Analysis of the Inorganic Constituents of Pulverized Coal on a Particle-By-Particle Basis Using a Computerized SEM-EDS System

"This paper is based on the development of the Maza-Austin technique for characterizing thousands of coal particles by approximate analysis of the major inorganic constituents Ca, Al, Si, K, Na, Fe and S in each particle. The technique has sufficient sensitivity to work on small volumes, is fast enough to handle large number of particles within a reasonable period of time, and consists of scanning electron microscopy combined with energy dispersive x-ray spectrometry (SEK-EDS). A computerized SEK-EDS system that allows the automated analysis of hundreds of coal particles per hour' is described. The analysis is based on the location and tracing of coal particles embedded in a polished specimen, the selection of sampling points on the traced particles, the collection of an x-ray spectrum from each sampled point for a defined period of time, and the processing of the x-ray intensity to calculate point composition using ZAF corrections. Procedures for correcting for dead time, background, detector efficiency and spectrum overlap are described. These corrections are essential if reasonably accurate analyses are to be obtained from unknown complex mixtures of the elements in a coal or fly ash particle. Examples are given. INTRODUCTION Coal is a complex mixture of organic and inorganic materials. The inorganic fraction is referred to as mineral matter (Gluskoter, 1978) but can occur in three modes: (i) an observable mineral species; (ii) as cations held by ion exchange to carboxylic groups; and (iii) in chelate complexes. The mineral matter content normally represents a significant proportion of a coal, with an average value of about 15% for US coals (Gluskoter, 1975). The standard methods of analysis of the mineral matter in coals are performed on a sample selected to represent the whole coal as closely as possible, and give the overall inorganic composition in the coal. Such analyses are of limited use for solving problems originating from the behavior of individual coal particles, whose compositions vary widely from the average value. This is particularly true in pulverized coal combustion, where the current methods of predicting the slagging and ·fouling potential of a coal are based on the bulk ash composition, but the fly ash particles which create the deposits originate from coal particles with large differences in mineral compo1ition. Some examples can be found where, in a comparison between two coals, the one with the smaller slagging potential according to the mean ash composition was the one that behaved worst in the furnace (Moza et al., 1980; Koza and Austin, 1983). The u1e of gravity fractionation to assess the slagging potential of pulverized coals has shown that conventional ASTM analysis is not sensitive to differences in the distribution of mineral matter when comparing coals (Borio and Narciso, 1979). Bryer (1979) concluded that coals should be examined for fouling and slagging potential by analyzing gravimetric and size-fractionated coal, preferably pulverized· and ground to the size expected at the mill outlet."