A model for the statistical distribution of microlithotypes in coal

Microlithotype composition of a coal sample is often summarized by examining a large number (~500) of subsamples of a grain mount and estimating proportions of vitrite, intermediates, and inertite, where, for samples we have investigated, the proportion of intermediates is generally less than 0.4. This suggests that most subsamples are either greater than 95% vitrinite or greater than 95% inertinite, so that the statistical distribution of vitrinite has most of its weight in its tails. Two distributions which may have this shape are the beta and the logistic normal, and these have been fitted to the microlithotype distribution of some coal samples. Parameters of these fitted distributions are related to the proportion of vitrinite in the sample and thickness of microscopic bands in the coal. For coals in the Sydney Basin, at least, it was found that the parameter relating to band thickness is approximately constant over a coal seam; therefore, fitting one or other of these distributions to such data leads to directly interpretable parameters.