Development of Water-Based Methods for Determining Coal Washability Data

Abstract

In Australia, coal washability analysis, which is used to obtain the cumulative yield—ash % relationship—has traditionally been generated using the float-sink method AS4156.1-1994 [1]. This technique relies on heavy organic liquids to fractionate coal and mineral particles based on particle density. These organic liquids are considered problematic for health, safety, and environmental reasons. This paper examines water-based methods for attaining coal washability data for −50 + 0.045 mm particles, covering a water fluidization method based on the Reflux Classifier, and two batch jigging devices. Data produced by the three methods are compared with results obtained using the float-sink method. The cumulative yield versus cumulative ash curves generated by all three methods were found to be in reasonable agreement with the data obtained using the float-sink method, but with a consistent discrepancy evident at relatively low ash levels. The discrepancy was attributed to the effects of dispersion, a phenomenon inherent in most particle separation processes. The fractionation achieved using the Reflux Classifier required further fractionation by size separation, while the two jigs generally required no fractionation by size provided the initial size range was not too broad. The Boner Jig, in which the vessel is moved up and down at a given frequency and amplitude, showed the best performance on the −16 + 0.25 mm size range. The Mintek Jig, in which the upward fluid flow is varied, showed the best performance on the −50 + 16 mm size range. For the ultrafine particles, covering the size range −0.25 + 0.045 mm, the Reflux Classifier proved to be the ideal choice, achieving higher yields than multistage flotation. It was also demonstrated, using the data from the Reflux Classifier, that the “true” washability data could be deduced mathematically by combining a model of the separation process, with the normalized partition curve [2].