A method for rapid estimation of processing behaviour based on ore texture

Abstract

Abstract A method for rapid evaluation of processing behaviour based on assessment of the heterogeneity of ore texture extracted from drill core images is described in this paper. The method involves three stages - drill core scanning in a hyperspectral scanning device to generate high resolution core images, image processing using texture spectrum analysis with cluster analysis to extract and classify ore texture, and modelling to correlate texture heterogeneity with processing behaviour. A case study performed on selected drill cores from a porphyry copper deposit demonstrates the application of the method in predicting the breakage behaviour of the ore by estimating the comminution parameter A*b for the entire set of drill cores from a subset of measured data. The results indicated a high degree of variability in breakage along the drill core, and that the Very Hard group (with an average A*b of 27) is dominated by the more heterogeneous texture while the Very Soft group (average A*b of 196) is dominated by the more homogeneous texture. The case study demonstrated the value of the method in providing a rapid and cost-effective approach to prediction of processing behaviour, which leverages high resolution textural data measured at core-scale to extract information which is relevant to processing, with a reduced need for physical test work. The method also allows a semi-automated and rapid evaluation of the processing variability of the ores from drill core samples using images which are being routinely collected on drill core. This improves geometallurgical characterisation by increasing the density of data across a deposit and providing a better understanding of the processing performance variability within the orebody.