Iron Ore Coarse Particle Characterisation: Towards Prediction of Particle Distribution in Gravity Separation Processing

Abstract

– The Limpopo and Northern Cape provinces of South Africa host hematitic iron ore deposits that, geologically, form part of the Transvaal Supergroup. Due to various geological processes that took place during the formation of the ore, textures are variable, and may be qualitatively described as massive, laminated, conglomeratic, brecciated, etc. These textures affect the separation efficiency during processing to upgrade low-grade ore by gravity separation. Mineralogy plays a crucial role during beneficiation; the obtained particle mineralogy can be linked to density classes to predict particle distribution during processing. Measures can thus be taken to improve the separation efficiency. Commonly used mineralogical techniques like automated scanning electron microscopy (AutoSEM) and optical microscopy, however, are not well-suited for coarse particle characterisation. For this study, therefore, the emerging technique, micro-X-ray fluorescence (micro-XRF) imaging, was investigated to produce elemental maps for texture characterisation on coarse particles (>6mm) of an Fe ore sample from Limpopo, together with X-ray diffraction (XRD) to characterise the coarse particle samples. The results show that the ore contains massive hematite as well as laminated hematite-quartz particles. These preliminary results predict that, for sink-float separation tests, massive hematite particles will be recovered at high density, but laminated hematite-gangue particles will be lost to the floats at different density classes, dependent on the ratio of hematite:gangue in the particles. Quantification of these effects is the next step in the study, towards establishing a predictive method for coarse particle distribution in gravity separation of Fe ore.