Geohydrological consequences associated with the post-mine closure flooding of dewatered dolomitic karst aquifers in the Far West Rand, South Africa

Abstract

English The thesis presents a hydrologic investigation into the effects of deep-level gold mining on the hydrogeological conditions of a fractured-karst aquifer system in the Far West Rand (FWR) goldfield, South Africa. The goldfield, located some 50 km south-west of Johannesburg, hosts some of the deepest mines worldwide. In contrast to most other goldfields mining here takes place below up to 1.2-km-thick karstified dolomite in which large volumes of groundwater are stored. Nearly impermeable vertical dykes subdivide this dolomitic karst aquifer into several individual ‘groundwater compartments’ displaying distinct differences in water table elevation between them. In an attempt to reduce the large influx of dolomitic groundwater into the underlying mine void, three of the compartments were dewatered by the mines through active pumping. These compartments are investigated. For the foreseeable future, when mining and the associated pumping stops and the mine void and overlying compartments will gradually fill-up with naturally infiltrating groundwater (here termed ‘re-watering’), predictions were made that a total of four compartments will be hydraulically linked through artificial holings in previously impermeable dykes created by mining. With water now flowing across the dykes that used to hydraulically separate adjacent compartments a so called ‘mega-compartment’ may form in which a single groundwater table spans across all four joined compartments. The main objective of this study is to test the validity of this hypothesis and investigate whether or not the predicted scenario is likely to occur. This is of significant practical importance for future water and land use as a ‘mega-compartment’ scenario would leave the final groundwater table far below its original elevation over large parts of the aquifer thus preventing that formerly strong yielding karst springs that dried up due to dewatering will ever flow again. It has also repercussions for the future water availability as well as the quality of ground- and surface water and ground-stability aspects. The results of the study are therefore of crucial importance for designing adequate water management and land-use strategies that proactively address the hydraulic effects of the inevitable mine closure. The prediction of the final post re-watering groundwater levels and groundwater flow directions and rates relies largely on historical data predominantly gathered during the active dewatering of the compartments. The spatial scale covered by the data is exceptionally large compared to similar activities elsewhere with the vertical groundwater drawdown approaching nearly a kilometre in depth affecting surrounding areas over distances of up to several kilometres. The data were analysed by a Darcy-based falling-head approach in order to predict the groundwater flow between compartments, termed inter-compartmental groundwater flow (IGF). Determining the IGF was a critical parameter, as its ratio to groundwater recharge was determined as the key factor for the establishment of the final groundwater levels of all the compartments. Groundwater recharge for re-watered compartments was assessed from a conceptual and semi-quantitative groundwater balance study. The results revealed that the IGF (between 0.7-5.4 Ml/d) will most likely be too low to hydraulically merge together the individual compartments as a mega-compartment. Recharge in re-watered compartments will probably be higher than in pre-mining times due to irreversible hydrogeological changes in the study area (e.g. sinkholes). However, to a large degree it will also depend on the manner of water usage and thus on management decisions. Furthermore the thesis showed that in principle it is possible to analyse the historical data originating from the large-scale dewatering of the dolomitic compartments in the same way as an ultra-large pumping test using ordinary analytical methods (e.g. Theis). The obtained values were 2468 m²/s for the average horizontal transmissivity and 0.67% for the storativity of the dolomitic aquifer. The methods used in the thesis to determine the hydraulic parameters and the groundwater flow are exclusively based on Darcy’s law, initially designed for the saturated flow in porous media, whose applicability to karst aquifers was tested. It turned out that methods generally delivered realistic results. However, persuasive evidence of their applicability exists only from the analysis of the vertical groundwater flow (IGF), whereas in the case of the pumping test analysis it was not possible to finally validate the methods.