University Free State
. 11111111111111111111111111111111111111111111111111111111111111·lllllllllllllllllli
34300004747840
Universiteit Vrystaat
3
llOl A\fW 6 Z
INVESTIGATION INTO THE GROUNDWATER
INTERACTION BETWEEN A DEEP COAL MINE
AND A DEEPER LYING GOLD MINE
by
Morne Burger
Submitted in fuifiIIment of the requirements for the degree
Magister Scientiae in the Faculty of Natural Sciences
and Agriculture, Department of Geohydrology,
University of the Free State, Bloemfontein, South Africa
Supervisor: Dr. P.D. Vermeulen
November 2010
DECLARA TION
November 201 0
I, Morne Burger, declare that the dissertation hereby submitted by me for the Master of Science
degree at the University of the Free State, is my own independent work and has not previously
been submitted by me at another University/Faculty. I further cede copyright of the thesis in
favour of the University of the Free State.
Morne Burger
ACKNOWLEDGEMENTS
This project was only possible with the co-operation of many individuals and institutions. I wish
to record my sincere thanks to the following:
• My faithful Father in heaven who has always answered my prayers.
• Sasol Secunda and Harmony Evander for their co-operation in this study.
• Prof. G. van Tonder and Dr. S. R. Dennis for their assistance and time.
• Mr. Eelco Lucas for his time and patience in WISH.
• The personnel and fellow students at the Institute for Groundwater Studies for their
support.
• My girlfriend Salumé for motivation and love.
• My parents and family, which includes the Calitz family, you have given me the strength
which I can call my own.
• Prof. Teuns Verschoor for believing in me when others did not.
• Most importantly, to Dr. Danie Vermeulen, for whom I have the utmost respect, you have
influenced me far more than the boundaries of this project.
Institute for Groundwater Studies 2010
TABLE OF CONTENTS
CHAPTER ONE: INTRODUCTION 1
1.1 BACKGROUND INFORMATlON 1
1.2 SCOPE OF WORK 2
1.3 ApPROACH TO RESEARCH 3
1.4 REGIONAL SETTING 3
1.4.1 Magisterial District and Regional Services Council 3
1.4.2 Direction and Distance to Nearest Towns 3
1.4.3 Su/face Infrastructure 5
1.5 PHYSIOGRAPHIC CONDITIONS 6
1.5.1 Data collection 6
1.5.2 Climate 7
1.5.3 Rainfall 7
1.5.4 Topography and Drainage 8
CHAPTER TWO: LITERATURE STUDy 10
2.1 INTRODUCTION 10
2.2 PREVIOUS WORK UNDERTAKEN IN THE STUDY AREA 10
2.3 HARMONY GOLD MINING Ev ANDER OPERATIONS 10
2.4 HARMONY GOLD MINING EVANDER: GEOLOGY 12
2.4.1 Witwatersrand Supergroup 12
2.4.2 Ventersdorp Supergroup 12
2.4.3 Transvaal Supergroup 13
2.4.4 Karoo Supergroup 13
2.4.5 Structural Geology 14
2.4.6 Faulting 14
2.4.7 Intrusives 14
2.5 HARMONY GOLD MINING EVANDER: HYDROGEOLOGY 15
2.5.1 Karoo Aquifer 15
2.5.2 Witwatersrand Aquifer 16
2.6 HARMONY GOLD MINING EVANDER: No. 6 SHAFT'S WATER LEVEL 17
2.7 SASOL COAL MINING SECUNDA 19
2.8 SASOL COAL MINING SECUNDA: GEOLOGY 19
2.8.1 Dolerite Intrusions 20
Table of Contents iii
Institute for Groundwater Studies 2010
2.9 SASOL COAL MINING SECUNDA: HYDROGEOLOGY 20
2.1 0 STABLE ENVIRONMENTAL ISOTOPES 24
2.10.1 General applications of stable environmental isotopes in hydrogeology 25
2.10.2 Fractionation 25
2.10.3 Application of stable isotopes 180 and 2H 26
CHAPTER THREE: MINING ACTIVITIES AND METHODS 29
3.1 SASOL SECUNDA COAL MINING COMPLEX 29
3.2 COAL MINING METHODS AND THEIR IMPACT ON GROUND WATER QUALITY AND QUANTITY 31
3.2.1 Board-and-pillar (BP) mining 32
3.2.2 High extraction (HE) Mining 32
3.3 WATER COMPARTMENTS 34
3.3.1 No water storage zones 34
3.4 HARMONY EVANDER GOLD MINING OPERATIONS 36
3.4.1 Gold Mining methods and their impact on groundwater 38
3.4.2 Causes and effects of gold mining methods 38
3.4.3 Prevention of possible groundwater interaction between Harmony Evander gold mining operations
and Sasol Secunda Middelbult (Block 8) 40
CHAPTER FOUR: GEOLOGY OF THE STUDY AREA 42
4.1 INTRODUCTlON 42
4.2 KAROO SUPERGROUP 47
4.2.1 Coal seam sedimentary processes 52
4.2.2 Coal seam contours 53
4.3 TRANSV AAL SUPERGROUP 54
4.3.1 Black Reef Formation 56
4.3.2 Olifants River Group 56
4.3.3 Pretoria Group 57
4.4 VENTERSDORP SUPERGROUP 57
4.4.1 The Klipriviersberg Group 57
4.5 WITWATERSRAND SUPERGROUP 59
4.5.1 Hospital Hill Subgroup 60
4.5.2 Government Subgroup 60
4.5.3 Turffontein Subgroup 60
4.5.4 Johannesburg Subgroup 60
4.6 BASEM.ENT COMPLEX 61
Table of Contents iv
Institute for Groundwater Studies 2010
4.7 STRUCTURAL GEOLOGY 61
4.8 INTRUSIVES 62
4.8.1 Karoo age intrusives 63
4.9 CONCEPTUAL MODEL OF GEOLOGY 64
CHAPTER FIVE: HYDROGEOLOGY OF THE STUDY AREA 66
5.1 INTRODUCTION 66
5.2 AQUIFER TYPES 67
5.2.1 Shallow perched aquifer (unconfined) 67
5.2.2 Shallow weathered zone Karoo aquifers (confined) 67
5.2.3 Deeper fractured Karoo aquifers (confined) 68
5.2.4 Coal mine artificial aquifer (confined) 69
5.2.5 Effective Porosities/ Storage of the Karoo aquifer(s) 69
5.2.6 Hydraulic conductivity of the Karoo aquifer(s) 70
5.2.7 Transvaal dolomitic aquifer (confined) 70
5.2.8 Gold mine artificial aquifer (confined) 71
5.2.9 Deep Witwatersrand aquifer (confined) 71
5.2.10 Effective porosities/ storage of the Witwatersrand aquifer(s) 72
5.2.11 Hydraulic Conductivity for the Witwatersrand aquifer 72
5.2.12 Summary of aquifers 73
5.3 AQUICLUDES/IMPERMEABLE LAYERS 74
5.3.1 Karoo Supergroup (Aquicludes) 74
5.3.2 Transvaal Supergroup (Aquiclude) 74
5.3.3 Ventersdorp Supergroup (Aquiclude) 75
5.4 PREFERENTIAL FLOW PATHS 77
5.4.1 Introduction 77
5.4.2 Geological preferential flow paths 78
5.4.3 Mining activities 80
5.5 WATER LEVELS 82
5.5.1 Karoo aquifer water levels 82
5.5.2 Witwatersrand aquifer groundwater levels 85
5.5.3 Groundwater pumping stations and their effect on the Witwatersrand aquifer's water level 88
CHAPTER SIX: POSSlliLE INTERACTION AREAS WITHIN THE STUDY AREA 93
6.1 POTENTIAL INTERACTION: MINING DEPTI-IS 95
6.1.1 Sasol coal mining 95
Table of Contents V
Institute for Groundwater Studies 2010
6.1.2 Harmony gold mining 95
6.1.3 Difference in mining depths 96
6.1.4 The area around Harmony's No. 1 Shaft 98
6.1.5 The area around Harmony's No. 3 Shaft 99
6.1.6 The area around Harmony's No. 9 Shaft 99
6.1.7 The area around Harmony's No. lO Shaft 100
6.2 POTENTIAL INTERACTION: EVANDER TAILINGS STORAGE FACILITIES 101
6.2.1 Subsidence due to High Extraction coal mining 102
6.3 POTENTIAL INTERACTION: GEOLOGY 108
6.4 POTENTIAL INTERACTION: PREFERENTIAL PATHWAYS 110
CHAPTER SEVEN: HYDRO-CHEMISTRY AND STABLE ENVIRONMENTAL ISOTOPES 112
7.1 ENVIRONMENTAL FACTORS AFFECTING GROUNDWATER CHEMISTRY 112
7.1.1 Climate 112
7.1.2 Geological Setting 112
7.1.3 Biochemical efJects 115
7.1.4 Other factors 115
7.2 SAMPLlNG 118
7.2 SAMPLING PROTOCOL 121
7.3 DATA INTERPRETATION TECHNIQUES 122
7.3.1 Chem ical parameters 122
7.4 HYDRO-CHEMICAL RESULTS INTERPRETATION 123
7.4.1 Electrical Conductivity (EC) 125
7.4.2 pH 129
7.4.3 Sodium (Na) 131
7.4.4 Sulphate (SO/) 135
7.4.5 Chloride (Cl) 138
7.4.6 Bicarbonate (HC03) 140
7.4.7 Piper diagrams 143
7.4.8 Piper diagramfor the study area 143
7.4.9 Expanded Durov Diagram 145
Stiff diagrams 150
7.5 DISCUSSION OF CHEMICAL RESULTS 151
7.5.1 Samples taken from boreholes on surface 151
7.5.2 Samples taken directly from underground 153
7.6 ENVIRONMENTAL STABLE ISOTOPES 156
Table of Contents vi
Institute for Groundwater Studies 2010
7.6.1 Application of environmental isotope and hydro-chemistry in this study 161
7.6.2 Stable environmental isotope data 162
7.6.3 Major ion chemistry and stable environmental isotopes 165
7.6.4 Depth and stable environmental isotopes 166
7.7 DISCUSSION OF ISOTOPE RESULTS 169
7.8 ISOTOPES AND HYDRO-CHEMISTRY CONCLUSIONS 169
7.8.1 Karoo aquifer origin samples recharged by precipitation (KBID, KB5D, WEID, WB3D, WB5D,
MHN47, MHR 45, MDBCasino and Sl-S5) 169
7.8.2 Witwatersrand samples recharged by influxfrom shallower Karoo aquifers (HI, H4, H6, H7) 170
7.8.3 Witwatersrand aqu ifer water recharged by influx from shallower Karoo aquifer with a long residence
time and possibly mixed with connate water or paleo-meteoric water (H2, H3, H5, HlO) 171
7.8.4 Groundwater interaction between the Karoo aquifer and the Witwatersrand aquifer based on hydro-
chemical interpretations 171
7.8.5 Limitations of sampling 172
CHAPTER EIGHT: DECANT 173
8.1 MINING AND RECHARGE IN SASOL'S MIDDELBULT (BLOCK 8) 173
8.2 GROUNDWATER SYSTEMS WITHIN THE COAL MINE 174
8.3 CONCEPTUAL DECANT MODEL FOR SASOLMIDDELBULT BLOCK 8 174
8.4 MINING AND RECHARGE IN HARMONY EVANDER GOLD MINING OPERATIONS 176
8.5 GROUNDWATER SYSTEMS WITHIN THE GOLD MINE 177
8.6 STUDY AREA CONCEPTUAL DECANT MODEL FOR THE GOLD MINE 177
8.6.1 Current conceptual modelfor decant in Harmony Evander gold mining operations under pumping
conditions 181
8.6.2 Conceptual model for decant in Harmony Evander gold mining operations (Post-closure) 181
8.7 TIME OF GOLD MINE PIEZOMETRIC LEVEL TO REACI·I COAL SEAM FLOOR 186
8.7.1 Assumptions 186
8.7.2 Estimate of the piezometric level to reach coal seam floor with a constant inflow rate 187
8.7.3 Estimate of piezometric level to reach coal seam floor applying Darcy's law 189
8.7.4 Conceptual scenario when pumping ceases 191
CHAPTER NINE: CONCLUSIONS AND RECOMMENDA TIONS 192
9.1 CONCLUSIONS 192
9.1.1 Isotopes and hydro-chemistry conclusions 197
9.1.2 Summary of possible interaction areas 199
9.1.3 Expected impact of coal mine on the water quantity and quality within the study area 201
Table of Contents vii
Institute for Groundwater Studies 2010
9.1.4 Expected impact of gold mine on the water quantity and quality within the study area 203
9.1.5 Current conceptual model for decant in Sasol coal m ining operations under operational (pumping)
conditions 204
9.1.6 Conceptual model for decant in Harmony Evander gold mining operations post-closure (no-pumping)
204
9.1.7 Timefor gold mine workings' voids to jlood 204
9.1.8 Limitations of this study 205
9.2 RECOMMENDATIONS 206
REFERENCES 208
APPENDICES 215
Table of Contents viii
Institute for Groundwater Studies 2010
LIST OF FIGURES
FIGURE 1: LOCATION OF THE STUDY AREA/INTERACTION AREA WITH ITS SURROUNDING TOWNS 4
FIGURE 2: LOCATION OF THE STUDY AREA AND WITH QUATERNARY CATCHMENTS C24D, C23G AND C23H 6
FIGURE 3: CLIMATE DATA FOR SECUNDA WEATHER STATION (SECUNDA WEATHER STATION, 04783303) 7
FIGURE 4: MONTHLY RAINFALL DATA FOR THE SECUNDA WEATHER STATION (SECUNDA WEATHER STATION,
04 783303) 7
FIGURE 5: TOPOGRAPHY OF THE STUDY AREA AND SURFACE DRAINAGE DIRECTIONS 9
FIGURE 6: HIGH EXTRACTION MINING PANELS OCCURRING CLOSE TO THE LESLIE AND WINKELHAAK SLlMES DAMS
.................................................................................................................................................................................... 22
FIGURE 7: HISTOGRAM OF THE SHALLOW (DRILLED 0-30M) KAROO AQUIFER'S WATER LEVELS 23
FIGURE 8: HISTOGRAM OF THE DEEP (DRILLED 80-150M) KAROO AQUIFER'S WATER LEVELS 23
FIGURE 9: HISTOGRAM OF THE DEEP & SHALLOW (DRILLED 0-150M) KAROO AQUIFER'S WATER LEVELS 24
FIGURE 10: METEORIC WATER LINES (ADAPTED FROM GAT, 1996) 27
FIGURE 11: MIDDELBULT LAYOUT OF CURRENT AND FUTURE MINING 30
FIGURE 12: BOARD-AND-PILLAR MINING FOLLOWED BY HIGH EXTRACTION (VERMEULEN & USHER, 2006) 31
FIGURE 13: COAL MINE WATER COMPARTMENTS AND NO WATER STORAGE ZONES 35
FIGURE 14: GOLD MINING AREA WITH SHAFTS AND VENTILATION SHAFTS 37
FIGURE 15: GOLD MINING STOPE FROM WHICH ORE CONTAINING GOLD IS EXTRACTED FROM THE KIMBERLEY REEF
.................................................................................................................................................................................... 38
FIGURE 16: GROUNDWATER FLOWING INTO THE GOLD MINE VIA EXPLORATION BOREHOLE IN A GOLD MINE SHOOTS
.................................................................................................................................................................................... 39
FIGURE 17: GEOLOGICAL MAP OF THE EVANDER BASIN 43
FIGURE 18: EXPLORATION BOREHOLE POSITIONS WITH IN THE STUDY AREA 45
FIGURE 19: GEOLOGICAL LOG FOR EXPLORATION BH 390 (0-1129M) 46
FIGURE 20: PHOTOGRAPHS OF TYPICAL SEDIMENTARY ROCKS FOUND IN KAROO SUPERGROUP OF THE STUDY
AREA 47
FIGURE 21: INTERPOLATED THICKNESS OF THE KAROO SUPERGROUP IN THE STUDY AREA 48
FIGURE 22: INTERPOLATED OVERBURDEN THICKNESS IN M ABOVE THE No. 4L COAL SEAM ROOF OF THE
MIDDELBULT BLOCK 8 AREA 50
FIGURE 23: NORTH-SOUTH CROSS SECTION ILLUSTRATING THE No. 4L COAL SEAM AND THE B4 AND B8 DOLERITE
SILLS 51
FIGURE 24: WEST-EAST CROSS SECTION ILLUSTRATING THE No 4L COAL SEAM AND THE B4 AND B8 DOLERITE
SILLS 51
Table of Contents ix
Institute for Groundwater Studies 2010
FIGURE 25: SURFACE VIEW OF THE CROSS SECTIONS 52
FIGURE 26: INTERPOLATED ELEVATIONS OF THE No. 4L COAL SEAM FLOOR IN M.A.M.S.L. 54
FIGURE 27: THINNING AND ABSENCE OF THE TRANSVAAL SUPERGROUP TOWARDS THE EAST-SOUTHEAST OF THE
STUDY AREA 55
FIGURE 28: THINNING OF THE TRANSVAAL SUPERGROUP NORTH TO SOUTH ACROSS THE STUDY AREA 56
FIGURE 29: INTERPOLATED THICKNESS OF THE VENTERSDORP SUPERGROUP 58
FIGURE 30: THINNING OF THE KLIPRIVIERBERG GROUP LAVAS TOWARDS THE EAST-SOUTHEAST OF THE STUDY
AREA 58
FIGURE 31: THINNING OF THE KLIPRIVIERBERG GROUP LAVAS TOWARDS THE SOUTH OF THE STUDY AREA 59
FIGURE 32: MAJOR FAULTS IN THE STUDY AREA 62
FIGURE 33: INTERPOLATED THICKNESS OF THE B4 DOLERITE SILL COVERING THE STUDY AREA 63
FIGURE 34: SURFACE VIEW OF THE DOLERITE DYKES IN THE STUDY AREA 64
FIGURE 35: CONCEPTUAL MODEL OF THE GEOLOGY OF THE STUDY AREA 65
FIGURE 36: CONCEPTUAL MODEL OF THE DIFFERENT AQUIFERS IN THE STUDY AREA (NOT TO SCALE) 67
FIGURE 37: POROSITY VALUES OF SELECTED ROCK TYPES (KRUSEMAN & DE RIDDER, 1994) 75
FIGURE 38: HYDRAULIC CONDUCTIVITY VALUES OF SELECTED ROCK TYPES (KRUSEMAN & DE RIDDER, 1994) 76
FIGURE 39: VENTERSDORP THICKNESS IN THE STUDY AREA 77
FIGURE 40: RELATIONSHIP BETWEEN FRACTURE APERTURE, FRACTURE SPACING AND AQUIFER HYDRAULIC
CONDUCTIVITY, FOR AN AQUIFER CONSISTING OF PLANAR, PARALLEL UN IFORM FRACTURES (COOK, 2003) .. 78
FIGURE 41: FRACTURES NEXT TO DYKE WITH WATER AT HARMONY EVANDER No. 8 SHAFT 11 LEVEL (1340 M
BELOW SURFACE) 79
FIGURE 42: FAULT BEARING WATER AT HARMONY EVANDER No. 8 SHAFT 11 LEVEL (1340 M BELOW SURFACE) .. 80
FIGURE 43: GOLD MINE SHAFT POSITIONS 81
FIGURE 44: KAROO AQUIFER WATER LEVELS 83
FIGURE 45: CORRELATION OF THE INTERPOLATED WATER LEVELS WITH SURFACE TOPOGRAPHY 84
FIGURE 46: INTERPOLATED GROUNDWATER ELEVATION MAP FOR THE KAROO AQUIFER WATER LEVELS (BAYESIAN
INTERPOLATION) 84
FIGURE 47: INTERPOLATED ELEVATIONS FOR THE No. 4L COAL SEAM ROOF 86
FIGURE 48: INTERPO.LATED THICKNESS OF THE OVERBURDEN ABOVE THE WITWATERSRAND SUPERGROUP 87
FIGURE 49: INTERPOLATED GROUNDWATER LEVELS AROUND THE GOLD MINE SHAFTS (INVERSE DISTANCE
WEIGHTING METHOD) 88
FIGURE 50: CONCEPTUAL MODEL OF PIEZOMETRIC LEVELS WITHIN A CONFINED AQUIFER 90
FIGURE 51: CONCEPTUAL MODEL OF THE WATER LEVELS IN THE WITWATERSRAND AQUIFER WHERE PUMPING
STATIONS HAVE BEEN DECOMMISSIONED 91
FIGURE 52: CONCEPTUAL MODEL OF THE WATER LEVELS IN THE WITWATERSRAND AQUIFER WHERE PUMPING
STATIONS ARE IN OPERATION 92
Table of Contents x
Institute for Groundwater Studies 2010
FIGURE 53: SASOL MINING'S MIDDELBULT (BLOCK 8) CURRENT AND FUTURE MINING OVER HARMONY EVANDER'S
GOLD MINING ACTIVITIES 94
FIGURE 54: OVERBURDEN THICKNESS IN METRES ABOVE THE No.4L COAL SEAM 95
FIGURE 55: GENERAL EVANDER GOLD MINING OPERATIONS ILLUSTRATING THEIR SHAFT SYSTEM (NOT TO SCALE).
(PRESENTATION 53 KINROSS, HARMONY EVANDER) 97
FIGURE 56: CROSS SECTION THROUGH EVANDER No.1 SHAFT 98
FIGURE 57: CROSS SECTION THROUGH EVANDER No. 2 AND No.3 SHAFT 99
FIGURE 58: CROSS SECTION THROUGH EVANDER No.9 SHAFT 100
FIGURE 59: EVANDER No. 10 SHAFT 101
FIGURE 60: HARMONY EVANDER GOLD MINE TAILINGS STORAGE FACILITIES 102
FIGURE 61: SCHEMATIC CROSS-SECTION OF A SUBSIDED HIGH EXTRACTED PANEL 103
FIGURE 62: POSSIBLE AREAS OF SUBSIDENCE RESULTING FROM HIGH EXTRACTION COAL MINING 105
FIGURE 63: KINROSS SLlMES DAM CONCEPTUAL CROSS SECTIONS 106
FIGURE 64: LESLIE SLlMES DAM CONCEPTUAL CROSS SECTIONS 106
FIGURE 65: WINKELHAAK SLlMES DAM CONCEPTUAL CROSS SECTIONS 107
FIGURE 66: B4 DOLERITE THICKNESS IN METRES WITH THE WEST-EAST AND NORTH-SOUTH SECTION LlNES 107
FIGURE 67: THINNING OF THE VENTERSDORP SUPERGROUP FROM NORTH TO SOUTH 108
FIGURE 68: THINNING OF THE VENTERSDORP SUPERGROUP FROM WEST TO EAST lO9
FIGURE 69: POTENTIAL INTERACTION AREA BASED ON GEOLOGY 109
FIGURE 70: SURFACE VIEW OF THE POTENTIAL INTERACTION AREA COMBINED WITH THE WEST -EAST AND NORTH-
SOUTH CROSS SECTIONS 110
FIGURE 71: SURFACE VIEW OF THE SAMPLING POSITIONS 119
FIGURE 72: UNDERGROUND VIEW OF THE SAMPLING POSITIONS 120
FIGURE 73: A SAMPLE BEING COLLECTED IN THE GOLD MINE NEXT TO A FRACTURE 121
FIGURE 74: UNDERGROUND SAMPLING POSITIONS WITH THE CURRENT AND FUTURE COAL MINING LAYOUT 124
FIGURE 75: DYKES AND FAULTS ACROSS THE STUDY AREA 126
FIGURE 76: EC VALUES OF SAMPLES COLLECTED IN THIS STUDY FROM SHALLOW (LEFT) TO DEEP (RIGHT) 127
FIGURE 77: ELECTRICAL CONDUCTIVITY VALUES FOR SAMPLES COLLECTED IN THIS STUDY 128
FIGURE 78: pH VALUES OF SAMPLES COLLECTED IN THIS STUDY FROM SHALLOW (LEFT) TO DEEP (RIGHT) 129
FIGURE 79: pH VALUES FOR SAMPLES COLLECTED IN THIS STUDY 130
FIGURE 80: SODIUM SATURATION OVER TIME 132
FIGURE 81: SODIUM CONCENTRATIONS FOR THE SAMPLES COLLECTED IN THIS STUDY FROM SHALLOW (LEFT) TO
DEEP (RIGHT) 133
FIGURE 82: SODIUM CONCENTRATIONS OF THE SAMPLES COLLECTED IN THIS STUDY 134
FIGURE 83: SULPHATE CONCENTRATIONS IN THE STUDY AREA FROM SHALLOW (LEFT) TO DEEP (RIGHT) 136
Table of Contents xi
Institute for Groundwater Studies 2010
FIGURE 84: SULPHATE CONCENTRATIONS FOR SAMPLES COLLECTED IN THIS STUDY 137
FIGURE 85: CHLORIDE CONCENTRATIONS IN THE STUDY AREA FROM SHALLOW (LEFT) TO DEEP (RIGHT) 138
FIGURE 86: CHLORIDE CONCENTRATIONS FOR SAMPLES COLLECTED IN THIS STUDY 139
FIGURE 87: BICARBONATE CONCENTRATIONS FROM SHALLOW TO DEEP 141
FIGURE 88: BICARBONATE CONCENTRATIONS FOR SAMPLES COLLECTED IN THIS STUDY 142
FIGURE 89: PIPER DIAGRAM, ADAPTED FROM PIPER (1944) 143
FIGURE 90: PIPER DIAGRAM FOR THE SAMPLES COLLECTED IN THIS STUDY 145
FIGURE 91: INTERPRETATIONS FROM THE EXPANDED DUROV DIAGRAM (1 ) 146
FIGURE 92: INTERPRETATIONS FROM THE EXPANDED DUROV DIAGRAM (2) 147
FIGURE 93: EXPANDED DUROV DIAGRAM FOR THE SAMPLES COLLECTED IN THE STUDY AREA 148
FIGURE 94: STIFF DIAGRAM FOR THE UNDERGROUND SAMPLES COLLECTED IN THIS STUDY 150
FIGURE 95: ORIGIN OF SAMPLES H6 AND H7 ACCORDING TO HYDRO-CHEMISTRY 156
FIGURE 96: SCHEMATIC OF ISOTOPE ENRICHMENT/DEPLETION IN THE VAPOR AND CONDENSATION PHASE 157
FIGURE 97: THE RELATIONSHIP BETWEEN t;,.180 AND t;,.D FOR PRETORIA RAINFALL (TALMA, 2003) 159
FIGURE 98: METEORIC WATER LINES (ADAPTED FROM GAT, 1996) 160
FIGURE 99: GENERALISED t;,.D VERSUS t;,.180 PLOT SHOWING THE WORLD RAINFALL (METEORIC) LINE AND
PROCESSES COMMONLY ASSOCIATED OF DEVIATION FROM THE RAINFALL LINE (FRITZ ET AL., 1980) 160
FIGURE 100: 0180 AND oD RELATIONSHIP ISOTOPE PLOT FOR THE SAMPLES COLLECTED IN THE STUDY 164
FIGURE 101: 0180 VS. ELECTRICAL CONDUCTIVrry PLOT 166
18
FIGURE 102: 0 0 VS. DEPTH FOR SAMPLES IN THE STUDY AREA 167
FIGURE 103: 8180 VS. DEPTH FOR UNDERGROUND SAMPLES ORIGINATING IN THE DIFFERENT MINES 168
FIGURE 104: NORTH-SOUTH CROSS-SECTION OF THE CONCEPTUALISATION OF THE DIFFERENT WATER LEVELS 175
FIGURE 105: THE INTERPOLATED THICKNESS OF THE OVERBURDEN COVERING THE WITWATERSRAND
SUPERGROUP 177
FIGURE 106: INTERPOLATED THICKNESS OF THE VENTERSDORP SUPERGROUP 178
FIGURE 107: CONCEPTUAL MODELS FOR PIEZOMETRIC LEVELS 180
FIGURE 108: COMPARTMENTALlSATION WITHIN THE WITWATERSRAND AQUIFER RESULTING FROM FAULTS AND
DYKES 182
FIGURE 109: POPLAR AND ROLSPRUrr PROJECT AREAS 183
FIGURE 110: CONCEPTUAL MODEL WITH THE INTERPOLATED WrrWATERSRAND AQUIFER WATER LEVELS 185
FIGURE 111: VOLUME OF VOIDS VERSUS ELEVATIONS 188
FIGURE 112: INFLUX OVER TIME 189
FIGURE 113: INFLUX wrru ELEVATIONS OVER TIME 190
FIGURE 114: INFLUX OVER TIME INDICATING YEARS FOR THE GOLD MINE TO FILL WITH WATER 190
Table of Contents xii
Institute for Groundwater Studies 2010
LIST OF TABLES
TABLE 1: No. 4L COAL SEAM ELEVATIONS THROUGHOUT MIDDELBULT BLOCK 8 29
TABLE 2: GENERALISED STRATIGRAPHY OF THE STUDY AREA 44
TABLE 3: THICKNESS INFORMATION OF THE DIFFERENT COAL SEAMS (ADAPTED FROM JMA, 2002, REPORT No:
JW98/02/8068) 49
TABLE 4: STATISTICS FOR RESULTS ON PACKER HYDRAULIC CONDUCTIVITY TESTING OF THE DWYKA TILLlTE
(HODGSON ET AL., 1998) 69
TABLE 5: TOTAL POROSITY FOR SAMPLES TAKEN AT BLOCK 8 JASPER MULLER ASSOCIATES CC (2002) 70
TABLE 6: SUMMARY OF THE AQUIFER PARAMETERS WITHIN THE STUDY AREA (VERMEULEN & DENNIS, 2009;
KRUSEMAN & DE RIDDER, 1994) 74
TABLE 7: PREFERENTIAL FLOW PATHS ENCOUNTERED IN THE STUDY AREA 82
TABLE 8: AVAILABLE WATER LEVELS FOR THE WITWATERSRAND AQUIFER. 85
TABLE 9: SHAFT'S AND PUMPING STATIONS OF THE GOLD MINE 89
TABLE 10: DIFFERENCE IN MIN ING DEPTHS 96
TABLE 11: CHARACTERISTICS, SOURCES AND INTERACTIONS OF ELEMENTS OCCURRING AS MAJOR CATIONS AND
ANIONS IN NATURAL WATERS (CONCENTRATIONS FROM ApPELO AND POSTMA, 2005) 117
TABLE 12: SAMPLING SUMMARY 118
TABLE 13: SUMMARY OF SAMPLES TAKEN FROM BOREHOLES ON SURFACE OF THE STUDY AREA 151
TABLE 14: SUMMARY OF SAMPLES TAKEN DIRECTLY FROM UNDERGROUND IN THIS STUDY 153
TABLE 15: HARMONY GOLD MINE DATA USED TO CALCULATE VOLUME OF VOIDS/FLOODING 186
TABLE 16: VALUES OF INDIVIDUAL SHAFTS AND LEVELS 187
Table of Contents xiii
Institute for Groundwater Studies 2010
LIST OF ABBREVIATIONS
A Area
ABA Acid Base Accounting
AP Acid Potential
ARD Acid Rock Drainage
BP Board and Pillar
DWAF Department of Water Affairs and Forrestry
DME Department of Mines and Energy
EC Electrical Conductivity
EIA Environmental Impact Assessment
EPA Environmental Planning and Assessment
EMP Environmental Management Programme
EMPR Environmental Management Program Report
GIS Geographic Information System
GMWL Global Mean Water Line
GPS Global Positioning System
HE High Extraction
i Gradient
IGS Institute for Groundwater Studies
K-value Hydraulic conductivity
MAR Mean Annual Run-off
NNP Net Neutralisation Potential
NP Neutralisation Potential
Abbreviations xiv
Institute for Groundwater Studies 2010
NWA National Water Act
Rr Reference standard
Rs Reference sample
S Storage coefficient
SMOW Standard Mean Ocean Water
Specific yield
TSF Tailings Storage Facility
TDS Total Dissolved Solids
WISH Windows Interpretation System for Hydrogeologist
Q Rate of inflow
'I
1
I
I
Abbreviations xv
Chapter One: Introduction 12010
Chapter One: Introduction
1.1 Background Information
Coal and gold mining in South Africa and more specifically the Mpumalanga Province have been
in existence for more than a century. Due to the scale of these operations in area and the time
period of operation, these activities have significantly altered the geohydrological environment of
the area. These alterations or impacts are not only limited to the operational life of the mines,
but are expected to continue for an indefinite number of years after mining has ceased. As such,
an understanding of the groundwater interaction between the coal and gold mines are to be
exercised in area where these two mining activities affect each other hydrogeologically.
One of the key areas of concern is the phenomenon of intermine flow existing between the
different mines. This has been studied by previous researchers, such as Grobbelaar (2001) and
Havenga (2002). After the closure of mines, it is expected that water in the mined-out areas will
flow along preferred pathways and accumulate in lower-lying areas (Grobbelaar, 2001). Over
time, these man-made voids will fill up with water and hydraulic gradients will be exerted onto
peripheral areas within mines, resulting in groundwater flow between mines and possibly to the
surface. This flow is referred to as intermine flow (Grobbelaar, 2001). Due to the potential long-
term impact of intermine flow in terms of water quantities and qualities, the Department of Water
Affairs and Forestry regards intermine flow as one of the most important challenges in the
mining industry.
In order to address these issues, individual mining houses undertake periodic and ongoing
research into water quantity and quality issues. Through the collaborative contribution of Sasol
Coal Mining (Pty) Limited, Secunda and Harmony Gold Mining (Pty) Limited, Evander this
research project was launched to address certain issues to provide a more quantitive solution to
these problems. This research project was done to determine the potential groundwater
interaction between Sasol Coal Mining's Middelbult (Block 8) Colliery, which extracts the coal
above the Harmony Evander Gold Mining Operations extracting the underlying gold deposits.
Groundwater interaction between a deep coal mine and a deeper lying gold mine 1
Chapter One: Introduction 12010
1.2 Scope ofWork
Impacts have to be assessed in terms of the potential environmental liabilities of each mining
operation and the impact on closure costs.
• It is appropriate that each party accepts and acknowledges its legal responsibility for
decanting or management of its seepage water into the activities or areas within the
jurisdiction of the other party. Each party has the responsibility to exercise a "duty of care" to
prevent any pollution/degradation or harm that may arise from their activities. This "duty of
care" entails taking reasonable steps to prevent the pollution/degradation from occurring
continuing and/or to minimise and remediate the effects as far as is practicably possible.
• It is therefore necessary to determine that whoever creates the risk remains responsible for
managing that risk and is liable for the impact or consequences of that risk.
The main aim of the proposed work is to determine the potential groundwater interaction
between Sasol Mining's Middelbult (Block 8) and Harmony Evander Operations including the
combined impacts on the environment. The impacts have to be assessed in terms of the
potential environmental liabilities of each mining operation and the impact on closure costs.
To determine the potential water groundwater interaction between the mines, the following
issues need to be addressed:
~ Understand the hydrological interaction, if any, between Middelbult and Harmony Evander
Operations.
• Is there any influence between the coal and gold mines during the operational phase?
• Is there any influence between the coal and gold mines during the decommissioning
phase and post closure?
• Determine the influence pre-operational phase (exploration done by both parties before
mining commenced).
• Identify preferred pathways that has been created by industry, which includes exploration
boreholes (pre- and during mining), shafts, boreholes, etc.
~ Determine (model) if decanting will occur:
• Will Middelbult decant?
• Will Harmony decant?
~ If decanting is going to occur, determine (model) the following:
• Qualities to be expected?
• Quantities to be expected?
Groundwater interaction between a deep coal mine and a deeper lying gold mine 2
Chapter One: Introduction 12010
~ Will seepage occur as a result of the coal and gold mining?
~ Identify where seepage is taking place and what is the cause.
1.3 Approach to Research
The first step towards achieving these aims was the collection and collation of available data.
This included large quantities of data from the different mines in different forms (documents,
Excel files, GIS files, etc.). Review the available literature on topics related to the research. This
information was combined into databases and formatted to be compatible with GIS systems,
such as the SURFER surface mapping system and Windows Interpretation System for
Hydrogeologist (WISH) software packages.
Samplings were done at strategic points on the surface as well as underground in each
individual mine. These samples were then submitted to laboratories for hydro-chemical and
isotope analysis and interpreted to "fingerprint" each specific sample's origin.
1.4 Regional Setting
1.4.1 Magisterial Districtand Regional Services Council
The study area/interaction, area is situated in the Mpumalanga Province north-northwest of the
town of Secunda (Figure 1). This area is part of the Highveld Coal Field (Havenga, 2002). It is
located in the Block 8 coal reserves owned by Sasol Mining (Pty) Limited which overlies the
deeper Harmony Evander (Pty) Limited gold mining operations. The study area falls within the
Highveld East Magisterial District and Govan Mbeki Regional Services Council.
1.4.2 Direction and Distance to Nearest Towns
Kinross lies to the north-northeast, approximately 12 km from the northern study area boundary,
whilst Embalenhle is situated on the southern boundary. Secunda is located to the southeast,
approximately 8 km from the southem study area boundary. Evander is located approximately in
the centre of the study area. The co-ordinates (Transverse Mercator: LO 29) of the investigated
study area are -20000 to 25000 East and -2920000 to -2945000 South (Figure 1).
Groundwater interaction between a deep coal mine and a deeper lying gold mine 3
Chapter One: Introduction 12010
-29
Gauteng
~.
".,..Ill
Mpumalanga
LEGEND
7ft. Interaction .aroe-arulidde-Ibult boundariesN Ha rm ony bou Ilda ry
lil 4
Figure 1: Location of the study arealinteraction area with its surrounding towns.
Groundwater interaction between a deep coal mine and a deeper lying gold mine 4
Chapter One: Introduction 12010
1.4.3 Surface Infrastructure
1.4.3.1 Roads
Several roads cross the study area, ranging in magnitude from national freeways such as the
N17 [(previously R29) and arterial routes (Leandra-Standerton Road (RSO)] to secondary and
farm roads (Middelbult Mine Block Expansion EMPR, 2003).
1.4.3.2 Power lines
A number of Eskom power lines cross the surface of the area (Middelbult Mine Block Expansion
EMPR,2003).
1.4.3.3 Railway lines
The Springs-Trichardt-Bethal railway line passes over the northern extent of the study area. The
railway line links to the Witbank-Richards Bay railway line (Middelbult Mine Block Expansion
EMPR,2003).
1.4.3.4 Pipelines
A number of Rand Water and gold mine water and slurry pipelines cross over the surface of the
area (Middelbult Mine Block Expansion EMPR, 2003).
1.4.3.5 Servitudes
A number of servitudes exist in the area for:
District roads;
Pipelines;
Eskom overhead electrical power distribution lines;
Telkom telephone lines;
1.4.3.6 Land Tenure and Use
The land adjacent to mining activities is utilised for agricultural purposes, mixed residential,
commercial, and industrial uses. Summer crops such as maize and sunflower are cultivated.
The rest of the land is utilised for sheep/cattle grazing and for game farming (Middelbult Mine
Block Expansion EMPR, 2003).
1.4.3.7 Catchment in which the study area is situated
The study area is located in the quaternary sub-catchments C24D, C23G and C23H of the
Waterval River Catchment drainage region as shown in Figure 2. Grootspruit, Winkelhaakspruit,
Groundwater interaction between a deep coal mine and a deeper lying gold mine S
Chapter One: Introduction 12010
and Trichardtspruit drain over the area and join the Waterval River upstream of the confluence
(Vermeulen & Dennis, 2009).
LEGEND
N Quaternary catchments
N"/SI.. Interaction areaMiddelbult boundaries
N Harmony boundary
Figure 2: Location of the study area and with quatemary catchments C24D, C23G and C23H.
1.5 Physiographic Conditions
1.5.1 Data collection
Climatic data for the study area was obtained from the South African Weather Service for the
Secunda weather station (Secunda, 04783303) which has accurate records of weather
conditions for more than 20 years.
Groundwater interaction between a deep coal mine and a deeper lying gold mine 6
Chapter One: Introduction 12010
1.5.2 Climate
The monthly distribution of average daily maximum temperatures shows that the average
midday temperatures for Secunda range from 16.4°e in June to 25.8°e in January (South
African Weather Service). The region is coldest during the month of June, when the mercury
drops to ooe on average during the night (Figure 3).
Average """day temperature ('Cj Ivtr.gt noght-tme temperature ('C) A .rage """a' (rM'I)
26 13 110
16~ _ o~ _
J F M A " J JAS 0 N 0 J F ~ A ~ J JAS 0 N 0 J FI AI J JASOljO
Figure 3: Climate data for Secunda weather station (Secunda Weather Station, 04783303).
1.5.3 Rainfall
Secunda normally receives about 560 mm of rain per year, with most rainfall occurring during
the summer. It receives its lowest rainfall (0 mm) in June and the highest (105 mm) in January
as shown in Figure 4 (Secunda Weather Station, 04783303).
Monthly rainfall
350
- 300E 250-E 200-ii 150 f-t:::"a=ii 100 1- -- I- r.-- - , ~ f-50 ltu t- I0 lit II II JL II lt II 11111.11 IIIIIIJ, 111,.1 IIII I II I III
q q LO ID ,.... ,..... 00 en. C.l 0 .-.I N rt.) rt) q LO~ ~ ~ ~ en, en ~ en Cl 0, 0 9 0 9 9 9 9'" 'C
c: .u.... ~ ~ c: .u... ~ ~ c: .... ~ ~ c: ..... c: .."...l.ru a._ m Q_0 « 0 « -«.ti 0 Q_0 « «ti U ::::J a._0 « «ti U0
Month
Figure 4: Monthly rainfall data for the Secunda weather station (Secunda Weather Station, 04783303).
Groundwater interaction between a deep coal mine and a deeper lying gold mine 7
Chapter One: Introduction 12010
1.5.4 Topography and Drainage
The topography of the study area is shown in Figure 5. The surface elevation varies from 1560
to 1695 m.a.m.s.1. The topography dips towards the rivers. According to Midgley et al., (1994),
the mean annual runoff (MAR) varies between 30 to 73 mm/a for the study area. The
groundwater contribution to base flow is estimated at approximately 4 to 7 mm/a (Midgley et al.,
1994).
Groundwater interaction between a deep coal mine and a deeper lying gold mine 8
Chapter One: Introduction 12010
Cc
.".11' 0 00'
"".,....",'~V 'l? .0~;\0 tt" 00
Ito. Ito. ;1 I'\.~ 0 0 CJ
,...,.. ~ 0'l7 0 0 0
,.. Ito. 0~l·"..a 0
1',.. .0 r;..O"l7 0 0
,.. 'IV ~ 0 tI". 0
Ito.".. ,..~,.. ~ {?'"
Ito.
Figure 5: Topography of the study area and surface drainage directions.
Groundwater interaction between a deep coal mine and a deeper lying gold mine 9
Chapter Two: Literature Study 12010
Chapter Two: Literature Study
2.1 Introduction
The literature study involved a review of all relevant available information, including Harmony
Gold Mining Evander and Sasol Coal Mining Secunda groundwater databases, all relevant
geological databases, previous sampling surveys, hydro-census, remote sensing data and a
literature survey. It also involved the application of environmental isotopes and hydro-chemistry
in hydrogeological aquifer conceptualisation.
A comprehensive literature search was carried out on the following:
• The geology ofthe study area;
• The hydrogeology of the study area, which included water levels;
• The application of environmental isotopes and hydro-chemistry in hydrogeological aquifer
conceptua lisation.
The major findings of the literature study are incorporated in the text, where appropriate,
throughout this report.
2.2 Previous work undertaken in the Study Area
Harmony Gold Mining Evander and Sasol Coal Mining Secunda operations respectively, both
have underground mining operations as well as surface infrastructure in the study area. Both
have in the past affected the environment and in future will affect the environment through
underground as well as surface mining activities.
Over the past two decades, a number of environmental as well as mining related studies have
been done on both mines. These studies have been done to assess and quantify the impact that
historical, current, and future mining activities have had or will have on the environment and
more specifically the groundwater environment of the study area.
2.3 Harmony Gold Mining Evander Operations
The Winkelhaak mines were formed in 1955, being the first gold-mining company in the Evander
Basin. The first gold was poured in 1958. In 1959, Bracken and Leslie mines were incorporated,
and poured their first gold in 1962. Kinross mine was formed in 1966, pouring its first gold the
Groundwater interaction between a deep coal mine and a deeper lying gold mine 10
Chapter Two: Literature Study 12010
following year. In 1996, the four mines were combined to form Evander Gold Mines and in 1997
were bought by Harmony Gold Mining Company Products (Harmony Evander Servitudes, 2008).
Previous studies done on Harmony Gold Mining Evander operations and used in this study are:
• [GCS] Groundwater Consulting Services (2002). Harmony gold mining company Evander
operations water quality status report. Project nr 2002-11-338.
• [GCS] Groundwater Consulting Services (1998). Evander goldmines hydrogeological
investigation final report, Johnstone, A. C. and Jansen H.C.
• Harmony Gold Mining Company (1994). "Kinross Mines Limited, Environmental
Management Programme Report."
• Harmony Gold Mining Company (1994). "Leslie Gold Mines Limited, Environmental
Management Programme Report."
• McKnight Geotechnical Consulting, C. Oliver, D. and De Bruyn, I. (2002). Evander Mine
Poplar Gold Project, Mpumalanga. A Review of the Geological Structural Model and
Assessment of possible Groundwater Ingress for Mining of Kimberley Reef. Internal Report
prepared for Rison Groundwater Consulting.
• Pritchard-Davies, E.W.D. (1962). Bulletin No. 77. "Sub- Karoo water table." Addendums 1-7.
• Rison Groundwater Consulting (2002). Harmony Gold Mining Limited- Evander operations:
Hydrogeological assessment of the potential risk of groundwater inflow into the Rolspruit
Project. Rison Groundwater Consulting, M.
• Rison Groundwater Consulting (2007). Harmony Gold Mining Limited - Evander operations:
Hydrogeological assessment of the groundwater table in Evander no. 6 shaft. Rison
Groundwater Consulting, M.
• Rison Groundwater Consulting (2003). Harmony Gold Mining Limited - Evander operations:
Hydrogeological assessment of the potential risk of groundwater inflow into the Poplar
Project. Rison Groundwater Consulting, M.
• Tweedie, E.B. (1986): The Evander Goldfield. In: Mineral Deposits of Southern Africa 1,
Anhaeusser C.R. & Maske, S. (eds.). Geological Society of South Africa, Johannesburg,
South Africa.
• Harmony Gold Mining Company (1988). "Winkelhaak Mines Limited, Environmental
Management Programme Report."
Groundwater interaction between a deep coal mine and a deeper lying gold mine 11
Chapter Two: Literature Study 12010
2.4 Harmony Gold Mining Evander: Geology
The following section is a summary of the geology as found by previous investigations,
especially the work done by Tweedie ef al. (1986) into the Evander Basin. Sediments of the
Witwatersrand Supergroup unconformably overlie the Archaean Basement. These
Witwatersrand sediments are then overlain by rocks of the Ventersdorp Supergroup and the
Transvaal and Karoo Superg roups (Tweedie ef el., 1986).
2.4.1 Witwatersrand Supergroup
The Witwatersrand Supergroup, as compiled from exploration boreholes done by Harmony, has
a total estimated thickness of approximately 1500 m in the Evander Basin (Winkelhaak Mine
EMPR, 1994). The Kimberley reef as described by Tweedie ef et., (1986) represents the distal
facies of a fluvial placer that was deposited by a system of braided streams which flowed down a
north-easterly dipping palaeoslope, as the sequence is divided into the basal West Rand Group
and the overlying Central Rand Group, which contains the auriferous gold deposits. The Central
Rand Group is subdivided into the lower Johannesburg Subgroup and the overlying Turfontein
Subgroup (Tweedie ef el., 1986). These formations consist predominantly of quartzite with
subordinate lava, shale and conglomerate. The Kimberley Shale separates the two units and is
marked by a basal unconformity. Immediately above this unconformity is a conglomerate layer
known as the Kimberley Reef. The primary focus for gold exploration in this area has been the
thinly developed Kimberley Reef conglomerate (Tweedie ef al., 1986). The Kimberley Reef
Horizon is situated at variable depths below surface. This variable depth is partly due to faulting,
but also due to sub-outcropping of the strata against the overlying Black Reef Formation. The
Witwatersrand stratigraphy overlying the Kimberley Reef has an average thickness of
approximately 261 m (Rison Groundwater Consulting, 2007). This becomes significantly less
towards the west (Rolspruit Project area), where the Kimberley Reef sub-outcrops against the
base of the Transvaal strata (Rison Groundwater Consulting, 2007).
2.4.2 Ventersdorp Supergroup
Overlying the Witwatersrand strata is a thick Supergroup of andesite (lava) of the Klipriviersberg
Group, Ventersdorp Supergroup. The Klipriviersberg Group thickness of 0 to approximately
1036 m thick is represented in the Evander Basin by a succession of superimposed flows of
greenish-grey, andesitic lavas (Tweedie ef et., 1986). Amygdaloidal, non-amygdaloidal and
porphyritic varieties are present, together with thin zones of greenish grey pyroclastic rocks.
Two well defined porphyritic units are present near the base of the Group. The contact between
the lavas of the Klipriviersberg Group and the underlying Witwatersrand sediments appears to
Groundwater interaction between a deep coal mine and a deeper lying gold mine 12
Chapter Two: Literature Study 12010
be conformable in the Evander Basin (Tweedie et st., 1986). The Supergroup, which consists of
numerous lava flows, attains a maximum thickness of approximately 1430 m between the Black
Reef unconformity and the Witwatersrand strata. The Ventersdorp lava package appears to be
thinning towards the southwest, where a minimum thickness of approximately 400 m was
recorded in exploration boreholes (Rison Groundwater Consulting, 2007).
The variable thickness of the Ventersdorp lava is partly attributed to the faulting in the area. It is
important to note that for the current mining area the Ventersdorp Supergroup mantles the
Witwatersrand Supergroup (Rison Groundwater Consulting, 2007). This is a very important
observation made from a groundwater point of view as it effectively means that the ventetsdorp
lavas fonn an impermeable barrier between the overlying (shallower) Karoo aquifelS and the
underlying (deeper) Witwatersrand aquifer.
2.4.3 Transvaal Supergroup
The Transvaal sediments overlie the Ventersdorp Supergroup and dip to the north and north-
east at 8° to 10°. The Transvaal Supergroup consists of a thin, basal quartzite and conglomerate
package belonging to the Black Reef Formation. The latter grades conformably upwards into
manganiferous, chert-poor dolomite of the Malmani Subgroup, Chuniespoort Group (Rison
Groundwater Consulting, 2007). According to McKnight Geotechnical Consulting Geotechnical
Consulting (2002), above the Chuniespoort Group is the Pretoria Group which consist of a basal
breccia-conglomerate unit known as the Bevets Conglomerate Formation and overlying the
ferruginous shale of the TimebalI Hill Formation (Rison Groundwater Consulting, 2007).
Overlying the shale is a volcanic unit of andesite that belongs to the Hekpoort Formation.
Several diabase sills intrude the Transvaal Supergroup. There are two main sills that formed
lensoid intrusions into the shale of the TimebalI Hill Formation. It is, however, unclear if these
features are continuous along dip and strike (Tweedie et al., 1986). The Transvaal Supergroup
rests unconformably on the Ventersdorp Supergroup and the strike orientation differs 90° to that
of the underlying strata (Rison Groundwater Consulting, 2007). During previous investigations of
the Rolspruit and Poplar Project areas situated to the north-northwest of the study area,
McKnight Geotechnical Consulting Geotechnical Consulting (2002) concluded that there are no
major displacements in the unconformity below the Black Reef Fonnation.
2.4.4 Karoo Supergroup
Directly overlying the Transvaal Supergroup are the glacial and deltaic sediments of the Karoo
Supergroup. The Supergroup is typically 150 m to 330 m thick (Rison Groundwater Consulting,
Groundwater interaction between a deep coal mine and a deeper lying gold mine 13
Chapter Two: Literature Study 12010
2007). An important feature that was highlighted by the McKnight Geotechnical Consulting
(2002) investigation is the presence of a paleo-glacial valley. The ice sheet of the Dwyka
glaciations was grounded in the study area and subsequently carved out broad "U" shaped
valleys. This event stripped out any old regolith developed on the underlying formation and in
particular it removed the karstification within the dolomite. The karstification forms the main
aquifer in dolomitic formations. Other than the paleo-glacial valley, the Karoo unconformity is not
disrupted by any major faults (Rison Groundwater Consulting, 2007).
In the study area, the sediments of the Vryheid Formation were invaded by doleritic magma that
formed thick sills and dykes. A prominent sill (known as the B4 sill) is present in the Evander
Basin, forming the general outcrop geology for the region. It is possible that the dykes and sills
within the dolomites are part of the same intrusion event, but there is not enough evidence to
suggest that these dykes are groundwater compartment forming such as is found in the West
and Far West Rand (Rison Groundwater Consulting, 2007).
2.4.5 Structural Geology
In assessing the groundwater occurrences during mining of the Kimberley Reef it is important to
understand the nature of faulting and in particular the age limits of such events. McKnight
Geotechnical Consulting (2002) concluded that both early compression tectonics and later
extensional tectonics affected both the Witwatersrand and Ventersdorp Supergroups in the
Evander Basin, but pre-dated the deposition of the Transvaal Supergroup (i.e. >2500 Ma). No
evidence of major post- Transvaal faulting was found that could potentially link the gold mine
underground workings with the overlying (shallower) Karoo aquifers (Rison Groundwater
Consulting,2007).
2.4.6 Faulting
Faulting is the most common form of deformation and both primary and secondary fault features
are recognised. The secondary faults are generally antithetic and throw down to the south.
They divide the gold mine into discrete structural blocks. Most common are south-downthrown
normal faults striking essentially northeast-southwest. Large losses of ground are associated
with this type of faulting (Tweedie et al., 1986).
2.4.7 Intrusives
The incidence and frequency of intrusives in the Witwatersrand sediments of the Evander
Goldfield are low (Tweedie eta/., 1986).
Groundwater interaction between a deep coal mine and a deeper lying gold mine 14
Chapter Two: Literature Study 12010
2.5 Harmony Gold Mining Evander: Hydrogeology
Previous hydrogeological studies undertaken by McKnight Geotechnical Consulting (2002),
Rison Groundwater Consulting: Rolspruit Project (2002) Rison Groundwater Consulting: Poplar
Project (2003) and Rison Groundwater Consulting No. 6 shaft (2007) in the Evander Basin
suggest that there are potentially two significant aquifers that may affect mining. These are as
described by Rison Groundwater Consulting (2007):
~ A perched aquifer situated in the weathered Karoo strata. This aquifer is restricted to a zone
of approximately 60 m below surface. This aquifer was classified as semi-confined, due to
the underlying impermeable base formations and is recharged by rainfall. The Karoo aquifer
is the main water supply to farmers in the region and the groundwater quality is generally
good. The current understanding of the hydrogeology of the region clearly indicates that this
aquifer poses no threat to the underground coal mine workings as hydraulic connections
between the aquifer and the coal mine are limited to selected dolerite dykes (Rison
Groundwater Consulting, 2007).
~ A deeper Witwatersrand aquifer. Gold mining in the region has shown that groundwater is
still encountered underground. This water may be derived from the overlying formations or it
could have been trapped in the sediments during deposition. The latter would be referred to
as connate water (Rison Groundwater Consulting, 2007).
Both aquifers are discussed and described by Rison Groundwater Consulting in Rison
Groundwater Consulting: Rolspruit Project (2002) Rison Groundwater Consulting: Poplar Project
(2003) and Rison Groundwater Consulting NO.6 shaft (2007). It should be noted that no
reference is made to the dolomite aquifer. This is due to the geological investigation done by
McKnight Geotechnical Consulting (2002) at the Rolspruit and Poplar Project areas, which has
shown that no significant dolomite aquifer has been preserved. The possibility of encountering
large volumes of dolomite water is therefore considered negligible. The McKnight Geotechnical
Consulting (2002) investigation has also shown that the Karoo aquifer is not connected to the
gold mine workings (Rison Groundwater Consulting, 2007).
2.5.1 Karoo Aquifer
Sasol undertook a very detailed hydrogeological investigation during its EMPR studies for the
Middelbult Colliery Block 8 expansion. During this study, 30 hydrogeological boreholes were
drilled and a detailed hydro census of some 170 privately owned boreholes was completed
(Rison Groundwater Consulting, 2007).
Groundwater interaction between a deep coal mine and a deeper lying gold mine 15
Chapter Two: Literature Study 12010
The drilling confirmed the presence of at least two phases of dolerite intrusions. The oldest
intrusive is typically a massive (B4) sill mostly restricted to the surface with a maximum
thickness of 49 m Jasper Muller Associates cc (2002) . A second intrusion is a porphyritic
dolerite (B8) sill that ranges in thickness from very thin to a maximum thickness of 18 m. This
dolerite sill features near vertical dykes, where it transfers from one horizontal plane to another.
This resulted in extensive geological/hydrogeological compartmentalisation Jasper Muller
Associates cc (2002) . Twenty-three dolerite intersections were recorded in twenty of the newly
drilled hydrogeological boreholes. Thirteen water strikes, associated with the host rock contacts
as well as the contacts between the weathered and fresh dolerite, were recorded along these
intersections. These water intersections constitute the two aquifers typically encountered within
the Karoo strata. The first aquifer is classified as a shallow, weathered aquifer in which the
primary water intersections are found on the weathered - fresh bedrock interface. The second
aquifer is classified as a deeper, fractured aquifer in which fracture flow dominates. The contacts
between dolerite intrusions and the host rock are the primary target for the drilling area in this
deeper aquifer. The two aquifers mayor may not be hydraulically connected (Rison
Groundwater Consulting, 2007).
The groundwater levels in the shallow, weathered Karoo aquifer varied between artesian and
11.04 m below surface, with a mean of 3.9 m below surface. The groundwater levels in the
deeper, fractured Karoo aquifer vary between 0.26 m below surface and 73.86 m below surface,
with a mean of 14.42 m below surface (Rison Groundwater Consulting, 2007).
Rison Groundwater Consulting: Poplar Project (2003) stated that the Karoo aquifers have not
been affected by the past 50 years of gold mining in the region, although large volumes of water
have been pumped from the underlying gold mines. This serves as a good indication of the poor
hydraulic connectivity of the Karoo aquifers to the gold mine workings (Rison Groundwater
Consulting, 2003).
2.5.2 WitwatersrandAquifer
Despite the deep geological setting of up to 2.5 km, groundwater is encountered in Harmony
Evander's gold mine workings. The McKnight Geotechnical Consulting (2002) report highlighted
the potential reasons for this.
);> McKnight Geotechnical Consulting (2002) established that the spacing of the exploration
boreholes and the subsequent geological interpretation do not allow for the detection of
smaller, potentially younger faults and joints. The presence of joints and fractures is
Groundwater interaction between a deep coal mine and a deeper lying gold mine 16
Chapter Two: Literature Study 12010
important in the sense that due to the fact that no displacement is associated with them they
go undetected, but they may still be water-bearing. If such features are present they may tap
water from the overlying Karoo aquifer. If they were major fractures they would have caused
the subsequent dewatering of the overlying Karoo aquifer. McKnight Geotechnical
Consulting (2002) also concluded that there is no evidence of this anywhere in the study
area and hence they stated that there are no major fractures or faults connecting the gold
mine workings with the overfying Karoo aquifer (McKnight Geotechnical Consulting et aI.,
2002).
~ A second theory is that connate water may be trapped in the Witwatersrand sediments as is
the case in the Free State Goldfields (Rison Groundwater Consulting, 2007). Significant
groundwater intersections have been recorded in this type of environment and Oryx Mine
was at one stage dewatering at a rate of 60 mega litres per day (Rison Groundwater
Consulting,2007).
Post-Karoo tectonic events along fracture planes and igneous intrusions of post-Karoo age
could form conduits for leakage from the Karoo aquifer to the deep Witwatersrand aquifer. Rison
Groundwater Consulting (2007) concluded that the deep Witwatersrand aquifer recharge from
surface is a negligible quantity and that the Karoo aquifer is unaffected by any dewatering of the
deeper gold mining in the Witwatersrand aquifer. Rison Groundwater Consulting (2007) also
concluded that the deeper Witwatersrand aquifer can then be regarded as a confined aquifer
that is not recharged by rainfall andlor near surface aquifers (Rison Groundwater Consulting,
2007).
2.6 Harmony Gold Mining Evander: No. 6 shaft's water level
Evander No. 6 shaft is located on the eastern boundary of the mining area, Rison Groundwater
Consulting (2007) conclude that No. 6 shaft it is totally isolated in the sense that there are no
connections to the rest of the mining areas and the groundwater table, and has remained at a
constant water level for the past few years. This assumption made by Rison Groundwater
Consulting (2007) is based that there is no direct connection to other mining activities. Rison
Groundwater Consulting (2007) also made the assumption that the No. 6 shaft could be
hydraulically connected to other parts of the gold mine workings, but they state the fact that the
groundwater table does not fluctuate indicates that there is no effect from neighboring mining
activities (Rison Groundwater Consulting, 2007). This suggests that this portion of the mine void
either does not receive any recharge or the recharge is equal to the loss through seepage
(possibly horizontal flow to neighboring mine voids?). The groundwater elevation in the shaft is
Groundwater interaction between a deep coal mine and a deeper lying gold mine 17
Chapter Two: Literature Study 12010
at approximately 925 m above mean sea level or 707 m below surface (Rison Groundwater
Consulting, 2007). Based on the Middelbult Mine: Block 8 Expansion EMPR, 2003 (2003) the
coal seam varies in elevation between 1363 m.a.m.s.1 to 1577 m.a.m.s.1 or 55 m below surface
to 269 m below surface (Middelbult Mine: Block 8 Expansion EMPR, 2003), thus the water level
in NO.6 shaft has not reached the coal mining level as the minimum difference between the coal
mining and the water level is currently at 438 m.
Rison Groundwater Consulting (2007) stated that the groundwater levels have changed very
little since 1969. It is important to note that their interpolated groundwater level in the vicinity of
Evander No. 6 Shaft was in the region of 929 m.a.m.s.1 at the end of 1969. It is currently at an
elevation of 925 m.a.m.s.1. Based on this observation, Rison Groundwater Consulting (2007)
concluded that the groundwater level in Evander No. 6Shaft has remained static for the past 35
years (Rison Groundwater Consulting, 2007).
Rison Groundwater Consulting (2007) stated that the fact that the Evander No. 6 Shaft is
situated close to the eastern extremity of the Witwatersrand aquifer suggests that this
groundwater elevation represents the edge of the dewatering cone and that the shaft can in fact
be representative of the final water level within the basin.
The groundwater levels measured by Rison Groundwater Consulting: Poplar Project (2003) at
the Poplar Project, situated at the opposite western extremity of the Witwatersrand aquifer, show
water levels varying between 1400 m.a.m.s.1 and 1600 m.a.m.S.1 or between 232 m below
surface and 32 m below surface. This part of the basin has remained unaffected by mining and
these elevations roughly equate to the original water levels encountered at the start of mining in
the Evander area (1533 m.a.m.s.1 or 99 m below surface) as listed by Pritchard-Davies (1962).
This indicates that the dewatering at Evander has had no influence on the groundwater levels in
the Poplar area, suggesting some compartmentalisation within the Witwatersrand aquifer that
might be due to the major northeast-southwest trending faults between the two areas. These
faults form impermeable barriers and allows for selected parts of the basin to be dewatered
independently (Rison Groundwater Consulting, 2007).
Rison Groundwater Consulting (2007) were therefore of the opinion that the groundwater level in
Evander No. 6 Shaft is representative of the final groundwater table of the Witwatersrand aquifer
within the study area (Rison Groundwater Consulting, 2007).
Groundwater interaction between a deep coal mine and a deeper lying gold mine 18
Chapter Two: Literature Study 12010
2.7 Sasol Coal MiningSecunda
The development of new collieries at Secunda was commissioned in 1975. From 1975 to the
present, numerous mining and environmental studies have been conducted on the mining area.
Previous studies and data used in this study include the following:
• Grobbelaar, R. (2001). "Quantification and management of intermine flow in the western
Witbank Coalfield: Implication for mine water volumes and quality." Unpublished M.Sc
thesis, Institute for Groundwater Studies, University of the Free State.
• Jasper Muller Associates cc. Van der Berg, J. (2002). "Compilation of geology and
groundwater inputs for the Middelbult black 8 EMPR - SASOL coal volume I and II - text
and appendices." Report prepared for Sasol Mining (Pty) Limited.
• Middelbult Mine: Block 8 Addendum Volume 1. (2003). "Environmental Management
Programme Report for the Middelbult Mine: Block 8 Expansion." Prepared for Sasol
Mining (Pty) Limited. Oryx Environmental. nr.OE63.
• Vermeulen, P.D., Dennis I. (2009). "Numerical Decant Models for Sasol Mining Colleries,
Secunda." Institute for Groundwater Studies. University of the Free State.
2.8 Sasol Coal MiningSecunda:Geology
Jasper Muller Associates cc (2002) describe the Sasol Secunda Coal Mines as being located in
the Vryheid Formation of the Ecca Group, Karoo Supergroup. These rocks consist primarily of
sandstones, shales and coal beds and are extensively intruded by dolerites of Jurassic age.
The coal seams in the Highveld Coalfield are mainly flat to gently undulating, with a very gentle
regional dip to the south (Grobbelaar, 2001). On the northern and western margins of the
coalfield, however, coal seam topography and distribution are commonly controlled by pre-
Karoo topography. Steeper dips are encountered where seams are situated against pre-Karoo
hills. The dolerites occur both as sills and linear dyke structures that may extend over tens of
kilometres. A large sill-like dolerite structure covers most of the study area. Apart from the
displacement of the sedimentary succession caused by intrusive dolerite sills, the Karoo strata
are not significantly faulted. Small faults (of less than one metre displacement) are present, but
rare. Larger faults have not as yet been identified, except for an east-west striking graben
structure with a down-throw of 22 m bisecting the south of the southem boundary of the study
area. Locally, the general lithological profile, up to and including the deepest mine-able coal
seam, comprises of:
~ Soft overburden
~ Hard overburden
Groundwater interaction between a deep coal mine and a deeper lying gold mine 19
Chapter Two: Literature Study 12010
~ No. 5 coal seam
~ Inter burden
~ No. 4L coal seam
Dolerite dykes and sills also appear over the study area Jasper Muller Associates cc (2002) .
2.8.1 Dolente Intrusions
Jasper Muller Associates cc (2002) stated that the Karoo sediments were invaded by two
phases of post-Karoo dolerite intrusions. The oldest intrusive, namely the B4, being a fine to
medium crystalline dole rite and is typically a massive sill that is mostly restricted to the surface,
with a maximum thickness of approximately 49 m Jasper Muller Associates cc (2002). This sill
is eroded away in the lower lying areas. Locally, the B4 is not only surface bound, but
transgresses the coal seams in a trough-like fashion to effectively compartmentalise these
portions of the reserve on the mining horizon. The B6, is a porphyritic dolerite, usually 3 m thick,
and intersects the coal seams less frequently than the B8 dolerite. Out of 615 exploration
boreholes, only one intersection was noted Jasper Muller Associates cc (2002) . The B8 dolerite
(a fine grained porphyritic dolerite) intruded later than the B4, along semi-planar features, with
the result that it is mainly exposed as dykes, i.e. almost vertical intrusives. The B8 ranges in
thickness from very thin to a maximum of 18 m. The prominent, east-west striking dyke or sub-
vertical sill, separating most of the Block 8 reserve from Middelbult Colliery, can be seen to
range in thickness between 7 m and 15 m. The B8 sill dolerite, approximately 18m in thickness,
features near vertical off-shoots (dykes), where it transfers from one horizontal plane to another.
These features occur predominantly along the planes of transference Jasper Muller Associates
cc (2002) .
Jasper Muller Associates cc (2002) describes the B12 as a light grey, fine grained porphyritic
dolerite with large needle-like phenocrysts, roughly ranging in thickness between 0, 12 m and 0,
75 m. Twenty-three dolerite intersections were recorded in twenty of the newly drilled
geohydrological boreholes. Thirteen water strikes, associated with host rock contacts as well as
the contact between weathered and fresh dolerite, were recorded along these intersections
Jasper Muller Associates cc (2002) .
2.9 Sasol Coal Mining Secunda: Hydrogeology
Previous hydrogeological studies done by Jasper Muller Associates cc (2002) in the study area
suggest that there is only one aquifer that may affect coal mining.
Groundwater interaction between a deep coal mine and a deeper lying gold mine 20
Chapter Two: Literature Study 12010
This aquifer is situated in the weathered or fractured Karoo strata. This aquifer is restricted to a
zone of approximately 60 m below surface. This aquifer is classified as semi-confined, due to
the underlying impermeable base formations and is recharged by rainfall. The Karoo aquifer is
the main water supply to farmers in the region and the groundwater quality is generally good
Jasper Muller Associates cc (2002) .
According to Jasper Muller Associates cc (2002), the displacement of the coal seams caused by
dolerite intrusion is seen to range from no displacement to not much more than the coal seam
thickness itself. The Block 8 underground reserve is largely separated from the existing
Middelbult Colliery, compartmentalised and sub-compartmentalised by a ± 15 m thick B8
dolerite sill roughly southwest-northeast in orientation Jasper Muller Associates cc (2002) . The
sill underlies the Middelbult Colliery close to the floor of the Karoo sediments before bending
upwards to vertical again, transgressing the coal seams before surfacing. To both the east and
the west, this sill has numerous sub-vertical split-offs that join up and split off again to
compartmentalise and sub-compartmentalise the Block 8 reserve Jasper Muller Associates cc
(2002) .
The larger Middelbult Colliery compartment is of particular significance, as a sizable portion of
the proposed Block 8 mine layout falls within this compartment. Jasper Muller Associates cc
(2002) state that of significance is the fact that both the Leslie and the Winkelhaak slimes dams
are located in this area. Some existing high extraction mining panels occur as close as 20-50 m
to the west of the southwestern corner of the slimes dam and 80-100 m to the east of the
Winkelhaak slimes dam, as seen in Figure 6. Manifested impacts relating ~?the de-watering of
the shallow weathered zone aquifer(s) over some of Middelbult Colliery's high extraction panels
are likely and have already been observed by Jasper Muller Associates cc (2002) in some of the
drilled monitoring boreholes.
Groundwater interaction between a deep coal mine and a deeper lying gold mine 21
Chapter Two: Literature Study 12010
LEGEND
N INlnlelulal slmn dam
LeslIe shmes dam
NlnteractlOn are-a
N IoIlddelbutt boundarIeS
N Harmon} bounoary
N 1.IIdO"lbu~ UlG
Figure 6: High extraction mining panels occurring close to the Leslie and Winkelhaak slimes dams.
During the geohydrological investigation undertaken by Jasper Muller Associates cc (2002), a
total of 30 boreholes were drilled specifically for geohydrological purposes. Two boreholes were
drilled per site, one shallow borehole 30 m deep to investigate the shallow weathered zone
aquifer(s), and one deep borehole ranging in depth between 80 and 150 m, to investigate the
deep fractured aquifer(s). The data of these boreholes is available in Jasper Muller Associates
cc (RSA) (2002) Compilation of geology and groundwater inputs for the Middelbult block 8
EMPR - SASOL coal volume I and 11- text and appendices. Van der Berg, J.
The following observations made in the Jasper Muller Associates cc (2002) report are important:
~ The depth to water level observed, varied between 0 m and 74 m, with a mean of 6.16 m
(Figure 7).
~ The depth to water level observed in the newly drilled shallow weathered zone boreholes,
varied between 0.27 m and 11.04 m, with a mean of 4.77 m (Figure 8).
~ The depth to water level observed in the newly drilled deep Karoo aquifer boreholes, varied
between 0.26 m and 73.86 m, with a mean of 14.42 m (Figure 9).
Groundwater interaction between a deep coal mine and a deeper lying gold mine 22
Chapter Two: Literature Study 2010
Shallow Karoo aquifer (weathered) water levels
1:.:ount_l<5
Mo'IPIf"I:; 1/4K
:S10~" - 2 no....
Mll"llmlJTl 0
25thrt'nenhl~IOI). I.71fJ
Perc .. ntrl ... 1""'''0:1'.'')'' !-..MO
15tnp£o~enlll C(.l~,=-':ol'OoI
Mo.,.. mum-é 9
;15.,., ':1 Meen:;::! 27(, Hl 5 2~2
05,.. Cl &,~ma" I (''S1 lo" aoa
~f'Iovo-DM,n9 1""'""41 "y T_I
,Q"..Squared- (, 2..,121(1
P·velV9 -= 0710:;1
_, ,I
:~
Shallow Karoo aquifer wDter levels
Figure 7: Histogram of the shallow (drilled 0-30m) Karoo aquifer's water levels.
12 Deep Karoo (fractured) water levels
CI,.UIl", 1~
Mean .. 1.. 301
::'t09V" .lO \.I:..
M,,-.'ll...,'_ 0 ZIJ
10 2'5IhPercertlle [01) -2 tf2tJ
r h Percentile (Median) .. " 7£10
75thPl(oorcoM""fll.C()3 •• 10 1.
MdXlmum" 7.! 60
g">% Cl M •• " :: 2 ~ te .2~
95,", Cl Slome = 1:5320 (CJ.33002
Ant1,. .....(Y"I-1 )Art,"O f'lnrm ...ury I _I
A-SQullr....J- 2 072
P-V.IU9 .. 0 OOC
o
~ ~
Deep Karoo (fractured) water levels
Figure 8: Histogram of the deep (drilled 80-150m) Karoo aquifer's water levels.
Groundwater interaction between a deep coal mine and a deeper lying gold mine 23
Chapter Two: Literature Study 12010
Karoo (shallow+deep) water levels
ClllJlll =30
MI'M.'In= '" 072
<::101"'\1~ l"i R?f'i
25 Mi ....m.. ...-n-O
~npE-rCen:lle(a'). IIIIl.I"5
hP.reenll (Med••n)s. ~8~
75thP.rc","II~(a3) .. 7 BOe
Mo;.lmum - 7~ ec
20
95'% Cl ME!itn '"' 32371" ,.. ~
9"'"'WtCISIgml!l= 12 444IQ 21 0(18
AI I'Jt!tloUl ~o.:.ll"'J NUlHh.hll/ Tl:nIl
A-~quered = "::I Jtiv
F"ValUE>= 0 0000
10
u
o
o ~ ~ ~
K;unn (-.;h:.llow...-d""pl ""."It,.r ,,...,ph,
Figure 9: Histogram of the deep & shallow (drilled 0-150m) Karoo aquifer's water levels.
2.10 Stable Environmental Isotopes
A detailed discussion on the application of stable environmental isotopes in hydrogeology is
available in Verhagen (1991, 1997 and 2000), Gat (1996) and Fritz et al., (1980). A literature
survey was done of the following:
• Fritz, P., Fontes, J., C.H., Elsevier, A. (1980). "Handbook of Environmental Isotope
Geochemistry, Volume 1, The Terrestrial Environment." New York. USA. p. 21-29, 283-
321,473-495.
• Gat, J.R. (1996). "Oxygen and hydrogen isotopes in the hydrologic cycle." Unpublished
paper, Department of Environmental Sciences and Energy Research, Weizmann
Institute of Science, 76100 Rehovot, Israel.
• Verhagen, B. Th., Geyh, M.A., Frëhlich, K. and Wirth, K. (1991). "Isotope hydrological
methods for the quantitative evaluation of groundwater resources in arid and semi-arid
areas." Research Report, Fed. Ministry for Econ. Coop. of the Fed, Republic of
Germany.
• Verhagen, B. Th. (1997). "Use of environmental isotopes in the characterisation of the
Table Mountain Sandstone Aquifer, South Africa." Progress Report to the Intemational
Atomic Energy Agency, Research Contract 8983/RO/Regular Budget Fund.
• Verhagen, B. Th. (2000). "Environmental isotope hydrology: Principles and application to
the geohydrology of the Karoo Basin." Karoo Aquifer Handbook, WRC, in preparation.
Groundwater interaction between a deep coal mine and a deeper lying gold mine 24
Chapter Two: Literature Study 12010
• Verhagen, B.Th. and Butler, M.J. (2004). "Isotope hydrology - An African Experience."
Geoscience Africa, Extended Abstracts, Vol. 2. Geological Society of South Africa,
Johannesburg. p. 671-672.
• Weaver, J.M.C., Talma A.S. and Cavé L.C. (1999). "Geochemistry and Isotopes for
Resource Evaluation in the Fractured Rock Aquifers of the Table Mountain Group.
Report No. 481/1/99, Water Research Commission, Pretoria.
2.10.1 General applications of stable environmental isotopes in hydrogeology
Stable isotopes 180 and 2H are unique for the investigation of hydrogeological processes.
Isotopes act as naturally occurring tracers in groundwater (commonly referred to as
environmental tracers) which can provide valuable information to the hydrogeologist about
aquifer characteristics and groundwater flow paths that would otherwise be difficult, if not
impossible, to establish (Verhagen ef al, 1991).
2.10.2 Fractionation
Molecules differ in mass, as the isotopes of the elements making up molecules differ in mass.
These mass differences influence factors such as vapour pressure, diffusivity, etc. During phase
processes, such as evaporation, condensation and exchange reactions, the abundance ratios of
the isotopes of individual elements change, or undergo fractionation (Verhagen, 2000). Such
processes usually take place at the interface of different phases or compounds containing the
same elements, such as vapour in contact with a liquid, a liquid in contact with a solid, etc.
These changes in abundance can be traced through natural and other systems (Verhagen,
2000).
The unit fractionation factor a is defined as: a = RA / RB, where RA and RB are the abundance
ratios of the rare (heavier) isotope to the more abundant (light) isotope in phases A and B
respectively. a Is temperature dependent.
As isotopic abundances and changes in these abundances are generally small, it is customary
to express these changes as fractional differences 8 in per mille with reference to the value of a
reference standard:
o = [ (Rs I Rr) - 1] x 1000 (%0) (Verhagen, 2000).
where the measured ratios for the heavy to the light isotope are Rs for the sample and Rr for the
reference standard, respectively.
Groundwater interaction between a deep coal mine and a deeper lying gold mine 25
Chapter Two: Literature Study IZ010
2.10.3 Application of stable isotopes 180 and 2H
18 2 .
The changes of 0 and H concentrations along groundwater flow paths is an effective tool to
determine the altitude of groundwater recharge, estimation of mixing proportions of different
sources or component flows and the relationships between groundwater and surface water (Gat,
1996).
180 and 18 162H are present in water in isotopic abundances (or ratios, R) of about 0/ 0 = 0.2 %
and 2H/1H = 0.015 %. In various combinations, these isotopes make up water molecules,
principally of masses 18, 19 and 20 (Verhagen, 2000).
Stable isotope abundance is normally expressed in terms of a ratio to the abundant isotope of
the same element or as positive or negative deviations of these isotope ratios from a standard,
similar to that of the radioactive isotopes (Verhagen, 2000).
The most important physical process resulting in variations in isotopic compositions of natural
water is vapour-liquid fractionation process during evaporation and condensation (Weaver et al.,
1996). The basic principle is that enrichment of the lighter isotope C60) occurs in the vapour
during evaporation, as opposed to loss of the heavy isotope e80) from the vapour first during
condensation (Weaver et al., 1996). These small changes can be expressed as a fractional
deviation 8 from a standard called SMOW (standard mean ocean water).
Values of 8 can be diagnostic of water from different recharge altitudes. Water vapour rising
from the ocean is depleted in the heavy isotopes (Weaver et al., 1996). Vapour masses moving
inland are subject to equilibrium isotopic exchange processes, with the continued depletion in
heavy isotopes in vapour travelling inland as a result of rainout. Consequently, the stable
isotopic content of meteoric water lies on a regression line:
&2H= s &180 + d (Weaver et al., 1996).
(where s is the slope, and the intercept d on the rYH axis the deuterium excess).
The line with s = 8 and d = +10 is called the global meteoric water line (Figure 10). The 2H -
excess is controlled by the deuterium in the vapour source region and the slope by evaporation
during rainfall and seasonal variations in precipitation. The position of any pair of 82H and 8180
values on this line for rainwater worldwide will depend on local climatological conditions
(temperature, latitude, altitude, and rainfall amount effects).
Groundwater interaction between a deep coal mine and a deeper lying gold mine 26
Chapter Two: Literature Study 12010
o
C,....
Figure 10: Meteoric water lines (Adapted from Gat, 1996).
Surface water bodies, such as dams, lakes and pans are enriched in the heavy isotope content
during evaporation. Light isotopes are preferentially transferred to the vapour phase (Weaver ef
al., 1996). The resulting surficial layer enriched in the heavy isotope is then readily mixed into
the bulk of the water body through convective processes. This evolutionary enrichment
produces 82H and 8180 values which lie to the right of the meteoric water line, and plot on an
evaporation line of lesser slope s (usually 4 to 5) and lower d than the GMWL (Figure 10).
Groundwater derived through infiltration from such water bodies will carry this distinctive
evaporative isotopic signal (Verhagen, 1997).
Evapo-transpiration losses from soil and groundwater generally occur under isotopic equilibrium,
i.e. without fractionation. In semi-arid environments such as the Klein Karoo, isotopic values of
groundwater derived directly from rain will therefore lie on or close to the GMWL. However, it will
tend to be displaced to more negative (or "lighter") 8 values than the weighted mean in rain, due
to the process of rainfall selectivity (Verhagen, 1997). Major rainfall events, which produce
isotopically lighter precipitation i.e. with more negative 8 values, may contribute proportionately
Groundwater interaction between a deep coal mine and a deeper lying gold mine 27
Chapter Two: Literature Study 12010
more to recharge. It is possible to distinguish different processes of rainfall recharge by different
sta bie isotope ranges on the GMWL (Gat, 1996).
Groundwater interaction between a deep coal mine and a deeper lying gold mine 28
Chapter Three: Mining Activities and Methods 2010
Chapter Three: Mining Activities and
Methods
3.1 Sasol Secunda Coal Mining Complex
Sasol Secunda Middelbult Mine Block 8 extracts coal from the economically viable No. 4L coal
seam situated at varying depths (Figure 11) throughout the study area (Table 1). The
underground workings of the Secunda Mining Complex consist of various layouts and
configurations. These configurations are a function of mine planning and method, and can
broadly be classified under board-and-pillar (BP) and high extraction (HE) methods (Middelbult
Mine: Block 8 Expansion EMPR, 2003). Shaft entrances, haulages to active sections and
ventilation channels can all be classified as types of board-and-pillar sections (Middelbult Mine:
Block 8 Expansion EMPR, 2003).
In board-and-pillar mining, coal pillars are left as support, although they may be extracted at a
later stage, resulting in high extraction mining. Middelbult has a mining height of approximately
3.5 m from the mining roof to the mining floor (Middelbult Mine: Block 8 Expansion EMPR,
2003). The mining layout of the Middelbult Block 8 area is shown in Figure 11.
Table 1: No. 4L coal seam elevations throughout Middelbult Block 8.
No. 4L coal seam elevations (m.a.m.s.l)
Seam Minimum Maximum Average
Roof 1450 1521 1490
Floor 1450 1520 1487
Surface 1560 1695 1600
Groundwater interaction between a deep coal mine and a deeper lying gold mine 29
Chapter Three: MiningActivities and Methods I 2010
LEGEND
rUil dd e-Ibou lt s hatts
Iritera cti on a rea
rUlidde-lbult bouneartas
~
/SI Harmony boundary
rulidde-ll>ult Future- Mining
j\/ rUlidde-lbult UJG
Figure 11: Middelbult layout of current and future mining.
Groundwater interaction between a deep coal mine and a deeper lying gold mine 30
Chapter Three: Mining Activities and Methods 2010
3.2 Coal mining methods and their impact on groundwater quality and
quantity
The common characteristic of board-and-pillar and more specifically high extraction coal mining
methods is that the removal of the coal seam results in the caving of the overlying strata into the
mined void. This disruption of the overlying rock mass may have a significant effect on the
hydrogeology, especially where high extraction has been done, resulting in subsidence
(Vermeulen & Usher, 2006). High extraction is when the board-and-pillar development is
followed by pillar extraction using a coal-cutting machine.
For shallow mines «200-300 m), the collapse usually propagates to the surface and establishes
a direct connection between shallow groundwater aquifers and the underlying mine workings
(Vermeulen & Usher, 2006). Although the specific technique used for high extraction does not
significantly influence the extent of the impact on the groundwater regime.
Figure 12: Board..and-pillar mining followed by high extraction (Vermeulen & Usher, 2006).
Groundwater interaction between a deep coal mine and a deeper lying gold mine 31
Chapter Three: Mining Activities and Methods 2010
3.2.1 Board-and-pillar (BP) mining
Recharge into board-and-pillar mining compartments is generally lower when compared to high
extraction compartments, due to the roof being kept intact through pillar support. This will result
in slower fill rates, allowing the build-up of oxidation products during the filling period on reactive
coal particle surfaces. Oxidation of pyrite is kinetically driven and, although it decreases over
time if no flushing takes place, it can generate significant amounts of oxidation products
(Vermeulen & Usher, 2006).
The Neutralisation Potential (NP) of a BP compartment is lower in comparison with a HE
compartment, since it is only the NP in the exposed coal that can actively participate in the
neutralisation of acidity released from pyrite oxidation. The lower recharge rates mean that the
rate of alkalinity added to the underground workings from the recharge water is also slower. This
means that BP compartments are not suitable for active flushing scenarios during the closure
phase (Vermeulen & Usher, 2006). Fracturing of the coal by board-and pillar methods
immediately results in an increased reactive surface area, which can significantly increase the
mass of acidity produced by Acid Rock Drainage (ARD) processes (Vermeulen & Usher, 2006).
BP compartments often await high extraction, allowing oxidation products to build up over time.
BP compartments can be turned into compartment dams, which will allow for the inundation of
compartment reactive surfaces, as well as make storage space available for water
management/reticulation purposes (Vermeulen & Usher, 2006).
3.2.2 High extraction (HE) Mining
Recharge into HE compartments is generally higher when compared to BP compartments due
to the fracturing of the overlying geology during goafing. This allows for higher recharge rates
and hence faster filling rates, reducing the time for oxidation product build-up. Where goafing
results in higher rainfall recharge, the alkalinity mobilised in the overlying geology can contribute
to the overall NP of the HE compartment. Neutralisation Potential (NP) associated with overlying
rocks is introduced in the compartment due to goafing. This NP can significantly contribute to
the alkalinity already present in the coal, as well as that associated with the recharge water
(Vermeulen & Usher, 2006).
Conversely, due to the presence of Karoo Supergroup marine sedimentary rocks in the roof, one
would expect a higher pyrite content, which will in turn become active once goafing results in the
exposure of fresh pyrite surfaces to oxygen. HE compartments can be sealed off and rapidly
filled with water to reduce the potential rate of pyrite oxidation to insignificant levels. During
Groundwater interaction between a deep coal mine and a deeper lying gold mine 32
Chapter Three: Mining Activities and Methods 2010
filling, the NP associated with roof material will become available for the neutralisation of water
acidity in the compartment (Vermeulen & Usher, 2006).
Acid Base Accounting (ABA) and various leaching tests were performed on 20 samples, using
the Modified Sobek (Lawrence) Method Jasper Muller Associates cc (2002) . The following
conclusions were made by Jasper Muller Associates cc (2002) with regard to the overall acid
generating potential of the overburden (Figure 22) above the No.4L coal seam:
• Paste pH levels measured indicate the presence of either excess base or acid material in
the overburden for the current (in-situ) situation. None of the samples had paste (initial)
pH-levels of lower than 7.77. This is an indication of the excess base material present in
the overburden at this stage. Groundwater draining initially into the underground
workings will display the effects of this excess base material, in the form of elevated
alkalinity values Jasper Muller Associates cc (2002) .
• A total S % calculation usually gives an indication of the sum-total of all sulphur species
present in the rock. This figure might include an entire range of sulphate species, sulfide
species, and organic sulphur species, some of which are only partly or not oxidised at all.
The total S calculated for Middelbult Block 8 does not give an overestimation of the
material available for oxidation, since only th~ reactive components were measured. The
range in total S % of all of the lithologies is relatively big (0.001 %-2.271 %), with an
average value of 0.370 %. This is an indication of the heterogeneity in terms of pyrite
mineralisation and distribution in the different overburdens Jasper Muller Associates cc
(2002) .
• The Acid Generation Potential (AP) gives an indication of the gross potential for
acidification per volume material. The range in AP is between 0.031 kgft CaC03 and
70.969 kgft CaC03, with an average value of 11.573 kgft CaC03. A number of 5
samples (25 %) showed elevated values above the average Jasper Muller Associates cc
(2002) .
• The Neutralisation Potential (NP) gives an indication of the total base potential available
to neutralise acidification. The range in NP is between 5.5 kgft CaC03 and 62.5 kg/t
CaC03, with an average value of 18.3 kg/t CaC03. A number of 8 samples (40 %)
showed elevated values above the average. The average value for all the samples is
higher than the values recorded for the Acid Generation Potential (AP) Jasper Muller
Associates cc (2002) .
Groundwater interaction between a deep coal mine and a deeper lying gold mine 33
Chapter Three: Mining Activities and Methods 2010
• The Nett Neutralisation Potential (NNP) is the total of NP - AP. A positive value means
excess base potential, a negative value excess acid potential. The range in NNP is
between -58.5 kg/t CaC03 and 62.5 kg/t CaC03, with an average value of 6.8 kg/t
CaC03. Overall, a positive NNP is present. The very large range in NNP indicates the
heterogeneity in the different stratigraphic (geochemical) units Jasper Muller Associates
cc (2002) .
3.3 Water Compartments
Middelbult Block is subdivided into water compartments shown Figure 13. In general, the water
in smaller compartments is easier to manage. Larger compartments create the risk that, when
filled to the decant elevation; certain areas of the compartment could be above the water level
due to the coal-seam topography. This will result in the active production of acidity in those
areas. Smaller compartments reduce the risk of failure, since smaller water volumes are
retained (Vermeulen & Dennis, 2009).
Smaller compartments, located in areas of high recharge, produce smaller volumes of water to
be managed in order to prevent an excessive build-up of head. If a compartment needs to be
emptied for whatever reason, smaller compartments will have smaller volumes to manage. The
negative aspect of smaller compartments is that they will have a higher cost of sealing
implication compared to larger compartments, due to the need for more seals (Vermeulen &
Dennis, 2009).
3.3.1 No water storage zones
AMinsim W analysis was done to determine the effect of gold mining on the coal horizon where
a small middling (>100 m) exists (Harmony/Sasol Working Group, 2005). Particular attention
was given to the tensile effect above the old gold workings. Results of the analysis show a small
area above the sub-outcrop of the gold workings has a tensile effect that extends into the coal
workings as shown in Figure 13. This zone has small horizontal stress change that can be
expected on the coal horizon in areas where the middling to the gold workings is less than
100 m. In this zone no water should be stored to prevent any possible groundwater interaction
from the coal mine workings into the gold mine workings (Figure 13).
Groundwater interaction between a deep coal mine and a deeper lying gold mine 34
Chapter Three: Mining Activities and Methods I 2010
LEGEND
rll1iddelbult shafts
Interaction area
Middelbouit bounearles
~
Harmony boundary
No water storage zones
~
/V Futu rE corn pa rtm ents
l/\'" 1/ Current Compartments
/ \/
-:10000 2.000
Figure 13: Coal mine water compartments and no water storage zones.
Groundwater interaction between a deep coal mine and a deeper lying gold mine 35
Chapter Three: Mining Activities and Methods 2010
3.4 Harmony Evander Gold Mining Operations
Harmony Evander is a deep-level gold mine which currently and in future will extract gold from
the economically viable Kimberley reef situated at variable depths underground. It is important
to note that the gold mine has no opencast or shallow workings « 200 m below surface). The
underground workings of Harmony Evander gold mining operations consist of various layouts
and configurations. These configurations are a function of mine planning and method, and can
broadly be classified under conventional mechanised face and strike gully stoping methods with
haulages, cross-cuts and ore passes (Leslie Gold Mines (EMPR), 1994).
The different depths of the gold mine are expected to have different effects on the quality of the
ground water. The extent to which the mining affects the quality of ground water must be
determined according to the different mining phases, the current status of the current mining
area and the way in which water is managed within the mining area.
1
Groundwater interaction between a deep coal mine and a deeper lying gold mine 3
Chapter Three: Mining Activities and Methods I 2010
LEGEND
Gold mine sh-afts
tntera cti on a rea
rulidd~lbult boundarie-s
Harmony boundary
Gold rt.l1ining Area
-3000
Figure 14: Gold mining area with shafts and ventilation shafts.
Groundwater interaction between a deep coal mine and a deeper lying gold mine 37
Chapter Three: Mining Activities and Methods 2010
3.4.1 Gold Mining methods and their impact on groundwater
Gold mining operations in the study area take place at depths from approximately 250 to 2500 m
below surface, using conventional mechanised face and gully stoping methods; the
development is down the footwall of the Kimberly Reef (Winkelhaak Mine EMPR, 1988).
Stoping takes place on the economically viable Kimberly Reef, where ore from underground
workings is then transported by overhead trolley lines from underground to the Metallurgical
Plant situated on the surface (Winkelhaak Mine EMPR, 1988).
Figure 15: Gold mining stope from which ore containing gold is extracted from the Kimberley reef.
3.4.2 Causes and effects of gold mining methods
The gold mining area is at a far greater depth than the coal and therefore no subsidence is
envisaged (Winkelhaak Mine EMPR, 1988). It is expected that over time pressure from the
overlying rock will compress the stope or even close the stope completely as shown in Figure 15
where the effect can be seen on the wooden stacks supporting the overlying rock mass.
Acid Rock Drainage (ARD) usually happens after deep gold mines have been abandoned. The
reason is deep sub-surface mines, when operational will have to keep pumping the groundwater
out of the mine (Figure 16). Once abandoned, however, the pumping stops and the mine floods.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 38
Chapter Three: Mining Activities and Methods 2010
This flooding is the initial step of ARD. Acidity is generated when metal sulfides are oxidised
after being exposed to air and water. Bacteria and archaeade microbes compose the metal-ions
faster. These microbes are found naturally in the rock, but their numbers are usually low in deep
gold mines due to limited oxygen and water (Freeze & Cherry, 1979). However, once they are in
an environment with an abundance of water and oxygen, they flourish (Freeze & Cherry, 1979).
Figure 16: Groundwater flowing into the gold mine via exploration borehole in a gold mine shoots.
Sulfide-bearing minerals, of which pyrite (FeS2) is the most abundant, will react with water in an
oxidising environment to form sulphuric acid (H2S04), Detailed reactions from Freeze & Cherry
(1979).
Equation 1
(1) FeS2 + 7/2 O2+ H20 => Fe 2++ zso,« + 2H+
Equation 2
(2) Fe2++ 1/402 + H+ => Fe3++ 1/2 H20
Equation 3
(3) Fe3++ 3H20 => Fe(OH)3 (yellow boy) + 3H+
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 39
Chapter Three: Mining Activities and Methods 12010
Equation 4
(4) FeS2 +14Fe3+ + 8H20 => 15Fe2++ 2S0/- + 16H+
The geochemistry of the geological regime dictates the degree and significance of impact on the
environment. The potential to generate acid rock drainage is largely determined by the presence
of sulphide minerals. However, acid drainage will not occur if either the sulphides are non-
reactive or sufficient alkaline materials are present to neutralise the acid (Vermeulen & Usher,
2006).
Static acid base accounting tests were conducted on samples of dust on rocks obtained from
mined strata in order to provide a qualitative predication of the net acidity, a function of the
relative content of acid generating minerals (sulphides) and acid consuming or neutralising
minerals (carbonates) (Winkelhaak Gold Mines (EMPR), 1988).When acid generating capability
exceeds acid consuming capacity, the pH of the water decreases with the consequent
mobilisation of sulphates and heavy metals, resulting in acid rock drainage (Winkelhaak Mine
EMPR, 1988).
Acid base accounting yielded a marginal negative (-0.10 kg/ton CaC03) net neutralising
potential (NNP), indicating a slight potential for acidity over the short term. Rain would wash the
rocks clean over the medium and long term. It was found that the rock rubble is hard and does
not weather over the medium to long term (110 years) (Winkelhaak Mine EMPR, 1988).
3.4.3 Prevention of possible groundwater interaction between Harmony Evander gold
mining operations and Sasol Secunda Middelbult (Block 8)
The two mines have specific regulations in place to prevent groundwater interaction between the
mines. These regulations are listed below and in their respective Environmental Management
Programmes.
• Barrier pillars of 100 m are left between all adjacent mining sections, to prevent the
migration of groundwater to low-lying areas and associated mining and safety problems
(Middelbult Mine; Block 8 Expansion EMPR, 2003).
• No mining activities will take place directly underneath the gold mine tailings storage
facilities (Middelbult Mine; Block 8 Expansion EMPR, 2003).
Cl The minimum distance that high extraction activities will take place from the perimeter of
the gold mine slimes dams is given as 300 m (Middelbult Mine; Block 8 Expansion
EMPR,2003).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 40
Chapter Three: Mining Activities and Methods 2010
• Coal mining activities will keep a barrier pillar of 50 to 100 m from vertical gold mine
shafts (Middelbult Mine; Block 8 Expansion EMPR, 2003).
• All gold mine exploration boreholes will be sealed off with a non-reactive material to
prevent the drainage of poor quality, low pH groundwater from the coal mine to the gold
mine (Middelbult Mine; Block 8 Expansion EMPR, 2003).
• The gold mine shafts will be sealed using design and methods specified by a competent
professional Civil Engineer (Leslie Mine EMPR, 1994).
• Any other excavations caused by mining operations will be back filled or graded to a safe
condition and rainfall run-off control measure in these areas will be implemented (Leslie
Mine EMPR, 1994).
• No water storage zones - In these zones no water should be stored to prevent any
possible groundwater interaction from the coal mine workings into the gold mine
workings. (Sasol/Harmony Working Group, 2005).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 41
Chapter Four: Geology of the Study Area 2010
Chapter Four: Geology of the Study
Area
4.1 Introduction
An important aspect of this investigation is the geology of the study area (Figure 17). It is the
major contributor in defining the various aquifers present within the different geological
lithologies. Geological processes also 'create preferential flow paths such as cracks, fissures,
joints and faults which allow for the rapid movement of groundwater in the subsurface (Cook,
2003). As the study area is situated in the well known Evander Basin, a detailed description of
the geology of the Evander Basin is included.
In the major part of the Basin sediments of the Witwatersrand Supergroup unconformably
overlie the Archaean Basement. These Witwatersrand sediments are in turn overlain by rocks of
the Ventersdorp Supergroup, Transvaal and Karoo Supergroups as shown in Table 2.
Three ages of intrusives are present in the Evander Basin, namely Bushveld, Ventersdorp and
Karoo. These intrusives are present as both dykes and sills across the study area. The Evander
Basin has undergone extensive structural dislocation and faulting is the most common form of
deformation, with both primary and secondary fault-features being observed in the study area
(Tweedie et et., 1986).
The following section describes the geology of the study area. The geology is described
according to the various Supergroups encountered in the study area from surface (Karoo
Supergroup) to deepest (Archaean Basement). Most of the geological description was based on
exploration borehole logs provided by Harmony and by work done by Tweedie et et, (1986).
The position of the exploration boreholes are shown in Figure 18 with a typical borehole log
shown in Figure 19.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 42
Chapter Four: Geologyof the Study Area I 2010
(t)
LEGEND
NN Inte ractl on :a reaMiddelbult boundaries
N Harmony boundary
N Evander Basin
/'\./1 Major Fault system
D Karoo Supergroup
D Transvaal Supergroup
Ventersdorp Superqroup
Witw,ate rs ran d SLIpe rg rou p
Basement corn pi ex
Figure 17: Geological map of the Evander Basin.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 43
Chapter Four: Geology of the Study Area 2010
Table 2: Generalised Stratigraphy of the Study Area.
PlEISTONE TO SUPERGROUP SUBGROUP SOil AND
RECENT AllUVIUM
Precambrian Karoo Supergroup Ecca Group Sandstones, car- Unconform ably
ca. 2000 x 106 years 183 - 22 m bonaceous shales and
coal seams
Dwyka Group Tillite
Precambrian Transvaal Pretoria Group Argilliccous quartzites, Unconformably
ca. 2300 x 106 years Supergroup up to black shales, and
1100 m tillites, with inter-
calated pyroclastic lava
flows.
Magnesium limestone,
Olifants River with chert bands.
Group
Quartzites and
carbonaceous
Black Reef shales, with a Unconform ably
formation poorly developed
basal conglome-
rate.
Precambrian Ventersdorp Klipriviersberg Grey-green Conformably
ca 2600 x 106 years Supergroup Group pyroclastics and
up to 1324 m amygdaloidal
andesitie lavas
Precambrian Witwatersrand Central Rand Quartzites, Unconform ably
ca 3100 x 106 years Supergroup up Group amygdaloidal
to 1500 m lava, shales,
and conglomerates,
including the Kimberley
Reef
Greywackes,
West Rand Group quartzites, and
magnetic shales
Archaean Shists, metasediments, Unconformably
basement and gneissose granites
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 44
Chapter Four: Geology of the Study Area I 2010
®
BH 390 log shown in
Figure 19
LEGEND
.. Explcratinn cc::"h:::I,,;
NN Int,,;rscti:::n 5r"sMi'do=lbult ac un';;sri:;;,
N HsrmC!n1, ccuncsry
Figure 18: Exploration borehole positions within the study area.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 45
Chapter Three: Mining Activities and Methods I 2010
Borehole Log . BH 390 (1 of 3) &1e'1de Log . BH390 (2 0(3) Borehole Log· BH 390 (3 of 3)
plh[m[ locality- X: 7158.47 Y:2929616.26 Z: 1578.11 Ph[m[ lDcali~-X: 1158.41 Y:292961626 Z: 1518.36
~~[m[ locality- X: 7158.47 Y: 2929616.26 Z: 1578.11
lIIIb;r Qob;J ~ Qdq
410 llhlkqy ~
820 224.n-l!I8JIllA~
m·IID.$COOlIIERATf:
'IID.$- 8l3Sl0UNlTmE
·8l3Sl-1lI8.11 COOlIIERATf:
51 461 D.ll-IlI8.lIiOUNlTmE
871 ·1lI8.lIi-1!il.18COOlIIERATf:
·1!il.\l-1Ii1.9I0UNlTmE
~- 819!1'COOlIIERATf:
'8l9!I!-IIIlJlIF~U:
·1IIlJlI-912.870UNlTmE
102 512 ·912.87-9131lIF~U:922
410 820 12ll
Figure 19: Geological log for exploration BH 390 (0-1129m)_
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 46
Chapter Four: Geology of the Study Area 2010
4.2 Karoo Supergroup
The Karoo Supergroup is a relatively thick and extensive group of rocks that was deposited on
the land and in fresh water, almost continuously from the Late Carboniferous to late Triassic
age. The Karoo Supergroup hosts all the South African coal deposits. The coal deposits were
formed in the great Gondwana basin (Tweedie ef el., 1986).
The study area is located on the north-eastern limit of the Main Karoo Basin, in which the
sediments were not horizontally disposed but dip to the south-west at a very low angle (Tweedie
ef el., 1986).
The Karoo Supergroup, with glacial and deltaic sediments in some places, directly overlies the
Transvaal, Ventersdorp and Witwatersrand Supergroup with a range of 50 m to 300 m at an
average of 210 m across the study area using the interpolated exploration boreholes (Figure
21). The lithology comprises of sedimentary rocks of the Dwyka Formation and Vryheid
Formation of the Ecca Group formed by erosion during and before the deposition of the Dwyka
Group of the Karoo Supergroup. These formations were deposited on an undulating glaciations
period. These rocks primarily consist of sandstones, shales and coal beds and are extensively
intruded by dolerites of Jurassic age. Figure 20 shows the sedimentary rocks encountered in the
Karoo Supergroup within the study area.
Figure 20: Photographs of typical sedimentary rocks found in Karoo Supergroup of the study area.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 47
Chapter Four: Geology of the Study Area 2010
1Ct
LEGE 0
NN Interaction areaMiddelbult boundaries
"IV Harmony boundary
N Karoo hickness
1€OOO
Figure 21: Interpolated thickness of the Karoo Supergroup in the study area.
The stratigraphic column consists of typical products of the Dwyka continental glaciation
process, represented by deposits of diamictite, glacio-Iacustrine varved siltstone, pebbly
mudstones and fluvio-glacial gravels and conglomerates. Clastic sediments and coal overlie
these rocks. The sediments were deposited in shallow marine and fluvio-deltaic environments
and the coal accumulated as peat in swamps and marshes associated within these sedimentary
environments (Tweedie ef al., 1986).
In general, three phases of sedimentation are recognised in the sedimentary of the Karoo
Supergroup of the study area by Tweedie ef al., (1986):
• The lower delta-dominated phase;
• A fluvial dominated coal zone and an upper deltaic succession;
• The lower succession commonly comprises one or more upward coarsening cycles of dark
micaceous siltstone.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 48
Chapter Four: Geology of the Study Area 2010
The fluvial dominated coal zone is about 75 m thick. In the study area there are only two coal
seams of significance as compared to the five coal seams of the Witbank Coalfield Jasper
Muller Associates cc (2002) . The two coal seams of significance are the No. 4L coal seam and
the No. 5L coal seam. Currently only the No. 4L coal seam is economically viable and is mined
by Sasol Coal Mining Secunda (Pty) Limited (Middelbult Mine: Block 8 Expansion EMPR, 2003).
The coal seams are usually separated by course-grained sandstone, fining to siltstone or shale
at the top. Glauconite sandstone, indicative of transgressive marine periods, is present above
the No. 4 Upper Coal Seam and forms useful stratigraphic markers. The approximate thickness
of the mainly sandstone overburden above the No. 4L coal seam is shown in Figure 22.
Locally, the general lithological profile, up to and including the deepest mine-able coal seam,
comprises of:
• Soft overburden;
• Hard overburden;
• NO.5coal seam;
• Inter burden;
• NO.4Lcoal seam.
Dolerite dykes and sills also appear over the study area Jasper Muller Associates cc (2002).
The coal seams are mainly flat to gently undulating, with a very gentle regional dip to the south.
The No. 4L coal seam has a maximum thickness of 6.60 m and a minimum of 4.14 m with an
average value of 5.69 m Jasper Muller Associates cc (2002) .
Table 3: Thickness information of the different coal seams (adapted from JMA, 2002, Report No: JW9810218068).
Values in (m) No. SLcoal seam (m) No. 4L coal seam (m)
Minimum 0.08 4.14
Maximum 5.32 6.6
Average 1.38 5.69
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 49
Chapter Four: Geology of the Study Area 2010
:...
oe:
LEGEND
NN. Interaction areaMiddelbult boundaries
N Harmony boundary
1EOOO
Figure 22: Interpolated overburden thickness in m above the No. 4L coal seam roofofthe Middelbult Block 8 area.
Apart from displacement of the sedimentary succession caused by intrusive dolerite sills, the
Karoo strata are not significantly faulted. Small faults - <1 m displacement - are present, but
larger faults have not as yet been identified in the study area. Jasper Muller Associates cc
(2002) . Intrusives are represented by dolerite, which covers large areas of the study area
(Figure 34). The intrusives are in the form of sills and dykes. The sills are present over large
portions of the study area as shown in Figure 23 and Figure 24. The sills are both conformable
and transgressive with respect to the sedimentary rocks, and are of the non-porphyritic types.
The intrusives are discussed in detail in Section 3.8.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 50
Chapter Four: Geology of the Study Area 2010
N 5
• Karoo
• B4 Dolerite
• 4L Coal seam
• B8 Dolerile
o Cross Sec
Figure 23: North-south cross section illustrating the No. 4L coal seam and the B4 and 88 dolerite sills.
E
• 4L Coal seam
• B8 Dolerile
.. ~.- , . ,]...
Figure 24: West-east cross section illustrating the No 4L coal seam and the B4 and 88 dolerite sills.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 51
Chapter Four: Geology of the Study Area 2010
The cross sections are
shown in Figure 23 and
Figure 24
LEGEr'JD
! ! I
Figure 25: Surface view ofthe cross sections.
An important feature of the Karoo Supergroup is the presence of a paleo-glacial valley. This
event stripped out any old regolith developed on the underlying formation and in particular it
removed the karstification within the dolomite (Tweedie ef et., 1986). The latter forms the main
aquifer in dolomitic formations. Other than the paleo-glacial valley, the Karoo unconformity is not
disrupted by any majorfaults (Tweedie ef al., 1986).
4.2.1 Coal seam sedimentary processes
In general, coal seams formed during periods of low sedimentary deposition when plant growth
and peat accumulation occurs at the same rate as the substrate. Such periods separate
depositional events. At the study area, these depositional Supergroups took place in a deep,
rapidly subsiding portion of the main Karoo Basin (Falcon, 1986).
According to Falcon (1986) the coal-forming environments are subdivided into two types.
• The paraglacial and the;
• epicontinental environments.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 52
Chapter Four: Geology of the Study Area 2010
The paraglacial environment is a basin edge phenomenon that developed around the northwest
and northern side of the basin as a result of the accumulation of glacia-Iacustrine and glacio-
fluvial sediments on the marine-covered surfaces of the Kaapvaal Craton (Falcon, 1986). These
sediments were deposited at the close of the Dwyka continental glaciation. The distribution of
the coal seams in this environment was controlled by post-Karoo topography, with peat
accumulation predominantly occurring in low-lying land protected from active sedimentation.
These coals are freshwater marsh in origin, with characteristically low sulphur content (Falcon,
1986).
The epicontinental coal forming environment is related to the major development of sediments
across the north-eastern stable shelf area of the Karoo Basin, and overlies the older paraglacial
sediments to the north (Falcon, 1986). This environment was locally subjected to marine or
brackish water inundation during delta switching, resulting in a variety of geochemical (pH,
salinity and Eh) environments. Such factors affect coal quality as well as the resultant sulphur
content of the coal considerably.
Within each coal seam, a range of local conditions controls the development of the coal type
and facies, such as topographical controls, the local paleo-environment, type of plant
community, rate of accumulation and nature and rate of plant degradation, or biochemical
coalification (Falcon, 1986).
The conditions for the formation of different coal facies varies with paleo-ecological
environments. The in-situ inertinic coals are found predominately in relatively dry environments,
viz. on high ground of valley flanks, levees of rivers, upper delta, and alluvial plane swamps.
Drift forms occur in reworked organic detritus pieces usually transported by water and deposited
on river banks, or as flooded deposits in deltaic orlacustrine environments (Falcon, 1986).
The vitrinite rich coals found at Secunda are usually associated with permanently wet or water-
logged conditions, as may be found on sea shores, or in fluviatile, lacustrine, or lower
deltaic/paralic back swamps (Falcon, 1986). Shallow open water, distal environments, lagoons,
ponds, and distal deltaic settings typically accumulate very fine wind- and water-borne organic
and inorganic matter - particularly algae, spores and pollens (Falcon 1986). Known as exinite,
these components impart high hydrogen and volatile matter content to the host sediment.
4.2.2 Coal seam contours
The No.4L coal seam slopes downward from a roof elevation ranging between 1488 m.a.m.s.1
and 1510 m.a.m.s.1 along the northern and eastern bounds of the Block 8 reserve, to a floor
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 53
Chapter Four: Geology of the Study Area 2010
elevation ranging between 1460 m.a.m.s.1 and 1465 m.a.m.s.1 along the central southern
boundary between the Block 8 reserve and Middelbult Colliery as shown in Figure 26 Jasper
Muller Associates cc (2002) .
The No.4 seam slopes downwards from the central southern boundary between the Block 8
reserve and Middelbult Colliery to a floor elevation of 1450 m.a.m.s.1 in the far south-western
portion of the reserve on Leeuwpan (Figure 26). From here is a rise up to 1475 m.a.m.s.1 along
the central western extent of Middelbult Colliery, towards a high of 1520 m.a.m.s.1 a little further
to the south-southwest (Figure 26).
Seam Floor (m)
1~20.0
1~1 .0
1 ~OO.O
1'9 .
1'80.0
1 70.0
1'60.0
LEGEND
No Inte-raction are-a
~ Midde-ll>ult boundaries
Figure 26: Interpolated elevations of the No. 4L coal seam floor in m.a.m.s.l.
4.3 Transvaal Supergroup
The Transvaal Supergroup is mostly present in the northern portions of the study area and has
been intersected both in surface boreholes and in gold mine shafts (Leslie Mine EMPR, 1994).
In the south-southeast of the study area, the sediments have been removed by pre-Karoo
erosion, as shown in Figure 27 and Figure 28. The sequence consists of a thin, basal quartzite
and conglomerate package belonging to the Black Reef Formation. The latter grades
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 54
Chapter Four: Geology of the Study Area 2010
conformably upwards into the manganiferous, chert-poor dolomite Malmani Subgroup of the
Chuniespoort Group (Tweedie ef et., 1986). Above the Chuniespoort Group is the Pretoria
Group, which consists of a basal breccia-conglomerate unit known as the Bevets Conglomerate
Formation, overlying ferruginous shale of the TimebalI Hill Formation. Overlying the shale is a
volcanic unit of andesite that belongs to the Hekpoort Formation (Tweedie ef et., 1986).
Several diabase sills intruded the Transvaal Supergroup. There are two main sills that formed
lensoid intrusions into the shale of the TimebalI Hill Formation. It is, however, unclear if these
features are continuous along dip and strike. The Transvaal Supergroup rests unconformably on
the Ventersdorp Supergroup and the strike orientation differs by 90° to that of the underlying
strata (Tweedie ef el., 1986).
w--------------------------------------------------------- E
- Faults
- Gold mine shaflsl"'ork~lQ~
Trersvaal
• Ven:ersdorp
w... te",.nd
o
Figure 27: Thinning and absence of the Transvaal SUpergroup towards the east-southeast of the study area.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 55
Chapter Four: Geology of the Study Area 2010
20
N ------------------------------------------------- S
10
5 seccer-ce to '~ard5 t"'tes:JLo:1t'oI f v-e
sn.d)' ,vea
Legend
• ".,. Dolc"looill - F01.l1b
• Karoo __ Gold mine .h.rt.lwol'1dng
• No4lc;:l.I ••• m
• ee DO.llw.iI
Transv •• 1
• V.m.rlldorp
Wltwfllt.,..rttnd
o 16500
Figure 28: Thinning of the Transvaal Supergroup north ID south across the study area.
4.3.1 Black Reef Fonnation
The Black Reef Formation is poorly developed in the study area and compromises a composite
sequence of quartzites and intercalated carbonaceaus shale with a basal conglomerate
(Tweedie ef al., 1986). The sediments, which attain a maximum thickness of 24 m, dip to the
north and northeast at 10° and appear to unconformably overlie the andesitic lavas of the
Ventersdorp Supergroup. Because of its northerly dip, the Black Reef Formation has been
truncated by the Karoo Supergroup and is only developed over the northeastern portion of the
basin (Tweedie et ei, 1986).
The stratigraphic thickness of the formation is highly variable, suggesting a highly dissected
Ventersdorp topography. As most of the faulting is pre-Transvaal in age, a certain amount of
relief in the Ventersdorp topography should be expected (Tweedie ef al., 1986).
4.3.2 Olifants River Group
4.3.2.1 Malmani Dolomite
The Malmani Dolomite is the only member of the Olifants River Group found in the study area
and dips to the north and northeast portion of the basin. The dolomite is generally massive, very
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 56
Chapter Four: Geology of the Study Area 2010
fine grained and varies in calor from light grey to medium-dark grey (Tweedie et et., 1986). A
number of Karoo-age dykes and sills are present in the upper dolomite. Most of the major
faulting is pre-Transvaal in age and therefore most of the major faults carry relatively little water
(Tweedie et aI., 1986).
4.3.3 Pretoria Group
The Pretoria Group sediments are northerly dipping and underlie the northern portions of the
study area (Tweedie et sl., 1986). There are three marker horizons that can be recognised:
• Bevel's Conglomerate,
• Daspoort Tillite,
• Ongeluk Volcanics.
The sediments of the Pretoria Group, as observed, are characterised by the lack of coarse-
clastic rock types and comprise a sequence of interbedded, light grey, banded siltstones, shales
and very occasional, micaceous, friable quartzites (Tweedie et ei, 1986).
4.4 Ventersdorp Supergroup
Overlying the Witwatersrand strata is a thick sequence of andesite (lava) of the Klipriviersberg
Group (Tweedie et el.. 1986).
4.4.1 The KlipriviersbergGroup
The Klipriviersberg Group of up to approximately 1324 m (Figure 29) thick is represented in the
study area by a succession of superimposed flows of greenish-grey, andesitic lavas (Tweedie et
et, 1986). Amygdaloidal, non-amygdaloidal, and porphyritic varieties are present, together with
thin zones of greenish-grey pyroclastic rocks (Tweedie et sl., 1986). Two well-defined
porphyritic units, developed near the base of the Group, constitute good marker horizons. The
contact between the lavas of the Klipriviersberg Group and the underlying Witwatersrand
sediments appear to be conformable in the study area. The variable thickness of the
Ventersdorp lava is partly attributed to the faulting in the area, but it is important to note that for
most of the study area the Ventersdorp Supergroup mantles the Witwatersrand Supergroup
(Tweedie et et; 1986). The exception is in the east-southeast of the study area where the Group
is very thin and in some areas sub-outcrops against the Karoo Supergroup as shown in Figure
30. In general, the Ventersdorp Supergroup thins from north to south in the study area, as
shown in Figure 31.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 57
Chapter Four: Geology of the Study Area 2010
ot::
LEGEND
NN. Interaction areaMiddelbult boundaries
lEOOO
Figure 29: Interpolated thickness of the Ventersdorp Supergroup.
\V E
-h nn ~Q Jf r-e
ventersccrc !i' 8S
t~,·.a'os tr-e east-
• e::k.. • _:",
S)~theBS: ~f the St~O,
KI':Q _ 300d~ .A.~";
• ~': ..~)I.t
13 39
Figure 30: Thinning ofthe Kliprivierberg Group lavas towards the east-southeast ofthe study area.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 58
Chapter Four: Geology of the Study Area 2010
2000
N s
CIl
E
~ 10.
E
5
Legend
• ?~Ot Ir.. sj - =.",tI;
KJf'OO _ Gc'r"-'l~,ftI
• ~oJ.~ .... ~
• Seo. MOoi
o
Figure 31: Thinning of the Kliprivierberg Group lavas towards the south of the study area.
4.5 WitwatersrandSupergroup
The Witwatersrand Supergroup is divided into the basal West Rand Group and the overlying
Central Rand Group, which contains the auriferous gold deposits (Tweedie ef aI., 1986). The
Central Rand Group is subdivided into the lower Johannesburg Subgroup and the overlying
Turfontein Subgroup. These formations predominantly consist of quartzite with subordinate lava,
shale and conglomerate. The Kimberley Shale separates the two units and is marked by a basal
unconformity. Immediately above this unconformity is a conglomerate layer known as the
Kimberley Reef (Tweedie ef et., 1986).
Although the sediments of the West rand Group in the study area are poorly documented, the
Jeppestown Subgroup is deemed to be missing due to the absence of the Crown Lava
Formation and the polymitic greywacke development in places at the top of the Government
Subgroup which appears to be very similar in lithological composition and stratigraphic position
to the "Blue Grit" of the Heidelberg-Balfour area (Tweedie ef al., 1986).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 59
Chapter Four: Geology of the Study Area 2010
The Central Rand Group in the Evander Goldfield comprises a composite sequence of
interbedded conglomerate and non-conglomerate quartzites that averages 650 m in thickness
(Tweedie ef al., 1986).
The Witwatersrand Supergroup is very thin by comparison with the Central Rand, a thinning of
approximately 3:1 having taken place in the Evander Basin (Tweedie ef et, 1986). Interbedded
within this sedimentary pile are two andesitic lava units, and the Kimberley shale is the only
argillite present. Many of the units occurring below the Kimberley Reef are lithologically very
similar to those occurring at a similar stratigraphic position in the East Rand Basin. However,
above the Kimberley Reef, sediments from the Evander Basin bear no resemblance to their
counterparts in the East Rand Basin. This is possibly related to their respective transport
directions and provenance (Tweedie ef et., 1986).
4.5.1 Hospital Hill Subgroup
The Hospital Hill Subgroup is characterised by interbedded shales and quartzites. Shales are
inter-grained, faintly to strongly banded and vary in color from greenish-black to maroon.
Magnetic and non-magnetic varieties are present. By contrast, the quartzites are generally sub-
siliceous, with fuchsite partings (Tweedie ef el.. 1986).
4.5.2 Government Subgroup
The Government Subgroup has interbedded sequences of shales, siltstones, quartzites and
polymitic greywacke. Shales are similar to those of the Hospital Hill Subgroup, but are generally
less magnetic and seldom exhibit maroon colouration. The quartzites are highly variable in
composition and are more argillaceous than their Hospital Hill counterparts (Tweedie ef et.,
1986).
4.5.3 Turffontein Subgroup
The Turffontein Subgroup consists predominantly of quartzite with subordinate lava, shale and
conglomerate. The Turffontein Subgroup also contains the thinly developed Kimberley Reef
conglomerate, which is the primary focus for gold exploration in the area (Tweedie ef et, 1986
4.5.4 Johannesburg Subgroup
The Johannesburg Subgroup consists predominantly of quartzite with subordinate lava, shale
and conglomerate (Tweedie ef al., 1986).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 60
Chapter Four: Geology of the Study Area 2010
4.6 Basement Complex
The Complex consists mainly of Basement Schists and Basement Granites (Tweedie et el.,
1986). The Basement Complex was not considered in this study area as it has a great depth
below the Witwatersrand and no data was available for interpretation.
4.7 StructuralGeology
The Evander Basin appears to be arcuate in shape. Minor folding occurs along the eastern and
western limbs of the basin. Folding in conjunction with faulting has resulted in the local
overturning of the Witwatersrand sediments in some places in the study area (Tweedie et el.,
1986). Peripheral faults are encountered across the study area, together with an associated
horst, graben and north-westerly plunging syncline to the south of the study area. The degree of
faulting and structural dislocation within the study area is extremely high (Tweedie et al., 1986).
A number of different fault types have been recognised by Tweedie et al. (1986) and these can
be broadly classified as being either primary or secondary. The former are presented by the
western and southern peripheral fault zones. The secondary faults comprise mostly normal
faults, mainly throwing down to the south-east. Early compression tectonics and later
extensional tectonics affected both the Witwatersrand and Ventersdorp Supergroups in the
Evander Basin, but pre-dated the deposition of the Transvaal Supergroup and hence the Karoo
Supergroup (i.e. >2500 Ma) (Tweedie et al., 1986)._The major faults covering the study area are
shown in Figure 32.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 61
Chapter Four: Geology of the Study Area 2010
LEGEND
NN In eraction areaMiddelbult boundaries
N Harmony boundary
~ Major fault system
16000
Figure 32: Major faults in the study area.
4.8 Intrusives
According to Tweedie et al., (1986) the incidence and frequency of intrusives in the
Witwatersrand sediments are low. This can be attributed to the following:
• Most of the faulting is of Ventersdorp age;
• A thick wedge of Ventersdorp lavas overlies much of the Witwatersrand Supergroup within
the Evander Basin;
• The proximity of the Bushveld Complex;
• The degree of structural dislocation exhibited by the Witwatersrand sediments;
• The abundance of dolerite sills of Karoo age.
Three different ages of intrusives are recognised in the Evander Basin (Tweedie et al., 1986):
• Karoo age intrusives;
• Ventersdorp age intrusives;
• Bushveld age intrusives.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 62
Chapter Four: Geology of the Study Area 2010
4.8.1 Karoo age intrusives
The Karoo Supergroup sediments were invaded by two phases of post-Karoo dolerite intrusions.
The oldest intrusive, namely the 84, a fine to medium crystalline dolerite, is typically a massive
sill, mostly restricted to the surface, with a thickness range of 49 m to 0 m at an average of 12 m
as shown in Figure 33. This sill is eroded in the lower-lying areas. The overburden thickness
covering the 84 dolerite sill is shown in Figure 33.
LEGEND
NN. lnterac ion areaMIddelbuit boundaries
N Harmony boundary
16000
Figure 33: Interpolated thickness of the B4 dolerite sill covering the study area.
Locally, the 84 is not only surface-bound, but also transgresses the coal seams in a trough-like
fashion to effectively compartmentalise these portions of the reserve on the mining horizon
Jasper Muller Associates cc (2002) . Refer to Figure 23 and Figure 24.
The B8 dolerite (a fine-grained porphyritic dole rite) intruded later than the B4, along semi-planar
features, with the result that it is mainly exposed as dykes, i.e. almost vertically intrusive. The B8
ranges in thickness from very thin to a maximum of 18 m. The prominent, east-west striking
dyke or sub-vertical sill, separating most of the Block 8 reserve from Middelbult Colliery, can be
seen to range in thickness between 7 m and 15 m Jasper Muller Associates cc (2002) . The
dolerites dykes across the study area are shown in Figure 34.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 63
Chapter Four: Geology ofthe Study Area 2010
The B8 sill dolerite, approximately 18 m in thickness, features near vertical off-shoots (dykes),
where it transfers from one horizontal plane to another. These features occur predominantly
along the planes of transference. This phenomenon results in extensive geological/
geohydrological compartmentalisation (Figure 23 and Figure 24).
LEGEND
NN In eraction areaMiddelbult boundaries
N Harmony boundary
/"\/ MIddelbuit OoIerite Dykes
16000
Figure 34: Surface view ofthe dolerite dykes in the study area.
4.9 Conceptual Model of Geology
A conceptual model of the geology was constructed using data from previous studies as well as
data from the 195 exploration borehole logs received from Harmony Gold Mining (Ltd) Evander
operations. The borehole logs were gathered in one Excel file and refined to be compatible with
WISH (Windows Interpretation System for Hydrogeologists). WISH interpolates the Excel data
by the distance weighting method to give an indication of the sub-surface geology across the
study area. Through the use of cross sections the conceptual model is displayed in Figure 35.
The cross sections were drawn to dissect the gold mine shafts were possible. The WISH
program does not allow for the interpolation of faults and therefore they were drawn in by hand.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 64
North West
CIl
E
co
E
Legend
• B4 Dolente si. - Faults 11623
• Ko"" _ Gold rrlne shlftslWOrtll\g'
• NO.4_coal saam
• B8 Dolente s
Transvaal
• Ventersdorp
Wrtwatersrand
o lOSS SedlOn 20146
Figure 35: Conceptual model of the geology of the study area.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 65
Chapter Five: Hydrogeology of the Study Area 2010
Chapter Five: Hydrogeology of the
Study Area
5.1 Introduction
Sasol Middelbult Mine (Block 8) overlies Harmony Evander gold mining operations. There are
different geological formations of different ages as discussed in Chapter 0 dividing the two
mines from each other. In some of these geological formations (e.g. Karoo Supergroup), aquifer
systems have developed due to their geological structure while other geological formations (e.g.
Ventersdorp Supergroup) have resulted in aquiclude or impermeable layers.
Currently, there are seven different aquifer types within the study area:
• shallow perched Karoo aquifers (unconfined);
• shallow weathered zone Karoo aquifers (semi-confined);
• deep Karoo fractured aquifers (zone below the weathered zone & confined);
• coal mine artificial aquifer (confined);
• Transvaal dolomitic aquifer (confined);
• gold mine artificial aquifer (confined);
• deep Witwatersrand aquifer (confined).
Conceptually, the aquifers are displayed in Figure 36. Over the medium to long term, as mining
continues, it is expected that the artificial aquifers created by coal and coal mining will expand
and will cover a much larger area than currently exists.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 66
Chapter Five: Hydrogeology of the Study Area 2010
Perc!'td~u~er
aquifer
system (+I. 200 ml
en eMCfP lD,aI (impea uicludes
Witwatersrand
aquifer system
Figure 36: Conceptual model of the different aquifers in the study area (not to scale).
5.2 Aquifer Types
5.2.1 Shallow perched aquifer (unconfined)
The aquifer is characteristic of clay, colluvium alluvium and weathered sandstone. Within this
unconfined (water table) aquifer, perched groundwater conditions often occur. These shallow
perched aquifers are essentially restricted to the soil (soft overburden) horizon within the study
area Jasper Muller Associates cc (2002) . The hydraulic conductivity value for the aquifer is
estimated at 1x10-5 mld to 0.10 mld with an average of 0.02 mld (Vermeulen & Dennis, 2009).
5.2.2 Shallow weathered zone Karoo aquifers (confined)
This aquifer comprises of white arenaceous sandstones located below the dolerite sheet but
above the coal seam horizon. The Ecca sediments are weathered to depths between 5 to 12 m
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 67
Chapter Five: Hydrogeology of the Study Area 2010
below surface throughout the area Jasper Muller Associates cc (2002) . The upper aquifer is
associated with this weathered zone and water is often found within a few metres of the surface.
The hydraulic conductivity value for the aquifer is estimated at 1x10-6 mid to 5x10-4 mid with an
average of 4x10-4 mld (Vermeulen & Dennis, 2009).
5.2.2.1 Recharge into the shallow weathered Karoo aquifer
The main source of recharge into the shallow aquifer is rainfall that infiltrates the aquifer through
the overlying unsaturated zone. Rainfall that manifests as surface run-off and drains to streams
may also subsequently enter the shallow aquifer by infiltrating the stream bed (Grobbelaar,
2001). Water impoundments and features such as tailings dams may constitute additional
recharge sources in certain areas.
The rainfall ultimately recharging the shallow aquifer is estimated at 3-5 % (Vermeulen &
Dennis, 2009). A higher proportion of infiltration may occur in areas where the natural
permeability is increased, such as the increased fracturing associated with high extraction
mining. Generally accepted values for recharge in high extraction areas are between 5 % and
7 % (Vermeulen & Dennis, 2009).
5.2.3 Deeper fractured Karoo aquifers (confined)
The pores within the Karoo and more specifically the Ecca sediments are too well-cemented to
allow any significant flow of water. All groundwater movement therefore occurs along secondary
structures, such as fractures and joints in the sediments. These structures are better developed
in competent rocks, such as sandstone, hence the better water-yielding properties of the latter
rock type (Hodgson etai., 1998).
It should be emphasised, however, that not all secondary structures are water-bearing. Many of
these structures are constricted because of compression forces that act within the earth's crust
(Hodgson et et., 1998). The chances of intersecting a water-bearing fracture by drilling decrease
rapidly with depth. At depths of more than 30 m, water-bearing fractures with significant yield
were observed to be spaced at 100 m or greater.
Dwyka Tillite occurs at the base of the aquifer. Packer testing of the Dwyka Tillite done by
Hodgson (1998) had a hydraulic conductivity distribution as indicated in Table 4.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 68
Chapter Five: Hydrogeology of the Study Area 2010
Table 4: Statistics for results on packer hydraulic conductivity testing of the Dwyka Tillite (Hodgson et al., 1998).
Statistics Dwyka Permeability
Mean (mId) 0,0034
Median (mId) 0,0024
Standard Deviation (mId) 0,0034
Minimum (mld) 0,0002
Maximum (mld) 0,0148
Number oftests 21
Due to its low hydraulic conductivity (Table 4), the Dwyka Tiltite forms a hydraulic barrier
(impermeable layer) between the overlying mining activities and the basal floor of the Karoo
Supergroup (Hodgson ef a/., 1998). The hydraulic conductivity value for the aquifer is estimated
at 1x10-5mld to 0.10 mld with an average of 0.02 mld (Vermeulen & Dennis, 2009).The
thickness of the deep aquifer depends on the depth of development of permeable zones and
permeable fracture systems in relation to the water level and can extend deeper than the coal
mine workings.
5.2.3.1 Recharge into the deeper Karoo fractured aquifer
Recharge of the deep Karoo aquifer occurs from the shallow Karoo aquifer through permeable
fracture systems that link the two aquifers. The natural distribution of such fracture systems is
highly variable, and the recharge of the deep aquifer is expected to be some orders of
magnitude lower than for the shallow aquifer. However, induced fracturing associated with
mining can extend from the deep aquifer up to the surface and provides a relatively direct and
highly permeable recharge route. The magnitude of recharge by this route depends on the
extent of mining and the nature of the induced fracture pattern (Vermeulen & Dennis, 2009).
5.2.4 Coal mine artificial aquifer (confined)
The artificial mine groundwater system associated with surrounding mined areas- the void zones
where mining has taken place either by board-and-pillar, high extraction andlor long wall mining
as well as subsided areas.
5.2.5 Effective Porositiesl Storage of the Karoo aquifer(s)
Twenty samples from the main sandstone unit in the Block 8 area at Middelbult were submitted
for porosity testing by Jasper Muller Associates (2002) Jasper Muller Associates cc (2002) .
The results are shown in Table 5. The large range in porosity for the fine- and medium-grained
sandstone is a function of the degree of pore-cementation. It also depends on the extent (depth)
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 69
Chapter Five: Hydrogeology of the Study Area 2010
of weathering. The difference in porosity between the different grain sizes of sandstone can be
seen in Table 5. The storativity of the Karoo aquifer(s) was calculated at a minimum of 0.005
and a maximum of 0.07.
Table 5: Total porosity for samples taken at Block 8 Jasper Muller Associates cc (2002).
Lithological unit Minimum (%) Maximum (%) Average (%)
Fine grained sandstone 0.3 9.9 4.1
Medium to coarse grained sandstone 7.7 14.4 10.1
Average for total 0.3 14.4 5.8
5.2.6 Hydraulic conductivity of the Karoo aquifer{s)
Slug tests were performed by Jasper Muller Associates (2002) in 13 of the 30 m deep shallow
boreholes and 14 of the deep boreholes, ranging in depth between 80 and 150 m, to determine
the hydraulic conductivity distribution within the saturated Karoo aquifers Jasper Muller
Associates cc (2002) .
Statistical analyses of packer tests, conducted at different depths in three of the deep boreholes
yielded the following results Jasper Muller Associates cc (2002) :
• A mean hydraulic conductivity of 4.3 x 10-3mid was calculated for fresh sandstonelsiltstone
intervals.
• A hydraulic conductivity of 1.56 x 10-2 mid was calculated for the 4 m fresh to slightly jointed
B4 dolerite test section (30-34 m).
• A hydraulic conductivity of 0.573 mid was calculated for the 4 m (fine-grained sandstone)
test section (60-64 m) across a water intersection.
Statistical assessment of hydraulic conductivities in South African hard rock aquifers indicate the
actual K-values to be somewhere between the geometric and harmonic mean Jasper Muller
Associates cc (2002) . A K-value of 0.02 mid is therefore proposed as realistic for the shallow
weathered Karoo aquifers at Block 8, while a value of 5 x 10-4mid is proposed for the deeper
fractured Karoo aquifers Jasper Muller Associates cc (2002).
5.2.7 Transvaal dolomitic aquifer (confined)
The Transvaal dolomitic aquifer is an aquifer that comprises the dolomitic formation that
underlies the Karoo rocks. The Karoo Supergroup has acted as a buffer to absorb weathering
events, and this prevented the development of any large scale karstification in the study area's
dolomites and its associated groundwater storage capacity. Furthermore, pre-Karoo glaciations
stripped out old karst scenery (Tweedie et et., 1986). As there is no or very little karstification in
the dolomites, the porosity and hydraulic conductivity values of the aquifer are very low (Figure
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 70
Chapter Five: Hydrogeology of the Study Area 2010
37 and Figure 38). Kruseman & De Ridder (1994) estimated that the Transvaal dolomitic aquifer
without karstification would have an effective porosity value of between 0 and 10% and a
hydraulic conductivity value of 10 x 10-5mis to 10 x 10-10mis (Figure 37 and Figure 38) and
Table 6 (Kruseman & De Ridder, 1994).
Few of the local farmers, if any, tap water from the aquifer beneath the Dwyka Formation Jasper
Muller Associates cc (2002) . The reasons for this, as stated by Van der Berg (2002) are:
• The great depth;
• Low-yielding character of the fractures;
• Inferior water quality, with high levels of fluoride, associated with granitic rocks;
• Low recharge characteristics of this aquifer because of the overlying impermeable Dwyka
Tillite.
The weathered zone near the surface of the dolomitic members of the Olifants River Group has
the highest groundwater potential in the area. However this is limited in extent (McKnight
Geotechnical Consulting, 2002). Previous studies have shown that the level of karstification is
negligible in the study area and as a result the hydraulic conductivity of the dolomite is greatly
reduced (Tweedie et al., 1986) Therefore, this aquifer can be regarded as insignificant as an
aquifer system in this specific study area. It is more likely to act as an aquiclude.
5.2.8 Gold mine artificial aquifer (confined)
The aquifer is overlain by the thick Ventersdorp Supergroup with its very low hydraulic
conductivity value resulting in it confining the Witwatersrand aquifer (Figure 29). Extensive
footwall development (haulages, cross-cuts, ore passes) for ore collection has allowed for an
artificial aquifer being created within the Witwatersrand that has a much higher hydraulic
conductivity value than the naturally occurring Witwatersrand aquifer surrounding it.
5.2.9 Deep Witwatersrand aquifer (confined)
The Witwatersrand aquifer occurs in siliceous sediments at the base of a thick, clastic
sedimentary sequence. The primary porosities of the sandstone sediments have been destroyed
by metamorphism, with the effect that the resultant quartzites are essentially impermeable due
to the precipitation of secondary quartz. However, the arenaceous members of the Supergroup
are structurally competent and fractures tend to remain open in the presence of water (McKnight
Geotechnical Consulting et al., 2002). The aquifer is overlain by the thick Ventersdorp
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 71
Chapter Five: Hydrogeology of the Study Area 2010
Supergroup (aquiclude) with its very low hydraulic conductivity value resulting in it confining the
Witwatersrand aquifer (Figure 29).
5.2.9.1 Recharge into the deep Witwatersrand aquifer
Groundwater is encountered within the gold mine workings of the Witwatersrand Supergroup,
the origin of which is thought to be a mixture of paleo-meteoric water, 2.0-2.3 (Ga) hydrothermal
fluid, connate water trapped in the Witwatersrand sediments during their deposition (Onstott ef
al., 2006) and influx from shallower aquifers. The paleo-meteoric water and 2.0-2.3 (Ga)
hydrothermal fluid of the Witwatersrand is thought to be finite within the Evander Basin and
therefore it could be exhausted (Rison Groundwater Consulting, 2007).
Recharge into the gold mine workings is thus influx from the shallow Karoo aquifer through
permeable fracture systems that link the two aquifers from the base of the Karoo aquifer into the
Witwatersrand aquifer where the Witwatersrand sub-outcrops against Karoo Supergroup or
where preferential paths through the Ventersdorp Supergroup intersect the gold mine. The
natural distribution of the fracture systems and preferential flow paths is highly variable, and the
recharge of the deep aquifer is expected to be lower than for the shallower Karoo aquifer (1-
3 %) or (5-7%) where subsidence has induced fractures (Vermeulen & Dennis, 2009). However,
induced fracturing or preferential flow paths associated with mining can extend from the deep
Witwatersrand aquifer up to the Karoo aquifer and provide a relatively direct and highly
permeable recharge route (Cook, 2003). The magnitude of recharge by this route depends on
the extent of mining and the nature of the induced fracture or preferential flow paths. Both of
these are the exception rather than the norm in the study area. The rate of recharge is therefore
estimated at less than 1 % of annual precipitation. The rate of recharge calculated from the rate
of inflow in the Witwatersrand aquifer (Table 15) was calculated at 0.5 % if the Evander Basin's
surface extent is used as the recharge area (Figure 17).
5.2.10 Effective porositiesl storage of the Witwatersrand aquifer(s)
Based on work done by Kruseman & De Ridder (1994) the effective porosities of the
Witwatersrand quartzites are a minimum of 0 to 5 %. The storativity value of the Witwatersrand
aquifer was calculated at 1.3 x 10-6 (Du Preez ef al., 2007).
5.2.11 Hydraulic Conductivity for the Witwatersrand aquifer
The hydraulic conductivity value for the Witwatersrand aquifer is very low due to metamorphism
of sedimentary rocks, which is estimated at 1 x 10-14 mis to 1 x 10-11 mis (Kruseman & De
Ridder, 1994).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 72
Chapter Five: Hydrogeology of the Study Area 2010
5.2.12 Summary of aquifers
There is dynamic interaction between the aquifers situated in the upper Karoo Supergroup as
there is a direct link with recharge from precipitation into the shallow perched aquifer and the
shallow weathered aquifer. This water then flows through preferential pathways (fractures) from
these aquifers into the deeper artificial coal mine aquifer and fractured Karoo aquifer. Therefore
the deeper Karoo aquifer is not directly recharged by precipitation but by influx from shallower
Karoo aquifers through preferential pathways naturally at a rate of 1 to 3% or by induced
fracturing resulting from subsidence at a rate of 5 to 7% of recharge. The water is prevented
from moving further downward at the base of the Karoo Supergroup by the Dwyka Tillite, and
further down by the Transvaal dolomites and the Ventersdorp Supergroup lavas that separate
the Karoo aquifer affected by coal mining from the Witwatersrand aquifer affected by gold
mining. The reason for the groundwater is due to the named layers low very low hydraulic
conductivities (Kruseman & De Ridder, 1994).
Therefore, recharge into the Witwatersrand aquifer can only occur where the Witwatersrand
aquifer sub-outcrops against the Karoo Supergroup and if preferential pathways exist
connecting the Witwatersrand aquifer with the Karoo aquifer through the Ventersdorp
Supergroup and Transvaal Supergroup. Same of the water in Witwatersrand aquifer could be a
mixture of paleo-meteoric water and 2.0-2.3 (Ga) hydrothermal fluid or connate water trapped in
the Witwatersrand sediments during their deposition (Onstott ef al., 2006). The source of this
water is thought to be finite within the Evander Basin and it could therefore be exhausted (Rison
Groundwater Consulting, 2007).
All recharge into the Witwatersrand aquifer is thus flux from preferential pathways connecting
the shallower lying Karoo aquifers, as no direct precipitation will recharge the Witwatersrand
aquifer or paleo-meteoric water from the Witwatersrand sediments. If no direct preferential
pathways exist, the influx will be relatively slow through the sub-surface and will be much less
than the recharge of 1-3 % in the Karoo aquifer and will be expected to be less than 1 % of
annual precipitation.
For the purposes of this study, the Karoo aquifers were regarded as one significant aquifer
situated above the Ventersdorp Supergroup and the deeper Witwatersrand aquifer situated
below the Ventersdorp Supergroup (refer to Figure 36).
Thus there are two major aquifers dominating the geohydrology of the study area namely:
• the Karoo aquifer (unconfined to confined),
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 73
Chapter Five: Hydrogeology of the Study Area 2010
• the deeper Witwatersrand aquifer (confined).
The shallower Karoo aquifer is regarded as an unconfined to confined aquifer and its recharge
can be linked directly to recharge by precipitation (1-3 % of annual precipitation and 5-7 %
where high extraction mining has occurred). The deeper Witwatersrand aquifer is regarded as a
confined aquifer that has no direct link of recharge to precipitation and has a very low flux from
shallower Karoo aquifers though preferential pathways or where the Witwatersrand Supergroup
sub-outcrops against the base of the Karoo Supergroup (>1 % of annual precipitation).
Table 6: Summary of the aquifer parameters within the study area (Vermeulen & Dennis, 2009; Kruseman & De
Ridder, 1994).
Supergroup Aquifer Recharge Hydraulic conductivity Effective Porosity Storativity (Ss)% (mld) % n)
Min. Max. Min. Max. Min. Max.
Perched 5
aquifer) 1x10·s 5x10·4 0.3 9.9 0.01 0.07
Weathered
Karoo aquiferl 3-5
Supergroup Fractured
aquifer 1x10·6 0.1 7.7 14.4 0 0.005
Goaf < 5 0.5 5
Coal Seam
(Before mined) 3-5 1x10·6 1xlO·' 0 10 0.001 0.01
Hydraulic conductivity
(mIs)
Transvaal Dolomites 3
Supergroup (Aquiclude) lxlO·6 lxlO·o.s 0 SO 0 1x10·6
Ventersdorp Ventersdorp >1
Supergroup lavas(Aquiclude) 1xlO·ll 1xlO-6 0 0 0.1 lxlO-6
Witwatersrand Witwatersrand >1
Supergroup aquifer lxlO-14 lxlO-ll 0 5 0 1.3xlO-6
5.3 Aquicludes/lmpermeable layers
5.3.1 Karoo Supergroup (Aquicludes)
The thick shale success ions within the Karoo Supergroup, the dolerite sills (K-value: 8.64 x
10-10 m/d) that are extensive over the area and the Dwyka Tillite (K-value: 0.0034 m/d) are all
low hydraulic conductivity to impermeable zones that may reduce the recharge to the underlying
aquifers (Hodgson ef al., 1998).
5.3.2 Transvaal Supergroup (Aquiclude)
The Karoo Supergroup has acted as a buffer to absorb weathering events; this prevented the
development of any large-scale karstification in the study area's dolomites and its associated
groundwater storage capacity. Furthermore, pre-Karoo glaciations have stripped out the old
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 74
Chapter Five: Hydrogeology of the Study Area 2010
karst scenery (Tweedie et al., 1986). As there is little or no karstification in the dolomites, the
porosity and hydraulic conductivity values of the aquifer are very low (Figure 37 and Figure 38).
Kruseman & De Ridder, (1994) estimated that the Transvaal dolomitic aquifer without
karstification will have an effective porosity value of between 0 and 10 and a hydraulic
conductivity value of 10 x 10.5 to 10 X 10.10 mis, as shown in Figure 37, Figure 38 and Table 6.
5.3.3 Ventersdorp Supergroup (Aquiclude)
The Ventersdorp lava in the study area is part of the Klipriviersberg lavas that were deposited as
a flood basalt seqeunce (Tweedie et el, 1986). Basalts if unweathered, as in the case of the
study area, have very low porosity values and very low hydraulic conductivity values, as shown
in Figure 37 and Figure 38 (Kruseman & De Ridder, 1994). Through the geological logs
interpreted, it does not appear as if the Ventersdorp Supergroup is permeable (weathered or
extensively fractured) and therefore the hydraulic conductivity value for the Ventersdorp lavas of
the study area can be regarded as 1 x 10-7to 1 x 10-am/s (Kruseman & De Ridder, 1994).
Porosity (effective porosity In gray)
Gravel r
cocree SlInd c-
Medium SlIM r
Fi"e 5 nd r
Slit, Loess
Till
Clay
Karst Limes-tone
LIn stone
SartdstOfl8 I
511t5100e
san
AII11ydllt8
ShaleL
Permeable ëesen
Fractured Boor k
Weath81~dGranite
Weathered Gal.Jbfo
BAMlt -
Crystlllllne ROCk-
o 20 40 60 80 100
Figure 37: Porosity values of selected rock types (Kruseman & De Ridder, 1994).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 75
Chapter Five: Hydrogeology of the Study Area 2010
Hydraulic Conductivity
Gravel I- -
coarse Sand f- -
Medium Sand t- -
Fine Sand t- -
Silt. loess f- -
Till f- -
Clay f- -
Marine Clay f- -
Karst Limestone f- -
Limestone f- -
Sandstone t- -
Siltstone I- -
Salt f- -
AnhydrHe t- -
Shale I- -
Permeable Basalt f- -
Fractured Bedrock t- -
Weëthered Granite f- -
Vofealhered Gabbro t- -
Basalt t- - -
Crystalline Bock f- -
10-7
meters/second
Figure 38: Hydraulic conductivity values of selected rock types (Kruseman & De Ridder, 1994).
The presence of the Ventersdorp lava is a critical issue, since it was found at other, similar
mines in the West rand Basin that groundwater inflow from the overlying aquifers is more
pronounced in areas where the lava is absent (Rison Groundwater Consulting, 2007).
Throughout the study area, the Ventersdorp lavas have a variable thickness (Figure 39), which
is partly attributed to the faulting in the area (Tweedie ef el., 1986). Important to note is that for
most of the study area the Ventersdorp Supergroup mantles the Witwatersrand Supergroup,
with the exception of the south-southeastern boundary of the interaction area, where the
Ventersdorp lavas are either very thin or non-existing (refer to Figure 30 and Figure 31). This
effectively means that the Ventersdorp lavas in most parts of the study area form an aquiclude,
restricting downward groundwater movement from the shallower Karoo aquifers towards the
deeper Witwatersrand aquifer.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 76
Chapter Five: Hydrogeology of the Study Area 2010
LEGEND
NN. Interaction areaMiddelbult boundaries
lEOOO
Figure 39: Ventersdorp thickness in the study area.
5.4 Preferential flow paths
5.4.1 Introduction
Preferential flow refers to the uneven and often rapid movement of water through subsurface
media, giving rise to heterogeneous conditions. This occurs in both the unsaturated and
saturated zones. Preferential flow paths can play a dominating role in the movement of water
and contaminants in the sub-surface (Cook, 2003). Figure 40 shows the relationship of fracture
aperture, fracture spacing and aquifer hydraulic conductivity, for an aquifer consisting of planar,
parallel uniform fractures (Cook, 2003). It indicates the wider and more frequent the preferential
pathways the higher the flow of water through a layer. As the coal mine lies above the gold
mine, the preferential pathways connecting the two mines will allow for flow from the coal mine
towards the deeper lying gold mine.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 77
Chapter Five: Hydrogeology of the Study Area 2010
E
1 10 100
Fracture Sp.acing (m)
Figure 40: Relationship between fracture aperture, fracture spacing and aquifer hydraulic conductivity, for an aquifer
consisting of planar, parallel uniform fractures (Cook, 2003).
In the study area some preferential flow are encountered. These flow paths may connect the
shallow Karoo aquifer with the deeper Witwatersrand aquifer though the Ventersdorp
Supergroup lavas. These pathways could have resulted from geological processes
(faults/fractures) or from mining activities (shafts/exploration boreholes).
5.4.2 Geological preferential flow paths
5.4.2.1 Fissures/Joints and Fractures (cracks)
Fissures/joints and fractures are planes along which stress has caused partial loss of cohesion
in the rock. Conventionally, a fracture or joint is defined as a plane where there is hardly any
visible movement parallel to the surface of the fracture (Cook, 2003).
Dykes can either be barriers or conductors for groundwater flow. How the dykes act depends on
their trends in relation to their hydraulic gradient and the density of the fracture system in the
dykes (Cook, 2003). Due to thermal effects, dykes can also cause fracturing of adjacent to rock.
Pumping tests indicate that dykes that are thicker than 10 m serve as groundwater barriers, but
those of smaller width are permeable, as they develop cooling joints and fractures (Kruseman &
De Ridder, 1994). Sills are nearly horizontal tabular bodies that commonly follow the bedding of
enclosing sedimentary rocks or lava flows. Some of the sills are very thick and extend over large
areas. Due to their low permeability, except when fractured, sills may support perched water
bodies. Groundwater flow next to dykes is controlled by two kinds of fractures, namely, vertical
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 78
Chapter Five: Hydrogeology of the Study Area 2010
and sub-vertical fractures with recharge occurring along these vertical fractures (Kruseman & De
Ridder, 1994).
In the study, only one water-bearing fracture in the gold mine (Table 7) next to a dyke was
encountered at Evander No. 8 shaft, 11 level approximately 1340 m below surface, as shown in
Figure 41. Early reports from the mines indicated that water- bearing fissures/fractures run dry
rapidly, suggesting a poorly connected fracture system and a lack of recharge (Rison
Groundwater Consulting, 2007).
The other water- bearing fractures (Table 7) were from the gold mine roof, which resulted from
planes along which stress has occurred causing a partial loss of cohesion in the rock and
allowing water to flow in these planes.
Figure 41: Fractures next 10dyke with water at Harmony Evander No. 8 shaft 11 level (1340 m below surface).
5.4.2.2 Faults
Faulting affected both the Witwatersrand and the Ventersdorp Supergroups in the study area,
but pre-dated the deposition of the Transvaal Supergroup (>2500 Ma) and the Karoo
Supergroup (Tweedie ef el., 1986). This was also highlighted in previous investigations of the
Rolspruit and Poplar Project areas (McKnight Geotechnical Consulting ef al., 2002). McKnight
Geotechnical Consulting Geotechnical Consulting (2002) also concluded that there are no major
displacements in the unconformity above the Black Reef Formation. In addition, McKnight
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 79
Chapter Five: Hydrogeology of the Study Area 2010
Geotechnical Consulting (2002) also concluded that the majority of the faults are under
compression and tight, meaning that the faults are most likely not conduits for groundwater flow.
In the study, two faults bearing water (Table 7) were encountered. They were at Evander No. 8
shaft, 11 level, 1340 m below surface (Figure 42) and at Evander No. 8 shaft, 18 level, 1830 m
below surface.
Figure 42: Fault bearing water at Harmony Evander No. 8 shaft 11 level (1340 m below surface).
5.4.3 Mining activities
5.4.3.1 Shafts
The deep level gold mining shafts have been cemented to significant depths (Leslie Mine
EMPR, 1994). If cracks occur within the cementation of the shafts, the shafts may act as
preferential flow paths for groundwater flow, connecting the shallower Karoo aquifer affected by
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 80
Chapter Five: Hydrogeology of the Study Area 2010
coal mining to the deeper lying Witwatersrand aquifer affected by gold mining. The positions of
the gold mine shafts are shown in Figure 43.
Gold mine shafts
lnteraction area
Figure 43: Gold mine shaft positions.
5.4.3.2 Exploration boreholes
Exploration boreholes intersecting the Karoo aquifer will act as preferential flow paths for
groundwater flow to the Witwatersrand aquifer if left unsealed. There are numerous exploration
boreholes covering the study area. The exploration boreholes with full geological logs are shown
in Figure 18. In this study, one water-bearing exploration borehole was found at No. 8 shaft, 15
level, 1620 m below ground level (Table 7). The movement of water from the surface downward
will be more rapid than in faults or fractures, as the exploration boreholes are drilled virtually
vertically in to the sub-surface, whereas faults and fractures are at different angles.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 81
Chapter Five: Hydrogeology of the Study Area 2010
Table 7: Preferential flow paths encountered in the study area.
Hl 1340 11 8 Water from fault
H2 1340 11 8 Water next to dyke
H3 1620 15 8 Water from exploration borehole
H4 1830 18 8 Water from fau It
HS 2152 24 8 Water dripping from roof stope
H6 282 1 9 Water dripping out of mine roof (minimal)
H7 282 1 9 Water dripping out of mine roof (minimal)
Hl0 634 8 2 Water dripping out of mine roof (minimal)
51 104 In coal mine Near Main shaft Borehole into mine workings
S2 113 In coal mine Near North Shaft Water dripping out of mine roof (minimal)
53 105 In coal mine Block 38 Fissure water from m ine floor
54 80 In coal mine Near West shaft Water dripping out of mine roof (minimal)
SS 55 Water dripping out of mine roof (minimal)
All were
5.5 Water levels
As there are different aquifers (unconfined to confined) in the study area, there are different
water levels, with the unconfined to confined. When a confined aquifer is intersected by a
borehole or a shaft, the water level in the borehole or shaft will rise and become a piezometric
level. This level will always try to come into equilibrium with the atmosphere or the piezometric
surface (piezometric surface = atmospheric pressure = 0). The piezometric surface is the
hypothetical surface of the hydraulic head in a confined or semi-confined aquifer (Freeze &
Cherry, 1979). The different water levels or piezometric levels within the aquifers will be
discussed sepa rate ly .
5.5.1 Karoo aquifer water levels
More than 240 shallow Karoo aquifer boreholes have been drilled across the study area in the
vicinity of the mines. The positions of the boreholes are shown in Figure 44. The groundwater
level is on average 3.9 m below surface, with a range of 0 to 69.6 m below surface.
In order to visualise the Karoo groundwater flow for the area, a water level contour map must be
generated. Although groundwater does not occur across over the entire study area, it gives a
good visualisation tool to interpret the general direction of groundwater flow and levels. An
interpolation technique, using the available data, was used to simulate groundwater levels over
the entire study area. The interpolation technique used is referred to as Bayesian interpolation,
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 82
Chapter Five: Hydrogeology of the Study Area 2010
where water levels are correlated with the surface topography. All available water levels were
plotted against topography, shown in Figure 45. The results indicate a correlation of 97 %
between the data sets. Therefore, Bayesian interpolation was valid and used to calculate water
levels for the entire study area. The interpolated water levels are shown in Figure 46.
It is important to note that the Karoo aquifers have not been affected by the past 50 years of gold
mining in the region, although large volumes of water have been pumped from the gold mines
(Rison Groundwater Consulting, 2007). This suggests a poor hydraulic connectivity from the
Karoo aquifer to the gold mine workings situated in the Witwatersrand aquifer, although some
dewatering of the Karoo aquifer has been observed in the vicinity of the Middelbult Colliery high
extraction panels, located to the south of the study area (Middelbult Mine: Block 8 Expansion
EMPR, 2003). To some extent, dewatering of this aquifer has occurred because of the pumping
in the gold mine. Here, the piezometric pressure in the deep Karoo aquifer is generally 10 to
50 m lower than that in the Karoo sediments (Hodgson et al., 1998).
LEGEND
NN. In eraction areaMiddelbult boundaries
N Harmony boundary
1EOOO
Figure 44: Karoo aquifer water levels.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 83
Chapter Five: Hydrogeology of the Study Area I 2010
In.
I Karoo aquifer water levelslE.
~ lS9D
~t
=-
lSO.
19. /
151iD 1650
lSaD 1510 1511. l5!IID lat. D.ID Uilll
.... II!!IIrIt ...... ,
1630
Figure 45: Correlation of the interpolated water
levels with surface topography.
o
Figure 46: Interpolated groundwater elevation map for the Karoo aquifer water levels (Bayesian interpolation).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 84
Chapter Five: Hydrogeology of the Study Area 2010
5.5.2 Witwatersrand aquifer groundwater levels
The Witwatersrand aquifer groundwater levels were found to be much more difficult to interpret
compared to the Karoo aquifer water levels. The reasons for this are:
• The great depth of the Witwatersrand aquifer;
• The pumping of groundwater from the underground gold mine workings to the surface,
lowering the water levels within the aquifer.
As no physical measurements of groundwater levels in the Witwatersrand aquifer were taken in
this study, the interpretations were based on previous data and reports, and on communications
with Harmony's environmental and geological department. The available groundwater levels are
displayed in Table 8.
Table 8: Available water levels for the Witwatersrand aquifer.
Water levels
Gold mine shaft name
Shaft level Metres above mean sea level Metres below surface
Shaft No.1 not available not available not available
Shaft NO.2 22 -143 1763
Shaft No.3 not available not available not available
Shaft No.S 22 -62 1682
Shaft No.6 9 925 701
Shaft NO.7 22 -497 2101
Shaft No.S 22 -466 2070
Shaft No.9 not available not available not available
Shaft NO.10 8 853 736
The first recorded Witwatersrand aquifer water level was measured at 1533 m.a.m.s.1. by
Pritchard-Davies (1962). Important to note is that this water level elevation is within the Karoo
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 85
Chapter Five: Hydrogeology of the Study Area 2010
Supergroup formation, just above the No. 4L coal seam roof, which ranges from 1450 m.a.m.s.1.
to 1521 m.a.m.s.l, at an average of 1490 m.a.m.s.1.
LEGEND
NNo Interaction areaMiddelbull boundaries
Figure 47: Interpolated elevations for the No. 4L coal seam roof.
The Witwatersrand Superg roup does not outcrop on the surface of the study area and is overlain
by the Ventersdorp, Transvaal and Karoo Supergroups (Tweedie ef al., 1986). The average
interpolated thickness of the overburden ranges from 2550 m to 254 m at an average of 1400 m.
This indicates that the orig inal water level measured by Pritchard-Davies (1962) is not a water
level but a piezometric level, as there are aquiclude(s) confining the Witwatersrand aquifer,
which results in the piezometric level of 1533 m.a.m.s.1. in the exploration borehole.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 86
Chapter Five: Hydrogeology of the Study Area 2010
LEGEND
NN Interaction areaMiddelbult boundaries
N Harmony boundary
IV Overburden Wits
Figure 48: Interpolated thickness of the overburden above the Witwatersrand Supergroup.
During the period 1956 to 1969, a dewatering cone developed around the mine workings and
the groundwater level was lowered to a deepest point of 1100 m below datum or 728 m.a.m.s.1.
(Rison Groundwater Consulting, 2007). Pumping in one mine affected the water levels in
adjacent mines. Groundwater drawdown therefore occurred at variable rates at the different
mines. The water level in Winkelhaak dropped by 305 m between August 1956 and July 1962.
The water level in Bracken dropped by 190 m between February 1961 and July 1962 and the
water level in Leslie dropped by 170 m between February 1961 and July 1962 (Pritchard-
Davies, 1962).
The values in Table 8 along with the top of the Witwatersrand Supergroup were used as the
current water levels around the gold mine shafts, shown in Figure 49. The top of the
Witwatersrand was derived from the exploration boreholes covering the study area and
discussed in the geological section. The top of the Witwatersrand was used as a water level, as
the base of the Ventersdorp Supergroup (aquiclude) confines the water level within the
Witwatersrand aquifer.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 87
Chapter Five: Hydrogeology of the Study Area 2010
Interpolation with the available data was used to simulate the Witwatersrand aquifer's water
levels over the entire study area. The interpolation technique used was the inverse distance
weighting method, which is deemed best for data with a large spatial variation. In Figure 49, the
effect of pumping on the Witwatersrand aquifer can be seen around the gold mine shafts.
Witwatersrand aquifer current water levels
-100
-200
-300
Figure 49: Interpolated groundwater levels around the gold mine shafts (Inverse distance weighting method).
5.5.3 Groundwater pumping stations and their effect on the Witwatersrand aquifer's
water level
In the gold mine, excess water accumulating in the mine workings is pumped to the surface
through a series of pumping stations. Currently, only two pumping stations are in operation: they
are the pumping station in No. 8 shaft and the pumping station in No. 7 shaft. No. 7 shaft's
pumping station is also the main pumping station within the gold mine. A summary of the gold
mine pumping stations is included in Table 9.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 88
Chapter Five: Hydrogeology of the Study Area 2010
Table 9: Shaft's and pumping stations ofthe gold mine.
Shaft name Shaft stat us Water pump stations Pump Station depths
Metres below Metres below Metres above
datum surface mean sea level
.'. , (1828.797
Winkelhaak No. 1 Decommissioned Water flows from 1 shaft
shaft (mothballed) towards 2 shaft where it is
652 444 1177
pumped out to surface
Winkelhaak No. 3 Decommissioned Water flows from 3 shaft
shaft (mothballed) towards 2 shaft where it is
pumped out to surface 698 437 1179
Winkelhaak No. 5 Decommissioned Water flows from 5 shaft
shaft (mothballed) towards 2 shaft where it is 1849 1640 -20
pumped out to surface
Winkelhaak No. Z Decommissioned Water from 1,3 and 5 shaft
shaft (mothballed) flows to 2 shafts from where
it is pumped to surface.
Pump decommissioned in
April 2010. Water from 1,2, 1471 1235 385
3 and 5 shaft is flowing
towards 7 shafts from
where it will be pumped to
surface.
Winkelhaak No. 6 Decommissioned Pump decommissioned in
shaft (mothballed) 1998 1471 1274 358
Kinross No. 7 shaft Decommissioned in Water from 7, 9 and 10 is
December 2009. currently being pumped to
surface. It is expected that
water from 2 shaft will
reach 7 shaft by 2011. 7 1857 1823 -219
shaft is the main pumping
station
Kinross No. 8 shaft In production Water pumped to surface
'. 2048 1855 -219
leslie No. 9 shaft Decommissioned Pump decommissioned in
(mothballed) October 2001. Excess water
flows to 7 shaft
unknown unknown unknown
Bracken No. 10 Decommissioned Pump decommissioned in
shaft 1994 1002 762 827
The shafts and the pumping stations are affecting the naturally occurring Witwatersrand aquifer
groundwater level either by lowering the groundwater level surrounding the shafts where
pumping stations are in operation or by creating piezometric levels in the shafts above the
Witwatersrand aquifer, where the pumping stations have been decommissioned for more than
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 89
Chapter Five: Hydrogeology of the Study Area 2010
10 years, as shown in Figure 52. The concept of a piezometric levels is shown in Figure 50
below.
Shaft puncturing the confined aquifer results in a
pressure release giving rise to a piezometric level
within the shaft
Figure 50: Conceptual model of piezometric levels within a confined aquifer.
The effect of the piezometric surface can also be seen in the conceptual model of the
interpolated groundwater levels in Figure 51. Both No. 6 and No. 10 shafts have been
decommissioned for over 10 years, which has resulted in the piezometric level in the confined
Witwatersrand aquifer rising above the confined water level within the aquifer. It will continue to
rise within the shafts until it reaches equilibrium below 1533 m.a.rn.s.l. (Pritchard-Davies, 1962),
which is expected to be the maximum the Witwatersrand water level can reach. Important to
note is that the 1533 m.a.m.s.1 recorded by Pritchard-Davies (1962) was before extensive
mining in the Witwatersrand Supergroup took place and that the Witwatersrand aquifer
(confined) was therefore under more pressure than currently exists due to the extensive voids
created by mining activities, which have decreased the pressure within the aquifer.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 90
2010
Coal seam floor 1450 to Witwaters rand aq uiter 0 rig inal 6 shaft = 925 mamsl (stable for 10 years)
1521 mamsl in study area water level = 1533 mamsl
10 shaft = S53mamst
E
Original piezometric level of {lold mine - 1533
mamsl
1:C
1CC
III
E
('II
E ~O
I Witwatersrand aquifer water
Legend level (top of Wits)
• B4 Dole rite sill - Faults
• Karco - Gold mine sha
• NO.4L coal seam
- Water levels
• B8 Dole rite sill
Transvaa
• Ventersdorp
Witv.atersrand
I I
-~C
,,~-::I-! -
Figure 51: Conceptual model of the water levels in the Witwatersrand aquifer where pumping stations have been decommissioned.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 91
Chapter Five: Hydrogeology of the Study Area I 2010
Witwatersrand aquifer
Original Piezometric level of gold Coal seam root at 1450tooriginal water level = 1533 1521mamsl in study area
SE
~
E wrtwatersrandlaquifer water
n::: level (top of
E
Shaft bottom - (-2i-53;;J,., _t.,
Legend
WI-(466) • 84 001.1t1Io'II - Foul!.Shaft bottom - (496) Witwatersrand aquifer water
Shaft bottom - (-516) • KAm!)
-- UukJ ,fJIt'. shaL
level (top of Wrts) • r40.4 ...C:>DIu,om - Wetet 1~'/ef!
• DTr'38nc~Y;"n ''''
• Venlo"""'"
-',...'.-:-.,-'--------------------L W"f:\IVAMrsrand _r--
.;':"'_
Figure 52: Conceptual model of the water levels in the Witwatersrand aquifer where pumping stations are in operation.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 92
Chapter Six: Possible Interaction Areas within the Study Area 2010
Chapter Six: Possible interaction areas
within the study area
The possible areas of interaction are where Sasol Mining's Middelbult (Block 8) Colliery is
currently and where it will in future extract coal from the No. 4L coal seam, overlying Harmony
Evander gold mining operations that mine the underlying gold-bearing Kimberley reef (Figure
53).
Sasol coal mining will expand its existing Middelbult Colliery into the Block 8 reserves to the
west and north of its current mining operations (Figure 53). The initial expansion will be to the
west from the existing underground coal mining, as shown Figure 53 (Middelbult Mine: Block 8
Expansion EMPR, 2003).
There is a potential risk that the coal mining will damage the geohydrological barriers
(aquicludes) that currently exist between the shallow Karoo aquifer where Sasol coal mining
takes place and the deeper Witwatersrand aquifer where Harmony gold mining operates.
Damage to these geohydrological barriers could result in an increase of groundwater influx from
surface (gold mine slimes dams) as well as the shallower Karoo aquifer affected by coal mining
into the deeper Witwatersrand aquifer affected by gold mining. This could affect both the quality
and quantity of groundwater in the study area.
In some areas, the potential risk for groundwater interaction is much higher than in others. The
potential risk for groundwater interaction is a function of the following:
• depth and areas of mining;
• tailings storage facilities (TSF);
• geology;
• preferential pathways.
If the above factors are considered, certain high risk areas for potential interaction can be
identified.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 93
Chapter Six: Possible Interaction Areas within the Study Area I 2010
LEGEND
Gold mine shafts
Middelbult shafts
Intera cti on .a rea
Middelbult boundaries
Middelbult Future I'I,ililling
Middelbult UlG
Gold n,lliningArea
Figure 53: Sasol Mining's Middelbult (Block 8) current and future mining over Harmony Evander's gold mining activities.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 94
Chapter Six: Possible Interaction Areas within the Study Area 2010
6.1 Potential Interaction: Mining depths
6.1.1 Sasol coal mining
The No.4L coal seam mined by Sasol is situated at an average depth of 80 m below surface in
the south and 150 m below surface in the north, or at elevations ranging between 1450 m.a.m.s.1
and 1521 m.a.m.s.1 in the different compartments resulting from the different dolerite sills
situated throughout the study area (Figure 54). The coal seam is overlain by the Karoo
Supergroup sediments compromising mainly sandstones and shales (Middelbult Mine: Block 8
Expansion EMPR, 2003). Figure 54 displays the interpolated contours of the overburden
thickness covering the No. 4L coal seam in the study area. The interpolated elevations for the
coal seam floor are shown in Figure 47.
.~:
e
z.
LEGEND
NN. Interae ion areaMiddelbult boundaries
Figure 54: Overburden thickness in metres above the No.4L coal seam.
6.1.2 Hannony gold mining
The Kimberly Reef containing the gold and platinum group metals are at varying depths below
surface is currently being mined. The deepest operations are 2500 m below surface, at
Harmony No. 8 shaft, and the shallowest are 240 m below surface, at Harmony No. 3 shaft. This
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 95
Chapter Six: Possible Interaction Areas within the Study Area 2010
variable depth is due to the general dip of the Kimberley reef to the north, of between 160 and
240 and the east-west trending faults, with downthrows to the south, which resulted in the blocks
being duplicated (Tweedie et el., 1986). In Figure 55, Harmony Evander's gold mining shaft
system is illustrated through a cross section.
6.1.3 Difference in mining depths
In some areas, the coal mining overlaps the gold mining, as shown in Figure 53 and Figure 55.
These areas have the highest risk of groundwater interaction where the coal mining and gold
mining are relatively close together (less than 300 m difference in mining depths) (see Table
10). In these areas, influx of water from the coal mine into the gold mine can result from the
close proximity of the mines through induced preferential flow paths.
The minimum difference in depth, of approximately 81 m between Sasol mining's No. 4L coal
seam and the shallowest gold mining workings, is at Harmony Evander's No. 9 shaft (A level)
and the maximum difference, of approximately 782.5 m, is at Evander's No. 8 shaft (8 level).
Table 10: Difference in mining depths.
Difference in elevation (m.a.m.s.l) of coal seam floor and shallowest gold mining operations through
the study area
Gold mine shaft
1 Shaft l479m 1347m (2 level) 132m
2 Shaft l483m 9l2m S7lm
3 Shaft l480m 93m
5 Shaft l482m 8S8m
7 Shaft l490m 849m 64lm
8 Shaft l489.5m 707 m 782.5m
9 Shaft l46lm 1380m (A level) 8lm
10 Shaft l470m 134Sm l2Sm
(m.a.m.s.l) metres above mean sea level
Currently, Sasol Middelbult Block 8 is only mining over Harmony gold mining in the
southemmost parts of the study area, with the largest area covering the southeastern part
(Figure 53). Future mining will cover most of the study area except for the central part of the
study area around Harmony Gold Mining's No. 7 and No. 8 shafts (Figure 53).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 96
10:=&11:,3,= '" &Sl.= ë= =&ï~ 1= Depth below ;;. ê=
datum (1828.797
."'< r= .T
I
,x~-rt __
•• , I. ..J._:: . ";y ..ti. 4 ='11 ..,t::'=:-n II.., "'"
.."I .',;,-"
·3r
.;?I J .'£i)
~ ·e~l1
c;y' ·11::11 ... 1 ·l1<f.m
~~
·1::- 'I----=....::....__--------=-,J--.1:3rn121"
~ .t3':'l1
9?i .1--113
l: .?J I .1J.;::- . _.. 13 ~h1
11?J' 41~_~j i.-:!I..·t:·l~
1: ?I,' ·lë·;;; l:. ~~~- ·1:~..,
,.J -1E"~
·..1....D~
I
qa «1
Figure 55: General Evander gold mining operations illustrating their shaft system (not to scale). (Presentation 53 Kinross, Harmony Evander).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 97
Chapter Six:Possible Interaction Areas within the Study Area 2010
Areas that have the highest potential for groundwater interaction according to difference in depth
are:
• The area around Harmony's No. 1 Shaft (Sasol future mining);
• The area around Harmony's No. 3 Shaft (Sasol current mining);
• The area around Harmony's No. 9 Shaft (Sasol future mining);
• The area around Harmony's No. 10 Shaft (Sasol future mining);
These areas are shown in Figure 56 to Figure 59.
6.1.4 The area around Harmony's No. 1 Shaft
Sasol Coal Mining in future will mine near Harmony's No. 1 Shaft, where the No. 4L coal seam
floor is situated at 1479 m.a.m.s.1 or 139 m below surface (Figure 54). Harmony Gold Mining's
closest mining operations at NO.1 shaft to the No. 4L coal seam is situated at Level 2, which is
1347 m.a.m.s.1 or 270 m below surface. The minimum difference between the No.4L coal seam
and the gold mine workings at NO.1 shaft is thus 132 m (Figure 56). As yet, no groundwater
interaction from the coal mine into the gold mine could be established in the No. 1 shaft area, as
no preferential pathways could be confirmed.
200
N --________________________________________________________ S
1#
• N04 e•• I".~
• sa cltn'A ,.
fI'lSruJ
• V.n· ~j.'I'
Wr.v."l!t1ltl."
500t:::~!!!!l~~~~L-----------------j
o Crc s Sect.en 2846
Figure 56: Cross section through Evander NO.1 shaft.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 98
Chapter Six: Possible Interaction Areas within the Study Area 2010
6.1.5 The area around Harmony's No. 3 Shaft
Sasol is currently mining near Harmony's No. 3 shaft, where the No. 4L coal seam floor is
situated at 1480 m.a.m.s.1 or 136 m below surface. Harmony gold mining's closest mining
operations at NO.3 shaft to the No. 4L coal seam is situated at Level 1, which is 1387 m.a.m.s.1
or 229 m below surface (Figure 54). The minimum difference between the No.4L coal seam and
the gold mine workings is thus 93 m at No. 3 shaft (Figure 57). As yet, no groundwater
interaction from the coal mine into the gold mine could be established in the No. 3 shaft area, as
no preferential pathways could be confirmed.
20 w -------------------------------------------------------- E
Tf1nsvlll
• i.nttBdorp
-1000L_==========:!.--------------------------------------------__.l
o
0055 Section 12000
Figure 57: Cross section through Evander No. 2 and NO.3shaft.
6.1.6 The area around Harmony's No. 9 Shaft
Sasol is proposing to mine in future near Harmony's No. 9 shaft, where the No. 4L coal seam
floor is situated at 1461 m.a.m.s.1 or 126 m below surface (Figure 54). Harmony Gold Mining's
mining operations at NO.9 shaft, closest to the No. 4L coal seam, is situated at A level which is
1380 m.a.m.s.1 or 207 m below surface. The minimum difference in metres between the No.4L
coal seam and the gold mine workings is approximately 81 m at No. 9 shaft (Figure 58). Two
preferential flow paths bearing water at No. 9 shaft (Level 1) directly underneath Level A (shown
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 99
Chapter Six: Possible Interaction Areas within the Study Area 2010
in Figure 55) were found in this study (Table 5). These fractures could link the Karoo aquifer
affected by coal mining to the gold mine workings in the Witwatersrand aquifer.
w --------------------------------------------------------- E
9#
~
E
CQ
E
Figure 58: Cross section through Evander No.9 shaft.
6.1.7 The area around Harmony's No. 10 Shaft
Sasol is currently mining near Harmony's No. 10 shaft, where the No. 4L coal seam floor is
situated at 1470 m.a.m.s.1 or 119 m below surface. Harmony Gold Mining's mining operations at
No.10 shaft, closest to the No. 4L coal seam, are situated at Level 1, which is 1345 m.a.m.s.1or
244 m below surface. The minimum difference between the No.4L coal seam and the gold mine
workings is approximately 125 m (Figure 59). As yet, no groundwater interaction from the coal
mine into the gold mine could be established in the No. 10 shaft area, as no preferential
pathways could be confirmed.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 100
Chapter Six: Possible Interaction Areas within the Study Area 2010
N S
~
E
• B4 Domme ,il _ ".utt.
• Karoo _ Gokt mm@ shaft Iwon
• No 4L cool .eom
• BSOolartt •• 1I
Transvaal
• Vonto"dorp
Wrtwot!rsrond
o 4 00
Figure 59: Evander No. 10 shaft.
6.2 Potential Interaction: Evander tailings storage facilities
The potential for groundwater interaction resulting from Harmony Evander Tailings Storage
Facilities (TSF) found on surface can be attributed to the extent of subsidence resulting from the
coal mining directly underlying or near the TSF. There has been coal mining underneath the
Winkelhaak and Leslie slimes dams, with proposed future coal mining underneath the Kinross
slimes dams, as shown in Figure 60.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 101
Chapter Six: Possible Interaction Areas within the Study Area 2010
LEGEND
Gold mine shafts
Winkelhaal slimes dam
Leslie slirnes dam
Kinross slimes dam
• lu1iddelbult shafts
NN Interaction areallIIiddelbult boundaries
,A lu1iddelbult Future Mining
N Middelbult U/G
Figure 60: Harmony Evander gold mine tailings storage facilities,
6.2.1 Subsidence due to High Extraction coal mining
As the coal seam that supports the overlying rock mass is removed, the immediate roof of the
seam fails and collapses, forming rubble fill (qoaf), The caved rock breaks up into blocks varying
in size from a few centimetres to tens of centimetres. These blocks rotate and pile up, partially
filling the mined void. A stage will be reached when the mine workings are filled due to bulking of
the rubble (Vermeulen & Usher, 2003).
As the coal mining face advances, higher-lying beds fall, settle and compress the caved layer.
The stress in this "secondary caving zone" depends on the degree of compression of the
"goafed" material. Typically, an extensive network of horizontal and vertical cracks develops,
although the degree of displacement is not such that the beds lose their relative positions
(Vermeulen & Usher, 2003).
Strata overlying the secondary caving zone bend and sag under their own weight and compress
the goaf. As a result, the strata part along bedding planes. Continuity of the beds is maintained,
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 102
Chapter Six: Possible Interaction Areas within the Study Area 2010
although some fractures develop from the stress of bending and sagging. The sagging will result
in tension zones, compression in the rock strata and the development of cracks (Figure 61).
I Phreatic aquifer
~
...__ ,..... --""""1 .",,- ....H..ig.-h- transmis,;,,;h...", ~I II
~ ~
--------. ,... <, "',. /' Low transmissivity(
'"~ .,... " Iy~ ~ --Interm-ediate-tran-smiss- -ivity -
CJ Soil
D Shale
D Sandstone (Main aquifer)
D- CoalHigh extraction panel with cracks extending to the surfacer--:-. Dewatering conesDirection of groundwater flow
Figure 61: Schematic cross-section of a subsided high extracted panel (Grobbelaar, 2001).
Pillar extraction will present the greatest risk to groundwater influx into the underground coal
mine due to potential surface subsidence as we" as the expansion of preferential flow paths
such as dolerite dykes, geological contact zones, dolerite si" contact zones, faults, fissures and
unsealed exploration boreholes situated in the study area (refer to Figure 18). Influx of
groundwater into the mine workings, as a result of high extraction mining will therefore be much
more severe than for board & pillar underground mining activities.
The main contributors to the influx volumes as described by Vermeulen & Usher (2003) are:
• Annual rainfall recharge;
• Lateral inflows from the surrounding rock mass - deep (fractured) Karoo aquifer;
• Water released from storage in the overlying rock mass - shallow and deep aquifers;
• Inflows related to geological features;
• Inflows from surrounding mined out areas.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 103
Chapter Six: Possible Interaction Areas within the Study Area 2010
The actual increase in groundwater influx to the coal mine will be a function of:
• The actual extent of surface depressions forming ponds on the surface;
• The occurrence of areas that have a higher recharge potential because of the type of soil
and vegetative cover; and
• The dimensions of apertures (cracks), and the infilling of these cracks with sediment.
A possible influx of groundwater into the coal mine could be as a result of board-and-pillar
mining into the coal mine from the 3 tailings facilities situated on the surface of the study area
and owned by Harmony Gold's Evander operations. These slimes dams are:
• Kinross slimes dam,
• Leslie slimes dam,
• Winkelhaak slimes dam.
The 3 tailings facilities with the possible subsidence are shown in Figure 62.
6.2.1.1 Kinross slimes dam
The Kinross slimes dam is situated in the centre of the study area with Sasol future mining
underneath it as shown in Figure 60. No subsidence is expected underneath the Kinross mine
as shown in Figure 62 and Figure 63. This is a result of no high extraction coal mining directly
underneath of near the Kinross slimes dam (Figure 63).
6.2.1.2 Leslie slimes dam
The Leslie slimes dam is situated to the south-southwest of the study area. Some existing high-
extraction coal mining panels occur as close as 20-50 m of the southwestern corner of the
slimes dam. The 84 dolerite sill restricting vertical groundwater is very thin to the southwest of
the slimes dam (Figure 64 and Figure 66).
Therefore subsidence is likely to the south-southwest of the Leslie slimes dam, as high-
extraction coal mining will occur and it is expected that preferential pathways (fractures) linking
the slimes dam with the underground coal mine working will be induced (Figure 64). Although
subsidence is expected near the Leslie slimes dam groundwater interaction from the slimes dam
could not be confirmed due to inaccessibility.
6.2.1.3 Winkelhaak slimes dam
The Winkelhaak slimes dam is situated south-southeast in the study area. Some existing high-
extraction coal mining panels occur as close as 50-80 m to the southeastern corner of the
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 104
Chapter Six: Possible Interaction Areas within the Study Area 2010
slimes. The 84 dolerite sill is very thin to the south-southeast of the slimes dam (Figure 65 and
Figure 66).
Therefore subsidence is most likely to occur southeast of the Winkelhaak slimes dam, as high
extraction coal mining will occur and it is expected that preferential pathways (fractures) linking
the slimes dam with the underground coal mine working will be induced (Figure 66). Although
subsidence is expected near the Winkelhaak slimes dam groundwater interaction from the
slimes dam could not be confirmed due to inaccessibility.
LEGEND
liIIiddelbult Possible Subsidense
Gold mine shafts
Winkelhaal slimes dam
Leslie slimes darn
Kinross slirnes dam
_ Middelbull shafts
NlY. lnteractlon areaMiddelbult boundaries
Middelbult Future Mining
/\j Middelbull U/G
Figure 62: Possible areas of subsidence resulting from high extraction coal mining.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine lOS
Chapter Six: Possible Interaction Areas within the Study Area I 2010
N S E W
1650
1600 /_ Kinross sllrnes dam _\=:::===--- - ._' . - --=
1550
1500
Legene
1450 .....
• &I~
• 4loeo.ts.ern --- ........
1400 _. 8100ent0
1400~ - ii.. _7':: I I
0 Cross Sa<:too 3500 _"" •..".. •.~.~ •..~ CrOS5 Sooc1.cn 4500
Figure 63: Kinross slimes dam conceptual cross sections.
N -----------------------------------5 W E
165~
4000
Figure 64: Leslie slimes dam conceptual cross sections.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 106
010
N ---------------------------------5 W E
..",-:..~"''''''''''
o Cross Seclton 4500
3000
Figure 65: Winkelhaak slimes dam conceptual cross sections.
I M tt,wa,,., ~___
~
NN Slime. darna section linea\Vinkelham slimes dam
N Le.lie slimes dam
Klnross slimes dam
NN Interaction areaMiddelbuU bound.mes
F~ure 66: B4 dolente thickness in metres with the west-east and north-south section lines.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 107
Chapter Six: Possible Interaction Areas within the Study Area 2010
6.3 Potential Interaction: Geology
Geology plays a major role in the possible groundwater interaction between the two mines
situated in the study area. It is expected that the Transvaal Supergroup dolomites and the
Ventersdorp Supergroup lavas will act as aquicludes restricting groundwater movement from the
overlying coal mining into the underlying gold mining (refer to section 5.3 Impermeable layers).
As a rule of thumb, it is safe to assume that a cover of 50 m (based on K-values, Kruseman &
De Ridder, 1994) or more of Ventersdorp Supergroup lavas in this study area will form an
impermeable layer restricting movement of groundwater from the overlying coal mine into the
underlying gold mine.
The Ventersdorp Supergroup lavas appear to thin towards the south-southeast of the study
area, with the exception of the southern boundary of the study area where the Ventersdorp
Supergroup lava is either very thin or sub-outcrop against the Karoo Supergroup as shown in
Figure 67 and Figure 68. In these areas the potential for groundwater interaction between the
two mines is much higher than in the areas where the Ventersdorp lava is thicker than 50 m.
2
N s
Legend
• B4Do SII - Flu
• Keroo _ Gold mmeshefts"'orbIQ~
• No 4Looel SU'Tl
• B8Do mUll
Transvaal
• Ve ta~dorp
Wll'AatelSllnd
·500~----------------------------------------------------------------~
o
Figure 67: Thinning ofthe Ventersdorp Supergroup from north to south.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 108
Chapter Six: Possible Interaction Areas within the Study Area 2010
W E
Th nn.nQ of t-e
Ventersnerp avss
towards t"'e east-
soU:~eastof Ire study
• B8Do~rited
rra sveal
• Venle""orp
o Cross Sec ion 13739
Figure 68: Thinning of the Ventersdorp Supergroup from west to east.
North West
Area
- Fauht 11623
ct! - Gold mlina .... 11.... 01""'._.• No'leoal •• am
E • ea OO"nl •• ill
• V.n,. ...do"
o CICSS St ~!I 20146
Figure 69: Potential Interaction area based on geology.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 109
Chapter Six: Possible Interaction Areas within the Study Area 2010
sections
in Figure 65
Golómlne snsfts
f\ Pote'1tl.1 IrtET'5Cion NEB
_ 1,1idd.lbult .h.tu
N Welt·Ent Se-aion
N I o<th-South SeClion
N Int~r.ction ilI.a_
~ Il1Idd.lbult t>oundlJI ..
N Hsrmony ouncs.ty
I lOl lOl! !
Figure 70: Surface view of the Potential Interaction Area combined with the west-east and north-south cross seerons.
6.4 Potential Interaction: Preferential pathways
Samples were collected from all water-bearing structures intersecting the gold mine and these
are summarised in Table 7_ It should be noted that these points are the only water- bearing
structures found in the study and therefore the only known pathways potentially connecting the
Karoo aquifer with the Witwatersrand aquifer. Also of importance from the mining layout and the
positions of the samples is that only the deep gold mine samples HS, H7 and H10 could be
affected by coal mining as future mining has not reached the positions of the other underground
gold mine samples as shown in Figure 74_All preferential pathways were physically and visually
inspected were possible throughout the gold mine.
• H1 and H4 - Water from faults intersecting the gold mine, the origin of this water is possibly
from the base of the Karoo aquifer where the Transvaal Supergroup (aquiclude) is absent
and where the faults went through the Ventersdorp Supergroup (aquiclude) and into the
Witwatersrand aquifer. Groundwater will move along these faults until they intersect the gold
mine workinqs. The position of the faults are at Evander No. 8 shaft, 11 level 1340 m below
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 110
Chapter Six: Possible Interaction Areas within the Study Area 2010
surface (Figure 42) and at Evander No. 8 shaft, 18 level 1830 m below surface. It unlikely
that the coal mining will reach the area where these faults are (Figure 53) and therefore it is
unlikely that the coal mine will have an impact on the gold mining through these specific
faults (Table 7).
• H2 - Water from fractures next to dyke and exposed in the gold mine workings the origin of
this sample will be where fractures next to the dyke allowed for groundwater flow (Cook,
2003). These fractures mayor may not be directly connected to surface (Cook, 2003) and
therefore recharge may occur in the sub-surface at different depths through the influx of
overlying aquifers or directly by surface precipitation. The position of the fractures next to a
dyke were encountered at Evander No. 8 shaft, 11 level approximately 1340 m below
surface shown in Figure 41. It is unlikely that the coal mining will reach the area where this
dyke is encountered (Figure 53) and therefore it is unlikely that the coal mine will have an
impact on the gold mining through this specific dyke (Table 7).
• H3 - Water from exploration borehole intersecting the gold mine workings. According to
Middelbult Mine, Block 8 Expansion EMPR (2003) all gold mine exploration boreholes will be
sealed off with a non-reactive material to prevent the drainage of groundwater from the coal
mine to the gold mine. It is unlikely that the coal mining will reach the area where this
borehole is situated (Figure 53). Therefore coal mining will have no impact on the gold
mining through this specific exploration borehole. The position of the exploration borehole is
at No. 8 shaft 15 level 1620 m below surface (Table 7).
• H5, HS, H7 and H10 - Water dripping out of the mine roof which origin is difficult to
determine. The origin of these samples is thought to be flux from fissures/fractures and joints
from overlying aquifers. The position of the samples can be seen in Table 7. All of these
positions can be affected by coal mining except H5 which is at Harmony No. 8 shaft and will
remain unaffected by coal mining as no coal mining will occur around No. 8 shaft area
(Figure 53).
• From communications with the coal and gold mine there is explorations borehole
underneath the Winkelhaak slimes dam leaking into the coal mine. The presence of this
borehole could not be confirmed as the area underneath the slimes dam is inaccessible due
to the flooding of the coal mine.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 111
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
Chapter Seven: Hydro-chemistry and
Stable Environmental Isotopes
.ID. Hem (1985) gives a detailed description of the chemical principles involved in naturally
occurring water as well as the behaviour of different elements in natural waters (see Hem, l-l.D;
1985. "Study and interpretation of the Chemical Characteristics of Natural Waters." United
States Geological Survey - Supply Paper, no. 2254). An understanding of the hydro-chemistry
of groundwater systems is required for any relevant interpretation. The following is an overview
of the principles and applications of natural water chemistry, gathered predominately from the
work of Hem (1985).
7.1 Environmental factors affecting groundwater chemistry
Solutes are derived from a number of sources, including rocks from the earth, the oceans,
animal and human activities. Water moves through the hydrogeological cycle, with solutes being
added and removed from solution at various stages of the cycle (Hem, 1985). Solutes present in
any form of solution represent the net effect of a series of chemical reactions. These include the
dissolving of material from solid phase, the altering of previously dissolved components and the
eliminating of certain components from solution by evaporation or ion-exchange reactions. The
concentrations of solutes as well as the reactions occurring in a given natural groundwater
system are influenced by factors such as climate, geological setting and biochemical effects
from plant, animal or human interaction. The most significant factor, however, is the geological
setting and its interaction with the groundwater of a given area (Hem, 1985).
7.1.1 Climate
Both temperature and precipitation influence weathering rates and hence the transfer of
components from a solid or gaseous phase into a liquid. The amount of rainfall and evaporation
are important parameters affecting the chemical composition of natural groundwater (Hem,
1995).
7.1.2 Geological Setting
Evaluating the composition of the rock types provides a good indication of the solutes available
for uptake in the hydrogeological cycle. Elements found in groundwater in the mining sector
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 112
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
come from a variety of sources, but most significantly from the host rocks through which the
water flows, particularly from the decomposition of minerals caused by weathering and
contamination caused by mining activities.
Weathering can be both physical and chemical. The mechanical (physical) breakdown of rocks
increases the surface to volume ratio, which in turn makes chemical weathering more effective
(Eby, 2004). Chemical weathering is the process that adds most types of soluble minerals to
natural waters.
There are four major types of chemical weathering as described by Eby (2004):
• Oxidation: the reaction with of oxygen with minerals in which reduced forms of elements are
oxidised is termed oxidation. For example, the reactions of oxygen with pyrite:
Equation 5
4FeS2 (Pyrite) + 1502 + 8H20» 2Fe203(Hematite) + 8H2S04
• Congruent dissolution by water: simple dissolution of minerals that are soluble in water. For
example, the simple dissolution of NaCI (halite) releases Na+and Cl into solution:
Equation 6
NaCI (halite) » Na++ Cl
• Congruent dissolution by acids: minerals are dissolved by acid attack, often organics formed
in the soil. For example, the congruent dissolution of calcite by acid attack releases Ca into
solution:
Equation 7
CaC03(Calcite) » H2C03 (aq) » Ca2++ 2HC03-
• Incongruent dissolution by acids: minerals breakdown into ions in solution as well as the
solid phase of different compositions. For example, the incongruent dissolution by acids of
plagioclase release Na into solution while forming kaolinite:
Equation 8
2NaAISbOa (Na-plagioclase) + 2H2C03 + 9H20 » AI2Si30s (kaolinite) + 2Na+ + 2HC03- +
4H4Si04 (aq)
Once in solution, various other mechanisms remove the solutes, again altering the chemistry.
Yechieli & Wood (2002) describe these mechanisms as:
• Precipitation of salts;
• Dissolution precipitation processes;
• lon-exchange and adsorption;
• Redox reactions.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 113
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
Precipitation is a major sink for solutes, causing a change in the residual brine. Throughout the
groundwater regime, minerals can be dissolved by water as well as precipitated out of the
groundwater depending on their solubility, thus adding or removing solutes from solution. lon-
exchange can have a large effect on the hydro-chemistry of groundwater (Appelo & Postma,
2005). Redox reactions associated with organic matter can have a considerable effect on
groundwater chemistry. Reduction of sulphates to sulphides is one of the most common
processes.
Certain physical aspects also play a role in the supply of solutes: aquifers consisting of porous
rocks more readily supply solutes, through factors such as:
• the size of the crystals within the rock;
• rock texture and porosity;
• regional structure formation and length of exposure or residence time.
Furthermore, the depth below surface to which the water is transported plays a key role in water
quality, since solubility increases with increasing temperature, as do dissolution rates (Eby,
2004).
Thus the geology of an area plays an important role in groundwater quality. Many sedimentary
and igneous rocks in the gold and coal mines contain reactive minerals that may create acidic
conditions when in contact with oxygen and water (Drever, 1997). These reactions are
collectively referred to as Acid Rock Drainage reactions or more commonly as ARD.
Sulfide-bearing minerals, of which pyrite (FeS2) is the most abundant, will react with water in an
oxidising environment to form sulphuric acid (H2S04). Detailed reactions are (Freeze & Cherry,
1979):
Equation 9
FeS2 + 7/2 O2 + H20 => Fe 2++ 2S0/- + 2W
Equation 10
Fe2++ 1/402 + H+ => Fe3++ 1/2 H20
Equation 11
Fe3+ + 3H20 => Fe (OH)3 (yellow boy) + 3W
Equation 12
FeS2 +14Fe3+ + 8H20 => 15Fe2++ 2S0/- + 16H+
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 114
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
Sulphuric acid is a strong acid that will mobilise most heavy metals, which will lead to
widespread contamination of the aquifer system. ARD is specifically associated with coal and
gold mines, as pyrites are abundant in carbonaceous material (Vermeulen & Usher, 2003).
Witwatersrand Supergroup sedimentary rocks in the gold mining area may contain water that
was trapped within pores during their deposition (McKnight Geotechnical Consulting et el.,
2002). This water is known as pore or connate water, and is usually briny as a result of
extensive ion-exchange reactions. As the pressure and temperature steadily increase (as more
material is deposited), the pre-water will be released from the pores and will cause an increase
in the salinity of the surrounding groundwater (Onstott et al., 2006).
Geological structures such as dykes are known to create separate groundwater compartments
in which the groundwater chemistry may differ from surrounding aquifers. Groundwater
compartments may receive less or more recharge than the surrounding aquifer, which will lead
to an increase or decrease in groundwater salinity. This can occur in gold mines as well as in
coal mines (Vermeulen & Usher, 2003).
7.1.3 Biochemical effects
Reactions involved in life processes such as the decay of organic matter occur not only in
freshwater bodies but also in groundwater bodies. The removal from, or addition of solutes to,
the water body can occur through many varying life forms (Onstott et st., 2006).
7.1.4 Other factors
7.1.4.1 Unsaturated zone
The unsaturated zone is defined as the area between the water table and the land surface. The
unsaturated zone is of great importance, as it plays a role in the recharge of an aquifer as well
as the quality of the groundwater (Freeze & Cherry, 1979). The unsaturated zone acts as a filter
that will filter out any solids in the water before it enters the aquifer. The volume of water that will
filter through the unsaturated zone is determined by its composition and thickness. If the
unsaturated zone is composed of clayey or other impermeable material, only a small volume of
water will filter through it and enter the aquifer. The thickness of the unsaturated zone is
determined by subtracting the surface elevation from the groundwater level elevation (Freeze &
Cherry, 1979).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 115
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
7.1.4.2 Aquifer recharge and groundwater residence time
The residence time of groundwater refers to the time that groundwater remains within an aquifer
before it is discharged into surface water bodies (Asano et al., 2002). An increase in the
residence time of groundwater will lead to an increase in groundwater salinity (Onstott et al.,
2006). When groundwater is in contact with reactive minerals, ion-exchange will cause the
groundwater to become progressively more saline. Under natural geohydrological conditions,
one can distinguish between groundwater within higher elevated areas and graundwater found
within valley bottoms or lower elevated areas. Groundwater found within the higher elevated
areas will plot within field one or two of the Expanded Durov diagram (Figure 91 and Figure 92),
which represent fresh, recently recharged groundwater that has undergone a minimum degree
of ion-exchange. The groundwater chemistry will therefore closely resemble the chemical
composition of fresh rainwater due to the limited degree of ion exchange. As the groundwater
migrates downstream, an increase in aquifer residence time will lead to an increase in the
degree of ion exchange, which will ultimately lead to higher salinity in the groundwater in lower
areas. A high hydraulic conductivity, good aquifer recharge, and sufficient groundwater
gradients will mobilise groundwater, which will minimise the amount of time that groundwater
spends within an aquifer and will lead to improved groundwater quality.
7.1.4.3 Human effects
Finally, and probably the most noteworthy, are the effects that humans have on the environment
and the composition of naturally occurring groundwater. The effects of mining both coal and gold
and other anthropogenic activities such as the disposal of mine waste often affect groundwater
by adding solutes to naturally occurring groundwater, often resulting in significant problems in
these scarce water resources. The sources and interactions of elements occurring as major
cations and anions in natural waters are summarised in Table 11.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 116
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
Table 11: Characteristics, sources and interactions of elements occurring as major cations and anions in natural
waters (Concentrations from Appelo and Postma, 2005).
Typical freshwater
Element General Source Ion interactions concentration in
mg/I
Pyroxenes, amfibioles,
Absorbed onto clay surfaces,
Most abundant alkali earth metal. feldspars, Carbonate
Calcium often involved with cation- 2to 80
Only one oxidation state mineral, dolomite and
exchange reactions
calcite, gypsum fluoride.
Alkaline earth metal. Only one Pyroxenes, amphiboles, Strongly absorbed onto clay
Magnesium oxidation state. Smaller than Na micas, chlorite, olivine, surface, involved in exchanged 1 to 50
and Ca, therefore more hydrated serpentine, dolomite reactions
Feldspars, clay minerals Absorbed onto clay surfaces,
Abundant alkali earth metal. Only
Sodium such as kaolinite, illite, often involved with cation- 2 to 50
one oxidation state
industrial wastes exchange reactions
Alkali metal, larger than Na.
Found in 1+ oxidation state. Once
Potassium in solution, tends to be K-feldspars, ash Absorbed onto clay su rfaces 0.5 to 80
reincorporated in solid,
weathering state
Feldspars, pyroxenes, also
Second only to oxygen in
Silicon more soluble forms of Absorbed onto mineral surfaces 0.5 to 30
abundance
quartz, such as opal or chert
Non-metal occurring in oxidation Metallic sulphides and
state, often found in high evaporates, e.g. pyrite and Sulfide oxidation, often forms
Sulphur
50,2' 1 to 500 as 50/oxidation state, or reduced gypsum, burning of fuels, complexes
form, 52. smelting of ores.
Subdued chemical behaviour,
Most abundant of halogens. Lower concentration in
does not take part in redox
Volatile element. Chloride ion Cl- rocks compared to other
Chloride reactions, not absorbed, forms 2 to 70
most significant oxidation state. rocks. Found in sedimentary
complex ions with seawater and
75 % of Cl occurs in oceans rocks and evaporttes.
brines at high CI-
May substitute -OH in minerals
Lightest element of halogen Fluorite, amphibole, micas,
Fluoride same size and charge, absorbed $1
group, F- is ion in solution apatite, ash
well onto gibsite
Humans, fertilisers,
In water as NO; NO,' NH: Crustal combustion of fossil fuels,
Nitrogen
% Often indicator of pollution 0.01 to 12 as NO,·rocks contain 25 of Nitrogen waste disposal (CN-),
sewerage.
Apatite, phosphorites.
Co-precipitation and adsorption
Phosphorus In water mainly as PO/ Humans, fertilisers, o to 2 as PO/·
onto minerals
sewerage, insecticides.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 117
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
7.2 Sampling
To determine the origin of groundwater in both mines, sampling was done at strategic positions
at each mine. Samples were taken from surface boreholes as well as underground in each
mine. Sampling at different depths within the mines allows water quality data to be collected for
both mines. The surface view of the sampling positions is shown in Figure 71 and the
underground view of sampling positions is shown in Figure 72. A summary of sampling
positions is displayed in Table 12.
Table 12: Sampling summary.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 118
Cl)
o Surface VI/ater 0 Ground Water
Lcode (Last value measured)
LEGEND
_ Snol UndergrGund Samples
D Sasol &. Harmony Samples from Surface
_ Harmony Undrg:round Samples
N Interaction area
N "'iddelbu~ beundarles
N Harmony boundary
Figure 71: Surface view of the sampling positions.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 119
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes I 2010
Depth below := €=
datum (1828.797)
·3..1~
..:....;:.,.·.S:3~
i.i/ ~':'ï
.~l'M
( 7 \ .. -'ltl - -31'-"[ ~3()'o ev ~~
~I I 1.IO~n ~ ' l'Qu:V-~~~;rn~
Slat -11~", '-- --' ...-..._. _..w".,
~.......ll..09",
l~ ~v-.l:")· .1!....~
-121:1""1 ~ ---=l.lli.:l IJ ~ _,:73-..,
.ll...!II....~"I
.!2u.._~.., ,
.I~"'>J.. ':~ .',1301
~---=.ll..19"",
..:.L~
.J:l.U. .....=ill7 ~. _'l,?~_..,
__ ::..-..!'..l.:-L.. ,~)-, ~.......Iii~"
.JZ.!!L~t ~H''''
.l!..!:L~~ ~~~1"'1
__,us)', ~~1", ~
_ ~15", .131".., p-v
~i
~...2ll)', ~ -'~!""'!"1.qcl.
~~j~~ Groundwaterflow direction in goldmine workings
Figure 72: Underground view of the sampling positions.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 120
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
7.2 Sampling Protocol
The same sampling protocol was applied during each sampling survey. Macro and trace
elements as well as the stable isotopes oxygen-18 to) and deuterium (2H) were sampled for in
this study.
Sampling of the hydro-chemical constituents was done by using new 250 ml plastic sampling
bottles. To avoid contamination, the sampling bottle was rinsed before each sample was taken,
first with sterilised water and then with water from the sampling point. Sample bottles were
sealed air tight and labeled appropriately. After sampling was completed, the samples were
stored in a cool (> 1DOe), dry storage place until delivered to the laboratory.
Oxygen-18 to) and Deuterium (2H) samples were taken in new 250 ml, amber glass bottles to
avoid contamination. The bottles were first rinsed with the sample water, then completely filled
and tightly sealed. The water was not treated in any way before or after sampling. Figure 73
illustrates how the samples were taken.
Figure 73: A sample being collected in the gold mine next 10a fracture.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 121
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
7.3 Data interpretation techniques
In order to characterise the origin of the groundwater within the various aquifers, the following
interpretation was carried out on the hydro-chemical data:
• Standard tri-linear diagrams, i.e. Piper and Expanded Durov diagrams were employed to
investigate the use of these diagrams to distinguish between groundwater originating from
surface or shallow Karoo aquifers or from deeper Witwatersrand aquifers.
• Various X-V column plots of ionic ratios were used to fingerprint groundwater types with
different flow paths.
There are numerous methods of displaying and defining the type of groundwater that occurs
within an area.
7.3.1 Chemical parameters
Although all major elements were analysed for only the elements that are expected to show
concentration increases or decreases within the coal and gold mining environment will be
discussed. All analysed elements will however be studied and discussed if anomalies do occur.
The chemical parameters that will be discussed in the document are the following:
• pH - pH is an important parameter, as it determines the solubility of metals in groundwater.
Most metals are immobile in groundwater in which the pH varies between 5 and 9 units. pH
decreases are also a good indication of the presence of acid rock drainage (ARD) reactions,
which occur when sulphide-bearing minerals such as pyrite are exposed to an oxidising
environment in the presence of groundwater and microbial organisms (Drever, 1997).
• Electrical Conductivity (EC) - the EC of groundwater is a good overall indication of the quality
of groundwater. EC increases are generally linked to pH decreases (Drever, 1997)
• Sodium (Na) - sodium is found in high concentrations in groundwater in the Karoo
Supergroup, due to its marine deposition environment. It is also found in groundwater
systems that have undergone a high degree of sodium ion-exchange (Eby, 2004).
• Sulphate (SO/-) - sulphate contamination is associated with the oxidation of sulphate-
bearing minerals such as pyrite (acid rock drainage reactions - ARD). As mentioned
previously, sulphate contamination will under most conditions lead to a decrease in the pH of
groundwater due to the formation of sulphuric acid (Vermeulen & Usher, 2003).
• Chloride (Cl) - elevated chloride concentrations are generally due to the inherent high
chloride concentrations that exist in some sedimentary rocks. High chloride concentrations
are also associated with old, stagnant groundwater that has undergone a reasonable degree
of ion-exchange (Eby, 2004).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 122
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
• Bicarbonate (HC03-) exists in aqueous solutions containing calcium (Ca2+),
bicarbonate (HC03-) and carbonate (Cot) ions, together with dissolved carbon dioxide
(C02) (Freeze & Cherry, 1979). The relative concentrations of these carbon-containing
species depend on the pH. Bicarbonate predominates within the range 6.36 to 10.25 in fresh
water (Freeze & Cherry, 1979). Most groundwater in a unconfined aquifer is in contact with
the atmosphere and absorbs carbon dioxide, and as these waters come into contact with
rocks and sediments, they acquire metal ions, most commonly, calcium and magnesium.
Therefore, most water near the surface (shallow, unconfined to semi-confined aquifers) can
be regarded as dilute solutions of these bicarbonates (Freeze & Cherry, 1979).
The positions of the sampling points for which groundwater chemistry data is available are
shown in Figure 71. Groundwater quality information was obtained for a total of 21 sampling
points. Only once-oft monitoring was done and therefore only once-oft data is available,
therefore no increasing or decreasing concentration trends can be identified, nor can
explanations for anomalies be given with a high degree of confidence.
7.4 Hydro-chemical results interpretation
The following interpretations were made from the chemical results. It is important to note from
the mining layout and the positions of the samples that only the deep gold mine samples HS, H7
and H10 might be aftected by coal mining, as future mining has not reached the position of the
other underground gold mine samples, as shown in Figure 74.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 123
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes I 2010
LEGEND
Sasol llnderqrourtd Samples
Harmolly Undrground Samples
Gold mine shafts
Middl:lbult shafts
lnteraction area
rulid dI:l bu lt DOUIl dari I:S
~ /\ ruliddelbult Futupe Mining
/\v/ ruliddelbult UJG
Figure 74: Underground sampling positions with the current and future coal mining layout.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 124
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
7.4.1 Electrical Conductivity (EC)
EC values are a good overall indication of groundwater quality and distribution. The EC values
of the sampling points range between 56.2 ms/m (SS) to 529 (S1) with an average value of 265
ms/m Figure 77 plots the EC values for the samples. The samples are plotted shallow to deep,
from left to right, in Figure 76.
EC is an indication of groundwater salinity. Groundwater salinity does not appear to increase
from the shallower Karoo aquifer samples (KB1D to MHR45) towards the deeper Witwatersrand
aquifer samples (H1 to H10) shown in Figure 76, where KB1D (189) shallowest and HS (169)
deepest are in the same order of magnitude, and this is unexpected. It can be attributed to the
following factors:
• Dolerite dykes and faulting (discussed in Structural Geology Section 4.7) are known to
create separate groundwater compartments in the study area (Figure 75). The groundwater
chemistry differs for the different compartments. Different groundwater compartments may
receive less or more recharge than the surrounding compartments, which will lead to an
increase or decrease in the groundwater salinity (Hodgson & Krantz, 1995). This can be
seen in samples MDBCasino (336), MHN47 (156) and MHR45 (449) all of which were taken
from the already water-filled underground coal mine workings. The maximum difference is
293 between MHR45 (449) and MHN47 (156), indicating the difference in chemical reactions
taking place in the different compartments as well as the different recharge potentia Is of the
different compartments (Hodgson & Krantz, 1995).
• Different aquifers and depths of the samples (discussed in Hydrogeology 5.2) within the
study area will lead to different recharge potentials and different chemical reactions taking
place within each aquifer. This can be seen in sample SS (56.2) taken from the shallow
Karoo aquifer compared to sample H1 (488) taken from the deeper Witwatersrand aquifer.
The difference in EC value is 431.8 ms/m between H1 (Witwatersrand) and MHN47 (Karoo).
• Contamination from Tailing Storage Facilities or other contamination sources may also affect
the EC of the samples especially the shallow Karoo aquifer samples taken downstream
and/or on preferred pathways near the gold mine tailings storage facilities. This can be seen
in samples WB1D (89.1) and WB3D (500), sample WB3D is situated downstream and in
close proximity of the Winkelhaak Slimes dam and shows elevated EC values compared to
WB1D situated upstream and unaffected by the Winkelhaak slimes dam (Figure 62).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 125
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
LEGEND
NN Interaction areaMiddelbult boundaries
N Harmony boundary
/'V Middelbult Fault system
/'V Middelbult Dolerite Dykes
11ïïI1ïïl' I
-4000 0 ~OOO 16000
Figure 75: Dykes and faults across the study area.
• The EC values of the gold mine samples H1 (488) and H4 (472) are much higher than the
other gold mine samples. Both these samples were taken from faults intersecting the gold
mine workings, suggesting that their origin is the base of the Karoo or Transvaal
Supergroup, as the faulting pre-dates the deposition of the Transvaal and hence the Karoo
Supergroup (Tweedie ef al., 1986). They were taken at 1380 m below surface (H1) at 1830
m below surface (H4). The Karoo Supergroup is approximately 210 m thick (Figure 21). This
shows that these samples had to travel approximately 1170 m (H1) and 1620 m (H4). As
these samples move along the fault they react with different minerals through the process of
dissolution. The more dissolution takes place, the higher the EC will be.
• H3 (177) was taken from an exploration borehole intersecting the gold mine workings. The
EC value is much lower compared to that of the fault samples H1 (488) and H4 (472). This
suggests that it has been mixed with precipitation as the borehole is directly linked to the
surface and the Karoo aquifer and that the water travelled much faster through the sub-
surface, subduing dissolution with rock minerals and therefore having a lower EC value.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 126
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
If the shallower Karoo Aquifer were well connected with the deeper Witwatersrand aquifer, there
would have been a significant increase of EC from the shallower Karoo aquifer samples to the
deeper Witwatersrand aquifer, which is not the case (Figure 76). In most of the samples, except
for samples H1 and H4, the EC value decreases (Figure 76).
ECvalues shallow to deep
600
500
-- 400EIII.5. 300
u
w 200
100 -Electrical conductivities
0
0 00 o 0 ~;;t; o M M r-- NU"l\Dr-- OM NM'<tU"lM U"l M M U"l ·sV:i: Vl Vl'<t Vl'<t::c::c M ::c ::c ::c ::c ::cco co co co co ra Z a:: ::c~ ~3: 3: 3: u ::c ::cco ~ ~
0
~
Sample Name
Figure 76: EC values of samples collected in this study from shallow (left) 10deep (right).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 127
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes I 2010
LEGEND
Interaction area
MiCildelbult boundaries
~
N Harm ony bound ary
Figure 77: Electrical conductivity values for sa ITllles collected in this study.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 128
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
7.4.2 pH
Groundwater pH appears to be basic throughout the samples in the study and ranges between
6.33 (MHR 45) to 8.8 (H2) with an average pH of 7.73, as shown in Figure 79. The neutralisation
of the pH is due to the naturally occurring marine carbonate minerals in the Karoo Supergroup
which are present in of the study area (Vermeulen & Usher, 2003). The lack of oxygen in the
deeper samples will also contribute to the basic pH (Freeze & Cherry, 1979). Basic groundwater
conditions are a good sign of the absence of Acid Rock Drainage reactions (Freeze & Cherry,
1979). Most metals are immobile with the basic pH values found in the study area. Groundwater
pH does not appear to increase or decrease from the shallower Karoo aquifer (samples KB1D
(7.4) towards the deeper Witwatersrand aquifer samples HS (8), as shown in Figure 78.
• The alkaline pH-level of the Karoo aquifer samples (KB1D, K5D, W1D, WB3D, WB5D, SS,
54, MDBCasino, S1, S3, MHN47, S2 and MHR45) suggests that the Karoo aquifer has
sufficient base potential to neutralise the generated acid at this point in time (Hodgson &
Krantz, 1995).
• The basic pH found in the deeper Witwatersrand aquifer samples (H1-H10) indicates that the
groundwater is deprived of oxygen and therefore basic, as oxygen is a major component for
lowering the pH in Acid Rock Drainage reactions, as shown in Equation 1 to Equation 4
(Freeze & Cherry, 1979).
pH shallow to deep
10
9 -~ r- ..... .....8 ..... " 1\/ ,/ :\./
III 7 '.JIl
CII
:::J 6
nl
> 5
:I: 4
Cl. 3 ~ f-
2
1
0
Cl Cl Cl Cl Cl LJ") 0 .-t Nl ,.... N LJ") 1.0 r-, 0 .-t N Nl q- LJ")
.~-t L~J")
LJ") ~ Z LJ") LJ") q- V) q-.-t Nl LJ") :I: :r: .-t :r: :r: :r: :r: :r:Cl:
3: co co Vi Z :I::r: :r:
3: 3: -uc ~ ~
co
Cl
~ Sample name
Figure 78: pH values of samples collected in this study from shallow (left) to deep (right).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 129
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes I 2010
'Ii . .·-.r' .'t... ".' ,, ~" .s ,. ~~ ~ .;O~ .I'C.' ..Jo" .~~ , • I.: - ,
, '...'.
..... :
."1"
.. '.""',-'.".:.,
=:a"..' ~ .. '.~,." ~» ·.ID,.~~,.·•.•~:I..,",r....."..w.rl.1.fj.~'·lI~~AI :./t" .'.J ':'_:, I.·i~.' "::' J • ~.~~I.'.I... 'jl ~.1 ... • .~~I .,.. ...... ,p.
" : . , ", .l:! "t <~~ ". ~..lP ~ ti. ." '~.' . ...
. tl.~. ~& I ~, .~.';.......' '..al, '. I1iIo. • lo.. . ~ •
..•. .•.' '" . .. <... .~ " 'l. 0........,. -~.r' .. ":a
• " • '\ .. II • i~
•. ..~:. ...! .,""C :~I..~t,~o;~,.~.
' . Y", • ",.'" .~. ~c. ','
, • . ~ :..,..~ J* • '. .,'! . J. ,c.....";" ~ ;'.
_.. '.':' ~, - . .~......0 .:'II", . ...... __ .~ • " .. ~, ','," -~~.... .."'-":, I. .. i.... ~ ~ JI....:~.. ,',,".;'!;;,.~ ..lic .•• ~' .[~ ;,.
'. ,f' .' .. _At . Ot •
.,•. ~J~. ..,. " ......~. UPIA:.l1lU=OrI.In" ~.. .. ', 0 , JJI1 '-':-
.... ;, "; \I , GrnEI .. ' f .1.' .;
' .~r..,.".tt:'· '. ", .' . '.~la.' ~,,_.LU~U~J ~\ol \:i-- G!Iffil' " "'.41" .';, ' ;J ..... ,GlJ'm ' -'"" :' ...
~'11;' '0,. _.... .' -., " " I' - ''':'_~' lo
. ,.. ~. -, ., .... 111:. . . ~ ....... '" Jo. ._.....'... I ".- ,.,~• .. , ~ " ·'r.'" ~0' 1nI!.1.iIIO.J~ ','0 J. .' i
"" '...... t ' ,. . _-.-_. _. .,-- ... LEGEND
•..- ~ . . ~;.r-'uua:n._ ~~1EJ1 "'J. ..~. • ~ "0 "
f. - •, CB. -B. ill ' fIIGHml .:..........1Hl'mI' •~ .: ' ..
'. ._..'t. ,.,',miIll '..
Interaction area
,- _""-- ~~I1 .. .... 0 •• "._
.,... )...,li .... ,.;.,"'io-"~...., ~. ..;.,- ~~I .. · , ~. - ,.', ,:., ~,.~~, ' ......... '\ f ..' _ _~' "_~'"1lJ.illllj~ ,5'::, ... ,qiddelbult boundaries•-~ ~4.., .' "J. ,. ~ .;, "#~~~~~~. .'.~..~I.').,~~d..~ '.~' J.,~'... ". ~l';t~ Harmony boundaryti' ,ii • - i loA..j -.:. •• N,',.• .-. '. .~' ~~'"''ill"' &I '. .'.'" .- ,-.. f. )... ,. Ir' "" _. ~ , 'Ill' ' ., ~. , ".J'.... ~~.~ 'N. •....
0, I ~ ., , fT..' ,ti;. • .. .'_. .- , .. _ 6 '.'. ~ 111.'):;". i.;t"'I'".~ T_.~._.," ,," ~:s \..'.., '.-. . ....., . _.
I .. - ~
, .
:?~....' , .. " .. , .• '"1 ' \
o :Drt~:~W.e.~~r 0 Oron,~W4!~r. ~V~I["~. 6C~";v,=Jr, oe:'JSCI0 04'-" <v!.tr:, 0( 1C,0:':1. v!.lr:.ooi4.c':1 or vl.ln"1o..C,:1
pHtus~ v!.IC' n,uc~~. :A."l 5,241 :2':':<5 ODD ClOOD ODD oOOD
I:!. CM.":""2 I
Figure 79: pH values for samples collected in this study.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 130
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
7.4.3 Sodium (Na)
Sodium is a highly soluble chemical element that reacts freely with groundwater in the sub-
surface. It is also found naturally in groundwater systems (Eby, 2004). It is expected that the
deeper the groundwater moves through the sub-surface, the more enriched in Na it will become
as it comes into contact with more minerals and more dissolution and cation-exchange will take
place (see Equation 6 and Equation 8). The groundwater found in the Witwatersrand aquifer
is also enriched in Na due to the long residence time of the groundwater, resulting in a longer
time for dissolution to take place (Onstott et et.. 2006). The cation-exchange capacity is the
capacity of a medium to preferentially release specific cations while binding other cations in their
place. In groundwater, this process is particularly prevalent in areas of dynamic groundwater
flow, such as the water-filled coal mining area, where calcium is adsorbed and sodium is
released. Dissolution of sodium, potassium and chloride-bearing minerals represents the first
interaction between the rock, hydrosphere and lithosphere following mining activity (Freeze &
Cherry, 1979).
Figure 81 indicates that the deeper the groundwater moves in the study area, the more
enriched in Na-concentration it becomes. The minimum Na-concentration is 91.34 mg/I (SS) and
maximum is 816.13 mg/I (H1), with an average of 394.47 mg/I, as shown in Figure 82. This is an
indication that the groundwater encountered within the deep Witwatersrand aquifer is either
derived from groundwater moving slowly through the sub-surface and reacting with Na-rich
minerals through dissolution and cation exchange until it reaches the end of the groundwater
cycle (Figure 80), or it is paleo-meteoric water that was trapped during the deposition of the
Witwatersrand sediments, which will have a high Na-concentration due to the deposition
environment and relatively long residence times (Onstott et a/., 2006).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 131
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
-- --S-atu-ration
I ItIc
0 I
+ro-' Disequilibrium quilibrium ..
l..- "1-
+c-'
<lJ I
U
c: I
0
U I
t-
(ij
:z J
0 Time
Figure 80: Sodium saturation over time.
• H1 (816) and H4 (717) (water from fault, Table 12) are higher in Na concentration compared
to the other gold mine samples (see also EC). As these samples move along the fault, they
react with different minerals through the process of dissolution and cation exchange. The
more dissolution takes place, the more cation exchange takes place and the higher the Na-
concentration will be.
• H3 (347) (water from a exploration borehole, Table 12) has a lower Na concentration than
samples H1 (816) and H4 (717) from the faults, which indicates that it has travelled at a
much faster rate from the Karoo aquifer to the Witwatersrand, giving it less time for
dissolution and therefore cation exchange.
• MDBCasino (595), S1 (729), S2 (691) and MHR45 (740) are all elevated in Na
concentration. They were samples taken within the coal mine filled with water. Surface water
that drains towards and into the collapsed columns above high extraction areas undergoes
base exchange. Calcium and magnesium are adsorbed onto clays within the sediments, in
exchange for sodium (Eby, 2004). Interstitial release of sodium from the shale occurs as the
water falls down the fractured columns (Hodgson & Krantz, 1995). The degree to which
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 132
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
cation exchange has affected the chemistry of the water is a function of the distance that the
water has travelled and the composition of the rock through which it has moved. The
resultant water that ends up on the coal seam horizon ranges from water almost totally
devoid of calcium and magnesium, to water that may be slightly enriched in sodium
(Hodgson & Krantz, 1995). The conclusion drawn from the samples is that the initial
character of water in coal extraction areas is one of high sodium water.
Na shallow to deep
900.00
800.00
700.00 .". r--
600.00 / 1\ I ~;:::::.
til) 500.00 ,j !\E \
znl 400.00 ,
I \ I 1\ I \
I I \ I \
300.00
200.00 "-,I -
I - ,
100.00
0.00 •~ • • • •
0 0 0 0 0 LI) ~ 0 .-t m r-. N LI) ID r-. 0 .-t LI).-t LI) m Vl Z Vl Vl.-t LI) '<t Vl '<t ::c ::c .-t::c ::c :
N:c :m:c :'<:ct ::c
:..:: :..:: Z Cl:3: al al V«i ::c ::c3: 3: U ~ ~
al
0
~
Sample name
Figure 81: Sodium concentrations for the samples collected in this study from shallow (left) to deep (right).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 133
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes I 2010
LEGEND
%. Interaction area.qiddelbult boundaries
N Harmony boundary
Figure 82: Sodium concentrations of the sarmes collected in this study.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 134
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
7.4.4 Sulphate (SOl-)
The sol concentration ranges from 4 mg/I (H2) to 2759 mg/I (S1), with an average value of
397.77 mg/I, as shown in Figure 84. Figure 83 shows that most of the shallower samples taken
from the Karoo aquifer (KB 5 D, WB3D, MDBCASINO, S1, S3, MHN47, MHR45) have a higher
SO/·concentration than the deeper samples (H1-H10) taken from the Witwatersrand aquifer.
• Samples taken from the Witwatersrand aquifer (H1-H10) were either water dripping from the
mine roof, fault or an exploration borehole. They have a much lower SO/- concentration than
the shallower Karoo aquifer samples. This can be attributed to a reducing environment (lack
of oxygen) that inhibits oxidising reactions with sulphate rich minerals in groundwater, as
explained in Equation 1 to Equation 4.
• The varying SO/- concentration could also be attributed to the varying geology and
compartmentalisation, which in some places mayor may not contain sulphate-rich minerals
with more or less water and oxygen.
• Samples WB3D, MDBCASINO, S1 and MHR45, all taken from the Karoo aquifer, have
relatively high concentrations of SO/- compared to the other samples taken in the Karoo
aquifer. WB3D's elevated concentration can be attributed to its close downstream proximity
to the Winkelhaak slimes dam, whereas elevated concentrations in samples MDBCASINO,
S1 and MHR 45 can be attributed to the oxidation of sulphate minerals within the coal mine
(Vermeulen & Usher, 2003).
• The elevated sulphate concentrations in samples MDBCasino (1024) and S1 (2749) indicate
that the coal seam has a high acid-generating potential (Hodgson & Krantz, 1995). However,
the alkaline pH-level of the samples suggests that the coal seam also has sufficient base
potential to neutralise the generated acid at this point in time. The sulphate concentrations in
the underground gold mine samples (H1-H10) are much lower than those of the shallow
Karoo aquifer samples (Figure 83). This is due to lack of oxygen and sulphate-bearing
minerals such as pyrite in the deeper gold mine inhibiting the reactions shown in Equation 1
to Equation 4. Static acid base accounting tests were conducted on samples of dust on
rocks obtained from mined strata in order to provide a qualitative predication of the net
acidity, a function of the relative content of acid-generating minerals (sulphides) and acid-
consuming or neutralising minerals (carbonates) (Winkelhaak Mine EMPR, 1988). When the
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 135
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
acid-generating capability exceeds the acid-consuming capacity, the pH of the water
decreases, with the consequent mobilisation of sulphates and heavy metals, resulting in acid
rock drainage (Winkelhaak Mine EMPR, 1988).
3000
2500 504 -shallow to deep
.:~::::.
2000
E
o:r 1500
0
III 1000
500
0
0 0 0 0 0 U'l ;1; 0 .-t ('I") l"'- N U'l ID l"'- Q .-t N ('I") o:r U'l.~-t U~'l VI.-t VI VI VI .-t('I") U'l Z o:r ao::r J: J: J: J: J: J: J:~ co co ~Vi
z J:
J: J:
~ ~ U ~co ~
0
~
Sample name
Figure 83: Sulphate concentrations in the study area from shallow (left) to deep (right).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 136
10
LEGEND
Interaction area
I.~iddelbult boundaries
~
N Harmony boundary
Figure 84: Sulphate concentrations forsampJes collected in this study.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 137
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
7.4.5 Chloride (Cl)
The chloride concentrations range from 24.8 mg/I (55) to 1612 (H4) mg/I, with an average value
of 476.9 mg/I. From the linear trend line there is an increase of Cl from shallow to deep. Cl has a
subdued chemical behaviour, meaning it does not take part in redox reactions and is not
absorbed by minerals (Appelo & Postma, 2005).
• Samples KB1D, KB5D and WB3D are elevated in Cl and this can be attributed to the
dissolution of the Cl-rich shales found in the Ecca formation of the Karoo Supergroup, which
contain a relatively high amount of Cl, combined with the fact that the major anion of
rainwater in the study area is Cl (USGS, 2004).
• The elevated Cl concentrations of H1 and H4 can be associated with groundwater that has
undergone a reasonable degree of ion-exchange and has reached the end of the
groundwater cycle (Phillips, 2009). As these samples are from a fault, it was expected that
they would have a high Cl concentration due to high dissolution and ion exchange taking
place as the water travels along the fault.
Cl shallow to deep
1800.0
1600.0
1400.0 r-
1200.0 Il
::::::..
Ol) 1000.0 J 1\ 1 1
E
V 800.0
7 li 1 \ I \
600.0 \ I \, I \ I \
400.0 \ /
200.0 \,/ 1..
0.0 ~ ..-./ ,\v•••• /. nI I I •I
Q Q Q Q Q LI) 0 .-t m r-, N LI) I.D ,..... 0 .-t N m ~ LI)
.-t LI) ;;Ii.-t m VlLI) Z Vl Vl ~ Vl a~:: J: J: .-t J: J: J: J: J:::..: ::..: 3: Z J:al al Vi J: J:
3: 3: 5 s ~
al
Q
~
Sample name
Figure 85: Chloride concentrations in the study area from shallow (left) to deep (right)
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 138
LEGEND
lnteractlon area
.qiddelbult boundaries
~
N Harmony boundary
Figure 86: Chloride concentrations for samples collected in this study.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 139
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
7.4.6 Bicarbonate (HC03-)
HC03- only exists in aqueous solutions containing calcium (Ca2+), bicarbonate (HC03-),
and carbonate (Cot) ions, together with dissolved carbon dioxide (C02) (Freeze & Cherry,
1979). The relative concentrations of these carbon-containing species depend on the pH.
Bicarbonate predominates within the pH range from 6.36 to 10.25 in fresh water (Freeze &
Cherry, 1979).
Most groundwater in an unconfined aquifer is in contact with the atmosphere and absorbs
carbon dioxide. As these waters come into contact with rocks and sediments, they acquire metal
ions, most commonly calcium and magnesium. Therefore, most water near the surface (shallow
unconfined to semi-confined aquifers) can be regarded as dilute solutions of these bicarbonates
(Freeze & Cherry, 1979).
CO2 occurs from atmosphere and is enriched in the soil, as precipitation moves through the
unsaturated zone and CO2 dissolves to form carbonic acid (Equation 13). Detailed reactions
from Freeze and Cherry (1979):
Equation 13
H20 + CO2 <=> H2C03-
Carbonic acid is a weak acid that dissociates to form bicarbonate (Equation 14).
Equation 14
H2C03 <=> W + HC03•
In
Figure 87, the decreasing trend of the bicarbonate concentrations can be seen in the samples
taken from the Karoo aquifer, indicating their relative recharge histories, as more bicarbonate is
in solution in the shallow Karoo aquifer, which is recharged by precipitation, compared to the
deeper Witwatersrand aquifer, which is recharged by influx from the Karoo aquifer. H1 has
o mg/I bicarbonate in solution compared to 52, which has 1585 mg/I bicarbonate in solution.
I) KB1D, KB5D, W1D, WB3D, WB5D, SS, S4, MDBCasino, S1, S3, MHN47, S2, MHR45, HS
and H7 - All have relatively high concentrations of bicarbonate, suggesting that they are
recharged by precipitation or water in the shallow unconfined Karoo aquifer that freely reacts
with the atmosphere. Although HS and H7 were not taken directly from the Karoo aquifer,
their bicarbonate concentration indicates that they are of Karoo aquifer origin.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 140
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
• H2, H3, HS H10 - have higher bicarbonate concentrations compared to the other gold mine
samples if HS and H7 are excluded. H3 is from an exploration borehole intersecting the gold
mine and therefore its bicarbonate concentration is expected to be a mixture of precipitation
and flux from confined aquifers. H2, HS, H10 represent an influx of confined aquifer water as
they were taken from fractures intersecting the gold mine. Their lower bicarbonate
concentrations compared to H3 are attributed to the reverse of Equation 14 in a closed
system (confined aquifer) where there is no CO2 present, resulting in the slow loss of HC03-
(Freeze & Cherry, 197). This explains the lower bicarbonate concentrations in H2 and HS. It
indicates that they are recharged by the shallower Karoo aquifer but that they have had a
long residence time in the sub-surface.
• H1 and H4 - Both these samples were taken from faults, and their bicarbonate
concentrations suggest that they are recharged by influx from a confined aquifer below the
No. 4L coal seam, presumably the deep fractured Karoo aquifer or from the Transvaal
dolomites.
HC03- shallow to deep
1800
1600
1400
~ 1200
E, 1000
t'I'I
0 800
U
:I: 600
400 -HC03-
200
0
0 0 0 0 0 U"I o::t 0 .-i M ,.... N U"I lO ,.... 0 .-i N M o::t U"I
.-i U"I .-i Vl Vl Z Vl VlM U"I zo::t Vl:..: :..: ao:::t :I: :I: .-i :I: :I: :I: :I: :I:3i: co co Vi :I:<l: :I: :I:3i: 3i: U ~co ~
0
~
Figure 87: Bicarbonate concentrations from shallow to deep.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 141
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes I 2010
o Surface Water 0 Ground Water
Bicarbonate [mgJ'l] (Last value measured)
LEGEND
NN. Interaction areaMicfdelbult boundaries
N Harmony boundary
Figure 88: Bicarbonate concentrations for samples collected in this study.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 142
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
7.4.7 Piper diagrams
The Piper diagram is a useful tool in providing a quick comparison of numerous individual
chemical analyses. The Piper diagram allows for both anion and cation compositions to be
represented on a single graph. In the Piper diagram, the ion concentrations are plotted as
percentages, with each point representing a chemical analysis (Piper, 1944).
Plot of water containing: Legend
50% Ca, 35%Mg, 15% Na+K 1::::::::::::::1 Calciwntmagnesium
80% T.AIk. 10% Cl and 10% S04 .... ... bicarbonate waters
E:::3 Sodium bicarbonate!
c::::::I chloride waters
r.·;:.-;:'.I Sodium chloride water.;
~ Calcium/sodium
sulphate waters
Ca Na+K T. Alk. Cl+N03 Ca Na+K T. Alk. Cl+N03
Plotting procedure on the Piper Diagram Examples of typical plotting positions
on the Piper Diagram, for water from
various environments
Figure 89: Piper diagram, adapted from Piper (1944).
7.4.8 Piper diagram for the study area
From Figure 90, samples 51, H4, H1, 53, MHR 45, WB5D and MDB CASINO are grouped
together in the Piper diagram, indicating sodium-sulphate dominated groundwater. Samples
H10, H2, H3 and H5 are grouped together, indicating sodium-chloride dominated groundwater
and samples 55, MHN47 and W1D are grouped together, indicating sodium-
bicarbonate/chloride dominated groundwater.
• Sodium-sulphate dominated samples (51, H4, H1, 53, MHR45, WB5D and MDBCASINO).
All these are shallow Karoo aquifer samples except for H1 and H4, which are samples
collected from faults intersecting the gold mine workings. This indicates that the H1 and H4
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 143
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
origin is from the base of the Karoo aquifer, as all faults are cut off at the base of the Karoo
Supergroup.
o Sodium-chloride dominated samples (H2, H3, H5 and H10). These are samples collected
from the gold mine in the Witwatersrand aquifer. Due to the high sodium and chloride
content they can be regarded as water that has reached the end of the water cycle or water
with relative long residence times that is unaffected by coal mining.
• Sodium-bicarbonate/chloride dominated samples (S5, MHN47 and W1D). These are
samples taken in the Karoo aquifer. These samples are typical of water that has been mixed
with recent precipitation. This is seen in the bicarbonate content. C02 in the atmosphere
combines with water to form carbonic acid, which gives rise to the bicarbonate in the
groundwater, as shown in the reaction in Equation 15.
Equation 15
H20 + CO2«» H2C03 (Eby, 2004)
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 144
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
Piper Diagram
.• 5s4s
o• SjS2
• SI
.•1.1I Rt:r.!
Cabons Anions
'14
• lil
•• o• 210.011
+-C3.oum- -Ch.on~e ....
Figure 90: Piper diagram for the samples collected in this study.
7.4.9 Expanded Durov Diagram
Each field of the Expanded Durov Diagram represents groundwater dominated by different
anions and cations, which in turn represents groundwater that is at different stages of the
hydrogeological cycle. The different fields of the EDD are indicated in Figure 91 and Figure 92,
after which a short description of each field follows. The expanded Durov diagram offers a
distinct advantage over the Piper diagram in that nine groundwater fields display the different
hydro-chemical types and processes on the diagram.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 145
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
Examples of typical plotting positions on the
Expanded Durov Diagram, for water from
various environments
Na
Waste water discharge
Unpolluted Unpolluted Irrigation return now
water water High extraction
underground coal mines
Vanadium extraction
Lime treatment Opencast coal
Gold mine water
of acid water mine water
Power stations
Domestic waste dumps
Seldom Seldom
Natural saline water
found found
Deep mine water
Cl
Figure 91: Interpretations from the Expanded Durov Diagram (1).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 146
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
Exp. Durov Diagram A
<G: '/\V/\ .~.1 2 3
0.-<<)4567 B 9
GIMO'l
• Field 1: Fresh, very clean recently recharged groundwater with HC03- and C03
dominated ions.
• Field 2: Field 2 represents fresh, clean, relatively young groundwater that has started
to undergo Mg ion-exchange, often found in dolomitic terrain.
• Field 3: This field indicates fresh, clean, relatively young groundwater that has
undergone Na ion-exchange (sometimes in Na-rich granites or other felsic rocks), or
because of contamination effects from a source rich in Na.
• Field 4: Fresh, recently recharged groundwater with HC03_ and C03 dominated ions
that has been in contact with a source of 504 contamination, or that has moved
through 504 enriched bedrock.
• Field 5: Groundwater that is usually a mix of different types - either clean water from
Fields 1 and 2 that has undergone 504 and NaCI mixing/contamination, or old
stagnant NaCI dominated water that has mixed with clean water.
• Field 6: Groundwater from Field 5 that has been in contact with a source rich in Na,
or old stagnant NaCI dominated water that resides in Na-rich host rock I material.
• Field 7: Water rarely plots in this field that indicates N03 or Cl enrichment, or
dissolution.
• Field 8: Groundwater that is usually a mix of different types - either clean water from
Fields 1 and 2 that has undergone 504, but especially Cl mixing / contamination, or
old stagnant NaCI dominated water that has mixed with water richer in Mg.
• Field 9: Very old, stagnant water that has reached the end ot the geohydrological
cycle (deserts, salty pans, etc.) or water that has moved a long time and/or distance
through the aquifer and has undergone significant ion-exchange.
Figure 92: Interpretations from the Expanded Durov Diagram (2).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 147
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
Expanded Durov Diagram
1,1g
• WBSD
WB3D
10
• ss
• S4
Na+K o• S3S2
• • Sl• I,IHR~5• .',Ilé7
• unsc SlO
804 •••• H4
• H3
• •o H2o H10
• Hl
Cl
Figure 93: Expanded Durov Diagram for the samples collected in the study area.
From the EDD displayed in Figure 93, certain groupings according to each Field mentioned can
be made:
• Field 3 - (W1D, MHN47, 54, SS, 52, H7 and H6). High extraction coal mining or fresh clean
relatively young water that has undergone a degree of Na- ion exchange, although the
2
reduction of SO4 • at depth can produce H S and CO in the case of Na+ dominant water.
2 2
Water emanating from this field in the study area has most probably undergone ion-
exchange reactions in reducing conditions on its flow path through the Ecca shales. All the
samples except H6 and H7 are samples taken from the Karoo aquifer. H6 and H7 represent
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 148
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
fracture water dripping out of the roof of the mine, indicating that it is of Karoo aquifer origin
and flows along the fracture until it intersects the gold mine workings.
• Field 6 - (WB5D, MHR45, MDBCA51NO and 51). These samples are indicative of water that
has been in contact with a source rich in Na, or old stagnant NaCI dominated groundwater
that has been influenced by Na-Cl host rock/material. As these samples were collected in
the Karoo aquifer. They will have high Na- concentrations through the processes of
dissolution and cation exchange, where calcium and magnesium are adsorbed onto clays
within the sediments, in exchange for sodium (Eby, 2004).
• Field 8 - (WB3D, K1D, K5D and 53). Groundwater that is usually a mix of different types,
either clean water from Fields 1 and 2 that has undergone 504, but especially Cl mixing, or
old stagnant NaCI-dominated groundwater that has mixed with sources richer in Mg. These
are all samples collected in the Karoo aquifer.
• Field 9 - (H3, H4, H1, H2, H10, H5). Very old stagnant water that has reached the end of
the geohydrological cycle or water that has moved a long time and/or distance though the
sub-surface and has undergone significant ion-exchange. End-point water is normally saline
water (old brackish water), which is true for the samples plotted in Field 9.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 149
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
Stiff diagrams
A Stiff diagram plots major cations and anions on opposite sides of a y-axis, with the x-axis
representing the concentration of the ions in meq/I. By plotting the ions in this manner, each
sample will have its own unique geometry (Appelo & Postma, 2005).
STIFF Diagrams
1 10 Hl 3 1"4
:C1C"::~':1 • t:':-:,,:: :':'=~l!1·1;"· .: ;_:1':':~1'1:~C :':-1U:!:'1'1~'"« ,.< :..::10.::3.:1·1::----c ". ,-; 0
l "" c. " :e " "" " :. ".0< I, 5;" 1 ~,. , T .,.
.~ Mn' rT, o~ -~
H5 0 7 K1:l ~:50
':'::1:':~1'1:_-":': ". :!t:.tl.:1-t::-c.: !!1ctlet·':_-':: = :':1C':l:1-1:'- ... ;':lCCJ.:1-t!.-«C 1--'; ",-. C -, .;
" c. l " :.: " =~ " 0,!,j4. " S:l' ~ :5:4- $>' I,
7~ -, moq. " T~ T'
r.1HR45 51 52
:!1C{:;:,' ,:--:c :':,::1:):,-,::-<0;: :'C1C:JZ1'1:'~:: ,~ v='t:~:-'I·1:.·t:- :.r''':i~ 0':'-:::c ,-; c '0< C'l ", c. " e. ", ::.. e. "'0< I, ~ ~,. !.~ , ...,. 1> .,.
Mf:' 0' Ol ~ ~ -!
53 54 55 N1 ::l ~lB 3:J
:':1:tl:l' ,:''-:c ::':CJ.:1-'::-O: :':':Cl:l"~ ';':"::1:1·1~ :":,CCl.:1·'::-'::e ~o( ,,-00.0"( .... .*,.~
W .... c. l 'C, 00 ,. =~ :.i'& .. " ".'" " ill .><
T~ mt:' O~ 7~ :., O~ O! ..._. O~
Figure 94: Stiff diagram for the underground samples collected in this study.
From the Stiff diagrams, we see that some samples are dominated by major anions or cations.
The samples that show a clear dominance are listed below:
• (H1, H4, H10, H2, H3 and HS) - Na+K and Cl dominated groundwater. Samples collected in
the gold mine. H2, H3, HS and H10 geometry is different than that of H1, H4 and H10,
suggesting that they are the same water but affected by different hydro-chemical reactions.
• (HS, H7, S4 and S2) - Na+K dominated groundwater with a high alkalinity. H6 and H7 were
collected in the gold mine whereas S4 and S2 were collected in the water-filled coal mine.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 150
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
• (MDBCASINO, MHR45 and 51) - Na+K and S04 dominated groundwater. Samples collected
in the water-filled coal mine. They will have high Na- concentrations though the processes of
dissolution and cation exchange. Calcium and magnesium are adsorbed onto clays within
the sediments, in exchange for sodium (Eby, 2004).
• (WB3D, K1D, K5D) - Cl and Mg dominated groundwater. Samples collected in the shallow
Karoo aquifer. The processes of dissolution and cation exchange are limited and the calcium
and magnesium has not been adsorbed onto clays within the sediments in exchange for
sodium (Eby, 2004).
7.5 Discussion of Chemical Results
From the chemical parameter plots, Piper diagram, Expanded Durov diagram and Stiff
diagrams, the origin of some of the groundwater samples can be "fingerprinted" according to
their composition. Their origin of the samples can be ''fingerprinted'' from the shallow Karoo
aquifer affected/unaffected by coal mining or from the deeper Witwatersrand aquifer. The
following section discusses the origin of the samples with reference to their hydro-chemical
composition and the different aquifers found in the study area.
7.5.1 Samples taken from boreholes on surface
Table 13: Summary of samples taken from boreholes on surface of the study area.
Borehole
name Y X Z
KB 1 D -2929816.932 7611.298 1591.324 28 30 3.5
KBS D 8886.292 1615.099 28 30 1.02
WBlD -2932446.352 10378.745 1589.468 28 30 1.6
WB3D -2933838.364 10023.214 1599.229 28 30 1.3
WBSD -2933225.999 11669.846 1621.612
Borehole
name y X Z
MHN47 -2935823.037 11478.324 1578.937 110 Into mine 108.19
MHR4S -2935046.371 10963.433 1600.137 130 Into mine 129.44
MDBCasino -2933683.916 15952.46 1581.346 100 Into mine 97.67
m.a.m.s.l- metres above mean sea level
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 151
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
According to the chemical parameters, Piper diagram, Durov diagram and Stiff diagrams, the
surface borehole samples are indicative of Karoo aquifer water that has been affected by both
underground coal and surface gold mining activities in the study area.
• We can distinguish between samples taken directly from the coal mine filled with water
(MHN47, MHR45, MDBCasino) through the chemical parameter of sodium (Na). Dissolution
and cation exchange has taken place in these samples, replacing the Mg and Ca. The Stiff
diagrams also show this characteristic Na+K dominated water, which is indicative of the coal
mine filled with water (Hodgson & Krantz, 1995).
• KB1D, KB5D, WB1D, WB3D, WB5D are samples taken from the shallow Karoo aquifer.
According to their Stiff diagram, they are characteristic of Mg and Cl dominated water. These
samples have been exposed to Cl mixing and are dominated by groundwater that has mixed
with sources richer in Mg, found in the shallow Karoo aquifer. The Cl concentrations can be
attributed to the dissolution of the Cl rich shales found in the Ecca formation of the Karoo
Supergroup, which contain a relatively high amount of Cl, combined with the fact that the
major anion of rainwater in the study area is er (USGS, 2004).
• S04 concentration is the main contributor to ARD and it is thought that Harmony's slimes
dam may contribute to ARD in Sasol's coal mine. From the samples analysed in this study
for S04 concentrations, it is clear that if there is a contribution of S04 concentrations from
Harmony gold mining's slimes dams it is minimal, as Sasol coal mining's samples
(MDBCA5INO and MHR45) are more than double the highest S04 concentrations found in
Harmony's highest sample (WB3D), as shown in Figure 83 and Figure 84.
• KB1D, KB5D, W1D, WB3D, WB5D, 55, 54, MDBCasino, 51, 53, MHN47, 52, MHR45, HS
and H7 - All have relatively high concentrations of bicarbonate, suggesting they are
recharged by precipitation or water in the shallow unconfined Karoo aquifer that freely reacts
with the atmosphere.
• Referring to the EDD and Piper diagram in Figure 93 and Figure 90, a correlation exists
between the surface borehole samples (KB1D, KB5D, WB1D, WB3D, WB5D) and the
underground samples (51, 52, 53, 54, 55, HS and H7) taken from both mines, as they plot
in areas of the same hydro-chemical characteristic. This is a confirmation that there is
dynamic interaction between the aquifers situated in the Karoo Supergroup as there is a
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 152
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
direct link with recharge from precipitation into the shallow perched aquifer and the shallow
weathered aquifer. This water then flows through preferential pathways from these aquifers
into the deeper artificial coal mine aquifer and into the fractured aquifer. Therefore, the
deeper Karoo aquifer is not directly recharged by precipitation but by flux from shallower
Karoo aquifers. Furthermore, Karoo aquifer flux flows into Sasol's underground coal mine
and in some places through fractures into Harmony's gold mine workings, as seen in
samples HS and H7 (fracture water dripping from roof - Table 14) taken from Level 1 at
Harmony's No. 9 shaft, which is in close proximity to the overlying Sasol coal mine workings
(Figure 95).
7.5.2 Samples taken directly from underground
Only HS, H7 and H10 could possibly be affected by coal mining, as shown in Figure 74. They
are the only samples taken where current coal mining is near the gold mine workings.
Table 14: Summary of samples taken directly from underground in this study.
Hl 1340 11 8 Water from fault
H2 1340 11 8 Water next to dyke
H3 1620 15 8 Water from exploration bore hole
H4 1830 18 8 Water from fault
HS 2152 24 8 Water dripping out of roof of stope
H6 282 1 9 Water dripping out of mine roof (minimal)
H7 282 1 9 Water dripping out of mine roof (minimal)
H10 635 8 3 Water dripping out of mine roof (minimal)
S1 104 In coal mine Near Main shaft Borehole into mine workings
S2 113 In coal mine Near North Shaft Water dripping out of mine roof (minimal)
S3 105 In coal mine Block 38 Fissurefrom floor
54 80 In coal mine Near West shaft Water dripping out of mine roof (minimal)
SS SS In coal mine Block 35 Water dripping out of mine roof (minimal)
According to the chemical parameters, Piper diagram, Durov diagram and Stiff diagrams, the
underground groundwater samples can be grouped together. The grouping is as follows:
• (H1, H2, H4, H5 and H10) - The Na- concentrations in these samples (Figure 81) generally
increases with depth, while their SO/- concentration is much more depleted than the
samples taken in the Karoo aquifer (Figure 83) These samples have the same hydro-
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 153
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
chemical characteristic of groundwater expected to be found in deep confined aquifers or
gold mines. The samples are a mixture of old stagnant water (possibly connate water or
paleo-meteoric water) and water that has travelled a long distance in the sub-surface with a
relatively long residence time and has reached the end of the groundwater cycle. H2, H3, HS
and H10 have higher concentrations of bicarbonate compared to the other gold mine
samples, if H6 and H7 are excluded (
• Figure 87).
• H3 is water from an exploration bore hole intersecting the gold mine and therefore its
bicarbonate concentration is expected to be higher, as it has been recharged by a mixture of
precipitation and influx from confined aquifer water. H2, HS and H10 (samples from
fractures) have lower bicarbonate concentrations compared to H3, which is attributed to the
reverse of Equation 14 in a closed system (confined aquifer) where there is no CO2 present,
resulting in the slow loss of HC03- (Freeze & Cherry, 1979). Therefore there are lower
bicarbonate concentrations in HS, H2 and H10. This indicates that they are recharged by the
shallower Karoo aquifer but that they have had a long residence time in the sub-surface.
• H4 and H1 - Both these samples were taken from faults, and their bicarbonate
concentrations suggest that they are recharged by influx from a confined aquifer below the
No. 4L coal seam, presumably the base of the fractured Karoo aquifer or from the Transvaal
dolomites.
• H1 and H4 are unaffected by coal mining (Table 14 and Figure 74). In the Stiff diagram
(Figure 94) they have a much more pronounced Na+K and Cl geometry compared to the
other samples taken in the gold mine. This indicates that H1 and H4 (water from fault)
originated from the base of the Karoo Supergroup and that groundwater in some faults
moves from the base of the Karoo aquifer, unaffected by coal mining, towards the deeper
Witwatersrand aquifer. They are unaffected by coal mining as coal mining has not reached
the area where these samples were taken (Figure 74).
• H3 - This sample was taken from an exploration borehole intersecting the gold mine
workings. It resembles water of Karoo aquifer origin that has travelled in the borehole
through the Ventersdorp Supergroup and reached the gold mine workings. The travelling
time of the sample is less than samples H1 and H4 taken from the faults, as its Na
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 154
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
concentration is much lower than samples H1 and H4, indicating that it has not undergone
the degree of ion-exchange these samples have, due to less time of exposure to minerals
and therefore less travelling time. As coal mining has not reached No. 8 shaft, it has
remained unaffected by coal mining, which is indicated by the EDD (Figure 93). If the coal
mining intersects the exploration borehole it will be sealed (Block 8 Expansion EMPR, 2005).
• 51, 52, 53, 54, SS, HS and H7 have been grouped together in Field 3 of the EDD (Figure
93). They also have the same geometry in the Stiff diagram (Figure 94). These samples are
groundwater expected to be found in high extraction coal mining. They have most probably
undergone ion-exchange reactions in reducing conditions on their flow path through the
Ecca shales. All the samples except HS and H7 are samples taken from the water-filled coal
mine in the Karoo aquifer. HS and H7 are fracture water dripping out of the roof of the mine
(Table 14). Through the EDD (Figure 93) and from their bicarbonate concentrations (
• Figure 87) it can be seen that they are of Karoo aquifer origin, affected by coal mining (water
dripping from roof, Table 14). They were taken from Level1 at Harmony's No. 9 shaft, which
is in close proximity to the overlying Sasol coal mine workings (Figure 95).
• H10 is also fracture water dripping out of the mine roof (Table 14). As it is near the current
coal mining of Middelbult Mine Block 8, it may be affected by coal mining (Figure 74).
According to the hydro-chemical data, however, there is no indication of H10 being
influenced by coal mining.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 155
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
No 4L coal seam 9#
ct!
E
- F•• 1Is
- Gold" .... hfts
Tr.'IYIlII
• V.ntersdo",
Figure 95: Origin of samples H6 and H7 according 10Hydro-chemistry.
7.6 Environmental Stable Isotopes
The use of stable isotope ratios of deuterium (ÓO) and oxygen (Ó18) in hydrogeology can be a
reliable indicator of the origins and movement of surface water as well as groundwater. Stable
isotopes have been successfully used to determine the relative ages of the contribution of "new
water" to groundwater flow in previous studies, such as Saayman ef al. (2003) and Weaver ef al.
(1999).
As water molecules are made up of hydrogen and oxygen atoms, the hydrogen (deuterium) and
oxygen (oxygen-18) stable isotopes can be seen to be of key importance in most groundwater
studies. These isotopes are commonly used to determine the average elevation of recharge for
a particular mass of groundwater. The ratios of deuterium (ÓO) and oxygen (Ó18) in the water
molecule undergo small changes through fractionation processes such as evaporation and
condensation (Fritz ef et., 1980).
The values for the individual ratios are expressed as ó values, which are defined as:
Equation 16
ó= [(RsampIJRref)-1x] 1000 %0
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 156
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
where 1 18 16Rsample and Rref are the isotope ratios eH/ H and 0/ 0) for the sample and a reference
standard respectively. The universal reference standard is the standard mean ocean water
(SMOW). On a graph of deuterium (50) against oxygen (518), with respect to SMOW, all current
precipitation (meteoric water) worldwide plots along a linear regression trend called the world
meteoric water line (WMWL) or the universal meteoric water line (UMWL). Their position on the
line will be determined by geographic, climatologic and physiographic factors. Their isotopic
ratios in groundwater are conservative in the sense that they are not readily altered by
processes in closed systems and in the saturated zone. Even where subsequent changes in
hydrochemistry might occur, the isotopic composition remains diagnostic of the origin of
groundwater. For example, during condensation of the water molecule the lighter isotope has a
higher vapor pressure and hence is enriched in the gas phase, while the heavier isotope forms
precipitation more readily (Figure 96).
D 0
D 0 D 0
Vvatertable ( eo fi ed)
ac ifer
Aquiclude
D 0
Co i ed acui er
DO"" "lig er isotope"
D 0 =" eevier isotope"
Figure 96: Schematic of isotope enrichment/depletion in the vapor and condensation phase.
Equilibrium phase changes affect deuterium and 180 in a constant ratio and their values are
linearly related in rain water according to a general worldwide equation:
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 157
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
Equation 17
50 = 8 x 5180+10 (Fritz ef et., 1980).
Oxygen and hydrogen are in some ways the perfect tracers of water because their
concentrations are not subjected to changes by interactions with the aquifer material (Unesco,
1973). Once underground and removed from zones where fractionation through evaporation and
condensation can occur, the isotope ratios are conservative and can only be affected by mixing.
When precipitation infiltrates to recharge groundwater, mixing in the unsaturated zone smoothes
the isotopic variations, meaning water in the saturated zone has a composition corresponding to
the mean isotopic composition of infiltration in the area. This may differ slightly from the mean
isotopic composition of precipitation, due to the fact that not all precipitation during the year
infiltrates in the same proportion (Unesco, 1973).
Analyses of deuterium (50) and oxygen (518) can be used to identify the probable source of
undergroundwater. If the isotopic composition of undergroundwater plots close to the meteoric
water line in a position similar to that of present-day precipitation in the same region, the water is
almost certainly meteoric (Figure 97). Although the isotopic composition of precipitation at a
particular location is approximately constant, it varies from season to season and from one
rainstorm to another. The variations can be used to identify the season at which most of the
recharge takes place (Orever, 1997).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 158
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
60
GMWL
40 y = 8x + 10
-3: 20
0 •
:!E
(J) 0 •
-Cl~Cl -20E LOCAL RAINFALL LINE
":i::s: y = 6·Z+ 8.4 ..,4 ••-G) -40:::s ~,
G) ,'
C -60
-80
-100
-16 -12 -8 -4 o 4 8
Oxygen-18 (%0 SMOW)
Figure 97: The relationship between 15180and I5Dfor Pretoria rainfall (Talma, 2003).
Surface water bodies such as dams, lakes and pans are enriched in the heavy isotope content
during evaporation (Figure 96). Light isotopes are preferentially transferred to the vapour phase.
The resulting surficial layer enriched in the heavy isotope is then readily mixed into the bulk of
the water body through convective processes (Figure 99). This evolutionary enrichment
produces 8D and 8180 values that lie to the right of the meteoric water line, and plot on an
evaporation line of lesser slopes (usually 4 to 5) and lower d than the GMWL. Groundwater
derived through infiltration from such water bodies will carry this distinctive evaporative isotopic
signal (Gat, 1996).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 159
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
o
O{; ,_
c crr cc s-ec
3
Oxygen 18 (per mil)
Figure 98: Meteoric water lines (adapted from Gat, 1996).
Ó 00/00
H'S Exchange
-:Gas Exchange Exchange withRock Minerals
Fig. 2. Generalized 5180 versus 50 plot showing processes
commonly responsible for deviation from the rainfall line.
(Modified from Fritz et al., 1976.)
Figure 99: Generalised liD versus li180 plot showing the world rainfall (meteoric) line and processes commonly
associated of deviation from the rainfall line (Fritz et al, 1980).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 160
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
Regional groundwater flow in the study area follows a complex system of interconnected,
structurally controlled flow paths, incorporating multiple lithologies with both fracture and matrix
dominated flow (Rison Groundwater Consulting, 2007). The use of oxygen and hydrogen
isotopic composition could prove to be a useful tool to "fingerprint" the origin of groundwater
found in the study area and more specifically the origin of the deeper groundwater found in the
gold mining areas of Harmony Evander's gold mining operations. The isotopes 50 and 5180
could assist in determining the type of water as well as whether any groundwater flux exists
between the mines. The thought process behind this is that if there is any water interaction
between Sasol Mining's Middelbult (Block 8) and Harmony Evander operations, the groundwater
in both mines will have the same relative isotopic composition, or if the contrary is true and there
is no groundwater interaction between the mines, that the relative isotopic composition may
differ.
7.6.1 Application of environmental isotope and hydro-chemistry in this study
Environmental isotopes and hydrochemistry data analysis methodologies were applied in this
study to establish the different groundwater types and hydraulic interconnection between the
shallower Karoo aquifer and the deeper Witwatersrand aquifer.
Specific applications of the environmental stable isotopes 8180 and 80 in this study are:
• To assist structural geological and hydrogeological data with the conceptualisation of the
aquifer systems in the study;
• To determine the existence of mine interflow systems between the coal and the gold mine by
comparing the 8180 and 80 signatures with a hydro-chemical analysis of groundwater
emanating from the underground mining activities of each mine.
In order to test the applicability to interpretation of hydro-chemical and environmental isotope
data analyses methodologies for the two aquifers, the data was interpreted with the following:
Standard groundwater chemistry diagrams: Piper and Expanded Durovand Stiff diagrams are
used to define the general study area's groundwater chemistry characteristics relative to other
groundwater types.
Diagnostic chemistry diagrams: In the format of X-V scatter plots or composition diagrams of
selected pairs of measured parameters. In general, positive correlations are indicated as straight
lines, representing mixing of end-members in different proportions. Clusters of data indicate
similar groundwater types or groundwater fed by the same aquifer.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 161
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
8180 versus 80 diagram: This "fingerprints" the different groundwater in the different aquifers,
by comparing their relative isotopic ratios.
7.6.2 Stable environmental isotope data
Figure 100 represents the 8180 and 80 relationships for groundwater of the study area. Given
the fact that residence times for groundwater in the study area are likely to be spatially variable,
infiltrating rainwater with a Karoo aquifer isotopic "signature" can coexist with groundwater with
a Witwatersrand isotopic "signature". The lack of connectivity within the aquifer(s) is also
suggested by the spatial distribution of the samples within the aquifer system. A rainfall
signature prevails in Karoo aquifer samples taken in the study area (Figure 100). These samples
are affected by a relatively large annual precipitation and apparent short residence times, since
their isotopic signature is very close to that of rainfall. Therefore, most of the shallow
groundwater samples most probably reflect the character of the largest rainfall events during the
previous year. By contrast, the Witwatersrand aquifer samples generally have a different
"signature" to that of rainfall (Figure 100). Apparently, rainfall events have not recharged the
deep aquifer over a short period of time to make a significant contribution to the isotopic
"fingerprint" of the deeper groundwater found in the Witwatersra nd aquifer (Figure 100). In deep
aquifers such as the Witwatersrand aquifer, complex flow patterns originating from the
geological structures often lead to difficult predictions of water origin(s), determination of the
main flow paths, and potential mixing of waters (Orever, 1997). All these uncertainties prevent
an efficient interpretation from isotopic composition data alone. A much more efficient way is to
combine isotopic ratios with the major ionic ratios and depth to "fingerprint" the origin of the
groundwater in the different aquifers.
From the data plotted in Figure 100 the following samples can be "fingerprinted":
• Rainfall selectivity: The singular rainwater sample can be disregarded as it was a once of
sampling event and may vary due to seasonal rainfall events (Orever, 1997).
• The deeper Witwatersrand aquifer samples (H1, H2, H3, H4 and H5) are more depleted in
8180 and 80 than the shallower Karoo aquifer (WB10, WB3D, MHR47, MHN 45, S1, S3, S4,
and S5). This is explained by different intake histories. The Witwatersrand aquifer is not
constantly recharged from precipitation, compared to the shallower Karoo aquifer, which is
recharged by precipitation (Orever, 1997). This means that the Karoo aquifer is generally
enriched in both 8180 and 80 (less negative) compared to the Witwatersrand aquifer.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 162
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
• H1 and H4 are more depleted in 8180 and 8D than the other gold mine samples (Figure 100).
H1 and H4 are samples taken from faults intersecting the gold mine workings (Table 14),
indicating water that has had a fair amount of exchange with rock minerals when it travelled
along the fault from the base of the Karoo and into the gold mine, thus depleting the 8180
and 8D (Figure 99).
• H7 is water dripping out of the roof of the gold mine. Its isotopic "signature" is unrelated to
the other gold mine samples. It plots with samples WB1D, WB3D, MHR47, MHN45, S1, S3,
54 and S5, which are samples taken from the coal mine and the Karoo aquifer. This
indicates that HTs origin is that of the Karoo aquifer affected by coal mining.
• The other samples were difficult to distinguish between as they plot in the same area on the
8180 and 8D relationship isotope plot (Figure 100) with no unique isotopic "signature". They
all plot close to the local rainfall line (Figure 100), suggesting that they have been affected in
some way by precipitation directly or indirectly (mixing) over a period of time.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 163
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes I 2010
• H ..t.. S 10 ,<111-,• KB • WB " " II -~----------------,MOB ,'" :. Local Rainfall line II0 MH , I1:::.. /-,'" :(", I
0 Rain Water GMWL Rapw!:!1.€! ,.,,;;,5' J L~_~~~~~!_+_~·J~- /1
,..,:,
/,
580 (permil) /.,.'~
-8 -7 ,¥"-6 -5 -4 -3 -2 ,r -1 1 2
..J
-5
-10
-15
,,,./
, ,i
, l
,i / .'
, I -20 -
-25
,,' ,82
" -30
Figure 100: 0180 and oD relationship isotope plot for the samples collected in the study.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 164
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
7.6.3 Major ion chemistry and stable environmental isotopes
Various X-Y scatter plots of ionic ratios and/or environmental isotopes are used to ''fingerprint''
18
the different origins of groundwater in the aquifers. One of these plots is the 8 0 versus EC plot.
18
From the 8 0 versus EC plot (Figure 101) it is possible to distinguish between samples
18
originating from the different aquifers. Although no direct relationship between 8 0 and EC
18
exists, combination of the information obtained independently from 8 0 and EC is significant.
18
The 8 0-signature provides information on the recharge whereas EC provides information on
dissolution processes in the aquifer, flow process and path. The Witwatersrand aquifer
18
groundwater clusters around 8 0 and EC values between -4.5 to -6.4 %0 and 200 to 500 mg/I,
respectively, whereas the Karoo aquifer's sample clusters around 180 and EC values between
-2.3 to -5.15 %0 and 89 to 500 mg/I, respectively (Figure 101).
From Figure 101, it is possible to distinguish the following three different groups of groundwater:
• S1, WB 3D, MHR 45 - Groundwater originating from the Karoo aquifer and which is
recharged by rainfall on a fairly constant basis with relative short residence time (Fritz ef al.,
1980).
• H1, H4 - Groundwater originating from the faults with a relative long travel time and
exchange with rock minerals, as seen in Figure 99 (Fritz ef et., 1980).
• H2, H3, HS, HS, H7, H10, S2, S3, 84, SS, KB1D, KB5D, WB1D, WB5D and MHN47 -
Groundwater that has undergone a fair amount of mixing and mineralisation. It is not
possible to determine which aquifer these samples originate from using isotopic
compositions alone. They all plot close to the local rainfall line (Figure 100), suggesting that
they have been affected in some way by precipitation directly or indirectly (mixing) over a
period of time.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 165
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
Oxygen-1S VS EC
\\ !t!f thlt tray! ~d to d!!p!r
j:t¥tater~rand 2qud~f tnrovIfl 1:101ts
• Gold underground ssrrples
• 52 GroUnd\Ht!rtnet h!sund!r:;o.ne miJcin:! d Á Coal underground samples
upos~dto h':f1 imnuntr.fm,n!f!' s!tiDn
KB10 H6 Á Surface ssrrples
H3t~~ H5 ~
M 1'147
• WSlD
·7 ·6 ·5 ·4 ·3 ·2 ·1 o
Oxygen-iS (per mil)
Figure 101: ()180 vs. electrical conductivity plot
7.6.4 Depth and stable environmental isotopes
In Figure 102, the samples were plotted from shallowest (KB1D) 28 m to deepest (HS) 2152 m
18
below surface. Figure 103 show the depleted 0 0 signature of the deeper gold mine samples,
suggesting that water was recharged at a much earlier time with a much longer flow path.
18
• Figure 102 and Figure 103 show the depleted 0 0 signature of the gold mine samples (H1,
H2, H3, H4, HS and H10) with depth. This indicates their long residence and flow paths to
reach the gold mine.
18
• Figure 102 and Figure 103 show that the 0 0 signature for H6 and H7 is in the same order
of magnitude as that of the coal mine samples. They are of Karoo aquifer origin and they are
the closest samples taken to the floor of the No. 4L coal seam. H1 and H4 are also believed
to be of Karoo aquifer origin but are much more depleted compared to H6 and H7. This is
attributed to H1 and H4 originate from a fault that connects the base of the Karoo aquifer
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 166
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
with the Witwatersrand aquifer and therefore a confined aquifer with a long travel time,
18
giving rise to their depleted 0 0 signature (Figure 98).
Oxygen-18 with depth (deeper left to right)
-2
-'Ë
~ -3
..9:
00
"eg:, -4
o~
-5
Sample name
-7 ~-------------------------------------------------------------
-Oxygen-18 (per mil)
18
Figure 102: ~ 0 vs. depth for samples in the study area.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 167
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
Gold mine samples
-7 -6 -5 -4 -3 -2 -1 0 -+-Gold mine
0 samples
H10 500 Qj
>
~f7 .2!
"C
./ c:::s
~ 1000 0...bO
3
0
~ Qj~l---A 1500 ..cEf3 .E
<E~ ....c.:.loo.. 2000 Q,
~ ~~ HS
2500
Oxygen -18 (per mil)
Coal mine samples
-6 -5 -4 -3 -2 -1 o
rv ""_Coal mine samples
-~
Qj
>
4a- .!!!
"C
c:
::s
~ SS o...
~rn~ bO
o3
Qj
..c
r-
t J' Sa- E.Es..:..nr. Q,
S CII
~ ~r"" C
~ --......_ !~a-
--.. MH ~45
..:4a-
Oxygen-18 (per mil)
18
Figure 103: I) 0 vs. depth for underground samples originating in the different mines.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 168
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
7.7 Discussion of isotope results
A rainfall signature prevails in the shallower groundwater samples (KB1D, KB5D, WB1D,
WB3D, WB5D, MDBCasino, 81, 82, 83, 54, 85) taken from the Karoo aquifer in the study area
(Figure 100). These samples are affected by a relatively large annual precipitation and
apparently short residence time, since their isotopic signature is very close to that of rainfall
(Figure 100). Therefore, most of the Karoo aquifer samples most probably reflect the character
of the largest rainfall events during the previous year. By contrast, the deep groundwater
samples generally have a different "signature" to that of rainfall (Figure 100). It is apparent that
rainfall events have not recharged the deep aquifer over a short period of time to make a
significant contribution to the isotopic "fingerprint" of the samples found in the Witwatersrand
aquifer.
18
This means that the Karoo aquifer is generally enriched in both 8 0 and 8D as shown in Figure
100 (less negative).The deeper Witwatersrand aquifer (as seen in samples H1-H5 and H10 in
18
Figure 103) is more depleted in 8 0 and 8D than the shallower Karoo aquifer seen in samples
WB1D, WB3D, MHR47, MHN 45,81,83, 54, 85 and H7 in Figure 103. This is explained by
different intake histories and long travelling times. The Witwatersrand aquifer is not directly
recharged by precipitation compared to the shallower Karoo aquifer, which is directly recharged
by precipitation.
. 18 .
H1 and H4 are the most depleted samples In 8 0 and 8D (Figure 100) as they were taken from
faults it suggest that reactions with rock minerals have taken place in the water as it travelled
18
along the fault depleting its 8 0 and 8D composition (Figure 99).
7.8 Isotopes and Hydro-chemistry conclusions
From the 21 water samples that were collected from different localities (Figure 71 and Figure 72)
widely dispersed over the study area covering a depth range of between 28 and 2152 m below
surface. Three different types of sub-surface water were encountered in the study area and can
be differentiated between.
7.8.1 Karoo aquifer origin samples recharged by precipitation (KB1D, KB5D, WB 1D,
WB3D, WB5D, MHN47, MHR 45, MDBCasino and 81-85).
The shallow Karoo aquifer samples and those affected by coal mine activities (KB1D, KB5D,
WB1D, WB3D, WB5D, MHN47, MHR 45, MDBCasino and 81-85) are generally lower in Na
concentrations (Figure 81) and higher in 804 concentrations (Figure 83) than the deeper
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 169
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
Witwatersrand aquifer samples. All have relatively high concentrations of bicarbonate,
suggesting they are recharged by precipitation or water in the shallow unconfined Karoo aquifer
that reacts freely with the atmosphere. They all plot close to the local rainfall line (Figure 100),
which also suggests that they have been affected in some way by precipitation, directly or
18
indirectly (mixing). They are relatively enriched in 8 0 and 80, suggesting that they have had a
relatively short residence time in the sub-surface (Figure 103).
7.8.2 Witwatersrand samples recharged by influx from shallower Karoo aquifers (H1,
H4, H6, H7).
Samples H1, H4, HS and H7 are also regarded as Karoo aquifer origin samples, based on the
elevated Cl concentrations of H1 and H4, shown in Figure 85 (ion-exchange). The Na-S04
grouping of the shallower samples (51, 53, MHR45, WB5D and MDBCasino) in the Piper
diagram (
18
Figure 90) have an isotopic signature shown in their depleted 8 0 and 80 (Figure 100) and their
unique Stiff diagram geometry (Figure 94), combined with the fact that they were taken from
faults intersecting the gold mine workings linking the base of the Karoo aquifer with the
Witwatersrand aquifer (Table 14).
18
H1 and H4 are the samples most depleted in 8 0 and 80 (Figure 100), as they were taken from
faults. This suggests that reactions with rock minerals have taken place in the water as it
18
travelled along the fault, depleting its 8 0 and 80 composition (Figure 99).
HS and H7 are of Karoo aquifer origin, based on their Stiff diagram (Figure 94) general geometry
being the same as the coal mine samples (51-55) and their EOO plotting position in Field 3 with
the coal mine samples (51-55), which are expected to be found in high extraction coal mining.
These samples have most probably undergone ion-exchange reactions in reducing conditions
on the water's flow path through the Ecca shales. HS and H7 are the samples taken closest to
the No. 4L coal seam (Figure 95). HS, H7 and H10 are also the only samples that could have
been affected by coal mining as coal mining has not reached the area of the other underground
gold mining samples (Figure 74).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 170
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
7.8.3 Witwatersrand aquifer water recharged by influx from shallower Karoo aquifer
with a long residence time and possibly mixed with connate water or paleo-
meteoric water (H2, H3, HS, H10)
Sample H2 (next to dyke), H3 (exploration borehole), HS (out of stope) and H10 (out of mine
18
roof) all show depleted 8 0 and 80 (Figure 100). They have relatively high Na- concentrations
(Figure 81) and low S04 concentrations (Figure 83) compared to Karoo aquifer samples, but
have lower Na- concentrations compared to the H1 and H4 samples (from faults), believed to be
water from the base of the Karoo aquifer. H2, H3, HS and H10 all plot within Field 9 of the EDD
which is very old, stagnant or paleo-meteoric water that has reached the end of the
geohydrological cycle or water that has moved a long time and/or distance though the sub-
surface and has undergone significant ion-exchange. H3's bicarbonate concentrations are a
mixture of precipitation and flux of confined aquifer water. H2, HS, and H10 represent a flux of
confined aquifer water, as they were taken from fractures intersecting the gold mine workings.
Their lower bicarbonate concentrations compared to H3 is attributed to the reverse of Equation
14 in a closed system (confined aquifer) where there is no CO2 present, resulting in the slow
loss of HC03- (Freeze & Cherry, 1979). This indicates that they are recharged by the shallower
Karoo aquifer but that they have had a long residence time in the sub-surface.
7.8.4 Groundwater interaction between the Karoo aquifer and the Witwatersrand
aquifer based on hydro-chemical interpretations
There is interaction between the aquifers, as seen in samples H1 H3 H4 H6 and H7, through
preferred pathways such as faults and fractures. Important is the fact that H1, H3 and H4
interaction occurs without the influence of the coal mine, as coal mining has not reached the
areas where these samples were taken.
H6, H7 and H10 were taken near current coal mining activities. Only H6 and H7 show a
"fingerprint" that is characteristic of a Karoo aquifer affected by coal mining. Therefore, the only
current interaction between the coal mine and the gold mine is at No. 9 shaft, Level 1 (Figure
95). Therefore not all preferential pathways bearing water from the Karoo aquifer have been
affected by coal mining.
It can be stated categorically that very little water was encountered in gold mines- indicating that
very little seeps through from the Karoo aquifer into the Witwatersrand aquifer.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 171
Chapter Seven: Hydro-chemistry and Environmental Stable Isotopes 2010
7.8.5 Limitations of sampling
The following are the limitations of sampling:
• Only once of sampling was done and therefore no trends over time could be established.
• Not all areas within the gold and coal mine were accessible. This was due to flooding of
some compartments within the coal mine and shafts that have been decommissioned
and unsafe within the gold mines.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 172
Chapter Eight: Decant 12010
Chapter Eight: Decant
After mining of the coal and gold in the study area, the mining voids resulting from the mining
will be left to flood with water. Therefore, this study considers the possibility of groundwater
decanting onto the surface and from the deeper situated gold mine into the shallower coal mine.
If decant in the study area will occur it will be a function of:
• Type of mining,
• Topography,
• Depth of mining,
• Interconnectivity of the different mines, and
• Piezometric levels of the different mines and different aquifers.
8.1 Mining and recharge in Sasol's Middelbult (Block 8)
The underground workings of Sasol's Middelbult Block 8 consist and in future will consist of
various layouts and configurations (refer to Section 3.1 Sasol coal mining complex, Figure 11).
These configurations can be broadly classified under board-and-pillar and high extraction
methods. The common characteristic of these mining methods especially high extraction is that
the removal of the coal seam results in the caving of the overlying strata into the mined void.
This disruption of the overlying rock mass has a significant effect on the hydrogeology,
especially where high extraction has been done, resulting in subsidence (refer to Section 6.2.1
Subsidence due to high extraction coal mining, Figure 62)
The overburden covering the No. 4L coal seam consists of Karoo Supergroup formations, which
are sandstone, shale, interbedded siltstone, mudstone and coal of variable thickness
(Vermeulen & Dennis, 2009). The total thickness of the overburden covering the No. 4L coal
seam ranges from 60 m to 180 m, at an average of 90 m (refer to Karoo Supergroup, Figure 22).
Dolerite intrusions in the form of dykes and sills are present over the entire study area (refer to
Section 4.8.1 Karoo age intrusives). Subsidence is most likely to occur where the coal seam is
closest to the surface and where the B4 dolerite sill covering the study area is very thin to non-
existent.
Recharge of between 5 to 7 % can be expected at the subsidence areas. Low recharge of
between 1 to 3 % can be expected where no subsidence occurs, with a tendency towards 1 %
where the B4 dolerite sill is thick (Vermeulen and Usher, 2006).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 173
Chapter Eight: Decant 12010
8.2 Groundwater systems within the coal mine
Three distinct and superimposed groundwater systems are present in the coal mining area of
the study area, as described in (Section 5.2 Aquifer types). They are the upper weathered Karoo
aquifer system, the lower fractured Karoo aquifer system and the artificial coal mining aquifer
system.
8.3 Conceptual decant model for Sasol Middelbult Block 8
After the mining voids have filled up with water, the piezometric level of the artificial coal mine
aquifer will rise with the storage coefficient value of the artificial coal mine aquifer (not the
specific yield) as conditions in the artificial mine aquifer have changed from an unconfined
aquifer to a confined aquifer (Vermeulen & Dennis, 2009). The influx of water from the overlying
aquifers into the mine aquifer will decrease as the two water levels approach each other, as
shown in Figure 104 below (Vermeulen & Dennis, 2009). The lowest point on the surface
through this particular section is at 1566 m.a.m.s.l, while the lowest point on the surface in the
Middelbult Block 8 area is at 1560 m.a.m.s.1. The highest point of the No. 4L coal seam roof
through this particular section is 1505 m.a.m.s.l, while the highest point in the Middelbult Block 8
area is at 1521 m.a.m.s.1. The minimum difference in height between the surface and the No. 4L
coal seam roof in the Middelbult Block 8 area is thus more than 39 m.
For groundwater affected by the coal mining activities and left to fill the mining voids to decant it
will have to overcome more than four bars of pressure (10 m = 1 bar), which is unlikely. In the
unlikely event that the piezometric level from the mine and the water level in the fractured zone
combine, the excess water will seep out as water that is unaffected by underground coal mining
activities (Vermeulen & Dennis, 2009). The average thickness of the overburden above the No.
4L coal seam of the Middelbult Mine Block 8 area is approximately 90 m (refer Karoo
Supergroup, Figure 22), while the average thickness of the overburden covering the Sasol
Secunda mining complex is more than 100 m (Vermeulen & Dennis, 2009).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 174
Chapter Eight: Decant I 2010
(Ij
LEGEND
NA/_ Uorth-South SectionN Interection are.Middelbuit boUnt1anes
".iddelbult Future ".inmg
I~/ "'Iddelbult UfG
Piezon"letric level for coal mine
- 1532 m.a.m.s.1
N s
V"J.. th .. re-cl K.a_r-oo aq ....jfl!f"
LO'W'e'st point
"I56.s ~!lil
1
CEl)14
CU
E
o 15382
c."'O •• s-.cc_.cr
Figure 104: North-south cross-section of the conceptualisation of the different water levels.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 175
Chapter Eight: Decant 12010
8.4 Mining and recharge in Harmony Evander gold mining operations
Evander is a deep-level gold mine, with no opencast or shallow workings. The underground
workings of Harmony Evander gold mining consist of various layouts and configurations (refer to
Section 3.4 Harmony Evander gold mining operations, Figure 14). These configurations are a
function of mine planning and method, and can broadly be classified under conventional
mechanised face and strike gully stoping methods. The overburden covering the viable
Kimberley Reef consist of Karoo super group formations which are sandstone, shale,
interbedded siltstone, mudstone and coal the Transvaal Supergroup consisting of dolomites,
shales and conglomerates and the Ventersdorp Supergroup consisting of basalts.(refer to
Chapter 4: Geology of the Study Area). The total thickness of the overburden covering the
Witwatersrand Supergroup in the study area ranges from 208 m to 2500 m at an average of
1392 m (Figure 105). Faulting is common throughout the study area (refer to Figure 32).
Naturally occurring recharge to the gold mine is considered less than 1 % due to the immense
thickness of the overburden covering the study area. Influx into the gold mine may occur from
the overlying Karoo aquifer through preferred pathways such as:
• Fractures,
• Faults,
• Shafts,
• Exploration boreholes.
Influx may also occur from the water trapped in the Witwatersrand sediments or from connate
water (Onstott ef el., 2006). This recharge is limited and could be depleted due to pumping
activities. Influx into gold mine voids will allow the mine to fill up with water, when the possibility
of decant to surface or the overlying coal mine may arise.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 176
Chapter Eight: Decant 12010
...
u...,
<- •
.i:
NN Interaction areaMiddelbult boundaries
N Harmony boundary
N OVerburden Wits
Figure 105: The interpolated thickness of the overburden covering the Witwatersrand Supergroup.
8.5 Groundwater systems within the gold mine
Two distinct and superimposed groundwater systems are present in the gold mining area of the
study area. They are the fractured Witwatersrand aquifer and the artificial gold mining aquifer,
which are both confined due to the overlying Ventersdorp Supergroup. (refer to Section 5.3
Aq uicludesll mpermeable layers).
8.6 Study area conceptual decant model for the gold mine
If undisturbed, the water level in the Witwatersrand aquifer (confined aquifer) will not rise above
the Ventersdorp Supergroup (lavas) as the Ventersdorp Supergroup acts as an aquiclude
restricting groundwater movement. However, shafts from gold mining operations have intruded
through the Ventersdorp Supergroup, puncturing the Witwatersrand aquifer and effectively
creating a piezometric level within the aquifer. The average thickness of the Ventersdorp
Supergroup is 577 m with a range of 0 m to 1324 m throughout the study area (Figure 106).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 177
Chapter Eight: Decant 12010
NNo Interaction areaMiddelbult boundaries
1EOOO
Figure 106: Interpolated thickness of the Ventersdorp Supergroup.
Earlier reports by Pritchard-Davies (1962), Mulholland (1982), Rison Groundwater Consulting
(2002), Rison Groundwater Consulting (2003) and Rison Groundwater Consulting (2003)
indicate that the original Witwatersrand aquifer water level was at approximately 1533 m.a.m.s.1
pre-mining where exploration boreholes intersected the Witwatersrand aquifer. The average
interpolated thickness of the overburden above the Witwatersrand aquifer has a maximum of
2550 m and a minimum of 254 m at an average of 1400 m. This indicates that the original water
level measured by Pritchard-Davies (1962) is not an actual water level but a piezometric level as
there are aquicludes confining the Witwatersrand aquifer resulting in the piezometric level of
1533 m.a.m.s.1. that is within the Karoo Supergroup.
The piezometric level is the hypothetical surface of the hydraulic head in a confined or semi-
confined aquifer. The piezometric level will be the closest to surface where the confined aquifer
is open to atmospheric pressure or where hydraulic pressure is equal to atmospheric pressure,
which is zero. A conceptual model of piezometric levels is shown in Figure 107. Important to
note is that shafts and exploration boreholes situated on the surface of the study area go
through the Ventersdorp Supergroup and intersect the Witwatersrand aquifer, creating
piezometric levels within the shafts and exploration boreholes, shown in Figure 110 where the
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 178
Chapter Eight: Decant 12010
water levels of No. 6 shaft and No. 10 shaft have risen above the confined water level within the
Witwatersrand aquifer. During the period 1956 to 1969, a dewatering cone developed around
the mine workings and the groundwater level was lowered to a deepest point of 841 m below
surface or 728 m.a.m.s.1. (Rison Groundwater Consulting, 2007). Pumping in one mine affected
the water levels in adjacent mines. Groundwater drawdown therefore occurred at variable rates
at the different mines. The water level in Winkelhaak dropped 305 m between August 1956 and
July 1962. The water level in Bracken dropped 190 m between February 1961 and July 1962
and the water level in Leslie dropped 170 m between February 1961 and July 1962 (Pritchard-
Davies, 1962). Currently, the water levels of No. 10 and No. 6 shaft, where the pumping stations
have been decommissioned for over 10 years (Table 9), is at 736 m below surface or 853
m.a.m.s.1 at No. 10 shaft and at 701 m.a.m.s.1. or 925 m below surface at No. 6 shaft. This
suggests that their piezometric levels have stabilised at these levels as a result of pressure
release in the confined Witwatersrand aquifer combined with the effects that pumping has had
on these two shafts, lowering the water level around the gold mining. Thus the inflow of water is
equal to the outflow in the gold mine if No. 10 and No. 6 shaft are considered.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 179
Chapter Eight: Decant 12010
~!:·r;! :' .....•. 0:.'"lo.
-:: 1'1:5 ~!.:.-..; ~!:~;'r ·t: \'/"11
·,t '.'~~ t ".''::
"':,'r.1!·S 'i-:
'.•u· !l!J ~t-!
"-.~..!...yr.t~.~(~
,: tTt'~ltn:.:,'
r!~-~!J:.!~:: :llt.:I;'!;
;;!;.!.T!: .:1553
T.TJ!
T;;-J
~!~~~~1t!:1.'i"'"rJ r;
::r·mv,.;t!r
Figure 107: Conceptual models for piezometric levels,
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 180
Chapter Eight: Decant 12010
8.6.1 Current conceptual model for decant in Harmony Evander gold mining operations
under pumping conditions
The original water level was recorded at approximately 1533 m.a.m.s.1 (Pritchard-Davies, 1962)
and the current water levels in No.6 shaft and No.10 shaft (Table 9) are at 925 m.a.m.s.1or 1226
m below surface (No.6 shaft) and 853 m.a.m.s.1 or 818 m below surface (No.10 shaft). This
indicates that rate of pumping (outflow) was higher than the rate of recharge (inflow), thus
lowering the piezometric levels in the Witwatersrand aquifer by approximately 608 m in No. 6
shaft and approximately 680 m in No. 10 shaft from the original piezometric level of
1533 rn.a.rn.s.l.
It is not expected that the current Witwatersrand aquifer water level (pumping level) will reach
the coal mine, as the current piezometric level (pumping level) of the Witwatersrand aquifer in
the study area is below 925 m.a.rn.s.l, The lowest point in the Middelbult Block 8 area of the coal
seam floor is at 1450 m.a.rn.s.l, The highest point of the current Witwatersrand aquifers
piezometric level (pumping level) is 925 m.a.m.s.1 (No. 6 shaft). Under current pumping
conditions, the minimum difference of the Witwatersrand aquifer's piezometric level (pumping
level) and the coal seam floor is 525 m.
8.6.2 Conceptual model for decant in Harmony Evander gold mining operations (Post-
closure)
The Witwatersrand aquifer water levels in the Poplar and Rolspruit Project areas situated to the
north-northwest of the study area have been recorded at between 1400 to 1600 m.a.m.s.1 (Rison
Groundwater Consulting, 2007). The project areas of Poplar and Rolspruit are shown in Figure
109. These water levels have been unaffected by mining and pumping and are in the same
order of magnitude as the water level of 1533 m.a.m.s.1 recorded by Pritchard-Davies (1962).
This suggests that:
• The dewatering at the Harmony Evander gold mine shafts has had no influence on the
groundwater levels in the Poplar and Rolspruit area, suggesting some compartmentalisation
within the Witwatersrand aquifer. With reference to Figure 108, it can be seen that there are
major NE-SW trending faults and dykes between the two areas and based on the geological
done by McKnight Geotechnical consulting (2002) evidence it is concluded that these faults
form impermeable barriers and allow selected parts of the basin to be dewatered
independently.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 181
Chapter Eight: Decant 12010
LEGEND
Rolspruit Project area
Goldmine shafts
Major Fault system
Middelbult Dolerite Dykes
Figure 108: Compartmentalisation within the Witwatersrand aquifer resulting from faults and dykes.
• The current water levels measured in the shafts are not water levels but pumping levels and
they can only rise to the piezometric surface of maximum 1533 m.a.m.s.1. found in the Karoo
Supergroup where Sasol Middelbult Block 8 coal mining activities operate which is unlikely
as there has been significant pressure decrease due to mining within the confined
Witwatersrand aquifer. The highest point of the No. 4L coal seam roof in the Middelbult
Block 8 area is at 1521 m.a.m.s.1. (Table 1).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 182
Chapter Eight: Decant 12010
LEGEND
Rolspruit Project area
~ Poplar Project area
N~ .Interaction areaMiddelbult boundaries
Harmony boundary
Goldmine shafts
1€OOO
Figure 109: Poplar and Rolspruit Project areas.
After the mining voids have filled up with water, the piezometric level of the artificial gold mine
aquifer will rise with the storage coefficient value of the artificial gold mine aquifer (not the
specific yield) as conditions in the artificial mine aquifer change from an unconfined aquifer to a
confined aquifer. The influx of water from the overlying aquifers into the mine aquifer will
decrease as the Witwatersrand piezometric level approaches the Karoo aquifers water level in
the gold mine shafts shown in Figure 110. The lowest point of the coal seam floor is at
1450 m.a.m.s.1. Once the water level in the shafts reaches this elevation, the Witwatersrand
aquifer's piezometric level and the Karoo aquifer's water level will become one. As the
piezometric level from the gold mine and the water level in the Karoo aquifer combine, the
excess water will seep out as water that is unaffected by underground gold mining activities
(Vermeulen & Dennis, 2009). For groundwater affected by the gold mining activities and left to
fill the mining voids to decant, it will have to overcome the same four bar of pressure as the coal
mine artificial aquifer, which is unlikely.
This concept is shown in Figure 110. Important to note is that the piezometric level will only rise
in the shafts and not through the Ventersdorp Supergroup, which is an aquiclude (confining)
layer that will restrict vertical groundwater flow. Although the original piezometric level was at
1533 m.a.m.s.l., the Witwatersrand aquifer piezometric level can never rise to 1533 m.a.m.s.1.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 183
Chapter Eight: Decant 12010
There are numerous voids and shafts from gold mining activities greatly reducing the pressure
within the confined Witwatersrand aquifer and thereby significantly lowering the Witwatersrand
aquifer's piezometric level.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 184
Chapter Eight: Decant IZOlO
(t
\lIE eAJ'IJ section
Gold rnI~ shafts
.aT-") eece éOJQ
Coal ..ealTl floor 14&0 to Wi1:waters.ran d aqui1'er
1520 mamsl in study are .. original water level = 1533
10 shaft - 8-53 mams.1
000 w E
Piezolnet..-ic level of gold mine - 1533 m.aros t
1500P tt - $ tr1 ;-_-_:-__""- r - ! c* ~:q~'
1000
-.
ct>
E....
E 500
I I- Witwatersrand aquirer wate,-- 84 Ooiente .,11 ---- Fe",tl. level (top of Wits)- Karoo - Gold min.I "04Lcoel ••• mOl - ea Wet.r l.v.l.Dolent •• ,11"T.rnen.s.v,e,a:llo,.,VhtwetersrendI I
-=00 s
0 16:00
Figure 110: Conceptual model with the interpolated Witwatersrand aquifer water levels.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 185
Chapter Eight: Decant 12010
8.7 Time of gold mine piezometric level to reach coal seam floor
The time for the piezometric level of the gold mine to reach the coal seam floor can be
calculated by using the total volume of voids per level and the total volume of inflow of the gold
mine (Table 15). The voids were calculated using the tons milled by the gold mine and dividing it
by it the specific gravity of most rocks (G ~ 2.65) given by Freeze & Cherry (1979).
Table 15: Harmony gold mine data used to calculate volume of voidslflooding.
Mine Complex Name Harmony gold mine total voids Volume of inflow into Number of shafts
(m3) mine per annum
(m3)
Winkelhaak 28223773.58 1170500 6
Kinross 24454339.62 2284290 2
Leslie 15635471.7 0 1
Bracken 9980000 0 1
Total 78293584.91 3454790 10
8.7.1 Assumptions
Due to lack of data, numerous assumptions had to be made to determine the flooding of the gold
mine. These assumptions are as follows:
• The gold mine only has tonnages for mining complexes and not for individual shafts or
levels. Therefore, percentages of volumes were calculated for each complex according to
the number of shafts and levels. These uniform values were then assigned to each level of
the individual shaft. The values are shown in Table 16. Furthermore the volume of the
individual shaft from the bottom of the coal seam roof to the shallowest gold mine workings
in each gold mine shaft was calculated by using 12 m x 12 m x difference in depth.
• The total volume of groundwater inflows was calculated using data from Harmony Gold
Mining. The estimated groundwater inflow was calculated at 3454790 m3 per annum or
9461 m3 per day (Table 15). This is in the same order of magnitude as calculated by Rison
Groundwater Consulting (2002) and Rison Groundwater Consulting (2003) at 8300 m3 per
day. This roughly equates to 0.5 % of annual recharge, which is the expected range of
recharge. The recharge was calculated by using the entire Evander Basin as the surface
area for recharge. As not all recharge will accumulate in the underground gold mine
workings, this is a very rough estimate.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 186
Chapter Eight:Decant 12010
• All the gold mine shafts are interconnected and will therefore be affected by the same
volume of influx. The assumption was made that the gold mine will fill according to elevation
and not according to level, meaning that the gold mine will fill uniformly and not according to
shaft, where the shaft will fill until the water rises to where the shaft is connected to another
shaft, from where it will flow to the adjacent shaft. This is very difficult to simulate and
requires a very intricate simulation model. Important to note is that this will affect the
different filling rates of the shafts and levels but not the final piezometric level of the gold
mine.
Table 16: Values of individual shafts and levels.
Volumes in m3 from shallowest gold mine
Shaft name Volumes in m3 Volumes in m3to each level gold to coal seam floor
Shaft No. 1 3669090.566 521440.3666 19008
Shaft No. 2 8184894.34 506416.8962 82224
Shaft No. 3 4515803.774 500267.9748 13392
Shaft No. 5 5080279.245 495672.7245 123552
Shaft No. 6 6773705.66 519016.1277 26496
Shaft No. 7 13694430.19 13602126.19 92304
Shaft No. 8 10759909.43 10738885.43 21024
Shaft No. 9 15635471.7 976487.9811 11664
Shaft No. 10 9980000 1106984.889 1679616
Total 78293584.91
8.7.2 Estimate of the piezometric level to reach coal seam floor with a constant inflow
rate
The time for the water level to reach the coal seam is estimated at 22 years and 6 months if
current pumping conditions are ceased and a constant rate of 3454790 m3 is used (Figure 111
and Figure 112). This scenario is unlikely as a rate of inflow will decrease with an increase in
elevation or decrease in gradient according to Darcy's law:
Equation 18
Q = K (dh/dl) A
Q = K (constant) iA (constant)
Therefore
Q cx i
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 187
Chapter Eight: Decant 12010
where:
Q = Rate of inflow
K = Hydraulic conductivity
A = Area
i= Gradient
If K and A remain constant, Q will be directly proportional to i. Therefore a decrease in elevation
will result in a decrease in inflow rate. Thus the time taken for the gold mine to flood will increase
significantly.
Harmony Evander Total Voids
2000
Coal seam foor = 1450 m.a.m.s.l
1500 t::::::::;:;;::::::::::::::::;;;;;.::;;;;;;;;;;;;;;;;;.::;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;:;;;;;~r-7:..==;8293584.91. 22.6
~
E 1000
~
E
e 500
1/1
C
0
-~;;
j 0
III 20 40 60 80 -+- Total volume of
voids vs. elevations
·500 --Milhon~
·1000
Volumeinm3
Figure 111: Volume of voids versus elevations.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 188
Chapter Eight: Decant 12010
Harmony Evander Total Voids
25
Coal seem floor = 1450 m.a.m s.1
20
..
,I
"II 15
II.
s
11
E 10
i=
~Vofumes of water to
5 fill mine vs. time
0
0 20 40 60 80
Volume in m3 Millions
Figure 112: Influx over time.
8.7.3 Estimate of piezometric level to reach coal seam floor applying Darcy's law
If the current inflow rate for the total Harmony gold mine is taken as 3454 790 m3 per annum
(Table 15) the time for Harmony Evander gold mine workings to flood to the coal seam floor of
1450 m.a.m.s.1will be 406 years. The piezometric level of the artificial gold mine aquifer will rise
with the storage coefficient value of the artificial gold mine aquifer (not the specific yield). The
inflow of water from the overlying aquifers into the mine aquifer will decrease as the piezometric
level in the shafts rises, decreasing the gradient and thereby decreasing the inflow rate
(Equation 18). The inflow of water in the shafts will be high in the beginning but will decrease
over time, as shown in Figure 113 and Figure 114.
The time for the water level to reach the coal seam is estimated at 406 years if current pumping
conditions are ceased at an inflow rate of 3 454 790 m3.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 189
Chapter Eight: Decant 12010
Harmony Evander Total Mine Voids
Coal seam floor = 1450 m.a.rn.s.l
",',JV
•• • 406,1300
J,JV •
ui
E ••••iii
E "-IV •
r::::
1/1 ...... •• • Elevations vs. TIme
r::::
0
~ 0.1 •• 10 100 1000
i"jj --~-•-~-5~,Jc•B,I•-+---------------------------------
!G5HG-~--------------------------------
Time in years
Figure 113: Influx with elevations over time.
Harmony EvanderTotal Mine Voids
~ • 406.7086916,
-------:,:_=_~-7-61-+---------------------------..-----t-L-J...j.J.J.84139
~ •
en
E •
"'"Il • V:_I oo .,> +Influx vs. ime
•
• +
0.1 1 10 100 1000
Time in years
Figure 114: Influx over time indicating years for the gold mne to fill with water.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 190
Chapter Eight: Decant 12010
8.7.4 Conceptual scenario when pumping ceases
With reference to Table 9, the following scenario is envisaged for Harmony gold mine once
pumping has ceased:
• Water in No. 1 shaft will fill to level12, from where it will flow to No. 2 shaft.
• Water in No. 3 shaft will fill to level 5, from where it will flow to No. 2 shaft.
• Water in No. 5 shaft will fill to level 22, from where it will flow to No. 2 shaft.
• The water from No. 2 shaft will flow from level22 towards No. 7 shaft level17.
• Water in No. 10 shaft will fill to level 2, from where it will flow to No. 8 shaft.
• Water in No. 9 shaft will fill to level12, from where it will flow to No. 8 shaft.
• Water in No. 8 shaft will fill to level12, from where it will flow to No. 7 shaft.
Therefore, NO.7 shaft will flood to the point where it is interconnected to the other shafts before
further flooding. Only when the other shafts are flooded to the same level, will further flooding in
No. 7 commence. When all the shafts are flooded to the same level (elevation), the piezometric
level in one shaft will fill simultaneously with the piezometric level of the other shafts.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 191
Chapter Nine: Conclusions and Recommendations 2010
Chapter Nine: Conclusions and
Recommendations
9.1 Conclusions
The objective and aim of the study was to determine potential groundwater interaction between
Sasol Mining's Middelbult (Block 8) and Harmony Evander Operations including the combined
impacts on the environment. Tools used to establish whether such interaction exists comprised
geological data, water sample chemical analysis, water sample stable isotope environmental
analysis, and geohydrological information. It is important to list all the various characteristic
indicators identified across various geological lithologies and the different aquifers. Of interest is
to briefly compare the water samples with each other, with thought given to sample location,
depth, mining activity and aquifer. The following overall conclusions are drawn from this
investigation:
The study area is located in the quaternary sub-catchments C24D, C23G and C23H of the
Waterval River Catchment drainage region as shown in Figure 2. Grootspruit, Winkelhaakspruit,
and Trichardtspruit drain over the area and join the Waterval River upstream of the confluence.
The surface elevation varies from 1560 to 1695 m.a.m.s.1. The topography dips towards the
rivers (Figure 5).
Sasol's mines are located in the Vryheid Formation of the Ecca Group, Karoo Supergroup.
These rocks consist primarily of sandstones, shales and coal beds and are extensively intruded
by dolerites of Jurassic age. The coal seams in the Highveld Coalfield are mainly flat to gently
undulating, with a very gentle regional dip to the south. Coal seam topography and distribution,
however, are commonly controlled by pre-Karoo topography on the northern and western
margins of the coalfield. Steeper dips are encountered where seams are situated against pre-
Karoo hills. The dolerites occur both as sills and linear dyke structures that may extend over
tens of kilometres. A large sill-like dolerite structure the B4 sill covers most of the study area.
Harmony's gold mining operations are situated in the Witwatersrand Supergroup, predominantly
consisting of quartzite with subordinate lava, shale and conglomerate. The Kimberley Shale
separates the two units and is marked by a basal unconformity. Immediately above this
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 192
Chapter Nine: Conclusions and Recommendations 2010
unconformity is a conglomerate layer known as the Kimberley Reef, which is mined by
Harmony.
In the major part of the study area, the Witwatersrand Supergroup unconformably overlies the
Archaean Basement. These Witwatersrand sediments are in turn overlain by rocks of the
Ventersdorp Supergroup, Transvaal and Karoo Supergroups. The study area has undergone
extensive structural dislocation and faulting is the most common form of deformation with both
primary and secondary fault-features being observed in the study area below the Karoo
Supergroup. For most of the study area, the Ventersdorp Supergroup farms a thick wedge
between the KaroolTransvaal Supergroup and the Witwatersrand Supergroup. The average
thickness of the Ventersdorp Supergroup is 577 m and ranges from 0 in the southeast to 1324 m
in the central part of the study area.
Early compression tectonics and later extensional tectonics affected both the Witwatersrand and
Ventersdorp Supergroups in the Evander Basin, but pre-dated the deposition of the Transvaal
Supergroup and hence the Karoo Supergroup (i.e. >2500 Ma). This resulted in all major fauhing
being "cut-ott" at the base of the Karoo Supergroup and Transvaal Supergroups.
There are seven different aquifer types within the study area:
• shallow perched Karoo aquifers (unconfined);
• shallow weathered zone Karoo aquifers (semi-confined);
• deep Karoo fractured aquifers (zone below the weathered zone & confined);
• coal mine artificial aquifer (confined);
• Transvaal dolomitic aquifer (confined);
• gold mine artificial aquifer (confined);
• deep Witwatersrand aquifer (confined).
There is dynamic interaction between the aquifers situated in the upper Karoo Supergroup as
there is a direct link with recharge from precipitation into the shallow perched aquifer and the
shallow weathered aquifer. This water flows through preferential pathways from these aquifers
into the deeper artificial coal mine aquifer and fractured Karoo aquifer. Therefore, the deeper
Karoo aquifer is not directly recharged by precipitation but by influx from shallower Karoo
aquifers. The water is prevented from moving further downward at the base of the Karoo
Supergroup by the Dwyka Tillite, and further down by the Transvaal dolomites and the
Ventersdorp Supergroup lavas that separate the Karoo aquifer affected by coal mining from the
Witwatersrand aquifer affected by gold mining.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 193
Chapter Nine: Conclusions and Recommendations 2010
Recharge in the coal mine of between 5 and 7 % can be expected at the subsided areas. Low
recharge of between 1 and 3 % can be expected where no subsidence occurs, with a tendency
towards 1 % where the B4 sill is thick.
Recharge in the gold mine of a maximum of 1 % can be expected in the gold mine due to the
thickness of the overburden ranging from 208 m to 2500 m at an average of 1392 m and the lack
of water-bearing structures encountered in this study. Calculated recharge from data of the gold
mine was in the range of 0.5 %.
For the purposes of this study, the Karoo aquifers were regarded as one significant aquifer
situated above the Ventersdorp Supergroup and the deeper Witwatersrand aquifer situated
below the Ventersdorp Supergroup was regarded as the other significant aquifer (refer to Figure
36).
Thus there are two major aquifers dominating the geohydrology of the study area namely:
• The Karoo aquifer (unconfined to confined),
• The deeper Witwatersrand aquifer (confined).
The shallower Karoo aquifer is regarded as an unconfined to confined aquifer and its recharge
can be linked directly to recharge by precipitation (1-3 % of annual precipitation and 5 % to 7 %
where high extraction mining has occurred). The deeper Witwatersrand aquifer is regarded as a
confined aquifer that has no direct link of recharge to precipitation and has a very low flux from
shallower Karoo aquifers though preferential pathways or where the Witwatersrand Supergroup
sub-outcrops against the base of the Karoo Supergroup (>1 % of annual precipitation).
The No.4L coal seam mined by Sasol is situated at an average depth of 80 m below surface in
the south and 150 m below surface in the north, or at elevations ranging between 1450 m.a.m.s.1
and 1521 m.a.m.s.1 throughout the different compartments resulting from the different dolerite
sills situated throughout the study area (Figure 54). The Kimberly Reef containing gold and
platinum group metals at varying depths below surface is currently being mined, the deepest
being ± 2500 m below surface at Harmony No. 8 shaft and the shallowest being 240 m below
surface at Harmony No. 3 shaft.
Currently, Sasol Middelbult Block 8 is only mining over Harmony gold mining in the
southemmost parts of the study area, with the largest area covering the southeastern part
(Figure 53). Future mining will cover most of the study area except for the central part of the
study area around Harmony gold mining No. 7 shaft and No. 8 shaft (Figure 53).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 194
Chapter Nine: Conclusions and Recommendations 2010
Sasol coal mining will mine in future near Harmony's No. 1 Shaft, where the No. 4L coal seam
floor is situated at 1479 m.a.m.s.1 or 139 m below surface (Figure 54). Harmony gold mining's
mining operations at NO.1 shaft, closest to the No. 4L coal seam, are situated at Level 2, which
is 1347 m.a.m.s.1 or 270 m below surface. The minimum difference between the No.4L coal
seam and the gold mine workings at NO.1 shaft is thus 132 m (Figure 56). As yet, no
groundwater interaction from the coal mine into the gold mine could be established in the No. 1
shaft area, as no preferential pathways could be confirmed.
Sasol is currently mining near Harmony's No. 3 Shaft where the No. 4L coal seam floor is
situated at 1480 m.a.m.s.1 or 136 m below surface. Harmony gold mining's mining operations at
NO.3 shaft, closest to where the No. 4L coal seam is situated at Level 1, which is 1387 m.a.m.s.1
or 229 m below surface (Figure 54). The minimum difference between the NO.4L coal seam and
the gold mine workings is thus 93 m at No. 3 shaft (Figure 57). As yet, no groundwater
interaction from the coal mine into the gold mine could be established in the No. 3 shaft area, as
no preferential pathways could be confirmed.
Sasol is proposing to mine in future near Harmony's No. 9 Shaft, where the No. 4L coal seam
floor is situated at 1461 m.a.m.s.1 or 126 m below surface (Figure 54). Harmony gold mining's
mining operations at NO.9 shaft, closest to the No. 4L coal seam, are situated at A Level, which
is 1380 m.a.m.s.1 or 207 m below surface. The minimum difference in metres between the No.4L
coal seam and the gold mine workings is approximately 81 m at No. 9 shaft (Figure 58). Two
preferential flow paths bearing water at No. 9 shaft (Level 1) directly underneath Level A (as
shown in Figure 55) were found in this study (Table 5). These fractures could link the Karoo
aquifer affected by coal mining to the gold mine workings in the Witwatersrand aquifer.
Sasol is currently mining near Harmony's No. 10 Shaft, where the No. 4L coal seam floor is
situated at 1470 m.a.m.s.1 or at approximately 119 m below surface. Harmony gold mining's
closest mining operations at NO.10 shaft to the No. 4L coal seam is situated at (1 level) which is
1345 m.a.m.s.1 or 244 m below surface. The minimum difference in metres between the No.4L
coal seam and the gold mine workings is approximately 125 m (Figure 59). As yet, no
groundwater interaction from the coal mine into the gold mine could be established in the No. 10
shaft area, as no preferential pathways could be confirmed.
The potential for groundwater interaction resulting from Harmony Evander Tailings Storage
Facilities (TSF) found on surface can be attributed to the extent of subsidence resulting from the
coal mining directly underlying or near the TSF. Currently, there is coal mining undemeath the
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 195
Chapter Nine: Conclusions and Recommendations 2010
Winkelhaak and Leslie slimes dams, with proposed future coal mining underneath the Kinross
slimes dam, shown in Figure 60.
The Kinross slimes dam is situated at the centre of the study area with Sasol future mining
underneath it as seen in Figure 60. No subsidence is expected underneath the Kinross mine as
seen in Figure 62 and Figure 63. This is a result of no high extraction coal mining directly
underneath or near the Kinross slimes dam (Figure 63).
Subsidence is likely to the south-southwest of the Leslie slimes dam as high extraction coal
mining will occur and it is expected that preferential pathways (fractures) linking the slimes dam
with the underground coal mine working may be induced (Figure 64). This is only a assumption
and further monitoring is required to confirm this.
Subsidence is most likely to the southeast of the Winkelhaak slimes dam, as high extraction
coal mining will occur and it is expected that preferential pathways (fractures) linking the slimes
dam with the underground coal mine working may be induced (Figure 66). This is only a
assumption and further monitoring is required to confirm this.
There is thought to be an exploration linking the Winkelhaak slimes dam with the coal mining
underneath the slimes dam. This borehole could not be confirmed as the coal mine is already
flooded in this area.
The Ventersdorp Supergroup lavas appear to thin towards the south-southeast of the study
area, with the exception of the southern boundary of the study area where the Ventersdorp
Supergroup lava is either very thin or sub-outcrops against the Karoo Supergroup, as seen in
Figure 67 and Figure 68. It was expected that groundwater interaction would occur here, but
from field work done no groundwater interaction was found here.
Samples were collected from all water-bearing structures intersecting the gold mine and these
are summarised in Table 7. It should be noted that these points are the only water-bearing
structures found in the study and therefore the only known pathways connecting the Karoo
aquifer with the Witwatersrand aquifer. There may be more structures in other parts of the mine
but due to safety, inaccessibility and mining activities it was not possible to locate them.
• H1 and H4 - Water from faults intersecting the gold mine. The origin of this water is from the
base of the Karoo aquifer where the Transvaal Supergroup (aquiclude) is absent and where
the faults went through the Ventersdorp Supergroup (aquiclude) and into the Witwatersrand
aquifer. Groundwater will move along these faults until they intersect the gold mine
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 196
Chapter Nine: Conclusions and Recommendations 2010
workings. The position of the faults are at Evander No. 8 shaft, 11 level 1340 m below
surface (Figure 42) and at Evander No. 8 shaft, 18 level 1830 m below surface. It unlikely
that the coal mining will reach the area where these faults are (Figure 53) and therefore it is
unlikely that the coal mine will have an impact on the gold mining through these specific
faults (Table 7).
• H2 - Water from dyke and exposed in the gold mine workings. The origin of this sample will
be where fractures next to the dyke allowed for groundwater flow (Cook, 2003). These
fractures mayor may not be directly connected to surface (Cook, 2003) and therefore
recharge may occur in the sub-surface at different depths through the influx of overlying
aquifers or directly by surface precipitation from surface. The position of the fractures next to
a dyke was encountered at Evander No. 8 shaft, Level 11, approximately 1340 m below
surface, as shown in Figure 41. It is unlikely that the coal mining will reach the area where
this dyke is encountered (Figure 53) and therefore it unlikely is that the coal mine will have
an impact on the gold mining through this specific dyke (Table 7).
• H3 - Water from an exploration borehole intersecting the gold mine workings. According to
Middelbult Mine, Block 8 Expansion EMPR (2003), all gold mine exploration boreholes will
be sealed off with a non-reactive material to prevent the drainage of groundwater from the
coal mine to the gold mine. (Figure 53). It unlikely that the coal mining will reach this area
where this borehole is situated, therefore coal mining will have no impact on the gold mining
through this specific exploration borehole. The position of the exploration borehole is at No.
8 shaft, Level 15, 1620 m below surface (Table 7).
• H5, HS, H7 and H10 - Water dripping out of the mine roof, the origin of which is difficult to
determine. The origin of these samples is thought to be flux from fissures/fractures and joints
from overlying aquifers. The position of the samples can be seen in Table 7. All of these
positions can be affected by coal mining except H5 which is at Harmony No. 8 shaft and will
remain unaffected by coal mining as no coal mining will occur around No. 8 shaft (Figure
53). HS and H7 were taken near current coal mining activities and show a "fingerprint" that
is characteristic of a Karoo aquifer affected by coal mining. Therefore the only current
interaction between the coal mine and the gold mine is at No. 9 shaft, Level1 (Figure 95).
9.1.1 Isotopes and hydro-chemistry conclusions
The 21 water samples were collected from different localities (Figure 71 and Figure 72) widely
dispersed over the study area and covering a depth range of between 28 and 2152 m below
surface. Three different types of sub-surface water were encountered in the study area and can
be differentiated between.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 197
Chapter Nine: Conclusions and Recommendations 2010
9.1.1.1 Karoo aquifer origin samples and recharged by precipitation (KB1D,
KB5D, WB1D, WB3D, WB5D, MHN47, MHR 45, MDBCasino and 51-SS
Samples from the shallow Karoo aquifer and those affected by coal mine activities (KB1D,
KB5D, WB1D, WB3D, WB5D, MHN47, MHR 45, MDBCasino and 51-55) are generally lower in
Na- concentrations (Figure 81) and higher in S04 concentrations (Figure 83) than the deeper
Witwatersrand aquifer samples. All have relative high concentrations of bicarbonate, suggesting
they are recharged by precipitation or water in the shallow unconfined Karoo aquifer that freely
reacts with the atmosphere. They all plot close to the local rainfall line (Figure 100) this also
suggests that they have been affected in some way by precipitation directly or indirectly
18
(mixing). They are relatively enriched in 8 0 and 8D, suggesting they have had a relative short
residence time in the sub-surface (Figure 103).
9.1.1.2 Witwatersrand samples recharged by influx from shallower Karoo aquifers
(Hl, H4, H6, H7)
Samples H1, H4, HS and H7 are also regarded as Karoo aquifer origin samples based on H1
and H4 elevated Cl concentration shown in Figure 85 (ion-exchange), Na-S04 grouping with the
shallower samples (51, 53, MHR45, WB5D and MDBCasino) in the Piper diagram (
18
Figure 90), their isotopic signature showing their depleted 8 0 and 8D (Figure 100), their unique
Stiff diagram geometry (Figure 94) combined with the fact that they were taken from faults
intersecting the gold mine workings linking the base of the Karoo aquifer with the Witwatersrand
aquifer (Table 14).
18
H1 and H4 are the samples most depleted in 8 0 and 8D (Figure 100), as they were taken from
faults. This suggests that reactions with rock minerals has taken place in the water as it travelled
18
along the fault, depleting its 8 0 and 8D composition (Figure 99).
HS and H7 are of Karoo aquifer origin based on their Stiff diagram (Figure 94) general geometry
being the same as the coal mine samples (51-55) and their EDD plotting position in Field 3 with
the coal mine samples (51-55) which are expected to be found in high extraction coal mining.
These samples have most probably undergone ion-exchange reactions in reducing conditions
on the water's flow path through the Ecca shales. HS and H7 are the samples taken closest to
the No. 4L coal seam (Figure 95). HS, H7 and H10 are also the only samples that could have
been affected by coal mining, as coal mining has not reached the area of the other underground
gold mining samples (Figure 74).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 198
Chapter Nine: Conclusions and Recommendations 2010
9.1.1.3 Witwatersrand aquifer water recharged by influx from shallower Karoo
aquifer with a long residence time and that may have mixed with connate
water or paleo-meteoric water (H2, H3, HS, H1D)
Sample H2 (next to dyke), H3 (exploration borehole), H5 (out of stope) and H1D (out of mine
18
roof) all show depleted 8 0 and 8D (Figure 100). They have relative high Na- concentrations
(Figure 81) and low S04 concentrations (Figure 83) compared to the Karoo aquifer samples, but
have lower in Na- concentration compared to the H1 and H4 samples (from faults), believed to
be water from the base of the Karoo aquifer. They (H2, H3, H5 and H10) all plot within Field 9 of
the EDD, which is very old, stagnant or paleo-meteoric water that has reached the end of the
geohydrological cycle or water that has moved a long time and/or distance though the sub-
surface and has undergone significant ion-exchange. H3's bicarbonate concentrations are a
mixture of precipitation and flux of confined aquifer water. H2, H5, and H10 represent a flux of
confined aquifer water as they were taken from fractures intersecting the gold mine workings.
Their lower bicarbonate concentrations compared to H3 are attributed to the reverse of Equation
14 in a closed system (confined aquifer) where there is no CO2 present, resulting in the slow
loss of HC03- (Freeze & Cherry, 1979). This indicates that they are recharged by the shallower
Karoo aquifer but that they have had a long residence time in the sub-surface.
9.1.1.4 Groundwater interaction between the Karoo aquifer and the
Witwatersrand aquifer based on hydro-chemical interpretations
There is interaction between the aquifers, as seen in samples H1 H3 H4 H6 and H7, through
preferred pathways such as faults and fractures. Important is that H1, H3 and H4 interaction is
occurring without the influence of the coal mine as coal mining has not reached the areas where
these samples were taken.
H6, H7 and H10 were taken near current coal mining activities. Only H6 and H7 show a
"fingerprint" that is characteristic of a Karoo aquifer affected by coal mining. Therefore the only
current interaction between the coal mine and the gold mine is at No. 9 shaft, Level1 (Figure 95)
and not all preferential pathways bearing water from the Karoo aquifer have been affected by
coal mining.
9.1.2 Summary of possible interaction areas
Based on the mining activities, geology, hydrogeology, hydro-chemical and isotope
interpretations in this study, the following can be stated. Groundwater interaction will possibly
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 199
Chapter Nine: Conclusions and Recommendations 2010
occur between the shallower coal mine and the deeper gold mine through preferential pathways
made by the coal mine in the areas listed below:
• The south-southeast corner of the Winkelhaak slimes dam, due to high extraction coal
mining;
• The south-southwest corner of the Leslie slimes dam, due to high extraction coal mining;
• South-southwest of Harmony Evander's No. 9 shaft, Level1, as seen in samples HS and
H7. Although the current interaction is minimal as the water dripping out of the roof is
less than one litre a day it may increase as future coal mining extends to a much larger
area around the shaft and the coal mine fills up with water;
Interaction occurs in the central part of the study area near Harmony Evander's No. 8 shaft
without the presence of coal mining, through faults, fractures next to dykes and exploration
boreholes intersecting the gold mine workings.
Thus the coal mine is not the only contributor to the interaction between the mines, as naturally
occurring structures allow for influx from the Karoo aquifer into the gold mine workings in the
Witwatersrand aquifer. There is no future coal mining envisaged around No. 8 shaft and
therefore the potential for groundwater in the Karoo aquifer affected by coal mining to interact
with gold mine around the No. 8 shaft is very low. 1
The coal mine will in future cover most of the gold mining area except for the central part around
Harmony Evander No. 8 and No. 7, it is expected that groundwater affected by the coal mining
in the Karoo aquifer will only move along naturally occurring structures that links the two mines,
from the base of the Karoo aquifer into the gold mine. This can only occur when structures that
links the Karoo aquifer with Witwatersrand aquifer is encountered, which was found seldom in
this study.
The possible interaction can therefore directly link to the number of structures that are water-
bearing in the study area as well as the area covered by coal-mining activities. From the field
work done in the underground gold mine, the number of structures that are water-bearing is the
exception rather than the norm. Therefore the potential for groundwater interaction from future
coal mining into the gold mine will be low as the Karoo aquifer is very poorly connected to the
Witwatersrand aquifer based on:
• Frequency of water-bearing structures encountered within the study area;
• Travel time of the Karoo aquifer water to reach the Witwatersrand aquifer;
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 200
Chapter Nine: Conclusions and Recommendations 2010
• Water level in the Karoo aquifer remains relatively constant;
• Rate of inflow in the gold mine remains relatively constant;
• Recharge in the gold mine is estimated at 0.5 %, which is less than 1-7 % of that in the coal
mine;
• Water-bearing structures encountered within the study area have low water volumes;
• Prevention of interaction measures of each individual mine.
9.1.3 Expected impact of coal mine on the water quantity and quality within the study
area
The amount of groundwater from surface into the coal is 1-3 % of annual recharge (Vermeulen &
Usher, 2006). A higher proportion of infiltration of 5-7 % may occur in areas where the natural
permeability is increased, such as the increased fracturing associated with high extraction
mining.
The rate of recharge if subsidence occurs under the gold mine slimes dams may therefore be
regarded as 5-7 % if preferential pathways are created with a direct link to the slimes dams. The
quality of this water will be poor as the ARD has been extensive in the slimes dams.
Recharge of the deep aquifer occurs from the shallow aquifer through permeable fracture
systems, faults and exploration boreholes that link the two aquifers. Induced fracturing
associated with mining can extend from the deep aquifer up to the surface, and provides a
relatively direct and highly permeable recharge route.
Groundwater draining into the mine workings will initially be of a good quality. The pH will be
alkaline due to the presence of bicarbonate species. However, once the groundwater reaches
the mine, the material that it comes into contact with will influence its quality.
The following sequence of chemical reactions is assumed to occur during the operational phase:
• The water seeping into the coal mine will generally be of good quality, except for suspended
solids present. Most, if not all of the water resulting from operations, will be used during the
operational phase. Isolated areas of water may be present, and will drain to the lowest point
of the mine.
• The water present will be alkaline, but the Total Dissolved Solids/EC content will increase
due to the contact with the coal floor/pillars.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 201
Chapter Nine: Conclusions and Recommendations 2010
• Groundwater will continue to percolate through the roof downward to the saturated areas.
This will lead to the mixing of initially alkaline to neutral groundwater with relatively stagnant,
alkaline groundwater on the mined horizon.
• The water liberated in the stratigraphical units above pillar extraction panels, as well as the
water seeping into the mine, will generally be of good quality, but contain suspended solids,
in the form of sediment and carbonaceaus material, e.g. shale and coal.
• Very little, if any, pyrite oxidation will take place in this saturated zone. The quality of this
groundwater will initially remain constant, providing that water is not continuously pumped
out of this zone.
• Pyrite oxidation will commence in the unsaturated areas of the mined void (if any), as well as
in the unsaturated zones of the gaaf. The rate of oxidation will vary considerably, depending
on the rate of ingress of groundwater (the residence time), the rate of ingress of oxygen and
the contact areas available for oxidation.
• This initial acidification will be neutralised by the natural buffering capacity in the overlying
rock. This will take place for many years, until all the neutralising potential is depleted.
Isolated areas of buffering depletion might take place quickly, but regional acidification in the
total gaaf area will not occur for many years.
• Groundwater will percolate through the unsaturated gaaf areas downward to the saturated
mined void. This will lead to the mixing of initially alkaline to neutral groundwater with relative
stagnant, alkaline groundwater on the mined horizon.
• Isolated areas of buffering depletion will take place. This will lead to the formation of acidic
conditions in the gaaf area, with low pH groundwater that will percolate downwards to the
saturated zones.
The following sequence of chemical reactions is most likely to occur post-closure of the coal
mine:
• Pyrite oxidation on the mined horizons will be extensive due to the slow flooding during the
post-closure phase.
• Initial acidification will be neutralised by the natural buffering capacity in the coal seam, as
well as from groundwater flooding the sections. This will take place over many years, until all
the neutralising potential is depleted.
• Poor quality groundwater will be present in the mine horizon when total flooding is
completed.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 202
Chapter Nine: Conclusions and Recommendations 2010
9.1.4 Expected impact of gold mine on the water quantity and quality within the study
area
The impact will consist of two aspects:
• The impact of deep level mining on the groundwater within the workings;
• The impact of gold mining and related activities on the Karoo aquifer.
Recharge into to the gold mine will be less than 1 % of annual precipitation.
Another aspect to consider is the effect that coal mining will have on the groundwater
encountered within the gold mining area, as it is expected that groundwater which originates
from the Karoo aquifer and is affected by coal mining could travel through preferential flow paths
from the Karoo aquifer downward to the gold mining which affects the Witwatersrand aquifer.
Deep level gold mining operations will result in a degree of contamination of groundwater within
the Witwatersrand aquifer. Most of the groundwater encountered within this aquifer during
operational gold mining operations will be pumped to surface (Winkelhaak Mine EMPR, 1988).
Groundwater draining into the mine workings will be of varying quality as it will travel for a
considerable distance within the subsurface to reach the gold mining activities. Sulphate-rich
minerals will give rise to a low pH (depending on oxygen and buffering capacity of the carbonate
minerals) while carbonate-rich minerals will give rise to a higher pH.
The following sequence of chemical reactions will occur during the operational phase:
• The water seeping into the gold mine will generally be of varying quality, except for
suspended solids present, which are expected to be at high levels due to the considerable
distance travelled. Most, if not all of the water will be pumped to surface during the
operational phase. Isolated areas of water may be present, however, and will drain to the
lowest point of the mine.
• Pyrite oxidation will commence in the mined void. The rate of oxidation will vary
considerably, depending on the rate of ingress of groundwater (the residence time), the
amount of oxygen available and the amount of sulphate-rich minerals that react with the
groundwater.
The following sequence of chemical reactions is most likely to occur post-closure of the gold
mine:
• Pyrite oxidation on the mined horizons will be extensive due to the slow flooding during the
post-closure phase.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 203
Chapter Nine: Conclusions and Recommendations 2010
• Poor quality groundwater will be present in the mine horizon when total flooding is
completed.
9.1.5 Current conceptual model for decant in Sasol coal mining operations under
operational (pumping) conditions
For groundwater affected by the coal mining activities and left to fill the mining voids to decant, it
will have to overcome more than four bars of pressure, which is unlikely. As the piezometric
level from the mine and the water level in the fractured zone combine, the excess water will
seep out as water that is unaffected by underground coal mining activities. The average
thickness of the overburden above the No. 4L coal seam of the Middelbult Block 8 area is
approximately 90 m (refer Karoo Supergroup, Figure 22) while the average thickness of the
overburden covering the Sasol Secunda mining complex is more than 100 m thick.
9.1.6 Conceptual model for decant in Harmony Evander gold mining operations post-
closure (no-pumping)
After the mining voids have flooded with water, the piezometric level of the artificial gold mine
aquifer will rise with the storage coefficient value of the artificial gold mine aquifer (not the
specific yield) as conditions in the artificial mine aquifer have changed from an unconfined
aquifer to a confined aquifer. The influx of water from the overlying aquifers into the mine aquifer
will decrease as the Witwatersrand piezometric level approaches the Karoo aquifers water level
in the gold mine shafts shown in Figure 110. The lowest point of the coal seam floor is at 1450
m.a.m.s.1 once the water level in the shafts reaches this elevation the Witwatersrand aquifer's
piezometric level and the Karoo aquifer's water level will become one. As the piezometric level
from the gold mine and the water level in the Karoo aquifer combine the excess water will seep
out as water that is unaffected by underground gold mining activities (Vermeulen & Dennis,
2009). For groundwater affected by the gold mining activities and left to fill the mining voids to
decant it will have to overcome the same four bar of pressure as the coal mine artificial aquifer,
which is unlikely.
9.1.7 Time for gold mine workings' voids to flood
If the current inflow rate for the total Harmony gold mine is taken as 3454790 m3 (Table 15) the
time for Harmony Evander gold mine workings to flood to the coal seam floor will be 406 years.
The piezometric level of the artificial gold mine aquifer will rise with the storage coefficient value
of the artificial gold mine aquifer (not the specific yield). The inflow of water from the overlying
aquifers into the mine aquifer will decrease as the piezometric level in the shafts rises,
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 204
Chapter Nine: Conclusions and Recommendations 2010
decreasing the gradient and thereby decreasing the inflow rate (Equation 18). The inflow of
water in the shafts will be a high in the beginning but will decrease over time, as shown in Figure
113 and Figure 114.
Therefore, the time for the water level to reach the coal seam is estimated at 406 years if current
pumping conditions are ceased at an inflow rate of 3 454 790 m3.
In general, a conservative approach was taken during this investigation. It is recommended that
the system be monitored and re-calibrated as further monitoring data are collected.
9.1.8 Limitations of this study
There were some limitations in this study and are listed below:
• Some of the old mining areas such as No. 1 shaft and No. 6 shaft are inaccessible as
they have been decommissioned making them unsafe to investigate for groundwater
interaction;
• Some of Middelbult Mining's area mostly to the south of the interaction area has already
been flooded with water making them inaccessible for investigation;
• Water level data from the decommissioned shafts are relatively old and is not constantly
measured, therefore water level data were heavily relied on from previous reports;
• The positions of exploration boreholes are difficult to determine as they are not marked
on surface;
• The fact that no one has marked were water-bearing structures are encountered or if
they were sufficiently plugged;
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 205
Chapter Nine: Conclusions and Recommendations 2010
9.2 Recommendations
The two mines have specific regulations in place to prevent groundwater interaction between the
mines. These regulations are listed below and in their respective Environmental Management
Programmes. It is of utmost importance that both mines comply with these regulations to
prevent any groundwater interaction between the two mines.
• The minimum distance that high extraction coal mining activities will take place from the
perimeter of the gold mine slimes dams is given as 300 m (Block 8 Expansion EMPR, 2003).
• No mining activities will take place directly underneath the gold mine tailings storage
facilities (Block 8 Expansion EMPR, 2003).
• Coal mining activities will keep a barrier pillar of 50 to 100 m from vertical gold mine shafts.
(Block 8 Expansion EMPR, 2003). By keeping a barrier pillar of 50 to 100 m from the shafts
there will be no influx of water from the coal mine into the gold mine shafts as the barrier will
act as an impermeable layer restricting groundwater flow.
• All gold mine exploration boreholes when intersected by the coal mine will be sealed off with
a non-reactive material to prevent the drainage of poor quality, low pH groundwater from the
coal mine to the gold mine (Block 8 Expansion EMPR, 2003).
• The gold mine shafts will be sealed using design and methods specified by a competent
professional Civil Engineer (Leslie Mine EMPR, 1994). This would significantly decrease the
ability of the shafts as possible recharge sources as well as preferential pathways linking the
two mines.
• The water levels in the decommissioned gold mine shafts should be measured bi-annually to
establish the exact final water level of the Witwatersrand aquifer. This data can then be used
in a numerical model to confirm the findings in this report.
• Influxes encountered during the intersection of water-bearing structures in both underground
mines can usually be plugged (grouted), or the influx water can be pumped away and dealt
with as part of the mine water balance.
• The rate of influx where faults, fractures and exploration boreholes or any other water-
bearing structure intersecting the gold mine workings should be measured to establish the
exact volumes of groundwater flowing in the gold mine through structures. This data can
then be used in a numerical model to confirm the findings in this report.
• The position of the exploration borehole intersecting the gold mine workings at No. 8 shaft
15 level should be located on surface. The borehole should then be sampled for hydro-
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 206
Chapter Nine: Conclusions and Recommendations 2010
chemical, stable environmental isotopes at surface and underground where it intersects the
gold mine workings. This should be done to determine the rate of flow from surface to the
gold mine workings. The borehole should be sealed if the coal mine intersects it.
• All drainage in the floor of the coal mine should be investigated and where possible sealed.
This will restrict vertical movement of water into the underlying formation and hence the
Witwatersrand aquifer.
• Design long-term water management schemes taking cognisance of neighbouring mining
activities.
• Design the mine layouts to retain as much of the mine water as possible in the underground
workings whilst mining.
• Flood mine workings (cease pumping) as soon as possible after closure to limit oxidation
reactions within both mines.
• Water balances should be done in a more dynamic manner, clearly identifying the source
water.
• The area around number No. 9 shaft should be monitored as interaction is occurring there.
• The areas around Leslie and Winkelhaak slimes dam should be monitored as possible
interaction from the gold mine slimes dam could affect the water quality in the Karoo aquifer.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 207
References 12010
References
Appelo, C.A.J., Postma, D. (2005). "Geochemistry, Groundwater and Pollution." (2nd edition).
A.A. Balkema Publishers, Rotterdam, Netherlands.
Asano, Y., Uchida, T., Ohte, N. (2002). "Residence times and flow paths of water in steep
unchannelled catchments." Tanakami, Japan. Journal of Hydrology. p 261, 173-192.
Bredenkamp, O.B., Van Der Westhuizen C., Wiegmans F.E., Kuhn C.M. (1986). "Groundwater
Supply Potential of Dolomite compartments west of Krugersdorp, South Africa." Technical
Report No. GH 3440, Directorate of Geohydrology, Department of Water Affairs.
Clark, I.O., Fritz, P. (1997). "Environmental isotopes in hydrogeology," Lewis Publishers, New
York, USA. p. 328-329.
Cook, P.G. (2003). "A guide to regional groundwater flow in fractured rock aquifers." CSIRO,
Land and Water, Glen Osmond, SA, Australia. p 1-10, 14-23.
Criss, R.E., Davisson, M.L. (1996). "Isotopic imaging of surface water/groundwater interactions,
Sacremento Valley." California. Journal of Hydrology, Vo1178. p. 205-222 .
Deer, W.A., Howie, R.A. Zussman, J. (1992). "Rock forming minerals, Single Chain Silicates."
Longman, Volume 2A London. p 696.
Drever, J.I. (1997). "The geochemistry of natural waters, Surface and Groundwater
Environments." Third edition, Prentice Hall, New Jersey.
Du Preez, M., Dennis, R., G.J. van Tonder. (2007). "Experimental Measurements of Specific
Storativity by the Determination of Rock Elastic Parameters." Institute for Groundwater
Studies, University of the Free State.
Eby, G.N. (2004). "Principles of Environmental Geochemistry." Thomson Brooks/Cole, Pacific
Grove, CA.
Falcon, R.M.S. (1986). A brief review of the origin, formation and distribution of coal in Southern
Africa. In: Mineral deposits of Southern Africa, I & II, Ahaeusser, C.R. and Maske, S. (eds.)
Geological Society of South Africa, Johannesburg. p. 1879-1899.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 208
References 12010
Freeze, R.A., Cherry, J.A. (1979). "Groundwater." Prentice-Hall, Inc. Englewood Cliffs, New
Jersey. p. 39 - 44, 47-49, 114-139.
Fritz, P., Fontes, J., C.H., Elsevier, A. (1980). "Handbook of Environmental Isotope
Geochemistry, Volume 1, The Terrestrial Environment." New York. USA. p. 21-29, 283-321,
473-495.
Gat, J.R. Carmi, I. (1970). "Evolution of the isotopic composition of atmospheric waters in the
Mediterranean Sea area." J. Geophys. Res 75:3039-48.
Gat, J.R. (1996). "Oxygen and hydrogen isotopes in the hydrologic cycle." Unpublished paper,
Department of Environmental Sciences and Energy Research, Weizmann Institute of
Science, 76100 Rehovot, Israel.
Grobbelaar, R. (2001). "Quantification and management of intermine flow in the western
Witbank Coalfield: Implication for mine water volumes and quality." Unpublished M.Sc
thesis, Institute for Groundwater Studies, University of the Free State.
[GCS] Groundwater Consulting Services. (2002). "Harmony gold mining company Evander
operations water quality status report." Project nr 2002-11-338.
[GCS] Groundwater Consulting Services. (1998). "Evander gold mines geohydrological
investigation final report, Johnstone, A.C. and Jansen H.C.
Harmony/Sasol Working Group (2005). Interaction between Gold and Coal mining. A Minism W
analysis. Ref no. ERE 8-27/2005.
Harmony Evander Gold Mining Servitudes (2008). Excel sheet from Harmony Evander's
Geological Department.
Harmony Gold Mining Company (1988). "Winkelhaak Mines Limited, Environmental
Management Programme Report."
Harmony Gold Mining Company (1994). "Leslie Gold Mines Limited, Environmental
Management Programme Report."
Harmony Gold Mining Company (1994). "Kinross Mines Limited, Environmental Management
Programme Report."
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 209
References 12010
Havenga, A. (2002). The long-term impact of intermine flow from Collieries in the Mpumalanga
Coalfields." Unpublished M.Sc thesis, Institute for Groundwater Studies, University of the
Free State.
Hem, l-l.D. (1985). "Study and interpretation of the Chemical Characteristics of Natural Waters."
United States Geological Survey - Supply Paper, no. 2254.
Hodgson F.o.I., Krantz R.M. (1995). "Investigation into the groundwater quality deterioration in
the Olifants River catchment above the Loskop dam with specialised investigation into the
Witbank dam sub-catchment." Institute for Groundwater Studies, University of the Free
State.
Hodgson, F.o.I., Grobbelaar, R. (1998). "Investigation into long-term water quantity and quality
trends at Ermelo Colliery." Confidential Report to Ingwe S.A.
Jasper Muller Associates cc. Van der Berg, J. (2002). "Compilation of geology and groundwater
inputs for the Middelbult block 8 EMPR - SASOL coal volume I and II - text and appendices."
Report prepared for Sasol Mining (Pty) Limited.
Johnson, M.R., Anahaeusser, C.R., Thomas, R.J. (2006). "The Geology of South Africa, Council
for Geoscience." Pretoria, RSA, p 155-208.
Kruseman, G.P., N.A. De Ridder (1994). "Analysis and Evaluation of Pumping Test Data." 2nd
Edition. Wageningen, Netherlands, ILRI Publication No. 47.
Lippmann, J. Stute, M. Torgesen, T. Moser, D.P. Hall, J. Lihung, L. Borcsik, M. Bellamy, R.E.S.
Onstott, T.C. (2003). "Dating ultra-deep mine waters with noble gases and 36CI,
Witwatersrand Basin, South Africa." Geochimica et Cosmochimica Acta. Vo1.67, Issue (23).
p.4597-4619.
McKnight Geotechnical Consulting, C. Oliver, D. De Bruyn, I. (2002). "Evander mine poplar gold
project, Mpumalanga, A review of the geological structural model and assessment of
possible groundwater ingress for mining of Kimberley Reef." Internal Report prepared for
Rison Groundwater Consulting.
Middelbult Mine: Block 8 Addendum Volume 1. (2003). "Environmental Management
Programme Report for the Middelbult Mine: Block 8 Expansion." Prepared for Sasol Mining
(Pty) Limited. Oryx Environmental. nr.OE63.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 210
References 12010
Midgley, D.C., Pitman W.v., Middleton B.J. (1994). "Surface water resources of South Africa"
1990. WRC Report No. 298/1/94. Water Research Commission, Pretoria.
Mulholland, G. (1982). "Poplar mine limited. The Potential ingress of water into the proposed
underground workings." Gencor Mines Geological Department. Bulletin No. 2462.
Onstott, T.C., Lin, Davidson, M, Mislowack, B. Boresik M. Hall, J. (2006). "The Origin and Age of
Biogeochemical Trends in Deep Fracture Water of the Witwatersrand Basin, South Africa."
Department of Geosciences, Princeton University, Princeton, New Jersey, USA.
Phillips, F.M. (2009). "Chlorine-36 dating of Old Groundwater." New Mexico Institute of Mining
and Technology. Socorro, NM, USA.
Piper, A. M. (1944). "A graphic procedure in the geo-chemical interpretation of water analyses."
Am. Geophys. Union Trans., 25. p. 914-923.
Pritchard-Davies, E.W.D. (1962). Bulletin No. 77. "Sub- Karoo water table." Addenda 1-7.
Rison Groundwater Consulting, M. (2002) "Harmony Gold Mining Limited - Evander operations:
Geohydrological assessment of the potential risk of groundwater inflow into the Rolspruit
r Project."
Rison Groundwater Consulting. (2003). "Harmony Gold Mining Limited - Evander operations:
Geohydrological assessment of the potential risk of groundwater inflow into the Poplar
Project."
Rison Groundwater Consulting, M. (2007). "Harmony Gold Mining Limited - Evander operations:
Geohydrological assessment of the groundwater table in Evander no. 6 shaft."
Saayman, l.C. Scott, D.F. Prinsloo, F.W. Moses, G. Weaver, J.M.C. Talma, S. (2003)
"Evaluation of the application of natural isotopes in the identification of the dominant stream
flow characheteristics".
Schwartz., F.W. Zhang, H. (2003). "Fundamentals of Groundwater." Ohio State University. p.
42-130.
Snyman, J.P. (1998) Coal. "The mineral resources of South Africa." Eds Wilson, M.G.C and
Anhaeusser, C.R. p. 136-161.
South African Weather Services (www.weathersa.co.za).
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 211
References 12010
Talma, A.S., Weaver, J.M.C. (2003). "Evaluation of groundwater flow patterns in fractured rock
aquifers using CFCs and Isotopes." Water Research Commission Report No. 1009/1/03,
South Africa.
Tweedie, E.B. (1986). "The Evander Goldfield". In: Mineral Deposits of Southem Africa 1.
Anhaeusser, C.R. & Maske, S. (eds). Geological Society of South Africa, Johannesburg,
South Africa. p. 705-730.
Unesco (1973). "Ground-water studies. An intemational guide for research and practice." In
Brown, R.H., Konoplyantsev, A.A, Ineson, J. and Kovalevsky, V.S. (eds). Nuclear techniques
in ground-water hydrology. Chapter 10.
USGS. (2004) "Arsenic, Nitrate, and Chloride in Groundwater." Oakland County, Michigan, USA.
Usher, B.H., Cruywagen, L-M., de Necker E., Hodgson, F.D.I. (2001). "On-site and Laboratory
Investigations of Spoil in Opencast Collieries and the Development of Acid Base Accounting
procedures." Final report to the Water Research Commission; Report no 2001.
Verhagen, B.Th., Geyh, M.A., Fróhlich, K. and Wirth, K. (1991). "Isotope hydrological methods
for the quantitative evaluation of groundwater resources in arid and semi-arid areas."
Research Report, Fed. Ministry for Econ. Coop. ofthe Fed, Republic of Germany.
Verhagen, B.Th. (1997). "Use of environmental isotopes in the characterisation of the Table
Mountain Sandstone Aquifer, South Africa." Progress Report to the International Atomic
Energy Agency, Research Contract 8983/RO/Regular Budget Fund.
Verhagen, B.Th. (2000). "Environmental isotope hydrology: Principles and application to the
geohydrology of the Karoo Basin." Karoo Aquifer Handbook, WRC, in preparation.
Verhagen, B.Th., Butler, M.J. (2004). "Isotope hydrology-An African Experience." Geoscience
Africa, Extended Abstracts, Vol. 2. Geological Society of South Africa, Johannesburg. p.
671-672.
Vermeulen, P.D., Usher, B.H. (2006). "Sulphate generation in South African Underground and
Opencast Colleries." Environmental Geology 49 (4): 552-569, February 2006.
Vermeulen, P.D., Usher, B.H. (2006.) "Recharge in South African underground colleries."
Journal of Institute of Mining and Metallurgy 106: 771-788, November 2006.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 212
References 12010
Vermeulen, P.D., Dennis I. (2009). "Numerical Decant Models for Sasol Mining Collieries,
Secunda." Institute for Groundwater Studies. University of the Free State.
Vermeulen, P.D. (2010). ''The assessment of mine rebound and decanting in deeper coal
mines." Proceedings - Mine Water and Innovative Thinking. International Mine Water
Association. Nova Scotia, Canada. p. 511-515.
Weaver, J.M.C., Talma A.S., Cavé L.C (1999). "Geochemistry and Isotopes for Resource
Evaluation in the Fractured Rock Aquifers of the Table Mountain Group. Report No.
481/1/99, Water Research Commission, Pretoria.
Yechieli, Y., Wood, W.W. (2002). "Hydrogeological processes in saline systems: playas
sabkhas and saline lakes." Earth Science Reviews, 58, pp. 343-365.
Groundwater Interaction between a deep coal mine and a deeper lying gold mine 213
Groundwater interaction between a deep coal mine and a deeper lying_gold mine
Abstract
After the closure of mines, it is expected that water in the mined-out areas will flow along
preferred pathways and accumulate in lower-lying areas (Grobbelaar, 2001). Over time, these
man-made voids will fill up with water and hydraulic gradients will be exerted on peripheral areas
within mines, resulting in groundwater flow between mines and possibly surface. This flow is
referred to as intermine flow (Grobbelaar, 2001). Due to the potential long-term impact of
intermine flow in terms of water quantities and qualities, the Department of Water Affairs and
Forestry regards intermine flow as one of the most important challenges in the mining industry.
Hydro-chemical and stable environmental isotope sampling were used to determine the
interaction between the Karoo aquifer and the Witwatersrand aquifer. Comparison between the
results showed that there is interaction between the Karoo and Witwatersrand aquifer, both
where coal mining is present and where coal mining is not present.
As the coal mine in future will cover most of the gold mining area except for the central part of
the study area, it is expected that groundwater affected by the coal mining in the Karoo aquifer 1
will move along naturally occurring structures from the base of the Karoo aquifer into the gold
mine workings if preferential pathways are encountered. This interaction can directly link to the
number of structures that are water-bearing in the study area and are covered by coal mining
activities. From the field work done in the underground gold mine, the number of structures that
are water-bearing are the exception rather than the norm. Therefore the potential for
groundwater interaction from future coal mining into the gold mine will be low.
Based on current inflow rates, the gold mine workings will take approximately 406 years to flood
and reach the bottom of the coal mine workings. It is highly unlikely that this flooding will decant
onto the surface through the coal mine, as the shallowest coal mine workings are at 39 m below
surface in the study area. In the unlikely event of decanting, polluted water originating from the
mining cavities will seep out as normal unpolluted springs at low points.
214
Appendices
Appendix A: Hydro-chemical Results
5iteName OateTimeMeas pH EC Ca Mg Na K PAlk MAlk F Cl N02(Nl Br N03(Nl
KlO 2010103/01 12:00 7.41 189 228.53 156.90 122.23 4.09 0 345 0.1881 834.0 -0.1 4.4398 0.0196
K50 2010103/01 12:00 7.21 383 263.74 230.81 184.02 4.99 0 229 0.2952 1060.0 -0.1 5.8736 0.0273
Wl0 2010103/01 12:00 7.78 89.1 27.51 16.63 137.90 4.18 0 297 0.3529 89.0 -0.01 0.4403 0.0676
WB30 2010103/01 12:00 7.26 500 402.26 326.19 139.74 38.26 0 325 0.2241 1190.0 -0.1 5.543 0.0541
WB50 2010103/01 12:00 8.16 172 46.67 42.48 247.54 6.46 0 253 0.2014 250.4 -0.1 1.288 0.0148
55 2010103/01 12:00 8.11 56.2 24.57 7.83 91.34 9.11 0 271 1.0095 24.8 0.0132 0.1096 0.0753
54 2010103/01 12:00 7.97 188 9.65 3.73 421.07 2.97 0 715 4.4182 190.3 0.287 0.5367 0.1473
MOBCA51NO 2010103/01 12:00 6.82 336 132.98 39.09 594.57 6.21 0 561 7.2704 217.6 -0.1 0.6127 0.039
51 2010103/01 12:00 7.6 529 334.08 166.44 729.04 9.78 0 258 1.2374 162.3 -0.1 0.2575 0.2489
53 2010103/01 12:00 7.85 265 148.40 84.07 265.35 6.80 0 331 0.2556 585.0 0.1085 2.574 0.0856
MHN47 2010103/01 12:00 7.55 156 55.29 26.06 257.61 4.20 0 539 1.1082 60.6 -0.1 0.4362 3.0489
52 2010103/01 12:00 8.54 272 2.73 1.82 690.65 2.66 1376 9.9881 141.1 -0.1 0.5596 0.0481
MHR45 2010103/01 12:00 6.33 449 246.91 68.87 739.77 11.13 0 266 1.1537 158.7 -0.1 0.3978 0.024
H6 2010103/01 12:00 8.68 196 8.06 5.29 438.08 2.29 757 8.9569 176.0 -0.01 1.1557 0.7427
H7 2010103/01 12:00 8.63 148 3.70 0.10 370.00 2.12 572 8.8198 168.2 -0.1 1.0097 0.1052
Hl0 2010103/01 12:00 8.04 198 34.59 3.17 382.07 2.53 0 209 5.4121 456.1 -0.1 1.5611 0.0249
Hl 2010103/01 12:00 6.75 488 135.11 0.91 816.13 3.76 0 25 2.837 1595.0 -0.1 7.5205 0.0457
H2 2010103/01 12:00 8.8 130 2.81 0.36 250.22 1.67 107 9.3553 323.9 -0.1 0.7323 0.0393
H3 2010103/01 12:00 8.26 177 7.99 3.49 347.17 1.80 0 244 8.494 358.1 -0.1 1.2966 0.1383
H4 2010103/01 12:00 6.7 472 211.09 3.11 717.23 5.18 0 8.85 2.3909 1612.0 -0.1 7.477 0.0533
H5 2010103/01 12:00 8.04 169 6.20 0.60 342.20 1.67 0 252 8.3895 362.4 -0.1 1.5202 0.0462
P04 504 AI Fe Mn CA Hard. Mg Hard. Tot.Hard. Ag AI As Ba Be Cd Co Cr Cu Fe
-1 58.818 0.028 0.039 1.039 571.3 643.3 1214.6 <0.010 0.010 <0.006 0.631 <0.001 <0.001 0.230 <0.006 0.016 0.027
-1 420.6812 0.022 0.019 0.004 659.3 946.3 1605.7 <0.010 0.005 <0.006 0.060 <0.001 <0.001 0.308 <0.006 0.006 0.010
-0.1 53.5297 0.018 0.020 0.066 68.8 68.2 137.0 <0.010 0.004 <0.006 0.101 <0.001 <0.001 0.014 <0.006 0.003 0.013
-1 921 0.040 0.030 0.017 1005.6 1337.4 2343.0 <0.010 0.008 <0.006 0.033 <0.001 <0.001 0.366 <0.006 0.013 0.014
-1 247.0905 0.021 0.016 0.005 116.7 174.2 290.9 <0.010 0.012 <0.006 0.071 <0.001 <0.001 0.008 <0.006 0.005 0.012
-0.1 5.317 0.016 0.036 0.006 61.4 32.1 93.5 <0.010 0.003 <0.006 0.425 <0.001 <0.001 <0.005 <0.006 0.002 0.030
-1 45.56 0.101 0.029 0.004 24.1 15.3 39.4 <0.010 0.013 <0.006 0.374 <0.001 <0.001 <0.005 <0.006 0.003 0.024
-1 1024 0.027 0.072 1.004 332.4 160.3 492.7 <0.010 0.010 <0.006 0.052 <0.001 <0.001 <0.005 <0.006 0.003 0.048
-1 2759 0.050 0.037 2.093 835.2 682.4 1517.6 <0.010 0.013 <0.006 0.015 <0.001 <0.001 0.024 <0.006 0.004 0.025
-1 213.3041 0.021 0.027 0.123 371.0 344.7 715.7 <0.010 0.004 <0.006 0.133 <0.001 <0.001 <0.005 <0.006 0.005 0.020
0.2542 200.6636 0.024 0.020 0.426 138.2 106.8 245.1 <0.010 0.013 <0.006 0.089 <0.001 <0.001 <0.005 <0.006 0.003 0.017
-1 4.3192 0.015 0.051 0.004 6.8 7.5 14.3 <0.010 0.005 <0.006 0.399 <0.001 <0.001 <0.005 <0.006 0.003 0.011
-1 2066 0.023 8.735 2.740 617.3 282.4 899.7 <0.010 0.007 <0.006 0.057 <0.001 <0.001 <0.005 <0.006 0.002 6.188
-0.1 44.4517 0.017 0.018 0.003 20.1 21.7 41.8 <0.010 0.010 <0.006 0.066 <0.001 <0.001 <0.005 <0.006 0.005 0.018
-1 54.1747 0.018 0.178 0.017 9.3 0.4 9.7 <0.010 0.001 <0.006 0.013 <0.001 <0.001 <0.005 <0.006 0.002 0.036
-1 121.7786 0.017 0.014 0.004 86.5 13.0 99.5 <0.010 0.005 0.018 0.036 <0.001 <0.001 <0.005 <0.006 0.003 0.010
-1 14.1217 0.023 0.032 0.009 337.8 3.7 341.5 <0.010 0.008 <0.006 1.182 <0.001 <0.001 <0.005 <0.006 0.003 0.010
-1 4.0017 0.017 0.016 0.004 7.0 1.5 8.5 <0.010 0.003 0.015 0.001 <0.001 <0.001 <0.005 <0.006 0.000 0.003
-1 67.456 0.700 0.024 0.045 20.0 14.3 34.3 <0.010 0.537 <0.006 0.020 <0.001 <0.001 0.101 <0.006 0.008 0.023
-1 14.9831 0.020 0.014 0.021 527.7 12.8 540.5 <0.010 0.003 <0.006 1.375 <0.001 <0.001 <0.005 <0.006 0.004 0.012
-1 12.968 0.022 0.012 0.004 15.5 2.5 18.0 <0.010 0.013 0.088 0.023 <0.001 0.003 <0.005 <0.006 0.004 0.015
Mn Mo Ni Pb Sb Se V Zn
0.937 <0.003 <0.010 <0.006 <0.006 0.020 <0.006 0.017
0.002 <0.003 <0.010 <0.006 <0.006 0.014 <0.006 0.015
0.074 <0.003 <0.010 <0.006 <0.006 0.016 <0.006 0.005
0.014 <0.003 <0.010 0.006 0.006 0.016 0.017 0.009
0.002 <0.003 <0.010 <0.006 <0.006 0.013 <0.006 0.007
0.003 <0.003 <0.010 <0.006 <0.006 0.013 <0.006 0.004
0.002 <0.003 0.019 <0.006 <0.006 0.021 <0.006 0.005
0.843 <0.003 <0.010 <0.006 <0.006 0.024 <0.006 0.007
1.507 <0.003 0.028 <0.006 <0.006 0.Q10 <0.006 0.069
0.135 0.017 <0.010 <0.006 <0.006 0.Q10 <0.006 0.017
0.422 0.004 <0.010 <0.006 <0.006 0.023 <0.006 0.008
0.002 <0.003 <0.010 <0.006 <0.006 0.028 <0.006 0.004 i
1.707 <0.003 <0.010 <0.006 <0.006 0.012 <0.006 0.011
0.001 0.005 <0.010 <0.006 <0.006 0.034 <0.006 0.005
0.003 0.003 <0.010 <0.006 <0.006 0.025 <0.006 0.002
0.001 0.018 <0.010 <0.006 <0.006 0.012 <0.006 0.005 I
0.008 <0.003 <0.010 0.006 0.006 <0.006 <0.006 0.012
I
0.000 <0.003 <0.010 <0.006 <0.006 <0.006 <0.006 0.002
0.054 0.003 0.409 <0.006 <0.006 0.010 <0.006 0.026
I
0.024 <0.003 <0.010 <0.006 <0.006 <0.006 <0.006 0.011 I
0.001 0.003 0.010 <0.006 <0.006 0.013 <0.006 0.007
Appendix B: Isotope Results
,-~
6"0
Sa'T1pleNo. Sa'T1pleNa'T1e .s'H Reportable Value (pennil) 6'H Stand ...d Deviation (permil) 6"0 Reportable Value (pe,,"il) Standard Delliation (permil) Analysis Date
1 H1 -25.86 0.63 -5.97 0.20 0S.0l-2010
Cl HCl -Cl6.82 1.01 -5.JO 0.Cl1 u!!-u~2010
3 H3 -26.72 1.23 -5.30 0.21 06-03-2010
4 H4 -28.26 0.99 ~.37 0.26 0S.0l-2010
5 H5 -26.08 1.30 -4.99 0.19 0S.0l-2010
6 H6 -26.03 0.92 -4.49 0.08 0S.0l-2010
7 H7 -9.67 1.39 -3.29 0.09 0S.0l-2010
8 H10 -15.82 1.18 -3.35 0.26 0S.0l-2010
9 S1 -10.43 1.38 -2.35 0.27 0S.0l-2010
10 S2 -28.31 0.14 -4.85 0.07 0S.0l-2010
11 S3 -1J.7M u.96 -2.96 0.09 0S.0l-2010
12 S4 -21.11 o.n -4.41 0.18 0!!-0l-2010
13 S5 -19.85 U.76 -4.46 0.17 0S.0l-2010
14 KB1D -24.78 1.33 -5.12 0.26 oe-os-zoio
15 KB5D -22.81 1.11 -4.70 0.18 uS.ul-2010
16 WB1D -21.73 1.D -4.27 0.22 06-0l-2010
17 WB3D -13.57 1.04 -3.J6 0.11 u!!-ul-2010
18 WB5D -25.52 1.69 -5.11 0.04 0S.0l-2010
19 MHN47 -14.36 1.50 -3.35 0.17 0S.0l-2010
20 MHR45 -9.72 1.10 -2.52 0.14 0S.0l-2010
21 MOBCasino -24.68 0.69 -4.57 0.28 0S.0l-2010
22 Rainwater 3.53 1.51 -1.95 0.25 06-0l-2010
LGR DT-1oo Accuracy of standards preparation
6"0 (permII)
6"0('f~)
-20 -15 -10 -5 5 -16.20 -16.00 -15.80 -15.60 -1MO -1520 -15.00
-115.0
a 10 -20 ,I 0 -115.5o IOISW 1o:0LLGG" sw RR2A2 _e -116.0• LGR2 spec
-GMWL -40 -116.5
o LGR2 . • 0 g-117.0 Co VSMOW2 0
• IA-R053 .- -60 Ë -117.5 ~
i -118.0
:I:
-00 'lo 0 -118.5
-119.0
0
-100 -119.5
. -120.0
-120
&"0(0/00}
6"0('f~)
-0.40 -0.:1) -0.20 -0.10 0.00 0.10 020 0.30
-10.6 -104 -10.2 -10.0 -9.8 -9.6 -9.4 -9.2 -9.0
1.0
-55.0
0 -56.0 0.5
-57.0 0.0
+ -58.0 -0.5g 0-.!!-59.0 C o"--1.0 f:I:0 ~.O 'lo -1.51:0 -61.0IA-R053 (lAD) .1
• o IA-R053(IAD)A_e • -2.0• IA-R053 (lAD) spec -62.0 1:0VSMOW2 DATA el -2.5
-63.0 OVSMOW2 DATA A_e
• VS MOW (IAEA) spec
Calibrated Current run Std Dev Current run Calibrated Current run Std Dev Current run
I
SBEEH ~2H (loo) ~2H (0/00) ~2H (0/00) 180(/00) ~'80 (loo) ~'80 (loo)
STANDARD
10 6.33 6.30 0.35 0.68 0_68 0.14
10/SW -32.21 -32.28 0.40 -4.88 -4.61 0.11
SW -71.56 -71.35 0.11 -10_02 -10.01 0.19
- ~ '---~-- ---- - - ~ --- - -- ---- __L_ ____
Appendix C: Harmony Gold mining data
Harmony Gold Mining Company
EVANDER GOLD MINE: WATER BALANCE DATA
Rainfall Average (Weather
Bureau) 134.0 94.0 78.0 46.0 19.0 7.0 8.0 10.0 25.0 78.0 128.0 120.0 747.0 62.3
Evaporation Average 1701. 141.
(Weather Bureau) 179.8 151.1 147.8 111.1 94.8 79.2 89.0 132.0 167.0 186.6 167.6 195.9 9 8
Mo Jan- Feb- Mar- Apr- May- Jun- Jul- Aug- Sep- Oct- Nov- Dec- Total Avel
nth 09 09 09 09 09 09 09 09 09 09 09 09 year mth
Winkelhaak Shafts
Comolex
1167. 154. m
Rainfall 364.1 157.7 91.8 4 33.6 28.5 0.3 20.8 12.5 117.7 195.2 141.7 9 4 m
163.8 70.96 12.82 52.96 63.76
Total Shaft areas (45 ha) 45 5 41.31 1.8 15.12 5 0.135 9.36 5.625 5 87.84 5 525.6 69.5 MI
32,06 29,00 23,00 21,41 30,35 28,31 24,58 15,58 25,73 28,48 31,20 18,41 308,1 26,3 To
Tons treated 1 3 0 9 9 7 1 1 5 2 6 3 57 71 ns
Vent Shaft Water
20.72 18.10 17.36 18.64 19.25
Down cast water in air 18.8 19.5 17.9 7 17.5 2 17.89 2 17.98 8 4 20.54 224.2 19.2 MI
41.57 37.15 35.28 34.91 36.52 36.67
Up cast water water in air 6 42.7 41.85 41.5 41.5 8 36.3 3 35.35 6 3 3 461.3 41.9 MI
22.77 20.77 19.05 17.92 16.26 17.26 16.13
Excess water in air 6 23.2 23.95 3 24 6 18.41 1 17.37 8 9 3 237.1 22.7 MI
IN
95.00 98.00 89.00 106.0 75.00 96.00 108.0 73.00 113.0 68.00 105.0 105.0 1131.
No 2 Shaft Fissure water 0 0 0 00 0 0 00 0 00 0 00 00 000 97.0 MI
39.50
No 2 Shaft Sludqe 3.600 3.900 3.300 2.900 4.100 3.700 3.400 3.100 3.800 2.900 2.400 2.400 0 3.4 MI
899.8 410.0 238.6 10.40 87.36 74.10 54.08 32.50 306.0 507.5 368.4 2989. 389.
Rainfall Tailinqs Dam 60 20 80 0 0 0 0.780 0 0 20 20 20 740 7 MI
UEmergency 30.45 18.29 10.35 17.17 12.47 105.6
dam 7 3 8.078 0.352 2.957 2.508 0.026 1.830 1.100 8 8 0 07 14.3 MI
21.80 12.29 72.44
Plant area 4 9.935 5.783 0.252 2.117 1.796 0.019 1.310 0.788 7.415 8 8.927 4 9.4 MI
Rainfall runoff Emergency 25.33 14.28 10.37 76.72
catchment 5 5.772 6.720 0.293 2.460 2.086 0.022 0.761 0.000 8.616 9 2 6 9.5 MI
Water from Kinross Met. 74.27 75.91 78.76 88.28 97.21 92.75 91.32 97.49 58.66 61.42 59.40 875.5
Plant to TFC 7 8 9 7 9 0 4 7 0.000 8 8 9 46 79.3 MI
Water from Kinross Kariba
dam 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.0 MI
18.49
Moister from rock hoisted 1.924 1.740 1.380 1.285 1.822 1.699 1.475 0.935 1.544 1.709 1.872 1.105 0 1.6 MI
Tot 1152. 623.5 431.7 209.7 273.0 274.6 205.0 232.5 152.7 463.6 721.9 568.1 5309. 604.
al 257 78 10 69 35 39 46 13 32 86 85 03 053 3 MI
OUT
80.51 88.83 64.32 83.08 64.39 36.77 60.23 80.45 86.92 79.35 888.5
Water to Kinross Met plant 90.22 9 8 2 7 73.4 8 8 6 6 9 7 40 81.0 MI
547.2 322.7 159.2 66.59 124.1 109.6 63.99 104.8 20.91 244.1 368.9 292.6 2425. 273.
Evaporation Tailings Dam 04 52 43 9 78 89 6 65 4 95 59 89 283 9 MI
UEmergency 14.21 12.92 10.79 12.53 11.99 13.20 11.54 119.6
dam 6 7.43 4 3 7.629 5.914 3.464 7.986 1 6 7 8 38 11.3 MI
UEmergency Dam to 167.8 63.59 109.0 18.44 20.96 15.15 66.31 41.03 84.00 668.4 113.
Leeuwpan 76 6 71 9 6 27.58 7 2 0 1 5 54.36 03 5 MI
Water Spillage into
W/haak spruit 5.5 3.5 0 0 0 0 0 0 0 0 0 0 9.000 2.3 MI
Change in Emergency 38.60
Dam level 35 0 -33.6 20 -18.2 8 30.4 -28.9 49.3 -23.4 -1.8 1.8 0 5.4 MI
Slime dam intersitual 292.2 145.7 95.23 29.60 55.37 50.05 27.63 45.47 109.4 170.6 128.3 1159. 140.
water 41 81 5 6 4 5 1 3 9.75 06 84 49 585 7 MI
Tot 1152. 623.5 431.7 209.7 273.0 274.6 205.0 232.5 152.7 463.6 721.9 568.1 5309. 604.
al 257 78 11 69 34 38 46 14 31 84 84 03 049 3 MI
Kinross Mine Complex
1098. 148. m
Rainfall 303.6 188.9 98.8 3 25 27 0.3 22.7 9.5 100.4 163.9 155.3 4 6 m
48.57 30.22 15.80 16.06 26.22 24.84
Total Shaft areas (16 ha) 6 4 8 0.48 4 4.32 0.048 3.632 1.5 4 4 8 175.7 23.8 MI
56,93 62,99 67,00 72,08 77,14 78,68 81,91 88,91 82,26 77,51 80,38 825,9 64,7 To
Tons Treated 9 7 0 1 1 3 9 1 5 8 65 7 06 54 ns
I
Vent Shaft Water
22.47 22.47 20.05 17.38 15.24 24.49 26.67 27.78
Down cast water in air 5 23.4 22.2 21.9 22.89 3 6 6 1 8 8 7 267.0 22.5 MI
41.97 37.88 37.99 39.11 90.91 92.71
Up cast water water in air 6 40.6 41.54 41.8 41.7 45.29 3 1 4 5 6 94.41 645.9 41.5 MI
19.50 22.81 17.82 20.60 23.87 66.41 66.03 66.62
1 17.2 19.34 19.9 18.81 7 7 5 3 7 8 3 379.0 19.0 MI
IN
123.7 84.96 105.8 124.3 120.2 105.9 119.4 109.6 108.8 106.5 47.22 62.00 1218. 104.
No 7 Shaft Fissure water 1 2 75 33 09 29 64 13 89 18 1 1 724 8 MI
991.0
No 8 Shaft Fissure water 84 84 87 80 84 84 87 85 85 68 72.21 90.87 80 83.8 MI
No 7 Shaft Sludge from 11.40 75.22
U/g 7.21 5.117 6.103 5.916 5.943 6 3.818 4.275 7.114 8.046 4.948 5.333 9 6.1 MI
__ ~_------c:_ __
73.13
Rand Water 7.7 5.258 ·6.899 6.202 5.938 5.316 5.771 4.849 5.156 7.669 8.035 4.341 4 6.5 MI
569.8 354.5 185.4 46.92 50.67 42.60 17.83 188.4 307.6 291.4 2061. 278.
Rainfall Tailings Dam 57 65 45 5.631 5 9 0.563 8 2 51 4 98 694 9 MI
18.82 11.71 10.16 68.10
Plant area 3 2 6.126 0.186 1.55 1.674 0.019 1.407 0.589 6.225 2 9.629 2 9.2 MI
32.18 20.02 10.47 10.64 17.37 16.46 116.4
Kariba Dam 2 3 3 0.318 2.65 2.862 0.032 2.406 1.007 2 3 2 30 15.7 MI
12.14 43.62
Toe dam 4 7.556 3.952 0.12 1 1.08 0.012 0.908 0.067 4.016 6.556 6.212 3 5.9 MI
Kariba Dam
Rainfall runoff catchment 2.131 1.326 0.694 0.021 0.176 0.19 0.002 0.08 0.067 0.705 1.151 1.09 7.633 1.0 MI
Toe Dam 100.5 13.60 146.9
catchment 22 1 7.114 0.216 1.8 1.944 0.022 4.903 0.38 7.229 4.72 4.473 24 30.4 MI
Sludgelwater from 80.51 88.83 64.32 83.08 64.39 36.77 60.23 80.45 86.92 79.35 888.5
Winkelhaak Met plant 90.22 9 8 2 7 73.4 8 8 6 6 9 7 40 81.0 MI
Water return from 469.0
Grootspruit (KM 4) 10 20 12 40 35 48 70 60 40 45 44 45 00 20.5 MI
53.47
Moister from rock hoisted 3.416 3.78 4.02 4.325 4.628 4.721 4.915 5.335 4.936 4.651 3.92 4.823 0 3.9 MI
Tot 1061. 692.4 524.5 331.5 392.9 391.2 356.0 358.1 331.2 537.6 614.8 621.0 6213. 652.
al 915 19 39 90 06 01 16 62 73 08 65 89 583 6 MI
OUT
To Winkelhaak Tailings 74.27 75.91 78.76 88.28 97.21 92.75 91.32 97.49 58.66 61.42 59.40 875.5
dam 7 8 9 7 9 0 4 7 0.000 8 8 9 46 79.3 MI
48.00
To Crushers 4.000 4.000 4.000 4.000 4.000 4.000 4.000 4.000 4.000 4.000 4.000 4.000 0 4.0 MI
404.4 210.5 57.18 48.94 61.03 70.38 10.22 13.78 131.1 132.2 79.90 175.7 1395. 180.
Evaporation Tailings Dam 37 97 7 0 8 1 6 9 10 02 5 32 544 3 MI
16.91 11.07 13.64 11.08 10.90 14.81 17.88 131.0
Kariba Dam 4 6 1 5 9.189 7.276 4.157 5.526 8.544 6 1 5 10 13.2 MI
55.45
Toe Dam 6.462 4.180 4.905 4.183 3.468 2.317 2.771 3.684 5.696 5.453 5.589 6.749 7 4.9 MI
Kariba pumped to 273.9 234.1 306.5 194.9 210.2 177.1 241.0 226.5 141.7 254.4 315.7 279.3 2855. 252.
Leeuwpan 51 12 86 58 44 94 48 45 37 55 16 11 857 4 MI
- - -
32.76 14.60 14.78 - 14.40 14.50 53.46 17.52
Change in Dam levels 0.000 0.000 0.660 0 0 0 9.800 0 3.800 0 0 1.880 0 -8.0 MI
Water spillage into 158.0 71.50 11.20 240.7
Grootspruit 00 0 0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 00 60.2 MI
Water to Winkelhaak
Emergency dam. 0 0 0 0 0 0 0 0 0 0 0 0 0.000 0.0 MI
Slime dam intersitual 123.8 81.03 47.59 12.89 22.34 22.50 12.28 21.52 37.00 57.42 79.95 76.12 594.5
water 75 5 2 8 8 2 9 1 2 4 6 2 64 66.4 MI
1061. 692.4 524.5 331.5 392.9 391.2 356.0 358.1 331.8 537.6 614.8 621.0 6214. 652.
916 18 40 91 06 00 15 62 89 08 65 88 198 6 MI
Leslie Mine Comolex
910.2 116. m I
Rainfall 237.7 127 99.8 1.8 26.3 50 3.2 19.3 9.8 87.3 117.5 130.5 00 6 m
27.30
Total Shaft areas (3 ha) 7.131 3.81 2.994 0.054 0.789 1.5 0.096 0.579 0.294 2.619 3.525 3.915 6 3.5 MI
IN
Water returned from 14.76 14.39 78.68
Grootspruit (BM 13) 9.000 5.300 2.100 5.400 3.000 6.500 4.350 3.402 2.179 0 5 8.300 6 5.5 MI
701.2 374.6 294.4 77.58 147.5 56.93 28.91 257.5 346.6 384.9 2685. 343.
Rainfall Tailings Dam 15 50 10 5.310 5 00 9.440 5 0 35 25 75 090 9 MI
30.42 16.25 12.77 11.17 15.04 16.70 116.5
Kariba Dam 6 6 4 0.230 3.366 6.400 0.410 2.470 1.254 4 0 4 04 14.9 MI
Rainfall runoff Kariba 15.79
Catchment area 4.176 2.231 1.753 0.032 0.462 0.879 0.056 0.145 0.172 1.534 2.064 2.293 7 2.0 MI
Irrigation water (Rand
water) 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.0 MI
744.8 398.4 311.0 10.97 84.41 161.2 14.25 62.95 32.51 285.0 378.1 412.2 2896. 366.
17 37 37 2 3 79 6 2 5 03 24 72 077 3 MI
OUT
490.7 262.2 205.3 54.00 103.0 39.63 20.11 179.4 242.6 269.3 1876. 240.
Evalloration Tailinqs Dam 08 18 29 3.696 3 93 6.593 4 7 73 36 06 806 5 MI I
10.98 11.47 95.23
Kariba Dam 5 7.895 9.986 8.34 5.895 4.376 4.711 5.799 7.006 9.27 9.501 3 7 9.3 MI I
Kariba pumped to
Leeuwpan 0 0 0 0 0 0 0 0 0 0 0 0 0.000 0.0 MI
Water Spillage from
Kariba into Grootspruit 0 0 0 0 0 0 0 0 0 0 0 0 0.000 0.0 MI
Slime dam intersitual 210.3 112.3 88.32 23.27 17.08 77.26 103.9 115.4 805.5 103.
water 65 95 3 1.593 6 44.25 2.832 1 8.673 1 88 93 30 2 MI
Change in Kariba Dam - 118.3
level 32.76 15.93 7.4 2.657 1.24 9.56 0.12 0.24 -3.28 19 22 16 13 13.4 MI
744.8 398.4 311.0 10.97 84.41 161.2 14.25 62.75 32.51 285.0 378.1 412.2 2895. 366.
18 38 38 2 4 79 6 4 6 04 25 72 886 3 MI
Leeuwpan
7696. 645.
Dam Area 647 645 645 643 642 643 641 639 635.5 639 639 638 500 0 ha
902.8 112. m
Rainfall (measured) 301.7 35.7 110 4.2 24.6 28.6 0.5 22 11.3 115.9 112.9 135.4 00 9 m
1690. 151. m
Evaporation (measured) 197 129 151.4 129.1 90.3 71.5 81.5 113.7 178 168.3 172.5 208.3 600 6 m
10.16
Evaporation Factor 0.82 0.96 0.98 0.7 0.93 0.74 0.87 0.98 0.65 0.82 0.87 0.84 0 0.9
IN
441.8 297.7 415.6 213.4 231.2 204.7 256.2 292.8 141.7 295.4 399.7 333.6 3524. 342.
Mine Water 27 08 57 07 10 74 05 57 37 86 21 71 260 1 MI
1951. 230.2 27.00 159.8 183.8 140.5 71.81 740.6 721.4 866.5 5806. 729.
Rainfall 999 65 709.5 6 58 98 3.205 8 2 01 31 6 715_ _7 MI
203.8 17.41 15.98 18.36 100.9 93.93 110.7 631.7
Run-off 05 5 70.62 0 5 1 0 0 0 6 3 2 99 73.0 MI
2597. 545.3 1195. 240.4 407.0 407.0 259.4 433.4 213.5 1137. 1215. 1310. 9962. 1144
631 88 777 13 53 33 10 37 49 047 085 951 774 .8 MI
Runoff from 1070 ha 50.00
catchment (%) 6.0 4.0 6.0 0.0 6.0 6 0 0 0 8 7 7 0 4.0 %
OUT
1045. 798.7 959.9 581.0 536.1 340.2 454.5 712.0 678.7 881.8 958.9 1119. 9067. 846.
Evaporation 164 68 99 79 45 11 01 12 14 58 79 821 251 3 MI
Spill + Release 0 0 0 0 0 0 0 0 0 0 0 0 0.000 0.0 MI
1045. 798.7 959.9 581.0 536.1 340.2 454.5 712.0 678.7 881.8 958.9 1119. 9067. 846.
164 68 99 79 45 11 01 12 14 58 79 821 251 3 MI
Total balance
5556. 2259. 2463. 792.7 1157. 1234. 834.7 1087. 730.0 2423. 2930. 2912. 24381 2768
Total in 620 822 063 44 407 152 28 064 69 344 059 415 .487 .1 MI
4004. 2513. 2227. 1133. 1286. 1167. 1029. 1365. 1195. 2168. 2673. 2721. 23486 2469
Total out 155 202 288 411 499 328 818 442 850 154 953 284 .384 .5 MI
- - - - - -
1552. 253.3 235.7 340.6 129.0 66.82 195.0 278.3 465.7 255.1 256.1 191.1 895.1 298.
Imbalance 465 80 75 67 92 4 90 78 81 90 06 31 03 5 MI
Imbalance equals on 137.2 m
Leeuwpan 239.9 -39.3 36.6 -53.0 -20.1 10.4 -30 -43.6 -73 40 40 30 02 11.4 m
Actual change in water 110.0 m
level (mm) 240 -40 10 -50 -20 10 -30 -50 -70 40 40 30 00 40.0 m
-
Actual change in water 1552. - - - - 444.8 719.6 259.
level (MI) 8 -258 64.5 321.5 128.4 64.3 192.3 319.5 5 255.6 255.6 191.4 50 5 MI
% imbalance 100 98 366 106 100.5 103.9 101.5 87.1 104.7 99.8 100.2 99.9
EVANDER GOLD MINES PRODUCTION STATS (up till30 th April 2010 )
Kinross Bracken
Winkelhaak Mine Mine Leslie Mine Mine
Tons Tons
Mined Tons milled Mined milled Mined Tons milled Mined milled
Fin Fin
Year m2 000 Year m2 000 m2 000 Fin Year m2 000
1958/ 1211287 1968/8 747180 823415
86 8 49442 5 7 28500 0 28200 1963/84 6489338 20912
1987 548120 2204 1986 541761 2155 290285 1382 1985 267115 1003
1988 584211 2162 1987 453868 1920 310764 1415 1986 264867 954
1989 524672 2015 1988 507749 2057 307874 1409 1987 227664 915
1990 526558 2106 1989 475436 2106 298048 1370 1988 224920 842
1991 436389 1925 1990 444989 2043 298427 1383 1989 184859 687
1992 448167 1898 1991 431836 1918 212014 1039 1990 143458 530
1993 426818 1756 1992 412882 1899 149616 578 1991 108372 388
*1994 267219 1161 1993 423980 1892 122014 418 1992 49223 185
1995 345230 1442 *1994 315277 1382 114203 404 61993 5245 31
1996 351002 1479 1995 381467 1672 119515 397 *1994 - -
1997 250275 1315 1996 370640 1598 99198 306 Total 7965061 26447
1998 92400 688 1997 355179 1648 117438 403
6 Underground
productionceased in
March 1993 .
1999 43992 221 1998 267517 1334 117565 424 Because
oflackofdetailone
can assume 600 -
700 Kg came from
2000 86581 438 1999 258015 1261 144777 474 plant
2001 92676 606 2000 271754 1221 91054 388 clean up .
2002 106616 692 2001 280795 1313 86131 469
2003 113548 669 2002 261851 1267 72099 293
2004 75374 644 2003 201462 1109 41651 196
2005 98230 493 2004 208358 1145 35076 138
2006 60476 338 2005 81487 1042 31463 139
2007 57287 323 2006 201102 939 36063 183
2008 58289 324 2007 195384 1111 4413 26
113338
2009 48811 306 2008 170871 956 38 41434
20*10 23795 146 2009 130491 776
1777961
Total 4 74793 I'!. 2010 89652 540
Total 152056 64804 --
0
0 C
9 ..O..'ot
IJ)
e ....
0 0N
I-
.... ..,
'otN
.0... N coEN
N
It)
C
(II
.>t:.
U
O...l
al
u
-ec~
z.2
::JU
011:':";
(!)-
II::Q.
W
C
Z
::J
...J
<
b
I-