ABSTRACT
Acid mine drainage (AMD) and associated metal(loid) and SO42- pollution of soil, surface water and groundwater is ubiquitously associated with tailings material generated by Au mining in the Witwatersrand Basin in South Africa. The individual geochemical processes responsible for the AMD generation in this tailings material are relatively well understood. What is less clear are how these different processes interact as a network within the tailings system. Process network modelling (PNM) is a tool that can be used to study such interactive and complex networks of geochemical processes, especially when stochastic methods, e.g. Monte Carlo simulation, are included in the model development. Secondary mineral phase supersaturation requirements from classical nucleation theory are also built into the model.
A PNM was developed for a tailings facility in the Witwatersrand gold basin focussing on pH, Fe(total) and SO42- concentrations in the tailings pore water and the relationship of these parameters to the dissolution of pyrite, O2 diffusion into the tailings, oxidation of Fe2+ and the precipitation of secondary minerals, specifically goethite and jarosite. The model indicated that AMD conditions develop fairly rapidly after the sulphidic material is exposed to the Earth’s oxygenated atmosphere. The concentration of H+, and hence the pH, in the tailings pore water is controlled by a number of feedbacks. The positive feedback, implying addition of H+, is the dissolution of pyrite and the precipitation of the secondary Fe3+-bearing minerals goethite and jarosite. Jarosite precipitation was shown to increase the median H+ addition rate by ~2%. The negative feedback, i.e. decrease in H+, is the oxidation of Fe2+ to Fe3+. This feedback loop produces a net excess of H+. Together with the buffer effect of goethite and jarosite precipitation, system steady-state conditions are eventually achieved with respect to pH.
The pore water SO42- concentration is controlled by the positive and negative feedback of pyrite dissolution and jarosite precipitation. This feedback loop produces a large excess of SO42- and steady state conditions can only be achieved if SO42- is physically removed from the tailings system, e.g. seepage to groundwater.
Oxidation of the Fe2+ produced by pyrite dissolution to Fe3+ is the only positive feedback for tailings pore water Fe3+ concentrations. The negative feedbacks are precipitation of goethite and jarosite and the oxidation of pyrite by Fe3+. The former effect is delayed as these phases first have to achieve a certain level of supersaturation in the tailings pore water solution before they can form. The precipitation of jarosite and goethite, by removing Fe3+ from solution, decreases the effect of Fe3+ pyrite oxidation causing O2 to remain the dominant oxidation mechanism. This feedback loop produces a small excess of Fe3+ over time, however, the model is very sensitive to other factors, e.g. O2 diffusion deeper into the tailings facility.