Coal Preparation » Fine Coal
The aim of this project was to adapt existing particle based flotation circuit simulation architecture, developed by the Julius Kruttschnitt Mineral Research Centre for metalliferous flotation circuits, as a tool that can be applied in fine coal flotation to forecast the likely impact changes in the feed particle characteristics will have on the yield and quality of the flotation product. This architecture uses data generated from the Coal Grain Analysis (CGA) methodology.
The ability to quickly and adequately respond to changing plant feed conditions remains a key component for improving the overall coal recovery at a specified product quality. In the flotation of fine coal, the surface chemistry of the particles is important in establishing the proportion of flotation recoverable coal.
CGA images obtained from full scale flotation circuits under projects C24045 and C25019 were re-analysed to quantify the particle perimeter composition, with the aim that this information would be correlated to the measured flotation rates of different particle grain types (archetypes).
The particle based architecture was applied to allow the distribution of different particle archetypes, per size class, to be tracked throughout a circuit during simulations.
The benefit of this approach to the industry is that it allows coal flotation losses to be tracked at the particle level and provides a means by which the inter-dependencies of coal preparation-related activities may be integrated in the future to optimise value in an all-inclusive manner. The tool will enable coal industry participants to make greater use of CGA data to monitor and quantify the influence of sub-optimal feed conditions. It also provides another means by which benefits from sensor data can be transferred to and accepted by the coal industry.
This research has shown that the particle surface composition approach, which was developed in base metal flotation can be applied to data obtained from CGA to simulate coal flotation. The CGA measurements can be processed to provide the required particle composition data (both surface and bulk).
The results for the project were mixed. For four sites, the correlation between simulated full-scale, and actual full scale performance were excellent. For three sites, there were substantial differences between the measured flotation behaviour and the results of the simulations in terms of both yield and composition of the concentrate. For the remaining two sites, concentrate composition was well predicted, but the simulated yields did not agree with the estimate of the 'measured' plant yield. The reasons for these mixed results are detailed within the report.