Coal characterization research: sample selection, preparation, and analyses

Procedures for estimating responses of coals in synthetic fuels conversion processes are practically nonexistent. Consequently, a new look at coal characterization and classification procedures is propitious. Meaningful understanding of the relationships between coal properties can be obtained only through the study of a large suite of carefully selected, prepared and analyzed coal samples. Therefore, a comprehensive coal characterization research program was begun at the Baytown, Texas, laboratory of Exxon Research and Engineering Company. Sixty-six fresh samples of coal representing the coalification band in an H/C vs O/C plot have been analyzed in detail. Eighty percent of the samples contain 80% vitrinite. Through meticulous preparation procedures designed to minimize exposure to air, through numerous cross-checks of validity of the analytical data, and through calculation of the data to a mineral-matter-free basis a data library has been assembled that is being used to define interrelationships between coal properties. Numerous multivariate correlations have been found indicating strong dependence of properties such as density, free swelling index and volatile matter on elemental composition. Such correlations promise to shed considerable light on the area of coal characterization and classification.