Characterisation of geotechnical, geochemical and metallurgical properties for mine-to-mill optimisation at Sandsloot open pit platinum mine, South Africa

Abstract

Sandsloot open pit mine extracts platinum, copper and nickel from the Platreef orebody in the Northern Limb of the Bushveld Complex. At present, it is the world's largest open pit platinum mine, but it is one of the smaller of six pits to be mined in this area by Potgietersrust Platinums Limited (PPRust). As mining progresses and the mine operations expand, sound knowledge of the subsurface ground, the orebody properties and their performance in the processing plant is imperative. An accurately characterised orebody, in terms of its geotechnical, geochemical and metallurgical properties, will facilitate the development of appropriate and cost effective mining practices and processing plant design. It will improve the prediction and performance of materials in each of the steps involved in converting the orebody rock mass into a concentrate /marketable product, which ultimately leads to mine-to-mill optimisation. This dissertation has focused on the geotechnical and metallurgical strength properties of the Platreef orebody at Sandsloot; the geochemical properties of the various lithologies and defined for each orebody rock type the associations and correlations within and between these properties. The petrography of the orebody material was studied to identify and classify the rock types. A rigorous testing programme was conducted and an extensive database of petrographic, geotechnical, geochemical and metallurgical information was assembled. The corresponding results within and across the different studies were grouped together and statistical analysis and interrogation of the data sets were performed. The research identified diverse rock types with contrasting physical properties in the Sandsloot open pit orebody. For each of the rock types the distinguishing characteristics and the variations in properties were identified. Significant differences and relationships between rock types for each property were assessed for predictability in mining. Statistical associations and correlations between the properties of each study were defined and models for predicting strength and rock type were developed.