Coal Preparation » Fine Coal
In coal flotation, particles of different components of the coal such as maceral groups and mineral matter and their associations have differing hydrophobic characteristics and therefore different flotation responses. By using a new coal grain analysis method for characterising individual grains, more detailed flotation performance analysis and modelling approaches have been developed. The method involves the use of microscopic imaging techniques to obtain estimates of size, compositional and density information on individual grains of fine coal. The density and composition partitioning of coal processed through different flotation systems provides an avenue to pinpoint the actual cause of poor process performance so that corrective action may be initiated. Samples obtained from audits of coal flotation circuits of three coal preparation plants were studied. The study showed that pure grains of coarse inertinite and composite particles containing inertinite had lower flotation responses than pure grains of vitrinite and vitrinite-rich grains. In one circuit processing material with a nominal top size of 0.500 mm and a significant proportion of coarser particles, considerable loss of combustibles, mainly coarse inertinite and composite particles was observed. For another flotation circuit treating a feed with a top-size of 0.250 mm in mechanical cells most of the combustibles of all grain classes were recovered. This has provided a more advanced diagnostic capability for fine coal cleaning circuits than was previously possible. This approach enabled flotation performance curves analogous to partition curves for density separators to be produced for flotation devices.
The information on grain size, density and composition has been used as input data to develop more detailed flotation process models to provide better predictions of process performance for both mechanical and column flotation devices.
This study also attempts to develop flotation models that respond to changes in the size and composition distributions of the feed particles. A particle size and composition dependent kinetic flotation model is used to generate data for varying feed particle composition, density and size distribution. The model is assessed for its ability to predict the effects due to changes in particle properties. Distributed flotation rate constants are used for fitting the effects of particle heterogeneity, accounting explicitly for particle composition. This has essentially allowed the concepts used in metalliferous mineral flotation with a focus on the flotation of specific mineral entities to be applied to coal flotation. This was not previously available. The model predictions have been compared with laboratory experimental data. The calculated component recoveries and rate constants compared well with the experimental data. Moreover, the model predictions were able to replicate the characteristic maximum in rate constant and recovery for the intermediate size classes as well as mimic previous experimental observation that coarse inertinite and inertinite-rich composite grains have lower rate constants than vitrinite and vitrinite-rich composite grains. A great need identified in this respect is a reliable component flotability estimation which was partially touched on but requires further work.
The coal grain analysis methodology has been used to generate washability information and the results compared with float-sink analysis data. The results showed very good agreement with the float sink results, which suggests that provided the maceral and mineral densities are known and that the maceral and mineral proportions are accurately measured in each grain then coal grain analysis provides a realistic alternative for determining the washability characteristics of fine coal compared to that achieved by using organic liquids. To date testing has been restricted to samples with a topsize of 1mm, and further testing is required to verify that the method can obtain reliable washability information on size fractions up to a topsize of approximately 4mm. The method was also able to provide washability information on ultra fine (eg -63 micron) samples that are difficult to analyse with organic liquids.
An experimentally determined relationship between %ash and % minerals allows the generation of estimated yield-ash curves which are useful in determining an estimated yield at a specified product ash for samples of fine coal.