TESTING ALTERNATIVE TECHNOLOGIES TO REPLACE CURRENT BELT SAMPLING METHODS

Current manual face sampling methods are unreliable, and it takes time to obtain results. Manual face sampling is also labour intensive and exposes people to unnecessary risks while they are sampling on the mining fronts. The AOK programme’s focus in this area is to investigate alternative sampling techniques, which could be potentially automated. This will ensure that quality information can be obtained timeously and safely, which will enhance the accuracy of analysis and planning.

The research team investigated and tested the automated mineralogy (AM) to establish if it can supplement grade assays and provide orebody characterisation. The team determined the element grades for assay laboratories. The next step was to determine the actual extractable grade based on element deportment. The AM data was used in each stage of the mining cycle, from greenfields to tailings. The research team conducted a baseline assessment of the laser-induced breakdown spectroscopy (LIBS) by testing gold and platinum samples and having them analysed, using LIBS scanning technology.

Selected results from LIBS analyses on a gold sample of 100 g/t indicated that as the particle size is decreased, fewer repeats are required to achieve a given level of accuracy. For a 100 g/t sampled on a 1000×1000 grid, the mean and standard deviation converge after about 60 measurements. For a 1000×1000 LIBS dots model, four different gold grades, the variance decreases as the grade increases.

This project was commissioned to the CSIR, University of Johannesburg and the University of the Witwatersrand. Researchers involved include CSIR principal geophysicist, Michael Van Schoor, CSIR candidate researcher, Zamaswazi Nkosi, University of Johannesburg mineralogist, Bradley Guy, University of Johannesburg geologist, Dr George Henry, and University of the Witwatersrand Director for the Centre for Mechanised Mining Systems, Dr Declan Vogt.