BENCH-SCALE PARTICLE SORTING TESTS: BEWARE OF THINGS UNKNOWN

摘要

Amenability of crushed material to particle sorting is typically assessed using a relatively limited quantity of material. As a consequence, the question of sample representativeness (especially in the context of mining operations) introduces a degree of uncertainty and, thus, leaves some ambiguity in interpretation of particle sorting amenability test results. In this study, a link between intrinsic material heterogeneity and the sorting potential is explored using computer simulations. The results indicate that statistical uncertainties for a relatively small sample size (ca. 100 rocks) showing a high amenability to particle sorting (i.e. high recovery into a low mass pull) can be significant. Under tested conditions, absolute deviations of -15.3/+10.7% in recovery and -12/+11.9% in mass pull were obtained for a particulate material with a high heterogeneity using a 95% confidence level relative to the expected values. It is argued that the number of particles used for sorting amenability tests should be based on the Theory of Sampling considerations. In particular, constitutional heterogeneity parameter of the batch of particles used during testing should be reported along with the estimated number of particles required to achieve a certain Fundamental Sampling Error threshold (e.g. 10 or 15 %).