14.RS: cir a*<° c. 2 Gqiek SulOjJ bu\ STATE OF ILLINOIS WILLIAM G. STRATTON, Governor DEPARTMENT OF REGISTRATION AND EDUCATION VERA M.BINKS, Director Influence of Coking Time on Expansion Pressure and Coke Quality H. W. Jackman R. L. Eissler R. J. Helfinstine DIVISION OF THE ILLINOIS STATE GEOLOGICAL SURVEY JOHN C. FRYE, Chief URBANA CIRCULAR 246 1958 ILLINOIS GEOLOGICAL SURVEY LIBRARY FEB 25 1958 ILLINOIS STATE GEOLOGICAL SURVEY 3 3051 00004 4556 INFLUENCE OF COKING TIME ON EXPANSION PRESSURE AND COKE QUALITY H. W. Jackman, R. L. Eissler, and R. J. Helfinstine ABSTRACT Pilot-plant coking tests on coal blends have shown that ex- pansion pressure of coke against the oven wall varies with coking time, but not always in a predictable manner. We concluded that when coals are to be judged on the basis of these tests, they should be made at the rate of coking to be used commercially. A standard test oven, and standard operating procedures, should be developed if expansion pressures determined indifferent laboratories are to be compared. As is well known, physical properties of coke vary with the rate of coking; this report shows trends over the wide range studied. INTRODUCTION The expansion pressure developed against the wall of a laboratory test oven by any blend of coals is an empirical value that may vary according to the design of the test oven or according to operating procedures used in the test. Just how much this pressure may vary for any specific coal blend can be de- termined only by experimentation. With some coals it has been shown to be ap- preciable. It is common knowledge that the pressure developed in a coke oven de- pends on the bulk density of the coal in the oven. Loosely packed coal tends to develop low pressure during carbonization. Conversely, an expanding coal blend charged at a high bulk density may produce damaging pressures on the walls of an oven, and for this reason bulk density generally is controlled in plant operation, either by maintaining coal pulverization within narrow limits, or by addition of moisture or oil to the coal. In order to determine the maximum pressure that might develop in commercial ovens, coal samples for expansion tests are commonly air dried so that tests may be made at the greatest bulk density that might prevail in any portion of a commercial coke oven. Other operating procedures also affect the pressure imposed on oven walls. Among these is the rate at which coal is carbonized, or, in the common terminol- ogy, the coking time. Russell (1949) reports a definite reduction in pressure as the flue temperature of the 12-inch oven is reduced and the coking time lengthened, Others have reported that variations in the rate of coking have an indefinite effect on expansion pressure, in some cases causing it to decrease and in some cases to increase (British Coke Research Assn. , 195 2). In view of this somewhat conflicting evidence , the Illinois State Geologi- cal Survey has tested coal blends over a wide range of coking times in its movable- wall coke oven. This oven (fig. 1) is 17 inches wide and has a capacity of 675 pounds of coal (Jackman, 1955). When operating the oven under temperature conditions simulating those of commercial ovens, we determine the expansion pressure of the blend and obtain a coke that closely duplicates the commercial product in physical and chemical properties. There were two practical reasons for this investigation. First, we had found, as did Russell, that certain coal blends expanded strongly when tested [1] ILLINOIS STATE GEOLOGICAL SURVEY t^ i^r Steel Fireclay brick Silicon carbide Insulating brick Alcor brick Castable Fig. 1. - Cross section of oven through brickwork. under usual laboratory procedures, but developed lower pressures when coked slowly, as for the production of foundry coke. We wished to determine whether this relationship is true of blends in general, or whether with certain blends the pressure may remain constant or increase at slower rates of coking, as noted in the British source cited. Also, knowing that coking time does affect the expansion pressure, we wished to show the need for standardizing test procedure in the operation of movable -wall ovens when comparisons of tests run in different labo- ratories are made, or when coals are to be accepted or rejected on the basis of a maximum pressure value such as the well known limit of two pounds per square inch. EXPANSION PRESSURE AND COKE QUALITY o to a. LU Q. en OD _l uf ^"^ ^— B C r 6 8 TIME (HOURS) 10 12 14 Fig. 2. - Representative pressure curves. Z5 2.0 1.5 -. 1.0 1.5 COAL BLEND 75% III. No. 5 25% Pocahontas Normal oven heating PEAK PRESSURES 0--0- f I AVERAGE PRESSURES -\-> b--' 10 15 20 COKING TIME. HRS. Fig. 3. - Expansion pressure vs. coking time. 25 The coals used in this investigation were obtained from both the Illinois and Appalachian fields. We wish to thank both the coal and steel companies in the Illinois -Indiana area that furnished certain of these coals, and the coal producers in Illinois and Virginia who furnished coals directly from the mines. PROCEDURE Tests on a number of coal blends, including those that develop very low pressure with practically no pressure peak, those that develop a dangerously high pressure peak, and the more normal, average type of blend are described. Blends thought to be suitable for both blast furnace coke and foundry coke were tested. 4 ILLINOIS STATE GEOLOGICAL SURVEY The 17-inch width of the experimental coke oven is equal to the average width of a commercial oven, so that the experimental rates of coking correspond directly with those of full size ovens. Coking times for individual blends were varied from a 12-hour minimum to a 28-hour maximum. The usual rate, however, was from 14 to 24 hours. Maximum and average expansion pressures over this range were noted and have been plotted to show trends. In addition, the physical properties and yields of coke were determined for each experimental oven test. These data on coke quality have been plotted to show general trends resulting from the variation in coking time. EXPERIMENTAL RESULTS Pressures in the experimental oven, recorded from the time coal is charged until after a pressure peak has been registered, tend to follow one of the curves shown in figure 2. Curve A represents the pressure produced during coking of a blend having fairly high expansion properties. Pressure generally increases rapidly for about four hours with blends of this type. After a period of gradual increase, a rapid rise in pressure occurs as the plastic zones meet. This is followed by a sudden pressure drop when the plastic en- velope breaks, releasing trapped gases. Curve B represents the pressure conditions when a blend with fairly low expansion properties, such as a mixture of Illinois coals with Pocahontas, is coked. Curve C depicts the extreme condition occurring when gases are not entrapped by the plastic envelope in sufficient quantity to produce a definite peak. Variations in curve C may occur in which a maximum pressure is reached within the first four hours, and is followed by a gradual pressure drop. In this case a final peak may or may not develop. For the purpose of discussion we have divided the blends studied here into two general groups: first, that in which the chief high-volatile constitu- ent is Illinois coal, and second that group in which the high-volatile coal comes chiefly from the eastern coal fields. The analyses and plastic properties of all coals used in blends are shown in table 1, and the expansion pressures, both maximum and average, for both groups of blends are shown in tables 2 and 3. Analyses of blends and the cokes produced are shown in table A of the appendix. Table 1. - Analyses and Plastic Properties of Individual Coals Analyses Moisture-free basis Coal M. V.M. F.C. Ash Sulfur F.S. Illinois No. 5 6.6 37.0 55.4 7.6 1.30 6 Illinois No. 6 9.1 38.0 54.3 7.7 1.01 5 Kentucky Elkhorn 5.0 38.2 57.5 4.3 0.99 5| West Virginia Eagle 5.4 29.5 64.8 5.7 0.66 9 Virginia medium-volatile 3.1 21.3 73.3 5.4 0.58 9 Pocahontas 4.0 17.0 75.9 7.1 0.87 9 Gieseler Fluidity Plastic Range Dial div. per min. °C. 58 77 37 75 144 65 3,000 100 1,020 92 10 61 EXPANSION PRESSURE AND COKE QUALITY Table 1. - Continued Plastic Properties Coal Illinois No. 5 Illinois No. 6 Kentucky Elkhorn West Virginia Eagle Virginia medium-volatile Pocahontas Group I - Illinois Coal Blends Illinois No. 5 Coal Blended with Pocahontas The first blend studied contained 75 percent Illinois No. 5 and 25 percent Pocahontas coals. This blend has a relatively low expansion pressure regard- less of the rate at which it is coked. In this study the blend was carbonized at a range of flue temperatures that caused complete coking in 14:00, 15:10, 17:00, 18:20, and 21:00 hours, respectively. Maximum and average wall pres- sures are plotted in figure 3. At what might be considered the usual coking time range for producing blast furnace fuel in a 17-inch oven, 15| hours to 19| hours, the maximum ex- pansion pressure for this blend was shown to range between approximately 1.45 and 1.0 pounds per square inch. Greatest pressures were recorded for coking times of 15 and 17 hours, and the pressure decreased during longer coking times. It was noted also that at the extremely fast coking time of 14 hours a lower pressure was exerted, probably because of the greater shrinkage and cracking of the coke structure taking place at this rapid rate of heating. To check this blend of coals further, a second series of coking tests was made with somewhat faster heat input to the coal at the start of each run. This procedure simulated the effect of charging coal into unusually hot ovens. A maximum pressure curve of the same general shape as before was obtained, but with slightly lower values (fig. 4). The greatest expansion pressure obtainable was 1.31 pounds per square inch. Pressures ranged from 1.25 to 0.95 pounds in the usual coking range. Illinois Coals Blended with Medium-Volatile Coal Next to be studied were two blends, one of 50 percent Illinois No. 5 coal and 50 percent medium -volatile coal from Virginia, and the other of 50 percent Illinois No. 6 coal and 50 percent of the same Virginia coal. The medium-volatile coal (21 percent V.M.) had a relatively high Gieseler fluidity, and like similar coals tended to exert less pressure during carbonization than the less volatile Pocahontas coals. The blend of this Virginia coal with the Illinois No. 5 produced an ex- ceptionally strong coke, but under no conditions was a wall pressure as high as 1.1 pounds per square inch obtained. In the usual coking range the maximum ILLINOIS STATE GEOLOGICAL SURVEY 2.5