Investigations into the prediction and modelling of total sulphur in coal seams, German creek mines, Central Queensland

Full text: The German Creek Formation contains five laterally persistent coal seam intervals known as the German Creek, Corvus, Tieri, Aquila and Pleiades in ascending order. The principal seam is the German Creek seam but the Aquila seam and, to a lesser extent, the Tieri seam have also been the focus of past open cut mining operations. The coal measures dip generally eastwards and the German Creek seam reaches depths of 600m to 800m at the eastern lease boundary. The majority of German Creek's production is from the German Creek seam, an ortho-bituminous, low ash, medium rank, hard coking coal, primarily for export. Most of the shallow coal in the German Creek sequence has been mined except in the north of the leases where thinner, poorer quality seams remain. Present day production is from three separate mines, German Creek Open-cut, Central and Southern Collieries. There is a systematic variation in rank across the German Creek mining lease, which is exemplified by a regular decrease in volatile matter in the coal from southwest to northeast. Because many of the coal's coking properties are governed by rank, there is a long-range reduction of coking parameters, such as fluidity and dilatation, towards the northeast. By comparison, the sulphur content of some coal seams can, on occasions, widely vary in both a vertical and lateral sense within and between seams, particularly near dykes and seam wash-out areas. This short-range variability creates problems in predicting the sulphur content of washed product coal, with the possibility of coal shipments to customers being rejected if sulphur levels exceed pre-defined limits. Therefore, if patterns in this variability could be determined and related to some geological property or process of formation of the coal seams, then areas of high sulphur might be predicted in advance of mining. The author instigated a background coal characterisation study, as it was seen as an integral part of ongoing investigations aimed at determining lithological and structural controls to the sulphur distribution in the coal. To this end, a representative selection of exploration slimcore and face channel samples from all production areas were subjected to coal petrographic, coal and parting mineral matter by XRD and or SEM using SIROQUANT techniques, and coal ash trace element analyses by ICP, AAS or XRF techniques. Additional information was gained by comparing quantitative XRD analyses (SIROQUANTTM) with mineral matter determinations obtained using SEDNORM. Various modelling methodologies were employed, ranging from geostatistics (ordinary kriging) to polygons of influence so as to best represent the short-range variations present in the seam data sets. Though this proved moderately satisfactory for raw in situ data, the influence of float sink washability testing was also examined, as it was found that borecore size influences the clean total sulphur result. Discrepancies between raw and clean sulphur values from the models so constructed still need to be reconciled. In non-cored boreholes it was hoped that some sensible use could be made of available downhole geophysical logging. Whilst providing a wealth of information on coal quality and thickness, data on sulphur distribution from these typical log responses was found to be limited. Trials will soon be conducted using CSIRO's SulphaLOG downhole geophysical logging sonde. This logging technology employs the prompt neutron-gamma method (neutron-capture), for the determination of sulphur and other elements in coal seams and partings. The characterisation study information from exploration and channel sampling enabled significant updating of the existing quality database to occur, which in turn allowed rigorous statistical analyses of available values (by seam) to be conducted. These analyses are essential building blocks to allow areas of high sulphur to be predicted in advance of mining. Eventually, a methodology for predicting sulphur mineral concentrations in advance of mining and reconciling 'as-sold' coal quality to that predicted will be developed. These prediction theories will be assessed at the coalface as mining progresses through high-sulphur zones. Copyright (1999) Geological Society of Australia