Towards Developing a Viable Circuit for Beneficiating Ultrafine Chrome and PGM Tailings to Recover Chromite

Abstract

Tharisa minerals is one of the operations that generates ultrafine tailings from its Platinum Group Metals (PGM) and chrome spiral recovery circuits that treat Middle Group (MG) chromitites package in the Western limb of the Bushveld Igneous Complex (BIC). The tailings generated by the process at Tharisa had a grind of 80% passing 75μm, of which about 50% was below 25μm. This work therefore aims at investigating the potential to recover chrome from such tailings using different approach. The mineralogical examination performed by TESCAN Integrated Mineral Analyzer (TIMA) an equivalent of Quantitative Evaluation of Minerals by Scanning Electron Microscopy (QEMSCAN), electron probe microanalysis, X-ray diffraction as well as wet chemical assay analysis by Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES) revealed that the bulk of the mineral in this feed material contained 9% Cr2O3, and the rest of gangue material being SiO2, Al2O3, MgO and CaO. In terms of mineral makeup, the feed material comprised of feldspar accounting for 41.7% of the mass followed by Enstatite-ferrosilite accounting for 29.2% of the mass. Chromite contributed 18% by mass and the balance were minor oxides. The chromite mass percentage increased as the particle size decreased from 106μm to 25μm indicating why conventional gravity equipment would not be effective in recovering this chrome. The degree of liberation of the particles in the feed was found to be 90% on average and the liberation increased as the particle size decreased. In terms of chrome deportment, it was found that chromite mineral is the only source of Cr whereas Fe which is an element making chromite with Cr was contributed by enstatite- ferrosilite, chromite and amphibole minerals. Bulk samples were used to investigate the effect of process variables on separation capabilities of three unit of separation specifically chosen to recover ultrafine chrome. These units are: (i) a wide radius shallow trough angle M1T53 spiral originally designed for ultrafine feed like coal, (ii) an L300 Wet High Intensity Magnetic Separator (WHIMS) belt magnet of 1.5mx0.3m dimensions fitted with a 1.3mm thick rubber belt rotating over a 9000 Gauss magnet cassette and (iii) a Holman-Wilfley Model 800 laboratory shaking table capable of handling ultrafine feed. Results from single units’ separation tests showed that for optimum recovery and grade attainment, the spiral requires a slurry of density 1.52t/m3, feed rate of 0.83t/h per start and concentrate cutter position set at 80 mm from the centre pole. The most significant factor among the three factors was concentrate cutter position followed by feed rate. For magnetic iv separation, the optimum slurry density was 1.38t/m3 feeding at 0.81 t/h at a belt speed of 0.5m/sec and fluidisation water of 74.4ml/sec. Of these factors, the most significant one was feed-rate followed by fluidisation water. For shaking table separation, the deck angle was the most significant factor, and the operating level was 4o. It was also found that wash water was important in terms of its effect on tailings grade. The optimum slurry density for the shaking table was 1.34t/m3, feed rate was 500ml/min (0.01428t/h) and wash-water was 140ml/sec per deck. For full circuit design, results from single unit tests placed the spiral as the first or rougher unit for recovery purposes. The magnet was suitable for lower grade streams hence it was placed as a scavenger of spiral middlings and tailings. The shaking table had the best upgrade as well as highest grades but was poor in mass yield. For these reasons, the shaking table was placed as the last or cleaner unit of the circuit to treat smaller high-grade streams and produce final grade. From this, two complete circuit configurations were set up to compare their effectiveness. The first configuration consisted of the spiral as the first unit producing concentrate of grade 14.41% Cr2O3 from a feed of 10.68% Cr2O3 with the middlings and tailings assaying 8.97% and 9.31% Cr2O3 respectively. The middlings and tailings were processed on the magnet and the magnetic fraction sent to a separate shaking table. The combined final concentrate from separate shaking tables (configuration 1) assayed 40.59%Cr2O3 at an overall yield and recovery of 5.81% and 19.01%, respectively. The second configuration consisted of only two concentration stages, i.e. the spiral separation and shaking table separation of the two size fractions, -25μm and +25μm. These two streams were fed to the spirals separately and the spiral concentrates fed to the shaking table separately. The fine spiral concentrate was processed on the shaking table at lower slurry density of 1.34t/m3 and lower feed rate of 0.150t/h compared to single units shaking table results. The combined final concentrate from the +25 μm and -25 μm circuits assayed 40.07% Cr2O3 at a yield and recovery of 7.00% and 29.28%, respectively. The better performance of circuit configuration 2 is due to the treatment of narrow size fractions separately a requirement for most gravity separation processes. Configuration 2 was chosen as the better circuit for this investigation. The costing of configuration 2 showed that the capital cost would be R718,376,073 with an operating cost of R128,716,294 per annum and annual production of metallurgical chrome of 259,512 tons, the net present value of the project is R2,550.5 million and a payback period of 11 months. From v a technical and economic point of view the chosen circuit is viable and it is a terrific opportunity for the operation that is currently losing this ultrafine material to the Tailings Storage Facility.