Abstract
Beginning with the crude attempts at coal sampling reported 100 years ago, the author reviews the slow progress made down to 1916, when Bailey’s experimental work on sampling was incorporated in A.S.T.M. Designation D-21, Standard Method of Sampling Coal for Analysis. By this standard, which is still official, the gross sample taken must be not less than 1000 lb; a large number of increments are required involving considerable time and expense, which has led in many cases to taking smaller samples, causing uncertainty in representativeness of the samples. In 1930, Grumell and Dunningham, in England, modified the method to include the effect of ash content and size of piece on occurrence of impurities. Far less stringent sampling rules were subsequently formulated. Many studies were then commenced on this subject, and about 1934, Subcommittee XIII of A.S.T.M. Committee D-5 began to reconsider specifications for taking gross samples, and tentative standards (D492-43T) have been prepared. The present paper is devoted to classifying missing data on coal sampling which are needed to supply information on (a) consist; (b) variability of ash of pieces by float-and-sink method, and by direct analysis; (c) variability of ash of increments; supplemented by (d) collection and analysis of gross samples of specified number of increments of constant weight to give confirmation of established relations between variables involved. Four sampling experiments are described from which principles are developed to bring out one of the important characteristics affecting the sampling of coal. This is the degree of mixing of the coal at the point of sampling.
