Principals of coal selection for Advanced Beneficiation

摘要

After mining and crushing some particles which report to the coal benefication plant can consist of essentially pure components of a single maceral or mineral phase, whilst others are composite particles that are comprised of varying amounts of macerals and minerals. The proportion of particles that are present as pure components or as composites will be a function of the characteristics of the coal and the particle size. In general, it is considered that size reduction will result in liberation and hence increased yield. The amount of liberation that occurs during crushing or grinding a coal is however coal specific. Hence, to optimise coal recovery during benefication it is necessary to understand the associations of the minerals and macerals in individual particles as these give a particular coal its washability (density distribution) and contribute to its surface chemistry characteristics. Over the past six years, CSIRO has used optical reflected light imaging and Coal Grain Analysis (CGA) to obtain and use compositional information on individual coal grains for liberation and flotation optimisation studies. The CGA system collects contiguous 14 bit colour images and mosaics them together to obtain reflectance and composition information on individual particles. Sufficient images are collected so that micron level detail is obtained on the mineral and maceral inclusions for a statistically relevant number of particles. Size and compositional information (the amount of vitrinite, inertinite, liptinite and mineral) is obtained for each particle and this information is used to estimate the density, ash value and hydrophobicity of each particle. This enables the washability and flotation attributes of the sample to be estimated. The particles are then classified as pure components (＞95% of a single phase) or as composites and the mass abundance, and if required the washability and hydrophobicity of each grain type, can then be determined. CGA analyses are generally conducted on a small representative subsample of -lmm material. However, it has been successfully used to characterise particles up to 4 mm in size, which report to the middlings circuit, and samples which have been crushed to less than 20 microns, for assistance in beneficiation of coals to produce low ash value products for uses such as in direct injection coal engines (DICE) or for use in carbon fuel cells. For advanced benefication applications, CGA enables liberation to be assessed and expected yield at different target ash% to be estimated. The CGA information also provides detail on the intrinsic and entrained minerals in a sample and hence benchmarks the lowest ash value which could be obtained if ail entrained minerals are removed from the sample at that topsize.