A Comparison of Centrifugal Float-Sin k Testing with Alternative Methods for Determining Grade-Yield Curves of Fine Coal

This paper describes a comparison of centrifugal float-sink testing with two alternative methods for characterizing the beneficiation potential of fine coal. The alternative methods evaluated were release analysis and scanning electron microscope automated image analysis (SEM-AIA). The primary objective of this study was to compare the extent of separation measured (or predicted in the case of SEM-AIA) by each of the three techniques in terms of their respective grade-yield curves. A secondary objective of the study was to compare the separation results of two surface-based fine-coal cleaning processes [selective agglomeration and selective hydrophobic coagulation (SHC)] with the grade-yield curves of the characterization techniques The study revealed that the grade-yield curves derived from the release analysis tests indicated significantly lower mineral matter rejection values for a given yield than those from either centrifugal float-sink testing or SEM-AIA. Grade-yield curves for SEM-AIA showed higher mineral matter rejection values for a given yield than those obtained for centrifugal float-sink testing, indicating that a greater level of cleaning was theoretically possible. Because SEM-AIA tends to overestimate liberation, leading to an artificially high predicted level of separation, it is believed that the maximum achievable grade-yield curve will, in all likelihood, fall somewhere between those produced by centrifugal float-sink testing and SEM-AIA Selective agglomeration and SHC were tested to evaluate their coal cleaning capabilities relative to the grade-yield curves generated by the three characterization techniques. Based on very limited data, at a fixed mineral matter rejection, the selective agglomeration and SHC processes achieved yields lower than those for centrifugal float-sink testing and SEM-AIA, as would be expected when comparing surface-based separation processes with density-based characterization techniques. Within experimental error, the selective agglomeration results were coincident with those obtained by release analysis. The SHC process could not be compared with release analysis because release analysis was not performed on the sample used for SHC testing.