RI 4735 Physical Properties Of Coke: Size And Its Measurement

Twelve samples of run-of-oven coke were taken at coke plants in the United States and Canada. Sampling procedure was directed toward obtaining the highest possible degree of uniformity in the gross samples, so that variations in results caused by sampling variations were at a practical minimum. Under these conditions the standard deviation of the screen analysis test on 500-pound samples is found to range from 1.2 percent for the 1-inch screen to 4.5 percent for the 5-inch screen. At present, the A.S.T.H. standard method of test requires that results be reported to the nearest 0.1 percent. This implies a higher than actual degree of precision, and revision of the report required by the method is suggested. Much larger Samples than are generally tested are required to reduce normal testing errors to a point where they are not appreciable fractions of the total variations found in a smoothly operating plant. For example, it is deduced that over 100,000 pounds of coke must be screened to determine if ordinary day-to-day variations in screen analysis have appreciable effects on the performance of a blast furnace.