35 g/cc), which is not adjustable without the manual addition of low-density additives such as polystyrene or other organic
bulking agents. When compared with ANFO, emulsions typically behave
more as “ideal” explosives, having higher velocities of detonation and lower
sensitivities to blasthole diameter (Figure 6.25). Additionally, the energy
distribution within an emulsion explosive differs dramatically from that of
ANFO, having a higher percentage of shock energy and a lower percentage
of heave energy. Due to the reduced gas production of emulsion, the overall
output energy can be less than that of ANFO even at a significantly higher
charge density.
Chemically sensitized or “gassed” emulsions are sensitized through the
production of gas bubbles due to a reaction between chemicals added to
the mixture immediately prior to or during pumping of the product into
a blasthole. The rate and degree of gassing are regulated by the amount
and injection location of the gassing agent or agents, the temperature of
the product, the hole diameter, and the length of the explosive column. Once
the emulsion is loaded in the blasthole, the chemical reaction takes place,
causing the product to increase in volume and thus reduce in density. The
desired density in the hole should be checked regularly during loading by
performing cup density checks using standard testing practices. A wide
range of in-hole product densities are available due to the easily adjustable
amount of gassing agent injected. Product density ranges from 0.8 to 1.2 g/cc
are common for chemically sensitized emulsions.
Drilling and Blasting
285
The fact that sensitization occurs upon loading into blastholes makes gassed
emulsions a preferred product to reduce storage and transportation restrictions. The presence of free-forming gas bubbles in comparison to glass microspheres also makes gassed emulsion more susceptible to desensitization under
shock conditions and largely unsuitable for highly confined blasting conditions
where product sensitivity can be a concern. Additional concerns with gassed
emulsions are the control of uncharged collar lengths due to mismanaged gassing rates or gassing agent amounts, quality control of the average charged
density, product waste, and the variable in-hole density profile due to differential gassing deep in the column from the weight of the explosive product.
6.5.6 Special ANFO and Emulsion Blends
Some customized products have been developed for specialty blasting conditions regularly experienced in underground stope blasting. These specialty
products use modified formulations of existing products such as ANFO or
emulsion to achieve specialized detonation characteristics. The most widely
used specialty products in underground blasting include buffered explosives
for resistance to thermally or chemically reactive ground and low-density
products to reduce blast-induced damage or extraneous blasting vibrations
in stope walls or outside the designed stope perimeter.
Buffered explosive products typically include a chemical agent to reduce
the sensitivity of an explosive to high temperatures or to reduce the reaction
of the explosive with sulfides in the rock mass or groundwater. Excessive
heat generated either through thermally active ground or through an exothermic chemical reaction between the explosive and the rock mass can lead
to premature detonation of blastholes or malfunctioning of initiation systems or charge boosters. Low-density ANFO or emulsion products typically
contain a low-density bulking agent such as polystyrene or other low-­density
organic materials to reduce the in-hole charged density. The reduction in
density and alteration of the detonation characteristics reduce the borehole
pressure and the associated damage around a blasthole. Charge densities
down to 0.3 g/cc are achievable in commercial low-density underground
specialty products.
6.6 Explosive Placement
Before placing explosive charges, blastholes are cleaned out using compressed air to remove any water, sludge, or drill cuttings to allow hole
depths to be accurately measured. An ANFO hose can be used to both clean
the holes and also measure the hole length. Prior to explosive charging,
286
Geotechnical Design for Sublevel Open Stoping
Nonel
Nonel
Uncharged collar
Uncharged collar
ANFO
ANFO
Primer + detonator
ANFO
Powergel
ANFO bag + powergel
(a)
Primer + detonator
ANFO bag
(b)
FIGURE 6.26
Charged blasthole geometries in open stoping.
breakthrough holes must be blocked near the toe. To block breakthrough
holes, high-energy emulsion explosive cartridges such as powergel can be
placed in a plastic bag, tied together, and dropped down the hole by a strong
chord that reaches the required depth. Additional powergel bags are then
cut open along their axes and dropped down the holes to block the breakthrough as they split on top of the initial plug (Figure 6.26a). Alternatively, an
empty ANFO bag or other material such as a ventilation bag is simply placed
at the end of the ANFO hose or a rope and lowered to position just above the
breakthrough depth (Figure 6.26b). Sticks or wedges attached to a rope or
inflatable air bags may also be used to block breakthrough holes. To confirm
the efficacy of the breakthrough blockage, the hole is checked for breathing
(air flow at the hole collar) or explosive leakage at the toe.
When used for charging downholes, ANFO products are either poured or
blow-loaded, depending upon the hole diameter. The charging density (q) for
pour-loaded ANFO is approximately 0.80–0.85 g/cm3. Small-hole diameters
are typically blow-loaded to guarantee a consistent explosive density, as
small pieces of rock may block the hole or create air pockets when pourloading. Similarly, all inclined holes, regardless of their diameter, are blowloaded. The charging density (q) for blow-loaded ANFO is approximately
0.90–0.95 g/cm3 due to prill breakage and increased compaction. Figure 6.27
shows uphole charging of ANFO within a typical longitudinal bench stope.
Blind (nonbreakthrough) wet holes are typically left until last when charging with ANFO. If the water cannot be removed using compressed air or
pumping, or if the hole will experience excessive sleep time prior to firing,
pumped or cartridge emulsion products are used instead of ANFO. In some
applications, holes are charged and not fired until adjacent filling operations
Drilling and Blasting
287
FIGURE 6.27
Bench stope blasting—blow-loading of ANFO using compressed air.
are under way. Although the sleep time of ANFO depends on the rock
­temperature, as a general rule charged holes should not be left unblasted
for over 2 weeks. Emulsion explosive products typically have a much longer
sleep time, unless there are reactive ground conditions.
6.6.1 Powder Factor
Traditionally, powder factor is an indirect measure of the explosive energy
being imparted to a rock mass per unit volume or weight blasted. It is calculated by dividing the weight of the explosives by the ring volume or the
tonnage that is expected to be broken. Because ring blasting is a dynamic
event and each rock mass is unique, the conventional definition of powder
factor has limited applications other than being an index for comparisons on
a global scale. The explosive quantity within each stope ring depends upon
the following factors:
•
•
•
•
•
•
The number of meters drilled
The blasthole diameter
The explosive type
The method of loading the holes (pour- or blow-loaded)
The number of meters charged
The tonnes broken
Typical powder factors for stope-blasting applications range from approximately 0.20 to 0.30 kg of explosive per tonne of ore for ring blasting, from
288
Geotechnical Design for Sublevel Open Stoping
0.20 to 0.50 kg/tonne for bench stoping, and from 1.4 to 1.8 kg/tonne for
cutoff blasting. Due to excessive confinement at the toes of blasthole rings,
higher localized toe powder factors are recommended to ensure adequate
breakage. This can be achieved by using high-density emulsion explosives
near the blasthole toes.
6.6.2 Energy Distribution
Conventional powder factor calculations only provide a number and do not
indicate the distribution of explosives within a ring design (Onederra and
Chitombo, 2007). This is especially critical for radial drilling, where it can be
very difficult to achieve an even distribution of explosives throughout a ring.
In addition, larger-diameter blastholes give a poor distribution of explosives
throughout the rock mass. The ring design described in Section 6.4.3 is an
attempt to achieve an even distribution of explosives, since the powder factor
is controlled by the burden and toe spacing chosen for each particular hole
diameter. Staggered charging in adjacent ring holes also attempts to evenly
distribute the explosive energy by minimizing the explosive concentration
near the blasthole collars where the hole spacing is reduced.
The conventional powder factor represents an average number over the
blasted volume and is unable to identify regions having excessive energy
such as collars near the hole. Research work at the Julius Kruschnitt Mineral
Research Centre (JKMRC), Brisbane, Australia, has developed a technique
to analyze powder factors within small cell areas, rather than a total area.
The JKMRC QFRAG dynamic powder factor calculation can be used to analyze explosive distribution within rings and to determine regions where
poor fragmentation may occur or where excessive damage is likely. The program QFRAG uses hole initiation timing to calculate the amount of breakage from each hole, and thus the energy required to achieve it. Because of
detonation scatter, several simulations are required to obtain an average of
powder factors from the analyzed design. The calculations are performed
for a plane parallel to the ring, at one burden distance away from the blastholes. Figure 6.28 shows explosive distributions from a ring of 140 mmdiameter blastholes, drilled on a 3.5 m burden and 6 and 7 m toe spacings,
respectively, and loaded with ANFO.
The conventional design powder factors for the rings shown in Figure 6.28
were 0.28 and 0.25 kg/tonne for the 6 and 7 m toe spacing rings, respectively.
The simulated initiation sequence was started from the bottom right with an
MS #4 delay, and 3 or 4 holes per delay number were fired to give an approximate maximum charge weight of 600 kg/delay. The hangingwall holes were
fired two numbers later than adjacent holes.
Additional research in the early 1990s at the JKMRC saw the development of the computer program 3 × 3WIN, which is capable of calculating a
4D powder distribution by considering the influence of initiation sequence.
The JKMRC 4D powder factor distribution tessellates points on a specified
289
Drilling and Blasting
6 m toe spacing
7 m toe spacing
Powder factor
kg/tonne
<0.167
0.167 to 0.208
0.208 to 0.250
>0.250
FIGURE 6.28
JKMRC cell powder factor distribution for two ring designs.
plane using a distance weighting calculation that includes a weighting with
respect to the time a deck detonates within the ring. Assumptions associated
with the location and detonation time of decks are considered to calculate
the 3D powder factor distribution and weighted using a factor called cooperation time. This is the time at which explosives from two decks in different
holes interact on a portion of the rock mass (AMIRA, 1993). The effects of
timing on explosive cooperation within a ring are shown in Figure 6.29.
6.7 Initiation Systems
In the modern explosives market, two main categories of initiation system
exist for underground development and production blasting applications.
These two systems are categorized by the type of delay element contained within the detonator. The two main types of modern delay element
are a controlled burning-front pyrotechnic element and an electronic
computer chip.
6.7.1 Pyrotechnic Delay Element Detonators
Pyrotechnic delay detonators are the most commonly used initiation systems in underground blasting. The well-known electric and shock-tube initiation systems contain pyrotechnic delay elements.
290
Geotechnical Design for Sublevel Open Stoping
Scale: kg/tonne
0.100 to 0.400
0.400 to 0.700
0.700 to 1.000
>1.000
(a)
(b)
FIGURE 6.29
(a) 3D powder factor distribution and (b) 4D powder factor distribution (cooperation
time 35 s) from the JKMRC program 3 × 3WIN. (From AMIRA, P93E advanced blasting
technology, Julius Kruttschnitt Mineral Research Centre Final Report (1990–1993), 1993.
With permission.)
Pyrotechnic delay elements are composed of a controlled length of a pyrotechnic material having a highly controlled burn rate between the initiation
line (downhole leg wires or shock tube) and the match head or primary
charge. The delay within the detonator is therefore controlled by a physical
burn process. The downhole timing accuracy of the detonator is controlled
by the quality control of the manufacture of the pyrotechnic delay material
and the length of the element. The timing accuracy of the entire pyrotechnic
system must also consider any variations in burn time or charge transfer
time for surface delays or connection mechanisms such as detonating cord
or shock tube for shock-tube systems, or sequential firing boards for electric
detonators.
6.7.2 Available Timing and Sources of Timing Error
for Pyrotechnic Delay Elements
The standard underground series of pyrotechnic detonators include two
basic timing configurations. These two systems are long-period delay intervals (LP, Table 6.6) or millisecond delay series (MS, Table 6.7). Extended delay
291
Drilling and Blasting
TABLE 6.6
Comparison of Dyno Nobel NONEL LP and Orica Exel LP Delay Intervals
Dyno Nobel NONEL
LP Series
Detonator
No.
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Orica Exel LP
Firing Time
(ms)
Detonator
#
Firing
Time (ms)
Detonator
#
Firing
Time (ms)
25
500
800
1100
1400
1700
2000
2300
2700
3100
3500
3900
4400
4900
5400
5900
6500
7200
8000
0
¼
½
¾
1
1¼
1½
2
2½
3
4
5
5½
6
7
8
9
10
11
0
100
200
300
400
500
600
800
1000
1200
1400
1600
1800
2000
2250
2500
3000
3500
4000
12
13
14
15
16
17
18
19
4500
5000
5500
6000
6500
7000
8000
9000
Sources: Dyno Nobel, NONEL® LP series, technical data sheet, Dyno Nobel Asia
Pacific, Brisbane, Queensland, Australia, 2007, Available at: www.dynonobel.com; Orica, Exel™ LP: non-electric, long delay detonator assembly,
technical data sheet, Orica Mining Services, Mansfield, Queensland,
Australia, 2008, Available at: www.oricaminingservices.com.
systems have also been developed to allow a greater range in delays for the
long duration complex blasts commonly experienced in sublevel open stope
mass-blasting applications. Additional delay periods between the specified
in-hole delay numbers can be achieved using hole-to-hole or ring-to-ring
delayed connector elements.
Standard accepted delay timing errors for pyrotechnic delay element systems is approximately ±2% due to differences between delay element batches,
temperature and humidity effects on shock tube and in-hole delay elements,
and nonuniform standardization for all lengths of manufactured detonator.
For short blast durations or use of long hole-to-hole delays, the probability
of out-of-sequence firing is minimal. The accuracy error does increase the
probability of out-of-sequence firing when long in-hole delays are used or
the charge-to-charge delay intervals are reduced by using MS connectors.
292
Geotechnical Design for Sublevel Open Stoping
TABLE 6.7
Interhole Delays and Detonating Times for Open Stope Blasting
Delay
No.
Time
(ms)
Interdelay
Interval
(ms)
Delay
No.
Time
(ms)
Interdelay
Interval
(ms)
3
4
5
6
7
8
8+(1)a
9
9+(1)
10
10+(1)
11
11+(1)
12
12+(1)
13
13+(1)
14
14+(1)
14+(2)
14+(3)
15
75
100
125
150
175
200
225
250
275
300
325
350
375
400
425
450
475
500
525
542
565
600
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
17
23
35
15+(1)
15+(2)a
15+(3)
16
16+(1)
16+(2)
16+(3)
17
17+(1)
17+(2)
17+(3)
18
18+(1)
18+(2)
19
19+(1)
19+(2)
19+(3)
20
20+(1)
20+(2)
20+(3)
625
642
665
700
725
742
765
800
825
842
865
950
975
992
1025
1050
1067
1092
1125
1150
1167
1192
23
17
23
35
25
17
23
35
25
17
23
85
25
17
33
25
17
25
33
25
17
25
a
Delay
No.
Time
(ms)
Interdelay
Interval
(ms)
21
21+(1)
21+(2)
21+(3)
22
22+(1)
22+(2)
22+(3)
23
23+(1)
23+(2)
23+(3)
24
24+(1)
24+(2)
24+(3)
25
25+(1)
25+(2)
25+(3)a
1225
1250
1267
1292
1400
1425
1442
1465
1675
1700
1717
1740
1950
1975
1992
2015
2275
2300
2317
2340
33
25
17
25
108
25
17
23
210
25
17
23
210
25
17
23
260
25
17
23
Standard millisecond delays (ms) and 25(1), 42(2), and 65(3) ms TLDs.
6.7.3 Electronic Delay Element Detonators
Electronic delay element detonators have been under development since
the 1980s and were launched into commercial use in the early 2000s. The
electronic delay element (microchip) in general replaces the pyrotechnic
element without significantly changing the design, dimensions, or physical properties of the detonator. The accepted error in electronic delay
detonators is typically ±0.1% with available delays from 0 to 20,000 ms in
predetermined or 1 ms intervals (e.g., Davey Bickford, 2008; Orica, 2010;
Dyno Nobel, 2011).
Previous research in open pit and underground mining has investigated
the impacts of accurate delay timing on muckpile fragmentation and mine
productivity (e.g., Tose and Baltus, 2002; Bartley and McClure, 2003; Grobler,
2003). The results of these studies largely indicate that accurate timing can
improve the uniformity of the fragmentation distribution and in many cases
Drilling and Blasting
293
FIGURE 6.30
Blasting a narrow vein uphole stope using signal tube initiation systems and detonating
chord.
decrease the mean particle size within the muckpile. Additional theories
on the useful application of millisecond-accurate electronic firing deal with
blast vibration reduction or frequency control, tailored timing for irregular
or complex blasthole patterns, and collision of stress waves to improve fragmentation in specific areas of a blast.
One standard practice for blasthole initiation in open stoping is to use
signal tube initiation system detonators placed down (or up) the hole as
shown in Figure 6.30. The signal tube initiation systems are attached to
loops of detonating cord for each ring. The detonating chord is then initiated by instantaneous electric detonators connected to a mine-wide stopeblasting circuit. When more than one ring is being detonated, each loop
of cord is linked to the next by a length of cord to provide security. Two
electric detonators are placed on the cord loop at each ring position. In
order to minimize damage from shrapnel cutting the signal tube initiation
system downlines, the electric detonators should be placed under sandbags
(Figure 6.31).
The signal tube initiation system delay detonators are initiated by a shock
wave passing through 3 mm-diameter plastic tubing, which is crimped onto
a detonator. The abrasion-resistant, flexible, and high tensile–strength plastic tube has a 1.5 mm bore that contains explosive material which transmits
a shock wave at 1.9 km/s. The shock front is capable of negotiating sharp
bends, kinks, and knots without rupturing the plastic tube. Hence, it cannot
side initiate any explosive and will minimize air blast. The reactive material
is initiated by detonating cord or electric detonators.
294
Signal tube
initiation
systems
M/R 3
M/R 2
M/R 1
Geotechnical Design for Sublevel Open Stoping
Signal tube
initiation
systems
Loop of
detonating
cord
Loop of
detonating
cord
Blastholes
Electric detonators
sandbagged
FIGURE 6.31
Typical multi-ring blasting hookup. (Courtesy of Mount Isa Mines, Mount Isa, Queensland,
Australia.)
Detonating cord is high-tensile, waterproof material, which has a core of
typically 4–10 g/m of pentaerythritol tetranitrate (PETN) enclosed in plastic tapes, natural and synthetic fibers, and an outer sleeve of plastic. The
3.9–5.1 mm-diameter cord is flexible, abrasion-resistant, and relatively insensitive to detonation due to friction, impact, and electrostatic discharge. The
cords have a large velocity of detonation (VOD) ranging from 6 to 7 km/s.
6.7.4 Priming
The conventional approach in ring design is to place two detonators and two
boosters at the bottom of each charged hole. This provides some insurance
in the event that one detonator does not initiate. Double priming is indicated
in the blasting plans by placing a circle around the delay number assigned
to each hole. In addition, for long charge lengths (exceeding 20 m), security
boosters are placed every 20 m along the charge axis. In practice, however,
the location of the boosters in a charged column is largely a function of ring
geometry and the location and orientation of large-scale geological discontinuities. Additional boosters are required in broken ground with high geological discontinuity connectivity, especially where large-scale faults may
allow water inflow into the explosive charges. Figure 6.32 shows a typical
setup where boosters are placed on both sides of faults to ensure initiation of
a charged column of explosive.
295
Drilling and Blasting
Fault
Fault
Fault
Double booster
Security booster
FIGURE 6.32
Booster location with respect to large-scale geological discontinuities. (After Rosengren, M.
and Jones, S., How can we improve fragmentation in the copper mine? Unpublished Mount Isa
Mines Limited Internal Report, 1992. With permission.)
Damage to walls in radial patterns toeing into walls may be increased by
the location of the boosters toward the end of the holes (near the excavation
boundary). Boosters provide high shock energy, which is required to initiate other explosives and, consequently, the local rock damage at that point
may be higher. If the boosters are moved along the charge axis (away from
the boundary), the local damage may be reduced. However, the explosive
column below the new booster location would reach full VOD, thus increasing the damage at the toe. In practice, boosters of holes toeing into walls are
placed 2–4 m from the bottom of the holes.
6.7.5 Sequencing and Timing
The fundamental objective of blasthole delay sequencing is to provide each
charge column with as many free faces as possible to break into. In ring firing,
this can be achieved not only by blasting holes toward a free face in the ring
burden direction, but also by providing each hole with at least one free face in
the direction of the adjacent blastholes in the same ring. Figure 6.33 shows the
concept of interhole and inter-ring delays in ring blasting for open stoping.
Interhole delays are usually kept to a minimum to optimize blasthole
interaction and enhance rock fragmentation. To achieve this, short period
296
Geotechnical Design for Sublevel Open Stoping
Inter-ring delay
Interhole delay
Inter-ring delay
FIGURE 6.33
Interhole and inter-ring delays in ring blasting. (From Langdon, C. and Duniam, P., Advances
in theory and application of non-electric initiation systems to 60 series extraction at the Mount
Lyell copper mine, in T. Golosinki, ed., Proceedings of the Sixth Underground Operators Conference,
Kalgoorlie, Western Australia, Australia, November 13–14, 1995, pp. 291–298, The AusIMM,
Melbourne, Victoria, Australia. With permission.)
delays (MS signal tube initiation systems series) are normally used. Because
of detonator scatter, ring-to-ring timing must be designed to avoid out-ofsequence firing, where holes in the second or third ring fire prior to holes in
the face row closest to the void.
A minimum delay time of 20 ms/m of burden is recommended for ringto-ring timing. However, for blasts having more than three rings detonating, the only way to eliminate the possibility of inter-ring misfires is to skip
one complete number in the MS series between two holes in two adjacent
rings. For example, a hole shadowed by a number 8 should be fired on a
number 10. Furthermore, it is important to minimize the total blast duration
within a stope firing. The longer the charged holes sit while other holes are
detonating, the greater the chance of blast malfunction due to hole dislocation, shearing, sympathetic detonation, or explosive desensitization. This is
particularly true in rock masses in which large-scale structures are present.
The standard MS signal tube initiation system series has 28 delay numbers
that can be used for open stope blasting. The practice of piggybacking using
trunk line detonators (TLD) at the hole collars can extend the range of delay
detonators to 55 numbers. TLDs are used to provide a delay between the
detonating cord and the signal tube initiation system delays placed down
the hole. For security reasons, two TLDs are used and all signal tube initiation system downlines (including security detonators) are connected to the
TLDs, which in turn are then hooked up to the detonating cord trunklines.
Three TLDs (having 25, 42, and 65 ms delays) can be introduced anywhere in
the standard MS range as shown in Table 6.7.
The use of TLDs reduces the interhole delay, thus enhancing fragmentation. However, they should not be used for inter-ring delays, as an out-ofsequence detonation may result. In addition, some TLD connections may
Drilling and Blasting
297
produce shrapnel, so they may have to be sandbagged or pushed into the
holes in uphole blasting. Furthermore, to minimize damage to the blast
hook-up (and avoid misfires), all downhole detonators should be burning
before the first hole in the blast is fired (total flame front). When all three TLD
delay numbers are used in a single blast, it is recommended to fire the initial
hole using delay number 4 (100 ms) to ensure all TLDs have fired and all
downhole delays have initiated before any fly-rock or shrapnel is produced.
In cases where blasthole rings toe into each other, neighboring holes in
the opposing rings should be fired on the same delay and each hole security
primed. This is the case when main downhole rings are combined with TUC
holes or when rings are fan-drilled from two drill drives, one on each side
of a stope. In cases where downholes toe into a horizontal hole drilled from
a sublevel below, they should be fired on the same number. All holes should
be security-primed and the boosters of the downhole charges pulled up 4 m
from the bottom of the holes.
Modern electronic detonators (Liu et al., 2002) allow a greater degree of delay
interval control by enabling the delays to be modified and programmed for
repetition if needed. The pyrotechnic component of the delays is replaced by
an electronic component that uses a miniature electronic timing circuit to ignite
the detonation charge. During detonator manufacturing, a delay sequence
number is built into each detonator. During blasting the detonators fire with
a constant delay interval between consecutive numbers and it is possible to
program desired time intervals to suit the rock mass conditions within a stope.
The following information should be recorded on ring section charge
plans prior to firing, with amendments recorded during charging:
•
•
•
•
The date the ring section was fired
The amount and explosive type used in each hole
The actual firing sequence used
Any problems encountered while charging and firing each ring
section
• The results of the blast including quality of fragmentation, misfires,
hole freezing, stope wall falloff, backbreak, etc.
The ring charge plans should be returned to the mine planning department
for stope reconciliation once the firing of a stope is complete.
Figure 6.34 shows a typical firing sequence in open stoping. The following
are the basic guidelines for ring blasting:
1.
2.
3.
4.
In any ring, the longest hole is fired first.
Rings fire from the bottom proceeding upward.
Interlocking toes fire simultaneously.
Footwall and hangingwall holes fire late in the sequence.
298
Geotechnical Design for Sublevel Open Stoping
13
11
9
12
12
2m
13
11
10
9
10
10
12
8
11
11
11
8
8
5
5
5
2
2
7
4
1
3
6
9
7
4
1
3
6
6
4
2
1
3
5
2m
1
Detonation sequence
FIGURE 6.34
Typical detonation sequence in open stoping. (Courtesy of Mount Isa Mines, Mount Isa,
Queensland, Australia.)
6.8 Raise and Cutoff Slot Blasting
Successful excavation of raises and COSs is critical in sublevel stoping, as
they provide free faces and voids into which the remaining ore in a stope is
blasted. Raises and slots are critical areas where significant rock mass damage can occur due to the high concentration of explosive energy utilized
to ensure the formation of the initial free face or void. Breakthrough holes
drilled parallel to the initial raise or LHW are implemented to create the COS
as shown in Figure 6.35.
6.8.1 Longhole Winzes
A LHW provides an initial void into which the COS is blasted. The fundamental principle of a LHW is similar to that of a large burn cut, where a number of large-diameter relief holes are left uncharged to provide an initial void
into which the charged shothole(s) can break. The key to successful LHW firing is to have adequate initial void and adequate delay times between holes.
LHWs are typically fired using long period delay detonators to allow the
broken rock to fall out of the winze before the next hole is detonated.
Drilling accuracy in a LHW is critical. Excessive drill deviation may cause
the following:
299
Drilling and Blasting
Main rings
Cut-off slot
Raise
or
LHW
FIGURE 6.35
Three-dimensional view of a slot within an open stope geometry.
1. Blastholes designed as void become ineffective and the winze at a particular horizon does not resemble the intended design (Figure 6.36).
2. Blastholes intersect each other, causing confusion or difficulties during charging. This may also create sympathetic detonation, desensitization, or hole dislocation, thus compromising breakage.
3. Blastholes having excessive burdens may not break out adequately
and cause damage around the winze.
4. Excessive movement near large-scale geological discontinuities may
block holes.
5. The final axis of the winze may be changed at some horizons.
Figure 6.37 shows a standard LHW design based on the premise that drill
setup is a major cause of drill deviation contributing to the issues discussed
earlier. A good LHW design should minimize drilling setup positions, have
300
Geotechnical Design for Sublevel Open Stoping
Top level
E
F
A
B
C
D
G
H
F
E
A
G
D
C
B
H
Breakthrough level
FIGURE 6.36
Deviation of drillholes in a LHW.
1.5 m
3.0 m
#12
#11
#6
#4
1.3 m
0.5
#8
#3
m
#9
#1
#5
#2
#10
89 mm blasthole
160 mm relief hole
#2 Detonation sequence
#7
#13
FIGURE 6.37
LHW pattern—Hilton Mine, Mount Isa Mines. (Courtesy of Mount Isa Mines, Mount Isa,
Queensland, Australia.)
301
Drilling and Blasting
enough large-diameter relief holes, and the charged cut holes should be positioned such that they are shielded from one another. The design shown in
Figure 6.37 is set up so that the winze can be started from a number of points
if required. In practice, however, more holes may have been drilled than
those shown in the pattern. These are either rebores for deviated drillholes
or additional easer holes around the winze.
Figure 6.38 shows an alternative LHW design having 17 charged holes of
73 mm diameter and 4 relief holes of 115 mm diameter. At Mount Isa Mines,
a typical 12 m-long winze similar to the one shown can be drilled in 36 h
with an Atlas Copco Simba H221 using T38 rods.
Figure 6.39 shows a typical LHW pattern that uses a Robbins 12MD raise
hole to provide a 660 mm-diameter initial void. This minimizes the potential
for drill deviation-related problems experienced with LHW. Nevertheless,
significant preparation work is required to use a 12MD machine, as the floor
has to be cleaned and a concrete slab poured prior to raise-boring. Following
the completion of the raise, an accurate survey pickup of the hole is required,
so that an appropriate number of easers can be drilled around the raise to
establish the COS. Data from Mount Isa Mines showed that when a stope
height was greater than 25 m, a 12MD became cost-effective in comparison
to the Simba for drilling an entire COS. Although difficult to quantify, it is
expected that reduced dilution may occur using a 12MD cutoff compared to
a Simba COS.
The maximum length for a LHW in sublevel stoping is around 40–50 m.
Conventionally, the holes are charged from the top of the winze, with 3–6 m
cuts fired each time starting at the bottom of the winze and moving up. If a
large-scale structure is present, the winze should be fired to the structure to
avoid falloff. If the winze freezes (malfunction), the holes must be washed
0.65 m
2m
0.3 m
73 mm blasthole
2m
115 mm relief hole
0.65 m
FIGURE 6.38
Typical LHW pattern, Lead Mine. (Courtesy of Mount Isa Mines, Mount Isa, Queensland,
Australia.)
302
Geotechnical Design for Sublevel Open Stoping
2.1
2
m
3.00 m
Relief hole
140 mm blasthole
0.60 m
6m
0.6
3.00 m
0.60 m
FIGURE 6.39
A typical blasthole pattern used in conjunction with a 0.66 m raise. (Courtesy of Mount Isa
Mines, Mount Isa, Queensland, Australia.)
and blown out using compressed air. If the holes cannot be reestablished,
then redrilling may be required. A typical advance for a 3 m × 2 m blasted
LHW ranges from 5 to 10 m per blast, depending upon the amount of hole
deviation present.
6.8.2 Cutoff Slots
The COS is the most important part of the in-stope extraction sequence,
as it provides the initial void into which the subsequent rings are fired
(Figure 6.40). The first ring adjacent to a COS will typically only break to the
width defined by the slot void, although slightly wider rings may be gradually fired into the initial slot to “gain ground.” A large increase in orebody
width over the stope length may require the inclusion of an additional COS
in the widest section of the orebody. In general, the decision on the location
of a COS is dependent upon orebody width, the fill type of adjacent stopes,
and access constraints. In narrow vein orebodies in which bench stoping is
practiced, the COS is commonly designed to extend across the full width of
each stope.
In order to successfully fire a COS, it is recommended to have orebody sills
open to full operating width above and below the slot. Fanning of blastholes
into unbroken ground to strip the slot is usually unsuccessful, as the blastholes are unlikely to pull to the full design depth. This problem becomes
exaggerated with increasing angles of advance (angle of arc) away from the
void (see Figure 6.41). To minimize the angle of arc that the drillholes are
required to cover, a drill rig has to be capable of drilling across the entire
orebody width, using as near to parallel holes as possible.
Drilling and Blasting
303
FIGURE 6.40
Longitudinal view into a COS, Mount Isa Mines. (Courtesy of Mount Isa Mines, Mount Isa,
Queensland, Australia.)
LHWs and COSs have the greatest concentration of explosive energy of
any area of a blast design. Consequently, the LHW position within a COS
should be as close as possible to the stope footwall to minimize blast damage to the stope hangingwall. A number of cutoff design rules are suggested
as follows:
1. Ideally, the COS should be positioned in the widest part of the
orebody.
2. In narrow orebodies, the drill drives at the COS position should be
stripped to full orebody width. In wide orebodies, the drilling drives
should be connected above and below the COS (Figure 6.35).
3. Drillholes should be drilled as close as parallel to the raise or LHW
as possible. When holes converge at the winze or raise area, they
must be fired together.
4. The COS must be enlarged by firing holes into the winze or raise, as
if the slot were a narrow orebody. This is usually done using a “dice
five” pattern with two lead holes firing first into the LHW, followed
by an easer hole.
304
Geotechnical Design for Sublevel Open Stoping
Ground opened
by cut-off slot
firing
10L
11E
FIGURE 6.41
Example of fanned holes in a COS not pulling to full depth. (Courtesy of Mount Isa Mines,
Mount Isa, Queensland, Australia.)
At Mount Isa Mines, some of the COSs are drilled using vertical holes of 140
mm diameter and up to 50 m in length. The holes are drilled on a “dice five”
pattern, with 1.5–1.8 m burden and 3.2–3.6 m spacing (Figure 6.42). The holes
are usually drilled to breakthrough on the sublevel below, which allows for
drill accuracy checks and draining of water and drill cuttings. Holes are
drilled parallel to the cutoff raise and, if applicable, the hangingwall and
footwall of the orebody. It is important that the collars and breakthroughs
are recorded by surveyors prior to designing the remaining stope blasthole
patterns or slot charge plans, as these may be significantly different from
the initial designed coordinates. Figure 6.43 shows the position of the COS
with respect to the main rings within two stoping geometries. Figure 6.44 is
a long section view showing the location of the slot with respect to the main
rings within the stope.
Stringing of COS holes during drilling ensures that the designed number of
holes have been drilled. This also helps to establish drilling accuracy. When
305
Drilling and Blasting
Dia. (mm) Burden
Area
Spacing
TPMD
PF
Length
Upholes
70
1.8 m
1.2 m
4.6
0.67
20 m
Downholes
140
1.8 m
3.6 m
12.4
0.74
40 m
FIGURE 6.42
Plan view of COS showing raise and blasthole positions. (Courtesy of Mount Isa Mines, Mount
Isa, Queensland, Australia.)
Filled stope
Expansion
rings
Expansion
rings 1–4
40 m
Expansion
rings
Cut-off slot
Cut-off slot
Ring 5
Ring 8
Ring 9
Ring 10
Ring 6
Mass blast
Mass blast
Ring 7
Ring 7
Ring 8
Ring 9
Ring 10
Ring 11
Ring 12
40 m
FIGURE 6.43
Plan view showing COS with respect to the main ring geometries. (Courtesy of Mount Isa
Mines, Mount Isa, Queensland, Australia.)
cutoff raises are raise-bored after the COS has been drilled, holes around the
raise must be strung to ensure an adequate number of holes are available to
break into the raise. However, the practice of raise-boring after cutoff drilling is not recommended.
COS blasting sequences depend upon the size of the raise used. For example, a 60 m COS lift having a 1.8 m-diameter raise may be fired in two 30 m
lifts. A slot with a 0.66 m-diameter raise can be blasted in 12–20 m lifts.
Conventionally, when a LHW has been fired to approximately half way,
blasting of the COS blasthole toes can start. A recommended lead-lag
306
4
Expansion
7
4
7
8
6
5
4
3
1
Cut-off slot
7
Massblast
Geotechnical Design for Sublevel Open Stoping
3
1
2
1 Blasting sequence within stope
FIGURE 6.44
Typical long section for a COS. (Courtesy of Mount Isa Mines, Mount Isa, Queensland,
Australia.)
25
23
24
21
22
5
22
7
25
23
21
3
1
2
8
4
6
22
FIGURE 6.45
Blasting a LHW and cutoff collars to breakthrough.
between the LHW advance and the COS advance is approximately 10 m.
When firing out a COS, up to 6 holes at a time are blasted when firing toes,
while up to 12 holes are blasted when firing collars (Figure 6.45). In some
cases, all slot toes are fired, with an arched geometry from the LHW to the
slot extents. If for some reason, cutoff toes are not fired prior to completion
of the LHW, up to 6 complete COS holes are fired at a time into the LHW
void. Both LHW and COS holes should be double-primed to ensure that they
detonate. In addition, when firing a COS and main rings or cleaner rings, a
maximum of 3 rings are recommended to ensure adequate void for the rings.
307
Drilling and Blasting
L4
1
Fa
ul
t
Uphole LHW
O392
Filled
1.7
%
2.0
%
20B
Bas
em
ent
Downhole LHW
2450 RL
21E
FIGURE 6.46
COS incorporating downhole and uphole LHWs. (Courtesy of Mount Isa Mines, Mount Isa,
Queensland, Australia.)
In some stoping geometries, uphole winzes may be required in conjunction
with uphole COS charging (Figure 6.46). Uphole winzes require a single firing,
and therefore their length usually does not exceed 15 m. If an uphole slot is to
be wider than the LHW, then additional stripping holes should be fired with
the LHW. If the LHW does not pull to the required depth, it is likely that a
recovery LHW will have to be drilled or the entire slot will have to be redrilled.
Figure 6.47 shows a 2 m × 3 m uphole LHW geometry used at Mount Isa Mines.
The relief holes in the figure are drilled 1.5 m deeper than the charged holes.
6.9 Trough Undercut Blasting
TUCs are designed in a similar manner to the main stope rings. However,
TUCs are usually drilled using 70–89 mm-diameter upholes inclined at 70°
and limited to a length of 15 m to allow conventional blow-loading of explosives. TUCs are shaped to promote the best rill angle at the stope drawpoints
308
Geotechnical Design for Sublevel Open Stoping
3.0 m
0.4 m
0.4 m
0.5 m
0.25 m
1.0 m
89 mm blasthole
0.8 m
150 mm relief hole
0.25 m
0.5 m
0.7 m
2.0 m
FIGURE 6.47
Typical uphole LHW geometry. (Courtesy of Mount Isa Mines, Mount Isa, Queensland,
Australia.)
while protecting the rock mass at the brows as much as possible. Inclined
(dumped) TUC rings are required, so that charging of holes at the edge of
a stope is not undertaken. TUC rings are designed to be small rings which
minimize the charging time and exposure of charge-up personnel to the slot
void or stope rill.
Figure 6.48 shows a 70 mm-diameter blasthole TUC, with the first row
of holes inclined at 50°, followed by inclinations of 60°, 70°, and 80°, with
the remainder of the rings at 90°. The collars for the first row of holes are
located 4 m away from the COS. A horizontal distance of 2.5 m is used for the
remainder of the rings.
Consecutive TUC rings are designed using staggered patterns, and the
holes must interlock with the toes of the downholes from the sublevel above.
A nominal toe overlap of 2 m is recommended.
6.10 Rock Diaphragm Blasting
The role of a diaphragm is to protect an adjacent fill mass (weak or uncemented) from blast damage from production blasting and help to minimize
fill failures. COSs should be designed to the width of the main ring firings,
and cleaner rings adjacent to the diaphragm can be used to optimize recovery
309
Drilling and Blasting
70 mm blasthole diameter ANFO loading
3m
2m
50°
70°
(a)
Toe spacing
(b)
2.5 m
3m
2.35
m
2m
Burden
FIGURE 6.48
(a) Cross section and (b) longitudinal section views of a TUC. (Courtesy of Mount Isa Mines,
Mount Isa, Queensland, Australia.)
and minimize blast damage to the diaphragm (see Figure 2.7). Cleaner rings
are drilled and blasted from diaphragm ring recesses located at the edges of
the stopes. Hole deviation can become a large problem while drilling cleaner
rings, as the holes are drilled from limited locations and are of excessive
length (Figure 6.49).
Some of the holes in cleaner rings and diaphragm rings are difficult to
clean out and prepare for charging, as the holes can be damaged from blasting the stope COS or main rings. In addition, insufficient information on
the exact location of the fill boundaries can lead to inadequate diaphragm
design thickness, further contributing to fill mass instability. Development
of the cavity-monitoring system has allowed final stope geometries to be
determined, although localized rock falls can occur prior to fill completion.
6.11 Mass Blasting
Mass blasts consist of multiple blasthole rings and can exceed 100,000 tonnes
per firing. The sequencing rules for mass blasting must take into account blast
vibration constraints, rock mass damage from overconfinement, delay variability, geometrical constraints, and major geological structures as follows:
1. The longest hole in a ring should be fired first to maximize the initial
void created at the start of the firing. Other holes are then sequentially stripped into the void created by the first hole.
2. Holes toeing into each other should be fired on the same number.
310
Geotechnical Design for Sublevel Open Stoping
1500 E
1450 E
2950
Meters drilled
346.4
Meters charged
188.4
Tonnes broken
7929.6
Tonnes/m drilled
22.8
Kg ANFO /ring 1957.2
/tonne
0.24
/m drilled 5.6
38.4
+3°
42.8
–10°
12
27
31
.1
35 0° 0.8
3 6°
–3
–3
9
13
18
.0
40 °
5
–2
12D
19
2
–4 7.5
3°
45.7
°
–15
3
45.
°
–20
8
23
40.3
–4°
12/L
2900
FIGURE 6.49
Cross section showing a typical cleaner ring charge geometry. (Courtesy of Mount Isa Mines,
Mount Isa, Queensland, Australia.)
3. A minimum of 30% expansion void is required prior to mass blasting to accommodate broken material swell.
4. A minimum delay time of 15–20 ms/m of burden is suggested
between successive rings, while changes of firing direction within a
mass blast require 100–200 ms delays.
5. Perimeter holes as well as holes parallel to stope crowns should
be fired a few numbers after adjacent holes within a ring to allow
adequate relief. This reduces confinement of explosive gases at the
stope boundaries and minimizes the likelihood of overbreak. The
delay between stope boundary holes and adjacent holes should follow the same rule as per burden, that is, 15–20 ms/m of toe spacing. Alternatively, one complete number in the detonation sequence
should be skipped between adjacent holes.
Drilling and Blasting
311
In order to increase the number of delays available for mass blasting, a
combination of trunk line delays and down-the-hole delays can be used as
discussed in Section 6.7.2. It is extremely important that all surface delays
be activated before the first hole in the first ring detonates to avoid cutoff.
Before a mass blast is undertaken, all reentry, inspection, ventilation, and
shift change procedures should be detailed. The advantages of mass blasting
include the following (Guilfoyle and Bradford, 1982):
1. Safer work conditions arise when charger and ring-firer crews are not
continually required to work near freshly blasted stope boundaries.
2. Improved rock fragmentation results from the shearing action of
interacting detonation charges and in-flight rock collisions.
3. A better utilization of resources is possible due to a concentrated and
semicontinuous charging operation which proceeds simultaneously
on a number of sublevels.
4. Because fewer individual firings are required, the problem of postblast falloff is reduced.
5. Large-scale structural discontinuities (such as faults, shears, etc.)
can be included within a single firing, thus minimizing ground
movement and the potential loss of blastholes.
6. Stable conditions can be maintained during slotting, initial ring
expansion, and charging of the mass blast, after which no need
exists for personnel to reenter the area.
7. Following a mass blast, passive support to the stope walls (rock or
fill masses) is provided by the broken ore. Up to three-quarters of the
stope may be filled by the broken ore following blasting.
8. Fewer individual blasts are likely to minimize damage to services
and other scheduled activities around the stopes.
9. The large broken ore tonnages from mass blasts allow uninterrupted
production at high extraction rates from the stope drawpoints.
The disadvantages of mass blasting may include the following:
1. Multiple-lift mass blasts are typically initiated from multiple access
drives. Should a cut-off occur, it may be difficult to gain reentry to
those areas.
2. Mass blasts often create a large change of geometry likely to redistribute significant stress around the stope boundaries. Stress changes
may induce rock noise and damage and a reentry period to the stoping area may be required, thus delaying production.
3. Any malfunction of the initiation system or explosive early in the
firing sequence can “freeze” the entire mass firing.
312
Geotechnical Design for Sublevel Open Stoping
4. Mass firing on top of broken stocks within a stope can lead to excessive ore compaction at the draw point.
5. Inadequate delays in main ring firings or between main firing levels
can lead to rapid overpressurization of the development drives from
ore block displacement, causing damage to the ventilation system.
6.11.1 Control of Ground Vibration
In addition to a correct and complete detonation sequence of all the holes
involved in a mass blast, minimization of damage to adjacent structures
(such as shafts, pillars, etc.) from excessive vibration is an important objective. Overpressure from blasting may also cause significant damage to ventilation systems. Consequently, the initiation sequences must be designed with a
charge weight per delay evenly distributed throughout the mass blast duration. The objective is to prevent periods of high explosive concentration within
the blast. Often, the quantity of explosive detonating within a specified time
interval is limited to 1000 kg. The optimum delay interval between successive
detonating charges to minimize wave interaction has been suggested by Heilig
(1999) to be half of the duration of vibration from a single blasthole charge. For
most underground rock types (for a distance up to 200 m from the blast) this
value has been determined by Heilig (1999) to be approximately 20 ms.
The effects of charge weight per delay distribution throughout a blast
can also be determined by blast monitoring to ensure that the number of
charges initiating per a 20 ms period are minimized. In mine sites where
a town or city is nearby, the standard practice is to monitor the surface
vibration from all stope blasts exceeding 100,000 tonnes. Monitoring experience suggests that the surface vibration values obtained from surface
monitoring are likely to change from place to place. It is possible that due
to large-scale geological discontinuities or the effects of mining voids or
fill masses, some sites may experience higher levels than those monitored
at shorter distances. Figure 6.50 shows the monitored peak particle velocity (PPV) from long-term surface blast monitoring at the Mount Isa Mines
lease boundary (approximately 1 km from the blast). At Mount Isa Mines,
the vibrations induced by stope blasting are generally acceptable to the
community. The historical level of complaints is very low and no damage to property has ever been linked to the large-scale underground blasting activities. Consequently, on the basis of the long-term data collected at
Mount Isa Mines, a suitable criterion which can be realistically achieved in
sublevel stoping is suggested as follows:
1. The surface PPV of 10 mm/s may be exceeded by up to 10% of the
total number of daily blasts.
2. The level cannot exceed 20 mm/s at any time, including during mass
blasts.
313
30
Mass blasts
25
20
15
10
5
450,000
400,000
350,000
300,000
250,000
200,000
150,000
100,000
50,000
0
0
Peak particle velocity (mm/s)
Drilling and Blasting
Tonnes per blast
FIGURE 6.50
Surface vibration levels from mass blasts at Mount Isa Mines. (Courtesy of Mount Isa Mines,
Mount Isa, Queensland, Australia.)
7
Rock Reinforcement and Support
7.1 Introduction
The objective of ground support is to maintain excavations safe and open for
their intended purpose and lifespan (Villaescusa, 1999a). In an open stoping
context, the effectiveness of a ground support strategy is important for two
main reasons: safety to personnel and equipment within the stope development, and achieving the most economical extraction of ore with minimal
dilution from the final stope walls.
The type of ground support required in a particular stope location is
dependent on several factors, including the available rock mass strength,
the geometry of the excavation, the stresses present in the rock, the blasting practices, and the weathering process (see Section 1.4). Two stabilization
techniques can be used to improve and maintain the load-bearing capacity
of a rock mass near the boundaries of an underground excavation (Windsor
and Thompson, 1992):
• Rock reinforcement
Reinforcement is considered to be exclusively systems of components
installed in boreholes drilled in a rock mass, for example, cementencapsulated threaded bar, friction stabilizers, and cable bolts. The
reinforcing elements are an integral part of a reinforced rock mass.
• Rock support
Support is considered to be exclusively systems of components that
are located on the exposed faces of excavations, for example, mesh,
straps, shotcrete, and steel arches. The supporting members are
external to the rock and respond to significant inward movement of
the rock mass surrounding an excavation.
The reinforcing elements provide effective stabilization by helping a rock
mass to support itself (Hoek and Brown, 1980). This is achieved by preventing unraveling and enhancing the self-interlocking properties of a rock mass.
A reinforcement pattern strengthens the exposed rock mass around an excavation by preventing the detachment of loose blocks and by increasing the
315
316
Geotechnical Design for Sublevel Open Stoping
shear strength of the geological discontinuities intersected by the reinforcing
elements. This results in a reinforced zone that helps to redistribute stresses
around the excavations and also minimizes dilation of preexisting geological discontinuities. Careful blasting and correct scaling reduce the amount
of loose rock that has to be supported, thus enhancing the self-stabilizing
behavior of a rock mass.
In sublevel open stoping mines, the primary form of excavation stabilization is provided by the reinforcement pattern installed within the various stope development excavations. Rock support, such as that provided by
mesh and shotcrete, is required to provide surface restraint within a reinforcement pattern at the excavation boundaries. The reinforcement controls
the overall excavation stability through keying, arching, or composite beam
reinforcing actions (Windsor and Thompson, 1992), while mesh or shotcrete
supports the small loose pieces of rock that can potentially detach within a
bolting pattern (Figure 7.1).
Ground support can be considered to consist of combinations of reinforcement and support systems. It is normal practice to design the reinforcement
to act with the support to form a ground support scheme (Windsor and
Thompson, 1992). That is, the support is restrained by a plate held in place
by the reinforcement system. If this interaction at the collar of the reinforcement system fails, then the ground support scheme will not be effective in
retaining the unstable rock. Another important aspect of the ground support
design is its overall response to the amount of rock mass deformation and
the rate at which this occurs.
FIGURE 7.1
Support and reinforcement of a highly stressed blocky rock mass.
317
Rock Reinforcement and Support
7.2 Terminology
A classification to describe the forms, functions, basic mechanics, and behavior of the different commercially available rock support and reinforcement
systems was developed by Thompson and Windsor (1992). The method classifies the existing reinforcement systems by dividing them into three basic
categories in order to explain the basic mechanisms of load transfer between
the reinforcing elements and a rock mass.
A description and comparison of devices within a particular category or
between separate categories is facilitated by the method. The categories are
shown in Figure 7.2 and are described as continuous mechanical coupled
(CMC), continuous friction coupled (CFC), and discrete mechanical and friction coupled (DMFC). Some typical reinforcing devices are grouped according to this classification in Table 7.1.
Type
Longitudinal view of reinforcement element
Unstable
surface
region
Stable
interior
region
Unstable
surface
region
Stable
interior
region
CMC
CFC
DMFC
Unstable
surface
region
Stable
interior
region
FIGURE 7.2
Classification of reinforcement action. (After Thompson, A.G. and Windsor, C.R., A classification system for reinforcement and its use in design, in T. Szwedzicki, G.R. Baird, and T.N.
Little, eds., Proceedings of the Western Australian Conference on Mining Geomechanics, Kalgoorlie,
Western Australia, Australia, June 8–10, 1992, pp. 115–125, Western Australian School of Mines,
Kalgoorlie, Western Australia, Australia.)
318
Geotechnical Design for Sublevel Open Stoping
TABLE 7.1
Classification of Typical Reinforcement Devices
Type
Description
CMC
Full-column cement-/resin-grouted bars (grouted CT bolt,
deformed bar, threaded bar, and fully grouted Posimix)
Cement-grouted cables (plain strand and modified geometry)
Friction stabilizers (split-set bolt, friction bolt, and Swellex)
Mechanical anchors (ungrouted CT and HGB bolts, expansion
shell, and slot and wedge)
Single cement/resin cartridge anchors (paddle bolt, deformed
bar, and debonded Posimix)
CFC
DMFC
Source: Thompson, A.G. and Windsor, C.R., A classification system for
reinforcement and its use in design, in T. Szwedzicki, G.R. Baird,
and T.N. Little, eds., Proceedings of the Western Australian
Conference on Mining Geomechanics, Kalgoorlie, June 8–10, 1992,
pp. 115–125, Western Australian School of Mines, Kalgoorlie,
Western Australia, Australia.
7.2.1 Continuous Mechanical Coupled
A CMC reinforcing element relies on a fixing agent, usually a cement- or
resin-based grout, which fills the annulus between the element and the borehole wall. The main function of the grout is to provide a mechanism for load
transfer between the rock mass and the reinforcing element.
The reinforcing elements are usually manufactured with variable crosssectional shapes in order to increase the element-to-grout bond strength.
A mechanical key is effectively created by the geometrical interference
between the element and the grout along the entire reinforcement length.
The element is defined as continuously coupled to the rock mass by way of
interlock with the grouting agent (Thompson and Windsor, 1992).
7.2.2 Continuous Friction Coupled
A CFC reinforcing element is installed in direct contact with the rock
mass. The mechanism of load transfer is a function of the frictional forces
developed between the reinforcing element and the borehole wall. The
load transfer is limited by the radial prestress set up during the initial element insertion. The bond strength is a function of the element diameter,
the borehole diameter, and any geometrical irregularities occurring at the
borehole wall.
The radial stress can be related to a force along the length of the reinforcing element and is achieved by deforming the cross-sectional area of the
element to suit the borehole. This can be achieved by either contracting an
oversized element section into an undersized borehole (friction stabilizer)
or by expanding an undersized element section into an oversized borehole
319
Rock Reinforcement and Support
(Swellex bolt). A modification of this reinforcing action can be achieved
by cement grouting of the split-set bolts as described by Villaescusa and
Wright (1997).
7.2.3 Discrete Mechanical and Friction Coupled
A DMFC device transfers load at two discrete points, namely, the borehole
collar and the anchor point, which is located at some depth into the borehole.
The length of the element between the two discrete points (plate and anchor)
is actually decoupled from the rock mass. The load transfer is then limited
to a relatively short anchor length. Load transfer at the anchor point can be
achieved by either mechanical (grouted anchor) or frictional means (expansion shell).
The strength of an expansion shell may be limited by the strength of the
rock at the borehole wall, and these devices are best suited to hard rock applications (Villaescusa and Wright, 1999). Grouted anchors may be used in soft
rock masses, where a high load transfer can be achieved over a short length,
provided that gloving by the resin cartridge does not occur (Villaescusa
et al., 2008).
7.2.4 Load Transfer Concept
The load transfer concept is one of the most fundamental concepts required
to completely understand the behavior of a reinforcing element. The concept shown in Figure 7.3 can be used to understand the stabilizing action of
all reinforcing devices and their effect on excavation stability. The concept
Frictional resistance and
mechanical interlock within
stable (interior) region
l
ica y
log nuit
o
Ge onti
c
dis
Embedment length
within stable region
Unstable
region
Movement
FIGURE 7.3
Load transfer and embedment length concepts.
Frictional resistance and
mechanical interlock within
unstable (wedge) region
(complemented by plate)
320
Geotechnical Design for Sublevel Open Stoping
can be explained by three basic individual components (Windsor and
Thompson, 1993):
1. Rock movement at the excavation boundary, which causes load transfer from the unstable rock, wedge, or slab to the reinforcing element
2. Transfer of load via the reinforcing element from the unstable portion to a stable interior region within the rock mass
3. Transfer of the reinforcing element load to the rock mass in the stable zone
Failure of a rock block or a layer of rock being stabilized may be associated
with any one of the three separate components of load transfer because of
insufficient steel capacity (rupture of the reinforcing element) or inadequate
bond strength (slippage).
7.2.5 Embedment Length Concept
Embedment length is the length of a reinforcing element on either side of
an active geological discontinuity defining a potentially unstable wedge or
block such as that shown in Figure 7.4. The critical embedment length is the
minimum length of reinforcement required to mobilize the full reinforcing
capacity of the system.
Short embedment lengths within an unstable region can be compensated
for by the fact that a properly matched face plate provides enough surface
restraint to mobilize the system capacity. Short embedment lengths within
FIGURE 7.4
Slippage within a stable region due to insufficient embedment length.
321
Rock Reinforcement and Support
the stable region are more critical, especially when a reinforcement element
is installed at an unfavorable angle with respect to the free surface.
7.2.6 Reinforcement Performance Indicators
A number of parameters may be used to characterize the performance of
different reinforcement systems. In the absence of being able simply to
simulate axial and shear loading of reinforcement, reinforcement performance is generally characterized by the force–displacement response of a
reinforcement system subjected to axial loading. Figure 7.5 shows a generic
force–­displacement response with annotations of a number of reinforcement
system performance indicators.
The performance indicators may be grouped as follows (Thompson
et al., 2012):
Force
• Force capacities
Fmax Maximum force.
Fres Residual force at maximum displacement.
• Displacement capacities
δp Displacement at maximum force.
δmax Maximum displacement.
• Stiffnesses
Kti Initial tangent stiffness.
Peak
Fmax
Residual
Fres
1
Kti
Ksp
1
1
δp
Ksr
Displacement
δmax
FIGURE 7.5
Force–displacement response for a generic reinforcement system subjected to axial loading.
(With kind permission from Springer Science + Business Media: Geotech. Geol. Eng., Ground
support terminology and classification: An update, 30, 2012, 553, Thompson, A.G., Villaescusa,
E., and Windsor, C.R.)
322
Geotechnical Design for Sublevel Open Stoping
Ksp Secant stiffness at maximum force.
Ksr Secant stiffness at maximum displacement.
• Energy absorption capacity
Energy absorption capacity is equivalent to the area between the
force–displacement curve and the displacement axis and is relevant
to the performance of reinforcement subjected to dynamic loading.
Ep Energy absorption to peak force.
Er Energy absorption at maximum displacement.
Other parameters may need to be considered if the reinforcement system
is loaded predominantly in shear. For example, it is known that strand is
more flexible when loaded in shear than a solid bar and can therefore sustain higher shear displacements. The ability of a reinforcement system to
sustain shear displacements is improved by de-coupling of the element from
the grout as it allows for axial displacement of the element to be distributed
over a longer length of the element near the discontinuity.
7.3 Ground Support Design
Ground support design in most stoping operations is based on previous
experience and evolves over a number of years. In many instances, there may
be nothing technically wrong with the designs, and the performance can be
considered to be acceptable. However, rock mass conditions usually change
with the progress of a mine (e.g., stresses increase as the depth of mining
increases and when the global extraction increases), and accordingly, ground
support performance may change and become unacceptable. Also, the experiential ground support measures may not be optimal. That is, the installed
reinforcement and support capacities may not satisfy the rock mass demand.
A formal ground support design procedure (Thompson et al., 2012)
attempts to
1. Identify the rock mass demand
2. Select reinforcement and support systems with appropriate characteristic responses
3. Specify their arrangement
The generic procedure consists of several distinct steps (Thompson et al., 2012):
1. Identify a mechanism of failure
2. Estimate the areal support demand
323
Rock Reinforcement and Support
3.
4.
5.
6.
Estimate the reinforcement length, force, and displacement demand
Estimate the energy demand
Select appropriate reinforcement and support systems
Propose arrangement of reinforcement and support systems and
evaluate
7. Specify the complete ground support scheme
This procedure may need to be applied to several different observed mechanisms of failure. In most instances, it is not possible to perform formal
designs because the rock mass variables that define demand cannot be quantified with any degree of confidence. However, the rock mass demand can
usually be defined qualitatively in terms of low, medium, high, very high,
and extremely high reaction pressure, surface displacement at failure, and
energy demands per meter square (Table 7.2). These qualitative descriptions
of rock mass demand can then be satisfied by reinforcement systems that can
be classified using corresponding ratings (see Figure 7.40).
The design process is more complicated when the rock mass experiences
seismicity and the ground support is subjected to dynamic loading. For
dynamic ground support design, it is necessary to consider the expected
nature of seismic events associated with slip on major structures or unstable
propagation of rock mass failure and their proximity to excavations where
reinforcement and support will be installed. Ideally, the design event must
be based on the history of seismic events at a particular mine and their correlation with other major influencing factors such as large faults and the stress
concentrations (induced by mining) relative to the rock mass strength (Kaiser
et al., 1996). This procedure assumes that the design event is remote from the
surfaces of an excavation. However, it is also possible that the event source is
in the immediate vicinity of an excavation wall. In this case, the mechanism
of failure will result in a different form of dynamic loading of the ground
support. It is worth noting that very high values of PPV have been measured
without associated rock failure and ejection (Fleetwood, 2010).
TABLE 7.2
Typical Rock Mass Demand for Ground Support Design
Demand Category
Low
Medium
High
Very high
Extremely high
Reaction
Pressure (kPa)
Surface
Displacement (mm)
Energy (kJ/m2)
<100
100–150
150–200
200–400
>400
<50
50–100
100–200
200–300
>300
<5
5–15
15–25
25–35
>35
Source: Modified from Thompson, A.G. et al., Geotech. Geol. Eng., 30(3), 553,
2012.
324
Geotechnical Design for Sublevel Open Stoping
7.3.1 Location of Failure due to Overstressing
The analysis of stresses around underground excavations in rock can be
accomplished using a number of different numerical analysis techniques.
These can range from simple linear elastic analyses performed in two dimensions to complex three-dimensional nonlinear analyses. For complex geometries, two-dimensional analyses cannot be expected to provide meaningful
guidance on the locations of failures. On the other hand, the latter types of
analyses can be expected to provide the most detail and understanding of
the changes in rock stresses as excavations are formed and extraction progresses. However, due to their complexity, they require significant resources
to be expended in terms of testing to obtain realistic material properties and
multiple back analyses to calibrate the models with documented on-site
observations and experience (Pardo and Villaescusa, 2012).
An intermediate approach is to use linear-elastic analyses in three dimensions (e.g., Wiles et al., 2004). This approach is able to model complex threedimensional models of excavations and sequences and to identify regions of
high-stress concentrations and volumes of rock where it might be expected
that the rock mass strength is exceeded (Figure 7.6). Again, however, it is
essential that the model is calibrated with documented experience. The
limitation of this approach is that the redistribution of stresses following
progressive rock mass failure cannot be determined. Nevertheless, the most
important outcome from the analyses is to identify areas that can be expected
to experience ground stability problems due to excessive stress.
Another methodology is that reported by Beck and Duplancic (2005). The
basis of this method is the three-dimensional nonlinear modeling computer
program Abaqus that can be used to predict the ground reaction curves at
distinct locations for different extraction sequences. The energy release associated with the ground reaction curve at the particular location can also be
predicted (Beck et al., 2010).
7.3.2 Depth of Failure: Stress or Strain Controlled
The depth of rock mass failure around excavations can be estimated by calculating the strength factors for the rock mass near excavation surfaces. The
aim of the analysis for intact rock failure is to determine the depth of failure
to provide estimates for both reinforcement length demand and reinforcement and support capacity demand. At this point, it is worth noting that it
may be possible to minimize or eliminate intact rock failure by modifying
the excavation from a flat back to an arched profile. Analyses have shown
that the stresses in the rock in the backs and shoulders of rectangular excavations are higher than when the excavation incorporates an arched profile
(Figure 7.7).
Martin et al. (1997) have found that the benefits of an arched back profile apply to both low- and high-stress environments. In both cases, the
325
Rock Reinforcement and Support
Strength
factor-A
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
Broken
Ground—2 m
UCS = 124 = –37°
Span—8 m
Anchoring zone
Reinforced
rock mass
Cracked zone
Broken–damaged zone
Span
FIGURE 7.6
Modeled damage zones for rock reinforcement design. (After Wiles, T. et al., Rock reinforcement
design for overstressed rock using three dimensional numerical modeling, in E. Villaescusa
and Y. Potvin, eds., Ground Support in Mining & Underground Construction, Proceedings of the
5th International Symposium on Ground Support, Perth, September 28–30, 2004, pp. 483–489,
Balkema, Leiden, the Netherlands.)
arch profile reduces the volume of failed rock that needs to be supported.
However, in intermediate stress environments, a flat back profile was found
to improve roof stability by forcing and restricting stress-induced failure to
the confined regions of the shoulders.
To predict the likely volume of failure, a particular site would require
estimates of the in situ stresses and in situ rock strength and stiffness. For
simple excavation shapes, graphical methods based on simple closed-form
analytical solutions based on elasticity theory such as those presented in
other text books (e.g., Hoek and Brown, 1980) could be used. Alternatively,
the use of stress analysis programs will allow the stress distribution
around the actual excavation shapes to be analyzed. Such an analysis could
incorporate an appropriate failure criterion (e.g., Wiles et al., 2004—see
Chapter 5). This failure criterion allows the depth of failure to be estimated.
326
Geotechnical Design for Sublevel Open Stoping
(a)
(b)
(c)
(d)
FIGURE 7.7
Excavation profiles for mine development in hard rock. (a) Square, (b) shanty back, (c) oval,
and (d) semicircular (flat floor).
The depth of failure coupled with an estimate of a bulking factor allows an
estimate to be made for the expansion of the rock surface. The estimates of
failure depth and volume provide the rock mass demands that need to be
satisfied by the reinforcement systems in terms of their length, force, and
displacement capacities and the support systems in terms of their force and
displacement capacities. The approach used by Beck and Duplancic (2005) is
to conduct a nonlinear stress analysis (using Abaqus) and then to define the
depth and volume of failure based on the calculated plastic strains.
7.3.3 Depth of Failure: Structurally Controlled
In structured rock masses, it is possible to estimate ranges of blocks sizes
formed from combinations of discontinuities with different orientations,
persistence, and spacing. Rock mass characterization techniques are requ­
ired to determine the likely sizes and shapes of the unstable blocks to be
supported by suitable reinforcement schemes. Depending upon the characteristics of the reinforcing scheme chosen, a suitable embedment length
that ensures full capacity of the system must be designed and installed.
Rock Reinforcement and Support
327
FIGURE 7.8
Large potentially unstable wedge reinforced with cable bolting.
A similar reasoning applies for cable bolt reinforcement of large unstable
wedges (Figure 7.8).
Over the years, a number of procedures for examining the stability of single and multiple blocks of rock have been developed. Readers are referred
to Thompson (2002) for a reasonably recent, comprehensive review of these
methods. Single reinforced block stability analyses may be performed with
the Rocscience program Unwedge or modules within the SAFEX package
developed by Thompson (2002). A probabilistic design method developed by
Windsor (1999) is also incorporated into the SAFEX package. This method
uses the variability of discontinuity set orientations, persistence, and spacing combined with the excavation geometry to predict the range of possible
block shapes and sizes. This information is then used to predict reinforcement lengths and the ground support pressure that needs to be provided.
The ITASCA three-dimensional distinct element program 3DEC can
be used to model the stability of block assemblies. The program allows for
the analysis of the kinematics of the interactions between blocks and can be
used to model failure mechanisms, stress redistributions, and the effects of
reinforcement. An alternative approach, again incorporated into the SAFEX
package, allows for the modeling of progressive unraveling in jointed rock
mass assemblies and the analysis of reinforced block stability. This approach
is described in detail by Thompson (2002).
328
Geotechnical Design for Sublevel Open Stoping
7.3.4 Ground Reaction Curve Concept
The concept of displacement demand and appropriate reinforcement are
best considered in terms of a ground reaction curve. Figure 7.9 shows a typical ground reaction curve (Windsor and Thompson, 1998), which is the
relationship between radial stress and radial displacement at the boundary
of an excavation. The radial direction is normal to the excavation surface.
The stress reduces from its value before excavation. For a stable excavation,
the radial stress will reduce to zero at a certain displacement. For unstable
excavation surfaces, a restraining force (from support and/or reinforcement) is required to maintain the rock mass stability and excavation shape.
Experience has shown that an equilibrium condition may be attained by
limiting displacements so that the rock assists in maintaining stability.
Large displacements are accompanied by rock mass loosening and may
lead to larger stabilizing force requirements as the volume of failure
expands.
Nonlinear numerical modeling methods can be used to quantify the
ground reaction curve for a given rock mass and excavation shape. It is
known that the displacement demand will be a function of the stress regime
and the mechanical and rheological properties of the rock. For example,
Characteristic force
Reinforced rock system response
Excitation characteristic
Reinforced system response
Rock system response
Mechanistic characteristic
Characteristic displacement
FIGURE 7.9
Ground reaction curve showing the reduction of force with increased displacement. (From
Windsor, C.R. and Thompson, A.G., Reinforcement systems—Mechanics, design and performance testing, in J. Orozco and J. Schmitter, eds., Proceedings of the Third North American Rock
Mechanics Symposium, Cancún, June 3–5, Int. J. Rock Mech. Min. Sci., 35, 4–5, Paper 076, 1998, 9pp.)
Rock Reinforcement and Support
329
failure in rocks that behave in a ductile manner is accompanied by significant postyield creep displacements. On the other hand, brittle rock failure
may be accompanied by a high-energy ejection of material at small displacements. The different types of rock mass behavior require support and reinforcement schemes with distinctly different characteristics.
Beck et al. (2010) have used the Abaqus program to predict ground reaction curves. If the curve can be predicted, then it is possible to design ground
support with an appropriate force–displacement response and capacities
so that the rock mass pressure, displacement, and energy demands are
satisfied.
7.3.5 Ground Support for Massive Rock and Low Stress
Massive rock masses are characterized by a limited number of discontinuity
sets with limited persistence and wide spacing between members of the sets.
As shown in Figure 7.10, the impersistent discontinuities do not intersect to
form distinct blocks of rock. Excavations formed in massive rock by drilling
and blasting can be expected to have some blast damage and localized surface instabilities that can be scaled down. However, clean profiles can result
from controlled drilling and blasting. This type of rock should not require
surface support or internal reinforcement at the time of forming the excavation. However, allowance should be made for changes in the stress conditions that might occur as a result of future mining.
7.3.6 Ground Support for Massive Rock and Moderate Stress
The creation of excavations in massive rock in a moderate premining stress
field may cause localized stress concentrations at distinct locations around
the excavation boundary. Figure 7.11 shows how stresses may cause failure
at one shoulder and the toe of the wall on the other side of the excavation.
These failures are induced by tensile cracking oriented normal to the minor
FIGURE 7.10
Massive rock with widely spaced discontinuities with limited persistence.
330
Geotechnical Design for Sublevel Open Stoping
Major
principal
stress
FIGURE 7.11
Localized crushing and spalling around an excavation.
principal stress. Postfailure control of the shoulder can be achieved with
either mesh or shotcrete and pattern bolting for restraint of the support. It is
also probable that failure at the toe of the wall may undercut the overlying
rock and propagate upward. Support restrained by reinforcement should be
used to maintain the stability of the toe.
7.3.7 Ground Support for Massive Rock and High Stress
High in situ or induced stress regimes may exceed the strength of the intact
rock and the rock mass. The failure modes may be similar to those shown
for moderate stress but may occur more violently due to the energy stored in
the rock mass. Also, in a highly stressed region of rock, sudden slip on major
discontinuities in the vicinity of the excavation is more likely with the associated release of energy in the form of pressure and shear waves that excite
the rock near the boundaries of excavations. These pressure and shear waves
cause changes in the local stresses and vibrations that may be sufficient to
initiate rock failure, loosening, and ejection as shown in Figure 7.12.
In these circumstances where all excavation surfaces are likely to be
affected, support using mesh-reinforced (embedded) shotcrete (Morton
et al., 2009b) restrained by reinforcement is suggested. The shotcrete is in
close contact with the rock and will provide immediate response to any
rock mass movements preceding the seismic event remote from the excavation. As indicated previously, small increases in the minor principal stress
normal to the excavation surfaces can increase the rock mass strength
and inhibit fracture propagation. However, should the rock fail violently,
the shotcrete may not have sufficient toughness to absorb the energy
331
Rock Reinforcement and Support
Removed by
crushing or
ejection
Vpp
Vs
Vp
FIGURE 7.12
Stress-induced violent failure with rock ejection.
associated with failure, and it may itself crack locally so that the localized displacements may not be able to be tolerated, even if the shotcrete is
reinforced with fibers (Morton et al., 2009b). A layer of mesh restrained by
bolts is flexible and is therefore able to sustain large rock mass movements
and retain the failed shotcrete (Figure 7.13). The types of bolts that are
used for shotcrete and mesh restraint may need to be specially designed
to allow for the bulking of the rock associated with the rock mass failure.
For example, the reinforcement element may need to be simply debonded
from the rock near the collar. If the potential displacements are larger than
the elongation of the element, specially designed anchors that slip may be
required. In both instances, the stiffness of response to rock mass movement is reduced and can result in rock mass loosening. If this is a concern, then a combination of stiff and flexible reinforcement systems may
be more appropriate.
(a)
(b)
FIGURE 7.13
Large deformation allowed by mesh-reinforced shotcrete. (a) Welded-wire and (b) woven-wire
reinforcement.
332
Geotechnical Design for Sublevel Open Stoping
7.3.8 Ground Support for Layered Rock and Low Stress
Stratified rock masses are characterized by continuous, approximately parallel planes with cross-jointing. In subhorizontally layered rock with crossjointing, the walls will be stable, but horizontal stresses are required to keep
the vertical joints closed and create vertical frictional resistance to downward displacement in the roof. Consequently, if the stresses are low, the frictional resistance is insufficient to prevent the rock between vertical joints
from falling. Progressive collapse can occur until a stable arch is formed as
shown in Figure 7.14.
The need for support will depend on the spacing between the vertical joints; that is, if the spacing is small, then mesh or shotcrete will be
required to span between the restraint provided by the reinforcement.
Reinforcement should be installed to intersect the vertical joints at an
oblique angle to improve the shear resistance. Otherwise, reinforcement
installed vertically may need to be longer to penetrate beyond the potential
height of the stable arch.
When the layering is dipping relative to an excavation, several different
failure modes are possible as shown in Figure 7.15. In the absence of crosscutting joints, cantilever beams are formed in the roof of the excavation.
Tensile stresses will form at the top of the cantilever, and cracks will form
near the abutment or shoulder. These cracks will penetrate the full depth
of the layer, and slabs of rock will fall into the excavation. This mode of
failure can be prevented by the installation of reinforcement to restrain the
free end of the cantilever. In the left wall, a toppling mode of failure may
occur, especially if blast damage (which is frequently observed) undercuts
the toe of the wall. This type of failure may be controlled by the installation for reinforcement angled upward to intersect the dipping layers. Mesh
support may be required if the layers are thin. If possible, the lowest row
of reinforcement should be installed horizontally to cross beyond the line
of intersection with the floor of the drive. Failure by sliding may occur in
Rock fall
FIGURE 7.14
Arch formation in layered rock masses.
Rock Reinforcement and Support
333
FIGURE 7.15
Flexural toppling and sliding in layered rock masses.
the right wall. This mode of failure may be controlled by the installation of
horizontal reinforcement that intersects the layers and improves the shear
strength. Reinforcement should not be installed parallel to the layers. Again,
mesh may be required for thin layers.
7.3.9 Ground Support for Layered Rock and Moderate Stress
For moderate stresses in layered rock, the failure mode may involve a
sequence of sagging, followed by buckling and eventual cracking, and
failure as shown in Figure 7.16. The initial bending (sagging) is initiated
by gravity forces. Following sagging, the induced stresses result in an
increase in bending moments at both the center of the span and at the
abutments. These bending moments result in tensile stresses at the lower
surface of the rock beam at the center of the span and at the upper surface
near the abutments. These tensile stresses result in crack propagation, and
eventually two distinct segments of the beam may form. This mechanism
Major
principal
stress
FIGURE 7.16
Buckling and cracking failure in bedded rock.
334
Geotechnical Design for Sublevel Open Stoping
of behavior is commonly referred to as the Voussoir beam, and its stability
is controlled by the strength and stiffness of the rock and the horizontal
stiffness at the shoulders of the drive (Diederichs and Kaiser, 1999). This
mechanism is most effectively controlled by attempting to improve the
shear resistance between the individual layers to form a thicker beam with
improved resistance to bending and buckling. The reinforcement system
resistance to shear for this mechanism is more important than the tensile
strength.
7.3.10 Ground Support for Layered Rock and High Stress
Figure 7.17 shows a layered rock mass in which the excavation is formed in
a lower stiffness rock than the layers above and below. Following extraction,
the less stiff rock will attempt to dilate more horizontally than the stiffer
rock. This differential dilation will result in shear stresses being developed
between the layers. This shear stress may result in vertical cracking in the
less stiff layers and tensile cracking and shear failures in the stiffer layers
near the shoulders and toes of the walls as shown.
The dilation of the walls in the less stiff rock can be controlled (but not
prevented) by the installation of horizontal reinforcement. The shear failures
and potential overbreak from the roof can be controlled by reinforcement
angled into the shoulders and vertically in the center of the span.
7.3.11 Ground Support for Jointed Rock and Low Stress
Jointed rock masses are characterized by the frequent occurrence of rock
discontinuities with variable persistence and spacing. The stability of blocks
in jointed rock is controlled by the forces acting on the blocks and the shear
strengths of the joints that form the faces. In many cases, at the time the
FIGURE 7.17
Tensile splitting, shearing, and sliding in bedded rock.
Rock Reinforcement and Support
335
FIGURE 7.18
Discrete large blocks falling or sliding from a rock mass.
excavation is formed, blocks are not fully formed; that is, the faces of the
blocks have intact rock bridges. The rock bridges may be strong enough to
maintain the stability of the blocks at this time. However, the changes in
stresses caused by the excavation may result in the preexisting discontinuities propagating through the rock bridges to create fully formed blocks. After
this time, the block stability is controlled by the orientations of the faces and
the shear strengths of the fully formed faces.
In a low-stress environment, the normal stresses acting across the joint
faces are low and therefore the frictional shear resistances are also low.
The shear stresses resisting sliding or falling are insufficient to prevent the
failure modes shown in Figure 7.18. In this figure, the discontinuities are
widely spaced and could be controlled by reinforcement installed normal to
the excavation faces. An estimate of the maximum block size is required
to enable an appropriate length of reinforcement element to be selected so
that it penetrates beyond the unstable block into a rock mass region that
is stable.
Alternatively, if the discontinuities are closely spaced as shown in
Figure 7.19, then surface support is also required to prevent unraveling and
progressive large-scale collapse. In a low-stress environment, mesh is sufficient to retain the volume of failed rock. However, as mesh does not provide immediate restraint to loosening, the volume of failure may be larger
and deeper than if shotcrete is used to provide immediate response to rock
mass loosening. This observation is an important factor when considering
the reinforcement length demand.
336
Geotechnical Design for Sublevel Open Stoping
FIGURE 7.19
Unraveling and progressive collapse of small blocks.
7.3.12 Ground Support for Jointed Rock and Moderate Stress
A jointed rock mass in a moderate stress field may behave similarly to a
massive rock. That is, the normal stresses acting across the joint surfaces
may result in the shear strengths being greater than the shear stresses acting. Whether this is the case or not depends on the orientations of the joint
surfaces relative to the orientations of the stresses. If the joints do not slide,
then tensile cracking may occur as shown in Figure 7.20. These tensile cracks
may coalesce and interact with the preexisting joints to form blocks that slide
and rotate and result in a general dilation of the rock mass and deformation
of the excavation boundaries.
Both support and reinforcement are required to control this type of rock
mass behavior. Shotcrete can be used for support together with reinforcement used for both restraint of the support and improved shear strength of
the joints.
FIGURE 7.20
Tensile cracking, crushing, sliding, and dilation.
Rock Reinforcement and Support
337
7.3.13 Ground Support for Jointed Rock and High Stress
As with moderate stress, a jointed rock mass may behave as a massive
rock. This again depends on the orientations of the discontinuities relative to the stresses. Consequently, the failure modes shown in Figure 7.21
are similar to those shown in Figure 7.12. However, the support and reinforcement requirements are different. Failure at the right shoulder would
result in a loss of horizontal confinement across the back, and the blocks
would then fail due to gravity loading. Failure at the toe of the wall would
result in undercutting of the blocks above and failure. Experience indicates
that mesh-reinforced (embedded) shotcrete would be required to provide
support with immediate (by the shotcrete) and postfailure (by the mesh)
response to rock mass deformations. Reinforcement would be used for both
restraint of the mesh-reinforced shotcrete and to stabilize rock discontinuities close to the excavation.
As indicated previously, the surfaces may be stable immediately following the creation of the excavation. However, in highly stressed rock,
cracks may gradually form and propagate with time. The area of cracks
per volume of rock may eventually exceed some critical value at which
time the rock will fail violently with fragments of rock ejected as shown
in Figure 7.22. The time after the formation of the excavation at which this
phenomenon occurs may range from seconds to weeks. These events are
therefore a major hazard in a mine, as ejection may occur before appropriate ground support is installed or may occur from the face of an excavation
during drilling and charging operations. Again, both mesh-reinforced
shotcrete and reinforcement are required in a rock mass susceptible to this
type of failure mode.
As with the other rock types and conditions described earlier, an excavation may be stable for the stresses acting locally. However, in a highly
stressed rock mass, there is a possibility of slip on discrete major structural
FIGURE 7.21
Crushing and spalling under high stress.
338
Geotechnical Design for Sublevel Open Stoping
FIGURE 7.22
Ejection of material due to stresses exceeding the strength of the rock at the boundary of an
excavation.
Vp
Vs
Vpp
Incident and reflected seismic waves
FIGURE 7.23
Detachment and ejection of a discrete block due to seismic waves from an event remote from
the excavation.
features some distance from the particular location of the excavation. As
indicated previously, these sudden, unpredictable slips can release energy
in the form of pressure and shear waves that eventually reach the excavation (Figure 7.23). These waves are reflected at the excavation boundary, and
stress changes result that may be sufficient to cause crack propagation, failure, and massive ejection of rock.
7.3.14 Design by Precedent Rules
Precedent rules can be applied in conjunction with the concepts detailed
in the previous sections. Brief details and discussion of the more common
systems in use are given in the following sections. It is left to the reader to
339
Rock Reinforcement and Support
examine the sources of information and to assess whether the particular rule
or system can be applied in any given circumstance.
Precedent rules, based on back analysis of reinforcement that was used
and found to be effective in civil engineering structures, were developed
in the late 1950s (Lang, 1961). Note that these rules do not take account of
the rock mass quality and the stress regimes in which the excavations were
formed. For reinforcement length and bolt spacing, the following rules have
been found to apply:
B
(B < 6 m)
2
(7.1)
B
H
(B > 18 m),† (H > 18 m)
4
5
(7.2)
L min = The largest of 2s, 2b,
or
L min = The largest of 2s, 2b,
L
smax = The smallest of or 1.5b
2
(7.3)
where
L is the bolt length
s is the bolt spacing
b is the mean block width
B and H are the excavation width and height, respectively
Various other suggestions for reinforcement length have been proposed as
follows (Rabcewicz, 1955; Pender et al., 1963; Benson et al., 1971; Cording
et al., 1971):
L = 0.3B
(7.4)
L = 1.829 + 0.0131B 2 , ≥ 3b
(7.5)
L = 0.25 to 0.30B
(7.6)
L = 0.35B
(7.7)
L = 0.1 to 0.5H
(7.8)
Choquet and Hadjigeorgiou (1993) presented a summary of the length estimates from various sources (e.g., Coates and Cochrane, 1970; Farmer and
Shelton, 1980; USACE, 1980; Laubscher, 1984). It is found that the predictions
of reinforcement lengths given by most of the expressions are reasonably
consistent and that more complicated expressions are not required. It is also
important to note that the actual required length of a reinforcement system
340
Geotechnical Design for Sublevel Open Stoping
will depend on the force demand and the load transfer mechanism. For
example, a CFC system (e.g., split set) will need to be longer than a DFMC
device (e.g., expansion shell anchor) to achieve the same force capacity.
For excavation crown pressure demand, Cording et al. (1971) suggested
that
Pc = nBg(kPa)
(7.9)
where
B is the crown span (m)
γ is the unit weight of the crown rock (30 kN/m3)
n is a constant ranging from 0.1 to 0.3, that is, for a span of 6 m; this formula predicts a crown pressure Pc in the range from ∼20 to ∼60 kPa
For walls, the pressure demand, Pw, is given by
Pw = mBg(kPa)
(7.10)
where
B is the wall span (m)
γ is the unit weight of the wall rock (30 kN/m3)
m is a constant ranging from 0.05 to 0.15
This equation indicates that, statically, the pressure demand for excavation
walls is about 50% of that for backs/roofs/crowns.
7.3.15 Design by Rock Mass Classification
The Q system (Barton et al., 1974; Grimstad and Barton, 1993) is probably
the most widely used rock mass classification system. However, it should
be used with caution, particularly in regard to some of the design expressions that have been developed. It is worthwhile noting that the database
was originally developed from case studies of civil engineering tunnels at
shallow depths.
The chart shown in Figure 5.7 is used to estimate ground support based
on the Q value and the span or height of an excavation surface. The Q system
incorporates relationships to estimate minimum reinforcement length. For
example, rock bolt lengths are estimated using
L =2+
0.15 B
ESR
where
B is the width or height of an excavation surface
ESR is the excavation support ratio (see Table 5.1)
(7.11)
341
Rock Reinforcement and Support
The value of ESR depends on the intended function of the excavation and
ranges from 0.8 for public infrastructure excavations to more than 1.5 for
mine excavations. As an example, for rock bolts in a 5 m by 5 m permanent development heading (ESR = 1.6), L = 2.5 m. This is in agreement with
the precedent mining practice. Note, however, that for individual blocks or
stress-driven failure, longer bolt lengths may be required.
Two formulae have been proposed as part of the Q system to calculate the
excavation roof/back/crown pressure demand:
200 Jn
(kPa) for Jn < 6 (0 to 2 sets)
3JrQ1/3
(7.12)
200
(kPa) for Jn > 6 (more than 2 joint sets)
JrQ1/3
(7.13)
Proof =
and
Proof =
In most rock masses, the Jn value will be greater than 6 and therefore
Table 7.3 shows the variation of Proof predicted using Equation 7.13 for rock
masses ranging in quality from very poor to good. Note that the SRF value
can range from 0.5 to 20 (see Figure 5.10) with corresponding large changes
in the Q value and predicted values of Proof.
A pressure demand of 312 kPa can be satisfied by twin strand cable bolts
(500 kN capacity) on 1.25 m by 1.25 m pattern. However, in poor quality rock,
this reinforcement would need to be complemented by a shotcrete layer to
retain the small block sizes. Similar calculations for other reinforcement systems can be made to satisfy the other pressure demands given in Table 7.3.
As the stress levels increase and the energy release accompanying failure
increases, it can be concluded that there is a need for increased force and
displacement capacities in both reinforcement and support.
TABLE 7.3
Examples of Roof Support Pressure as a
Function of Q Value
Rock Quality
Parameter
RQD
Jn
Jr
Ja
Jw
SRF
Q
Proof
Very Poor
Poor
Fair
Good
25
12
1
4
1
2
0.26
312
50
12
1
2
1
2
1.04
197
75
9
1.5
1
1
2
6.25
73
95
9
2
1
1
2
10.6
46
342
Geotechnical Design for Sublevel Open Stoping
The choice of an appropriate stiffness is an inherently difficult task when
based simply on a classification such as the Q system. A higher-stiffness element can arrest rock movement with less displacement. However, the penalty
is a higher force generated in the element. On the other hand, a low-stiffness
element allows for greater displacement but may not absorb the released
energy before the rock mass has significantly loosened to a point where serviceability requirements mean that failure has effectively occurred.
One way of estimating displacement demand may be simply to assume that
the stress change (from pre-mining to post-mining) occurs over a length (L)
observed for a particular mine site, and the rock mass deformation modulus
(Em) may be estimated from one of the several expressions available in the
literature (e.g., Serafim and Pereira, 1983; Hoek and Brown, 1997; Zhang and
Einstein, 2004; Hoek and Diederichs, 2006):
Em = 10
Em =
RMR-10
40
sc
10
100
(GPa) (for sc > 100 MPa)
GSI-10
40
(GPa) (for sc < 100 MPa)
(7.14)
(7.15)
where
RMR is defined by Bieniawski (1976)
GSI is the Geological Strength Index introduced by Hoek (1994) (see also,
Hoek et al., 1995)
The most up-to-date of these expressions is probably that due to Hoek and
Diederichs (2006) and given by Equation 4.21.
The displacement, δ, is then given by
d=
Ds
L
Em
(7.16)
For example, if the average stress decrease is 40 MPa in a rock mass with
E = 50 GPa over a depth of 2 m, δ = 1.6 mm. On the basis of experience, this
displacement is considered to be unrealistically low for a highly stressed
rock mass where loosening could be expected to accompany destressing.
An alternative approach is to use the plastic strain obtained from nonlinear stress analysis. If the plastic strain is assumed to be about 5% over a 2 m
depth, then the associated excavation wall displacement is about 100 mm.
Another approach might be to consider the depth of failure and the bulking associated with rock mass failure and dilation. For example, if the depth
of failure is observed to be approximately 1.5 m and the volume increase
associated with failure is assumed to be say 20%, then an excavation wall
would move about 300 mm. This magnitude of boundary displacement is
considered to be more reasonable.
343
Rock Reinforcement and Support
7.3.16 Reinforcement Layout
Several assumptions are implicit in the approach detailed in the preceding
sections:
1. Reinforcement is justified (in terms of both safety and production
requirements) and economically viable.
2. The reinforcement can be installed evenly spaced within the excavation surface associated with the failure volume.
3. The reinforcement will actually pass beyond the failure volume.
All these assumptions may usually be satisfied within most stope development excavations.
The average square spacing (s meters) can be determined from
s=
C
p
(7.17)
where
C is the reinforcement design capacity (kN), not necessarily the maximum
force capacity
p is the pressure demand (kPa)
For rectangular patterns
rs =
C
p
(7.18)
where
r is the spacing within a ring
s is the strike spacing of rings
7.3.17 Energy Release
Conceptually, rock fails violently when the unloading stiffness of the surrounding rock mass is softer than the unloading stiffness of the volume of
failing rock (Jaeger and Cook, 1976; Brady and Brown, 2004). It may be possible to precondition the rock mass so that these conditions do not occur. That
is, the intact rock needs to be damaged prior to the formation of the excavation so that these conditions do not occur. Preconditioning of the rock mass
has been used successfully at many mines (e.g., Board and Fairhurst, 1983;
Chacon et al., 2004).
344
Geotechnical Design for Sublevel Open Stoping
South African and Canadian workers have provided a number of examples of the range of velocities, typical masses, and kinetic energies that have
been measured or estimated for dynamic failure. For example, it has been
suggested that the kinetic energy is generally in the range 20–30 kJ/m2 with
a maximum velocity of 1.5–2 m/s and a displacement demand of about 150
mm. Other authors have suggested that kinetic energy may be up to 25 kJ/m2
with velocities of ejection of 2–3 m/s.
Ortlepp (1992) has inferred that block velocities after dynamic failure may
be considerably higher than these values, having measured an ejection velocity of about 7.5 m/s after a displacement of about 50 mm.
The data provided earlier can be used to design ground support schemes
that have the necessary energy and displacement capacities to survive violent rock mass failures. It is worth noting that the energy dissipation depends
on both the ability of the ground support to deform and the system force
capacity. Displacement is particularly important. For example, although a
reinforcement system may have large displacement capacities, it may cause
the rock mass to disintegrate to the point where the support system may
not be able to hold the broken rock. Systems that absorb large amounts of
energy, but allow large deformations are not really suitable for excavation
stability. The objective should be moderate, say 100–200 mm, reinforcement
displacement that is compatible with stable surface support systems (mesh
and shotcrete) at the boundaries of excavations.
7.3.18 Rock Mass Demand
The required force–displacement response and capacities of reinforcement should ideally be matched to the rock mass demand. This rock mass
demand may be applied directly from the rock mass or through the support that is retained by the reinforcement. In almost every case, this rock
mass demand is very difficult to quantify (the possible exception to this is
the reinforcement of a discrete fully formed block). On the other hand, the
demand may change with time for some rock masses. For example, a stiff
response may be required in the short term to minimize rock mass loosening, while in the longer term, the reinforcement system may be required
to absorb large displacements as the block size reduces and the rock mass
creeps (Figure 7.24). In this case, a single reinforcement system may not be
able to provide both the short- and long-term properties required to satisfy
the rock mass demand. This also applies to areas that may be susceptible to
sudden failure of the rock mass due to overstressing where the requirement
of the reinforcement system to absorb energy may be incompatible with
the short-term requirement to provide a stiff response to static rock mass
movement and the ability to sustain the displacements associated with rock
mass bulking.
Support demand is even more difficult to predict due to the fact that the
rock mass characteristics may change with mining and time. For example, a
Rock Reinforcement and Support
345
FIGURE 7.24
Observed damage near the boundary of an excavation in hard rock under very high stress.
massive rock mass may change to a broken rock mass following failure due to
overstressing (Figure 7.24). Initially, there is apparently no demand for surface support (or even for reinforcement). Following failure, there is a definite
need for surface support to retain the broken rock and the need for the support to be restrained by the associated reinforcement.
7.4 Rock Bolting of Open Stope Development Drives
A selection of typical reinforcing elements will be discussed in accordance
with the classification of reinforcement presented in Section 7.2. An understanding of the different elements is important, as no single reinforcing
scheme is likely to match the range of observed ground behavior at a particular mine site. This is because of the likely range of failure mechanisms
that can be experienced throughout a stope extraction process.
Reinforcement systems may be broadly characterized as rock bolts or
cable bolts according to their length. Rock bolts are generally less than
3 m long while cable bolts are longer than about 5 m. The mechanical
properties vary widely as do the installation requirements. In general,
rock bolts may be classified as one-pass or two-pass systems. It has been
found that one-pass systems are preferred in many mines. However, the
installation procedure may not be compatible with the requirement to also
restrain mesh, and the mechanical properties may not be appropriate for
346
Geotechnical Design for Sublevel Open Stoping
the expected rock mass demand (in terms of one or more of force capacity,
displacement capacity, or energy absorption).
7.4.1 Continuous Mechanical Coupled Rock Bolts
CMC rock bolts rely on a grouting element that fills the annulus between the
element and the borehole wall. The strength of the system is a function of
the nominal element capacity, the grout strength, and the active embedment
length. The coupling agent can be either a cement- or a resin-based grout.
7.4.1.1 Cement-Encapsulated Threaded Bar
A typical cement-encapsulated threaded bar consists of a 2–3 m long, 20–25
mm diameter corrugated bar that is grouted along its entire length. The bolts
are usually manufactured with a variable cross-sectional shape to provide
effective geometrical interference between the grout and the bolt surface.
The geometrical interference creates a mechanical interlock that extends over
the entire length of the element. Figure 7.25 shows a cross section through a
typical bolt and its components.
The critical embedment length for a typical water cement ratio of 0.35 is
approximately 0.5 m. A dense grout mix increases the bond strength, both in
the bolt-to-grout contact and in the grout-to-rock contact. Each bolt provides
long-term reinforcement exceeding 15 tons/m of embedment (Figure 7.26).
However, this depends upon the strength of the grout mix, with the main
cause of failure observed being slippage (shear failure) at the bolt–grout
interface (Figure 7.27).
Cement-encapsulated rock bolts can be used for long-term reinforcement
in areas where stress-related damage is expected, or where weathering
effects over a long period of time would make an ungrouted point-anchored
rock bolt unreliable. Experience also suggests that the system may be too
stiff to be used in rock masses likely to undergo large deformations or sudden movement. Cracking of the grout across a geological discontinuity may
cause corrosive damage to the rock bolts, due to water being able to reach
the steel bar, and sometimes the bar is galvanized prior to installation. In
addition, this rock bolt may be susceptible to blast damage (flying rock hitting the exposed fine thread at the plate end) when installed very close to
an active face. A coarse threaded bar can be used to overcome this problem.
However, coarse threads do not allow active restraint to be maintained and
can fail under dynamic loading (Player, 2012). The reasons for this are, first,
the short free lengths between the internal and external fixtures, which
means that small axial displacements result in larger strains (and accordingly larger force changes) than with a longer free length, and second, a
coarse thread has a large helix angle, which means that a nut can rotate more
easily than on a fine thread (e.g., standard metric thread) that has a smaller
helix angle.
347
Rock Reinforcement and Support
Hemispherical
plate
Nut welded with
hemispherical
washer
150 mm M20 thread
Right hand
33–45 mm hole
(grouted)
FIGURE 7.25
Typical components of a cement-encapsulated threaded bar. (Courtesy of Mount Isa Mines,
Mount Isa, Queensland, Australia.)
7.4.1.2 Resin-Encapsulated Threaded Bar
In cases where immediate reinforcement is required, resin-encapsulated
threaded bars can be used (Kaiser et al., 1996; Mikula, 2004; Varden, 2005).
In order to install rock bolts that are coupled with resin along their entire
lengths, it is necessary to insert multiple resin cartridges with sufficient volume of resin to fill the annulus between the rock bolt and the borehole.
The typical rock bolts being used in underground hard rock mines have
been modified from the rock bolts used in coal mining industry. The modifications have been necessary due to the need to drill larger hole diameters
with the type of equipment used in hard rock metalliferous mines. The modification is mainly in the form of paddles or the use of a spring welded onto
the end section of the rock bolts. Figure 7.28 shows the anchor sections for a
24 mm diameter Posimix bolt with a spiral arrangement and a 27 mm diameter Secura bolt showing a paddle arrangement. The Posimix spiral is 3 mm
in diameter and has a length of 500 mm. The Posimix system is designed
348
Geotechnical Design for Sublevel Open Stoping
1 m (double) embedment length
0.35 W/C–7 days strength
250
Load (kN)
200
150
100
50
0
0
10
20
Displacement (mm)
40
30
FIGURE 7.26
Typical load–displacement response for cement-encapsulated threaded bar.
Local
grout failure
Shear failure
Local crushing
Shear loading
Joint opening
Joint opening
Local crushing
Shear failure
Axial loading
FIGURE 7.27
Failure by slippage at the bolt–grout interface. (Courtesy of Mount Isa Mines, Mount Isa,
Queensland, Australia.)
Rock Reinforcement and Support
349
(a)
(b)
FIGURE 7.28
(a) Posimix and (b) Secura bolts showing spiral and paddle mixing arrangements, respectively.
to push the resin cartridge plastic to the end of the hole. Additionally, the
system allows the rock bolt to be centrally located in the hole allowing even
distribution and mixing of the resin. The Secura paddles are 29.2 mm wide
and are sheared into the end of the bolt for the purpose of mixing resin.
The introduction of mechanized resin-anchored bolting using jumbos has
been difficult to implement economically due to the high cost of resin transport and storage: Depending upon the local weather, underground temperatures, and humidities, this may require the use of refrigerated trucks and
surface and underground storage facilities. Other problems include speed of
bolt installation, including the ability to install mesh on a single pass, poor
matching of bolt diameter to jumbo-drilled hole diameters, as well as operator skills.
In the case of resin-encapsulated rock bolts, experience from in situ pull
testing shows that high transfer loads can be achieved over short embedment
lengths. However, cartridge resin systems may suffer from either underspinning or overspinning. Underspinning results in poor mixing and low resin
grout strength, often at the critical anchor end of the hole. In some cases, the
resin will never set. Overspinning during installation can result in shearing
of the partially cured resin. This results in a reduced bonded area and lower
load transfer. In addition, gloving of the rock bolts by the plastic packaging
may occur completely eliminating load transfer along the bolt axis (Mould
et al., 2004; Villaescusa et al., 2008).
The performance and ultimate capacity of a reinforcement scheme can be
affected by substandard installation practices. However, in CMC schemes,
faulty installations are difficult to detect, given that the only visible parts of
an installed element are the plate, nut, and a short length of the bolt indicating the orientation of installation with respect to an excavation wall. Thus,
for a fully resin-encapsulated threaded rebar, it is very difficult to determine
350
Geotechnical Design for Sublevel Open Stoping
FIGURE 7.29
Bolt overcoring showing negligible resin migration within large shear zones.
the actual bonded length (bolt encapsulation) along the entire axis of the
bolt. In addition, because the full steel capacity may be mobilized with very
short embedment lengths of good quality resin, pull testing of exposed collar lengths within a fully encapsulated element is almost meaningless. Pull
testing provides only an indication of resin effectiveness at the collar or at
the first (unknown) location along a rock bolt axis where the resin is working effectively. It provides only a definite indication of poor installation in
cases where the entire length of resin-encapsulated reinforcement fails at
well below its designed capacity.
Examination of the entire length of a fully encapsulated rock bolt in situ
can be achieved by overcoring of the reinforcement element (Hassell and
Villaescusa, 2005; Villaescusa et al., 2008). Rock bolt overcoring not only
allows the recovery of the element, but also provides a clear view of the surrounding rock mass and a better understanding of the rock bolt system/rock
mass interaction (Figure 7.29). It provides a range of information, including
location and frequency of geological discontinuities, overall rock mass conditions, bolt encapsulation, load transfer along the bolt axis, and corrosion
effects.
Overcoring in broken ground or shear zones shows that very little resin
migration occurs in jumbo-installed resin-encapsulated bolts. The resin simply fills the annulus between the bolt and the borehole. Because of its viscosity, the resin is unable to penetrate the rock mass fissures and voids.
In comparison, significant cement migration has been observed during overcoring of cement-encapsulated bolts in poor ground conditions
(Figure 7.30). The degree of rock mass interlocking using cement grout
is superior to that achieved by resin grouting or friction stabilizers.
Interlocking around an underground excavation has been suggested as
an important mechanism to allow the rock mass to be self-supporting
(Windsor and Thompson, 1993).
Figure 7.31 shows overcoring results for 27 mm Secura bolts installed
in basalt at the Bullant Mine near Kalgoorlie, Western Australia, using
351
Rock Reinforcement and Support
(a)
(b)
FIGURE 7.30
Broken rock mass interlocking (a) friction rock stabilizers and (b) cement-encapsulated rock
bolts.
250
Collar region
Middle region
Toe region
Relative residual load (kN)
200
150
100
50
0
Embedment length (300 mm)
Secura M27–35 mm hole
Secura M27–33 mm hole
300 mm embedment length
FIGURE 7.31
Load transfer variability along bolt axis for resin-encapsulated Secura bolts. (From Varden,
R., A methodology for selection of resin-grouted bolts, MEngSc thesis, Western Australian
School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia,
2005, 113pp.)
33 and 35 mm holes. Similar embedment lengths (300 mm) were tested.
The results show that similar strengths were found for the collar and toe
regions, with increased strengths for the middle regions where resin mixing appears to be more effective. The residual loads were measured at
15 mm displacement.
352
Geotechnical Design for Sublevel Open Stoping
7.4.2 Continuous Friction Coupled Rock Bolts
CFC elements rely on the load transfer resulting from friction between the
reinforcement element and a borehole wall. The actual strength per meter of
embedment length of a CFC element is limited by the radial prestress setup
during installation.
7.4.2.1 Split-Tube Friction Rock Stabilizers
Thin-walled 47 mm diameter galvanized friction stabilizers are extensively
used as reinforcement for stope development access. Such reinforcement elements generally have a nominal wall thickness of 3 mm and are mechanically installed using jumbos. This type of rock bolt consists of a hollow rolled
tube having a slot along its entire length, which is driven into a drilled hole
of smaller diameter. It relies on friction between the tube and the rock to
provide reinforcement (Figure 7.32).
Friction bolts are simple and quick to install, while standing up to blast
vibrations relatively well. However, they have a very low initial bond
strength per meter of embedment length. A capacity of approximately 4–5
tons/m of embedment has been established for 46–47 mm diameter elements
(Figure 7.33). This may be insufficient to guarantee effective reinforcement of
wedges, blocks, and slabs potentially formed within the immediate backs of
excavations.
The initial bond strength is developed during bolt insertion, where the
drillhole tolerance with respect to bolt diameter is small and is likely to control the available frictional forces along the bolt length. In soft ground, the
driving time to completely install a bolt is sometimes reduced indicating an
even lower initial bond strength per meter of embedment length.
Despite their low bond strength limitation and their susceptibility to corrosion (Hassell and Villaescusa, 2005), split-set bolts are used extensively
FIGURE 7.32
Schematic of the installation process for friction rock stabilizers.
353
Rock Reinforcement and Support
Spilt set bolts (SS46)
Ungrouted strength
14
12
Load (tonnes)
10
8
6
4
Thalanga mine
Stawell mine
Hilton mine 1991
Hilton mine 1996
2
0
0.0
0.5
1.0 1.5 2.0 2.5 3.0
Embedment length (m)
3.5
4.0
Ungrouted SS46
1 m of embedment length
8
7
Load (tonnes)
6
5
4
3
2
1
0
0
2
4
Deformation (mm)
6
8
FIGURE 7.33
Load transfer for fully coupled friction bolts. (From Villaescusa, E. and Wright, J., Permanent
excavation reinforcement using cement-grouted split set bolts, Proceedings of the AusIMM,
No. 1, 1997, pp. 65–69. With permission.)
354
Geotechnical Design for Sublevel Open Stoping
throughout the mining industry even for permanent back reinforcement in
blocky ground. This is because of the advantages that the system has to
offer. These can be listed as
1. Immediate reinforcement to the face where damage from development blasting is minimal.
2. Low-cost mechanized bolt and mesh installation with a minimum of
components.
3. Excavations can be meshed at a later date by installing a short friction bolt (having a smaller diameter) inside a previously installed
friction bolt element.
4. Rock bolts can be installed into partially collapsed holes, providing
reinforcement in poor ground conditions and reducing the number
of holes that require redrilling.
5. In some cases, corrosion resistance can be minimized with the use of
galvanized or stainless steel elements.
A disadvantage is that the load transfer for a split tube friction rock stabilizer is usually limited to values that are usually insufficient to mobilize
the force capacity of the element. This is particularly so if the borehole is
oversized. It is worth noting that an undersized borehole may cause yield
of the steel cross section. If shear occurs across the borehole in which a
split tube rock bolt has been installed, then sliding in the toe region may
be prevented, and the rock mass movement may be sufficient to cause the
welded ring to be sheared off with a loss of the plate at the collar. Also, it is
important to note that split tube bolts are susceptible to corrosion damage.
Figure 7.34 shows a number of overcored friction bolts ranging in age from
1 to 5 years. Laboratory pull testing was devised to investigate the loss of
frictional capacity due to corrosion over various embedment lengths in the
range of 250–500 mm (Hassell and Villaescusa, 2005). Figure 7.35 clearly
shows a loss of load-bearing capacity due to corrosion; the moderately corroded elements generally have twice the load-bearing capacity of their
highly and severely corroded counterparts.
7.4.3 Discrete Mechanical or Friction Coupled Rock Bolts
Discrete mechanical or friction coupled (DMFC) elements are point
anchored and rely on load transfer over a relatively short interval of their
total length. Chemical grouting is used to provide a frictional coupling element that in most cases is less than 500 mm in length. Mechanical coupling
is provided by expansion-type anchorages that are shorter than 200 mm in
length. External fittings such as face plates are an essential component of a
DMFC system.
Rock Reinforcement and Support
355
FIGURE 7.34
Overcored friction bolt elements, ready for pull testing. (From Hassell, R.C., Corrosion of rock
reinforcement in underground excavations, PhD thesis, Western Australian School of Mines,
Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 2007, 277pp.)
7.4.3.1 Expansion Shell Rock Bolts
These discrete frictional coupled elements consist of 16–25 mm diameter
steel bars (of varying lengths) that are installed with point anchor expansion
shells in conjunction with face plates (Figure 7.36). The tension to the bolts
is provided by tightening a nut hemispherical washer and a plate against
the rock on the exposed ends of the bolts. Mechanically anchored bolts are
capable of providing very reliable anchorage in hard rock applications where
the rock mass has a high uniaxial compressive strength.
One of the main disadvantages of mechanically anchored rock bolts is that
if the anchor slips or the rock breaks around the plate, the capacity of the
bolt drops to zero and the rock around the bolt can fail. In some cases, short
threaded lengths (at the plate end) make the tightening of the plate against
the rock very difficult to achieve, especially in uneven rock faces. The standard point anchor systems can be susceptible to corrosion and may not be
effective in heavily broken rock masses in which an anchor point cannot be
secured.
356
Geotechnical Design for Sublevel Open Stoping
60
Corrosion rate
Severe
High
Moderate
Light
Load (kN)
50
40
FB5a
FB1
FBx
30
20
FB6
FB8
FB3
10
0
0
2
4
6
Displacement (mm)
8
10
12
FIGURE 7.35
Galvanized 47 mm diameter friction bolts—400–500 mm embedment lengths. (From Hassell,
R.C., Corrosion of rock reinforcement in underground excavations, PhD thesis, Western
Australian School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia,
Australia, 2007, 277pp.)
Expansion shell
Bolt
Hemispherical plate
Hardened washer
Nut
FIGURE 7.36
Components of an expansion shell rock bolt. (Courtesy of Mount Isa Mines, Mount Isa,
Queensland, Australia.)
357
Rock Reinforcement and Support
Kanowna Belle mine
CT bolts 2.4 m long
Load (tonnes)
25
25
20
20
15
15
10
10
FW conglomerate
FW conglomerate
FW conglomerate
5
0
0
5
10
15 20 25 30 35
Displacement (mm)
Cannington mine
CT bolts 3 m long
Schist
Pegmatite
FW zinc
5
40
0
0
10
20
30
40
Displacement (mm)
50
60
FIGURE 7.37
Load–displacement responses for expansion shell–anchored bolts. (From Villaescusa, E. and
Wright, J., Reinforcement of underground excavations using the CT bolt, in E. Villaescusa,
C.R. Windsor, and A.G. Thompson, eds., Rock Support and Reinforcement Practice in Mining,
Proceedings of the International Symposium on Ground Support, Kalgoorlie, Western Australia,
Australia, 15–17 March, 1999, pp. 109–115. Rotterdam, the Netherlands, A.A. Balkema.)
The initial bolt installation can be mechanized to provide an immediate
reinforcement force of approximately 10–15 tonnes. However, because of
the short internal coupling, the actual point anchor strength is limited by
the strength of the rock around the borehole (Figure 7.37). Point-anchored
bolts tend to slip progressively due to blast vibrations when installed very
close to an active face. An initial tension of approximately 7 tonnes is often
used to reduce subsequent loosening due to blast vibrations.
7.4.4 Rock Bolts with Yielding Mechanisms
The term yielding has been introduced and accepted by others as the appropriate terminology for rock bolt systems that have high energy dissipation
capacities. Unfortunately, this term does not distinguish between systems
that involve true material yield of the element or sliding movement at the
anchored section of a rock bolt system. These high energy dissipation systems can be classified as follows:
1. Those involving mainly anchor slip relative to an internal fixture at
a force less than the yield strength of the element.
2. Those involving mainly material yield in a decoupled length
between discrete fixed anchors.
3. Those involving a combination of anchor slip and element yield.
In the interests of clarity and concentrating on principles rather than specific
products, it is worthwhile to review the rock bolts that have been developed
358
Geotechnical Design for Sublevel Open Stoping
to address problems associated with dynamic loading and the large displacement and energy-dissipation capacities required to maintain excavation
stability. This is an area of contemporary interest and development (Player
et al., 2004, 2009; Thompson et al., 2004).
An early attempt to improve load transfer for strand-based cable bolts, while
providing increased elongation between anchors, was reported by Schmuck
(1979). A similar system with decoupling of the strand between fixed anchors
was reported by Matthews et al. (1983) and demonstrated to be effective in
maintaining the stability of highly stressed open stope crown pillars. The
decoupling was achieved either by plastic sleeves or, more simply, by coating cable bolt strand with plastic paint. A recent development, the D-Bolt
(Li, 2010), can be considered to have evolved from these earlier ideas of using
the element elongation to dissipate energy between discrete fixed anchors.
Conway et al. (1975) tested a mechanical anchor that allowed for sliding
of a standard rock bolt through a fixed smooth bore die and reported that this
system was developed in South Africa by Ortlepp and Read (1970). Thus, the
Garford Solid Dynamic Bolt (Varden et al., 2008) and Roofex (Neugebauer,
2008) developed during the last decade can be considered to be commercial
products that have evolved from these much earlier ideas.
Another example of using element sliding relative to the internal fixture
was the cone bolt developed at the CSIR in South Africa (Jager, 1992). The
cone bolt is believed to be the first bolt designed to use a sliding mechanism
to dissipate energy. The bolt consists of a plain bar with an expanded cross
section at the toe end and a thread, nut, washer, and plate at the collar as
shown in Figure 7.38. The expanded cross section of the bar is designed to
provide resistance to pull out that is controlled by the strength and stiffness
of the cement grout that encapsulates the bolt within a borehole. The shaft of
the bolt is coated with saponified wax so that there is little or no resistance
to movement of the bolt relative to the cement grout. The initial prototypes
Decoupled (waxed)
length
Cement grout
Cone anchor
Cement grout
Smooth shaft with
wax coating
Thread, nut and
spherical base
washer
Steel domed
plate
FIGURE 7.38
Cone bolt anchor designed to pull through cement grout and increase displacement and
energy absorption capacities.
359
Rock Reinforcement and Support
were manufactured from 16 mm diameter bar, and the majority of testing
was performed on these bolts. Subsequent to the final development of the
original cone bolt, demand for higher-capacity elements resulted in a version
based on 22 mm diameter plain bar. It is believed that only limited testing
has been performed for this bolt (Player, 2012).
The design of the cement grout should be such that the anchor ploughs
(pulls) through the grout column at a force less than the yield strength of the
bolt. Both static and dynamic tests have shown that this is not the case with
strong grouts, and much of the elongation is stretch of the bar (Player, 2012).
It is therefore critical that both
• The cement grout properties are designed to ensure that the cone
pulls through the grout at a force less than the yield capacity of
the bar
• The equipment and procedures used for mixing and placing the
cement grout paste in a borehole result in consistent strength and
stiffness of the hardened cement grout
Figure 7.39 shows a number of dynamic testing results obtained by Player
(2012). The energy dissipation ranged from 10 to 60 kJ, with a 25 kJ dissipation
achieved at approximately 150 mm of displacement (Player, 2012).
More recently, a modified cone bolt was developed in Canada. This bolt,
like the original cone bolt, has an expanded end but is designed to be
300
Dynamic force (kN)
250
200
150
100
50
0
0
50
100
150
200
250
Deformation (mm)
300
350
FIGURE 7.39
Performance of 22 mm diameter cone bolts. (From Player, J.R., Dynamic testing of rock reinforcement systems, PhD thesis, Western Australian School of Mines, Curtin University of
Technology, Kalgoorlie, Western Australia, Australia, 2012, 501pp.)
360
Geotechnical Design for Sublevel Open Stoping
encapsulated with resin grout from a two-component cartridge that is mixed
during installation. The breaking of the cartridge and mixing of the resin are
aided by a flat paddle attached to the expanded end. The reported results
(e.g., Simser et al., 2002; Gaudreau et al., 2004) show that this bolt performs
either by gross anchor displacement or element extension, but sometimes by
a combination of both mechanisms. The fact that the bolt eventually breaks
suggests that the ploughing effect eventually ceases.
An important consideration for any high energy-dissipation strategy is
to test the full reinforcement system, including anchors, bolts, and plate/­
hemispherical nut assemblies together. Systems that dissipate large amounts of
energy, but allow large deformations are not suitable. The objective should be
to limit the reinforcement displacement, such that it is compatible with stable
surface support systems (mesh and shotcrete) at the boundaries of excavations.
In order to enable dynamic rock reinforcement design, the rock mass
demand in terms of the ranges of displacement and energy presented in
Table 7.2 has been combined with the WA School of Mines reinforcement
dynamic capacity database (Player, 2012). The suggested design chart for
rock reinforcement under dynamic loading is shown in Figure 7.40. For
each rock mass demand category (Table 7.2), the corresponding ranges of
displacement and energy were used to define a region (shown as a box)
that has been labeled low, medium, high, and very high energy demand.
For each region, the acceptable bolts should have similar displacement
compatibility, while providing higher energy dissipation. That is, for each
demand region, the recommended appropriate reinforcement would plot
within the green (design) region.
At this time, research on complete ground support schemes that include
compatible support and reinforcement systems in terms of displacement
compatibility is ongoing. Nevertheless, displacement at failure exceeding 300 mm is deemed very significant, given the typical bulking factors that follow dynamic rock mass failure at an excavation boundary
(Figure 7.41).
7.5 Cable Bolting of Open Stope Walls
Cable bolt reinforcement is used to stabilize large single blocks or wedges
formed in the backs and walls of stope development infrastructure. In
addition, cable bolts provide effective reinforcement of stope walls where
normal rock bolts would prove geometrically inadequate due to their short
embedment lengths. For stope wall reinforcement, the cable bolts are usually installed from drilling drives internal to a stope void. The main objective is to stabilize the rock mass around a stope before the stope is extracted.
As an alternative to installing cable bolts from stope drill drives, special
Energy dissipated (kJ)
0
10
20
30
40
50
0
Low
100
Medium
High
Very
high
200
300
400
Deformation at failure (mm)
Very significant
damage to surface
support
500
600
700
FIGURE 7.40
Design of rock reinforcement under dynamic loading. (Data from Player, J.R., Dynamic testing of rock reinforcement systems, PhD thesis, Western
Australian School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 2012, 501pp.)
2.4 m 550 MPa 20 mm threaded bar—T20
2.4 m 550 MPa 20 mm threaded bar—T20—no plate
2.4 m 550 MPa 20 mm threaded bar—T20
2.4 m 550 MPa 20 mm threaded bar—Secura T20—resin
2.4 m 550 MPa 23 mm threaded bar—Secura R27—resin
2.4 m 550 MPa 25 mm threaded bar—JTECH—resin-SE
3.0 m 550 MPa 20 mm threaded bar—T20—1.6 m centrally decoupled
mine nut
3.0 m 550 MPa 20 mm threaded bar—T20—1.6 m centrally decoupled
integrated nut/washer
2.4 m 550 MPa 20 mm threaded bar—T20—1.0 m centrally decoupled
Posimix bolt—resin
3.0 m 280 MPa 22 mm threaded bar—Saferock—four buffer
3.0 m 280 MPa 22 mm threaded bar—Saferock—two buffer
2.2 m 280 MPa 22 mm threaded bar—Saferock—HC (weak grout)
2.4 m 580 MPa 22 mm Garford solid yielding bolt version 1
2.4 m 580 MPa 22 mm Garford solid yielding bolt version 2
2.4 m 580 MPa 22 mm Garford solid yielding bolt version 2—resin
2.4 m 580 MPa 22 mm Garford solid yielding bolt version 2—resin
2.4 m 400 MPa 22 mm cone bolt >40 MPa grout
2.4 m 400 MPa 22 mm cone bolt >40 MPa LE grout
2.4 m 400 MPa 22 mm cone bolt >40 MPa HE grout
2.4 m 400 MPa 22 mm cone bolt 25 MPa grout
3.0 m Roofex 12.5 mm—cement grout
3.0 m 450 MPa D-Bolt 22 mm—cement grout
3.0 m Yield-Lok 17.2 mm—775 mm yield length—cement grout
2.6 m Cable bolt-A 15.2 mm –plain strand—2.0 m toe anchor rupture
2.6 m Cable bolt-A 15.2 mm –plain strand—1.5 m toe anchor
toe slid
2.6 m Cable bolt-A 15.2 mm—plain strand—0.6 m collar slid
3.4 m Cable bolt-A 15.2 mm—plain strand—1.7 m centrally debonded
3.4 m Garford yielding cable bolt - Version 2
3.0 m Cable bolt-C 15.2 mm—plain strand—two buffer LC
3.0 m Cable bolt-C 15.2 mm—plain strand—four buffer LC
3.0 m Cable bolt-C 15.2 mm—plain strand—damaged wire
2.4 m 47 mm split tube bolt—1.8 m average toe anchor
2.2 m Inflatable bolt—1.5 m average toe anchor
Failure by rupture
High-impact testing
ion
Reinforcement types
R
c
for
Re
in
s
de
nt
em
e
eg
ign
r
m
oc
k
d
an
em
ass
d
60
Rock Reinforcement and Support
361
362
Geotechnical Design for Sublevel Open Stoping
FIGURE 7.41
Example of extremely high rock mass demand, where reinforcement failure was followed by
mesh loading, rock mass bulking, and load transfer to other bolts.
drives can be developed around a stoping block solely for cable bolt installation. To decrease cost and increase the reinforcing effectiveness, such
horizontal drives are usually located at the same vertical horizon as the
drilling sublevels and 10–15 m away from a planned stope wall location
(Figure 7.42). However, special cable bolting drives are not normally used in
most open stoping operations.
6 m long bulbed cable bolts
22B1 S
Panel
CMS
section
FIGURE 7.42
Cable bolt reinforcement and resulting stope crown. (From Villaescusa, E. et al., An integrated approach to the extraction of the Rio Grande Silver/Lead/Zinc orebodies at Mount
Isa, in Singhal et al., ed., Proceedings of the Fourth International Symposium on Mine Planning
& Equipment Selection, Balkema, Calgary, Alberta, Canada, October 31–November 3, 1995,
pp. 277–283.
Rock Reinforcement and Support
363
7.5.1 Cable Bolt Reinforcement Mechanisms
The cable bolt reinforcement system is made up of four components
(Windsor, 2004):
• Rock mass
• Element (strands)
• Internal fixture (cement grout)
• External fixture (plate and barrel and wedge anchor)
Stope wall responses can be measured during stoping to develop a better
understanding of cable bolt/rock mass interaction (Bywater and Fuller, 1983;
Greenelsh, 1985; Hutchinson and Diederichs, 1996). Additionally, assessment of cable bolt reinforcement effectiveness can be based on visual interpretation of stope wall photographs (see Figure 1.13) and the survey of the
resulting stope voids (see Chapter 9). Oddie and Pascoe (2005) have reported
results for stope crowns at the Olympic Dam mine, where significant reductions in the resulting depth of failure were achieved with the use of cable
bolting (Figure 7.43).
The main mechanisms that apply during cable bolt reinforcement are as
follows:
• Application of compression to improve resistance against shear and
tension across preexisting geological discontinuities.
• Creation of a composite beam of several layers of strata (when the
cables are installed in bedded rock). The stability can be improved if
individual bands can be grouped together to form a much stronger
composite beam. Cable bolting can be used to minimize bedding
slip along strike and dip adjacent to the stope walls.
• Anchoring unstable zones to stable or solid ground, while providing
large retention capabilities.
• Minimization of large excavation deformations, arising in part from
rock mass relaxation at the mid-stope spans.
For open stoping, the stabilization process requires the implementation of
surface support and rock bolts to create a strong membrane along the walls
of the drilling drives. Cable bolt rings are spaced every 2–3 m, and rock bolts
can be installed between rings. The reinforced skin is tied into better-quality
rock further into the rock mass by the longer cable bolts (Figure 7.44). The
reinforcement length is typically taken as the depth of unstable rock around
a stope plus a specified length for anchorage. In practice, the length of a typical cable bolt length for stope wall reinforcement ranges from 6 to 10 m.
Cable bolt spacing is designed to provide a static capacity equal to the dead
0
10
20
30
40
50
60
70
80
0
10
30
40
20
Stope width (m)
50
HR = 4
HR = 6
HR = 8
HR = 10
HR = 12
HR = 14
HR = 16
Maximum
depth of failure (m)
0–5
5–10
10–15
>15
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
Stope width (m)
Cable bolted
50
HR = 4
HR = 6
HR = 8
HR = 10
HR = 12
HR = 14
HR = 16
FIGURE 7.43
Cable bolt reinforcement and stope crown performance at Olympic Dam mine. (From Oddie, M.E. and Pascoe, M.J., Stope performance at Olympic
Dam Mine, Proceedings of the Ninth Underground Operators’ Conference, Perth, Western Australia, Australia, March 7–9, 2005, pp. 265–272, The AusIMM,
Melbourne, Victoria, Australia. With permission.)
Stope length (m)
90
Unreinforced
Stope length (m)
100
364
Geotechnical Design for Sublevel Open Stoping
365
Rock Reinforcement and Support
Anchorage
zone
Cab
Unsupported
Rock
reinforced
zone
le bo
lt
Cable
bolts
Mesh
Rock
bolt
Initial
Final
shotcrete shotcrete Detailed view
layer
layer
FIGURE 7.44
Deep cable bolt anchorage of stope walls.
weight of the failed material. For twin strand cable bolts, the spacing is typically 1.5–2 m within each ring.
The underlying design philosophy is to increase the density of cable bolt
reinforcement within the exposed stope walls (Figure 7.45) in an attempt to
stabilize a surface band along the walls of the drill drives (Rauert, 1995). The
overall result is to minimize the deformation of the final exposed stope walls.
Zone of intense cable bolting
Unsupported span
FIGURE 7.45
Zone of intense cable bolting at a stope drill drive. (Courtesy of Mount Isa Mines, Mount Isa,
Queensland, Australia.)
366
Geotechnical Design for Sublevel Open Stoping
7.5.2 Cable Bolt Types
The cable bolts utilized in sublevel open stoping consists of a seven-wire,
stress-relieved, high-tensile steel strand with plain (round) wires. Six wires are
laid helically around a slightly larger diameter central (king) wire. The regular
15.2 mm diameter strand can be produced to provide a number of grades that
provide differing yield and ultimate load capacities. Standard single strands
have a minimum yield force capacity of 213 kN and a minimum breaking
force of 250 kN. Single or twin strand cables may be used for stope and bench
hangingwall reinforcement, while twin strand cables are used for permanent
back reinforcement. Figure 7.46 shows some of the typical cable bolting geometries used in the mining industry (Windsor and Thompson, 1993).
7.5.2.1 Plain Strand Cable Bolts
Plain strand cable bolts may or may not have a high rate of load transfer
(measured in terms of force per unit embedment length). This will depend
on the cleanliness of the strand prior to grouting and the quality of the
cement grout (Villaescusa et al., 1992, Figure 7.47). These cable bolts may also
suffer from a significant reduction in the rate of load transfer if the borehole
confinement (stress) reduces (Hyett et al., 1995). Consequently, plain cables
installed in areas where the rock mass deteriorates due to the mining process may fail by slippage without developing any significant loads before
failure. However, plain cables are very effective in supporting stope walls
(Bywater and Fuller, 1983; Villaescusa et al., 1992).
7.5.2.2 Modified Strand Cable Bolts
Two types of strands that have been modified to cause a variation in cross
section along their length are known as birdcaged and bulbed strands.
Longitudinal section
Single plain strand
Cross section
Twin birdcaged
Twin plain strand and spacers
Birdcaged—7 wire
Twin bulbed
Bulbed
FIGURE 7.46
Typical cable bolting geometries. (From Windsor, C.R. and Thompson, A.G., Rock reinforcement—Technology, testing, design and evaluation, in J.A. Hudson et al., eds., Comprehensive
Rock Engineering, vol. 4, 1993, pp. 451–484, Oxford, U.K., Pergamon.)
367
Rock Reinforcement and Support
300
250
Load (kN)
200
150
Double (1 m) embedment test
grout w/c ratio
0.30
0.35
0.40
0.45
0.50
0.55
100
50
0
0
5
10
15
20
25
30
35
Displacement (mm)
40
45
50
55
FIGURE 7.47
Influence of cement grout on the load transfer of single strand plain cables.
Both types of strands result in more effective load transfer between the
strand and the cement grout. The more effective load transfer is reflected
by the need for a shorter embedment length, in which to transfer the strand
capacity, and higher values for the force–displacement response stiffness.
Figure 7.48 shows a schematic example of a twin bulbed cable geometry
used for development excavation reinforcement in which the installed
bulb density is 4/m. This bulb density provides a stiff reinforcement likely
to minimize the movement of the reinforced blocks in the back of an excavation (Figure 7.49). The optimal bulb diameter ranges from 29 to 31 mm,
thereby facilitating the use of thick cement grouts that can effectively penetrate the bulbs.
Bulb diameter: 29–31 mm, overall cable length: 6 m ± 5 mm, tail length 0.5 m ± 5 mm,
0.5 m
Tail to be plated
Bulb density
4 bulbs/m
FIGURE 7.48
Schematic of twin strand bulbed cable used for back reinforcement in hard rock.
368
Geotechnical Design for Sublevel Open Stoping
600
500
Load (kN)
400
300
Double (1 m) embedment test
0.45 grout w/c ratio
200
2 b/m + plain
2 b/m + 1 b/m
100
0
2 b/m + 2 b/m
10
0
20
30
Displacement (mm)
40
50
FIGURE 7.49
Laboratory performance of twin bulbed strand.
7.5.2.3 Debonded Plain Strand Cable Bolts
A debonded plain strand cable bolt requires 0.6–1.5 m of bulbed strand at
the toe of the hole to establish an acceptable anchorage capacity (Figure 7.50).
The response of the anchor will be relatively stiff. However, the overall
response will not be stiff due to the extension of the free length between
the anchor and the collar. Therefore, to provide stiff, near-surface restraint
Fixed anchor
length
Borehole
Decoupled
length
Cement
grout
FIGURE 7.50
Debonded plain strand cable.
De-coupling
sleeve
Coupled length
Strand
Barrel and
wedge
anchor
Steel
plate
Rock Reinforcement and Support
369
to minimize rock mass loosening will require the installation of additional
stiff rock bolts. One possible advantage of decoupled strand is that it can
cope better with shear displacement across the axis of the borehole than
fully coupled strand and solid bar.
7.5.2.4 Cable Bolt Plates
A plain strand cable bolt generally requires a plate to be effective in retaining rock. Plates are required when it is not possible to ensure sufficient load
transfer near the excavations, especially when large-scale structures are
present (Figure 7.51). Bulbed strand should also be plated where possible,
but is more likely to be effective where it is not possible to get access to the
strand (i.e., cable bolts installed in stope hangingwalls prior to stoping).
The use of barrel and wedge anchors to restrain plates, straps, and
mesh in cable bolt reinforcing applications commenced in the early 1980s
in Australian mines (Thompson, 2004). Recent developments in cable bolt
design have meant an increased reliance on anchors being serviceable for
long periods of time, especially for applications where the strand is decoupled from the cement grout.
A barrel and wedge anchor is essential when a length of strand is decoupled at the collar. In twin decoupled strand cable bolts, it is necessary to have
barrel and wedge anchors on both strands to achieve the full capacity of the
system (Figure 7.52). It is also necessary to have a clear understanding and
appropriate procedures to ensure that anchors are installed correctly and
perform according to specifications (Thompson, 2004; Hassell et al., 2006).
FIGURE 7.51
Plain strand cables installed with no plates unable to retain unstable blocks in a stope crown.
370
Geotechnical Design for Sublevel Open Stoping
FIGURE 7.52
Barrel and wedge anchor on both cable bolt strands.
7.6 Cable Bolt Corrosion
Corrosion is one of the major factors determining which reinforcement type
can be used as the permanent support. Corrosion reduces the capacity and
life expectancy of ground support, creating a number of safety concerns
and operational difficulties in underground mining (Villaescusa et al., 2008,
Figure 7.53). Furthermore, corrosion has been found to be partly responsible
for 29% of all rock bolt failures and 25% of all cable bolt failures during rock
falls within the Australian mining industry (Potvin et al., 2001).
7.6.1 Corrosivity of Cable Bolt Strands
Cement-grouted cable bolts are capable of high load transfer capacity and
resistance to corrosion damage. This resistance is provided by the protective alkaline environment of the cement grout and the physical barrier it
provides from the surrounding environment. However, experience has
shown that corrosion begins once the cement grout barrier is removed.
This occurs by cracking of the grout column due to ground movement,
blast damage, or in sections where the element is exposed from inadequate
installation.
Rock Reinforcement and Support
371
FIGURE 7.53
Severely corroded cement-grouted cable bolt.
In an effort to better understand the response of cement-grouted strand to
corrosion attack following cracking of the grout column and infiltration of
groundwater, a number of laboratory experiments, including data collection
at eight Australian mines, have been reported by Hassell (2007). The research
concluded that at least a 2 mm crack width is needed before significant corrosion occurs. Variables such as pH, temperature, total dissolved solids (TDS),
dissolved oxygen, flow rate, and time were analyzed. A very good direct linear
relationship was found between the dissolved oxygen and the measured corrosion rates (Figure 7.54). Dissolved oxygen content was found to be directly
related to the temperature and the salinity of the water. Thus, with one parameter, three controlling variables can be taken into account (Hassell, 2007). The
good correlation between TDS and corrosion rate is partly due to the temperatures being similar and having a comparable effect on the corrosion rate.
In general, a reduction in the rate of corrosion over time was observed.
This is due to the corrosion products partly inhibiting further corrosion. This
rate becomes constant after a certain period of time, depending on the environmental conditions.
The rate of groundwater flow affects the corrosion rate by two processes.
Increases in the flow rate simultaneously increase the rate at which dissolved
oxygen is brought in contact with the steel surface. This provides more available oxygen for the electrochemical process, and thus higher rates of corrosion occur. Higher flow rates also increase the level of physical erosion of the
corrosion products and reduce the thickness of the partially protective cover
increasing the corrosion rate.
372
Geotechnical Design for Sublevel Open Stoping
Corrosion rate (mm/year)
1.4
Corrosion environment CR = 0.5528 (DO) – 0.9267
Mine D
R2 = 0.988
Mine G
1.2
Mine C
Mine F
Mine H
Mine A
1.0
0.8
0.6
0.4
0.2
0.0
0
1
2
3
Dissolved oxygen (mg/L)
4
5
FIGURE 7.54
Dissolved oxygen versus corrosion rates for a number of Australian mines. (From Hassell, R.C.,
Corrosion of rock reinforcement in underground excavations, PhD thesis, Western Australian
School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia,
2007, 277pp.)
Table 7.4 shows the corrosivity classification for groundwater-affected
hard rock conditions found in Australian underground mines as proposed
by Hassell (2007). The classification considers two factors in determining the
corrosivity of the groundwater: dissolved oxygen content as measured in
situ from a dissolved oxygen probe and the groundwater flow conditions as
illustrated in Figure 7.55. Uniform corrosion rates for HA300 grade steel can
then be estimated for different environments.
The classification provides a range of possible corrosion rates for a specific
dissolved oxygen content and groundwater flow. As the groundwater condition is obtained from qualitative observation rather than quantitative assessment, this variation in values is necessary. Projection of the corrosion rates
for measurements of dissolved oxygen less than 1.5 and greater than 4.5 is
uncertain due to insufficient data. The given corrosion rates are for uniform
corrosion only. However, it is appropriate to assume that pitting corrosion
will increase with higher rates of uniform corrosion. The classification does
not take into account the rock mass quality. It is assumed that if the classification is to be applicable, the reinforcement will intersect water-bearing discontinuities. In addition, the rock mass damage from stress redistribution is
373
Rock Reinforcement and Support
TABLE 7.4
Maximum Corrosion Rates for HA300 Steel in GroundwaterAffected Australian Hard Rock Mining Environments
Strong flow—large continuous water flow from a large fault or many
fractures
Dissolved oxygen (mg/L)
1–2
2–3
3–4
4—5
Corrosion rate (mm/year)
<0.12
0.12–0.36
0.36–0.58
0.58–0.8
Flowing—water flows from fractures
Dissolved oxygen (mg/L)
1–2
2–3
Corrosion rate (mm/year)
<0.09
0.090–0.225
3–4
0.225–0.365
Dripping—numerous drips and trickling of water from fractures
Dissolved oxygen (mg/L)
1–2
2–3
3–4
Corrosion rate (mm/year)
<0.06
0.060–0.105
0.105–0.160
4–5
0.365–0.50
4–5
0.16–0.20
Wet—rock mass discolored. Dripping from fractures moderately common
Dissolved oxygen (mg/L)
1–2
2–3
3–4
4–5
Corrosion rate (mm/year)
<0.04
0.040–0.075
0.075–0.100
0.10–0.12
Damp—rock mass is discolored from dry rock mass. Very minor drips
Dissolved oxygen (mg/L)
1–2
2–3
3–4
4–5
Corrosion rate (mm/year)
<0.02
0.020–0.030
0.030–0.040
0.04–0.05
Source: Hassell, R.C., Corrosion of rock reinforcement in underground excavations, PhD thesis, Western Australian School of Mines, Curtin
University of Technology, Kalgoorlie, Western Australia, Australia,
2007, 277pp.
expected to increase the permeability within the zones where reinforcement
is utilized.
Approximate minimum and maximum service lives have been measured
from corrosion chamber experiments (Hassell, 2007). The service life is estimated as the material loss required to cause failure of the strand loaded to
175 kN or approximately 17.5 tonnes, a 30% decrease in the original capacity
of 250 kN. Groundwater is assumed to be present, and it is assumed that
either cracking of the grout column has occurred or grout encapsulation is
poor. Comparing the measured service lives to the corresponding corrosion
rates of the simulated environment calculated using the corrosivity classification, estimates can be made to the expected minimum and maximum
service lives (<17.5 kN) of 15.2 mm diameter black strand across a range of
corrosion rates as shown in Figure 7.56.
It is estimated that even in the most corrosive conditions observed in
underground mines, cable strand will last at least 1 year once cracks along
the element axis have formed. This figure is much higher than the expected
life of uncoated barrel and wedge anchors, which is found to be approximately 7 months at comparatively corrosive conditions (Hassell et al., 2006).
374
Geotechnical Design for Sublevel Open Stoping
0.7
0.6
Corrosion rate (mm/year)
Strong flow
(corrosion
chambers)
Groundwater
Flowing (chambers)
Flowing (mine site)
Dripping (mine site)
Wet (mine site)
Damp (mine site)
0.5
0.3
Flowing (assumed
boundary between
strong flow and
flowing)
0.2
Flowing (mine site)
Dripping (mine site)
0.4
0.1
0
Wet (mine site)
Damp (mine site)
1
1.5
2
2.5
3
3.5
4
4.5
Dissolved oxygen (mg/L)
FIGURE 7.55
Rate of corrosion in coupons grouped by groundwater flow. (From Hassell, R.C., Corrosion
of rock reinforcement in underground excavations, PhD thesis, Western Australian School of
Mines, Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 2007, 277pp.)
7.6.2 Corrosivity of Cable Bolt Anchors
Corrosion of barrel and wedge anchors and the consequences for the entire
cable bolt performance are poorly understood, despite the common use of
cable bolts in Australian underground mines since the 1970s. The use of
barrel and wedge anchors to restrain plates, straps, and mesh in cable bolt
reinforcing applications commenced in Australian mines in the early 1980s
(Thompson, 2004). Anchor failures after short time durations and under
low loads have been observed in several underground mines (Figure 7.57).
Failure is often characterized by the barrel and wedge remaining intact after
being found on the floors of drives with no evidence of strand rupture.
The corrosion of barrel and wedge anchors is intrinsically linked to the
environment in which they are installed. Circumstances in which groundwater is flowing or dripping over the exposed end of the reinforcement are
considerably more corrosive than dry environments.
In an attempt to better understand the behavior of cable bolt anchors,
various barrel and wedge anchor configurations were placed within a
corrosion chamber to simulate underground environmental conditions.
Laboratory pull tests were used to determine the force–displacement
responses and the influence of corrosion on the load-bearing capacities of
375
Rock Reinforcement and Support
1200
Esti
1000
mat
ed m
axim
um
ser v
ice
life
Mine environment
Mine G
Mine D
Mine C
Mine H
Mine F
Mine A
Time (days)
800
Esti
600
mat
ed m
inim
um
ser v
ice
life
400
200
0
0
0.1
0.2
0.3
0.4
0.5
Corrosion rate (mm/year)
0.6
0.7
FIGURE 7.56
Service life estimates for cable strand in strong groundwater flow Australian mining environments. (From Hassell, R.C., Corrosion of rock reinforcement in underground excavations,
PhD thesis, Western Australian School of Mines, Curtin University of Technology, Kalgoorlie,
Western Australia, Australia, 2007, 277pp.)
the anchors (Hassell et al., 2006). A number of corrosion protection methods were trialed. The methods used included galvanizing of the barrel as
well as three simple and common corrosion inhibitors: grease, bitumen, and
wax. Some tests were conducted with galvanized strand. The samples were
placed in a corrosion chamber with some anchors left outside the chambers
for noncorroded reference testing. Subsequent testing of the samples was
undertaken after 3, 7, and 10 months of exposure (Figure 7.58).
The experiments showed that failure occurred at the wedge/strand interface with the strand pulling through the anchor and was associated with
small wedge movement relative to the barrel. Importantly, failure took place
at significantly lower loads than strand force capacity, ranging from 22 to
111 kN (Figure 7.59). The internal section of the failed barrel and wedge
anchor shown in Figure 7.60 displayed a buildup of corrosion products on
the internal surface of the barrel together with shearing of the wedge teeth.
376
Geotechnical Design for Sublevel Open Stoping
FIGURE 7.57
Barrel and wedge slippage failure at low load due to corrosion.
FIGURE 7.58
Hemispherical barrel and three-part wedge anchor with compact strand before and after
placement in corrosion chambers.
Corrosion products on the internal surface of the barrel increase the frictional resistance at the barrel/wedge interface preventing sliding of the
wedge relative to the barrel. This in turn prevents the wedge from gripping
the strand. That is, the increase in normal force that results from wedge slip
does not occur, and this means that load must be transferred by the shear
resistance of the wedge teeth. It will be shown that this area loaded in shear
is very small, and the result is shear failure of the teeth. This allows the
377
Rock Reinforcement and Support
250
Typical anchor
strength (design)
Load (kN)
200
150
100
Failed anchor
50
0
0
5
10
15
20
Displacement (mm)
25
30
FIGURE 7.59
Performance of hemispherical barrel and three-part wedge anchor after 7 months in corrosion
chamber.
(a)
(b)
FIGURE 7.60
Internal condition of (a) failed and stable (b) barrel and three-part wedge anchor.
378
Geotechnical Design for Sublevel Open Stoping
strand to slip at loads significantly less than the design capacity associated
with the tensile strength of the strand.
Anchors that were coated with grease, wax, or bitumen had significantly
fewer instances of failure (Hassell et al., 2006). Consequently, it is recommended that barrel and wedge corrosion protection systems such as a long
life lubricant at the barrel/wedge interface and barrier coatings are applied
following installation.
7.7 Cement Grouting of Cable Bolts
Grouting is the procedure by which a hole drilled into the boundary of an
excavation is filled with a cement paste to set the reinforcement element hard
inside the rock mass. This allows load transfer from a potentially unstable
section of the rock mass (at the excavation boundary) to a stable section
deep into the rock mass through the reinforcing element as described earlier for the load transfer concept (Section 7.2.4). The strength of the grout is
critical in order to minimize the length of embedment needed to mobilize
the ultimate steel tendon capacity of a particular reinforcement system. In
general, failure by slippage at the steel–grout interface will be experienced
when weak grouts are used. Alternatively, rupture of the tendons can be
envisaged when using thick, strong grouts in conjunction with stiff reinforcement systems.
7.7.1 Collar to Toe Grouting
Collar to toe grouting methods were developed in conjunction with pistonbased cement grouting pumps. The water/cement (w/c) ratio for such cement
grouts can range from 0.40 to 0.55. This method requires a breather tube,
generally of 13 mm (inside diameter), to be attached to a cable bolt element
before it is installed into a hole. A permanent collar packing is placed at the
hole collar for the wet grout to be kept inside the hole. A short grouting hose
of approximately 1.0 m in length is also placed permanently in the collar
of the hole. The grout is pumped through the grouting hose, and when it
reaches the upper end of the breather tube, it begins to flow back through
this tube. When full flow of grout emerges from the end of the breather tube,
it indicates that full encapsulation of the steel tendon has been achieved
(Figure 7.61).
A typical piston-based pumping system usually requires that the mixing
and pumping are carried out within the same container. This is called a onestage grouting system. Mixing of the grout is achieved using paddles that
rotate around a vertical axis (Figure 7.62). This may create settlement of the
cement particles into the area where the pumping is taking place, that is, the
379
Rock Reinforcement and Support
Ungrouted
end (15 cm)
Steel spring
anchor
Permanent breather
tube taped to
cable (13 mm ID)
Fully encapsulated
reinforcement
Grout hose
(20 mm ID)
Collar
packing
Tail to
plate
FIGURE 7.61
Conventional collar to toe grouting method.
Grout
pump
Grout mixer
Valve
Delivery
hose
Mixing
drum
Pumping
chamber
Grout
flow
FIGURE 7.62
A conventional one-stage piston-based grouting system.
380
Geotechnical Design for Sublevel Open Stoping
bottom of the mixing tank, potentially reducing pumpability if the grout is
too thick. Having a single container can lead to changes in water/cement
ratios during the grouting operations.
In addition, no accurate devices to measure the amount of water being
poured into the mixing tank are fitted to most conventional one-stage
grouting systems. Consequently, following an initial mix design in which
the water cement ratio was probably correct, additional water can be added
(while still pumping and grouting) before a corresponding amount of
cement is added to the mix. This problem can be avoided if each mix is
pumped separately.
7.7.2 Toe to Collar Grouting
Toe to collar grouting consists of inserting a cable bolt without a breather
tube into the hole to be grouted (the need for the collar plug is also eliminated), followed by the subsequent grouting of the cable by means of a selfretracting grouting hose (Figure 7.63). The grout pushes the grouting hose
out of the hole as the grouting process is undertaken. The optimal grouting
rate is such that a self-retracting hose should be in minimal contact with an
advancing grout paste inside the hole. In order to achieve this, the typical
water/cement ratios required usually range from 0.32 to 0.35.
This method of grouting provides many advantages, including a much
faster initial cable placement and pregrouting preparation times, savings on
materials, faster rates of grouting, and a significant increase in grout strength.
In some cases, increased strength of the grout may effectively decrease the
Thick grout
Self-retracting
high-pressure
hose (>19 mm ID)
Initial set-up:
no breather tube,
no collar plug.
Fully encapsulated
reinforcement
Hose pushed out by the grout.
Uniform grout flow ensures
reinforcement encapsulation.
FIGURE 7.63
Toe-to-collar cement grouting method.
Final set-up: hose pulled out
(only 30 cm of hose in contact
with grout at all times).
Rock Reinforcement and Support
381
required embedment length to achieve the nominal steel failure capacity.
Strong and thicker grouts do not leak into voids and crevices encountered
along the drillhole axis, thereby minimizing wastage of cement.
This technique has been implemented successfully mainly because of the
use of modern two-stage grout pumps that allow a high degree of quality
control on the mixing, pumping, and water/cement ratios used. Two stage
means that the machine has separate mixing and pumping containers that
can be operated simultaneously or independently, thereby significantly
increasing productivity (Figure 7.64).
The ability to mix independently of pumping allows a constant water–
cement ratio to be achieved throughout a grouting operation. This enables
a high degree of quality control, as an accurate water meter allows water
addition to be controlled to a precision of one-tenth of a liter. Cement, water,
and additives are mixed in a horizontal paddle mixer and then discharged
into the lower hopper where a variable speed drive coupled to a rotor–stator
pump (mono pump) discharges the grout at the desired rate.
A disadvantage of this method is the potential for poor cable bolt encapsulation that may result if the grouting operator retrieves the hose while grouting is being done. Consequently, in order to avoid potential encapsulation
problems, the grouting hose can just be left in place (with no retracting) and
cut when the grout reports to the collar of the hole.
Toe to collar grouting is also carried out during mechanized installation of cable bolts using a cable bolter (Figure 7.65). The process consists of
drilling of holes followed by cement grouting using a self-retracting hose.
FIGURE 7.64
A two-stage mono pump cement grouting machine.
(e)
(d)
(f)
(c)
FIGURE 7.65
Mechanized toe to collar grouting method using the Tamrock Cabolter. (a) General view of Cabolter, (b) drilling of holes, (c) grouting of holes, (d) cable
bolt rill, (e) grout mixer, and (f) cable bolt inserter.
(b)
(a)
382
Geotechnical Design for Sublevel Open Stoping
383
Rock Reinforcement and Support
The rate of grouting and hose retraction is mechanized, thus ensuring that
no gaps are left along the axes of the holes. Once the holes are grouted, the
cable bolts are inserted into the holes that are full of grout. The cable bolts
are then mechanically cut in place with a tail left to be plated at a later stage.
Regardless of the grout method used, several issues require consideration during the selection of the most appropriate grout mix design to suit
a particular operation. The volume of grout that can be efficiently mixed
and the ability of a machine to mix the required water/cement ratio in a
reasonable amount of time must be considered. In general, the use of additives is recommended for efficient grouting of cable bolts. Additives prevent
segregation of water and cement, while reducing grout shrinkage during
curing. Preventing water–cement segregation at the toe end of the hole is
very important in order to achieve the required anchorage according to the
load transfer concept.
7.8 Support Systems
As with reinforcement systems, the types of support systems are presented
in generic terms and discussed in terms of the parameters related to materials and dimensions. All steel-based products (i.e., plates and mesh) may
be supplied with a zinc (galvanizing) coating, which should be specified to
be consistent with the rock bolts used for restraint. The precise effects of
galvanic reactions between dissimilar metallic surfaces are unknown but
qualitatively are known to accelerate the loss of the zinc coating.
7.8.1 Plates
Plates may be supplied as flat or proprietary domed plates or as large
deformed profile and combination plates (Figure 7.66). The dome in a plate
serves several purposes—to increase bending stiffness compared with a flat
(a)
(b)
(c)
FIGURE 7.66
Typical plates used for ground support. (a) Flat plate, (b) dome plate, and (c) combination plate.
384
Geotechnical Design for Sublevel Open Stoping
plate of the corresponding thickness, to account for nonperpendicular alignment with the rock face, and to facilitate the use of spherical washers.
The dome also provides a small level of positive restraint to the rock face
and/or mesh beneath the plate. This compensates for small relative movements between the end of the rock bolt and the rock, which tend to result in
a decrease in rock bolt tension. Flat plates are generally thicker (and therefore stiffer in bending) than the deformed profile plates. The use of plates
exceeding 300 mm by 300 mm in conjunction with mesh is questionable
other than to increase the number of mesh wires restrained by the plate and
for the prevention of premature rupture of the wires by the sharp edges of
flat plates.
7.8.2 Straps
Straps cannot generally be installed to provide any active restraint to the
rock mass between the reinforcement. The possible exception to this statement is in the case of the convex rock surfaces associated with corners of
intersections, pillars, or stope brows (Figure 7.67). In these cases, the straps
can be installed to be in contact with the rock between the reinforcement.
(a)
(b)
FIGURE 7.67
Strap support in conjunction with reinforcing elements. (a) Across structural features and
(b) across a stope brow.
Rock Reinforcement and Support
385
Straps may comprise either W-profile or mesh and should be installed across
the smaller excavation span, that is, across rather than along a drive width.
7.8.3 Mesh
Steel wire mesh is a key component of the ground support required to maintain the load-bearing capacity of a rock mass near the boundaries of an
underground excavation (Villaescusa, 1999b). While rock bolts are likely to
control the overall excavation stability through keying, arching, or composite beam reinforcement actions, mesh is installed to retain small, loose pieces
of rock or shotcrete that maybe detached within a bolting pattern. Rock mass
deterioration within a bolting pattern can arise from intense jointing, blast
damage, weathering, or excessive tangential stress changes. Mesh support is
effective in building up a back pressure to inhibit further slabbing within a
bolting pattern. Mesh loading mechanisms can be either uniformly distributed loading forces as in rock bulking, or point loading by loose individual
rock blocks (Figure 7.68). Ultimately, the role of mesh is to respond to significant inward movement of the rock mass surrounding an excavation and to
transfer the load to the reinforcement systems (Thompson et al., 2012).
Steel wire mesh for ground support is available in various configurations. The most common types are welded mesh, consisting of straight wires
arranged in a rectangular or square grid and welded together, and chain
link mesh that consists of regularly bent wires that are woven together and
interconnected mechanically (Figure 7.69).
The welded mesh may have different wire diameters at different spacings and be supplied in various sheet sizes. The most common configuration
consists of 5.6 mm diameter wires spaced at 100 mm centers. The wire may
have a smooth or deformed profile. These configurations of mesh used for
surface support in mines have changed little in the last 25 years or more. The
changes (in Australia) have been driven by civil engineering applications
and not the mining industry. The changes have mainly been associated with
FIGURE 7.68
Mesh support in highly stressed rock masses.
386
(a)
Geotechnical Design for Sublevel Open Stoping
(b)
FIGURE 7.69
Different types of mesh configurations. (a) Welded mesh and (b) woven (chain link) mesh.
material properties (i.e., yield and ultimate force capacities and elongation
capacity) and wire diameters and surface condition (i.e., smooth or deformed
wire). The deformed wire has better load transfer capacity than the smooth
wire when embedded in concrete slabs. This is also a consideration for meshreinforced shotcrete, depending on the sequence of mesh then shotcrete or
shotcrete then mesh.
Sheets are generally 2.4 m wide (the maximum that may be specified) with
variable lengths, commonly 3.6 m and up to 6 m. Larger sheets generally
cause handling and placement problems. The mechanical handling and
installation of welded mesh are shown in Figure 7.70.
In the past, rolls of weld mesh or woven mesh required manual installation from either a scissor lift or an IT basket (Figure 7.71). More recently,
an automated roll mesh handler for the application of high-tensile chain link
mesh was developed and successfully tested in Australia (Coates et al., 2009).
The handler is compatible with all commercial multiboom jumbo drilling
equipment, applying mesh from a cassette system. The mesh handler with
the mesh roll is mounted on one boom and the drilling/bolting element
mounted on the other boom allowing the application and pinning of the
high-tensile mesh simultaneously (Figure 7.72).
The length of the woven mesh can be cut to suit the width of the opening
being supported. A perceived problem with this mesh is that it may unravel
when one wire is broken. However, this does not appear to have been a
problem at mines such El Teniente Mine in Chile. A mesh manufactured by
Geobrugg in Switzerland overcomes some of the problems usually associated with woven mesh rolls. The Geobrugg mesh is an assembly of highstrength wires that results in mesh that is stiff across the width but can be
rolled in the other direction (Roth et al., 2004).
7.8.3.1 Mesh Testing
In the assessment of any ground support system, the relationship between
displacement, force, and energy must be assessed in relation to the expected
387
Rock Reinforcement and Support
(a)
(b)
(c)
FIGURE 7.70
Mechanical handling and installation of weld mesh. (a) Surface storage, (b) decline transportation, and (c) jumbo installation.
(a)
(b)
FIGURE 7.71
Manual installation of weld and woven mesh. (a) Scissor lift and (b) IT basket.
388
Geotechnical Design for Sublevel Open Stoping
FIGURE 7.72
Mechanical installation of woven mesh. (From Coates R. et al., Fully mechanized installation
of high tensile chain link mesh for surface support in tunnels, in P. Dight, ed., Proceedings of the
First International Conference Safe and Rapid Development, May 6–7, 2009, pp. 165–172, Australian
Center for Geomechanics, Perth, Western Australia, Australia.)
ground reactions. The energy absorption is a function of force and displacement. Displacement is influenced, sometimes significantly, by the number of
failures within a sample. For this reason, analysis of the mesh types reported
here has been undertaken at rupture. Rupture may or may not correspond
to the peak force achieved during a test, but the variability of a sample once
rupture has occurred means that detailed analysis with strong conclusions
cannot always be achieved.
The results presented here have been obtained using the Western
Australian School of Mines (WASM) static and dynamic testing facilities
for ground support elements (Player et al., 2008; Morton, 2009). The WASM
static facility consists of two steel frames: a lower frame used to support the
sample and an upper frame used to provide a loading reaction (Figure 7.73).
A mesh sample (1.3 m by 1.3 m) is restrained within a stiff frame that rests
on the support frame. The boundary conditions attempt to simulate the continuation of the material beyond the limited sample boundary. The restraint
system consists of high-tensile bar, eye nuts, and D shackles passing through
a perimeter frame at allocated points to simulate a number of boundary conditions. A screw feed jack is mounted on the reaction frame. The screw feed
jack is driven at a constant speed (4 mm/min) and allows large displacements to be imposed on the mesh. Load is applied to the mesh through a
spherical seat, to a 300 mm2, 35 mm thick hardened steel plate. The force is
measured using a 50-tonne load cell mounted behind the loading point. Data
acquisition is undertaken at two samples per second. Testing of a sample can
take up to an hour to complete.
The WASM dynamic test facility for mesh is shown in Figure 7.74.
Samples are loaded using the momentum transfer concept (Player et al.,
2004; Player, 2012). The mesh testing frame is bolted to a drop beam while
the 1.3 m by 1.3 m mesh sample is held in place using threaded bar, shackles, and eye bolts in the same configuration as the static test arrangement.
389
Rock Reinforcement and Support
Load
bearing
beam
Instrument
panel
LVDTs
Load shaft
Load cell
Sample frame (1.4 m × 1.4 m )
FIGURE 7.73
Details of WASM static test facility for surface support elements. (From Morton, E.C., Static
testing of large scale ground support panels, MSc thesis, Western Australian School of Mines,
Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 2009, 250pp.)
A loading mass is placed into the center of the restrained mesh. The loading mass consists of a pyramid-shaped bag filled with a known mass of
steel balls (0.5 or 1 tonne). The loading area of the bag is 650 mm × 650 mm.
A wooden prop is placed between the loading mass and the drop beam
to prevent the mass floating during the initial free fall period. The drop
beam and attached assembly are dropped from a specific height to generate dynamic loading on the sample. Computer software, advanced instrumentation, and a high-speed camera are used to record the test data. Data
acquisition is undertaken at 25,000 samples per second. Testing is completed in less than a second.
7.8.3.2 Mesh Force and Displacement
The failure mechanism of welded wire mesh is a measure of the mesh
quality. Three different welded wire mesh failure modes have been identified during laboratory testing. These can be described as shear failure
at the weld points, failure at the heat-affected zone (HAZ), and tensile
failure of the wire (Figure 7.75). Failure at the weld is an indication of the
weld technology and quality control (dirty electrodes or dirty wire) during mesh manufacturing. Failure at the HAZ is caused by weakening of
the wire during the welding process due to excessive weld head pressure
390
Geotechnical Design for Sublevel Open Stoping
Helicopter release hook
Drop
beam
Buffers
Loading
mass
Sample frame (1.4 m × 1.4 m )
FIGURE 7.74
WASM dynamic test facility for surface support elements.
(a)
(b)
(c)
FIGURE 7.75
Welded wire mesh failure mechanisms. (a) L–R tensile wire failure, (b) weld failure, and (c) failure of the wire through the HAZ. (From Morton, E.C., Static testing of large scale ground support panels, MSc thesis, Western Australian School of Mines, Curtin University of Technology,
Kalgoorlie, Western Australia, Australia, 2009, 250pp.)
391
Rock Reinforcement and Support
300
160
Woven (chain-link) mesh
140
Dynamic force (kN)
Static force (kN)
120
100
80
60
40
Welded mesh
(a)
0
100
200
300
Static displacement (mm)
200
150
100
Welded mesh
50
20
0
Woven (chain-link) mesh
250
0
(b)
0
100
200
300
Dynamic displacement (mm)
FIGURE 7.76
(a) Typical static and (b) dynamic reactions for welded wire mesh and woven wire mesh.
(From Villaescusa, E. et al., A database of static and dynamic energy absorption of mesh for
rock support, Proceedings of the 2012 Australian Mining Technology Conference, CRC Mining,
Perth, Western Australia, Australia, October 8–10, 2012b, pp. 27–34.)
and temperature, while tensile failure of the wire is controlled by the wire
manufacturing process. For ground support, the preferred mode of failure is at the HAZ or through the wire. Consequently, the weld strength
must be designed to have a strength at least equal to that of the line wire
strength (Villaescusa, 1999b).
Only one failure mechanism has been observed for the woven wire mesh.
The mesh fails on the edge of the loading area either as a result of the loading weight cutting through the wires or as a result of the wires cutting each
other at a link. This failure mechanism limits the accuracy of testing and
causes some variability in the results. Generally, only one or two strands
break, which does not constitute a complete destruction of the mesh. Typical
force–displacement reaction curves for welded wire mesh and chain link
mesh are shown in Figure 7.76.
Figure 7.77 shows the WASM static database for galvanized weld mesh
strength and deformability. The effect of wire diameter and failure mode can
be clearly seen. The variability shown is due to the different dimensions and
manufacturers of the products tested. Figure 7.78 shows detailed results for
5.6 mm diameter galvanized welded wire mesh where failure mode and corrosion effects are shown to significantly influence the results (Hassell et al.,
2010). Static results for woven mesh are shown in Figure 7.79. The high overall capacity offers a potential for improvement compared with conventional
weld mesh. Furthermore, woven mesh installation can be fully mechanized,
thereby potentially increasing productivity and development rates.
392
Geotechnical Design for Sublevel Open Stoping
60
Static rupture force (kN)
Galvanized welded wire mesh
30
20
10
0
(a)
All failure modes, = 5.6 mm
Wire failure, = 5.0 mm
Weld failure, = 5.0 mm
Wire failure, = 4.95 mm
Wire failure, = 4.85 mm
Wire-weld failure, = 4.75 mm
Wire-weld failure, = 4.65 mm
40
0
2.0
40
60
80
100
120 140
Static displacement (mm)
160
180
200
220
160
180
200
220
Galvanized welded wire mesh
1.8
Static rupture force (kJ)
20
All failure modes, = 5.6 mm
Wire failure, = 5.0 mm
Weld failure, = 5.0 mm
Wire failure, = 4.95 mm
Wire failure, = 4.85 mm
Wire-weld failure, = 4.75 mm
Wire-weld failure, = 4.65 mm
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0
(b)
0
20
40
60
80
100 120 140
Static displacement (mm)
FIGURE 7.77
(a) Galvanized welded wire mesh strength and (b) deformability as a function of diameter.
(From Villaescusa, E. et al., A database of static and dynamic energy absorption of mesh for
rock support, Proceedings of the 2012 Australian Mining Technology Conference, CRC Mining,
Perth, Western Australia, Australia, October 8–10, 2012b, pp. 27–34.)
393
Rock Reinforcement and Support
60
Galvanized welded wire mesh
Noncorroded, wire failure, = 5.6 mm
Static rupture force (kN)
50
Noncorroded, HAZ failure, = 5.6 mm
Noncorroded, weld failure, = 5.6 mm
Lightly corroded, wire failure, 5.0 mm < < 5.6 mm
40
Lightly corroded, HAZ failure, 5.0 mm < < 5.6 mm
Moderately corroded, wire failure, 4.5 mm < < 5.0 mm
Highly corroded, wire failure, 4.0 mm < < 4.5 mm
Highly corroded, HAZ failure, 4.0 mm < < 4.5 mm
30
Highly corroded, weld failure, 4.0 mm < < 4.5 mm
Severely corroded, wire failure, < 4.0 mm
20
10
0
0
20
40
60
(a)
2.0
100
120
140
160
180
200
220
160
180
200
220
Galvanized welded wire mesh
1.8
Static rupture energy (kJ)
80
Static displacement (mm)
Noncorroded, wire failure, = 5.6 mm
1.6
Noncorroded, HAZ failure, = 5.6 mm
Noncorroded, weld failure, = 5.6 mm
Lightly corroded, wire failure, 5.0 mm < < 5.6 mm
1.4
Lightly corroded, HAZ failure, 5.0 mm < < 5.6 mm
Moderately corroded, wire failure, 4.5 mm < < 5.0 mm
Highly corroded, wire failure, 4.0 mm < < 4.5 mm
Highly corroded, HAZ failure, 4.0 mm < < 4.5 mm
1.2
1.0
0.8
Highly corroded, weld failure, 4.0 mm < < 4.5 mm
Severely corroded, wire failure, < 4.0 mm
0.6
0.4
0.2
0
(b)
0
20
40
60
80
100
120
140
Static displacement (mm)
FIGURE 7.78
(a) Galvanized 5.6 mm diameter welded wire mesh strength and (b) deformability. (From
Villaescusa, E. et al., A database of static and dynamic energy absorption of mesh for rock
support, Proceedings of the 2012 Australian Mining Technology Conference, CRC Mining, Perth,
Western Australia, Australia, October 8–10, 2012b, pp. 27–34.)
394
Geotechnical Design for Sublevel Open Stoping
180
High-tensile woven wire mesh
Static rupture force (kN)
160
Product A, = 4.0 mm
Product B, = 4.0 mm
Product C, = 4.0 mm
Product A, = 3.0 mm
Product B, = 3.0 mm
Product D, = 2.0 mm
10006, = 5.0 mm
140
120
100
80
60
40
20
0
0
50
100
(a)
14
200
250
300
350
400
300
350
400
High-tensile woven wire mesh
12
Static rupture energy (kJ)
150
Static displacement (mm)
Product A, = 4.0 mm
Product B, = 4.0 mm
Product C, = 4.0 mm
Product A, = 3.0 mm
Product B, = 3.0 mm
Product D, = 2.0 mm
10006, = 5.0 mm
10
8
6
4
2
0
(b)
0
50
100
150
200
250
Static displacement (mm)
FIGURE 7.79
(a) Woven wire mesh static strength and (b) deformability for a number of products. (From
Villaescusa, E. et al., A database of static and dynamic energy absorption of mesh for rock
support, Proceedings of the 2012 Australian Mining Technology Conference, CRC Mining, Perth,
Western Australia, Australia, October 8–10, 2012b, pp. 27–34.)
Rock Reinforcement and Support
395
Figure 7.80 shows the WASM dynamic mesh strength and deformability
database. As for the static database results, the woven mesh can absorb more
energy than the welded mesh. The large variability in the woven mesh results
is due in part to the different products tested. It is also noted that large deformations were allowed and that the compatibility with the reinforcement systems used as part of a complete ground support scheme must be considered.
For mesh-reinforced shotcrete, having an embedded mesh that allows high
deformation in discrete areas, where the shotcrete cracks, is beneficial for high
energy absorption.
7.8.4 Thin Spray on Liners
Various polymeric materials have been, or are currently being, developed
for use as thin spray on liners (TSLs; Archibald and DeGagne, 2001). These
liners (TSL) have the potential to serve the function of areal coverage for
scat control as currently achieved by mesh and shotcrete layers. However,
a thin polymeric liner (a few mm thick) cannot be considered to be capable
of replacing the structural strength and stiffness of a shotcrete layer (several centimeter thick), particularly on concave surfaces where they will be
required to react initially in compression due to rock mass loosening. It
would also appear that some of the TSL materials have very poor tensile
strength and can be torn easily by hand. In addition, creep may be a serious
problem and is a property that has not been yet investigated.
Currently, some TSLs have problems in the mining environment associated with toxicity and the need for isolation from other mining activities.
Toxicity is usually associated with fast-setting materials whereas the nontoxic materials have slower strength gain. Consequently, it is considered that
TSLs at this stage of their development are not an option for ground support
and are not considered further within this book.
7.8.5 Shotcrete Layers
Shotcrete is a surface support technique in which a specially mixed concrete
is sprayed at high speed onto rock excavation surfaces to achieve rock mass
integrity and therefore load-carrying capacity. The benefits of using shotcrete
compared with other ground support schemes have been demonstrated particularly where the rock mass is of poor quality, has short stand-up times,
and is easily disturbed when attempting to scale or to drill boreholes for
installation of reinforcement and mesh restraint. In sublevel open stoping,
shotcrete can be used for a variety of conditions such as support for stope
development, drawpoints, and during re-habilitation for pillar mining. Wet
mix (as opposed to dry mix) shotcrete is now widely accepted in mines
throughout the world, particularly those prone to violent rock failure due to
induced stress changes.
396
Geotechnical Design for Sublevel Open Stoping
300
Galvanized welded wire mesh, = 5 .6 mm
Black welded wire mesh, = 5 .0 mm
High-tensile woven wire mesh, product A, = 4 . 0 mm
High-tensile woven wire mesh, product B, = 4 . 0 mm
10006 woven wire mesh, = 5 .0 mm
Dynamic rupture force (kN)
250
200
150
100
50
0
(a)
0
22
100
150
200
250
300
350
Dynamic displacement (mm)
400
450
500
400
450
500
Galvanized welded wire mesh, = 5 .6 mm
Black welded wire mesh, = 5 .0 mm
High-tensile woven wire mesh, product A, = 4 . 0 mm
High-tensile woven wire mesh, product B, = 4 . 0 mm
10006 woven wire mesh, = 5 .0 mm
20
Dynamic rupture energy (kJ)
50
18
16
14
12
10
8
6
4
2
0
(b)
0
50
100
150
200
250
300
350
Dynamic displacement (mm)
FIGURE 7.80
(a) Woven wire mesh dynamic strength and (b) deformability.
7.8.5.1 Shotcrete Support Mechanisms
Studies by Holmgren (1976) and Fernandez-Delgado et al. (1976) concluded
that adhesion loss and flexure are the main modes of shotcrete failure.
A further review of shotcrete capacity in blocky ground under static conditions conducted by Barrett and McCreath (1995) identified six failure
mechanisms, namely, adhesion loss, direct shear, flexural failure, punching shear, compressive failure, and tensile failure (Figure 7.81). Such failure
397
Rock Reinforcement and Support
Adhesion loss
Flexural failure
Direct shear failure
Punching shear failure
Compressive failure
Tensile failure
FIGURE 7.81
Shotcrete failure mechanisms. (After Barrett, S.V.L and McCreath, D.R., Tunn. Undergr. Space
Technol., 10, 79, 1995. With permission.)
mechanisms are generally not well understood, and further research is
required to understand the complexities of rock–shotcrete interaction
(Morton et al., 2009b).
7.8.5.2 Shotcrete Reaction to Transverse Loading
Rock bolt plates and shotcrete are in contact with the rock surface and can
provide confinement and immediate resistance to movement. This is different from straps and mesh that are usually only in contact with the rock at
the positions of restraint and therefore allow (in some cases very significant)
rock movement before providing restraint against further rock movement.
For this reason, mesh alone may not be suitable for the control of a rock
mass susceptible to violent failure due to overstressing. When shotcrete does
not prevent rock failure, the energy absorbed is accompanied by a loss in
the intimate contact with the rock and cracking to form slabs of shotcrete
(Figure 7.82). The crack widths may exceed the length of the any internal
fiber reinforcement, and therefore mesh is the only way of retaining the slabs
of shotcrete that are not directly held by the reinforcing elements.
398
Geotechnical Design for Sublevel Open Stoping
FIGURE 7.82
Unstable slabs of rock and shotcrete.
7.8.5.3 Shotcrete Reaction in Tension
Shotcrete may also act as a membrane in tension. The main disadvantage of
shotcrete in tension is cracking at small strains/differential displacements.
Although shotcrete performance is improved through the addition of polypropylene or steel fibers, the tensile strength (following rupture) is a function
of how well the fibers transfer load across cracks (Figure 7.83). This depends
on the length and number of fibers, the strength of the fibers, their orientation, and the load transfer between the fiber and the shotcrete matrix.
Morton et al. (2009a,b) describe laboratory experiments to determine the
force–displacement properties of large-scale (1.3 m × 1.3 m) fiber-reinforced
shotcrete panels (Fernandez-Delgado et al., 1976; Kaiser and Tannant, 2001)
that were statically loaded simulating punching shear failure (Figure 7.84).
The results enable a comparison with results from a mesh testing program
similar to that described on Section 7.8.3.2. Testing was conducted using a
mix design containing a cement content of approximately 15%, similar water
cement ratios, and 6 kg of polypropylene fibers per cubic meter. However,
slightly different chemical admixtures and aggregates were used for the two
results presented in Figure 7.85. Results for samples with the same thickness
and the same curing time are shown. The failure mode was a combination
of flexural failure and adhesion loss between the substrate and the shotcrete,
with the initial stiff reaction being followed by rupture of the shotcrete.
Shotcrete postrupture behavior is difficult to characterize, as it is dependent
upon the failure mode, the shotcrete thickness, and the type of reinforcing.
7.8.5.4 Shotcrete Reaction in Compression
Support systems designed to be efficient when loaded in tension have poor
performance when loaded in compression. Straps and mesh tend to buckle
when loaded in compression. A thin skin will have negligible strength and
stiffness when compared with the rock with which it is in contact.
Rock Reinforcement and Support
399
FIGURE 7.83
Shotcrete tensile crack opening exceeding the fiber length.
Shotcrete-based products have appropriate properties for membrane
action in compression in terms of both strength and stiffness. Plain shotcrete
is susceptible to cracking due to both shrinkage and any distortion caused
by rock movements. The resistance of shotcrete to cracking is dramatically
improved by the addition of either polypropylene or steel fibers to the mix or
when used in conjunction with mesh.
In addition to transverse loading, rock movements may also cause distortion in the plane of the support. These distortions produce shear forces that
may in turn cause shear or tension cracks (Figure 7.86). Mesh or shotcrete
reinforced with either fibers or mesh can sustain in-plane distortion.
7.8.5.5 Shotcrete Toughness
Toughness is defined as the ability of a support system to absorb energy and
to deform plastically before failing. Toughness is used to assess the support
system where the transient forces immediately after failure would be sufficient to cause support failure if the system did not deform until the rock
force demand reduces to an acceptable level.
Figure 7.87 shows some conceptual force–displacement responses for
various configurations of mesh and shotcrete. The energy absorption can
400
Geotechnical Design for Sublevel Open Stoping
FIGURE 7.84
Sample preparation and testing of large-scale shotcrete panels.
be determined by calculating the area under the force–displacement curve.
Determining energy at an arbitrary displacement is not indicative of the
energy capacity of shotcrete. In order to effectively assess the energy absorption capacity, the cumulative energy absorption variation with central
displacement should be considered. Figure 7.88 shows cumulative energy
absorption for polypropylene fiber–reinforced samples from similar mixes
having different thicknesses and curing ages (Morton et al., 2009b).
With the reintroduction of shotcrete into mining in Australia and elsewhere in the world in the early to mid-1990s, the emphasis in assessing fiberreinforced shotcrete was on the first crack strength. This was driven by the
requirement in civil engineering not to have cracks mainly for aesthetic reasons. It is worth noting that a similar requirement is used for the assessment
of glass fiber–reinforced cement sheets used in civil construction. The highest
first crack strength was found to be directly related to the strength of the fiber
and load transfer to the shotcrete matrix. This was clearly demonstrated by
fibers with essentially the same shape but lower tensile strength having lower
first crack strength. In more recent years, the attitude in mining has changed,
and this has led to the widespread adoption of plastic fibers. Tests have shown
that the overall toughness of plastic fiber–reinforced shotcrete can match that
401
Rock Reinforcement and Support
30
25
Force (kN)
20
15
10
5
0
Site 1 60 mm 7 days
Site 2 60 mm 7 days
0
10
20
30
40
50
60
Displacement at loading point (mm)
FIGURE 7.85
Shotcrete punching shear failure with polypropylene fibers. (From Morton, E.C. et al.,
Determination of energy absorption capabilities of large scale shotcrete panels, in F. Amberg
and K.F. Garshol, eds., Shotcrete for Underground Support XI, Proceedings of the 2009 ECI Conference
on Shotcrete for Underground Support, Davos, Switzerland, June 7–10, Paper 6, 2009b, 20pp.)
of high–tensile strength steel fiber–reinforced shotcrete. However, it is important to note that mesh has a superior response to lateral loading when compared with both steel and plastic fiber–reinforced shotcrete.
Unreinforced shotcrete has low toughness and zero residual strength. The
addition of fibers to a shotcrete mix improves the toughness. However, meshreinforced shotcrete is preferable to plastic or polypropylene fiber–reinforced
shotcrete and steel fiber–reinforced shotcrete where large rock displacements occur after rock failure or accompany creep of ductile rock. That is,
the limit of total displacements that fiber-reinforced shotcrete can sustain
is significantly less than those of mesh-reinforced shotcrete (Figure 7.89).
Accordingly, the energy absorption capacity of mesh-reinforced shotcrete is
also much greater. The results shown are for a mix having 30 kg of steel
fiber per cubic meter and 5.6 mm galvanized weld mesh at 100 mm × 100 mm
openings. Following the test, only a small portion of the mesh could be seen
at the base of the fracture (see Figure 7.13); consequently, the displacement
capacity of the sample is potentially much greater than the results indicated.
The use of mesh and shotcrete has the advantages of providing an immediate response to rock mass movement (due to the shotcrete) and large displacement capacity (due to the mesh). However, a large difference in energy
402
Geotechnical Design for Sublevel Open Stoping
FIGURE 7.86
Shotcrete failure due to movement at a stope brow.
Lateral force
High-strength
steel fibers
High-elongation
plastic fibers
Mesh
Lateral displacement
FIGURE 7.87
Conceptual force–displacement responses of laterally loaded surface support systems. (With
kind permission from Springer Science+Business Media: Geotech. Geol. Eng., Ground support
terminology and classification: An update, 30, 2012, 553, Thompson, A.G., Villaescusa, E., and
Windsor, C.R.)
403
Rock Reinforcement and Support
3.0
2.5
2.5
1.5
1.0
2.0
1.5
1.0
0.5
0.0
60 mm 7 days
70 mm 7 days
85 mm 7 days
3.0
Energy (kJ)
2.0
Energy (kJ)
3.5
35 mm 5 days
40 mm 4 h
60 mm 7 days
140 mm 14 days
160 mm 24 h
0.5
0
20
40
60
0.0
80
Displacement at loading point (mm)
0
20
40
60
80
Displacement at loading point (mm)
FIGURE 7.88
Fiber (polypropylene)-reinforced shotcrete cumulative energy results. (From Morton, E.C.
et al., Determination of energy absorption capabilities of large scale shotcrete panels, in
F. Amberg and K.F. Garshol, eds., Shotcrete for Underground Support XI, Proceedings of the 2009
ECI Conference on Shotcrete for Underground Support, Davos, Switzerland, June 7–10, 2009b,
Paper 6, 20pp.)
120
Mesh-reinforced 105 mm 7 days
Fiber-reinforced 80 mm 24 h
Fiber-reinforced 100 mm 6 days
Fiber-reinforced 110 mm 7 days
Force (kN)
100
80
60
14
12
8
6
40
4
20
2
0
0.0
50.0
100.0
150.0
Displacement at loading point (mm)
Mesh-reinforced 105 mm 7 days
Fiber-reinforced 80 mm 24 h
Fiber-reinforced 100 mm 6 days
Fiber-reinforced 110 mm 7 days
10
Energy (kJ)
140
0
0.0
50.0
100.0
150.0
Displacement at loading point (mm)
FIGURE 7.89
Mesh- and fiber-reinforced shotcrete force–displacement and energy results. (From Morton, E.C.
et al., Determination of energy absorption capabilities of large scale shotcrete panels, in F. Amberg
and K.F. Garshol, eds., Shotcrete for Underground Support XI, Proceedings of the 2009 ECI Conference
on Shotcrete for Underground Support, Davos, Switzerland, June 7–10, 2009b, Paper 6, 20pp.)
404
Geotechnical Design for Sublevel Open Stoping
FIGURE 7.90
Mesh retaining failed shotcrete (above) and stable mesh-reinforced shotcrete (below).
absorption is expected when the mesh is exposed compared to when the
mesh is embedded. Mesh exposed does not participate in the stabilization process at the same time as the shotcrete layer, and the support fails
significantly earlier (Figure 7.90). The use of mesh-embedded reinforced
shotcrete under severe dynamic loading may result in some ejection of the
exposed shotcrete (Figure 7.91), and a second layer of mesh may be required
to ensure safety.
FIGURE 7.91
Dynamic ejection of exposed shotcrete within a mesh-reinforced layer.
8
Mine Fill
8.1 Introduction
Fill consists of materials such as waste rock, aggregates, sand, or classified
mill tailings, which are placed underground to fill voids created by openstope mining. The use of fill contributes to waste disposal, which in turn
helps the environment by reducing the sizes of the required tailings dams.
Operationally, depending upon the detailed stoping method used, fill may
provide a working floor, a side wall, and/or a working back (Figure 8.1). Brady
and Brown (2004) have proposed three support mechanisms for fill-rock mass
interaction (Figure 8.2). First, in destressed rock, the fill provides kinematic
constraint to key blocks formed at the stope boundaries. Second, the passive
resistance of a fill mass is mobilized locally by dilation of fractured rock and
rigid body displacements at the stope wall boundaries. Third, displacement
of an entire stope wall confines a fill mass, which in turn provides global
support to a large area, such as a secondary stope wall or pillar. Thus, a fill
mass provides superficial, local, and global support to the stope walls (Brady
and Brown, 2004).
Sublevel open stoping with primary and secondary extraction requires
tight filling of the stoping voids by means of free-standing cemented fill
masses. In addition, tight fill allows subsidence control in orebodies having
large footprints. Failure of an exposed fill mass is likely to lead to broken ore
contamination. However, stability of a fill mass is difficult to predict accurately. It is a function of the fill type and related properties, the method of fill
placement, the degree of arching and confinement, and also the dimensions
of fill mass exposure (Bloss, 1992). In addition, a long time may elapse between
placement and fill exposure, especially in very large open stope geometries,
making it difficult to optimize ultimate fill performance (Figure 8.3).
A large number of fill types and applications exist in sublevel open stoping. In general, cemented fill is required to recover ore from pillars and
achieve a high extraction ratio. Cemented fill is essential in checkerboard
stope extraction sequences within massive orebodies (Figure 8.4) and also
in tabular orebodies having a primary and secondary extraction sequence.
Cemented fill is also required for continuous stope extraction sequences.
405
406
Geotechnical Design for Sublevel Open Stoping
FIGURE 8.1
Crown pillar recovery for uphole bench stoping under cemented hydraulic fill (CHF). (Courtesy
of Mount Isa Mines, Mount Isa, Queensland, Australia.)
On the other hand, unconsolidated, dry fill is used in conjunction with
bottom-up, narrow bench stoping extractions (Villaescusa et al., 1994;
Villaescusa and Kuganathan, 1998).
8.2 Unconsolidated Rock Fill
Rock fill (RF) is the simplest form of mine fill consisting of waste rock
dumped into a stope void. The material can be sourced from a surface
excavation or from development mining waste. Depending upon the size
distribution, the material can be delivered throughout a stoping block
through large-diameter boreholes or fill raises or is transported using conveyor belts or trucks. Distribution to individual stopes is normally achieved
using mobile equipment.
The material forms an unconsolidated fill mass and large particle sizes
should be avoided to eliminate large void spaces within the fill mass.
407
Dilation of
fractured rock
Mine Fill
Fill
mass
Internal fill
stress
Block
displacement
(a)
Stope wall
closure
Fill
mass
Ps
Pn
Stope wall
supportelastic
response
of fill
(b)
Stope wall
closure
(c)
FIGURE 8.2
Ground support mechanisms due to mine fill. (a) De-stressed rock mass constraint on
surface blocks, (b) fractured and jointed rock mass support forces mobilized locally, and
(c) global stope wall support fill compression due to stope wall closure. (From Brady, B.H.G.
and Brown, E.T., Rock Mechanics for Underground Mining, 3rd edn., Kluwer, Dordrecht, the
Netherlands, 2004, 628pp.)
Legend sequence:
Year(s) produced
Year(s) filled
Hangingwall
conveyor drive
High-grade
orebody limit
Crosscut
conveyor drive
Footwall
conveyor drive
Cemented
Uncemented
#3
1986–1987
1988
Intermediate
sublevel access
#8
1994–
1995
Producing
Fill
pass
Future stope
Silica
dolomite
#1
1985–1989
1989
#4
1987–1990
1990
#6
1992–
1993
1994
#7
1993
1993
#5
1992
1992
#2
1986–1987
1990–1994
Greenstone basement contact
100 m
FIGURE 8.3
Cross section showing time lag between fill placement and exposure. (From Bloss, M.L., Miner.
Resour. Eng., 5, 23, 1996.)
408
Geotechnical Design for Sublevel Open Stoping
FIGURE 8.4
Details of tertiary stope extraction where more than one fill mass is exposed.
Exposure of this fill may lead to dilution. However, in some cases it becomes
fundamental to a mining method as at the Mount Charlotte mine (Ulla, 1997)
where a top-down stope extraction strategy, with stoping under unconsolidated fill has been implemented (see Figure 3.2).
Unconsolidated RF forms a cone according to the rill angle and the
location of fill placement within the stope geometry. Figure 8.5 shows an
example of RF placement for a massive, isolated stope, in which no further
FIGURE 8.5
Unconsolidated RF delivered through a raise.
Mine Fill
409
FIGURE 8.6
Unconsolidated RF unable to provide tight support to a stope hangingwall.
wall exposures would take place. In tabular orebodies, the RF rill angle of
approximately 37°–42° makes it very difficult to achieve tight fill against a
steeply dipping stope hangingwall (Figure 8.6).
8.2.1 Rock Fill for Bench Stope Support
The success of the bench stoping method largely depends upon the level
of understanding of unsupported wall exposures, the application of remote
mucking technology, drilling and blasting optimization, and the appropriate
use of fill technology (Villaescusa et al., 1994). An extraction strategy related
to the maximum stable length that can be safely exposed, and the type of
fill to be used is usually identified during the initial design stages. In most
cases, permanent infrastructure such as ramp access configurations are also
fixed very early on, leaving the extraction strategy as the only flexible (and
most important) parameter to be optimized during the subsequent production stages (Villaescusa and Kuganathan, 1998).
In bottom-up (up-dip) bench extractions (see Section 2.6), fill provides a
working floor for mucking and also helps to stabilize the exposed spans by
minimizing deformation and dynamic loading of the excavated walls from
blasting. Following extraction of an economic length of a steeply dipping
orebody, the void created by a bench stope can be filled with dry fill (waste)
to the floor of the drill drive, which becomes the new extraction horizon
on the next lift as indicated in Figure 1.14. Dry RF can be used to minimize
deformations (and optimize stability) while the benches are being extracted,
provided that the fill can be kept sufficiently far away to minimize dilution
of the broken ore by fill at the interface.
410
Geotechnical Design for Sublevel Open Stoping
Filling
Cablebolted area
Production blasting
Unsupported
hangingwall area
Stope walls supported
by fill mass
Most likely stable length
Blasted ore
Extraction
FIGURE 8.7
Blasted ore–RF interface. (From Villaescusa, E. and Kuganathan, K., Backfill for bench stoping
operations, in M.L. Bloss, ed., Minefill 98, Proceedings of the 6th International Symposium on Mining
with Backfill, Brisbane, Queensland, Australia, April 14–16, 1998, pp. 179–184, The AusIMM,
Melbourne, Victoria, Australia. With permission.)
Empirical stability charts such as the stability graph method (see Chapter 5)
can be used to determine the maximum unsupported strike lengths, which
can be safely exposed during continuous filling operations. An optimal use
of the “critical strike length” concept would ensure that excessive dilution
does not occur during production blasting, where the blasted material may
be thrown on top of closely located backfill rills (Figure 8.7), contributing to
contamination of the ore during mucking.
The support provided by RF minimizes the deformations at the exposed
unsupported bench stope hangingwalls either as the stope is being
extracted or following bench completion. Hangingwall deformation
data collected from properly located multiple point extensometers have
shown that unconsolidated RF effectively stops the large-scale deformation of unsupported hangingwall layers during bench stoping (Figure 8.8,
Villaescusa, 1996).
Geotechnical instrumentation has also been used to determine the
dynamic response of a stope wall as a bench stope is extracted and filled
progressively. Table 8.1 shows a frequency analysis of instrumented walls
using triaxial arrays of geophones, indicating that the wall of a filled stope
(using dry RF) behaves like a closed wall (i.e., intact solid ground, where no
void has been created). All the blast vibration data were collected at approximately 5, 9, and 13 m into the hangingwall of a stope (Villaescusa et al., 1994).
The beneficial impact of the fill in stabilizing the rock mass surrounding a
stope void is very clear from the data presented in Table 8.1. Promptly placed
RF appears to reduce the dynamic loading caused by blasting, thus enhancing the overall regional rock mass stability.
411
Mine Fill
5FP1 extensometer 1
30
Stope
blastings
Fill introduced here
Anchor depth into H/W:
Deformation (mm)
25
A1—0.5 m
20
A2—1.5 m
A3—2.5 m
15
A4—3.5 m
10
A5—7.5 m
5
0
2/22/93
A6—Ref
3/14/93
4/3/93
4/23/93
5/13/93
6/2/93
6/22/93
Date
FIGURE 8.8
Influence of RF on a bench stope hangingwall deformation. (From Villaescusa, E., Trans. Inst.
Min. Metall. Sect. A Mining Industry, 105, A1, 1996.)
TABLE 8.1
Dynamic Response of a Rock Mass as Rock Filling Proceeds
Fill Status
No closed walls (1.5 m burden)
No closed walls (3 m burden)
No; 6 m open span
No; 9 m open span
No; 15 m open span
Stope empty; 15 m open span
Stope 1/2 filled
Stope 3/4 filled
Stope 5/6 filled
Stope filled
Stope filled
Dominant
Frequency (Hz)
Average
Frequency (Hz)
Number of
Data Points
10–20
40–50
30–50
90–100
100–110
100–130
100–110
—
10–20
40–50
30–40
31
52
45
88
94
114
86
71
28
38
29
17
8
7
5
84
9
7
6
5
5
8
Source: Villaescusa, E., Quantifying open stope performance, in A. Karzulovic and
M.A. Alfaro, eds., Proceedings of the MassMin 2004, Santiago, Chile, August
22–25, 2004, pp. 96–104, Chilean Engineering Institute, Santiago, Chile.
412
Geotechnical Design for Sublevel Open Stoping
8.3 Cemented Rock Fill
Cemented rock fill (CRF) consists of dry rock that is mixed with cemented
slurry at the top of a stope (Yu and Counter, 1983; Grice, 1989). The method is
suitable for multiple-lift open stopes, where a high-performance cemented fill
is required to achieve high production targets and fast cycle times. The rock
material may be crushed and screened or constitute run-of-mine material that
is transported and placed in a nonsaturated state. Depending upon the particle
size distribution, the material can be delivered through fill holes and passes,
using conveyor belts or trucks (with or without slingers). The method has a
low capital cost. However, the maximum particle size of the aggregates used
to prepare RF has a major impact on the capital cost of establishing a fill plant,
as well as on the operational cost of the fill material preparation. Extracted
stopes are sealed with fill fences or barricades and the CRF is dumped into
stope voids where a cementitious matrix is added to the waste rock (Figure
8.9). As the fill mass consolidates, it can be exposed, achieving a very high
strength and stiffness leading to high free-standing walls (Bloss, 1992).
When the fill materials can be delivered through small-diameter drill
holes, this allows a better distribution of fill within a stope void. However,
the CRF is known to segregate around a dump point. Grice (1989) describes
a differential filling technique in which the cemented hydraulic beach is
Rock fill
crushing and
screening
Rock fill pile
at fill pass
Copper
tailings
Slurry fill
preparation
plant
Copper
concentrator
Rock quarry
Cemented
hydraulic fill
borehole
Rock fill pass
CHF pipes at
underground
Underground rock fill
conveyor system
Mount Isa Mines composite fill system
Rock fill core
and CHF beach
Blasted
copper ore
FIGURE 8.9
Conceptual CRF distribution system. (From Kuganathan, K. and Neindorf, L.B., Backfill
technology development at Mount Isa Mines between 1995 and 2005, Proceedings of the Ninth
AusIMM Underground Operators Conference, Perth, Western Australia, Australia, March 7–9,
2005, pp. 173–183, The AusIMM, Melbourne, Victoria, Australia. With permission.)
Mine Fill
413
FIGURE 8.10
Differential RF—CRF impact cone and cemented hydraulic beaches (40 m × 40 m stope
plan area). (Courtesy of Mount Isa Mines, Mount Isa, Queensland, Australia.)
placed against a future stope exposure, while the main void is filled with
aggregate (Figure 8.10). As the stoping depth increases, the raises used for
material delivery can wear out.
CRF can also be used within bench-stoping extraction sequences as
described by Saw et al. (2011). Within this context, a mix of a crushed and
screened waste rock from mine development, general-purpose cement, and
fresh water can be used in order to provide the following functions:
• Regional stability to a bench stope surrounding rock mass.
• Allowance for undercutting of sill-stoping levels once a bottom-up
sequence reaches the top of a stoping panel.
• Retention of unconsolidated waste rock in the back half of each stope
while achieving a free-standing face that facilitates the removal of
an adjacent stope (Figure 8.11). This also provides resilience to slot
firing activities within close proximity to the CRF.
8.3.1 Cemented Aggregate Fill
Cemented aggregate fill (CAF) is created when crushed rock is added to
cemented hydraulic fill (CHF) and dumped into a stoping void (Bloss, 1992,
1996; Farsangi and Hara, 1993; Bloss and Greenwood, 1998, see Figure 8.12).
The relationships between aggregates, tailings, and cement dosage, and the
effect of the addition of admixtures on the workability of backfill are very
important. Cowling et al. (1989) reported an example of aggregate added
at 20%–25% by weight and the product reticulated and delivered through
pipelines and boreholes at a pulp density of about 70% weight percent solids.
414
Geotechnical Design for Sublevel Open Stoping
Filling stop point
Direction of advancing stope front
Uncemented rock fill mass
Unmined orebody
Slot raise holes
Waste rock bound
15 m
CRF
15 m
FIGURE 8.11
Schematic of cemented and uncemented rock-filled portions of a bench stope. (From Saw, H.
et al., Characterisation of cemented rock fill materials for the Cosmos nickel mine, Western
Australia, in C. Leung and K.T. Wan, eds., Proceedings of the International Conference on Advances
in Construction Materials through Science and Engineering, Hong Kong, China, September 5–7,
2011, RILEM, Bagneux, Paris, France.)
FIGURE 8.12
CAF dumped from the stope footwall at the Kanowna Belle Mine, Western Australia.
Mine Fill
415
FIGURE 8.13
CAF exposure during secondary stope extraction at the Bronzewing Mine.
Figure 8.13 shows a secondary stope extraction which exposed very stable
CAF at the Bronzewing Mine, Western Australia.
General-purpose portland cement is used to prepare CAF according to the
requirements of a particular mine site. Typical cement rates used are 3%–7%
by the weight of solids, with cement dosage accounting for a high proportion in the cost of cemented aggregate. Therefore, it is very important that
the use of alternative cost-effective binding agents be considered. The use of
pozzolans as binding agents, partially or fully replacing ordinary portland
cement in CRF or other types of cemented fill, has been widely practiced. For
example, Mount Isa Mines used 1.3% cement and 2.6% copper reverberatory
furnace slag in their CRF and CHF to achieve a designed uniaxial compressive strength (UCS) of 1 MPa at 56 curing days (Grice, 1989). With fill of this
design, exposures of 40 m wide by in excess of 200 m high have been proved
stable (Bloss, 1992). Many additional examples can be found worldwide of the
use of pozzolans partially as CRF binders.
Aggregate composition and size distribution play a significant role in fill
strength creation and fill mass structure. Well-designed aggregate composition and size distribution can maximize the fill strength and minimize
segregation of the fill mass during fill placement. Generally, strong rocks
are used to produce aggregates for CAF as it is commonly recognized that
416
Geotechnical Design for Sublevel Open Stoping
the stronger the aggregates, the higher the strength of the CAF. However,
laboratory tests indicate that the strength of CAF made of a combination
of strong aggregate and relatively weak aggregate which generates more
fines is noticeably higher than that obtained from a fill made solely of either
strong or weak aggregate (Golosinski et al., 1997). The generation of fines
during CAF preparation, delivery, and placement at a mine site can produce
a mix that is densely packed resulting in a higher bulk density and a lower
void ratio, thus increasing the UCS strength.
The effect of the addition of sand to CRF was studied in detail for Kidd
Creek Mine’s fill practice. Yu and Counter (1983) reported that a 5% addition of sand to the CRF significantly reduced segregation of coarse aggregate and resulted in a 40% increase in compressive strength. When more
sand was added, the increase in the surface area of the aggregate, which
must be coated with the same amount of cement paste, caused the strength
to decrease monotonically. Around many mine sites there is often either no
sand available or an insufficient supply for mine fill purposes. It is therefore
necessary to make use of tailings as an alternative, normally having a much
finer particle size than sand. The optimal addition of tailings to the CAF
remains unknown and needs to be investigated and specified.
Figure 8.14 shows the effect of tailings introduction into a CAF mix. The
results indicate that for fill samples, 100 mm in diameter and 200 mm in
10
Failure stress (MPa)
(at a confining pressure of 200 kPa)
9
8
7
6
5
4
3
2
1
0
0
10
20
30
Tailings percentage (%)
40
50
FIGURE 8.14
Influence of tailings addition on the strength of CRF. (From Wang, C. and Villaescusa, E.,
Backfill research at the Western Australian School of Mines, in G. Chitombo, ed., Proceedings of
the MassMin 2000, Brisbane, Queensland, Australia, October 29–November 2, 2000, pp. 735–743,
The AusIMM, Melbourne, Victoria, Australia. With permission.)
Mine Fill
417
length, made of aggregates having a nominal maximum particle size of
20 mm (at a cement dosage of 4%), the highest strength was achieved with a
tailings addition of 10%. With the same cement dosage, for both cases where
the RF had no tailings or had more than 10% tailings, a lower strength was
achieved. This is because mixing tailings into a CRF with no tailings had a
poor cement coating effect, whereas a high tailings addition increased the
surface area of the particles that required cement coating and binding.
Previous research has shown that the total dissolved solids (TDS) influence
cement strength and that the CAF results depend upon the amount of TDS
and the chemical composition of the groundwater (Wang and Villaescusa,
2000). The goal of water content determination is to achieve an optimal water
addition for a fill with a slump equivalent to 200 mm. However, in cases
where the tailings percentage is low, leakage of free water may occur when
a fill mixture is prepared. In cases where tailings are used to make a fill
mixture, the principle for determining water addition is to ensure a proper
distribution of cement slurry through the aggregates.
It is also important to ensure that setting of the CAF mixture takes
place after the fill is placed. Otherwise, if the setting and consolidation of
cemented fill mixture take place during its transportation, the pour of fill
into a stope void will impose a detrimental impact on the strength development of the cemented fill mass. The workability of CAF in conjunction with
the utilization of admixtures needs to focus on the influence of admixtures
(Weatherwax et al., 2011) on fill slump, water addition, strength development,
and setting time of cement.
Figure 8.15 shows the effects of an admixture dosage ranging from 0.3%
to 0.9% for a mixture of cemented aggregate/tailings fill. The results are for
cemented tailings/aggregate fill with a recipe of tailing:aggregate:cement
equal to 32:64:4. Admixture dosages used were 0.3%, 0.6%, and 0.9% by weight
of cement. An increase in the failure stress of around 35% was achieved when
a 0.3% and 0.6% admixture was used for both cases of 100 and 300 kPa confining pressures. A 0.9% admixture dosage increased the failure stress by 55%
for both tests (100 and 300 kPa confining pressures). The slump of the mixture with 0.6% admixture was 205 mm, which was 5 mm higher than that for
other samples. In addition, the water contents used to make the fill mixtures
to achieve a relatively equal slump of 200 mm for 0%, 0.3%, 0.6%, and 0.9%
admixture were 28.9%, 25.2%, 25.6%, and 23.5%, respectively. This clearly indicates the marked effect of water reduction through the use of an admixture.
In summary, the following factors are likely to contribute to the performance
of CAF:
• Cement content
• Percentage of extra fines (tailings or sand)
• Quality and quantity of cement alternatives such as ground slag and
fly ash if used as binding agents
418
Geotechnical Design for Sublevel Open Stoping
5.0
Confining pressure
Failure stress (MPa)
4.5
300 kPa
4.0
200 kPa
3.5
100 kPa
3.0
2.5
2.0
0
0.15
0.3
0.45
0.6
0.75
0.9
Admixture dosage (% by wt cement)
FIGURE 8.15
Influence of admixture on cemented tailings RF. (From Wang, C. and Villaescusa, E., Backfill
research at the Western Australian School of Mines, in G. Chitombo, ed., Proceedings of the
MassMin 2000, Brisbane, Queensland, Australia, October 29–November 2, 2000, pp. 735–743,
The AusIMM, Melbourne, Victoria, Australia. With permission.)
• Water/cement ratio of the cement slurry or cemented hydraulic
slurry
• Nature and quality of admixtures
• Degree of mixing between the cement slurry and fill aggregates
• Composition and quality of aggregates
• Aggregate size distribution
• Segregation of material during transport and placement
8.4 Hydraulic Fill
Hydraulic fill (HF) consisting of classified mill tailings from which the fine
fractions have been removed (Figure 8.16) is one of the most effective methods available to support an open stope void (Thomas et al., 1979). It can be
placed either cemented or uncemented, with material sourced from a surface
plant. The material can also be sourced from a sand fill plant where dry sand
is slurried for piping underground. Cement can be added to the fill in order
to provide strength and fill rill control, so that pillars adjacent to filled stopes
can be recovered without undue dilution from the fill.
Gravity is the prime conveyor and slurried fill is piped underground. Vertical
pipelines can be made of steel, or boreholes may be used. Level distribution
419
Mine Fill
11.8%
solids
TSF2
16.5%
solids
Secondary
cyclones
Primary deslime
cyclones
30% solids
32% solids
Cement silo
40% solids
62% solids
Vortex mixer
76% solids
50%–55%
solids
FIGURE 8.16
Schematic representation of a CHF process plant. (From Winder, K., The introduction of
cemented hydraulic fill to the Gossan Hill Mine, MEngSc in Mining Geomechanics thesis,
Curtin University of Technology, Kalgoorlie, Western Australia, Australia, 2006.)
lines are used to transport the fill from the main vertical lines or borehole to
the stopes to be filled (Figure 8.17). Gravity feed is adequate unless excessive
horizontal distances are involved. Fill density is around 70%–75% solids by
weight (45%–50% by volume) and can settle during transport. The recommended operating velocity is approximately 6 m/s.
Once a fill-handling system is installed, a minimum amount of time
and labor is required for filling the stope voids. However, the HF requires
(a)
(b)
FIGURE 8.17
HF of (a) bench and (b) multiple-lift open stopes in progress. (Courtesy of Mount Isa Mines,
Mount Isa, Queensland, Australia.)
420
Geotechnical Design for Sublevel Open Stoping
reasonable water permeability or percolation rates to allow water draining,
so that the material does not remain in a fluid state. A significant aspect of
fill placement is the stope preparation prior to filling. Sealing off of any stope
development is required, and adequate provision must be made for water
removal. Instrumentation of fill lines and fill barricades is also required
(Winder, 2006).
For uncemented hydraulic fill (UHF), a size distribution having no more
than 8% (solids by weight) passing 20 μm is required to allow adequate
permeability following placement. UHF is used where a fill mass would
not be exposed by future mining, given that it is not likely to develop the
strength required to support its own weight. CHF typically has no more than
10% (solids by weight) passing 10 μm. The addition of a binder allows the
fill to gain sufficient strength to support its own weight when exposed by a
stoping sequence.
UHF consists of a porous mass in which the excess transported water drains
from the stope void when placed. Strength is developed by interparticle friction and confinement from stope wall closure. Therefore, when the walls are
removed, a fill mass will become unstable. For CHF, Grice (1998) has reported
that a cement addition of approximately 6% by dry weight will achieve an
unconfined compressive strength exceeding 750 kPa within 28 days curing.
In practice, for each operation, cement addition varies slightly due to the
different tailings mineralogy.
Figure 8.18 shows a void created by the completion of a secondary stope
extraction. A slight arching (across the orebody) of CHF fill masses previously
placed within the adjacent primary stopes can be observed. CHF allows for
100% extraction of large multiple-lift open stopes, as shown in the example of
the long section from the Golden Grove Mine, Western Australia (Figure 8.19).
FIGURE 8.18
Secondary stope extraction showing slight arching of the two CHF surfaces. (Courtesy of
Mount Isa Mines, Mount Isa, Queensland, Australia.)
421
Mine Fill
371c66
Stope
40 m
40 m
120 m
FIGURE 8.19
Long section view of a copper orebody—Golden Grove Mine, Western Australia. (From
Winder, K., The introduction of cemented hydraulic fill to the Gossan Hill Mine, MEngSc in
Mining Geomechanics thesis, Curtin University of Technology, Kalgoorlie, Western Australia,
Australia, 2006.)
The successful use of HF requires large amounts of transport water.
Excess water can cause significant problems on stope barricade levels, as it
drains out and may carry cement and other fines (slimes) out of the stopes,
potentially resulting in loss of strength and substantial pumping costs.
For HF to start working within a stope, it must be dewatered, so that the
particles of the fill can come in contact and interlock. Excess water must
be removed from a stope before more fill is added. When CHF is used, the
main fill dewatering is through percolation and if the rate is too slow, the
excess slurry water will not be drained quickly enough and pooling will
occur at the stope surface. Figure 8.20 shows an example where the pooling water level was kept below the stope barricade and water drainage
occurred prior to another fill run.
An advantage of HF is the simplicity of production and reticulation. Fill
materials such as sand and deslimed tailings are readily available and
easy to mix, resulting in a relatively low production cost, with cement content constituting the largest cost component of the fill production process.
Experience and knowledge of the effects of fines content on flow properties have resulted in the development of dependable HF placement methods
(Grice, 1989, 2005b; Winder, 2006; Archibald et al., 2011).
422
Geotechnical Design for Sublevel Open Stoping
FIGURE 8.20
Evidence of ponded water with respect to permeable fill barricade. (Courtesy of Mount Isa
Mines, Mount Isa, Queensland, Australia.)
8.5 Cemented Paste Fill
Cemented paste fill (CPF) represents a variation of hydraulically placed
fill that uses total mill tailings that have been dewatered (Figure 8.21). The
material is transported and placed using boreholes and pipeline distribution
systems and contains in excess of 80%–85% mass solids by weight (Landriault
and Goard, 1987). Cement is required in all placements, and sand (or
aggregate) and water are added to develop specific rheological and strength
characteristics. The liquids in the fill and drain water blend with the solids
after placement; however, the cement hydration process generally consumes
any excess water. Paste fill develops good strength characteristics compared
to other fill types. However, the method has high capital costs to facilitate the
transportation and placement of plug flow material.
The required size distribution for paste fill is that at least 15% (solids
by weight) must pass 20 μm in order to achieve the rheological properties
required for pipeline or borehole placement (Figure 8.22).
CPF strength is a function of cement content which can range from 1% to
10%. The strength gain depends significantly upon the water/cement ratio
(Figure 8.23). Figure 8.24 shows the CPF strength in comparison to other fill
types. Early strength development potentially reduces stope cycle times.
CPF consists of a non-segregating slurry, which means that even when
stationary, the fill remains in a homogeneous single phase. This is due to the
size and content of the solids that can retain the water in the mix, which has
negligible excess water. Preparation of the tailing materials by dewatering is
423
Mine Fill
To tailings
disposal
No.4
concentrator
Thickener
Cement silo
Cyclones
Belt filter
Wet fill
processing
plant
Pugmill
Gob hopper
Underground reticulation system
Pastefill processing and reticulation system
Filled stopes
FIGURE 8.21
Example of paste fill plant process flow and components. (From Kuganathan, K. and Neindorf,
L.B., Backfill technology development at Mount Isa Mines between 1995 and 2005, Proceedings
of the 9th AusIMM Underground Operators Conference, Perth, Western Australia, Australia,
March 7–9, 2005, pp. 173–183, The AusIMM, Melbourne, Victoria, Australia. With permission.)
FIGURE 8.22
Paste fill delivery and reticulation using pipelines.
required, as the mineral processing is usually undertaken using high water
contents. The tailings are raised to the required density using a process that
includes thickening and filtering (Earl, 2003; Faulkner, 2005). The slump test
is used to ensure that a given paste plant is producing paste at the required
density. Typical slumps for paste fill range from 150 to 250 mm.
424
Geotechnical Design for Sublevel Open Stoping
Failure strength (MPa)
2
7% cement in solids
1.6
5% cement in solids
1.2
0.8
3% cement in solids
0.4
0
0
7
Curing time (days)
14
28
FIGURE 8.23
Comparative strength—UCS versus time for different CPF types. (From Saw, H. and
Villaescusa, E., Research on the mechanical properties of minefill: Influences of material particle size, chemical and mineral composition, binder and mixing water, in H.J. Ilgner, ed., Minefill
2011, Proceedings of the 10th International Symposium on Mining with Backfill, Cape Town, South
Africa, March 21–25, 2011, pp. 143–152, SAIMM, Johannesburg, South Africa.)
3.0
CAF
CHF
CPF
2.5
UCS (MPa)
2.0
1.5
1.0
0.5
0.0
0
7
14
21
28
35
Curing (days)
42
49
56
63
FIGURE 8.24
UCS development for fill mixes having 4% cement. (From Saw, H. and Villaescusa, E.,
Geotechnical properties of mine fill, in C.F. Leung, S.H. Goh, and R.F. Chen, eds., Proceedings
of the 18th South East Asian Geotechnical & Inaugural AGSSEA Conference, Singapore, May 29–31,
2013, Research Publishing, Singapore.)
425
Mine Fill
Yield stress is the stress at the limit of elastic behavior describing the
rheology of a paste fill. In other words, it is the minimum stress required to
initiate paste flow at almost zero shear rate. Understanding the relationship
between the yield stress and the solids percentage is essential for the design
of a paste fill transportation system. A proper transportation system enables
the delivery of CPF from the surface to underground at the highest solids percentage. Direct yield stress measurements can be undertaken using a method
suggested by Nguyen and Boger (1985) and using a Haake VT550 viscometer.
The vane shear stress is calculated as being uniformly distributed within
the cylindrical CPF samples. Yield stresses are measured immediately after
mixing, that is, about 5–10 min after binder and water contact. The vane is
rotated at a shear rate of 0.5 rpm for 100 s and the stress is recorded during
that period. The peak stress is reported as the yield stress. Standard conical
slump tests in accordance with Australian Standard AS 1012.3.1 can be also
conducted on different CPF mixes. Typical yield stress, correlations with solids percentage, and slump for different mixes are presented in Figures 8.25
and 8.26. Slightly different correlations can be established for different mixes.
The capital cost for a paste fill plant can be high due to the specialized
machinery and the instrumentation required for monitoring the water content and pumpability. Higher operating costs are incurred during plant commissioning. In addition, as total tailings are used, a high proportion of fines
is likely to cause problems during filtering. However, as total tailings can be
used, this removes the need for classifying before preparation of the fill for
2000
1800
Yield stress (Pa)
1600
1400
1200
1000
800
600
400
200
0
65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83
Solid (%)
FIGURE 8.25
Typical correlation between CPF solids density and yield stress for different mine sites. (From
Saw, H. and Villaescusa, E., Geotechnical properties of mine fill, in C.F. Leung, S.H. Goh,
and R.F. Chen, eds., Proceedings of the 18th South East Asian Geotechnical & Inaugural AGSSEA
Conference, Singapore, May 29–31, 2013, Research Publishing, Singapore.)
426
Geotechnical Design for Sublevel Open Stoping
600
Yield stress (Pa)
500
400
300
200
100
0
150
175
200
225
Slump (mm)
250
275
300
FIGURE 8.26
Typical correlation between CPF solids density and slump for different mine sites. (From
Saw, H. and Villaescusa, E., Research on the mechanical properties of minefill: Influences
of material particle size, chemical and mineral composition, binder and mixing water, in
H.J. Ilgner, ed., Minefill 2011, Proceedings of the 10th International Symposium on Mining with
Backfill, Cape Town, South Africa, March 21–25, 2011, pp. 143–152, SAIMM, Johannesburg,
South Africa.)
placement into the stope voids. An environmental advantage results from
having a high proportion of the metallurgical waste disposed underground.
The overall paste distribution system must be engineered to withstand high
pressures while having the flexibility to deliver fill to all the stoping locations within a particular mine.
Figure 8.27 shows an example of a stable fill mass where continuous
extraction using a top-down bench-stoping sequence mining under paste
fill was implemented at the Junction Mine at Kambalda, Western Australia.
Figure 8.28 shows an exposed paste fill wall at the extraction horizon of a
multiple-lift open stope.
8.6 Open Stope Fill Operations Systems
The selection process for an effective fill operation system is always influenced by local experience and the availability of materials at a particular mine
site. Cowling (1998) has recommended a series of steps to properly identify
and progressively eliminate alternatives. The initial stage is to engineer a system which is then followed by optimization, leading to long-term cost control.
427
Mine Fill
Cemented
paste fill
Stope
void
FIGURE 8.27
Example of bench stoping under CPF at the Junction Mine.
FIGURE 8.28
Front view of a stable CPF wall exposed during secondary stoping.
8.6.1 Material Preparation
Material preparation often involves a fill-processing plant or station that
is located on the surface, where the total size mill tailings are dewatered,
resized, and mixed. For CHF, resizing usually involves removing the −10 μm
fraction to produce a free-draining fill. Moist tailings are reconstituted
with water for CPF and rock crushing is undertaken to achieve the desired
428
Geotechnical Design for Sublevel Open Stoping
aggregate particle size distribution for CAF. The resulting tailings or aggregates are mixed with binding agents such as cement and pozzolans. The mix
proportion is calculated by dry mass of the components. Uniformity of fill
material supply is required at all times to ensure continuity of production.
8.6.1.1 Chemistry and Mineralogy
The chemical and mineral composition of fill material is likely to influence
the ultimate strength development of a fill mass. Some minerals can undergo
oxidation, hydration, or carbonization when exposed to the humid conditions of a fill mass. The presence of pyrite in a fill material has been known
to lead to exothermic reactions that can ignite their sulfur content and start a
self-sustaining fire within a stoping void (AMIRA, 1995). Clay and micas are
flat platy minerals that can reduce the percolation rate, and due to their large
surface area, require additional binder contents to achieve target strengths.
Analyzing results is complex because of the effects of material grinding,
which can break down the crystal structure of some minerals present and
cause difficulties during the identification of the minerals. Table 8.2 shows
the mineral composition of typical mine tailings obtained using x-ray diffraction (XRD) methods. The results show that tailings mainly contain
quartz, feldspar, mica, clay minerals, sulphide minerals, and carbonate minerals. Some minerals are not favorable for cement hydration. In addition, the
presence of clay minerals (chlorite, illite, and kaolin) and sulphide minerals
(pyrite and pyrrhotite) would be expected to reduce the strength of fill for a
given cement type and dosage. On the other hand, the presence of carbonate minerals (calcite and dolomite) would favorably increase the mine fill
strength.
8.6.1.2 Particle Size Distribution
The particle size distribution of the fill material is a key controlling factor
in the engineering properties of a fill mass. Hence, one of the first steps in
preparing the materials for use in fill is a sizing process to remove fines from
the tailings. Depending upon the fill type, thickeners, crushers, screens, and
cyclones are used in the sizing process. Other sizing methods involve optical
imaging using laser light. The size distribution range controls the density,
void ratio, and porosity of a fill mass. Figure 8.29 shows the sizing curves
for a number of fill types. In general, for CRF applications, the particle sizes
exceeding 10 mm are classified as coarse aggregates.
8.6.1.3 Binders
In most cases, a fill mass must have enough strength to support its own
weight following exposure. The strength is gained from binders that become
a key ingredient of a fill mix. The most common fill binders are cement or
—
—
—
—
29
<2
4
—
—
—
—
—
23
24
—
—
18
—
—
30
Gold
Tailings-1
19
—
—
—
—
—
—
—
9
3
21
10
10
—
Lead–Zinc–
Silver Tailings
—
—
43
—
9
5
16
—
—
—
—
—
—
—
—
—
27
Gold
Tailings-2
51
—
19
—
—
—
15
—
—
—
3
—
—
—
7
—
4
Copper
Tailings-1
11
17
23
—
—
—
3
—
—
—
—
—
—
—
37
—
9
Copper
Tailings-2
9
3
23
—
—
—
—
—
—
—
—
—
6
—
50
—
9
Copper
Tailings-3
—
—
45
—
—
11
31
5
—
—
—
—
—
—
—
—
8
Copper
Tailings-4
Source: Saw, H. et al., Characterisation of cemented rock fill materials for the Cosmos nickel mine, Western Australia, in
C. Leung, and K.T. Wan, eds., Proceedings of the International Conference on Advances in Construction Materials
through Science and Engineering, Hong Kong, China, September 5–7, 2011, RILEM, Bagneux, Paris, France.
a Favorable mineral for cement hydration.
b Unfavorable mineral for cement hydration.
Amphibole
Ankerite
Calcitea
Chlorite
Dolomitea
Gypsumb
Halite
Illiteb
K Feldspar
Kaolinb
Kyanite
Magnetite
Muscovite
Plagioclase
feldspar (Albite)
Pyriteb
Pyrrhotiteb
Quartz
Mineral
Typical Mineral Composition of Tailing Materials
TABLE 8.2
Mine Fill
429
430
Geotechnical Design for Sublevel Open Stoping
Cummulative (% passing)
100
90
CPF mill tailings
CHF mill tailings
80
CAF aggregate
70
CRF waste rock
60
50
40
30
20
10
0
0.0001
0.001
0.01
0.1
1
10
100
1000
Particle size (mm)
FIGURE 8.29
Typical particle size distributions for different fill types.
natural pozzolans which control the time-dependent fill strength development. Strength gain will continue as long as unhydrated cement and water
is present. However, because of mining schedules, there is often a limited
amount of time available for the fill to cure. In order to achieve the required
strength within a stope extraction period, the correct water/cement ratio
must be known and adhered to as closely as possible.
In general, the choice of binder depends upon the required strength and
durability requirements of a particular fill operation. According to Bogue
(1955), the main compounds in the different types of cement and pozzolans
can be estimated using XRD scan results. As shown in Figure 8.30, the major
cement components are tricalcium silicate (3CaO·SiO2) and dicalcium silicate (2CaO·SiO2). Both react with water to produce calcium silicate hydrate
(C-S-H) and calcium hydroxide (CH). The strength development is due to the
formation of C-S-H. CH can react with aggressive chemicals in tailings and
saline water in some underground mines, lowering the durability of minefill (Wang and Villaescusa, 2001). Therefore, a cost-effective mix design with
optimum strength can be achieved by selecting or blending the right binder
for given tailings and mixing water.
Cement cost often constitutes a large proportion of the total fill production
cost. Mine operators use combinations of other materials that have
cementitious properties to reduce the amount of cement needed (Grice,
1989). Alternate binders include fly ash, granulated iron blast-furnace slag,
and silica fume.
431
Mine Fill
100
Type I—Ordinary Portland cement
Type II—Modified cement
Type III—Rapid-hardening Portland cement
Type IV—Low heat Portland cement
Type V—Sulfate-resisting Portland cement
GP cement-A
GP cement-B
GP cement-C
GP/FA blended-A
GP/FA blended-B
GP/FA blended-C
GB slag
Portland/slag blended
90
80
Compound (%)
70
60
50
40
30
20
10
0
C3S
C2S
C3A
C4AF
Compound
C3S—Tricalcium silicate (3CaO . SiO2)
C2S—Dicalcium silicate (2CaO . SiO2)
C3A—Tricalcium aluminate (3CaO . Al2O3)
C4AF—Tetracalcium aluminoferrite (4CaO . Al2O3 . Fe2O3)
FIGURE 8.30
Composition of the main compounds for a number of cement types. (From Saw, H. and
Villaescusa, E., Research on the mechanical properties of minefill: Influences of material particle size, chemical and mineral composition, binder and mixing water, in H.J. Ilgner, ed., Minefill
2011, Proceedings of the 10th International Symposium on Mining with Backfill, Cape Town, South
Africa, March 21–25, 2011, pp. 143–152, SAIMM, Johannesburg, South Africa.)
8.6.1.4 Admixtures
Chemical admixtures can be used to influence the setting and consolidation
of cemented fill mixtures during their transportation and placement into
stope voids (Weatherwax et al., 2011). Fill slump tests to study water addition, cement workability, and strength development time in conjunction with
admixture dosage need to be considered for a mix design. Typical admixtures include accelerants, dispersants, stabilizers, and activators. Accelerants
can be used to increase the speed of hydration of the cement and provide a
faster early strength and facilitate quicker stope cycle times. Dispersants can
improve fill flow characteristics by freeing up available water, resulting in
additional workability and fluidity and more efficient hydration. Stabilizers
form a protective barrier around the mineral constituents of cement, delaying
hydration. Activators break those barriers allowing hydration to commence.
8.6.1.5 Mixing Water
The mixing water has three main functions: (1) reacting with the cement
powder, thus producing hydration; (2) acting as a lubricant, contributing to
432
Geotechnical Design for Sublevel Open Stoping
the workability of the fresh mixture; and (3) securing the necessary space in
the paste for the development of hydration products (Popovics, 1992). The
amount of water in a fill mix can have a significant effect on fill strength
development. However, cement hydration is not the only controlling factor,
given that the method of fill placement also influences the water content
within a fill mix. For example, for every ton of fill placed at a density of 65%
(solids by weight), 0.2 tons of water will require drainage (Cowling, 1998).
Research conducted by Wang and Villaescusa (2001), Li et al. (2004), and
Benzaazoua et al. (2004) has shown that impurities in the mixing water can
cause a strength reduction in any type of mine fill. The impurities can either
be dissolved or suspended in the water. The amount of strength reduction
can change with the type of tailings and the binder dosage used. In certain
cases, contaminated water can be used for fill purposes by mixing it with
fresh water. Nevertheless, it is important to determine whether the impurities may lead to strength reduction.
The water/cement ratio is an important factor influencing the resulting
fill strength. However, it is difficult to optimize, as it is a function of the
fill method and placement into a stope void. Fill needs to be workable to
achieve appropriate flow properties for reticulation and must also be able
to be consolidated and shaped into different forms. A balance of mix design
and reticulation will achieve the required flow properties and strength gain.
8.6.1.6 Mix Design
Mine fill design largely depends on the availability of constituent materials and their physical and chemical properties, the required fresh properties
(flowability), strength, and durability. A typical mix design for CPF, CHF,
CAF, and CHF is shown in Table 8.3.
The required mine fill strength is a function of the mining method, geometry of orebody and stope, and the possible failure modes. Mitchell and
Roettger (1989) describe the potential failure modes of cemented mine fill
used to support steeply dipping ore zones. Failure modes include sliding,
crushing, flexure, and caving. Sliding can occur due to low frictional resistance between the fill and a rock wall. Crushing occurs when the induced
stress exceeds the compressive strength of the fill mass. Flexural failure
occurs when the fill mass has a low tensile strength, caving can be a result of
arching, and rotational failure may occur due to low shearing resistance at
the rock wall. When mine fill is considered as a roof slab, the analysis methods developed by Evans (1941) and later modified by Beer and Meek (1982)
can be applied. In addition, a method for roof design procedure considering
plane strain has been described by Brady and Brown (2004).
The mechanical properties for fill design are usually determined by laboratory testing. The most common tests are the UCS test and the triaxial (unconsolidated undrained) test. Strength development is a function of the type of
fill material (tailings or waste rock), cement type, cement dosage, water, solids
433
Mine Fill
TABLE 8.3
Typical Mine Fill Mix Design
Description
Tailings (%)
Waste rock <2 to 300 mm
diameter (kg/m3)
Coarse aggregate
10–40 mm diameter (%)
Sand (%)
Cement (%)
Solids (%)
Water/cement ratio
CPF
CHF
CAF
CRF
96
—
94
—
—
—
—
2017
—
—
86
—
—
4
70
10
—
6
76
5
10
4
—
2
—
5
—
2
Source: Saw, H. and Villaescusa, E., Geotechnical properties of mine fill, in C.F. Leung, S.H. Goh, and
R.F. Chen, eds., Proceedings of the 18th South East
Asian Geotechnical & Inaugural AGSSEA Confer­
ence, Singapore, May 29–31, 2013, Research
Publishing, Singapore.
percentage, water/cement ratio, curing time, and temperature. The typical
UCS of CPF at 28 days ranges from 0.4 to 1.7 MPa. The UCS of CHF and CAF is
about 1 and 2.5 MPa, respectively. The typical uniaxial tensile strength (UTS)
of CPF at 28 days ranges from 0.1 to 0.3 MPa and the UTS of CAF ranges from
0.2 to 0.8 MPa. The shear strength of a mine fill is usually obtained by unconsolidated undrained triaxial compression testing. Occasionally, consolidated
undrained and consolidated drained tests are conducted to determine the
effective stress parameters used in analyses of fill mass stability and in the
design of fill barricade systems (Kuganathan, 2005; Helinski et al., 2011a).
Typical shear strengths of mine fills are presented in Table 8.4.
TABLE 8.4
Typical Shear Strengths of Mine Fills
Total Stress
Mine
Fill Type
CPF
CPF
CPF
CAF
CAF
Effective Stress
Curing
(Days)
Test
Method
Cohesion
(c) (kPa)
Friction (Φ)
(Degrees)
Cohesion
(c′) (kPa)
Friction (Φ′)
(Degrees)
28
2
2
106
93
UU
CU
CD
UU
UU
208
—
—
400
1450
39
—
—
32
44
—
147
85
—
—
—
31
38
—
—
Source: Saw, H. and Villaescusa, E., Geotechnical properties of mine fill, in C.F. Leung,
S.H. Goh, and R.F. Chen, eds., Proceedings of the 18th South East Asian Geotechnical
& Inaugural AGSSEA Conference, Singapore, May 29–31, 2013, Research Publishing,
Singapore.
434
Geotechnical Design for Sublevel Open Stoping
8.6.2 Stope Preparation
Following completion of stope production and cleaning, the resulting void
needs to be prepared for fill material delivery. A water drainage system by
means of fill barricades must be established at each of the stope access drives
and drawpoints. Fill barricades are defined as permeable, free-draining
structures used to initially support the fill mass (Grice, 1989).
8.6.2.1 Design Criteria for Fill Barricades
Different fill methods and types will require slightly different functions
from a barricade system. Nevertheless, a similar approach is required for
their safe and effective design and construction. This includes size, position,
loading, materials, and curing.
The size of a drive in which a barricade is located influences the pressure on the barricade. Modern stope access drives and drawpoints are constructed using large machinery to increase production, hence increasing
the size of the drives. This requires larger and more robust barricades to
support the pressures acting on larger areas. Positioning can also affect the
load on fill barricades, with those built closer to a stope brow experiencing
higher loads (Mitchell et al., 1975). When access geometry allows, the common practice is to build a barricade at least a drive’s width away from a stope
brow. The likely direction of loading must be taken into account to construct
the barricade, so that the barricades are built perpendicular to the drives to
transfer load to the surrounding rock. The material used for barricade construction also contributes to the barricade performance. Materials include
timber, threaded bar, steel I beams, welded wire mesh, bricks, shotcrete, and
waste rock. A combination of good quality materials along with local experience can produce an inexpensive barricade that provides the level of sill
support and drainage required. Barricades made of shotcrete or bricks must
be given time to cure prior to fill delivery.
8.6.2.2 CHF Barricades
Stope preparation for HF requires that a strong but permeable barricade
be constructed following stope completion and before fill commencement
(Figure 8.31a). Barricades effectively seal stope voids, so that slurry, mud, and
the fill mass does not cause a mud rush. However, a considerable amount of
excess water needs to be drained (Figure 8.31b).
The barricades can be constructed in several ways, including using permeable masonry blocks. In addition, timber or steel frames and welded
wire mesh that are sprayed with shotcrete and have drainage points placed
throughout the barricade can also be used. The application of polyethylene
drain pipes to allow the water to percolate through the fill leaving the fines
behind has been described by Kuganathan and Neindorf (2005). In order to
435
Mine Fill
(a)
(b)
FIGURE 8.31
Example of fill barricades (a) prior to fill and (b) during initial stope filling. (From
Winder, K., The introduction of cemented hydraulic fill to the Gossan Hill Mine, MEngSc in
Mining Geomechanics thesis, Curtin University of Technology, Kalgoorlie, Western Australia,
Australia, 2006.)
control the excess water, barricades are built with drainage pools where the
water is gathered and pumped to the surface.
Barricades are subjected to low pore water and earth pressures and are
designed for applied pressures of around 100–200 kPa. Grice (1989) describes
porous bricks as being a common construction material. However, blinding of bricks, as the fines try to decant with excess water, can cause problems (Kuganathan, 2002). A standard barricade is 400–500 mm thick. Water
within the fill applies pressure radially, while the earth pressure is a function of the fill particle weight. The horizontal load is typically one-third of
the vertical load and arching reduces the vertical load (Thomas et al., 1979;
Helinski et al., 2011b). Kuganathan (2005) stated that arching (Figure 8.32)
allows a fill mass of 500 kPa strength to withstand exposures exceeding
100 m in height, whereas if no arching were present, a fill strength of 2 MPa
would be required.
Stope voids cannot simply be filled in one pass and the fill rate must be
designed to control the pooling of water, the hydraulic head, and the pore
water pressure acting against the stope barricades. This results in a noncontinuous filling process, which increases the stope cycle time. The first pour
of HF is usually either close to the height of the barricade or just above that
level. The fill is left for a period of time, so that the solids can settle (consolidate) and the majority of the water drains, thus providing a solid footing that
reduces the load on the barricades as filling continues.
In some cases, when the water level on the top of the fill exceeds 1.5 m,
filling is usually ceased until more water is drained (Landriault, 2001). Water
pooling can lead to the fine binder particles decanting with the excess water
as well as gravity separation between fine and coarse particles. Consequently,
436
Geotechnical Design for Sublevel Open Stoping
Barricade
FIGURE 8.32
Qualitative stress field within a fill mass. (From Kuganathan, K., Geomechanics of mine fill, in
Y. Potvin, E. Thomas, and A. Fourie, eds., Handbook on Mine Fill, ACG, Perth, Western Australia,
Australia, 2005, pp. 23–47. With permission from ACG.)
layers of strong fill containing the binder and bands of tailings that have
little strength can be formed. Importantly, pooling may cause sections of the
fill to be saturated, potentially leading to piping failures, while in secondary
stopes, where no binder is added, liquefaction can occur.
Fill liquefaction occurs when the pore pressure is equal to the normal
stress, resulting in zero effective stress. It can result from an earthquake,
blasting, or any other dynamic loading activity that suddenly reduces the
drainage pathway while increasing the hydraulic head pressure. A sudden
increase in pressure can cause a fill barricade to fail. Under conditions of no
effective stress, uncemented fill mixtures can flow like a liquid, as they cannot resist a sudden increase in applied stress.
Erosion pipe failures are caused when excess water forms an erosion pipe
through which water flows instead of percolating through the fill mass
(Grice, 1989, 2005a; Bloss and Chen, 1998). Hydrostatic pressure can build up,
loading the barricade, which fails through flexural bending. Fill slurry leaks
from the failure gap or crack and through erosion creates a larger pipe/hole
(Figure 8.33). This potential problem can be prevented through engineering
design, quality construction, and effective geotechnical monitoring, leading
to a safe filling strategy (Winder, 2006). Fill should be placed at a suitable rate
while the mix should have a solids density exceeding 70% to prevent excessive ponding of water. Furthermore, the barricade should be free-draining,
but have no leaks that can promote pipe erosion (Figure 8.34). Figure 8.35
shows a modern masonry barricade under construction and the various
drainage pipes installed (Winder, 2006).
437
Mine Fill
Surface pooling
Fill surface
Saturated fill zone
Erosion pipe
Barricade
failure
FIGURE 8.33
Schematic of erosion pipe failure and details of failed barricade. (From Grice, A.G., Fill
research at Mount Isa Mines Limited, in F.P. Hassani, M.J. Scoble, and T.R. Yu, eds., Innovations
in Mining Backfill Technology, Proceedings of the Fourth International Symposium on Mining with
Backfill, Montreal, Quebec, Canada, October 2–5, 1989, pp. 15–22, Balkema, Rotterdam, the
Netherlands.)
(a)
(b)
FIGURE 8.34
Example of (a) free-draining and (b) fill-leaking barricades. (Courtesy of Mount Isa Mines,
Mount Isa, Queensland, Australia.)
8.6.2.3 CRF Barricades
Although the majority of a filled CRF stope is occupied by waste rock, this
rock is mixed with cementitious slurry having a high water content, thus
requiring drainage. The positioning of the barricade will determine if waste
rock is likely to load the barricade. The closer a barricade is to a stope void,
438
Geotechnical Design for Sublevel Open Stoping
Threaded bar
shear pins
Decant lines
from level above
Drain pipes
Concrete footing
FIGURE 8.35
Masonry fill barricade under construction showing draining devices. (From Winder, K.,
The introduction of cemented hydraulic fill to the Gossan Hill Mine, MEngSc in Mining
Geomechanics thesis, Curtin University of Technology, Kalgoorlie, Western Australia,
Australia, 2006.)
the more likely it will have to act as a structural wall and support some of the
waste rock. Otherwise, it will become a conventional barricade and can be
made of timber structure, welded wire mesh, and shotcrete to provide a seal.
Drainage can be achieved by placing piping into the barricade.
In cases where the waste rock will be supported, threaded bar shear
pins can be installed into the drive walls to achieve a more robust structure. Steel beams and welded wire mesh with shotcrete provide a strong
structure with sufficient load-bearing capacity and free-draining capability
as described earlier.
8.6.2.4 CPF Barricades
In theory, CPF is placed with minimal excess water. In practice, however,
due to fill distribution, placement, and segregation, some drainage must be
accounted for. Therefore, the barricades must be designed to account for fill
weight, hydraulic head, and some drainage. The simplest way to construct a
barricade involves placing a waste rock plug at a stope brow and then shotcreting over a number of gypsum bags as shown in Figure 8.36. Waste rock
barricades of approximately 3 m height, in conjunction with welded wire
mesh and hessian backing sheet, erected on top of the bund and sprayed
Mine Fill
439
FIGURE 8.36
Typical CPF barricade.
with shotcrete have been reported by Foster et al. (2011). However, undercutting by the mining method led to failure at the interface of the waste rock
bund and the base of the shotcrete. Consequently, a barricade design consisting of welded wire mesh, hessian backing sheet, and shotcrete to the full
dimensions of the drive was then implemented.
Nienbauer (2011) has described the routine construction of paste fill barricades in which the material is supplied to the mine as a kit. Tubular telescopic steel formwork is assembled and attached to the rock using short rock
bolts. A layer of welded wire steel mesh, a layer of hessian, and another layer
of welded wire steel mesh are attached in sucession to the steel frame. The
assembly is cut to fit the profile of the excavation and three 100 mm-thick layers
of shotcrete are sprayed to form the barricade. Vibrating wire piezometers
are installed inside the barricade, where pressures of approximately 40 kPa
have been reported.
As noted previously, filling a stope in stages reduces the load on a barricade.
Therefore, the first stage consists of filling a stope to approximately 1–3 m
above its brow. The fill is then allowed to cure to approximately 150 kPa, thus
providing a bottom plug, so that the rest of the stope can be filled without
placing large loads on the barricades.
8.6.3 Material Delivery
Sublevel open stoping requires a high rate of fill placement into the
open voids. This reduces the available options for fill delivery systems,
with boreholes and pipelines being the best methods for rapid fill placement. The requirements of such reticulation systems change according
440
Geotechnical Design for Sublevel Open Stoping
to the materials being transported. However, common to all systems that
transport slurries is the requirement to be regularly flushed to clean the
boreholes and pipelines.
8.6.3.1 Rock Fill Passes
RF is usually transported dry to the underground stope delivery point. In
the case of CRF, mixing of the aggregate and slurry usually occurs as they
are discharged into the stope. The aggregate is usually transported underground through a fill pass or fill raise from the surface with a number of
factors affecting efficiency. The most important are particle size distribution (fine and coarse material), required volume of fill, water inflow, and
vertical opening instability at a particular mine site. Particle size attrition
due to vertical fall can also occur. Typical fill raise diameters range from
2 to 3 m.
The best method of reducing interlock arch blockage (Landriault, 2001)
is to screen for oversized particles on the surface. Experience has shown
that when the fill pass diameter to maximum particle size exceeds 5, the
frequency of interlocking arches is very low. In addition, blockage can
also occur when a cohesive arch is formed. This is the result of fine, sticky,
particles adhering to one another within the fill pass, thereby blocking
the flow. Water inflow can also affect the fill flow within a pass. The fill
is usually transported dry to minimize the risk of blockages. However, if
water comes into contact with fill during transportation, it increases not
only the risk of blockage, but also the moisture content and the water/
cement ratio.
Aggregates are transported from surface to underground through fill
passes to a distribution point. From this point, the aggregate is transported
by trucks or conveyor belts to the individual stopes. A fill shute is used at
the distribution point that allows the flow of materials to be controlled and
maintain the fill pass either choked or empty. Choked fill passes experience
less wear and attrition, but risk blockage.
8.6.3.2 Slurry Fill Passes
Hydraulic and paste fill are transported underground through reticulation
systems consisting of boreholes and pipelines. Boreholes can be lined with
steel pipes where the rock mass is altered, weathered, or highly jointed.
Ideally, two boreholes are drilled, so that if one is blocked, the backup hole
can be used. The hole collar must be located where a drill rig can promptly
redrill it if a blockage does occur.
Slurry is fed underground using gravity, with the hydraulic head generated used to provide the energy required to deliver the slurry into the stoping void. The reticulation system reaches equilibrium when the static head in
the vertical direction matches the friction losses in the horizontal direction.
Mine Fill
441
If pumping is required, centrifugal and positive displacement pumps which
can maintain a constant feed can be used to reduce pipeline wear.
Slurry velocity is an important factor in pipeline wear and free-fall sections must be minimized if possible. On the other hand, paste fill does not
have a critical flow velocity and will not settle even when stationary in the
system, where it remains until the vertical head equals the yield stress and
flow will commence.
The greatest pressure is experienced at the toe of a borehole, where
horizontal distribution begins. High-pressure, heavy wear-resistant components should be used to account for excessive wear at the pipeline distribution corners. Pipelines used on mine levels to distribute the fill are
usually steel or rubber-lined pipes. Closer to a stope, high-density, flexible,
polyethylene piping that can be readily positioned at the stope placement
point can be used.
8.6.4 Fill Placement
Fill placement into a stope void can cause segregation, thus reducing the ultimate strength at eventual exposure. Fill is placed from either a pass or stope
access drift and allowed to fall over a vertical distance. The segregation is a
function of the stope geometry and the angle of fill placement into the resulting stoping void (Landriault, 2001). The segregation and placement effects
for the final fill pour can vary for different fill materials.
8.6.4.1 CHF Placement
The most critical issue in the placement of HF is the rate of material delivery
into the stope void. The hydraulic head pressure rises very quickly due to
the high amount of water present (40%–50% by volume). Safe fill rates are
determined by the fill plant production capacity, stope geometry, barricade
permeability, and strength and practical experience at a particular mine site
(Winder, 2006). In general, the initial pour is allowed to drain and cure to
form a plug or footing for the rest of the stope. This allows the barricade to
just provide drainage instead of drainage and fill support. Some fill separation is likely to occur as a result of the water content and particle size distribution, as well as any free fall distances experienced during transport and
placement into the stope void.
8.6.4.2 CRF Placement
The solid (RF) and fluid (CHF) components of CRF are placed into the stope
voids simultaneously. The recommended approach is to direct the CRF mixture to the center of the void. If the CRF is placed to fill preferentially one side
of the void, then the CHF will drain away from the RF (Bloss, 1992, 1996). The
central placing of the CRF mixture will form a cone of RF with a CHF beach
442
Geotechnical Design for Sublevel Open Stoping
FIGURE 8.37
CHF beach formation following CRF placement. (Courtesy of Mount Isa Mines, Mount Isa,
Queensland, Australia.)
forming around the outside against other future stope walls (Figure 8.37).
The outside CHF beaches ensure future mass stability, as the central core
will likely have loose and highly compacted and cemented fill zones (Barrett
and Cowling, 1980).
The ratio of RF to CHF is also an important factor to minimize the
formation of uncemented fill near a stope boundary. The average ratio ranges
from 1.5 to 2 (Grice, 1989), with high ratios being required to achieve low fill
costs. A difficulty in achieving a high ratio and low cost arises when a stope
void is not square in plan, as it is difficult to vertically and centrally fill nonsymmetrical stope shapes. Stope design must ensure that the RF cone is not
in contact with a stope boundary that will be subsequently exposed.
8.6.4.3 CPF Placement
The rate of CPF placement must be controlled, so that the hydraulic head
does not cause failure of the stope barricades. An initial pour of paste fill is
allowed to cure to provide a stable footing for the rest of the stope. Given that
little excess water is produced by CPF, once the footing is set, and given the
nonsettling nature of the fill, the mix should stay in a homogeneous, stable
state. In practice, a small amount of water will be forced out and some minor
drainage may be required. In addition, ensuring tight fill against the back
with CPF is an important issue in some cases.
Mine Fill
443
8.7 Fill Monitoring and Quality Control
Monitoring of the fill operations process is required to produce a fill consistently achieving the required properties. The key stages of the fill process
include material supply, fill plant, reticulation, placement, and barricade performance. Problems can occur at every stage and may be compounded due
to the peculiarities and conditions of a particular mine site. Grice (2005a,b)
has identified a number of issues that require monitoring including fill supply and fill plant monitoring. For fill supply, the tailings mineralogy, changes
in the metallurgical process, and binder supplies were identified. For the fill
plant, the fines content, the slurry density, the cement content, the solids
mass flow, and the flushing water quantities were identified.
8.7.1 Fill Supply
Changes in the tailings mineralogy can result from variability of ore sources
or processing of adjacent orebodies, and could change the resulting fill
mass strength. In addition, the particle size distribution may change with
mill throughput changes, resulting in coarser or finer grind sizes. Regular
testing of processing, washing, and mixing water is required to assess the
potential effects on short- and long-term fill strength. Chemical additives
may be needed to account for any effects of water quality. Importantly, the
effects of any change of cement supplier or source must be quantified using
laboratory testing.
Quality assurance of the fill product is focused on ensuring that the particle size distribution is suitable to ensure adequate permeability. Winder
(2006) reported a quality control process in which samples were taken once
during a shift and stored until the end of the month when between 3 and 10
samples were selected to be sent off site for laser sizing. Figure 8.38 shows
some of the laser sizing results of the HF produced.
8.7.2 Fill Plant
Monitoring of the particle size distribution resulting from the desliming process is required. Changes can occur due to changes in throughput, treatment of different ore types, or wear of the sieves and cyclones.
Blockage of cement may occur in silos or during transport and lumping
can result in overaddition, thus increasing the cost. Density and volumetric flow gauges are required to monitor full mass balances of inputs and
outputs. Flushing water is used to regularly clean the fill plant and this
may affect the water/cement ratio of the mix. The final mix product should
be tested using slump tests with samples taken to a laboratory for uniaxial
compressive testing.
444
Geotechnical Design for Sublevel Open Stoping
30
Cummulative (% passing)
25
20
9-Jun
24-Jun
30-Jun
8-Jul
11-Jul
27-Jul
Criteria
15
10
5
0
0
10
20
30
40
50
60
Size (microns)
FIGURE 8.38
Monitored particle size distribution for CHF at the Scuddles Mine, Western Australia. (From
Winder, K., The introduction of cemented hydraulic fill to the Gossan Hill Mine, MEngSc in
Mining Geomechanics thesis, Curtin University of Technology, Kalgoorlie, Western Australia,
Australia, 2006.)
Monitoring of an HF plant operation involves real-time data acquisition
using pipe density gauges. The fill plant output volume must be closely monitored to control the hydraulic head formed in the stopes being filled. Winder
(2006) has reported Marcy scale checks every 3 h and regular cyclone checks
to ensure that the pressure was sufficient. Winder (2006) also described daily
reports showing the following:
• Average solids density placed for the last 24 h
• The dry tons of HF produced
• The system run hours
8.7.3 Fill Reticulation
Monitoring systems are required to detect any tailing loss in boreholes or
pipelines. Leakage or blockage points can be formed and can be detected
using pressure-monitoring devices. Fill samples taken at the stope placement
point determine if the system is coming into contact with additional water
during transportation.
8.7.4 Fill Placement
Cavity monitoring of resulting stope voids determines the actual stope
shapes and volumes, enabling the required amount of fill and filling time
Mine Fill
445
FIGURE 8.39
Water flowing over a V-notch weir installed in a fill barricade sump. (From Winder, K.,
The introduction of cemented hydraulic fill to the Gossan Hill Mine, MEngSc in Mining
Geomechanics thesis, Curtin University of Technology, Kalgoorlie, Western Australia,
Australia, 2006.)
to be estimated. Visual assessments can be used to monitor any amount
of water pooling on the stope surface, prior to any subsequent fill pours.
Measurements of actual fill height can be compared with fill plant output
and observations of water pooling. This provides an initial assessment of fill
water percolation and drainage. In addition, piezometers and earth pressure
cells can be placed at known locations within a stope to monitor and control
fill placement (Winder, 2006).
8.7.5 Barricade Performance
Visual inspection of a barricade can provide a good, initial indication of its
strength and performance. Pore water and earth pressure cells can be placed
at the stope barricades to quantify and monitor their loading. V-notch weirs
can be used to collect the water draining from a barricade to help quantify
the fill mass permeability (Figure 8.39). For CPF and CRF, any leakage from
a barricade can be used to assess the effects of a fill mass self-weight. Water
leakage from the walls and backs of the stope access drives provide an indication of water ingress into a stope.
9
Dilution Control
9.1 Introduction
Dilution is defined as the low-grade material (waste or fill) that is mined and
processed together with the ore stream, thus reducing its value. Ore loss
refers to any unrecoverable economic ore left inside a designed stope boundary. This ore may be broken, in place as pillars, or not properly blasted. Any
valuable ore not recovered by the mineral processing system also constitutes ore loss. Dilution and ore loss are always defined and quantified with
respect to an idealized (planned) stope boundary. However, dilution is not
always defined in an identical fashion. The two most widely used equations
to quantify dilution are (Pakalnis et al., 1995):
Waste tonnes mined
Ore tonnes mined
(9.1)
Waste tonnes mined
Ore tonnes mined + Waste tonnes mined
(9.2)
Dilution 1 =
Dilution 2 =
Equation 9.1 is more sensitive to stope wall slough. In addition, a number of
other definitions have been provided by Pakalnis et al. (1995):
Undiluted in situ grade from drillholes
Sampled assay grade at drawpoint
(9.3)
Dilution 4 =
Undiluted in situ grade reserves
Mill headgrades from same tonnage
(9.4)
Dilution 5 =
Tonnage mucked - Tonnage blasted
Tonnage blasted
(9.5)
Dilution 3 =
447
448
Geotechnical Design for Sublevel Open Stoping
Dilution 6 =
Footwall slough (m) + Hangingwall slough (m)
Orebody width (m)
Dilution 7 = Fill (Tonnage actually placed - Calculated to fill void)
(9.6)
(9.7)
In order to quantify dilution, an orebody must be properly delineated and
the extracted volumes must be effectively measured. Waste rock that is
left inside a stope (selective mucking) is often not considered as dilution.
In addition, high dilution may not necessarily mean a low mining recovery. However, when dilution decreases, there is a higher risk of leaving ore
behind (Figure 9.1). In some cases where dilution decreases, an increase in
ore loss is also experienced as shown in Figure 9.2.
In general, dilution can be divided into three categories: internal, external,
and ore loss (Figure 9.3). Internal dilution usually refers to the low-grade
material contained within the boundaries of an extracted stope. It can be
caused by insufficient internal delineation of waste pockets within an
orebody. It also occurs in situations where the mining method dictates a
minimum width of extraction. External dilution refers to the waste material that comes into the ore stream from sources located outside the planned
stope boundaries (Villaescusa, 1995). Low-grade material from stope wall
overbreak, contamination from backfill, and mucking of waste from stope
floors are typical examples of external dilution.
100
Recovery (%)
96
92
88
84
80
0
20
40
60
80
Dilution (%)
FIGURE 9.1
Relationship between recovery and dilution at the Hemlo Gold Mine, Canada. (From
Andersen, B. and Grebenc, B., Controlling dilution at the Golden Giant Mine, CIM Mine
Operator’s Conference, Timmins, Ontario, Canada, 1995, Paper 4, 14pp.)
449
Dilution Control
40
HW
35
FW
Total
Dilution (%)
30
25
20
15
10
5
0
0
2
4
6
8
10
Ore loss (%)
12
14
16
18
FIGURE 9.2
Relationship between dilution and ore loss at a Western Australian Mine.
Mine dilution
External
Unplanned
Instability
contamination
mining methods
Internal
Planned
Nature of
mineralization
mining methods
Ore loss
Geological
Exploration
orebody
delineation
FIGURE 9.3
A general classification of dilution. (From Villaescusa, E., Geotechnical design for dilution
control in underground mining, in R.K. Singhal, ed., Proceedings of the Seventh International
Symposium on Mine Planning & Equipment Selection, Calgary, Alberta, Canada, October 5–9, 1998,
Balkema, Rotterdam, the Netherlands, pp. 141–149.)
9.2 Types of Dilution
9.2.1 Internal Dilution
Internal dilution is the amount of waste rock material that is planned to be
blasted, mucked, hauled, hoisted, and processed along with the economical
450
Geotechnical Design for Sublevel Open Stoping
Overcut
Overcut
Blasting
outline
Planned
dilution
Waste
Undercut
Ore
Undercut
10500 N
10475 N
10450 N
FIGURE 9.4
Internal dilution from waste pockets in sublevel open stope mining.
Narrow
orebody
Planned
dilution
FIGURE 9.5
Internal dilution when the mining method dictates a minimum width of extraction.
material included within a designed stope boundary. Planned dilution usually occurs due to complex orebody shapes or due to the occurrence of waste
rock zones contained within an orebody. In such cases, the stopes may be
designed to include waste rock outside the delineated orebody contacts or
within the stope itself (Figure 9.4). Planned dilution can also be caused by
waste rock included within a stope boundary (usually the footwall) to force
favorable flow of broken ore to the drawpoints. In the case of narrow orebodies, the width of the mining equipment often requires widening of mining
blocks and the inclusion of waste rock within a stope boundary (Figure 9.5).
9.2.2 External Dilution
External (unplanned) dilution occurs when material below the cutoff grade
may be drilled, blasted, loaded, transported, and processed in the concentrator along with the planned material. Waste material such as rock or
fill sloughing from unstable stope walls contributes to external ­dilution
(Figure 9.6). External dilution does not include ore grade material that
sloughs into a stope from adjacent stopes or pillars.
Dilution Control
451
FIGURE 9.6
External dilution from a hangingwall failure in a large underground bench stope.
9.2.3 Geological Dilution
Geological dilution refers to the waste rock or ore losses incurred during
the exploration and orebody delineation stages, when only an estimated
model of the orebody can be developed. A geological model is based on limited information and is unlikely to coincide exactly with the real orebody.
Therefore, the delineated orebody boundaries are likely to exclude ore and
to include waste (Figure 9.7). The magnitude of this problem is a function
of the sampling pattern for the mineralization type under study. Geological
dilution may comprise up to one-third of the total dilution depending upon
orebody complexity (Lappalainen and Pitkajarvi, 1996).
9.2.4 Ore Loss
Ore loss refers to the economic material that is left in place within the boundaries of a planned stope. Planned ore diaphragms (ore skins), unbroken stope
452
Geotechnical Design for Sublevel Open Stoping
Waste
Mineralization lost
Ore
Nonmineralization
gained
Ore
Waste
FIGURE 9.7
Nonselective stoping at mineralization boundaries. (From Price, I., Towards optimal mining,
AMIRA Annual Meeting, Kalgoorlie, Western Australia, Australia, 1993, pp. 82–94.)
areas due to insufficient blast breakage, nonrecoverable pillars left to arrest
stope wall instability, and insufficient mucking of broken ore within stope
floors are typical examples of ore loss.
Ore losses in sublevel open stoping arise from insufficient breakage around
stope corners, especially when the stope is located near the boundaries of an
orebody. Excessive hole deviation at the toes of very long holes may create large
burdens that are difficult to break up by the explosive charges. This is particularly true when the blasthole rings toe into the walls of an orebody (Figure 9.8).
9.3 Economic Impact of Dilution
The detrimental impact of dilution to the economics of the mining industry
has been well documented. Puhakka (1991) and Elbrond (1994) have recognized that waste rock dilution and ore loss exist during geological modeling
and evaluation, and influence decisions regarding cutoff grade, design of the
mining method, stoping, and ore concentration. Nevertheless, this is not a
completely cumulative process because dilution and ore loss may be compensated for during subsequent stages of the overall mining and concentration process. A conceptual diagram proposed by Elbrond (1994) indicates the
complexity of the problem (Figure 9.9).
Dilution is a source of direct cost as waste or backfill material is blasted,
mucked, transported, crushed, hoisted, processed, and stored as tailings.
If excessive dilution occurs during stope production, a need for secondary
453
Dilution Control
Oversize
blocks
Dilution
Ore loss
Oversize blocks
FIGURE 9.8
Effects of drilling and blasting on dilution, fragmentation, and ore loss. (From Lappalainen, P.,
Personal communication, 1997.)
drilling blasting may arise. Figure 9.10 shows that handling of oversize dilution material at the stope drawpoints significantly affects productivity.
Dilution is also a source of indirect cost as the dilution material may
adversely affect the metal recoveries and concentrate grades. A lost opportunity may result from directing resources to handling waste (as opposed to
ore) for the mill feed. Furthermore, ore processing facilities will be engaged
for material that contributes very little to the final useful metal production.
In most cases, mining and milling capacities are limited; these capacities
are affected by the displacement of ore by waste within the overall mining
and processing facilities. In some cases, the cutoff grade must be increased
to maintain mill head feed grade (Figure 9.11). Dilution may also cause an
overall decrease in the net present value, as the quantity of the total metal
produced may be reduced (Figure 9.12).
9.4 Parameters Influencing Dilution
The most common parameters influencing dilution and ore losses in underground mining are listed in Table 9.1. Five key stages ranging from an initial
orebody delineation program to the final extraction stage have been identified
454
Geotechnical Design for Sublevel Open Stoping
The real but unknown deposit
Ore lost
Estimated deposit
Internal dilution
Ore lost
Estimated deposit after the
decision of cut-off grade
Dilution
Ore lost in pillars
Deposit after mine design
Dilution from mining method
Ore drilled, blasted, but lost
Mined deposit
Dilution drilled, blasted, loaded, and
transported to the concentrator
Overbreak
Ore which becomes concentrate
Ore lost in tailings
Dilution treated by the concentrator
LHD shift/blockholer shift ratio
FIGURE 9.9
Sequence of dilution and ore loss. (From Elbrond, J., CIM Bulletin, 87, 131, 1994. With permission.)
16
12
8
4
0
0
20
40
Dilution (%)
60
80
FIGURE 9.10
A correlation between dilution and increased handling of oversized muck at the Hemlo Mine.
(From Andersen, B. and Grebenc, B., Controlling dilution at the Golden Giant Mine, CIM Mine
Operator’s Conference, Timmins, Ontario, Canada, 1995, Paper 4, 14pp.)
455
Dilution Control
Effect of mine grade on mill performance
Concentrate grade (% Ni)
14
13
Ore grade (% Ni)
1.5
2.0
3.0
12
11
10
9
8
7
6
85
90
95
Recovery to concentrate (% Ni)
100
Costs avoided by not moving 1 ton of dilution
Mine cost
$3.71
Mine cost (direct)
$2.31
Mine cost (indirect)
$16.00
$22.02
FIGURE 9.11
Detrimental effects of dilution at Inco Manitoba. (From Ashcroft, J.W., Dilution: A total quality
improvement opportunity, 93rd Annual General Meeting of CIM, Vancouver, British Columbia,
Canada, April 28–May 2, 1991, 47pp.)
within the mine design process. Management issues are also included, given
that in some cases they represent the most critical factor controlling dilution
(Ashcroft, 1991).
9.4.1 Dilution at the Orebody Delineation Stage
Orebody delineation is the process that establishes the size, shape, grades,
tonnage, and mineral inventory for the ensuing mining process. Efficient,
effective, and accurate delineation of a deposit is required to design a mine
in a manner that maximizes recovery, minimizes dilution, and optimizes
safety. Dilution cannot be planned or minimized if detailed geological
and geotechnical information is not available. Experience indicates that
increasing the information density is likely to decrease dilution and ore loss
(Figure 9.13). In cases where the stope geology is not well delineated, the
interpreted ore outlines are usually regular; the presence of waste inclusions
is then likely to remain unknown.
Improved orebody delineation can be achieved with the potential application of geophysical logging of percussion-drilled holes. Increased sampling of
456
Geotechnical Design for Sublevel Open Stoping
Operating margin
Millions USD/year
20
2.3% Ni
2.0% Ni
1.8% Ni
Mill Feed
500 kton/year
Extra
capacity
available
15
10
Resource
grade: 2.5% Ni
5
0
0
10
20
30
40
Dilution (%)
FIGURE 9.12
Effects of dilution on operating profit at Outokumpu mines. (From Lappalainen, P. and
Pitkajarvi, J., Dilution control at Outokumpu mines, Proceedings of the Nickel ‘96, Mineral to
Market, Kalgoorlie, Western Australia, Australia, November 27–29, 1996, pp. 25–29, AusIMM,
Melbourne, Victoria, Australia. With permission.)
an orebody boundary (Figure 9.14) would occur by designating an optimum
percentage of a delineation drilling budget for geophysical logging of percussion-drilled holes. Geophysical properties have the potential to be extrapolated hole-to-hole in order to provide a better estimate of the size and shape
of an orebody. Once the geophysical tools are calibrated, increased logging
productivity may be achieved since assaying is not required. Unfortunately,
geophysical logging is affected by the uncertainty in the interpretation of
lithology and grade from geophysical data.
9.4.2 Dilution at the Design and Sequencing Stages
At this stage, several extraction strategies to minimize dilution/ore loss can
be studied in advance to choose the best design alternative. Stable stope
and ore outlines are superimposed in order to detect volumes of waste rock
inside, and ore outside, the stope limits. Wall instability and any relevant
remedial measures are also identified. A stope shape must be drillable and
stable, and the walls must ensure proper flow of broken ore to the stope
drawpoints (Figure 9.15). Extraction factors that account for dilution as well
as economic studies in conjunction with stability analysis can be performed
to evaluate different design options.
Mine engineering, geology, and operating personnel should have a direct
input into this stage of the design. Back analysis from adjacent stopes based
on cavity monitoring system (CMS) surveys (Miller et al., 1992), drill and
Dilution Control
457
TABLE 9.1
Parameters Influencing Dilution
Orebody delineation
Under sampling of orebody boundaries
Errors in decisions regarding cutoff grades
Down hole survey errors
Lack of geotechnical characterization
Design and sequencing
Poorly designed infrastructure
Poor stope design (dimensions)
Lack of proper stope sequencing
Lack of economical assessment
Stope development
Nonalignment of sill horizons
Poor geological control during mining
Mining not following geological markups
Inappropriate reinforcement schemes
Drilling and blasting
Poor initial markup of holes
Setup, collaring, and deviation of blastholes
Incorrect choice of blasting patterns, sequences, and explosive types
Production stages
Mucking of backfill floors
Mucking of falloffs and stope wall failures
Contamination of broken ore by backfill
Leaving broken ore inside the stopes
Poor management of waste rock (tipped into the ore stream)
Mine management
Lack of supervision and communication
Excessive turnover of personnel
Limited time for planning
Lack of stope performance reviews
No documentation and proper training
Performance indicators based on quantity (focus on tonnes as
opposed to metal content)
Lack of leadership and vision
blast design, and general experience in the area should be used. The overall
design is enhanced when a planning engineer has sufficient time for drilling, blasting and ground support optimization, schedule modifications, and
other issues. In order to minimize the detrimental impact of stress redistributions, the mining sequences must be designed to avoid leaving blocks
458
Geotechnical Design for Sublevel Open Stoping
Ni > 1.0%
Stope
outline
Enonkoski
nickel deposit
Stope
outline
Section
K = 34.776
FIGURE 9.13
Effects of increased sampling in orebody delineation. (From Lappalainen, P. and Pitkajarvi, J.,
Dilution control at Outokumpu mines, Proceedings of the Nickel ‘96, Mineral to Market, Kalgoorlie,
Western Australia, Australia, November 27–29, 1996, pp. 25–29, AusIMM, Melbourne, Victoria,
Australia. With permission.)
or pillars of highly stressed rock and also to limit the number of openings
within the future pillar areas. Numerical modeling can be used to identify
areas of high and low stresses or sudden stress changes.
9.4.3 Dilution at the Stope Development Stages
Drive location has been shown to be critical for dilution control. Undercut
of stope walls by the access drill drives is likely to control the mechanical
behavior at the stope boundaries (Figure 9.16). Drive shape and size also
influence stope wall undercut. Incorrect positioning of sill drive turnouts
off access crosscuts may also create stope wall undercut leading to dilution.
Crosscuts need to be mapped, sampled, and interpreted prior to developing
the sill drives along an orebody. In cases where assay information is required
prior to sill turnout, a prompt assay turnaround is critical to maintain development productivity.
The quality (and quantity) of the geological face mapping of development is critical to minimize stope wall undercuts. Geologists should
highlight any overbreak beyond an established mining width. Prompt feedback to the operating personnel undertaking the development mining is
required. Routine geotechnical mapping of development faces must also be
459
Dilution Control
a
Real orebody
boundary
Grade of orebody
1
8a
4a
2a
a
a/2
Information
density
FIGURE 9.14
Effect of information density on head grade. (From Lappalainen, P. and Pitkajarvi, J., Dilution
control at Outokumpu mines, Proceedings of the Nickel ‘96, Mineral to Market, Kalgoorlie,
Western Australia, Australia, November 27–29, 1996, pp. 25–29, AusIMM, Melbourne, Victoria,
Australia. With permission.)
undertaken. Good control of drilling and perimeter blasting techniques can
be used to reduce wall damage in development access in order to minimize
stope wall undercut.
9.4.4 Dilution at the Production Drilling and Blasting Stages
The blasting process involves the interaction of the rock mass, the explosives, the initiation sequences, and the drill hole patterns. Consequently, a
blast design should account for the interaction of the existing development,
equipment, orebody boundary, and stope outline. Geological, geotechnical,
operational, and extraction design issues must also be considered. Blasting
performance is affected by the orebody geometry and drilling limitations
in terms of hole length and accuracy. Explosive consumption and performance determines the quality of fragmentation. However, an increase in the
specific consumption of explosive may also increase the damage to the host
rocks, increasing external dilution. The effects of blasting on stability can be
460
Geotechnical Design for Sublevel Open Stoping
CMS #1
Ore lost
Dilution
CMS #2
Ore lost
CMS #3
FIGURE 9.15
Constructed stope shape that does not allow free flow of broken ore.
determined based on measurements of blast vibrations, hole deviation, hole
angle, and the distance of the holes to the exposed stope walls. Figure 9.17
shows an example of a 3-1-3 drilling pattern likely to lead to completely different outcomes mainly on the basis of drill deviation and the related confinement during the detonation process.
9.4.5 Dilution at the Production Stages
Even at this relatively late stage, dilution and ore losses can still be minimized. Information from percussion blastholes can be used to locate zones
of waste within an orebody, thus enhancing orebody delineation. The blast
design could be revised based on the detailed information regarding zones
of ore and waste. Some holes might not be blasted (i.e., leaving a pillar), or
additional holes may be drilled. Drill cutting data can be used to identify the
ore-waste contact in production holes. However, these task-intensive operations (sampling, bagging, and assaying) are prone to inaccuracies, and the
turnaround time for the data analysis is often too slow for practical use.
In practice, information about the ore-waste contact is seldom acquired in
the production stages without the use of properly calibrated single-hole geophysical tools. An advantage of single-hole geophysics is that information is
461
Dilution Control
9800 E
4666 level
Metasediments
4633 level
g
Fra
lu
nta
me
nit
Planned stope
boundary
Final stope
boundary
4600 level
Section 10300
looking west
FIGURE 9.16
Typical section of Hemlo Mine showing hangingwall undercut of the development mining
drive. (From Andersen, B. and Grebenc, B., Controlling dilution at the Golden Giant Mine, CIM
Mine Operator’s Conference, Timmins, Ontario, Canada, 1995, Paper 4, 14pp.)
FIGURE 9.17
Same drilling pattern with different outcomes due to drill deviation.
immediately available, significantly reducing turnaround time. This is particularly beneficial in situations in which severe blasthole deviation is occurring, and the exact location of the ore-waste contact is undefined.
An example is the OMS-logg geophysical logging system developed by
Outokumpu for quick orebody boundary definition. The system uses percussion-drilled production blastholes, and the results are obtained almost
462
Geotechnical Design for Sublevel Open Stoping
Ni assay
%
8.0
Calibrated Ni assay vs. gamma-gamma (rock density)
High-grade
disseminated
6.4
4.8
3.2
1.6
300
Low-grade disseminated
315
330
Distance (m)
345
360
FIGURE 9.18
Ni grade correlated to density at the Black Swan deposit, using the OMS-logg system. (From
Lappalainen, P. and Pitkajarvi, J., Dilution control at Outokumpu mines, Proceedings of the
Nickel ‘96, Mineral to Market, Kalgoorlie, Western Australia, Australia, November 27–29, 1996,
pp. 25–29, The AusIMM, Melbourne, Victoria, Australia. With permission.)
in real time. The system measures physical characteristics of the rock
around the boreholes. Density, magnetic susceptibility, electric conductivity,
and radioactivity are measured. A calibration for grade can be established
based on the different physical characteristics of ore and waste in sulfidic
Ni-orebodies (Figure 9.18). Measurements can be taken in 51–76 mm holes.
The logging rate ranges from 10 to 20 m/min.
In bench stoping, inspection and floor preparation before firing and mucking commence, minimize ore contamination during subsequent mucking. In bottom-up extraction sequences, the load-haul-dump (LHD) units
may dig holes in the floor and dilute ore with fill. Mucking units may also
ramp up and leave broken ore in the stope floors, and in cases where fill
support is required before bench completion along strike, contamination at
the fill-blasted ore interface may also occur. A training program on draw
point inspection for grade, ore contamination, and stope status (stability)
is required to control dilution. The stopes must be inspected several times
through a mucking shift to check the LHD tramming route and the state of a
stope. The condition of the hangingwall, footwall, and back must be assessed
during these inspections. Any significant falloff, overbreak, or underbreak
should be recorded, given that variations from planned designs could affect
stability and place at risk further extraction in adjacent stopes.
A stope performance review must be undertaken following the completion
of production blasting. These reviews are needed to improve performance
and to determine what lessons can be learnt and what improvements can be
made. Geology, mine planning, and operations personnel must be involved.
The performance review compares the laser (CMS) surveyed void with the
463
Dilution Control
planned stope void (see Figure 9.15). The differences can be due to blasting
overbreak, stope wall failures, pillar failures, and insufficient breakage. The
variations from the planned volumes are used to determine actual tonnage
and to estimate the extraction grade for each stope. These can be used to
undertake the final economic analysis and to optimize future extraction in
similar conditions.
9.4.6 Dilution Issues for Mine Management
Although geologists, engineers, and operators are involved in the mine
design process, mine managers must ultimately be accountable for the success of a dilution control plan (Figure 9.19). Dilution control and ore losses
must be managed within a global program of optimization for cost control
and increased safety (Figure 9.20). The choice of an option that minimizes
dilution may disrupt scheduling, and low levels of dilution could be sometimes justified in the context of a particular total mining scenario.
In some cases, dilution and ore loss are not assessed because the geology
and related costs are not sufficiently well known. At best, critical decisions
are simply based on the experience of the drilling and blasting designer. In
other cases, when a decision is taken, experience and rules of thumb are used
instead of calculations based on grade. This is often due to a lack of real
data. Management must develop performance indicators based on quality
rather than quantity with specific focus on metal tonnes, overbreak, and dilution control. Mine managers must recognize the potential for improvement
within their own mine environment. Most of the required understanding of
what comprises dilution and the tools to quantify it already exist.
Variables influencing mine dilution
Input
Process variables
Management
Geologists
Drillers
Samplers
Blasters
Geologists
Attitude
Grade variation
Drilling accuracy
Sampling accuracy
Wall quality
Interpretation
of ore contacts
Design layout
Engineers
Rock
Mechanics
Sandfill
Bolters
Stress conditions
movement
Fill competency
Ground support
practices
Process Dilution
15
Management attitude
Orebody definition
Blasting design
10
Fill quality
Ground control
5
Others
0
FIGURE 9.19
Factors controlling dilution at the Inco—Thompson Mine. (From Ashcroft, J.W., Dilution: A
total quality improvement opportunity, 93rd Annual General Meeting of CIM, Vancouver, British
Columbia, Canada, April 28–May 2, 1991, 47pp.)
464
Geotechnical Design for Sublevel Open Stoping
Dilution (%)
Process out
of control
Firefighting
Quality
control
Quality
improvement
Quality
control
FIGURE 9.20
Quality improvement through quality control. (From Ashcroft, J.W., Dilution: A total quality
improvement opportunity, 93rd Annual General Meeting of CIM, Vancouver, British Columbia,
Canada, April 28–May 2, 1991, 47pp.)
9.5 Cavity Monitoring System
The CMS was developed by the Noranda Technology Centre, Montreal,
Canada, in order to measure fill dilution at some of the Noranda’s Mining
Group Operations (Miller et al., 1992). The system was developed as an
alternative to conventional stope surveys that had proven time-consuming,
unsafe, and sometimes restricted by line-of-sight problems (Figure 9.21).
Conventional surveys may cause considerable delays to a production
cycle, while the quality of the data may be affected by lack of adequate
access. This is particularly true in large open stopes, where up to 1 week may
be required to collect, reduce, and evaluate data for each stope (Gilbertson,
1995). Surveyor safety is also a primary consideration, as the conventional
equipment is placed very close to a stope edge in order to ensure the maximum void coverage. In addition, accurate surveys of long excavations such
as orepasses are not possible with conventional systems.
The CMS system was developed to determine prompt and accurate threedimensional information on the volume and shape of empty voids, such
as an extracted stope (Miller et al., 1992). The instrument uses a laser survey range finder integrated within a motorized scanning head that can be
suspended inside a stope to obtain survey data remotely (Figure 9.22). The
system is able to measure the volumes of stopes, orepasses, cavings, etc.
465
Dilution Control
Equipment very close
to edge of stope
Significant shot
overlap requires
considerable editing
Stope requires
access until the
full survey has
been completed
Some areas cannot
be surveyed due to
loss of sight
Drawpoints
FIGURE 9.21
Conventional stope survey set-up using total station methods. (From Gilbertson, R.J., The
applicability of the caving measurement system at the Olympic Dam Operations, Proceedings
of the Sixth Underground Operators Conference, T. Golosinski, ed., Kalgoorlie, Western Australia,
Australia, Publication Series No. 7/95, The Australasian Institute of Mining and Metallurgy,
Melbourne, Victoria, Australia, November 13–14, 1995, pp. 245–252. With permission.)
Distances ­ranging up to 250 m without retro-reflectors can be surveyed, and
the system operates well in the dust of a typical mining environment.
The system provides a complete window of accessibility to an extracted
stope and can be used to determine failures in waste and fill, overbreak in
ore, muckpile shapes, as well as ore left unbroken inside an empty stope.
Back analysis of stope performance is critical to validate the rock mechanics
assumptions predicting wall behavior. Using a CMS system, the nature and
extent of any failure geometries can be established in order to ensure safety
and to provide information to optimize the production of future sources.
466
Geotechnical Design for Sublevel Open Stoping
CMS
covers
entire
stope void
with
one set-up
point
Drawpoints
Fill barricade
FIGURE 9.22
CMS laser survey arrangement of an open stope. (Modified after Gilbertson, R.J., The applicability of the caving measurement system at the Olympic Dam Operations, Proceedings of
the Sixth Underground Operators Conference, T. Golosinski, ed., Kalgoorlie, Western Australia,
Australia, Publication Series No. 7/95, The Australasian Institute of Mining and Metallurgy,
Melbourne, Victoria, Australia, November 13–14, 1995, pp. 245–252. With permission.)
The output from the CMS system consists of three-dimensional coordinates that can be used to build wireframe meshes and to calculate volumes.
Appraisal of stope performance is undertaken by calculating total surveyed
volumes and total design volumes. The surveyed and planned stope outlines
can be compared in order to calculate dilution or ore losses. The validity
of conventional ore grade reconciliation factors using routine clerical information such as a comparison between design and hoist/haulage-reconciled
tonnes can also be tested using the CMS system. More realistic dilution factors can also be determined. Stope performance can be measured directly
(Table 9.2), not just back-calculated from estimated grades and mill-feed
reconciliation.
9.6 Dilution Control Plan
The objectives for dilution control must be based on the realities of a particular mining system and its economics. A dilution control action plan must
include definition and identification of the dilution sources, including a strategy for measurements and implementation of corrective actions. Realistic
targets for dilution reduction over both the short and long term must be set.
The success of the program will rely on regular communication to all mining
personnel of the planned targets and their economic importance.
75,025
76,922
154,114
70,618
66,877
62,636
47,598
26,793
80,923
60,514
56,934
58,176
116,584
68,694
166,698
91,785
55,320
21,327
1,357,538
Plan
Tonnes
67,786
77,740
153,119
65,693
71,223
68,525
54,147
25,324
77,008
103,765
61,088
65,944
146,203
60,532
201,454
115,192
54,713
26,745
1,496,201
Mined
Tonnes
873
1,715
6,982
4,926
4,105
6,717
3,888
505
1,483
41,005
6,969
11,390
17,988
1,632
2,633
6,326
755
1,380
121,272 (8.93)
HW Waste
1,521
326
6,475
294
1,837
2,689
29,955 (2.21)
1,059
2,952
1,145
267
2,646
1,791
2,748
568
440
1,790
1,407
FW Waste
5,417
985
582
33,084 (2.44)
570
654
4,629
1,441
12,991
798
1,753
944
2,320
Backfill
Mined
1,041
74,119 (5.46)
2,408
45
1,569
4,036
1,433
322
473
2,746
1,991
1,132
16
8,905
2,753
27,157
18,092
Overbreak
2.58
7.29
6.78
7.35
12.25
34.32
15.62
10.55
2.38
70.72
15.71
20.70
20.70
2.85
5.46
13.11
6.47
21.81
13.58
Dilution
(%)
87.78
90.64
94.11
83.52
90.56
93.53
97.47
100.00
100.00
94.44
91.12
92.62
97.09
95.86
99.11
92.63
91.94
95.50
94.01
Recovery
(%)
0.00
3.13
0.03
2.22
6.03
2.29
0.68
1.77
3.39
3.29
1.99
0.03
7.64
4.01
16.29
19.71
0.00
4.88
5.46
Overbreak
(%)
Source: Andersen, B. and Grebenc, B., Controlling dilution at the Golden Giant Mine, CIM Mine Operator’s Conference, Timmins, Ontario, Canada,
1995, Paper 4, 14pp.
450-Q5
450-Q6
450-Q7/8
450-Q9
450-Q10
450-Q11
440-H2
440-H4
453-2W
460 Q9/Q10
450-Q13
450-Q12
466-3/4W
466-0
466-D1
456-5W
450-8W
460-Q11
TOTAL (% of
Planned)
Stope Name
Example of a Stope Extraction Summary
TABLE 9.2
Dilution Control
467
468
Geotechnical Design for Sublevel Open Stoping
The initial step is to establish and briefly document existing procedures
and practices, identifying areas and issues that may require change. Potential
solutions and improvements are identified, with costings, timing, and priorities assigned. As a general guideline, the mining personnel should be divided
into three groups that may include a planning, an operational, and a general group. The range of topics to be studied by the planning group includes
orebody delineation, stope design, rock mechanics, equipment, economics, and
downstream mineral processing analysis. The operational group could consider survey markups, development mining, longhole drilling and blasting,
production mucking, ground support, and fill placement and performance.
Finally, the general group could consider the schedules, ore accounting, stockpile management, stope performance review, and contracts/bonus system.
The suggested steps for the implementation of the dilution control plan
include an initial dilution meeting between geology, planning, operations, concentrator, and senior management. Following that initial meeting, the group is
divided into smaller groups in order to identify the dilution-related issues and
come up with solutions. It is critical that the groups are integrated together
in the best way so that they do achieve their objectives. For example, it is not
essential for individuals to remain within their normal fields of expertise. The
groups may benefit when outsiders present a different viewpoint on the operations. In addition, the groups should be able to include outside expertise as
required. The groups should determine their own operating strategies and
should be expected to produce a monthly progress report on their activities.
9.6.1 Stope Performance Review
A stope performance review is undertaken as a technical audit of the stope
design process. The review is carried out during the stope extraction (after
each firing) to monitor the conditions at the exposed stope walls, including
backbreak, underbreak, and broken ore fragmentation. The purpose of the
review is to determine any variations from a planned stope design extraction
strategy. To achieve this, a series of stope surveys can be carried out after each
significant firing and also following the completion of all firings (Figure 9.23).
The performance review provides a mechanism to record observations
from operators and technical personnel in order to indicate problems and
successes during the stope extraction life. A database highlighting lessons
to be learnt and improvements to be made can be prepared for each stope.
Table 9.3 shows some (by no means exhaustive) of the typical problems and
possible solutions encountered in open stoping. In addition to those problems, stopes left open over long periods of time may be influenced by timedependent regional fault behavior. Stress redistribution, production blasting,
and backfill drainage from adjacent stopes are likely to influence stope stability over a period of time. Blast damage and the effects of water from fill
can be transmitted along common fault structures intersecting a number
of stopes. Instability may create difficult remote mucking conditions due to
469
Dilution Control
Strike length
44 m 37 m 33 m 26 m
0m
Retreat
en
k
Bro
ore
FIGURE 9.23
Longitudinal section view of a large-scale bench stope showing consecutive surveys indicating minimal backbreak and the angle of repose of the broken ore.
large material falling off into the stope. These delays (stope production tails)
actually extend the stope life, which in turn may contribute to more overbreak and more production mucking delays.
Production profiles are usually shown as histograms of daily mucked volume. The data presented in Figure 9.24 show how highly stressed stopes
in which large rock falls occur slow down productivity. Also, since dilution is defined as any material that is extracted beyond the boundaries of a
designed orebody outline, a comparison of mucked versus designed volume
can be used to estimate dilution as shown in Figure 9.25.
With the advent of the CMS stope survey technique, information about the
actual variations from a designed stope shape can be routinely obtained and
used analytically to calculate dilution and stope wall depth of failure and to
determine structural control by large-scale geological discontinuities at the
stope boundaries. Contours of depth of failure can be determined by filling
the CMS wireframes with blocks and using the stope orientation information
to orient the block model such that the Y direction of the blocks is perpendicular to the stope walls, the X direction is parallel to the strike or width, and the
Z direction is parallel to the dip or height of the stopes as shown in Figure 9.26.
The block model can then be interrogated using the orebody boundaries
and the CMS wireframes. The blocks inside the CMS wireframe, yet outside the orebody boundaries (depth of failure), need to be determined. Once
the thickness for each column of blocks in the Y direction is calculated,
470
Geotechnical Design for Sublevel Open Stoping
TABLE 9.3
Example of Potential Problems and Solutions in Open Stoping
Open Stope
Activity
Rock mass
characterization
Stope design
Potential Problem
Design may not be stable
Different domain for design within
stope boundaries
Insufficient information
Major discontinuities intersect
stope walls
Design by default
Tonnage and grade do not match the
design
Stope access is not in the appropriate
location
Orebody delineation does not match
the geological interpretation
Excessive development in waste
Operators not following the design
Drilling and
blasting
Production
mucking
Stope survey
Excessive hole deviation
Not following design
Not drilling to required depth
Poor workmanship due
to bonus driven
Explosive malfunctioning
Area of low or high powder factor
Stope wall falloff
Inability to establish failuretriggering mechanism
Orepass hang-up
Large fragmentation/falloff
Long tramming distances
Poor reporting practices
Poor drawpoint condition
Continuous falloff inside the stope
Ability to survey as stope is extracted
Limited access
Poor ventilation, laser beam cannot
shoot through
Falloff may damage equipment
Potential Solution
Back-analyze previously
extracted stopes
Geological/engineering
judgment
More geological mapping
Consider firing sequences and
cablebolt reinforcement
Better preparation job—use
databases of stope performance
Better geological interpretation
needed
Better planning
More definition drilling, consider
geophysical techniques
Optimize the block design
Spot check and quality control,
better communication with
production
Down hole surveys, better
operator skills, laser alignment
Efficient supervision
Efficient supervision
May not be a short-term solution
Review pattern
Use modeling blasting software
Less aggressive design?
Use information from seismic
system
Limit intake size (use screen)
Optimize drilling and blasting
Improve block design
More personnel training
Support and reinforcement
Exclusion periods
Communication with survey
department
Establish stope access doors
Improve ventilation
Wait until ground stabilizes
471
Stope virtually empty
Further fall-off
Fines from slot firing reached
after cleaning big rocks
Large rocks observed in drawpoints
Clean up large
rocks before
mass blast
Large rocks observed in North drawpoint
5,000
MR 3-5 17B-16B 16B deterioration obvious
MR 4&5 13-14L
TUC 17L(1-3)
10,000
MR 1and 2 17B-16B
15,000
MR 3 13-14L
Tonnes
20,000
TUC 17L(4-8)
25,000
All firings Main fall-off
completed experienced
Fall-off
started
MR 1 and 2 13-14L
30,000
No problems to open slot
Slot firings
Fine muck
Dilution Control
1995–1996
1994–1995
Time
FIGURE 9.24
Production profile from a highly stressed stope at the end of a stoping block.
7000
6000
Volume (m3)
5000
4000
3000
2000
Blasted
1000
0
6-Nov
Mucked
11-Nov
16-Nov
21-Nov
26-Nov
1-Dec
Date
FIGURE 9.25
Cumulative plot of time versus volume for fired and mucked volumes.
6-Dec
11-Dec
472
Geotechnical Design for Sublevel Open Stoping
FIGURE 9.26
A CMS wireframe filled with 0.25 m × 0.25 m × 0.25 m blocks.
the information can then be contoured using 0.5 m intervals as shown in
Figure 9.27. Information from stope wall depth of failure can be used to
assess stope performance and provide instability criteria to predict future
stope performance. Confirmation of stope design reliability can be made by
back analyzing quantitative performance assessment criteria, such as depth
of failure against hydraulic radius (Figure 9.28).
The economic impact of dilution can readily be linked to stope wall depth
of failure. Beyond a critical stope dimension, larger failure depths are likely
to be experienced. On the other hand, reductions in the critical spans may
require additional pillars, leading to ore loss. The balance between additional pillars and the detrimental effects of failures can be established only
using an economic model of dilution.
In order to ensure that the actual stope performance information is used
to the best advantage, and to improve future designs, the details of stope
design and its underlying assumptions can be documented in a stope atlas.
Here, the history of a stope’s performance is recorded from the initial firing
through to final stope completion. The information contained varies depending upon the stoping practices at a particular mine site. The issues outlined
in Table 9.4 may be addressed.
9.7 Scale-Independent Measures of Stope Performance
Stope performance assessment can be undertaken using a number of measures ranging from subjective qualitative terms to quantitative measures,
473
Dilution Control
0.5 m
1.0 m
Stope outline
0.5 m
4.5
m
m
m
m
4.0
3.5
3.0
2.5
m
2.0 m
0.5 m
1.0 m
aterial
1.5 m
1.5 m 1.0 m
Broken m
Transverse drawpoint system
FIGURE 9.27
Longitudinal view of hangingwall depth of failure contours showing structurally controlled
failure.
such as equivalent linear overbreak/slough (ELOS) (Clark and Pakalnis,
1997), or depth of failure (Villaescusa, 2004). However, such conventional
stope performance indicators fail to adequately capture geometrical factors
related to the underlying failure modes (Cepuritis, 2011b). For example, failed
arched shapes may be related to weak rock masses, while blocky failure surfaces may indicate control by specific geological structures.
Typically, stope wall performance is analyzed by means of volume, area
or depth of instability, or ore loss. Cepuritis (2011b) suggested that a better
characterization of the performance is achieved by considering the location,
orientation, size, and shape of the stope wall under/overbreak. Two failures
are of the same size and shape if, after rotation and translation, they match
perfectly. Cepuritis (2011b) suggested that scale independency is the required
characteristic for a suitable geometrical comparison that is unaffected by
474
Geotechnical Design for Sublevel Open Stoping
9
HW FW
Depth of failure (m)
8
Total
7
6
5
4
3
2
1
0
3
4
5
6
7
8
9
Hydraulic radius (m)
10
11
12
FIGURE 9.28
Stope depth of failure for increasing hydraulic radius.
changes in the scale of an object. Therefore, a comparing measure should be
represented by a nondimensional or unitless value.
9.7.1 Conventional Measures
Clark and Pakalnis (1997) used the stope surface area and compared it to the
volume of overbreak or underbreak as follows:
ELOS =
S
VOB
AS
(9.8)
ELLO =
S
VUB
AS
(9.9)
where
ELOS is defined as the equivalent linear overbreak/slough
ELLO is the equivalent linear ore loss
S
S
VOB
and VUB
are the volume of overbreak and underbreak, respectively
AS is the surface area of a particular stope wall
According to Clark and Pakalnis (1997), a perceived advantage of ELOS
and ELLO is that, contrary to dilution calculations (e.g., Equation 9.7),
these stope performance indicators are independent of stope width, allowing comparison between different mining operations and orebodies. The
rationale is that both measures can be plotted against hydraulic radius.
475
Dilution Control
TABLE 9.4
Suggested Stope Performance Assessment Summary
Stope Performance Review
Stope Name:
Material
Ore (t)
Grade (%)
Internal dilution (%)
External dilution (%)
Underbreak (%)
Fill dilution (%)
Designed
By:
Date:
Actual
Tonnes mucked
—
—
Geology:
The effects of major geological structures, rock types, and properties
Reasons for any difference between design and actual grade and tonnes
Development:
Problems and concerns regarding ground conditions
Performance of ground support
Drilling:
Whether any holes or ring section could not be drilled as planned, set-up, or deviation
problems. Reasons for variation from design
Blasting:
Any problems encountered with charging, firing, or design sequence
The results of the blast, for example, fragmentation, misfires, freezing of holes, induced
failures
Production mucking:
Ventilation problems or otherwise with chosen circuit. Drawpoint and orepass conditions.
Broken ore left in base of stope?
Backfill:
Condition of fill passes, filling times, and cement ratios used, any problems encountered
Rock mechanics:
Stope and adjacent development stability. Timing of failures, and features that contributed to
dilution, effects of blasting, structure, and stress
Exposure and stability of adjacent fill masses
Planning and design:
General comments on original versus actual extraction. Recommended changes to design
procedure. Financial analysis of stope extraction
Also, as the rock mass quality decreases, there should be a corresponding increase in the observed overbreak, in this case represented by the
ELOS parameter (Clark and Pakalnis, 1997). However, as pointed out by
Cepuritis (2011b), ELOS and ELLO are both dimensional parameters that
are functions of the geometry (shape or size) of both the instability/ore loss
and the stope surface.
476
Geotechnical Design for Sublevel Open Stoping
9.7.2 Circularity Measures
Cepuritis (2011b) utilized the polygonal lines defined by the intersection of
the instability/ore loss volume with a planned stope surface to define a twodimensional shape measure of stope wall performance as follows:
Circularity =
4pA
P2
(9.10)
where A and P are the total area and total perimeter, respectively, of the
closed polygonal line(s) of intersection. Complex and irregular/elongated
shapes, having a large number of sides, tend to have a low circularity (below
0.4). On the other hand, compact objects such as regular/polyhedral tend to
have increased values, with elliptical to circular shapes having a value of
circularity exceeding 0.7 (Cepuritis, 2011b).
Cepuritis (2011b) utilized a circularity measure to characterize the twodimensional shape of an instability/ore loss as it intersects a stope surface, as well as the shape of the stope wall under investigation. The ratio
between the circularity of an over/underbreak and the circularity of a
stope wall provides a measure of how similar these two shapes are, as
follows:
CROB =
COB
CS
(9.11)
where
COB is the circularity of overbreak
CS is the circularity of the stope surface
When the circularity ratio (CROB ) is near unity, the two-dimensional shapes of
both the wall instability (or ore loss) and the stope surface are very similar.
9.7.3 Extensivity Measures
Cepuritis (2011b) introduced a measure to assess how extensive the twodimensional area of instability/ore loss is with respect to the stope wall
under investigation. The extensivity is given by
Extensivity =
A OB
AS
(9.12)
where AOB is the area of overbreak. An extensivity value approaching
unity indicates that the instability covers the majority of a stope wall.
477
Dilution Control
1.0
Example
overbreak
shape
Circular
0.9
0.8
Polyhedral
Circularity
0.7
Stope surface
polygon
0.6
0.5
Irregular
0.4
0.3
Elongated
0.2
0.1
0.0
Highly irregular/discontinuous
Sparse
0.0
0.1
0.2
0.3
0.4
0.5
Extensive
0.6
0.7
0.8
0.9
Extensivity
FIGURE 9.29
Circularity versus extensivity for an example stope surface shape. (From Cepuritis, P.M., Int. J.
Rock Mech. Min. Sci., 48, 1188, 2011b. With permission.)
For ­similar-shaped and -sized stope surfaces, this can provide a relative
measure of the size of overbreak (Cepuritis, 2011b).
Figure 9.29 shows an example plot of circularity versus extensivity for a
variety of example instability shapes. It is noted that the total intersected
areas and perimeters were utilized to calculate the circularity measure.
9.7.4 Hemisphericity Measures
Cepuritis (2011b) considered the flat intersectional area of instability and
compared it to the volume of a hemisphere in order to describe the threedimensional shape of under/overbreak. A scale-independent measure of the
three-dimensional shape of instability/ore loss is given by
Hemisphericity =
(3V S / 2p)
(A / p)
3/2
where
VS is the intersected volume of instability/ore loss
A is the intersected area with a stope wall under consideration
(9.13)
478
Geotechnical Design for Sublevel Open Stoping
This will result in unity for a hemisphere. High values indicate instability
having an elongated semi-ellipsoid shape, with a major axis perpendicular to the base area. Values lower than unity indicate flatter, platy shapes.
Cepuritis (2008) suggested that the three-dimensional shape of instability/
ore loss is dependent, to some extent, on the two-dimensional intersectional
area with the stope wall. For example, as this area becomes more elongated
or irregular, the ability to generate deeper prismatic shapes decreases.
A comparison of instability between different stope walls must consider
the relative shapes and coverage of over/underbreak across the respective
stope surfaces. Instability that is deep and arcuate in shape and covers an
entire stope surface represents worse stope performance conditions than
those represented by instability that is thin and platy in shape and covers only a small portion of the stope wall. In this regard, hemisphericity
and extensivity of instability/ore loss for a stope wall can be evaluated relative to the volume of a hemisphere having a 100% extensivity, as follows
(Cepuritis, 2008):
3/2
ÊExtensivity ˆ
Relative volume = 2p ¥ Hemisphericity Á
˜
p
Ë
¯
(9.14)
The relative volume can be used to quantify and subsequently classify relative
stope performance, irrespective of scale. Cepuritis (2011b) proposed the stope
performance classification, based on relative volume, shown in Table 9.5.
9.7.5 Cannington Mine Example
The scale-independent measures defined earlier have been applied to the
back analysis of stope performance from 76 stope surfaces at the Cannington
TABLE 9.5
Stope Performance Classification
Based on Relative Volume
Relative
Volume
Stope Performance
Classification
<0.02
0.02–0.05
0.05–0.1
0.1–0.2
0.2–0.5
>0.5
Very good
Good
Fair
Poor
Very poor
Exceptionally poor
Source: Cepuritis, P.M., Int. J. Rock
Mech. Min. Sci., 48, 1188, 2011b.
With permission.
479
Dilution Control
Relative volume
1.0
0.0–0.02
0.02–0.05
0.05–0.1
0.1–0.2
0.2–0.5
D
E
0.8
0.8
Hemisphericity
B
Hemisphericity
1.0
0.6
0.4
C
1.0
0.8
0.6
0.4
Circularity
(a)
A
B
0.2
D
0.4
F
0.0
0.2
0.4
0.6
1.0
0.8
0.0
vity
nsi
e
Ext
(b)
C
D
E
C
0.2
0.2
A
B
0.6
F
A
0.0
0.2
0.4
0.6
0.8
1.0
Extensivity
E
F
CMS
design
(c)
FIGURE 9.30
Stope instability data, Cannington Mine, Queensland, Australia. (a) Hemisphericity, circularity, and extensivity; (b) hemisphericity versus extensivity; and (c) rescaled example stope
surfaces shown in elevation and cross section with CMS and design profiles. (From Cepuritis,
P.M., Int. J. Rock Mech. Min. Sci., 48, 1188, 2011b. With permission.)
Mine, Queensland, Australia (Coles, 2007). The data were collected from a
number of stoping blocks with differing rock mass conditions and cableboltreinforcing patterns. Figure 9.30 shows the results of various shape measures
from the back-analyzed CMS geometries and stope design surfaces at the
Cannington Mine. A selected number of examples labeled A–F are represented graphically in Figure 9.30c.
A summary of the shape measures for the example stope surfaces, together
with a brief description, is shown in Table 9.6. The results show that the classifications based on the proposed shape measures are in good agreement
with the surveyed instability geometries (Cepuritis, 2011b).
480
Geotechnical Design for Sublevel Open Stoping
TABLE 9.6
Summary of Overbreak Shape Measures and Performance Classification (e.g., Stope
Surfaces Shown in Figure 9.30)
Extensivity
Circularity
Hemisphericity
Relative
Volume
Shape and Performance
Classification
A
0.06
0.66
0.09
0.001
B
0.51
0.56
0.58
0.239
C
0.44
0.22
0.20
0.068
D
0.30
0.26
0.47
0.087
E
0.61
0.58
0.21
0.116
F
0.18
0.09
0.05
0.005
Sparse, polyhedral, platy
to shallow—very good
performance
Moderately extensive,
irregular to polyhedral,
very deep—very poor
performance
Sparse, highly irregular/
discontinuous,
moderately deep—fair
performance
Sparse to moderately
extensive, elongated/
irregular, very
deep—fair performance
Moderately extensive,
irregular, moderately
deep—poor
performance
Sparse, highly irregular/
discontinuous,
shallow—very good
performance
Sample
Source: Cepuritis, P.M., Int. J. Rock Mech. Min. Sci., 48(7), 1188, 2011b. With permission.
References
Abel, J. F. 1983. MN 321 rock mechanics handouts: Unpublished lecture notes. Mining
Engineering Department, Colorado School of Mines, Golden, CO.
Alexander, E. G. and Fabjanczyk, M. W. 1981. Extraction design using open stopes
for pillar recovery in the 1100 ore body at Mount Isa. In D. R. Stewart (ed.),
Design and Operation of Caving and Sublevel Stoping Mines, pp. 437–458. New
York: SME-AIME.
AMIRA. 1993. P93E advanced blasting technology, Julius Kruttschnitt Mineral
Research Centre Final Report (1990–1993). The University of Queensland,
Brisbane, Queensland, Australia: Julius Kruttschnitt Mineral Research Centre.
AMIRA, 1995. Sulphide dust explosions. Vol 1—Cause and prevention in development headings, Project P316–P316A (1990–1994), 206p.
Andersen, B. and Grebenc, B. 1995. Controlling dilution at the Golden Giant Mine. CIM
Mine Operator’s Conference, Timmins, Ontario, Canada, January, Paper 4, 14pp.
Andrieux, P., McKenzie, C., Heilig, J., and Drolet, A. 1994. The impact of blasting
on excavation design—A geomechanics approach. In Proceedings of the 10th
Symposium on Explosives and Blasting Research, ISEE, Austin, TX, pp. 107–119.
Archibald, J. F., De Souza, E. M., and Beauchamp, L. 2011. Safe Canadian practices in
mine backfill operations. In H. J. Ilgner (ed.), Proceedings of the 10th International
Symposium on Mining with Backfill, Cape Town, South Africa, 21–25 March,
pp. 249–256. Johannesburg, South Africa: SAIMM.
Archibald, J. F. and DeGagne, D. O. 2001. Spray-on lining support in Canadian underground mining—A research summary. Journal of the Canadian Institute of Mining,
Metallurgy and Petroleum (CIM Bulletin), 94(1050): 49–56.
Ashcroft, J. W. 1991. Dilution: A total quality improvement opportunity. In 93rd Annual
General Meeting, CIM, Vancouver, British Columbia, Canada, 28 April – 2 May,
47pp.
Baczynski, N. R. P. 1974. Structure and hanging wall stability in the lead orebodies at
the Mount Isa Mine. MSc thesis, James Cook University of North Queensland,
Townsville, Queensland, Australia, 227pp.
Baczynski, N. R. P. 1980. Rock mass characterization and its application to assessment
of unsupported underground openings. PhD thesis, University of Melbourne,
Melbourne, Victoria, Australia, 233pp.
Baecher, G. B. and Lanney, N. A. 1978. Trace length biases in joint surveys. In Proceedings
of the 19th Symposium on Rock Mechanics, Stateline, NV, 1–3 May, pp. 56–65. Reno,
NV: University of Nevada.
Baldwin, N. 1998. The implementation of geotechnical mapping at Bronzewing
Gold Mine, WA. Bachelor of Engineering (Mining Geology) thesis, Western
Australian School of Mines, Curtin University of Technology, Perth, Western
Australia, Australia, 96pp.
Bampfield, H. A. and Morrey, W. B. 1984. Emulsion explosives. Presented to CIM Third
Coal Operators’ Conference, Fernie, British Columbia, Canada, 26–28 September,
Paper No. 3.
481
482
References
Barrett, J. R. and Cowling, R. 1980. Investigations of cemented fill stability in 1100
orebody, Mount Isa Mines Limited, Queensland, Australia. Transactions, Institution
of Mining and Metallurgy, Section A: Mining Industry, 89(July): A118–A128.
Barrett, S. V. L. and McCreath, D. R. 1995. Shotcrete support design in blocky ground:
Towards a deterministic approach. Tunnelling and Underground Space Technology,
10(1):–79–89.
Bartley, D. A. and McClure, R. 2003. Further field applications of electronic detonator technology. Fragblast: International Journal for Fragmentation and Blasting, 7(1):
13–22.
Barton, N. 1988. Rock mass classification and tunnel reinforcement selection using
the Q-system. In L. Kirkaldie (ed.), Rock Classification Systems for Engineering
Purposes: ASTM Special Technical Publication 984, pp. 59–88. Philadelphia, PA:
ASTM International.
Barton, N. 2002. Some new Q-value correlations to assist in site characterization and
tunnel design. International Journal of Rock Mechanics and Mining Sciences, 39(2):
185–216.
Barton, N., By, T. L., Chryssanthakis, P., Tunbridge, L., Kristiansen, J., Løset, F., Bhasin,
R. K., Westerdahl, H., and Vik, G. 1992. Comparison of prediction and performance
for a 62 m span sports hall in jointed gneiss. In G. Barla (ed.), Fourth International
Rock Mechanics and Rock Engineering Conference, Torino, Italy, pp. 17.1–17.15.
Barton, N., Lein, R., and Lunde, J. 1974. Engineering classification of rock masses for
design of tunnel support. Rock Mechanics, 6(4): 189–236.
Bawden, W. F. 1993. The use of rock mechanics principles in Canadian hard rock
mine design. In J. A. Hudson, E. T. Brown, C. Fairhurst, and E. Hoek (eds.),
Comprehensive Rock Engineering, pp. 247–290. Oxford, U.K.: Pergamon.
Bawden, W. F., Lausch, P., and Dennison, S. 2000. Lessons in control of mine costs from
instrumented cable bolt support. In J. Girard, M. Leibman, C. Breeds, and T. Doe
(eds.), Proceedings of the Fourth North American Rock Mechanics Symposium, Seattle,
WA, 31 July–3 August, pp. 633–642. Rotterdam, the Netherlands: A.A. Balkema.
Bawden, W. F., Nantel, J., and Sprott, D. 1988. Practical rock engineering in the optimisation of stope dimensions—Application and cost effectiveness. CIM Bulletin,
82(926): 63–70.
Bawden, W. F., Sauriol, G., Milne, D., and Germain, P. 1989. Practical rock engineering
stope design case histories from Noranda Minerals Inc. CIM Bulletin, 82(927):
37–45.
Beck, D., Brady, B. H. G., and Grant, D. R. 1997. Induced stress and microseismicity
in the 3000 orebody, Mount Isa. Geotechnical and Geological Engineering, 15(3):
221–233.
Beck, D. A. and Duplancic, P. 2005. Forecasting performance and achieving performance indicators in high stress and seismically active mining environments.
In Y. Potvin and M. Hudyma (eds.), Controlling Seismic Risk, Proceedings of
the Sixth International Symposium on Rockburst and Seismicity in Mines, Perth,
Western Australia, Australia, 9–11 March, pp. 409–417. Perth, Western Australia,
Australia: Australian Centre for Geomechanics.
Beck, D. A., Kassbohm, S., and Putzar, G. 2010. Multi-scale simulation of ground
support designs for extreme tunnel closure. In Y. Potvin (ed.), Caving 2010,
Proceedings of the Second International Symposium on Block and Sublevel Caving,
Perth, Western Australia, Australia, 20–22 April, pp. 441–453. Perth, Western
Australia, Australia: ACG.
References
483
Beck, D. A. and Sandy, M. P. 2003. Mine sequencing for high recovery in Western
Australian mines. In Proceedings of the 12th International Symposium on Mine
Planning and Equipment Selection, Kalgoorlie, Western Australia, Australia, 23–25
April, pp. 137–144. Melbourne, Victoria, Australia: The AusIMM.
Beer, G. and Meek, J. L. 1982. Design curves for roofs and hanging-walls in bedded
rock based on “voussoir” beam and plate solutions. Transactions, Institution of
Mining and Metallurgy, Section A: Mining Industry, 91(Jan): A18–A22.
Benson, R. P., Conlon, J., Merritt, A. H., Joli-Coeur, P., and Deere, D. U. 1971. Rock
mechanics at Churchill Falls. In K. S. Lane (ed.), Proceedings of the Symposium
on Underground Rock Chambers, Phoenix, AZ, 13–14 January, pp. 407–486.
New York: ASCE.
Benzaazoua, M., Fall, M., and Belem, T. 2004. A contribution to the understanding
the hardening process of cemented pastefill. Mining Engineering, 17(2): 141–152.
Bieniawski, Z. T. 1973. Engineering classifications of jointed rock masses. Transactions,
Institution of Civil Engineers of South Africa, 15(12): 335–344.
Bieniawski, Z. T. 1976. Rock mass classifications in rock engineering. In
Z. T. Bieniawski (ed.), Proceedings of the Symposium on Exploration for Rock
Engineering, Johannesburg, South Africa, 1–5 November, pp. 97–106. Cape
Town, South Africa: A.A. Balkema.
Bieniawski, Z. T. 1978. Determining rock mass deformability: Experience from case histories. International Journal of Rock Mechanics and Mining Sciences, 15(5): 237–247.
Bieniawski, Z. T. 1989. Engineering Rock Mass Classification, 251pp. New York: John
Wiley.
Bloss, M. L. 1992. Prediction of cemented rock fill stability—Design procedures
and modelling techniques. PhD thesis, University of Queensland, Brisbane,
Queensland, Australia.
Bloss, M. L. 1996. Evolution of cemented rock fill at Mount Isa Mines Limited. Mineral
Resources Engineering, 5(1): 23–43.
Bloss, M. L. and Chen, J. 1998. Drainage research at Mount Isa Mines Limited
1992–1997. In M. Bloss (ed.), Minefill’98, Proceedings of the Sixth International
Symposium on Mining with Backfill, Brisbane, Queensland, Australia, 14–16 April,
pp. 111–116. Melbourne, Victoria, Australia: The AusIMM.
Bloss, M. L. and Greenwood, A. G. 1998. Cemented rock fill research at Mount Isa
Mines Limited 1992–1997. In M. L. Bloss (ed.), Minefill’98, Proceedings of the Sixth
International Symposium on Mining with Backfill, Brisbane, Queensland, Australia,
14–16 April, pp. 207–215. Melbourne, Victoria, Australia: The AusIMM.
Bloss, M. L. and Morland, R. 1995. Influence of backfill stability on stope production in
the Copper Mine at Mount Isa Mines. In Proceedings of the AusIMM Underground
Operators’ Conference, Kalgoorlie, Western Australia, Australia, 13–14 November,
pp. 237–241. Melbourne, Victoria, Australia: The AusIMM.
Board, M. P. and Fairhurst, C. 1983. Rockburst control through destressing—
A case example. In Rockbursts: Prediction and Control, London, U.K., 20 October,
pp. 91–101. London, U.K.: IMM.
Bogue, R. H. 1955. Chemistry of Portland Cement. New York: Reinhold.
Brady, B. H. G. 1977. An analysis of rock behaviour in an experimental stoping block
at the Mount Isa Mine, Queensland, Australia. International Journal of Rock
Mechanics and Mining Sciences, 14(2): 59–66.
Brady, B. H. G. and Brown, E. T. 2004. Rock Mechanics for Underground Mining, 3rd
edn., 628pp. Dordrecht, the Netherlands: Kluwer.
484
References
Bridges, M. C. 1983. Review of open stope mining in Australia. In C. O. Brawner (ed.),
Stability in Underground Mining, pp. 813–828. New York: AIME.
Broch, E. and Franklin, J. A. 1972. The point-load strength test. International Journal of
Rock Mechanics and Mining Sciences, 9(6): 669–697.
Bronkhorst, D., Rheault, J., and Ley, G. 1993. Geotechnical principles governing mine
design at the Williams Mine. In W. F. Bawden and J. F. Archibald (eds.), Innovative
Mine Design for the 21st Century, Proceedings of the International Congress on Mine
Design, Kingston, Ontario, Canada, 23–26 August, pp. 433–442. Rotterdam, the
Netherlands: A.A. Balkema.
Brown, E. T. (ed.) 1981. Rock Characterization, Testing and Monitoring—ISRM Suggested
Methods, 211pp. Oxford, U.K.: Pergamon.
Brown, E. T. 2007. Block Caving Geomechanics, 2nd edn., 696pp. Brisbane, Queensland,
Australia: JKMRC.
Brzovic, A. 2010. Characterization of primary copper ore for caving at the El Teniente
Mine, Chile. PhD thesis, Curtin University of Technology, Perth, Western
Australia, Australia, 273pp.
Bywater, S., Cowling, R., and Black, B. N. 1983. Stress measurements and analysis for mine planning. In Proceedings of the Fifth Congress, International Society
for Rock Mechanics, Melbourne, Victoria, Australia, 11–15 April, pp. D29–D37.
Rotterdam, the Netherlands: A.A. Balkema.
Bywater, S. and Fuller, P. G. 1983. Cable support of lead open stope hangingwalls
at Mount Isa Mines Limited. In O. Stephansson (ed.), Rock Bolting: Theory and
Application in Mining and Underground Construction, Proceedings of the International
Symposium on Rock Bolting, Abisko, Sweden, 28 August–2 September, pp. 539–
555. Rotterdam, the Netherlands: A.A. Balkema.
Calizaya, F. 2013. Personal communication.
Call, R. D. 1972. Analysis of geologic structure for open pit slope design. PhD thesis,
University of Arizona, Tucson, AZ.
Call, R. D., Savely, J. P., and Nicholas, D. E. 1976. Estimation of joint set characteristics from surface mapping data. In W. A. Hustrulid (ed.), Monograph on Rock
Mechanics, Applications in Mining, pp. 65–73. New York: AIME.
Cameron, A. and Paley, N. 1992. Assessment of blasting to reduce damage in B704
bench stope at Mount Isa Mines. In T. Szwedzicki, G. R. Baird, and T. N. Little
(eds.), Proceedings of the Western Australian Conference on Mining Geomechanics,
Kalgoorlie, Western Australia, Australia, 8–10 June, pp. 375–383. Kalgoorlie,
Western Australia, Australia: Western Australian School of Mines.
Cepuritis, P. M. 1987. Hampton Boulder haulage shaft geotechnical study. MEngSc
thesis, Western Australian School of Mines, Curtin University of Technology,
Perth, Western Australia, Australia, 74pp.
Cepuritis, P. M. 2004. Three-dimensional rock mass characterisation for the design of
excavations and estimation of ground support requirements. In E. Villaescusa and
Y. Potvin (eds.), Ground Support in Mining and Underground Construction, Proceedings
of the Fifth International Symposium on Ground Support, Perth, Western Australia,
Australia, 28–30 September, pp. 115–127. Leiden, the Netherlands: A.A. Balkema.
Cepuritis, P. M. 2008. Techniques to assist in back analysis and assess open stope performance. In H. Schunnesson and E. Nordlund (eds.), Proceedings of the MassMin
2008, Luleå, Sweden, 9–11 June, pp. 203–212. Luleå, Sweden: Luleå University
of Technology.
References
485
Cepuritis, P. M. 2011a. An integrated approach to span design in open stope mining. PhD thesis, Western Australian School of Mines, Curtin University of
Technology, Perth, Western Australia, Australia, 473pp.
Cepuritis, P. M. 2011b. The use of scale independent shape measures to assess open
stope performance. International Journal of Rock Mechanics and Mining Sciences,
48(7): 1188–1195.
Cepuritis, P. M., Villaescusa, E., Beck, D. A., and Varden, R. 2010. Back analysis of
over-break in a longhole open stope operation using non-linear elasto-plastic
numerical modelling. In Proceedings of the 44th US Symposium of Rock Mechanics
and Fifth Canada—US Rock Mechanics Symposium, Salt Lake City, UT, 27–30 June,
Paper ARMA 10–124, 11pp (on CD-ROM).
Cepuritis, P. M., Villaescusa, E., and Lachenicht, R. 2007. Back analysis and performance of Block A long hole open stopes—Kanowna Belle Gold Mine. In
E. Eberhardt, D. Stead, and T. Morrison (eds.), Rock Mechanics: Meeting Society’s
Challenges and Demands, Proceedings of the First Canada—US Rock Mechanics
Symposium, Vancouver, British Columbia, Canada, 27–31 May, pp. 1431–1439.
Leiden, the Netherlands: Taylor & Francis.
Chacon, E., Barrera, V., Jeffrey, R., and van As, A. 2004. Hydraulic fracturing used to
precondition ore and reduce fragment size for block caving. In: Proceedings of
the. MassMin 2004, Santiago, Chile, 22–25 August, pp. 529–534. Santiago, Chile:
Instituto de Ingenieros de Chile.
Chileshe, B. C. and Kulkarni, R. V. 1995. Planning and engineering of the Konkola
Deep Mining Project, Konkola Mine, Zambia. The Institution of Mining and
Metallurgy, London, U.K., pp. 115–136.
Choquet, P. and Hadjigeorgiou, J. 1993. The design of support for underground
excavations. In J. A. Hudson, E. T. Brown, C. Fairhurst, and E. Hoek (eds.),
Comprehensive Rock Engineering, vol. 4, pp. 313–348. Oxford, U.K.: Pergamon.
Clark, L. M. and Pakalnis, R. C. 1997. An empirical design approach for estimating
unplanned dilution from open stope hangingwalls and footwalls. Presented at
99th CIM Annual General Meeting, New Frontiers for the Next Century, Vancouver,
British Columbia, Canada, 27 April–1 May.
Coates, D. F. and Cochrane, T. S. 1970. Development of design specifications for
rock bolting from research in Canadian Mines, Research Report R224, Mining
Research Centre, Energy, Mines & Resources Canada, 30pp.
Coates, R., Brown, S., Bucher, R., and Roth, A. 2009. Fully mechanized installation
of high tensile chain link mesh for surface support in tunnels. In P. Dight (ed.),
Proceedings of the First International Conference Safe and Rapid Development, Perth,
Western Australia, Australia, 6–7 May, pp. 165–172. Perth, Western Australia,
Australia: Australian Centre for Geomechanics.
Coles, D. 2007. Performance of open stopes at BHP-Billiton Cannington Mine, BEng
thesis, Western Australian School on Mines, Curtin University of Technology,
Kalgoorlie, Western Australia, Australia, 161pp.
Conway, J. P., Dar, S. M., Stears, J. H., and Williams, P. C. 1975. Laboratory studies of
yielding rock bolts. Washington, DC: US Department of the Interior Bureau of
Mines Report of Investigations 8058.
Cording, E. J., Hendron, A. J., and Deere, D. U. 1971. Rock engineering for underground caverns. In Proceedings of the Symposium on Underground Rock Chambers,
Phoenix, AZ, 13–14 January, pp. 567–600. New York: ASCE.
486
References
Cowling, R. 1998. Twenty-five years of mine filling: Developments and directions.
In M. L. Bloss (ed.), Minefill’98, Proceedings of the Sixth International Symposium
on Mining with Backfill, Brisbane, Queensland, Australia, 14–16 April, pp. 3–10.
Melbourne, Victoria, Australia: The AusIMM.
Cowling, R., Voboril, A., Isaacs, L. T., Meek, J. L., and Beer, G. 1989. Computer models
for improved fill performance. In F. P. Hassani, M. J. Scoble, and T. R. Yu (eds.),
Innovations in Mining Backfill Technology, Proceedings of the Fourth International
Symposium on Mining with Backfill, Montreal, Quebec, Canada, 2–5 October,
pp. 165–170. Rotterdam, the Netherlands: A.A. Balkema.
Davey Bickford, 2008. Daveytronic® III. Technical data sheet, Davey Bickford North
America. Available at www.daveybickford.com (accessed December 2012).
De Vries, R. 2013. Sublevel shrinkage at the Mount Wright underground gold mine.
MEngSc thesis, Western Australian School of Mines, Curtin University of
Technology, Perth, Western Australia, Australia, 89pp.
Deere, D. U. 1964. Technical description of rock cores. Rock Mechanics and Engineering
Geology, 1(1): 17–22.
Deere, D. U., Hendron, A. J., Patton, F. D., and Cording, E. J. 1967. Design of surface
and near-surface construction in rock. In C. Fairhurst (ed.), Failure and Breakage
of Rock, Proceedings of the Eighth Rock Mechanics Symposium, Minneapolis, MN,
15–17 September 1966, pp. 237–302. New York: AIME.
Dempers, G. D., Seymour, C. R. W., and Harris, M. B. 2010. Optimising geotechnical
logging to accurately represent the geotechnical environment. In P. Hagan and
S. Saydam (eds.), Proceedings of the Second Australasian Ground Control in Mining
Conference, Sydney, New South Wales, Australia, 23–24 November, pp. 85–94.
Melbourne, Victoria, Australia: The AusIMM.
Diederichs, M. S. and Kaiser, P. K. 1999. Stability of large excavations in laminated
hard rock masses: The voussoir analogue revisited. International Journal of Rock
Mechanics and Mining Sciences, 36(1): 97–117.
Du Pont. 1977. Blaster’s Handbook, 16th edn. Wilmington, DE: E I du Pont de Nemours
& Co.
Dyno Nobel. 2007. NONEL® LP series. Technical data sheet, Dyno Nobel Asia Pacific.
Available at www.dynonobel.com (accessed December 2012).
Dyno Nobel. 2011. Digishot™ plus. Technical information sheet, Dyno Nobel Inc.
Available at www.dynonobel.com (accessed December 2012).
Earl, J. 2003. Optimisation of pastefill recipe design at Junction Gold Mine, St Ives
Gold Mines, Kambalda. Mining Engineering Undergraduate thesis, Curtin
University of Technology, Kalgoorlie, Western Australia, Australia.
Elbrond, J. 1994. Economic effects of ore loss and rock dilution. CIM Bulletin, 87(978):
131–134.
Evans, W. H. 1941. The strength of undermined strata. Transactions, Institution of
Mining and Metallurgy, 50: 475–532.
Farmer, I. W. and Shelton, P. D. 1980. Review of underground rock reinforcement systems. Transactions Institution of Mining and Metallurgy, Section A: Mining Industry,
89: A68–A83.
Farsangi, P. and Hara, A. 1993. Consolidated rockfill design and quality control at
Kidd Creek Mines. CIM Bulletin, 86(972): 68–74.
Faulkner, J. 2005. Mine fill in open stoping operations. BEng thesis, Western Australian
School of Mines, Curtin University of Technology, Kalgoorlie, Western Australia,
Australia, 77pp.
References
487
Fernandez-Delgado, G., Mahar, J., and Parker, H. 1976. Structural behaviour of thin
shotcrete liners obtained from large scale tests. Shotcrete for ground support.
In Proceedings of the Engineering Foundation Conference, ACI, Asilomar, CA, 4–8
October, pp. 399–442.
Fleetwood, K. 2010. Near-field blast vibration monitoring and analysis for prediction of blast damage in sub-level open stoping. PhD thesis, Western Australian
School of Mines, Curtin University of Technology, Perth, Western Australia,
Australia, 350pp.
Foster, J., Jessop, S., and Andrews, P. 2011. Case study: Underhand cut and fill stoping
using cemented tailings pastefill at the Lanfranchi Nickel Mine. In H. J. Ilgner
(ed.), MineFill 2011, Proceedings of the 10th International Symposium on Mining
with Backfill, Cape Town, South Africa, 21–25 March, pp. 363–372. Johannesburg,
South Africa: SAIMM.
Gaudreau, D., Aubertin, M., and Simon, R. 2004. Performance assessment of tendon
support systems submitted to dynamic loading. In E. Villaescusa and Y. Potvin
(eds.), Ground Support in Mining and Underground Construction, Proceedings of
the Fifth International Symposium on Ground Support, Perth, Western Australia,
Australia, 28–30 September, pp. 299–312. Leiden, the Netherlands: A.A. Balkema.
Gilbertson, R. J. 1995. The applicability of the caving measurement system at the
Olympic Dam Operations. In T. Golosinski (ed.), Proceedings of the Sixth
Underground Operators Conference, Kalgoorlie, Western Australia, Australia,
13–14 November, pp. 245–252. Melbourne, Victoria, Australia: The Australasian
Institute of Mining and Metallurgy, Publication Series No 7/95.
Goddard, I. A. 1981. The development of open stoping in lead orebodies at Mount
Isa Mines Ltd. In D. R. Stewart (ed.), Design and Operation of Caving and Sublevel
Stoping Mines, pp. 509–528. New York: SME-AIME.
Golosinski, T. S., Wang, C., and Ganeswarran, R. 1997. Investigation on the composition and properties of cemented rockfill. Mineral Resources Engineering, 6(4):
221–227.
Grant, D. and De Kruijff, S. 2000. Mount Isa Mines—1100 Orebody, 35 years on. In
G. Chitombo (ed.), Proceedings of the MassMin 2000, Brisbane, Queensland,
Australia, 29 October–2 November, pp. 591–600. Melbourne, Victoria, Australia:
The AusIMM.
Greenelsh, R. W. 1985. The N663 stope experiment at Mount Isa Mine. International
Journal of Mining Engineering, 3(3): 183–194.
Grice, A. G. 1989. Fill research at Mount Isa Mines Limited. In F. P. Hassani, M. J. Scoble,
and T. R. Yu (eds.), Innovations in Mining Backfill Technology, Proceedings of the
Fourth International Symposium on Mining with Backfill, Montreal, Quebec,
Canada, 2–5 October, pp. 15–22. Rotterdam, the Netherlands: A.A. Balkema.
Grice, A. G. 1998. Stability of hydraulic fill barricades. In M. Bloss (ed.), Minefill’98,
Proceedings of the Sixth International Symposium on Mining with Backfill, Brisbane,
Queensland, Australia, 14–16 April, pp. 117–120. Melbourne, Victoria, Australia:
The AusIMM.
Grice, A. G. 2005a. Fluid mechanics of mine fill. In Y. Potvin, E. Thomas, and A. Fourie
(eds.), Handbook on Mine Fill, pp. 51–63. Perth, Western Australia, Australia:
ACG.
Grice, A. G. 2005b. Introduction to hydraulic fill. In Y. Potvin, E. Thomas, and
A. Fourie (eds.), Handbook on Mine Fill, pp. 67–79. Perth, Western Australia,
Australia: ACG.
488
References
Grice, A. G. 1999. Mine backfill—Course notes for the masters of engineering science
in mining geomechanics at the Western Australian School of Mines. Curtin
University of Technology, Perth, Western Australia, Australia.
Grimstad, E. and Barton, N. 1993. Updating the Q-System for NMT. In R. Kompen,
O. A. Opsahl, and K. R. Berg (eds.), Proceedings of the International Symposium
on Sprayed Concrete—Modern Use of Wet Mix Sprayed Concrete for Underground
Support, Fagernes, Norway, 17–21 October, pp. 46–66. Oslo, Norway: Norwegian
Concrete Association.
Grobler, H. P. 2003. Using electronic detonators to improve all-round blasting performances. Fragblast: International Journal for Fragmentation and Blasting, 7(1): 1–12.
Guilfoyle, K. S. and Bradford, N. L. 1982. Production blasting at Isa Mine. In Proceedings
of the Underground Operators’ Conference, West Coast Tasmania Branch, Hobart,
Australia, pp. 105–116. Melbourne, Victoria, Australia: The AusIMM.
Hamrin, H. 1982. Choosing an underground mining method. In W. A. Hustrulid (ed.),
Underground Mining Methods Handbook, pp. 88–112. New York: SME-AIME.
Hamrin, H. 1993. Precision drilling extends the range of long hole blasting. In
G. Almgren, U. Kumar, and N. Vagenas (eds.), Proceedings of the Second
International Symposium on Mine Mechanization and Automation, Luleå, Sweden,
7–10 June, pp. 143–151. Rotterdam, the Netherlands: A.A. Balkema.
Hassell, R. and Villaescusa, E. 2005. Overcoring techniques to assess in situ corrosion of
galvanized friction bolts. In S. S. Peng et al. (eds.), Proceedings of the 24th International
Conference on Ground Control in Mining, Morgantown, WV, 2–4 August, pp. 349–356.
Hassell, R., Villaescusa, E., Ravikumar, A., and Cordova, M. 2010. Development of
a corrosivity classification for weld mesh support at the Cannington Mine. In
P. Hagan and S. Saydam (eds.), Proceedings of the Second Australasian Conference
on Ground Control in Mining, Sydney, New South Wales, Australia, 23–24
November, 5p. Melbourne, Victoria, Australia: The AusIMM.
Hassell, R. C. 2007. Corrosion of rock reinforcement in underground excavations. PhD
thesis, Western Australian School of Mines, Curtin University of Technology,
Perth, Western Australia, Australia, 277pp.
Hassell, R. C., Villaescusa, E., and Thompson, A. G. 2006. Testing and evaluation of
corrosion on cable bolt anchors. In D. P. Yale, S. C. Holtz, C. Breeds, and U. Ozbay
(eds.), Golden Rocks 2006, 50 Years of Rock Mechanics—Landmarks and Challenges,
Proceedings of the 41st US Symposium on Rock Mechanics, Golden, MO, 17–21 June,
Paper 06-996, 11pp. Alexandria, VA: American Rock Mechanics Association.
Hedley, D. G. F. and Grant, F. 1972. Stope and pillar design at the Elliot Lake uranium
mines. CIM Bulletin, 65(723): 37–44.
Heilig, J. 1999. Blast engineering—Course notes for the masters of engineering science in mining geomechanics. Kalgoorlie, Western Australia, Australia: Western
Australian School of Mines.
Helinski, M., Fahey, M., and Fourie, A. 2011a. A framework for designing and managing cemented mine backfill. In H. J. Ilgner (ed.), MineFill 2011, Proceedings of the
10th International Symposium on Mining with Backfill, Cape Town, South Africa,
21–25 March, pp. 261–274. Johannesburg, South Africa: SAIMM.
Helinski, M., Wines, D., Revell, M., and Sainsbury, D. 2011b. Critical factors influencing the capacity of arched fibrecrete bulkheads and waste rock barricades. In
H. J. Ilgner (ed.), MineFill 2011, Proceedings of the 10th International Symposium
on Mining with Backfill, Cape Town, South Africa, 21–25 March, pp. 293–303.
Johannesburg, South Africa: SAIMM.
References
489
Hills, P. B. and Gearing, W. G. 1993. Gold ore mining by the Porgera Joint Venture
at Porgera, Papua New Guinea. In J. T. Woodcock and J. K. Hamilton (eds.),
Proceedings of the Australasian Mining and Metallurgy, pp. 897–902. Melbourne,
Victoria, Australia: The AusIMM, Chapter 12: Gold.
Hoek, E. 1994. Strength of rock and rock masses. ISRM News Journal, 2(2): 4–16.
Hoek, E. 2007. Rock mass properties. In Practical Rock Engineering. Available online at
http://www.rocscience.com/hoek/corner/11_Rock_mass_properties.pdf.
Hoek, E. and Brown, E. T. 1980. Underground Excavations in Rock, 527pp. London, U.K.:
IMM.
Hoek, E. and Brown, E. T. 1997. Practical estimates of rock mass strength. International
Journal of Rock Mechanics and Mining Sciences, 34(8): 1165–1186.
Hoek, E., Carranza-Torres, C., and Corkum, B. 2002. Hoek–Brown failure criterion 2002 ed. In R. Hammah et al. (eds.), Mining and Tunnelling Innovation and
Opportunity, Proceedings of the Fifth North American Rock Mechanics Symposium
and 17th Tunnel Association of Canada Conference, Toronto, Ontario, Canada, 7–10
July, vol. 1, pp. 267–273. Toronto, Ontario, Canada: University of Toronto Press.
Hoek, E. and Diederichs, M. S. 2006. Empirical estimation of rock mass modulus.
International Journal of Rock Mechanics and Mining Sciences, 43(2): 203–215.
Hoek, E., Kaiser, P. K., and Bawden, W. F. 1995. Support of Underground Excavations in
Hard Rock, 215pp. Rotterdam, the Netherlands: A.A. Balkema.
Holmgren, J. 1976. Thin shotcrete layers subject to punch loads. Shotcrete for ground
support. In Proceedings of the Engineering Foundation Conference, Asilomar, CA,
4–8 October, pp. 443–459.
Hudson, J. A. and Harrison, J. P. 1997. Engineering Rock Mechanics: An Introduction to
the Principles, 444pp. Oxford, U.K.: Pergamon.
Hudson, J. A. and Priest, S. D. 1979. Discontinuities and rock mass geometry.
International Journal of Rock Mechanics and Mining Sciences, 16(6): 339–362.
Hudson, J. A. and Priest, S. D. 1983. Discontinuity frequency in rock masses.
International Journal of Rock Mechanics and Mining Sciences, 20(2): 73–89.
Hudyma, M. 1988. Development of empirical rib pillar failure criterion for open stope
mining. MASc thesis, Department of Mining and Mineral Processing, University
of British Columbia, Vancouver, British Columbia, Canada.
Hutchinson, D. J. and Diederichs, M. S. 1996. Cablebolting in Underground Mines,
406pp. Richmond, British Columbia, Canada: Bitech Publishers.
Hyett, A. J., Bawden, W. F., MacSporran, G. R., and Moosavi, M. 1995. A constitutive
law for bond failure of fully-grouted cable bolts using a modified Hoek cell.
International Journal of Rock Mechanics and Mining Sciences, 32(1): 11–36.
ISRM. 1985. Suggested method for determining point load strength. International Journal
of Rock Mechanics and Mining Sciences and Geomechanics Abstracts, 22(2): 51–60.
Jaeger, J. C. and Cook, N. G. W. 1976. Fundamentals of Rock Mechanics, 3rd edn., 593pp.
London, U.K.: Chapman & Hall.
Jager, A. J. 1992. Two new support units for the control of rockburst damage. In P. K. Kaiser and D. R. McCreath (eds.), Rock Support in Mining and
Underground Construction, Proceedings of the International Symposium on Rock
Support, Sudbury, Ontario, Canada, 16–19 June, pp. 621–631. Rotterdam, the
Netherlands: A.A. Balkema.
Jing, L. 2003. A review of techniques, advances and outstanding issues in numerical
modelling for rock mechanics and rock engineering. International Journal of Rock
Mechanics and Mining Sciences, 40(3): 283–353.
490
References
Jing, L. and Stephansson, O. 2007. Fundamentals of Discrete Element Methods for Rock
Engineering: Theory and Applications, 545pp. Amsterdam, the Netherlands:
Elsevier.
Kaesehagen, M. R. and Boffey, R. H. 1998. Development of the Osborne Mine—
With a focus on technical and operational aspects. In Proceedings of the Seventh
Underground Operators’ Conference, Townsville, Queensland, Australia, 30 June–3
July, pp. 29–37. Melbourne, Victoria, Australia: The AusIMM.
Kaiser, P. K. and Tannant, D. 2001. The role of shotcrete in hard rock mines. In
W. Hustrulid and R. Bullock (eds.), Underground Mining Methods: Engineering
Fundamentals and International Case Studies, pp. 579–592. Littleton, CO: Colorado
Society for Mining, Metallurgy and Exploration.
Kaiser, P. K., Tannant, D. D., and McCreath, D. R. 1996. Drift support in burst-prone
ground. CIM Bulletin, 89(990): 131–138.
Kleine, T., Cameron, A., Townson, P., and Paley, N. 1992. Facing up to deviant blastholes. In Proceedings of the Fifth Underground Operators Conference, 14–16 July,
pp. 61–66. Ballarat, Victoria, Australia: The AusIMM.
Krauland, N. and Soder, P. E. 1987. Determining pillar strength from pillar failure
observations. Engineering and Mining Journal, 8: 34–40.
Kropp, W. and Villaescusa, E. 1996. Development of mining practices in the Lead/
Zinc Mine, Mount Isa. In Proceedings of the Ninth Australian Tunnelling Conference,
Sydney, New South Wales, Australia, 27–29 August, pp. 461–466.
Kuganathan, K. 2002. A method to design efficient mine backfill drainage systems
to improve safety and stability of backfill bulkheads and fills. In Growing Our
Underground Operations, Proceedings of the Eighth AusIMM Underground Operators’
Conference, Townsville, Queensland, Australia, 29–31 July, pp. 181–190.
Melbourne, Victoria, Australia: The AusIMM.
Kuganathan, K. 2005. Geomechanics of mine fill. In Y. Potvin, E. Thomas, and
A. Fourie (eds.), Handbook on Mine Fill, pp. 23–47. Perth, Western Australia,
Australia: ACG.
Kuganathan, K. and Neindorf, L. B. 2005. Backfill technology development at
Mount Isa Mines between 1995 and 2005. In Proceedings of the Ninth AusIMM
Underground Operators Conference, Perth, Western Australia, Australia, 7–9
March, pp. 173–183. Melbourne, Victoria, Australia: The AusIMM.
Kulatilake, P. H. S. W. and Wu, T. H. 1984. The density of discontinuity traces in sampling windows. International Journal of Rock Mechanics and Mining Sciences, 21(6):
345–347.
Landmark, J. and Villaescusa, E. 1992. Geotechnical mapping at Mount Isa Mines.
In T. Szwedzicki, G. R. Baird, and T. N. Little (eds.), Proceedings of the Western
Australian Conference on Mining Geomechanics, Kalgoorlie, Western Australia,
Australia, 8–10 June, pp. 329–333. Kalgoorlie, Western Australia, Australia:
Western Australian School of Mines.
Landriault, D. 2001. Backfill in underground mining. In W. A. Hustrulid and
R. L. Bullock (eds.), Underground Mining Methods: Engineering Fundamentals and
International Case Studies, pp. 601–614. Littleton, CO: SME.
Landriault, D. and Goard, B. 1987. Research into high density backfill placement
methods by the Ontario Division of INCO Limited. CIM Bulletin, 80(897):
46–50.
Lang, T. A. 1961. Theory and practice of rockbolting. Transactions, Society of Mining
Engineers of AIME, 220: 333–348.
References
491
Langdon, C. and Duniam, P. 1995. Advances in theory and application of non-electric
initiation systems to 60 series extraction at the Mount Lyell Copper Mine. In
T. Golosinski (ed.), Proceedings of the Sixth Underground Operators Conference,
Kalgoorlie, Western Australia, Australia, 13–14 November, pp. 291–298.
Melbourne, Victoria, Australia: The AusIMM.
Lappalainen, P. 1997. Personal communication.
Lappalainen, P. and Pitkajarvi, J. 1996. Dilution control at Outokumpu mines. In
Proceedings of the Nickel ‘96, Mineral to Market, Kalgoorlie, Western Australia,
Australia, 27–29 November, pp. 25–29. Melbourne, Victoria, Australia: The
AusIMM.
Laslett, G. M. 1982. Censoring and edge effects in area and line transect sampling of
rock joint traces. Journal of Mathematical Geology, 14(2): 125–140.
Laubscher, D. H. 1984. Design aspects and effectiveness of support systems in different mining conditions. Transactions, Institution of Mining and Metallurgy, Section
A: Mining Industry, 93(April): A70–A81.
Laubscher, D. H. 1993. Planning mass mining operations. In J. A. Hudson, E. T.
Brown, C. Fairhurst, and E. Hoek (eds.), Comprehensive Rock Engineering, vol. 2,
pp. 547–583. Oxford, U.K.: Pergamon.
Laubscher, D. H. and Taylor, H. W. 1976. The importance of geomechanics classification
of jointed rock masses in mining operations. In Z. T. Bieniawski (ed.), Proceedings
of the Symposium on Exploration for Rock Engineering, Johannesburg, South Africa,
1–5 November, pp. 119–128. Cape Town, South Africa: A.A. Balkema.
Lauffer, H. 1958. Gebirgsklassifizierung fur den Stollenbau. Geologie und Bauwesen,
24(1): 46–51.
Lauffer, H. 1960. Die neure Entwicklung der Stollenbautechnik. Österreichische
Zeitschrift, 3: 13–24.
Li, C. C. 2010. A new energy-absorbing bolt for rock support in high stress rock masses.
International Journal of Rock Mechanics and Mining Sciences, 47(3): 396–404.
Li, J., Villaescusa, E., Tyler, D., and McGrath, T. 2004. Factors influencing the quality of
minefill for underground support. Proceedings of the 8th International Conference
on Minefill, pp. 248–252. Beijing, China: The Nonferrous Metals Society of China.
Liu, Q., Ellis, B., and Proudfoot, D. 2002. Application of electronic detonators for complex blasts at Brunswick Mine in Canada. In X. Wang (ed.), Proceedings of the
Seventh International Symposium on Rock Fragmentation by Blasting—FRAGBLAST
7, Beijing, China, 11–15 August, pp. 37–46. Beijing, China: Metallurgical Industry
Press.
Livingston, C. W. 1956. Fundamental concepts of rock failure. Quarterly Colorado
School of Mines, vol. 51., Golden, CO.
Logan, A. S., Villaescusa, E., Stampton, V. R., Struthers, M. A., and Bloss, M. L. 1993.
Geotechnical instrumentation and ground behaviour at Mount Isa. In T. Szwedzicki
(ed.), Geotechnical Instrumentation and Monitoring in Open Pit and Underground
Mining, Proceedings of the Australian Conference, Kalgoorlie, Western Australia,
Australia, 21–23 June, pp. 321–329. Rotterdam, the Netherlands: A.A. Balkema.
Lopez Jimeno, C., Lopez Jimeno, E., and Ayala Carceda, F. J. 1995. Drilling and Blasting
of Rocks, 410pp (trans. B. Y. Visser de Ramiro). Rotterdam, the Netherlands:
A.A. Balkema.
Lunder, P. 1994. Hard rock pillar strength estimation: An applied empirical approach.
MASc thesis, University of British Columbia, Vancouver, British Columbia,
Canada, 166pp.
492
References
Lunder, P. and Pakalnis, R. 1997. Determination of the strength of hard rock mine pillars. CIM Bulletin, 90(1013): 51–55.
Magee, D. L. 2005. Geometric back analysis of CMS stope surveys at Kanowna Belle
Gold Mine. BEng thesis, Western Australian School of Mines, Curtin University
of Technology, Kalgoorlie, Western Australia, Australia, 114pp.
Marek, J. M. 1975. Open pit slope design—Ten Mile Cirque, Climax, Colorado School
of Mines. Master thesis, Colorado School of Mines, Golden, CO, 51pp.
Marinos, P. G., Marinos, V., and Hoek, E. 2007. The geological strength index (GSI):
A characterization tool for assessing engineering properties of rock masses. In
C. Mark, R. Pakalnis, and R. J. Tuchman (eds.), Proceedings International Workshop
on Rock Mass Classification for Underground Mining, Vancouver, British Columbia,
Canada, 31 May, NIOSH IC 9498, pp. 87–94. Pittsburgh, PA: US Department of
Health & Human Services.
Martin, C. D., Read, R. S., and Martino, J. B. 1997. Observations of brittle failure
around a circular test tunnel. International Journal of Rock Mechanics and Mining
Sciences, 34(7): 1065–1073.
Mathews, K. E. 1978. Design of underground mining layouts. Underground Space, 2(4):
195–209.
Mathews, K. E., Hoek, E., Wyllie, D. C., and Stewart, S. B. V. 1980. Prediction of Stable
Excavation Spans for Mining at Depths below 1,000 Metres in Hard Rock. Ottawa,
Ontario, Canada: Golder Associates Report to Canada Centre for Mining and
Energy Technology (CANMET), Department of Energy and Resources.
Mathews, K. E. and Kaesehagen, F. 1973. The development and design of a cemented
rock filling system at the Mount Isa Mine Australia. In E. G. Thomas (ed.),
Proceedings of the Jubilee Symposium on Mining with Backfill, Mount Isa, Queensland,
Australia, 1973, pp. 13–23. Melbourne, Victoria, Australia: The AusIMM.
Mathis, J. I. 1988. Development and verification of a three dimensional rock joint
model. PhD thesis, Luleå University, Luleå, Sweden.
Matthews, S. M., Worotnicki, G., and Tillmann, V. H. 1983. A modified cable bolt
system for the support of underground openings. In Proceedings of the AusIMM
Annual Conference, Broken Hill, New South Wales, Australia, pp. 243–255.
Mawdesley, C., Trueman, R., and Whiten, W. 2001. Extending the Mathews stability
graph for open-stope design. Transactions, Institution of Mining and Metallurgy,
Section A: Mining Technology, 110: A27–A39.
McHugh, C. 2005. Introduction of autonomous loaders to Olympic Dam Operations,
Australia. In Proceedings of the Ninth Underground Operators Conference, Perth,
Western Australia, Australia, 7–9 March, pp. 127–132. Melbourne, Victoria,
Australia: The AusIMM.
Mikula, P. A. 2004. Changing to the Posimix4 resin bolt for jumbo and QuickChem™
at Mt Charlotte mine. In E. Villaescusa and Y. Potvin (eds.), Ground Support
in Mining and Underground Construction, Proceedings of the Fifth International
Symposium on Ground Support, Perth, Western Australia, Australia, 28–30
September, pp. 211–219. Leiden, the Netherlands: A.A. Balkema.
Mikula, P. A. and Lee, M. F. 2000. Bulk low-grade mining at Mount Charlotte Mine.
In G. Chitombo (ed.), Proceedings of the MassMin 2000, Brisbane, Queensland,
Australia, 29 October–2 November, pp. 623–635. Melbourne, Victoria, Australia:
The AusIMM.
Miller, F., Potvin, Y., and Jacob, D. 1992. Laser measurement of open stope dilution.
CIM Bulletin, 85(962): 96–102.
References
493
Milne, D. and Gendron, A. 1990. Borehole camera monitoring for safety and design.
Presented at 92nd CIM Annual General Meeting, Ottawa, Ontario, Canada, 6–10
May, 13pp.
Milne, D., Germain, P., Grant, D., and Noble, P. 1991. Systematic rock mass
characterization for underground mine design. In W. Wittke (ed.), Proceedings
of the Seventh International Congress on Rock Mechanics, Aachen, Germany,
16–20 September, vol. 1, pp. 293–298. Rotterdam, the Netherlands: A. A.
Balkema.
Mine Modelling, 2013. Web site: http://www.map3d.com/.
Mitchell, R. J. and Roettger, J. J. 1989. Analysis and modelling of sill pillars. In
F. P. Hassani, M. J. Scoble, and T. R. Yu (eds.), Innovations in Mining Backfill
Technology, Proceedings of the Fourth International Symposium on Mining with
Backfill, Montreal, Quebec, Canada, 2–5 October, pp. 53–62. Rotterdam, the
Netherlands: A.A. Balkema.
Mitchell, R. J., Smith, R., and Libby, D. 1975. Bulkhead pressure due to cemented
hydraulic backfills. Canadian Geotechnical Journal, 12(3): 362–371.
Morrison, D. M. 1996. Deep hardrock mining—The future. CIM Bulletin, 89(1000):
46–51.
Morton, E. C. 2009. Static testing of large scale ground support panels. MSc thesis,
Western Australian School of Mines, Curtin University of Technology, Perth,
Western Australia, Australia, 250pp.
Morton, E. C., Thompson, A. G., and Villaescusa, E. 2009a. The performance of
mesh, shotcrete and membranes for surface ground support. In M. Diederichs
and G. Grasselli (eds.), Rock Engineering in Difficult Conditions, Proceedings
of the Third Canada—US Rock Mechanics Symposium and 20th Canadian Rock
Mechanics Symposium, Toronto, Ontario, Canada, 11–13 May, Paper 4022, 13pp
(on CD-ROM).
Morton, E. C., Villaescusa, E., and Thompson, A. G. 2009b. Determination of energy
absorption capabilities of large scale shotcrete panels. In F. Amberg and
K. F. Garshol (eds.), Shotcrete for Underground Support XI, Proceedings of the 2009
ECI Conference on Shotcrete for Underground Support, Davos, Switzerland, 7–10
June, Paper 6, 20pp.
Mould, R. J., Campbell, R. N., and MacGregor, S. A. 2004. Extent and mechanisms
of gloving and unmixed resin in fully encapsulated roof bolts and a review
of recent developments. In E. Villaescusa and Y. Potvin (eds.), Ground Support
in Mining and Underground Construction, Proceedings of the Fifth International
Symposium on Ground Support, Perth, Western Australia, Australia, 28–30
September, pp. 231–242. Leiden, the Netherlands: A.A. Balkema.
Neindorf, L. B. and Karunatillake, G. S. B. 2000. George Fisher Mine—Feasibility and
construction. In G. Chitombo (ed.), Proceedings of the MassMin 2000, Brisbane,
Queensland, Australia, 29 October–2 November, pp. 601–609. Melbourne,
Victoria, Australia: The AusIMM.
Neugebauer, E. 2008. Ready for Roofex—A new way to tackle safety in underground
operations. Mining and underground construction 3-08, pp. 24–25.
Nguyen, Q. D. and Boger, D. V. 1985. Direct yield stress measurements with the vane
method. Journal of Rheology, 29(3): 335–347.
Nickson, S. D. 1992. Cablebolt support guidelines for underground hard rock
mine operations. MASc thesis (unpublished), University of British Columbia,
Vancouver, British Columbia, Canada.
494
References
Nienbauer, L. A. 2011. Stable paste-fill walls in a large open stope at Sunrise Dam
Gold Mine. In H. J. Ilgner (ed.), MineFill 2011, Proceedings of the 10th International
Symposium on Mining with Backfill, Cape Town, South Africa, 21–25 March,
pp. 275–283. Johannesburg, South Africa: SAIMM.
Oddie, M. E. and Pascoe, M. J. 2005. Stope performance at Olympic Dam Mine.
In Proceedings of the Ninth Underground Operators’ Conference, Perth, Western
Australia, Australia, 7–9 March, pp. 265–272. Melbourne, Victoria, Australia:
The AusIMM.
Oñederra, I. and Chitombo, G. 2007. A design methodology for underground
ring blasting. Mining Technology, Transactions of the Institutions of Mining and
Metallurgy: Section A, 116(4): 180–195.
Orica, 2008. Exel™ LP: Non-electric, long delay detonator assembly. Technical data
sheet. Orica Mining Services. Available at www.oricaminingservices.com
(accessed December 2012).
Orica, 2010. i-kon™ Electronic detonator. Technical data sheet. Orica Mining Services.
Available at www.oricaminingservices.com (accessed December 2012).
Ortlepp, W. D. 1992. Impulsive-load testing of tunnel support. In P. K. Kaiser and
D. R. McCreath (eds.), Rock Support in Mining and Underground Construction,
Proceedings of the International Symposium on Rock Support, Sudbury, Ontario,
Canada, 16–19 June, pp. 675–682. Rotterdam, the Netherlands: A.A. Balkema.
Ortlepp, W. D. and Read, J. J. 1970. Yieldable rock bolts for shock loading and grouted
bolts for faster rock stabilization. The Mines Magazine, 60(3): 12–17.
Pahl, P. J. 1981. Estimating the mean length of discontinuity traces. International
Journal of Rock Mechanics and Mining Sciences, 18(3): 221–228.
Pakalnis, R. 2002. Empirical design methods—UBC geomechanics an update. In
R. Hammah, W. Bawden, J. Curran, and M. Telesnicki (eds.), Mining and
Tunnelling Innovation and Opportunity, Proceedings of the Fifth North American
Rock Mechanics Symposium and 17th Tunnelling Association of Canada Conference,
Toronto, Ontario, Canada, 7–10 July, pp. 203–210. Toronto, Ontario, Canada:
University of Toronto.
Pakalnis, R., Poulin, R., and Hadjigeorgiou, J. 1995. Quantifying the cost of dilution to
underground mines. Mining Engineering, 47(12): 1136–1141.
Pardo, C. and Villaescusa, E. 2012. Methodology for back analysis of intensive
rock mass damage at the El Teniente Mine. Presented at MassMin 2012, Sixth
International Conference and Exhibition on Mass Mining, Sudbury, Ontario,
Canada, 10–14 June, Paper 6844, 9pp.
Parker, B. 2004. Geotechnical study of shallow dipping orebodies—Lights of Israel
Underground Gold Mine. BEng thesis, Mining Engineering Department, WA
School of Mines, Curtin University of Technology, Perth, Western Australia,
Australia, 70pp.
Pender, E. B., Hosking, A. D., and Mattner, R. H. 1963. Grouted rock bolts for permanent support of major underground works. Journal, Institution of Engineers,
Australia, 35(7–8): 129–150 (including discussion).
Platford, R. E., Mundy, D. C., and Bonyai, C. G. 1989. Dilution control at Inco
Thompson mine, Manitoba Division. In Proceedings of the 10th Canadian
Underground Operators Conference, 9pp.
Player, J. R. 2012. Dynamic testing of rock reinforcement systems. PhD thesis, Western
Australian School of Mines, Curtin University of Technology, Perth, Western
Australia, Australia, 501pp.
References
495
Player, J. R., Morton, E. C., Thompson, A. G., and Villaescusa, E. 2008. Static and
dynamic testing of steel wire mesh for mining applications of subsurface support. In T. R. Stacey and D. F. Malan (eds.), Proceedings of the Sixth International
Symposium on Ground Support in Mining and Civil Engineering Construction, Cape
Town, South Africa, 30 March–3 April, pp. 693–706. Johannesburg, South Africa:
SAIMM.
Player, J. R., Thompson, A. G., and Villaescusa, E. 2009. Dynamic testing of threadbar
used for rock reinforcement. In M. Diederichs and G. Grasselli (eds.), Rock
Engineering in Difficult Conditions, Proceedings of the Third Canada—US Rock
Mechanics Symposium, Toronto, Ontario, Canada, 9–15 May, Paper 4030 (on
CD-ROM), 14pp. Montreal, Quebec, Canada: CIM.
Player, J. R., Villaescusa, E., and Thompson, A. G. 2004. Dynamic testing of rock
reinforcement using the momentum transfer concept. In E. Villaescusa and
Y. Potvin (eds.), Ground Support in Mining and Underground Construction,
Proceedings of the Fifth International Symposium on Ground Support, Perth, Western
Australia, Australia, 28–30 September, pp. 281–288. Leiden, the Netherlands:
A.A. Balkema.
Popovics, S. 1992. Concrete Materials: Properties, Specifications and Testing, 2nd edn.
Park Ridge, NJ: Noyes Publications.
Potvin, Y. 1988. Empirical open-stope design in Canada. PhD thesis, University of
British Columbia, Vancouver, British Columbia, Canada, 350pp.
Potvin, Y., Hedin, P., Sandy, M., Rosengren, K., and Rosengren, M. D. 2001. Towards
the elimination of rockfall fatalities in Australian Mines, MERIWA Report No
223 on Project No M341, 54pp. Perth, Western Australia, Australia: ACG.
Potvin, Y. and Hudyma, M. 2000. Open stope mining in Canada. In G. Chitombo (ed.),
Proceedings of the MassMin 2000, Brisbane, Queensland, Australia, 29 October–2
November, pp. 661–674. Melbourne, Victoria, Australia: The AusIMM.
Potvin, Y., Hudyma, M., and Miller, H. D. S. 1989. Design guidelines for open stope
support. CIM Bulletin, 82(926): 53–62.
Price, I. 1993. Towards optimal mining. In AMIRA Annual Meeting, Kalgoorlie,
Western Australia, Australia, pp. 82–94.
Priest, S. D. 1983. Computer generation of inclined hemisphere projections.
International Journal of Rock Mechanics and Mining Sciences and Geomechanics
Abstracts, 20(1): 43–47.
Priest, S. D. 1985. Hemispherical Projection Methods in Rock Mechanics, 124pp. London,
U.K.: Allen & Unwin.
Priest, S. D. and Hudson, J. A. 1976. Discontinuity spacings in rock. International
Journal of Rock Mechanics and Mining Sciences, 13(5): 135–148.
Puhakka, R. 1991. Geological Waste Rock Dilution. Espoo, Finland: Finish Association of
Mining & Metallurgical Engineers.
Puhakka, R. 1997. Underground Drilling and Loading Handbook. Helsinki, Finland:
Tamrock Corporation.
Rabcewicz, L. V. 1955. Bolted support for tunnels. Mine and Quarry Engineering, Part I,
March, pp. 111–116; Part II, April, pp. 153–160.
Rauert, N. 1995. The development of the Rib-Roc cable bolt reinforcement system at
the Pasminco Mining Broken Hill operations. In T. S. Golosinski (ed.), Proceedings
of the Sixth Underground Operators’ Conference, Kalgoorlie, Western Australia,
Australia, 13–14 November, pp. 17–26. Melbourne, Victoria, Australia: The
AusIMM.
496
References
Reusch, F. and Beck, D. 2007. Simulating shaft and crusher damage in deep mines. In
Y. Potvin (ed.), Proceedings of the Fourth International Seminar on Deep and High
Stress Mining, Perth, Western Australia, Australia, 7–8 November, pp. 65–79.
Perth, Western Australia, Australia: Australian Centre for Geomechanics, pp.
65–79.
Rocscience Inc., 1990. Website: http://www.rocscience.com/.
Rosengren, M. and Jones, S. 1992. How can we improve fragmentation in the Copper
Mine. Unpublished Mount Isa Mines Limited Internal Report.
Roth, A., Windsor, C. R., Coxon, J., and de Vries, R. 2004. Performance assessment of
high-tensile steel wire mesh for ground support under seismic conditions. In
E. Villaescusa and Y. Potvin (eds.), Ground Support in Mining and Underground
Construction, Proceedings of the Fifth International Symposium on Ground Support,
Perth, Western Australia, Australia, 28–30 September, pp. 589–594. Leiden, the
Netherlands: A.A. Balkema.
Rustan, A. 1990. Burden, spacing and borehole diameter at rock blasting. In Proceedings
of the Third International Symposium on Rock Fragmentation by Blasting, Brisbane,
Queensland, Australia, 26–31 August, pp. 303–310. Melbourne, Victoria, Australia:
The AusIMM.
Salamon, M. D. G. 1970. Stability, instability and the design of mine workings.
International Journal of Rock Mechanics and Mining Sciences, 7(6): 613–631.
Savely, J. P. 1972. Orientation and engineering properties of jointing in the Sierra pit,
Arizona. MS thesis (unpublished), University of Arizona, Tucson, AZ.
Saw, H., Prentice, S., and Villaescusa, E. 2011. Characterisation of cemented rock
fill materials for the Cosmos nickel mine, Western Australia. In C. Leung
and K. T. Wan (eds.), Proceedings of the International Conference on Advances in
Construction Materials through Science and Engineering, Hong Kong, China, 5–7
September. Bagneux, France: RILEM.
Saw, H. and Villaescusa, E. 2011. Research on the mechanical properties of minefill:
Influences of material particle size, chemical and mineral composition, binder
and mixing water. In H. J. Ilgner (ed.), Minefill 2011, Proceedings of the 10th
International Symposium on Mining with Backfill, Cape Town, South Africa, 21–25
March, pp. 143–152. Johannesburg, South Africa: SAIMM.
Saw, H. and Villaescusa, E. 2013. Geotechnical properties of mine fill. In C. F. Leung, S.
H. Goh, and R. F. Chen (eds.), Proceedings of the 18th South East Asian Geotechnical
and Inaugural AGSSEA Conference, Singapore, 29–31 May. Singapore: Research
Publishing.
Schmuck, C. H. 1979. Cable bolting at the Homestake gold mine. Mining Engineering,
31(12): 1677–1681.
Schubert, C. J. and Villaescusa, E. 1998. An approach to hard rock pillar design at
the McArthur River Mine. In Proceedings of the AusIMM Annual Conference—
The Mining Cycle, Mount Isa, Queensland, Australia, 19–23 April, pp. 255–259.
Melbourne, Victoria, Australia: The AusIMM.
Schunnesson, H. and Holme, K. 1997. Drill monitoring for geological mine planning
in the Viscaria copper mine, Sweden. CIM Bulletin, 90(1013): 83–89.
Serafim, J. L. and Periera, J. P. 1983. Considerations of the geomechanical classification of Bieniawski. In Proceedings of the International Symposium on Engineering
Geology and Underground Construction, Lisbon, Portugal, September, vol. 1,
pp. 33–42. Lisbon, Portugal: SPG/LNEC.
References
497
Simser, B., Andrieux, P., and Gaudreau, D. 2002. Rockburst support at Noranda’s
Brunswick Mine, Bathurst, New Brunswick. In R. Hammah, W. Bawden,
J. Curran, and M. Telesnicki (eds.), Mining and Tunnelling Innovation and
Opportunity, Proceedings of the Fifth North American Rock Mechanics Symposium
and 17th Tunnelling Association of Canada Conference, Toronto, Ontario, Canada,
7–10 July, pp. 805–813. Toronto, Ontario, Canada: University of Toronto.
Sjöberg, J. 1992. Failure modes and pillar behaviour in the Zinkgruvan Mine.
In J. R. Tillerson, and W. R. Wawersik (eds.), Proceedings of the 33rd US Rock
Mechanics Symposium, Santa Fe, New Mexico, 3–5 June, pp. 491–500. Rotterdam,
the Netherlands: A.A. Balkema.
Stacey, T. R. 1981. A simple extension strain criterion for fracture of brittle rock.
International Journal of Rock Mechanics and Mining Sciences, 81(6): 469–474.
Stewart, S. B. V. and Forsyth, W. W. 1995. The Mathews’ method for open stope
design. CIM Bulletin, 88(992): 45–53.
Suorineni, F. T. 2012. A critical review of the stability graph method for open stope
design. Presented at MassMin 2012, Sixth International Conference and Exhibition
on Mass Mining, Sudbury, Ontario, Canada, 10–14 June, Paper 6990, 8pp (on
CD-ROM).
Suorineni, F. T., Kaiser, P. K., and Tannant, D. D. 2001. Likelihood statistic for interpretation of stability graph for open stope design. International Journal of Rock
Mechanics and Mining Sciences, 38(5): 735–744.
Teasdale, P. 2001. Open stoping mining method of mining at WMC Resources Gold
Business Unit operations, design process and operating practices. MEngSc
thesis, Western Australian School of Mines, Curtin University of Technology,
Perth, Western Australia, Australia, 68pp.
Terzaghi, K. 1962. Stability of steep slopes on hard unweathered rock. Géotechnique,
12(4): 251–270.
Terzaghi, R. D. 1965. Sources of error in joint surveys. Géotechnique, 15(3): 287–303.
Thomas, G. and Earl, A. 1999. The application of second-generation stope optimisation tools in underground cut-off grade analysis. In Proceedings Strategic
Mine Planning Conference, pp. 1–6. Melbourne, Victoria, Australia: Whittle
Programming Ltd.
Thomas, E. G., Nantel, L. H., and Notley, K. R. 1979. Fill Technology in Underground
Metalliferous Mines, 293pp. Kingston, Ontario, Canada: International Academic
Services.
Thompson, A. G. 2002. Stability assessment and reinforcement of block assemblies near underground excavations. In R. Hammah, W. Bawden, J. Curan,
and M. Telesnicki (eds.), Mining and Tunnelling Innovation and Opportunity,
Proceedings of the Fifth North American Rock Mechanics Symposium and 17th
Tunnelling Association of Canada Conference, Toronto, Ontario, Canada, 7–10 July,
vol. 2, pp. 1439–1446. Toronto, Ontario, Canada: University of Toronto.
Thompson, A. G. 2004. Performance of cable bolt anchors—An update. In
A. Karzulovic and M. A. Alfaro (eds.), Proceedings of the MassMin 2004, Santiago,
Chile, 22–25 August, pp. 317–323. Santiago, Chile: Chilean Engineering
Institute.
Thompson, A. G., Player, J. R., and Villaescusa, E. 2004. Simulation and analysis of
dynamically loaded reinforcement systems. In E. Villaescusa and Y. Potvin
(eds.), Ground Support in Mining and Underground Construction, Proceedings
498
References
of the Fifth International Symposium on Ground Support, Perth, Western
Australia, Australia, 28–30 September, pp. 341–355. Leiden, the Netherlands:
A.A. Balkema.
Thomson, C. and Villaescusa, E. 2011. Open stoping at Golden Grove under high
stress conditions. In P. Hagan (ed.), Mining Education Australia, Research Projects
Review 2011, The AusIMM, pp. 55–59.
Thompson, A. G., Villaescusa, E., and Windsor, C. R. 2012. Ground support terminology and classification: An update. Geotechnical and Geological Engineering, 30(3):
553–580.
Thompson, A. G. and Windsor, C. R. 1992. A classification system for reinforcement and its use in design. In T. Szwedzicki, G. R. Baird, and T. N. Little
(eds.), Proceedings of the Western Australian Conference on Mining Geomechanics,
Kalgoorlie, Western Australia, Australia, 8–10 June, pp. 115–125. Kalgoorlie,
Western Australia, Australia: Western Australian School of Mines.
Tose, S. S. and Baltus, C. 2002. Comparison of the fragmentation results from the use
of electronic detonators and shock tube initiation systems under similar mining
conditions. In X. Wang (ed.), Proceedings of the Seventh International Symposium
on Rock Fragmentation by Blasting—FRAGBLAST 7, Beijing, China, 11–15 August,
pp. 50–53. Beijing, China: Metallurgical Industry Press.
Trotter, D. A. 1991. Vertical crater retreat mining in the Sudbury Basin. Mining Science
and Technology, 13: 131–143.
Trout, P. L. 1997. Formulation and application of new underground mine scheduling models. PhD thesis, The University of Queensland, Brisbane, Queensland,
Australia, 344pp.
Tucker, G., Herbert, D., and Robinson, J. 1998. Bench stoping at Mount Isa Mine,
Mount Isa, Queensland. In Proceedings of the Seventh Underground Operators
Conference, 30 June–3 July, pp. 135–147. Townsville, Queensland, Australia: The
AusIMM.
Ulla, Z. 1997. Applicability of the Mathews stability graph for evaluating stability of
open stopes at the Mount Charlotte Mine. MEngSc thesis, Curtin University of
Technology, Kalgoorlie, Western Australia, Australia.
Ulusay, R. and Hudson, J. A. (eds.) 2007. The Complete ISRM Suggested Methods for
Rock Characterization, Testing and Monitoring: 1974–2006, 628pp. Ankara, Turkey:
ISRM Turkish National Group.
US Army Corps of Engineers. 1980. Engineering and Design: Rock Reinforcement.
Engineering Manual EM1110-1-2907. Washington, DC: USACE.
Varden, R. 2005. A methodology for selection of resin grouted bolts. MEngSc thesis,
Western Australian School of Mines, Curtin University of Technology, Perth,
Western Australia, Australia, 113pp.
Varden, R., Lachenicht, L., Player, J., Thompson, A., and Villaescusa, E. 2008.
Development and implementation of the Garford dynamic bolt at the Kanowna
Belle Mine. In Proceedings of the 10th Underground Operators’ Conference,
Launceston, Tasmania, Australia, 14–16 April, pp. 95–102. Melbourne, Victoria,
Australia: The AusIMM.
Villaescusa, E. 1987. Slope stability analysis at La Caridad Mine, Nacozari, Sonora
Mexico. Master thesis, Colorado School of Mines, Golden, CO, 120pp.
Villaescusa, E. 1991. A three dimensional model of rock jointing. PhD thesis (unpublished), University of Queensland, Brisbane, Queensland, Australia.
References
499
Villaescusa, E. 1995. Sources of external dilution in underground sublevel and bench
stoping. In Proceedings of the EXPLO 95: Exploring the Role of Rock Breakage in
Mining and Quarrying, Brisbane, Queensland, Australia, 4–7 September,
pp. 217–223. Melbourne, Victoria, Australia: The AusIMM.
Villaescusa, E. 1996. Excavation design for bench stoping at Mount Isa mine,
Queensland, Australia. Transactions, Institution of Mining and Metallurgy, Section
A: Mining Industry, 105: A1–A10.
Villaescusa, E. 1998. Geotechnical design for dilution control in underground mining. In R. K. Singhal (ed.), Proceedings of the Seventh International Symposium on
Mine Planning and Equipment Selection, Calgary, Alberta, Canada, 5–9 October,
pp. 141–149. Rotterdam, the Netherlands: A.A. Balkema.
Villaescusa, E. 1999a. Keynote lecture: The reinforcement process in underground
mining. In E. Villaescusa, C. R. Windsor, and A. G. Thompson (eds.), Rock
Support and Reinforcement Practice in Mining, Proceedings of the International
Symposium on Ground Support, Kalgoorlie, Western Australia, Australia, 15–17
March, pp. 245–257. Rotterdam, the Netherlands: A.A. Balkema.
Villaescusa, E. 1999b. Laboratory testing of welded mesh for rock support. In
E. Villaescusa, C. R. Windsor, and A. G. Thompson (eds.), Rock Support and
Reinforcement Practice in Mining, Proceedings of the International Symposium
on Ground Support, Kalgoorlie, Western Australia, Australia, 15–17 March,
pp. 155–159. Rotterdam, the Netherlands: A.A. Balkema.
Villaescusa, E. 2000. A review of sublevel stoping. In G. Chitombo (ed.), Proceedings
of the MassMin 2000, Brisbane, Queensland, Australia, 29 October–2 November,
pp. 577–590. Melbourne, Victoria, Australia: The AusIMM.
Villaescusa, E. 2003. Keynote lecture: Extraction sequences in sublevel stoping. In
Proceedings of the 12th International Symposium on Mine Planning and Equipment
Selection, Kalgoorlie, Western Australia, Australia, 23–25 April, pp. 9–18.
Melbourne, Victoria, Australia: The AusIMM.
Villaescusa, E. 2004. Quantifying open stope performance. In A. Karzulovic and
M. A. Alfaro (eds.), Proceedings of the MassMin 2004, Santiago, Chile, 22–25
August, pp. 96–104. Santiago, Chile: Chilean Engineering Institute.
Villaescusa, E., Azua, J. M., Player, J. R., Thompson, A. G., and Morton, E. C. 2012b.
A database of static and dynamic energy absorption of mesh for rock support. In Proceedings of the 2012 Australian Mining Technology Conference, Perth,
Western Australia, Australia, 8–10 October, pp. 27–34. Perth, Western Australia,
Australia: CRC Mining.
Villaescusa, E. and Brown, E. T. 1990. Characterizing joint spatial correlation using
geostatistical methods. In N. Barton and O. Stephansson (eds.), Rock Joints,
Proceedings of the International Symposium, Loen, Norway, 4–6 June, pp. 115–122.
Rotterdam, the Netherlands: A.A. Balkema.
Villaescusa, E. and Cepuritis, P. 2005. The effects of large scale faults in narrow
vein stope stability—A case study. Rock Mechanics for Energy, Mineral and
Infrastructure Development in the Northern Region, Proceedings of the 40th US
Rock Mechanics Symposium, Anchorage, AK, 25–29 June, Paper 714, 6pp (on
CD-ROM).
Villaescusa, E., Cepuritis, P., Li, J., Heilig, J., Wiles T., and Lund, T. 2003a. Open stope
design and sequences at great depth at Kanowna Belle. Unpublished Research
Report for Placer Dome Asia Pacific, 217p.
500
References
Villaescusa, E., Hassell, R., and Thompson, A. G. 2008. Development of a corrosivity classification for cement grouted cable strand in underground hard rock
mining excavations. In T. R. Stacey and D. F. Malan (eds.), Proceedings of the
Sixth International Symposium on Ground Support in Mining and Civil Engineering
Construction, Cape Town, South Africa, 30 March–3 April, pp. 461–476.
Johannesburg, South Africa: SAIMM.
Villaescusa, E., Karunatillake, G., and Li, T. 1995. An integrated approach to the extraction of the Rio Grande Silver/Lead/Zinc orebodies at Mount Isa. In Singhal
et al. (eds.), Proceedings of the Fourth International Symposium on Mine Planning
and Equipment Selection, Calgary, Alberta, Canada, 31 October–3 November,
pp. 277–283. Rotterdam, the Netherlands: A.A. Balkema.
Villaescusa, E. and Kuganathan, K. 1998. Backfill for bench stoping operations. In
M. L. Bloss (ed.), Minefill 98, Proceedings of the Sixth International Symposium on
Mining with Backfill, Brisbane, Queensland, Australia, 14–16 April, pp. 179–184.
Melbourne, Victoria, Australia: The AusIMM.
Villaescusa, E. and Neindorf, L. 2000. Damage to stope walls from underground blasting. Proceedings of the Structure and Materials, 8: 129–140.
Villaescusa, E., Neindorf, L. B., and Cunningham, J. 1994. Bench stoping of lead/zinc
orebodies at Mount Isa Mines Limited. In New Horizons in Resource Handling
and Geoengineering, Proceedings of the Joint MMIJ/AusIMM Symposium, Ube,
Yamaguchi, 1–5 October, pp. 351–359.
Villaescusa, E., Oñederra, I., and Scott, C. 2004. Blast induced damage and dynamic
behaviour of hangingwalls in bench stoping. Fragblast, 8(1): 23–40.
Villaescusa, E., Sandy, M. P., and Bywater, S. 1992. Ground support investigation and
practices at Mount Isa. In P. K. Kaiser and D. R. McCreath (eds.), Rock Support in
Mining and Underground Construction, Proceedings of the International Symposium
on Rock Support, Sudbury, Ontario, Canada, 16–19 June, pp. 185–193. Rotterdam,
the Netherlands: A.A. Balkema.
Villaescusa, E, Seto, M., and Baird, G. 2002. Stress measurements from oriented core.
International Journal of Rock Mechanics and Mining Sciences, 39(5): 603–615.
Villaescusa, E., Varden, R., and Hassell, R. 2008. Quantifying the performance of resin
anchored rock bolts in the Australian underground hard rock mining industry.
International Journal of Rock Mechanics and Mining Sciences, 45(1): 94–102.
Villaescusa, E., Windsor, C. R., Baird, G., and Seto, M. 2003b. Experimental verification of the Acoustic Emission technique of in-situ rock stress tensor measurement. In K. Sugawara, Y. Obara, and A. Sato (eds.), Rock Stress: Proceedings of the
Third International Symposium on Rock Stress, Kumamoto, Japan, 4–6 November,
pp. 395–402. Lisse, the Netherlands: A.A. Balkema.
Villaescusa, E., Windsor, C. R., and Machuca, L. 2012a. Stress measurements from oriented core—A decade of results. Presented at MassMin 2012, Sixth International
Conference and Exhibition on Mass Mining, Sudbury, Ontario, Canada, 10–14 June,
Paper 6842, 9pp (on CD-ROM).
Villaescusa, E. and Wright, J. 1997. Permanent excavation reinforcement using cement
grouted split set bolts. Proceedings of the AusIMM, 1: 65–69.
Villaescusa, E. and Wright, J. 1999. Reinforcement of underground excavations using
the CT bolt. In E. Villaescusa, C. R. Windsor, and A. G. Thompson (eds.), Rock
Support and Reinforcement Practice in Mining, Proceedings of the International
Symposium on Ground Support, Kalgoorlie, Western Australia, Australia, 15–17
March, pp. 109–115. Rotterdam, the Netherlands: A.A. Balkema.
References
501
Von Kimmelmann, M. R., Hyde, B., and Madgwick, R. J. 1984. The use of computer applications at BCL Limited in planning pillar extraction and the design of mining layouts. In E. T. Brown and J. A. Hudson (eds.), Design and Performance of Underground
Excavations, Proceedings of the ISRM International Symposium, Cambridge, U.K., 3–6
September, pp. 53–63. London, U.K.: British Geotechnical Society.
Wang, C. and Villaescusa, E. 2000. Backfill research at the Western Australian School
of Mines. In G. Chitombo (ed.), Proceedings of the MassMin 2000, Brisbane,
Queensland, Australia, 29 October–2 November, pp. 735–743. Melbourne,
Victoria, Australia: The AusIMM.
Wang, C. and Villaescusa, E. 2001. Influence of water salinity on the properties of
cemented tailings backfill. Transactions, Institution of Mining and Metallurgy,
Section A: Mining Industry, 110(1): 62–65.
Warburton, P. M. 1980. A stereological interpretation of joint trace data. International
Journal of Rock Mechanics and Mining Sciences, 17(4): 181–190.
Wawersik, W. R. and Fairhurst, C. 1970. A study of brittle rock fracture in laboratory
compression tests. International Journal of Rock Mechanics and Mining Sciences,
7(5): 561–575.
Weatherwax, T., Evans, R., and Hafeez, F. 2011. The utilization of paste backfill admixtures at Barrick gold mines. In H. J. Ilgner (ed.), MineFill 2011, Proceedings of the
10th International Symposium on Mining with Backfill, Cape Town, South Africa,
21–25 March, pp. 169–174. Johannesburg, South Africa: SAIMM.
Wiles, T. D. 2001. Map3D course notes. Masters of Mining Geomechanics, Western
Australian School of Mines, Mine Modelling Pty Ltd., Mount Eliza, Victoria,
Australia, 124pp.
Wiles, T. D. 2006. Reliability of numerical modelling predictions. International Journal
of Rock Mechanics and Mining Sciences, 43(3): 454–472.
Wiles, T., Villaescusa, E., and Windsor, C. R. 2004. Rock reinforcement design for overstressed rock using three dimensional numerical modelling. In E. Villaescusa and
Y. Potvin (eds.), Ground Support in Mining and Underground Construction, Proceedings
of the Fifth International Symposium on Ground Support, Perth, Western Australia,
Australia, 28–30 September, pp. 483–489. Leiden, the Netherlands: A.A. Balkema.
Winder, K. 2006. The introduction of cemented hydraulic fill to the Gossan Hill Mine.
MEngSc in Mining Geomechanics thesis, Curtin University of Technology,
Kalgoorlie, Western Australia, Australia.
Windsor, C. R. 1995. Rock mass characterization—A course on structural characterization and structural analysis. Course notes for the Masters of Engineering
Science in Mining Geomechanics, Western Australian School of Mines, Mount
Eliza, Victoria, Australia, 462pp.
Windsor, C. R. 1999. Keynote lecture: Systematic design of reinforcement and support schemes for excavations in jointed rock. In E. Villaescusa, C. R. Windsor,
and A. G. Thompson (eds.), Rock Support and Reinforcement Practice in Mining,
Proceedings of the International Symposium on Ground Support, Kalgoorlie, Western
Australia, Australia, 15–17 March, pp. 35–58. Rotterdam, the Netherlands:
A.A. Balkema.
Windsor, C. R. 2004. A review of long, high capacity reinforcing systems used in rock
engineering. In E. Villaescusa and Y. Potvin (eds.), Ground Support in Mining
and Underground Construction, Proceedings of the Fifth International Symposium on
Ground Support, Perth, Western Australia, Australia, 28–30 September, pp. 17–41.
Leiden, the Netherlands: A.A. Balkema.
502
References
Windsor, C. R. 2009. Rock stress measurements in Earth’s upper crust. In P. Knights
and P. Lever (eds.), Proceedings 2009 Australian Mining Technology Conference,
Technology Solutions for Challenging Financial Times, Brisbane, Queensland,
Australia, 27–28 October, 24pp (on CD-ROM). Melbourne, Victoria, Australia:
The AusIMM.
Windsor, C., Cavieres, P., Villaescusa, E., and Periera, J. 2006. Reconciliation of strain,
structure and stress in the El Teniente mine regions, Chile. In M. Lu, C. C. Li,
H. Kjørholt, and H. Dahle (eds.), In-Situ Rock Stress: Measurement, Interpretation
and Application, Proceedings of the International Symposium on In-Situ Rock Stress,
Trondheim, Norway, 10–21 June, pp. 244–260. Leiden, the Netherlands: Taylor &
Francis.
Windsor, C. R. and Thompson, A. G. 1992. Reinforcement design for jointed rock
masses. In J. R. Tillerson and W. R. Wawersik (eds.), Proceedings of the 33rd US
Symposium Rock Mechanics, Santa Fe, NM, 3–5 June, pp. 521–530. Rotterdam, the
Netherlands: A.A. Balkema.
Windsor, C. R. and Thompson, A. G. 1993. Rock reinforcement—Technology, testing,
design and evaluation. In J. A. Hudson, E. T. Brown, C. Fairhurst, and E. Hoek
(eds.), Comprehensive Rock Engineering, vol. 4, pp. 451–484. Oxford, U.K.: Pergamon.
Windsor, C. R. and Thompson, A. G. 1998. Reinforcement systems—Mechanics,
design and performance testing. In J. Orozco and J. Schmitter (eds.), Proceedings
of the Third North American Rock Mechanics Symposium, Cancún, Mexico, 3–5
June, International Journal of Rock Mechanics and Mining Sciences, 35(4–5), Paper
076, 9pp (on CD-ROM).
Windsor, C. R., Villaescusa, E., Funatsu, T., and Lachenicht, R. 2007. Measurement of
the regional and local stress field along a 10 km strike of the Zuleika Shear Zone
in the Kundana gold mining province of Western Australia. In E. Eberhardt,
D. Stead, and T. Morrison (eds.), Rock Mechanics: Meeting Society’s Challenges
and Demands, Proceedings of the First Canada—US Symposium Rock Mechanics,
Vancouver, British Columbia, Canada, 27–31 May, 7pp. London, U.K.: Taylor
& Francis.
Yu, T. R. and Counter, D. B. 1983. Backfill practices and technology at Kidd Creek
Mines. CIM Bulletin, 76(856): 56–65.
Zhang, L. and Einstein, H. H. 2004. Using RQD to estimate the deformation modulus
of rock masses. International Journal of Rock Mechanics and Mining Sciences, 41:
337–341.
Mining EnginEEring
“…a specialist textbook for mining courses at the advanced undergraduate and postgraduate levels [and] an authoritative, practically
oriented reference work for those involved in the industry, both in
mining operations and as consulting engineers.”
—From the Foreword by Edwin T. Brown, AC, Senior Consultant, Golder Associates Pty Ltd,
Brisbane and Emeritus Professor, University of Queensland, Australia
The first comprehensive work on one of the most important underground
mining methods worldwide, Geotechnical Design for Sublevel Open Stoping
presents topics according to the conventional sublevel stoping process
used by most mining houses, in which a sublevel stoping geometry is chosen
for a particular mining method, equipment availability, and work force
experience. Summarizing state-of-the-art practices encountered during his
25+ years of experience at industry-leading underground mines, the author:
• Covers the design and operation of sublevel open stoping,
including variants such as bench stoping
• Discusses increases in sublevel spacing due to advances in
the drilling of longer and accurate production holes, as well as
advances in explosive types, charges, and initiation systems
• Considers improvements in slot rising through vertical crater
retreat, inverse drop rise, and raise boring
• Devotes a chapter to rock mass characterization, since increases
in sublevel spacing have meant that larger, unsupported stope
walls must stand without collapsing
• Describes methodologies to design optimum open spans and
pillars, rock reinforcement of development access and stope walls,
and fill masses to support the resulting stope voids
• Reviews the sequencing of stoping blocks to minimize in situ
stress concentrations
• Examines dilution control action plans and techniques to
back-analyze and optimize stope wall performance
Featuring numerous case studies from the world-renowned Mount Isa Mines
and examples from underground mines in Western Australia, Geotechnical
Design for Sublevel Open Stoping is both a practical reference for industry
and a specialized textbook for advanced undergraduate and postgraduate
mining studies.
K21696
ISBN-13: 978-1-4822-1188-7
90000
9 781482 211887