Geochemical investigation of union collery underground mine workings, Mpumalanga, South Africa
The defunct Union Colliery located in the Vryheid Coalfields of South Africa is decanting excess mine water after decades of mining activities without any measures to prevent or control acid generation. Mine waters further indicate elevated concentrations of metal(loid)s. The mine has however made attempts to investigate a small pilot treatment system having a barium carbonate [BaCO3] dispersed alkaline substrate for use in the at least partial remediation of mine drainage. This dissertation investigates potential geochemical risks from the geological mine waste resulting from underground coal mining with the aim to quantify these risks and evaluate long-term mine waste drainage stability. The evolution in the geochemistry of mine impacted waters as it evolves from the primary source minerals until final discharge is subject to site specific influences and necessitates site specific evaluation. A multidisciplinary, integrated assessment approach was implemented that included hydrogeochemical, mineralogical and geochemical assessment, and finally geochemical modelling techniques to predict the quality of water emanating from the mine waste over the long-term. Various data sets and sources were compiled and processed and served as input into geochemical models of the expected impacts on the conceptual mine waste environments. Forward reaction path models allowed the evaluation of the current system with thermodynamic and chemical reaction processes leading to the hydro-chemical composition of the groundwater and mine decant water as observed in the monitoring data sets. A review of the assessments that were undertaken to determine the long-term water quality risks after mine closure and comments of potential water management strategies thereon are presented with special focus on quantification of risks by means of forward reaction path models. XRD analysis revealed that the mine waste comprises of varying proportions of primary, secondary and clay minerals with calcite as the dominant carbonate phase and pyrite the dominant sulphide bearing mineral. ABA assessments indicated that the coal discard material have high likelihood to produce acid, whereas the hanging wall coal sample collected from the Susanna compartment suggested neutral conditions. It is however important to contextualise the ABA results with the hydrogeochemical conditions of the study area and existing water monitoring data. For hydro-chemical assessment and modelling purposes the area was divided into different hydrogeochemical components, each representing different systems exposed to dynamic processes. From the water quality monitoring data sets, baseline hydrogeochemical conditions for the study area, including the different hydrogeochemical components, could be inferred. Based on the hydro-chemical plot assessment, sodium was identified as the dominant cation in the groundwater, whereas calcium dominates in the mining-influenced waters (MIW). Bicarbonate is a significant anion along with sulphate in the groundwater, surface water and eastern operation components; whereas sulphate is the dominant anion in the western operations. The ion distributions in the different hydrogeochemical components are expected considering the mineralogy of the mine waste material. Processes such as sulphide mineral oxidation and carbonate dissolution as an attempt to neutralise any acidic environments formed, are responsible for the elevated sulphate, bicarbonate and calcium concentrations. As expected in a mining affected area where sulphidic waste was produced, there is a distinct trend in all areas, except for groundwater in the unmined land, to be or to become sulphate dominated. A bimodal pH distribution was observed between the hydrogeochemical components which is a phenomenon observed in many mining influenced areas. The western operations decant water are distinctly acidic, whereas the other hydrogeochemical areas indicate circum-neutral to alkaline conditions. Based on the site characterisation and hydrogeochemical and geochemical assessments, representative conceptual models were developed as a foundation for the kinetic geochemical models. The main purpose of the geochemical models was to reduce uncertainty through constant sensitivity analyses, and to facilitate understanding of the interaction of geochemical processes with a focus on metal(loid) mobility to a level at which potential liability can be quantified. The weathering models simulated for the coal discard dumps and eastern operation components suggested that circum-neutral to alkaline pH will prevail in the long-term which may be attributed to weathering reactions of the carbonate and mafic minerals within the local geology that does not allow acidic environments to develop. Manganese and selenium remain in solution after the simulation period and therefore create several risks. Selenium specifically may reach concentrations in the groundwater that may pose high level risk to humans. Therefore, although alkaline conditions prevail in the groundwater and MIW, selenium may remain mobile in the long-term. The western operation weathering models indicated acidic conditions throughout the simulation period which may be attributed to the depletion of carbonate minerals along flow paths. Associated with the acidic nature of the decant water are elevated concentrations of manganese, aluminium, lead, nickel and selenium that may reach concentrations that pose high level risk to several aquatic tropic levels as well as humans. Main risks associated with the mine waste included contamination of shallow groundwater resources as well as ecological impacts and regional impacts of major river systems associated with decant of MIW at the surface. The model findings suggest that risks will remain in the long-term and it therefore invokes certain mitigation implications. The study highlighted the importance of representative and complete data sets of especially the mineralogy of the mine waste which is used as vital inputs to the geochemical models. As the study lacked mineralogical input of especially the underground floor and roof lithologies, it is recommended for the mine that the study be furthered by extensive solid waste sampling of the underground waste material. This will lead to better calibration, and hence higher confidence outcomes of the current geochemical models. Conventional active and passive treatment methods may be successful to reduce current risks, especially of that posed by the acidity in the western operations, however, as these methods may not be feasible in the long-term (>100 years) due to waste generated by lime dosing or desalination plants, the evaluation of in situ methods such as bioremediation is recommended.