Technical Market Support » General
The project’s objective was to develop a reliable handling index that could be used to predict whether handling issues were likely for different coals. A comprehensive suite of thirty-three coals (coking, thermal and PCI coals) from seventeen NSW and QLD coalmines was assessed for the routinely determined coal properties of moisture content, size distribution, ash abundance and ash composition. Thirteen of the coals were also assessed for mineral composition. A soil chemistry test that had been previously used with some success to differentiate high clay content of power station feed coals was investigated for its suitability for determining clay contents of product coals. Each coal was also tested with the CSIRO Handleability test at its as received moisture content. In all up to seven tests were conducted for each coal at consolidating pressures from 1.6 to 10.0 t/m2. Previous work with this apparatus had determined that coals which were free flowing after the application of a consolidating pressure of 7.6 t/m2 did not cause problems when unloading at Queensland coal ports.
The results obtained with the CSIRO handleability test were compared to other handleability tests that are used in Australia (Jenike type direct shear tester and the Durham Cone) and overseas (Edinburgh Cohesive Tester (ECT) and the Johanson Indicizer). Good agreement was obtained for five PCI coals that were assessed by the CSIRO handleability test, the ECT and the Johanson Indicizer but lesser agreement was obtained on a wider range of coals types (three PCI coals, two thermal coals, and one coking coal) with the CSIRO Handleability test, the Jenike Flow Factor and the Durham Cone.
The relationships between coal properties and handleability were found to be complex and a simple handleability index was not able to be developed. A non linear data analysis technique, CSIRO’s implementation of Self Organising Maps (CSOM), was able to use the commonly determined coal properties of size distribution, moisture (total, free and bound) and ash abundance and ash constituents to cluster the coals in a way that replicated quite well their handleability performance.
This means that it should be possible to predict the handleability of other coals from these common coal tests. As mineral composition is not routinely determined the mineral information obtained for the thirteen samples was not used for the handleability predictions.
Six of the seven most significant parameters for handleability performance were size related (mass% less than 0.5 mm). Interestingly, total moisture ranked as a more significant parameter than either free moisture or bound moisture. For this suite of mainly mine product coals the total ash % was a less significant parameter for handleability, than some of the ash constituents. Of the ash constituents SO3, Al2O3, SiO2, CaO and BaO, were most significant and TiO2, P2O5, Na2O and SrO were least significant. The soil test for determining clay content was found to be of no use for identifying product coals that may have sufficiently high clay contents that they gave handleability problems. This project focussed on coals that were prepared for the export market and generally
had lower ash contents (and presumably clay contents) than the coals that are used in domestic power stations.
The CSIRO handleability test is a good method of quantifying handleability, especially the unloading performance of bottom dump rail wagons. It is ideally suited for example, for evaluating the unloading characteristics of newly developed product coals, prior to the commencement of railing operations. The test uses approximately 90kgs of product coal and takes approximately one day to complete. Consequentially it may be less applicable when quick and inexpensive results are required.
A non linear data analysis technique such as CSOM is able to use simple coal properties of size distribution, moisture and ash content and ash constituents to cluster the information in a way that replicated coal handleability performance and hence can be used to provide an assessment of the handleability characteristics of untested coals. This data analysis method may have applicability to other coal quality issues.