8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
1/56
REVIEW RECORD SHEET
Project Short Title: ....................................................................................................................................................
Project No: ...................................................................... Document Ref. No: ......................................................
Document Title:...........................................................................................................................................................
Document Type: Report Draft Report
Letter Report Technical Memo
Other (specify) ...............................................................................................................
Project Manager: ............................................................ Project Director:...........................................................
CHECKLIST
YES NO* N/A
1. Have you read the brief/scope/TOR ? _______ ______ _______
2. Does the document meet the promised scope of works in the brief ? _______ ______ _______
3. Is the purpose and scope adequately stated? _______ ______ _______
4. Have the following been checked to the extent appropriate for this
project:
data analyses
calculations
spreadsheets
field data
drawings, figures, sketches
tables
data transfers
_______
_______
_______
_______
_______
_______
_______
______
______
______
______
______
______
______
_______
_______
_______
_______
_______
_______
_______
5. Has analytical uncertainty been adequately addressed?
6. Are conclusions and recommendations logical and defensible?
7. Are all necessary references identified?
8. Are limitations or conditions in the use of the data, recommendations, orconclusions discussed where necessary, and have appropriate disclaimers beenincluded?
9. Has the Project Director been consulted about this report
COMMENTS
ll checklist items have been satisfactorily addressed
Any No answers must be explained
eviewed by: ________________________________ ________________________________________
Print Name Signature
ate: ____________________________________________
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
2/56
Golde r Assoc iates Africa (Pty) LtdReg . No. 2002/007104/07
JOHANNESBURGPO Box 6001 Halfwa y House 1685Sou th AfricaTha nd a na ni Pa rk, Ma tuka C loseHalfway Ga rd ens, MidrandTel + (27) (0)11 254-4800Fa x + (27) (0)11 315-0317http://www.golder.com
Direc tors : P Onley (Au stralia), FR Suthe rlan d, AM va n Niekerk, JA Wate s
OFFICESIN JOHA NNESBURG, DURBAN, A USTRALIA, INDONESIA, NEW ZEALAND , PEOPLESREPUBLIC OF C HINA, PHILIPPINES, SINGAPORE
OFFICESAC ROSSAFRICA, ASIA, AUSTRALASIA, EUROPE, NO RTH AM ERICA , SOUTH AM ERICA
REPORT ON
HYDROGEOLOGICAL INVESTIGATION FOR
THE PROPOSED HEIDELBERG OPENCAST
MINE
NUMERIC MODELLING OF PIT INFLOWSAND GROUNDWATER IMPACTS
Report No : 7475/8297/7/G
Submitted to:
Anglo Coal Project Services
Private Bag X9
Leraatsfontein
DISTRIBUTION:
2 Copies - Anglo Coal Project Services1 Copy - Golder Associates Africa (Pty) Ltd Library
April 2006 7475
http://www.golder.com/http://www.golder.com/http://www.golder.com/8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
3/56
April 2006 i 7475/8297/7/G
GOLDER ASSOCIATES
EXECUTIVE SUMMARY
Anglo Coal Projects Services (ACPS) initiated the feasibility of developing the Heidelberg opencast
mine with an operational life of approximately twelve years for long-term coal supply to Eskoms
Grootvlei power station, which is to be re-commissioned.
Golder Associates Africa (GAA) was appointed to determine the possible impacts that the proposed
opencast mine will have on the groundwater regime in the area. A key aspect of the Heidelberg in-pit
water balance is the determination of groundwater inflows into the opencast workings during
operation and post mine closure.
Golder conducted a groundwater study to develop an understanding of the existing groundwater
environment and to evaluate the changes to this environment, as a result of pit impacts.
The main objectives of the groundwater modelling study are to:
Construct and calibrate a numerical groundwater flow model, to accurately simulate groundwater
flow through the aquifers
Simulate inflow rates into the mining sections using the preliminary mine plans in conjunction
with the calibrated groundwater flow model
Determine the necessity of dedicated dewatering well fields around mining sections
Simulate the potential influence of the Suikerbosrand River on mining
Simulate the extent of the influence of mining operations on the hydrogeological regime in the
study area.
Determine the influence of mining on community boreholes.
Simulate the effect of mine closure on groundwater levels and recommend an operational pit
groundwater/surface water level elevation to prevent decant of groundwater / surface water intothe Suikerbosrand River.
Identify the need for any additional hydrogeological field investigations in order to reduce
uncertainties identified from the hydrogeological flow modelling investigation conducted.
The results of the investigation are presented in this report and incorporate results of the
hydrogeological and geochemical study into the groundwater model.
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
4/56
April 2006 ii 7475/8297/7/G
GOLDER ASSOCIATES
A numerical groundwater model using the FeFlow code was constructed for the proposed Heidelberg
opencast pit. The conceptual model, model assumptions, construction, and results are presented in
Sections 12 - 14 of the report.
The key findings of the study can be summarised under:
Alternative A: Worst Case Scenario
The average influx of groundwater during operation will approximate 1 820 m3 /d, reduced to
1 160 m3/d (36 % reduction on average) with active dewatering of nine (9) dewatering wells and
825 m3/d (55 % reduction on average) with active dewatering of seventeen (17) dewatering wells
(see Figure S1 below).
The natural filling of the pit, to an elevation of 1476 mamsl (1 m below the base of the
Suikerbosrand River) will take approximately 1 980 days or more than 5 years after closure. The
average flow rate will be at 841 m3/d, reducing over time as driven by the rising head conditions
within the pit, to average flows of 452 m3 /d. If the opencast pit is allowed to fill above the
recommended waterlevel of 1476 mamsl, outflow of contaminated water to the Suikerbosrand
River could approximate > 280 m3/d.
There will be an increase in flows through the alluvium to the pit, as the pit development
approaches the Suikerbosrand River. Average flows will approximate 171 m3 /d without
dewatering and roughly contribute 9 % of groundwater inflow to the pit. Active dewatering will
increase the flow through the alluvium to an average of 214 m3/d and 265 m
3/d, for the 9 and 17
dewatering well scenarios respectively (Scenario 2).
Operation of the opencast pit will lead to the development of a dewatered cone of depression
extending to the Suikerbosrand River in the East, 1 400 m North, 615 m to the South and 410 m to
the West, with no active dewatering (Scenario 1). The footprint area will be influenced by the
nine (9) dewatering well field, operating six (6) months prior to pit development for Scenario 2.
This will result in a more extensive area of influence that could approximate 1 470m to the North,
720 m to the South and 570 m to the West.
The extent of dewatering by increasing the number of dewatering wells to seventeen (17), indicate
a cone of depression developing to 1 500 m North, 800 m South, 700 m West and to theSuikerbosrand River on the East. Operation of a 17 dewatering well field will ensure a more
effective barrier to groundwater inflow into the opencast pit.
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
5/56
April 2006 iii 7475/8297/7/G
GOLDER ASSOCIATES
Figure S1: Alternative A Pit Inflows
Alternative B: Best Case Scenario
The average influx of groundwater during operation of the Heidelberg opencast pit will
approximate 1 060 m3/d, reduced to 272 m
3/d (74 % reduction on average) with active dewatering
of nine (9) dewatering wells and 37 m3/d (97 % reduction on average) with active dewatering of
seventeen (17) dewatering wells (see Figure S2 below).
The natural filling of the pit, to an elevation of 1476 mamsl will take approximately 3 960 days,
almost 11 years after closure. The average flow rate will be at 617 m3 /d, reducing over time to
average flows of 442 m3 /d. If the opencast pit is allowed to fill above the recommended water
level of 1476 mamsl, outflow of contaminated water to the Suikerbosrand River could
approximate > 280 m3/d.
The average alluvial flow will approximate 76 m3 /d without dewatering and roughly contribute
7 % of groundwater inflow to the pit. Active dewatering will increase the flow through the
alluvium to an average of 98 m3 /d and 121 m
3 /d, for the 9 and 17 dewatering well scenarios
respectively (Scenario 2).
0
500
1000
1500
2000
2500
3000
3500
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650 4015 4380
Time (d)
Inflow
(m3/d) Scenario 1
Scenario 2_9 Wells
Scenario 2_17 Wells
Pump_9 Well
Pump_17 Well
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
6/56
April 2006 iv 7475/8297/7/G
GOLDER ASSOCIATES
Operation of the opencast pit will lead to a similar development of a dewatered cone as for
Alternative A, but extending to 710 m South and 860 m to the East of the opencast footprint.
Scenario 2 indicates an approximate area of influence extending to 860 m South and 960 m east.
By characterising the Suikerbosrand River as a gaining stream only, the developed dewateringcone will extend up to 980 m South and 1 125 m East, with the increase of dewatering wells to
seventeen (17).
There is no indication that the development of the dewatering cone, during operation of the
Heidelberg opencast pit, will influence current water supply boreholes, under hydraulic continuity
or no hydraulic continuity conditions (Alternative A or B).
Water quality from the deeper semi-confined aquifer is of a good quality. Most of the pumped water
for mine dewatering purposes during the operational phase will be derived from this aquifer. It can
therefore be safely assumed that the pumped water can be discharged into the Suikerbosrand Riverwithout any detrimental effect to the environment. In-pit management practices during the operational
phase should be such that no potential contaminated surface water be allowed the time to infiltrate
into the groundwater system.
Figure S2: Alternative B - Pit Inflows
0
500
1000
1500
2000
2500
3000
3500
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650 4015 4380
Time (d)
Inflow(
m3/d) Scenario 1
Scenario 2_9 Wells
Scenario 2_17 Wells
Pump_9 Wells
Pump_17 Wells
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
7/56
April 2006 v 7475/8297/7/G
GOLDER ASSOCIATES
Key uncertainties in the study are:
The extent of influence from the peat area (south-west of the proposed Heidelberg opencast pit) to
flows in the groundwater environment.
The hydrogeological properties of the Suikerbosrand riverbed.
The study aimed at developing a first order understanding of the groundwater impacts associated with
the Heidelberg opencast pit operation. To narrow down the envelope of possibility, the numeric model
results should be re-evaluated, once additional information is available from construction of the
dewatering well field and extension of the hydrocensus.
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
8/56
April 2006 vi 7475/8297/7/G
GOLDER ASSOCIATES
TABLE OF CONTENTS
SECTION PAGE
1 INTRODUCTION.................................................................................... 12 OBJECTIVES......................................................................................... 13 METHODOLOGY................................................................................... 24 LOCATION............................................................................................. 25 TOPOGRAPHY...................................................................................... 26 DRAINAGE ............................................................................................ 47 RAINFALL ACROSS STUDY AREA....................................................... 48 EVAPORATION ..................................................................................... 49 GEOLOGY............................................................................................. 5
9.1 Regional Geology.............................................................................. 59.2 Site Specific Geology ........................................................................ 5
10 HYDROGEOLOGY ................................................................................ 710.1 Aquifers............................................................................................. 710.2 Groundwater Flow Directions and Gradients ..................................... 810.3 Water level depth below surface and blow yields............................... 810.4 Aquifer test analyses and Hydraulic Parameters................................ 8
11 GROUNDWATER FLOW MODEL DEVELOPMENT.............................. 812 CONCEPTUAL MODEL....................................................................... 10
12.1 Modelling assumptions .................................................................... 1012.2 Software Selection (Modelling Code)............................................... 11
13 CONSTRUCTION OF THE NUMERIC FLOW MODEL ........................ 1113.1 Finite Element Grid.......................................................................... 11
13.2 Model Boundaries............................................................................ 1313.2.1 Model Boundary Conditions............................................... 14
13.3 Calibration ....................................................................................... 1414 UTILIZING THE GROUNDWATER FLOW MODEL TO ADDRESS
THE STUDY OBJECTIVES.................................................................. 1814.1 Introduction...................................................................................... 1814.2 Operational Mining Phase ............................................................... 18
14.2.1 Mining method ...................................................................1814.3 Modelled Scenarios......................................................................... 2014.4 Scenario Discussion ........................................................................ 21
14.4.1 Operational Phase............................................................. 2114.4.2 Developing Cone of Depression......................................... 24
14.4.3 Closure Phase...................................................................3114.4.4 Contribution from Alluvium................................................. 34
14.5 Groundwater Quality........................................................................ 3414.5.1 Local Background Groundwater Quality
Concentrations................................................................... 3415 CONCLUSIONS AND RECOMMENDATIONS..................................... 3916 REFERENCES..................................................................................... 44
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
9/56
April 2006 vii 7475/8297/7/G
GOLDER ASSOCIATES
LIST OF FIGURES
Figure 1: General Geographical Reference Map .................................................................................3
Figure 2: Monthly Rainfall Record for Heidelberg (1901 2000)....................................................... 4
Figure 3: Monthly Evaporation Record for Heidelberg (1901 2000) ................................................ 5
Figure 4: General Geological Reference Map.....................................................................................6
Figure 5: Schematic Representation of the Heidelberg Environment................................................... 7
Figure 6: Regional Piezometric Water levels and Flow Directions...................................................... 9
Figure 7: Model Matrix (Finite Element Grid).................................................................................. 12
Figure 8: Three Dimensional Model Domain ................................................................................... 13
Figure 9: Calibrated Water levels.....................................................................................................15
Figure 10: Geological Representation of Modelled Area and Water Levels ...................................... 15
Figure 11: Frequency Distribution of Conductivities for Shallow Aquifer system............................. 16
Figure 12: Frequency Distribution of Conductivities for Deep Aquifer system ................................. 17
Figure 13: Transient Calibration of Borehole D................................................................................18
Figure 14: Proposed Heidelberg Opencast Pit Development & Dewatering Well Field ..................... 19
Figure 15: Pit Inflows and Pump Rates for Alternative A .................................................................22
Figure 16: Pit Inflows and Pump Rates for Alternative B .................................................................23
Figure 17: Cone of Depression Scenario 1 (Alternative A) ............................................................25
Figure 18: Cone of Depression Scenario 1 (Alternative B) ............................................................26
Figure 19: Cone of Depression Scenario 2 (Alternative A) [9 Well Dewatering]............................ 27
Figure 20: Cone of Depression Scenario 2 (Alternative B) [9 Well Dewatering] ............................ 28
Figure 21: Cone of Depression Scenario 2 (Alternative A) [17 Well Dewatering].......................... 29
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
10/56
April 2006 viii 7475/8297/7/G
GOLDER ASSOCIATES
Figure 22: Cone of Depression Scenario 2 (Alternative B) [17 Well Dewatering] .......................... 30
Figure 23: Head in Pit vs. Fill Rate for Alternative A....................................................................... 32
Figure 24: Head in Pit vs. Fill Rate for Alternative B ....................................................................... 33
Figure 25: Alluvial Inflows for Alternative A...................................................................................35
Figure 26: Alluvial Inflows for Alternative B ................................................................................... 36
Figure 27: Shallow Aquifer Chemistry.............................................................................................38
Figure 28: Deep Aquifer Chemistry .................................................................................................38
Figure 29: Groundwater Flow Lines at 1476 mamsl ......................................................................... 39
LIST OF TABLES
Table 1: Calibrated Hydrogeological Parameters..............................................................................16
Table 2: Modelled Operational Timeframe.......................................................................................20
Table 3: Pit Water Elevation & Inflow Rates at Closure................................................................... 31
Table 4: Background Groundwater Qualities.................................................................................... 37
Table 5: Comparative Scenario Results for Pit Operation.................................................................43
LIST OF APPENDICES
Appendix A Groundwater Figures
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
11/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
1
1 INTRODUCTION
Anglo Coal Projects Services (ACPS) initiated the feasibility of developing the Heidelberg opencast
mine with an operational life of approximately twelve years for long-term coal supply to Eskoms
Grootvlei power station, which is to be re-commissioned.
Golder Associates Africa (GAA) was appointed to determine the possible impacts that the proposed
opencast mine will have on the groundwater regime in the area. A key aspect of the Heidelberg in-pit
water balance is the determination of groundwater inflows into the opencast workings during
operation and post mine closure.
To characterise the prevailing hydrogeological conditions in the area of the proposed opencast mine,
Golder launched a hydrogeological field investigation to characterise the aquifer(s) within which the
proposed opencast mine will function. A conceptual hydrogeological model was subsequently
developed for the study area based on results forthcoming from this field investigation (GAA Report
No: 7475/8235/4/G, March 2006).
GAA was appointed to construct and calibrate a numerical flow model based on the developed
conceptual model (GAA Report No: 7475/8235/4/G, March 2006), to determine the impacts that the
proposed mine will have on the groundwater regime of the area. The model is used to simulate the
groundwater flow through the groundwater system.
This report describes the modelling process and modelled impacts of the opencast mine on the
groundwater regime in the area investigated.
2 OBJECTIVES
The main objectives of the groundwater modelling study are to:
Construct and calibrate a numerical groundwater flow model, to accurately simulate groundwater
flow through the aquifers
Simulate inflow rates into the mining sections using the preliminary mine plans in conjunction
with the calibrated groundwater flow model
Determine the necessity of dedicated dewatering wellfields around mining sections
Simulate the potential influence of the Suikerbosrand River on mining
Simulate the extent of the influence of mining operations on the hydrogeological regime in the
study area.
Determine the influence of mining on community boreholes.
Simulate the effect of mine closure on groundwater levels and recommend an operational pit
groundwater/surface water level elevation to prevent decant of groundwater / surface water into
the Suikerbosrand River.
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
12/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
2
Identify the need for any additional hydrogeological field investigations in order to reduce
uncertainties identified from the hydrogeological flow modelling investigation conducted.
3 METHODOLOGY
The approach followed in this groundwater modelling study was to focus only on the proposed
Heidelberg opencast pit and its immediate environment, based on available data.
The study was therefore aimed at creating a first-order understanding of groundwater impacts through
numeric modelling and to develop knowledge gaps and recommendations for refinement of the model
in the future.
In order to achieve the objectives of the investigation the following methodology was adopted:
A desk study was conducted on information made available by Anglo Coal and the previous work
undertaken (WMB Report No: 3758/1713/1/W, March 1999)
Development of a conceptual model for the Heidelberg opencast pit, based on the desk study and
information available from Golder (GAA Report No: 7475/8235/4/G, March 2006)
Selection of the most appropriate modelling code
Model construction and set-up
Calibration of the model for steady state groundwater flow conditions, based on average water
levels measured during the hydrocensus conducted in 2005 across the study area
Calibration of the model for transient state groundwater flow conditions based on historic aquifer
test data collected from deep boreholes (WMB Report No: 3758/1713/1/W, March 1999)
Conduct model simulations in order to evaluate the groundwater impact associated with the
proposed development of the Heidelberg opencast pit over time
4 LOCATION
The proposed opencast is to be located on portion 6 of the farm Elandsfontein 412IR approximately
12 km south-east of Heidelberg Gauteng province in South Africa. The mine will be bound to the east
by the Suikerbosrand River and to the south by the R549, Heidelberg to Deneysville road (Figure 1).
Also indicated on Figure 1, is the location of the boreholes identified during the hydrocensus
conducted by Golder (2005), as a general geographical reference.
5 TOPOGRAPHY
The studied area falls within the South Rand Coal field comprising Karoo sediments. The area is
surrounded by hills of the Witwatersrand Supergroup and the Ventersdorp Group.
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
13/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
3
Figure 1: General Geographical Reference Map
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
14/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
4
The Suikerbosrand River rises in the eastern portion of the Vaal River catchment near Devon and
Leandra, from where it flows in a south-westerly direction, before turning to the north-west upstream
of its confluence with the Blesbokspruit (GAA Report No: 7475/8290/6/W, April 2006).
6 DRAINAGE
The natural groundwater level is mostly controlled by the local topography, with groundwater
draining towards surface stream channels. Below its confluence with the Blesbokspruit, the
Suikerbosrand River flows in a south-westerly direction for about half its length before turning to a
westerly course until its confluence with the Vaal River at Vereeniging. This portion of the catchment
is also largely undeveloped and dominated by grassland and dry land agriculture (GAA Report No:
7475/8290/6/W, April 2006).
7 RAINFALL ACROSS STUDY AREA
The rainfall record for Heidelberg (1901 2000) is depicted on a monthly basis in Figure 2. The
average mean annual precipitation (MAP) for Heidelberg is 687 mm/a. Rainfall is strongly seasonal
with most rain occurring in the summer period (October to April) The peak rainfall months are
December and January (GAA Report No: 7475/8290/6/W, April 2006).
Figure 2: Monthly Rainfall Record for Heidelberg (1901 2000)
8 EVAPORATION
The mean annual evaporation (MAE) of Heidelberg equals 1625 mm/a. The evaporation record is
depicted on a monthly basis in Figure 3. The highest Class A-pan monthly evaporation is in January
(range 180 mm to 260 mm) and the lowest evaporation is in June (80 mm to 110 mm) (GAA ReportNo: 7475/8290/6/W, April 2006).
0
50
100
150
200
250
300
350
400
450
500
1900 1905 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000
Date (year)
Rain(mm/m)
MMP
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
15/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
5
Figure 3: Monthly Evaporation Record for Heidelberg (1901 2000)
9 GEOLOGY
9.1 Regional Geology
The studied area falls within the South Rand Coal field comprising Karoo sediments. The area is
surrounded by hills of the Witwatersrand Supergroup and the Ventersdorp Group. Figure 4 shows the
regional geology of the study area as obtained from the 1:250 000 map of Heidelberg
(2628 East Rand). Generally the geology comprises sandstone, mudstone, shale and coal of the
Vryheid formation in the Karoo Supergroup. Geological structures include the Suikerbos and
Malanskraal fault to the south and east of the study area.
Regionally the coal deposit at Heidelberg comprises an infilled glacial valley on the edge of the South
Rand Karoo basin. The deposit itself is therefore relatively localised and elongated NW SE. It
underlies the flood plain of the Suikerbosrand River and also the area to the west of the River where
the topography rises.
The relatively straight North - South drainage of the Suikerbosrand River, indicates the river channel
to be structurally controlled.
9.2 Site Specific Geology
Locally the flood plain in the study area is characterised by a shallow soil profile, located on top of
the Karoo strata. The flood plain is the product of hillwash from the surrounding hills and the
alluvium associated with the Suikerbosrand River. The shallow soil profile comprises dark, soft tofirm, sandy clay and clay rich silty sands.
0
20
40
60
80
100
120
140
160
180
200
1 2 3 4 5 6 7 8 9 10 11 12
Month
Evaporation(mm/m)
MME
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
16/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
6
Figure 4: General Geological Reference Map
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
17/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
7
The modelled area is underlain by a layer of alluvium and sandy clays ( 6 m thick), overlaying shales
and sandstones ( 6 m thick). The Karoo rocks are 27 m thick and contain the No. 3 coal seam with
the largest economic potential.
A schematic cross section of the geological environment is presented in Figure 5.
Figure 5: Schematic Representation of the Heidelberg Environment
10 HYDROGEOLOGY
10.1 Aquifers
A shallow aquifer system is present in the floodplain of the Suikerbosrand River and surrounding
hills, comprised of alluvial material and hill-wash. These sandy clays are saturated and limited in
extend, which makes it an inefficient semi-confined aquifer, as it still allows seepage through to the
underlying geology, developing a perched water level at shallow depth (at the base of residual or
transported soils) above the unweathered rock layers.
A deeper aquifer system is associated with the fractured Karoo sandstone and coal seams and will be
the main system contributing to groundwater flow into the proposed Heidelberg opencast pit. This
deeper aquifer is confined to semi-confined, non-continuous and multi-layered, causing elevated
piezometric heads and in some instances artesian conditions.
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
18/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
8
10.2 Groundwater Flow Directions and Gradients
The natural groundwater level is mostly controlled by the local topography, with groundwater
draining towards surface stream channels.
The average water level distribution was estimated by the hydrocensus conducted and has been used
to calibrate the model for steady state conditions as represented in Figure 6. In general groundwater in
the study area flows towards the Suikerbosrand River (from the East) and the Blesbokspruit (from the
North).
The measured groundwater elevations, indicate that groundwater levels mirror the topography
(see Figure 6) and confirms that the groundwater flow is to the Suikerbosrand River and
Blesbokspruit at a gradient of approximately 1:300. Locally no preferential pathways for the
movement of groundwater away from the site, along fracture zones or deep weathered zones, have
been identified.
Artesian conditions are present in a peat area to the south-west of the proposed opencast pit, where
groundwater flows from subsurface to the ground surface, due to confining to semi-confining
conditions within the deeper aquifer system.
10.3 Water level depth below surface and blow yields
The newly drilled shallow boreholes (GAA Report No: 7475/8235/4/G, March 2006) in the study
area, indicating water levels of 7 to 9 meters below ground level (mbgl) with blow yields ranging
from
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
19/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
9
Figure 6: Regional Piezometric Water levels and Flow Directions
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
20/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
10
WMB report (now Golder Associates Africa (Pty)Ltd), 3758/1713/1/W (March 1999): Heidelberg
Opencast Coal Mine Specialist Studies for Surface and Groundwater
Hydrocensus, drilling and aquifer testing data (GAA Report No: 7475/8235/4/G, March 2006)
Current Heidelberg pit layout data in digital format, as provided by Anglo Coal
Lithological boundaries as from the geological database, distinguishing between soft (alluvium &
sandy clays), hard (sandstone & shale) and base of the No. 3 coal seam (coal seams &
sandstone / shale layers)
This data was used in the development of a conceptual model and finite element numerical model,
simulating the impact on groundwater from the operation of the proposed Heidelberg opencast coal
mine, starting in 2007 to 2018.
12 CONCEPTUAL MODEL
Groundwater flow modelling depends on the physical properties of the site. For a numerical model to
be useful as an assessment tool, it is necessary to integrate the physical geometry and properties of the
site into the model. Controlling factors are:
The geology, as well as the hydrogeological properties of the aquifer system
The topography and relief
Surface hydrology and precipitation
A conceptual model reduces the actual problem and domain, to an acceptable simplification, based on
a set of assumptions.
12.1 Modelling assumptions
The modelling assumptions used in the development of the conceptual model for the Heidelberg
opencast pit groundwater system, included:
The model boundary is represented by topographical highs (along water sheds) based on a digitalterrain model constructed from the 1:50 000 topographical maps of the study area (2627 & 2628).
Thickness of the modelling layers were based on an average thickness of the soft, hard and floor
of the No. 3 coal seam data, provided by Anglo. These layers are underlain by pre-Karoo
basement rock
Alluvial zones were spatially assigned according to the 1:250 000 geological map (2628 East
Rand). Exploration drilling across the footprint of the proposed opencast pit area confirmed that
no paleo channel is present
Operational life of the Heidelberg opencast pit is from 2007 to 2018 (12 years). After operation,the void within the pit will be allowed to naturally fill with water
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
21/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
11
12.2 Software Selection (Modelling Code)
FEFLOW is the software package selected for this modelling study. It provides a sophisticated three-
dimensional finite element modelling environment and allows interactive simulation of groundwater
flow systems within the subsurface. The software is also capable of simulating cross-sectional, fluiddensity-coupled, thermohaline or uncoupled systems. It also handles variably saturated, transient or
steady state flow, mass and heat transport systems in 3D. These systems could be simulated with or
without one or more free surfaces.
The main advantage of using the finite element approach (above the finite difference approach) is that
the boundaries of various features can be accurately represented.
13 CONSTRUCTION OF THE NUMERIC FLOW MODEL
A groundwater flow model was developed for the Heidelberg opencast pit in order to simulate
operational and post closure groundwater flow conditions. The model was calibrated for pre-mining
steady state conditions; these served as starting heads for the transient simulations, in which the effect
of the opencast pit on the groundwater environment was considered.
13.1 Finite Element Grid
The FEFLOW pre-processing software was used to generate a 6-noded triangular prism element
network across the area investigated (Figure 7).
The grid mesh consists of 300 448 elements and 190 005 nodes. Refinement of the grid mesh (finer
density, closer nodal spacing) was specified along the footprint area of the proposed Heidelberg
opencast pit, where a more accurate solution of groundwater flow is required.
The model consists of four (4) layers as schematically shown in Figure 5:
The first layer with a thickness of 6 m represents alluvial and sandy clay horizons, where these are
present. Elsewhere sandstone and shale formations were assigned to this layer where they outcrop
according the geology map of the area
The second layer represents sandstone and shale, approximately 6 m thick
The third layer represents sandstone and the bottom of the No. 3 Coal Seam, approximately 5 m
thick
The fourth layer represents fractured sandstone , approximately 22 m thick
A three dimensional view of the modelling grid is provided in Figure 8.
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
22/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
12
Figure 7: Model Matrix (Finite Element Grid)
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
23/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
13
Figure 8: Three Dimensional Model Domain
13.2 Model Boundaries
A surface catchment approach was adopted to reduce or eliminate the influence of incorrectly
specified boundary conditions in the numerical model (model boundaries are selected on localwatersheds, i.e., the model covers a larger area and thus the extent of potential dewatering impacts
will not be constrained by the size of the modelling domain). The natural catchment boundaries were
determined from an investigation of the local topography maps and subsequently assigned to the
model.
The model boundary includes the catchments of the Suikerbosrand River and Blesbokspruit
respectively to the east and north of the proposed mine site (Figure 7).
The modelled area covers 1 150 800 000 m2, with the proposed Heidelberg opencast pit covering
1 191 350 m2.
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
24/56
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
25/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
15
Figure 9: Calibrated Water levels
Figure 10: Geological Representation of Modelled Area and Water Levels
R = 0.95
1460
1480
1500
1520
1540
1560
1580
1460 1480 1500 1520 1540 1560 1580
Measured (mamsl)
Calculated(mamsl)
Observed
Modelled
R2=0.9043
-80000 -75000 -70000 -65000 -60000 -55000 -50000 -45000
-2965000
-2960000
-2955000
-2950000
-2945000
-2940000
-2935000
-2930000
-2925000
-2920000
Suikerbosrand River
Blesbokspruit
Pit
A
B
C
Zone:
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
26/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
16
A sensitivity analysis was conducted on the material underlying the alluvium, as a measure to address
the uncertainty in the properties of the material underlying the Suikerbosrand River. This analysis did
not have a significant impact on the calibration result of the model with regards to the water level
distribution.
Table 1: Calibrated Hydrogeological Parameters
Hydraulic Conductivity (m/d)Layer
Thickness
(m)Zone
AZone
BZone
C
Storage
Compressibility
(Operation)
1 6 0.24 0.11 0.03 1.28 E-4
2 6 0.24 0.11 0.03 1.28 E-4
3 5 0.19 0.11 0.03 1.28 E-4
4 22 0.24 0.11 0.03 1.28 E-4
Aquifer Test - 0.001-0.8 0.02-1.3 - -
Recharge
(% MAP)- >3
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
27/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
17
The calibrated groundwater flow model was further validated using transient abstraction and water
level data obtained from the aquifer tests conducted by Wates Meiring & Barnard (1997) [now Golder
Associates Africa (Pty)Ltd]. This was conducted in order to ensure that the model is sufficiently
calibrated on a small scale around the proposed footprint area of the Heidelberg opencast pit.
Figure 12: Frequency Distribution of Conductivities for Deep Aquifer system
A well-calibrated model will ensure that reliable estimates of flow rates and impacts on the
groundwater regime are approximated during operation of the Heidelberg opencast pit. A storage
coefficient value of 0.005 was assumed for the whole system modelled. This is reconcilable with
values for storativity typically encountered for Karoo aquifers.
Figure 13 shows the simulated and observed water level response for borehole D during a constant
discharge aquifer test. It is clear from the results obtained, and taking into account borehole losses,
that the flow model is capable of simulating the real aquifer test data sufficiently well to allow
realistic simulation predictions.
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3
Hydraulic Conductivity (m/d)
0
1
2
3
4
F
equency
Calculated values from aquifer tests
Value used in model for shale
Value used in model for sandstone
Value used in model for sandstone
Value used in model for sandstone
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
28/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
18
Figure 13: Transient Calibration of Borehole D
14 UTILIZING THE GROUNDWATER FLOW MODEL TO ADDRESS THE
STUDY OBJECTIVES
14.1 Introduction
The first objective of this study (see Section 2) is to develop and calibrate a numerical flow model
(discussed in Section 12) that accurately simulates groundwater flow through the aquifer systems to
and from the proposed Heidelberg opencast pit. The construction and calibration of the flow model
were discussed in detail in Section 13. This section describes the simulated inflows and impact
assessment on the groundwater environment for the operational and post closure phase of the
Heidelberg opencast pit with and without a dedicated pit dewatering system.
14.2 Operational Mining Phase
14.2.1 Mining method
The British Box-cut method of strip-mining using truck and shovel equipment will be used during the
operational phase with a planned production life of the mine at 12 years, indicated by Figure 14.
For each strip the topsoil is removed, followed by the overburden, and finally the coal. Spoils will be
loaded on trucks and selectively dumped in the void left by mining, leveled and the topsoil will be
replaced, leveled and re-vegetated.
1476
1476
1477
1477
1478
1478
1479
1479
1480
1480
0 0.2 0.4 0.6 0.8 1 1.2 1.4Time (d)
Elevation(mamsl)
Test
Model
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
29/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
19
Figure 14: Proposed Heidelberg Opencast Pit Development & Dewatering Well Field
With the British Box-cut method, rehabilitation is undertaken much sooner after the coal is extracted
and the area of disturbance at any one given time is much smaller. It also allows for more flexibility
during operation for materials handling.
The area that will comprise the active mining area and ramps is relatively small, due to the zigzag
nature of the mining operation.
This proposed mining method was introduced to the model on a monthly basis, with the use of
constant head conditions, as a function of the lowest block elevation and constrained as to only allow
water to be removed by the heads once active mining commence in a particular block.
Legend
2007\Jan - 2007\Dec
2008\Jan - 2008\Dec2009\Jan - 2009\Dec
2010\Jan - 2010\Dec
2011\Jan - 2011\Dec
2012\Jan - 2012\Dec
2013\Jan - 2013\Dec
2014\Jan - 2014\Dec
2015\Jan - 2015\Dec
2016\Jan - 2016\Dec
2017\Jan - 2017\Dec
2018\Jan - 2018\Dec
-2946000 -2946000
-71000
-71000
-71395mE
-71395mE
-70059mE
-70059mE
-2945000mN
-2946617mN -2946617mN
-2945000mN
Dewatering Borehole Locations
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
30/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
20
Operational timeframes is provided in Table 2, as proposed by Anglo Coal.
Table 2: Modelled Operational Timeframe
Rehabilitation Start of operation date Completion date
Block 1 2007/01/01 2007/12/31
Block 2 2008/01/01 2008/12/31
Block 3 2009/01/01 2009/12/31
Block 4 2010/01/01 2010/12/31
Block 5 2011/01/01 2011/12/31
Block 6 2012/01/01 2012/12/31
Block 7 2013/01/01 2013/12/31
Block 8 2014/01/01 2014/12/31
Block 9 2015/01/01 2015/12/31
Block 10 2016/01/01 2016/12/31
Block 11 2017/01/01 2017/12/31
Block 12 2018/01/01 2018/12/31
14.3 Modelled Scenarios
Two alternatives were modelled to simulate the groundwater inflows to the proposed Heidelberg
opencast pit and the subsequent impact on the surrounding groundwater environment. These two
alternatives are:
Alternative A It is assumed that the constant heads specified on the Suikerbosrand River and
Blesbokspruit can supply water into the modelling domain or remove water from the modelling
domain depending on the hydraulic gradient present. This situation would represent a worst case
scenario in terms of modelled inflow into the pit (for the set of calibrated hydraulic properties) as
additional water is derived into the model through the constant head boundary conditions specified on
the rivers.
Alternative B It is assumed that the constant heads specified on the Suikerbosrand River and
Blesbokspruit can only remove water from the modelling domain depending on the hydraulic gradient
present. This situation would represent a best case scenario in terms of modelled inflow into the pit
(for the set of calibrated hydraulic properties) as no additional water is derived from outside the model
domain through river inflow.
These two different approaches (scenarios) were adopted to address the uncertainty as to whether the
Suikerbosrand River is in hydraulic continuity with the underlying strata or not.
For each of the above described alternatives simulations were conducted for a Base Case scenario
(Scenario 1) and a Base Case plus dewatering from dedicated dewatering boreholes scenario
(Scenario 2):
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
31/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
21
Scenario 1 Base Case: Development of the Heidelberg opencast pit by truck and shovel
method, from 2007 till 2018, excluding a dedicated dewatering well field.
Scenario 2 Base Case plus Dewatering: Develop the Heidelberg opencast pit by truck and
shovel method, from 2007 till 2018, including a dedicated dewatering well field consisting of 9and 17 dewatering boreholes.
Assumptions made for Scenario 1 (Operational Phase): Operation of the Heidelberg opencast pit
will initiate in 2007 and function by truck and shovel method on a monthly time step over a period of
twelve years. Completion of the Heidelberg opencast pit is in 2018.
Assumptions made for Scenario 2 (Operational Phase): The same assumptions have been made as
for Scenario 1, with the following addition: Operation of the Heidelberg opencast pit will include a
dewatering well field, represented by nine (9) and seventeen (17) fully penetrating extraction
boreholes on the perimeter of the opencast footprint area, actively dewatering the system from
6 months prior to operation in 2007 at a pump rate of approximately two litres per second (2 l/s) per
borehole.
The initial estimation of the 9 borehole dewatering field was based on the average inflows into the pit
(approximating 1 800 m3/d), estimated from Scenario 1 and also taking into account the six months of
active dewatering prior to operation. This related to an approximate extraction rate of 1 555 m3/d,
from the nine wells, pumping at 2 l/s. Boreholes were then added between the initial 9 dewatering
wells, to more effectively overlay the dewatering interference from the well field, around the
perimeter of the proposed opencast footprint. This established the 17 dewatering wells, pumping at2 l/s (approximately 2 938 m
3/d).
Assumptions made for the Post closure phase: Changes in recharge conditions (15 % of MAP) and
hydraulic conductivities (8.64 m/d) are assumed after 2018 over the opencast footprint, from which
date the opencast pit was allowed to naturally fill with water. Natural filling of the opencast pit was
without the operation of the dewatering well field, considering both alternatives over a 12 year period.
14.4 Scenario Discussion
14.4.1 Operational Phase
Groundwater flows into the pit, over the operational period are represented in Figure 15 and Figure 16
for Alternative A and Alternative B respectively. A comparative evaluation of the base case scenario
(Scenario 1) and the dewatering well field (Scenario 2), by interpretation of the graphical results from
these figures indicate that the average influx of water into the pit will be approximately 1 820 m3/d,
reduced to 1 160 m3 /d (36 % reduction on average) with active dewatering of nine (9) dewatering
wells and 825 m3/d (55 % reduction on average) with active dewatering of seventeen (17) dewatering
wells, for Alternative A (see Table 5).
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
32/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
22
Figure 15: Pit Inflows and Pump Rates for Alternative A
0
500
1000
1500
2000
2500
3000
3500
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650 4015 4380
Time (d)
Inflow(
m3/d) Scenario 1
Scenario 2_9 Wells
Scenario 2_17 Wells
Pump_9 Well
Pump_17 Well
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
33/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
23
Figure 16: Pit Inflows and Pump Rates for Alternative B
0
500
1000
1500
2000
2500
3000
3500
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650 4015 4380
Time (d)
Inflow(
m3/d) Scenario 1
Scenario 2_9 Wells
Scenario 2_17 Wells
Pump_9 Wells
Pump_17 Wells
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
34/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
24
For Alternative B the average influx of water into the pit will be approximately 1 060 m3/d, reduced to
272 m3 /d (74 % reduction on average) with active dewatering of nine (9) dewatering wells and
37 m3 /d (97 % reduction on average) with active dewatering of seventeen (17) dewatering wells
(see Table 5).
The average pump rate will approximate 1 136 m3/d to 1 120 m
3/d with nine (9) dewatering wells and
1 829 m3/d to 1 726 m
3/d with seventeen (17) dewatering wells for Alternative A & B respectively.
From the simulations conducted for the operational phase the following is evident:
The difference in the inflows to the pit is significant between Alternative A (river is in hydraulic
continuity to allow inflow or outflow of water from the system) and Alternative B (river is not in
hydraulic continuity and only allow the outflow of water from the system). The model is therefore
highly sensitive as to the uncertainty that exists whether the river is in hydraulic continuity with
the underlying aquifer.
Dewatering using dedicated abstraction boreholes in advance of mining will significantly reduce
inflow of groundwater into the mine workings.
A trend of incremental groundwater inflow increases is observed over the 12 year operational life
of the pit, as a direct result of generally increases in floor elevations over time as the pit develops.
14.4.2 Developing Cone of Depression
The development of the cone of depression (impact on the surrounding groundwater environment)from 2007, with the start of the Heidelberg opencast pit, to the year 2018, are presented in
Appendix A [indicated on the sequential figures is the groundwater drawdown in meter(s) from the
calibrated steady state water level = Cone of Dewatering].
This development of the cone of depression around the footprint of the Heidelberg opencast pit over
time is indicating that the development will follow the progression of the mine, starting in the South-
West in 2007 and extending to the North until 2018.
The cone of depression will extend to the Suikerbosrand River in the East and could have a zone of
influence approximating 1 400 m to the North, 615 m to the South and 410 m to the West, with no
active dewatering for Alternative A, as indicated in Figure 17 (Scenario 1). Alternative B indicates
that the zone of influence from dewatering could extend up to 710 m in the South and 860 m to the
East of the opencast footprint (Figure 18).
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
35/56
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
36/56
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
37/56
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
38/56
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
39/56
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
40/56
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
41/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
31
14.4.3 Closure Phase
Natural refilling of the opencast pit from the year 2018, to an elevation of 1476 mamsl, one metre
below the base of the Suikerbosrand River (1477 mamsl), is proposed to ensure that any possible
contaminated water in the opencast pit will not flow into the Suikerbosrand River and will takeapproximately 1 980 days or more than 5 years, for Alternative A and approximately 3 960 days,
almost 11 years for Alternative B (Appendix A).
The average natural filling of the pit will be relatively high at approximately 840 m3/d and 617 m
3/d,
reducing over time, as driven by the reduction in head differential between outside aquifer and inside
pit head conditions to average flows of 452 m3 /d and 442 m
3 /d, as represented by Figure 23 and
Figure 24 for Alternatives A & B respectively.
Table 3 summarises the natural refilling of the Heidelberg opencast pit under constant head and steady
state conditions. It should be noted that where the flows are expressed as negative values, the flows
are not into the pit, but out of the pit and into the Suikerbosrand River.
Table 3: Pit Water Elevation & Inflow Rates at Closure
Alternative A
(River/Aquifer Hydraulic
Continuity)
Alternative B
(No River/Aquifer
Hydraulic Continuity)Pit Water
Elevation
(mamsl)AverageInflow
(m3/d)
Duration(Days)
AverageInflow
(m3/d)
Duration(Days)
1473 841 1 260 617 2 880
1474 712 1 440 580 3 2401475 582 1 680 523 3 600
1476 452 1 980 442 3 960
1477 323 2 340 317 4 380
1478 -67 (out) 2 790 -70 (out) -
1479 -171 (out) 3 510 -175 (out) -
1480 -280 (out) - -284 (out) -
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
42/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
32
Figure 23: Head in Pit vs. Fill Rate for Alternative A
1455
1460
1465
1470
1475
1480
1485
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Time (d)
Elevation(mamsl)
-1000
0
1000
2000
3000
4000
5000
Flux(m3/d)
Head
Inflow
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
43/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
33
Figure 24: Head in Pit vs. Fill Rate for Alternative B
1455
1460
1465
1470
1475
1480
1485
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Time (d)
Elevation(mamsl)
0
500
1000
1500
2000
2500
3000
3500
Flux(m3/d)
Head
Inflow
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
44/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
34
14.4.4 Contribution from Alluvium
Figure 25 and Figure 26 show the simulated contribution from modelling Layer 1 (shallow aquifer)
into the Heidelberg opencast pit for Alternatives A & B respectively. It is clear that the inflows
increase as the pit development approaches the Suikerbosrand River, under river/aquifer hydrauliccontinuity conditions (Figure 25), as well as an increased flow, due to increased pumping activity
from a 9 and 17 well dewatering field. Average flows from the alluvium, for Alternative A,
approximates 171 m3/d without dewatering and roughly contributes 9 % of groundwater inflows to the
pit. Active dewatering will increase the flow through the alluvium to an average of 214 m3 /d and
265 m3/d, for the 9 and 17 well scenarios respectively.
Average flows from the alluvium, for Alternative B (Figure 26), approximates 76 m3 /d without
dewatering and roughly contributes 7 % of groundwater inflows to the pit. Active dewatering will
increase the flow through the alluvium to an average of 98 m3/d and 121 m
3/d, for the 9 and 17 well
scenarios respectively.
A sensitivity analysis was conducted whereby the hydraulic conductivity of the zone below the
alluvial material (Zone A in Figure 10) was reduced from 0,24 m/d to 0,11 m/d. This simulation was
conducted for Alternative A where river/aquifer hydraulic continuity is assumed. The simulation
indicates a reduction in pit inflow from 1 820 m3/d to 1 250 m3/d a reduction of approximately
31%. This result clearly emphasizes the effect that uncertainty in hydraulic parameters might have on
inflow calculations (and general groundwater impact). The reader is referred again to the range in
field measurements of hydraulic parameters depicted in Figure 11 and Figure 12 vs. the actual values
used in the modelling exercise.
14.5 Groundwater Quality
14.5.1 Local Background Groundwater Quality Concentrations
The pumped water from active dewatering during operation of the Heidelberg opencast pit is to be
discharged into the Suikerbosrand River. It is therefore important to estimate the quality of the water
to be discharged.
Local background groundwater qualities were determined from analysis of current available data for
both the shallow and deep aquifer system.
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
45/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
35
Figure 25: Alluvial Inflows for Alternative A
0
50
100
150
200
250
300
350
400
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650 4015 4380
Time (d)
Inflow(
m3/d)
Scenario 1
Scenario 2_9 Wells
Scenario 2_17 Wells
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
46/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
36
Figure 26: Alluvial Inflows for Alternative B
0
50
100
150
200
250
300
350
400
0 365 730 1095 1460 1825 2190 2555 2920 3285 3650 4015 4380
Time (d)
Inflow(
m3/d)
Scenario 1
Scenario 2_9 Wells
Scenario 2_17 Wells
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
47/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
37
It is clear from Table 4 that unacceptably high concentrations of aluminium and iron are present in the
shallow aquifer system, as evaluated according to the South African Water Quality Guidelines for
Domestic Use. It should also be noted that sodium and sulphate concentrations have been measured at
an unacceptable level. However, the deep aquifer system indicates groundwater of a good quality.
As most of the dewatering water will be derived from the deeper aquifer system during operation, it is
expected that the quality of the water to be discharged should be of an acceptable quality.
Statistical graphs (Box and Whisker plots) showing the chemical distribution of the shallow and deep
aquifer systems is represented in Figure 27 and Figure 28 respectively.
The final in pit void of the Heidelberg opencast pit will be allowed to refill during closure, to a level
one (1) meter below the base of the Suikerbosrand River (1477 mamsl) at 1476 mamsl. This
management strategy will reduce the risk of contaminated water flowing from the pit into the adjacent
Suikerbosrand River, but also allows the pit to function as a sink to groundwater flow.
Figure 29 shows the groundwater flow lines (obtained from particle tracking) for the situation where
the post closure pit water level elevation is maintained at an elevation of 1476 mamsl. It is evident
from this figure that all flow is contained within the pit area.
Any possible contaminated in-pit water will therefore be theoretically contained within the footprint
of the opencast pit, as a result of this proposed in-pit water management strategy.
Table 4: Background Groundwater Qualities
Shallow Aquifer System Deep Aquifer SystemElement Unit
Min Ave Max Min Ave Max
DomesticGuideline
Value
pH 6.40 7.31 8.30 6.38 7.39 8.82 9
Calcium - Ca mg/l 11.80 28.54 73.00 2.60 7.64 11.80 80
Magnesium - Mg mg/l 2.00 17.33 44.00 2.00 4.70 7.10 70
Sodium - Na mg/l 11.80 190.39 503.00 3.60 30.68 65.00 400
Sulphate - SO4 mg/l 4.20 73.72 646.00 1.00 2.13 4.00 400
Aluminium - Al mg/l 0.02 5.84 55.00 0.10 0.10 0.10 0.5
Nickel - Ni mg/l 0.00 0.07 0.63 0.00 0.00 0.00 1
Manganese - Mn mg/l 0.00 0.16 1.00 0.01 0.01 0.01 1Iron - Fe mg/l 0.01 3.02 22.00 0.01 0.06 0.39 1
Zinc - Zn mg/l 0.01 0.03 0.26 0.00 0.00 0.00 10
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
48/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
38
Figure 27: Shallow Aquifer Chemistry
Figure 28: Deep Aquifer Chemistry
0.001
0.01
0.1
1
10
100
1000
Concentration(mg/l)
Ca Mg Na SO4 Al Ni Mn Fe Zn
0.1
1
10
100
Conce
ntration(mg/l)
Ca Mg Na SO4 Fe
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
49/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
39
Figure 29: Groundwater Flow Lines at 1476 mamsl
15 CONCLUSIONS AND RECOMMENDATIONS
It was prudent to use a conservative approach to the numeric modelling, in view of the sensitivity of
the hydrogeological properties of the underlying riverbed material and location of the proposed
Heidelberg opencast pit to the Suikerbosrand River.
The modelled results were based on the proposed alternatives for each scenario evaluated, to
determine the groundwater inflows to the proposed Heidelberg opencast pit and the impact on the
surrounding groundwater environment, during operation and post closure.
The proposed alternatives to the scenarios address the uncertainty in the hydraulic parameters of thematerial underlying the alluvium. This uncertainty relates to the hydraulic continuity (Alternative A)
or no hydraulic continuity (Alternative B) of the Suikerbosrand River and Blesbokspruit to the
shallow aquifer system. The proposed alternatives were based on the presence of clayey soils that
could restrain water flowing from the river to the opencast pit.
Boundary conditions representing the Suikerbosrand River and Blesbokspruit were assigned to either
allow water to discharge or recharge the system from the constant head nodes, depending on
surrounding head conditions for Alternative A (simulating a worst case scenario) or to be discharged
from the system should a positive hydraulic gradient exist towards the constant head nodes and where
the gradient is zero or negative no water would be added into the system for Alternative B (simulates
a best case scenario).
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
50/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
40
The following key conclusions can be drawn from the numerical groundwater modelling presented in
this report.
Alternative A: Worst Case Scenario
The average influx of groundwater during operation will approximate 1 820 m3 /d, reduced to
1 160 m3/d (36 % reduction on average) with active dewatering of nine (9) dewatering wells and
825 m3/d (55 % reduction on average) with active dewatering of seventeen (17) dewatering wells.
The natural filling of the pit, to an elevation of 1476 mamsl (1 m below the base of the
Suikerbosrand River) will take approximately 1 980 days or more than 5 years after closure. The
average flow rate will be at 841 m3/d, reducing over time as driven by the rising head conditions
within the pit, to average flows of 452 m3 /d. If the opencast pit is allowed to fill above the
recommended waterlevel of 1476 mamsl, outflow of contaminated water to the Suikerbosrand
River could approximate > 280 m3/d.
There will be an increase in flows through the alluvium to the pit, as the pit development
approaches the Suikerbosrand River. Average flows will approximate 171 m3 /d without
dewatering and roughly contribute 9 % of groundwater inflow to the pit. Active dewatering will
increase the flow through the alluvium to an average of 214 m3/d and 265 m
3/d, for the 9 and 17
dewatering well scenarios respectively (Scenario 2).
Operation of the opencast pit will lead to the development of a dewatered cone of depression
extending to the Suikerbosrand River in the East, 1 400 m North, 615 m to the South and 410 m to
the West, with no active dewatering (Scenario 1). The footprint area will be influenced by the
nine (9) dewatering well field, operating six (6) months prior to pit development for Scenario 2.This will result in a more extensive area of influence that could approximate 1 470m to the North,
720 m to the South and 570 m to the West.
The extent of dewatering by increasing the number of dewatering wells to seventeen (17), indicate
a cone of depression developing to 1 500 m North, 800 m South, 700 m West and to the
Suikerbosrand River on the East. Operation of a 17 dewatering well field will ensure a more
effective barrier to groundwater inflow into the opencast pit.
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
51/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
41
Alternative B: Best Case Scenario
The average influx of groundwater during operation of the Heidelberg opencast pit will
approximate 1 060 m3/d, reduced to 272 m
3/d (74 % reduction on average) with active dewatering
of nine (9) dewatering wells and 37 m
3
/d (97 % reduction on average) with active dewatering ofseventeen (17) dewatering wells.
The natural filling of the pit, to an elevation of 1476 mamsl will take approximately 3 960 days,
almost 11 years after closure. The average flow rate will be at 617 m3 /d, reducing over time to
average flows of 442 m3 /d. If the opencast pit is allowed to fill above the recommended
waterlevel of 1476 mamsl, outflow of contaminated water to the Suikerbosrand River could
approximate > 280 m3/d.
The average alluvial flow will approximate 76 m3 /d without dewatering and roughly contribute
7 % of groundwater inflow to the pit. Active dewatering will increase the flow through the
alluvium to an average of 98 m3 /d and 121 m3 /d, for the 9 and 17 dewatering well scenarios
respectively (Scenario 2).
Operation of the opencast pit will lead to a similar development of a dewatered cone as for
Alternative A, but extending to 710 m South and 860 m to the East of the opencast footprint.
Scenario 2 indicates an approximate area of influence extending to 860 m South and 960 m East.
By characterising the Suikerbosrand River as a gaining stream only, the developed dewatering
cone will extend up to 980 m South and 1 125 m East, with the increase of dewatering wells to
seventeen (17).
There is no indication that the development of the dewatering cone, during operation of the
Heidelberg opencast pit, will influence current water supply boreholes, under hydraulic continuity
or no hydraulic continuity conditions (Alternative A or B).
Water quality from the deeper semi-confined aquifer is of a good quality. Most of the pumped water
for mine dewatering purposes during the operational phase will be derived from this aquifer. It can
therefore be safely assumed that the pumped water can be discharged into the Suikerbosrand River
without any detrimental effect to the environment. In-pit management practices during the operational
phase should be such that no potential contaminated surface water be allowed the time to infiltrate
into the groundwater system.
Key uncertainties in the study are:
The extent of influence from the peat area (south-west of the proposed Heidelberg opencast pit) to
flows in the groundwater environment.
The hydrogeological properties of the Suikerbosrand riverbed.
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
52/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
42
The study was aimed at developing a first order understanding of the groundwater impacts associated
with the Heidelberg opencast pit operation. In order to develop more confidence in the impacts as well
as evaluate suitable dewatering measures, the numeric model results should be re-evaluated, once
additional information is available to recalibrate the current model.
The recommended action to address the key uncertainty is:
To conduct pumping tests at the identified peat area, to estimate a more accurate groundwater
contribution from the area, to the Heidelberg opencast pit. The development of the dewatering
well field will provide this additional information that should be used for refinement of the model.
To estimate the hydrogeological properties of the Suikerbosrand riverbed by conducting a river
profile investigation.
Additional dewatering boreholes (8) should be allocated for and possibly positioned between the
proposed 9 dewatering wells around the footprint of the proposed Heidelberg opencast pit.
A summary of the numeric modelling results is presented in Table 5 below.
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
53/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
43
Table 5: Comparative Scenario Results for Pit Operation
Alternative A (Worst Case) Alternative B (Best Case)
Scenario 1(Base Case)
Scenario 2(9 Well
Dewatering)
Scenario 2(17 Well
Dewatering)
Scenario 1(Base Case)
Scenario 2(9 Well
Dewatering)
Scenario 2(17 Well
Dewatering)Annual Block Units
Average
Inflow
Average
Inflow
Average
Inflow
% InflowReduction
from using
9-17
DewateringWells
Average
Inflow
Average
Inflow
Average
Inflow
% InflowReduction
from using
9-17
DewateringWells
Block 1 (2007) m3 /d 827.64 452.94 299.01 45-64 722.16 309.34 100.80 57-86
Block 2 (2008) m3 /d 994.72 500.39 307.52 50-69 803.82 266.37 46.95 67-94
Block 3 (2009) m3 /d 1181.72 587.29 351.58 50-70 957.13 301.37 77.23 69-92
Block 4 (2010) m3 /d 1182.62 637.63 416.27 46-65 930.93 290.45 90.53 69-90
Block 5 (2011) m3 /d 1302.52 770.19 548.03 41-58 1017.42 333.40 85.80 67-92
Block 6 (2012) m3 /d 1440.36 847.42 600.22 41-58 1016.60 247.66 34.28 76-97
Block 7 (2013) m3 /d 1965.90 1342.57 1030.16 32-48 1044.88 249.87 9.42 76-99
Block 8 (2014) m3 /d 2193.07 1501.52 1144.21 32-48 1142.68 267.66 5.15 77-100
Block 9 (2015) m3 /d 2299.18 1479.17 1100.29 36-52 1180.92 202.44 0.01 83-100
Block 10 (2016) m3 /d 2799.96 1808.75 1201.40 35-57 1591.69 391.26 0.00 75-100
Block 11 (2017) m3 /d 2687.91 1843.17 1333.68 31-50 1257.54 262.81 0.00 79-100
Block 12 (2018) m3 /d 2875.98 2037.41 1512.56 29-47 1134.01 166.38 0.00 85-100
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
54/56
April 2006 7475/8297/7/G
GOLDER ASSOCIATES
44
16 REFERENCES
WMB report (now Golder Associates Africa (Pty)Ltd), (March 1999): Heidelberg Opencast Coal
Mine Specialist Studies for Surface and Groundwater, Report No: 3758/1713/1/W
Golder Associates Africa (Pty)Ltd Report, (March 2006): Hydrogeological Investigation for the
Proposed Heidelberg Opencast Mine, Report No: 7475/8235/4/G
Golder Associates Africa (Pty)Ltd Report, (April 2006):Heidelberg Surface Water Specialist Report,
Report No: 7475/8290/6/W
GOLDER ASSOCIATES AFRICA (PTY) LTD
H Marais H van Rensburg G L HubertG:\PROJECTS\7475 - HEIDELBERG EIA\REPORTS AND PRESENTATIONS\7475-8297-7-G NUMERIC GROUNDWATER MODELLING\7475-8297-7-G GROUNDWATER
NUMERIC MODELLING RPT.DOC
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
55/56
GOLDER ASSOCIATES
APPENDIX A
Groundwater Figures
8/3/2019 8543-Volume 2-1.4 - Groundwater Numeric Modelling
56/56