Study of Flotation Process Variables Combining Metallurgical and Pentlandite Liberation Data

Size by size recoveries and pentlandite liberation data from image analysis were used to evaluate plant tests of pentlandite flotation and pentlandite/pyrrhotite middlings flotation from pyrrhotite in a circuit treating low grade magnetic intermediates at the Strathcona Mill. Raising the pH with lime gave improved selectivity for liberated pentlandite but there were major losses of middlings unless the circuit was pulled hard indicating the need for finer regrinding. However, size by size recoveries at the normal regrind showed that, while the pentlandite recovery improved in the finer ranges, the pyrrhotite recovery also increased leading to little selectivity in any size range. A much finer regrind gave greatly improved selectivity in all size classes suggesting that 'surface polishing' as well as liberation effects were being seen. Using liberation data from image analysis of the products, surface polishing accounted for over 99 % of the improvement in selectivity. Unreasonable recoveries of more highly liberated particles were observed for liberation analysis by volume fraction but showed the expected trend in recovery when the analysis was based on surface liberation. The surface liberation data showed that the improved pentlandite recoveries after finer regrinding were to more than 90 % caused by improved liberation with the remainder contributed by surface polishing effects. The practical implications of this work are that i) the comparison of particle grades with product grades allows to estimate the effect of regrinding on co-flotation/entrainment, ii) surface liberation models should be used for evaluating recoverabilities by flotation and iii) regrind circuits may be set up to ensure that all particles go through the mill or that only the coarse fraction is reground depending on the surface scrubbing found to be required.