A Process Mineralogical Study on the effect of Alteration on the Flotation of Great Dyke Platinum Group Element (PGE) Ores

Ores from the same deposit may exhibit extensive variability in their mineralogy and texture. The ability to quantify this variability linked to metallurgical performance is one of the primary goals of process mineralogy and geometallurgy. Ultimately this information can be used to inform decisions around all core activities of mining and processing. This study focusses on identifying the key mineralogical differences between three Great Dyke platinum group element (PGE) ores in Zimbabwe. These ores are known to be characterized by extensive oxidation and alteration resulting in numerous metallurgical challenges in recovering the PGE. The behaviour of three different ores sampled along the strike of the Great Dyke is compared, focusing on mineralogical composition, rheological characteristics and batch flotation performance. The contribution of the differences in mineralogy (bulk mineralogy, base metal sulfide (BMS) liberation and association, and naturally floating gangue) to processing challenges and potential opportunities to manage these was considered. It was noted that slight differences in mineralogy, particularly BMS liberation and association, yielded notable differences in copper, nickel, platinum and palladium recoveries. The most oxidized ore was found to have lower recoveries due to the oxidation of the BMS, though a deeper understanding of the oxidation and flotation behaviour of PGEs (and platinum group minerals - PGMs) is still necessary. Through the mineralogical analysis of the batch flotation concentrates it was observed that more finely disseminated and yet locked (unliberated) talc resulted in higher amounts of naturally floating gangue (NFG). The effect of 3 polymeric carboxymethyl cellulose (CMC) depressants, differing in degree of substitution, was also evaluated in terms of their ability to depress the naturally floating gangue and mitigate any rheological complexities that may be associated with these ores, through the electrostatic repulsion of the negatively charged carboxylate groups. There was no significant advantage of one depressant over the others in the batch flotation tests or in the rheology tests. The more oxidized ore was found to contain relatively low amounts of phyllosilicate minerals and, therefore, no rheological problem was present that would have required a chemical solution. There were no noticeable differences in the rheology of the slurries of the 3 ores. This was likely to be due to the dampening or buffering effect of the high proportion of minerals that do not contribute to rheological complexity. Changing of depressant type also had no effect in this case possibly due to the same reasons. In addition to this, the region after which the rheological complexity of all 3 ore types begins to increase exponentially is from 30-35 vol.% solids concentration (60-65 wt.% for an ore with a specific gravity of 3.3). It is therefore advisable for Great Dyke operations not to exceed these solids concentrations as this would exacerbate the processing challenges associated with rheological complexity. The use of such solids concentrations during flotation is however unlikely though this may be the case in other parts of the processing circuit, e.g. comminution, and should thus be noted. The decoupling of the terms referring to alteration (that is oxidation and hydrolysis/hydration) is also presented in this study together with the effects of these different types of alteration on the processing of PGE ores. Oxidation affects the valuable minerals and thus flotation recoveries whilst hydrolysis/hydration acts on the gangue minerals and therefore mainly affects concentrate grade. The more oxidized ore sample in this study had undergone the oxidation type of alteration, rather than hydrolysis/hydration and the processing challenge associated with it lies not in the gangue but with the valuable minerals. Finally, it was shown that investigating an ore’s characteristics solely on mineralogy may not necessarily give a full prediction of the ore’s response but the linking of the mineralogical characterization with metallurgical test work gives a more holistic view