Turner, J.V., Barr, A.D., Challen, R.P., Johnson, S.L., Townley, L.R., Wright, K.D., Woodbury, R.J., Watson, G.D., Bartle, G.A., and Gailitis, V. (1996). Groundwater supply to the mining industry in the WA goldfields. Minerals and Energy Research Institute of Western Australia. Report No. 154, 292 pp.
Groundwater allocation and use in the Eastern Goldfields: Data on hypersaline groundwater allocation and use between 1985 and 1994 have been compiled and reported. Data from 1991 indicate paleochannel groundwater use in the study area was about 12 x 106 m3 per year. More recent assessments place this figure closer to 16 x 106 m3 per year.
Hydraulic conductivity measurements of bounding formations: Hydraulic conductivity measurements conducted by triaxial testing revealed extremely low conductivities for the aquitard clays. Slug tests conducted in the field revealed higher conductivities and suggest that structural features such as cracks and fissures in the clays may play a significant role in the storage and release of groundwater from the aquitard into the aquifer.
Regional groundwater conditions - recharge response to heavy rainfall.: The network of continuous water level recorders has been expanded and now twenty three regional monitoring bores into producing aquifers and are being continuously logged. Over the entire study region we have only one water level logger in a paleochannel aquifer that does not show the effects of significant drawdown caused by pumping. This is at GSWA transect KRE in the western end of the paleochannel system. A recharge response to heavy rainfall associated with Cyclone Bobby in early 1995 was observed in the paleochannel aquifer at GSWA transect KRU and is the only instance we have observed of recharge response to rainfall in a paleochannel aquifer.
However, even this response may be due to recovery following the heavy rains as alternative water sources are substituted for groundwater when surface waters are abundant following rain.
Hydrogeochemical and environmental isotope data: Conclusions on the application of chlorine-36 and carbon-14 environmental isotopes have been completed and are presented in two conference papers in Appendices to this report. Some additional data on uranium and sulphur isotopes in groundwaters were collected and in general, confirm trends observed in earlier work. Detailed analysis of this data will be presented in research papers at a later date.
Chloride and hydrogeochemical profiles: New data on chloride concentrations in groundwater along the North Yindarlgooda paleochannel tributary revealed a previously unknown but systematic trend of decreasing salinity. This trend is highly suggestive of the inflow of fresher groundwater from as yet unidentified tributary paleochannel(s). Inspection of the Kurnalpi hydrogeological sheet suggests such tributaries occur near Yarri road and near 22 Mile Dam. Hydrogeochemical profiles in shallow unsaturated zones were obtained from drilling and coring. Qualitatively, these form and high solute concentrations in these profiles confirm our earlier conclusions that recharge to groundwater via regional rainfall-infiltration is not a significant recharge mechanism. A detailed analysis of these profiles will be developed and published later.
Evaporation rates from salt lake surfaces: The development of a portable evaporation dome for measurement of evaporation from salt lake surfaces has been completed. The device has been used to measure evaporation at three locations in the study area and results have been used in the groundwater modelling.
The model: A quasi-two-dimensional model was constructed using a number of vertical one-dimensional aquitard models coupled with a single one-dimensional aquifer model to represent a paleochannel. The model allows for a moving water table within the aquitard using a specific yield model. The model was verified against field measurements of observed drawdowns in nested piezometers screened in clays in a salt lake bed overlying a producing aquifer. Verification was also obtained by comparison with a published analytical model.
Modelling results: The time taken for the effect of pumping in the aquifer to affect the water table in the aquitard was found to be of the order of several months. This was also demonstrated when pumping from a radial aquifer, the aquifer reaches equilibrium in the order of days, whereas it takes the aquitard almost 30 years to reach the same level.
The sources of water derived from pumping were shown to vary over the time scale of days to years. The maximum percentage of water derived from the aquitard was found to be about 8% after pumping for a period of about one month. With increasing time, i.e. of the order of several years, from the commencement of pumping the proportion of pumped water derived from the fixed-head boundaries of the aquifer becomes dominant. Thus our previous estimates indicating that 50 to 60% of the pumped water was derived from aquitard leakage is an overestimate. We must conclude, therefore, that the additional groundwater resource evident by lower than expected drawdowns in producing aquifers is only partially due to leakage from confining formations. The remaining additional resource must be being derived from undefined or unknown minor paleochannel tributaries. An example of this has been identified from the regional pattern of hydrogeochemistry in the Yindarlgooda North paleochannel, where a decrease in salinity has been attributed to inflow from two tributary sources to the north of the main channel.
The optimum method of pumping from the paleochannel system to maximise the yield from the aquitard was found to make use of a larger number of pumps pumping at low rates rather than a small number of pumps pumping at a greater rate for the same net pumpage. Better results were also found when the pumps were spread further apart.
The coupled flow and solute transport model found that the final concentration in the aquifer would depend only on the evaporation if the system was fully mixed. This has previously been studied using the distributed parameter model, and the explanation for the general trend of increasing salinity in the regional groundwaters from west to east was confirmed by the solute transport modelling, with evaporative loss of groundwater through salt lakes being the dominant process causing the increasing salinities.
Flooding and persistence of salt lakes: Periods of significant rainfall were classified into major and minor flooding events based on the rainfall intensity, and examination of rainfall data from 1939 to 1992 determined that approximately two major and one minor flooding events occur each year. Although the frequency of flooding is primarily dependent on the extent and intensity of rainfall, factors such as antecedent conditions, topographic controls and catchment areas are also influential.
Critical factors which affect the persistence of flooding are the catchment area, AC/AL (Catchment Area : Salt Lake Area), geological nature of the catchment, topographic gradient and position. Black Flag Lake is the most persistent of the studied lake systems, which can be attributed to the large catchment area and its topographic position with respect to the surface drainage. The persistence of water bodies within Lake Rebecca and Lake Lefroy is related to the extensive lake area which itself acts as a catchment.
Volumes of water in flooded salt lakes: The estimated volume of flood water from the 1989 rainfall event ranged between 5.5 x 105 m3 for White Flag Lake to 10.8 x 106 m3 for Lake Rebecca. The economic potential of extracting surface water is restricted by the frequency of significant rainfall events. Analysis of historical rainfall data determined that flooding events greater than 1989 occur approximately once every two years.
In late 1993 a survey was conducted to determine the bathymetry of Black Flag lake and thereby establish ground truth and quantify associated errors with the remote sensing analysis. The bathymetry was determined using Magellan GPS units and a plumb bob for position and water depth determination. The volume and area of inundation were accurately calculated at 3.08 x 105 m3 and 2.86 x 106 m3 respectively. The 1989 flooding event was corrected to give a maximum volume of 1.32 x 106 m3 based on image and ground truth data. Algal bed and erosion berm heights respectively were utilised to estimate the maximum 1992 inundation at 4.25 x 106 m3 following significant rainfall events.
Economic significance of surface water use: Such volumes of surface water impoundments are significant in terms of their process water use, and significant potential appears to be available for cost saving on metallurgical reagent use and as a source water for artificial recharge of the paleochannel aquifers.
Model Results: Solving the model using a paleochannel groundwater stock equal to the original estimate of 6.25 x 108 kL produced the optimal time paths of ore and groundwater stock depletion. For all three of the initial ore stocks, ore extraction equals the plant capacity during the early periods. Subsequent extractions decline. For all of the three initial ore stocks, approximately 90% of the ore stock was mined. Because of the constant water to ore ratio in processing, groundwater abstractions followed a similar path to ore extractions.
Rates of groundwater depletion: The paleochannel groundwater resource was depleted and, hence, scarce for the two highest ore stocks, with groundwater depletion after 26 and 24 years for the medium and high ore stocks respectively. The alternative water source was used for 5% and 29% of ore processing for initial ore stocks of 4.5 x 108 and 6.0 x 108 tonnes, respectively. For the highest initial ore stock, the paleochannel water resource was depleted while ore extraction was equal to the plant capacity.
Time paths of resource use with 25% and 50% larger initial paleochannel groundwater stocks were calculated. With the initial stock larger by 25%, the paleochannel water resource is scarce only for the highest initial ore stock, being depleted after 31 years. The alternative water resource is used for approximately 10% of total ore processing. With the initial paleochannel water stock larger by 50%, the resource is relatively abundant for all initial ore stocks.
Benefit of improved resource knowledge: Knowledge of additional groundwater stocks brought forward ore and water extractions slightly in scenarios where a switch to the alternative water resource occurred after the mining rate has been reduced to below the plant capacity.
Regional and local scale implications: The conclusions of this study are based upon an analysis which assumes that the water and ore resources are homogeneous in terms of quality and geographical location relative to points of use. In practice, the exhaustion of groundwater stocks can be expected to occur on a local scale, depending on the extent of groundwater extraction from a given location. The purpose of this analysis was to provide a regional, aggregated indication of the likely lifetimes of the groundwater resource.
Copyright © 2005 by Lloyd Townley