INTERRELATIONSHIPS BETWEEN COAL COMPOSITIONAL PARAMETERS

Analytical data from a set of 66 coal samples, ranging in rank from lignitic to low-volatile bituminuous, were used to define relationships between coal properties. Unusual care was taken in selecting, collecting, preparing and analyzing this set of premium samples providing a high degree of confidence in the conclusions obtained. It was found that univariate relationships between properties, which might seem to convey useful information, and generally of limited operational value and can lead to erroneous conclusions. However, multivariate correlations can predict a number of properties. Specific energy, volatile matter yield, and density (by helium pycnometry) can be estimated from the elemental composition.