A geometallurgical study of the quartz dominant ore varieties in a polymetallic base metal sulphide deposit, Northern Cape, South Africa
Thesis (MSc)--Stellenbosch University, 2020.
ENGLISH ABSTRACT: Geologically complex deposits require integrated approaches to characterisation, extraction and processing of ores in order to maximise the yield of the deposit. In this study, a complex polymetallic (Cu-Pb-Zn-Ag) base metal sulphide deposit was examined to evaluate whether the different ore bodies should be grouped into geometallurgical domains. Geometallurgy is the practice of incorporating the entire value chain, and in this case, means the incorporation of geological knowledge of a deposit with metallurgical/process mineralogical knowledge to create a predictive, all-encompassing, model of a deposit. This idea was investigated using a Cu-Pb-Zn deposit in the Northern Cape of South Africa. The deposit consists of three ore bodies, namely: the Upper Ore Body (UOB), the Garnet Quartzite Ore Body (GQOB) and the Lower Ore Body (LOB). The UOB consists of magnetite-dominated ores and the other two ore bodies are quartz-dominant ores. The two quartz-dominant ore bodies consist of three different ore types: (1) garnet quartzite (GQOB), (2) mineralised schist (LOB) and (3) sulphidic quartzite (LOB) (abbreviated as ores G, H and I respectively). In this study, an in-depth investigation into the feed mineralogy/morphology, minerals processing responses and flotation concentrate mineralogy/morphology was conducted to test the legitimacy of creating one quartz-dominant geometallurgical domain. Mineralogical analysis of the feed was performed using QEMSCAN on samples milled to achieve a P80 of 65% passing 75 μm. Mineralogical analysis of the feed showed that the three ores were not as similar as originally proposed. Ore G had a unique feed mineralogy compared to ores H and I (which were similar to each other). Ore G consists of high (5.3 wt. %) amounts of chalcopyrite compared to that of ores H and I (1.1 wt. % and 1 wt. % respectively). In contrast ores, H and I consist of far higher (9.9 wt. % and 10.1 wt. % respectively) modal amounts of galena compared to the low amounts found in ore G (0.7 wt. %). Together with the differences in modal abundances of economic sulphides (ES) in the three ores, the two ore bodies had distinct sulphide gangue (SG) and non-sulphide gangue (NSG) populations. The NSG component of the three ores was, as expected, dominated by quartz, however, ore G has high amounts of garnet (10.2 wt. %) and magnetite (23 wt. %), whereas ores H and I has high amounts of mica (8.5 wt. % and 12.8 wt. %, respectively) and barite (8.7 wt. % in ore I). The mineral liberations and associations of ore G is also unique from ores H and I. ore type, similar total solids vs water recoveries were obtained for each ore type. Further investigation into the flotation performance of the three ores was done by analysing the flotation concentrates using ICP-OES to create elemental grade vs. recovery and metal mass vs water recovery profiles. Through the analysis of these results, it was determined that the ores again showed a similar grouping to the feed mineralogy (GQOB ores distinct from LOB ores). The analyses of the flotation concentrate showed that the target elements in the quartz-dominant ores (Cu and Pb) were being recovered as expected (in chalcopyrite and galena, respectively). The high iron recovery in the flotation concentrate (in particular of ore G) was a result of true flotation concentration of pyrite. This concentrate analysis once again showed that the three quartz-dominant ores were not as similar as originally thought. Ore G had a far higher pyrite content (56.7 wt. %) than ores H and I (29.8 wt. % and 20.7 wt. %, respectively) which were similar. As expected, ore G had a high chalcopyrite content (30.7 wt. %) in the concentrate, whereas ores H and I had a high galena content (57.4 wt. % and 67.3 wt. %) in the concentrate. Based on the above information, it is proposed to make two geometallurgical domains from the three quartz-dominant ores, one made up solely of ore G and the other made up of the combination of ores H and I. Implementation of geometallurgical domaining of this Cu-Pb-Zn deposit would require further analysis of the feasibility of mining and processing ores in such a way to preserve the ore domaining throughout the mining chain. On existing mine sites this would likely prove challenging but the results of this study suggest that such an approach could improve the financial return on the mining activities. In the case of this deposit, the ore domaining proposed for the quartz-dominant ores needs to take into consideration the UOB which is magnetite-dominant. Should the difference between the ores of the UOB and the quartz-dominant ores be more significant than the difference between the two proposed quartz-dominant domains, then the subdivision of the quartz-dominant ores might not make economic sense. Nevertheless, this study proposes that the quartz-dominant ores should be split into two domains aimed at grouping together ore types that exhibit similar processing responses.
AFRIKAANSE OPSOMMING: Geen opsomming