THE COLLIERY GUARDIAN. 590 ________________________________________________________ DETERMINATION OF MOISTURE IN COKE.* By A. C. Fieldner and W. A. Selvig. The experiments described below were undertaken, in the course of fuel investigations made by the Bureau of Mines, with the purpose of ascertaining the most rapid and simple method for determining the moisture in coke with sufficient accuracy for use in specifications for the purchase of coke by the Govern- ment, by industrial concerns, or by private individuals. For the determination of moisture in coal, the com- mittee on coal analysis of the American Society for Testing Materials and the American Chemical Society recommend heating the pulverised sample (60-mesh) for one hour at 104 degs. to 110 degs. Cent, in an oven through which is passed a rapid current of air dried by passing through concentrated sulphuric acid. Under the same conditions 20-mesh samples must be heated 1| hrs. In order to obtain comparable results, the method specified must be strictly followed, especially with coals normally having a high moisture content. Experiments have shown that such coals will easily give off 1 per cent, more moisture in an atmosphere of circulating dry air than in an atmosphere of more or less stagnant air in which there is an appreciable partial pressure of water vapour. It is also essential, in order to obtain the correct moisture content, to pulverise the coal to a fineness of at least 10-mesh before the final moisture determination. The best practice is to air-dry the coal at | or J in. size, and then to pulverise it for the final determination of residual moisture, thus determining the water in two stages. Undoubtedly the above method for the determina- tion of moisture in coal gives results of the same degree of accuracy when applied to coke. However, owing to the much greater stability of coke as com- pared to coal, it is believed that a simpler method gives accurate results. Coal is a complex mixture of extremely sensitive organic compounds containing a considerable propor- tion of hydrogen and oxygen. These compounds are easily altered by heating and by oxidation. On the other hand, coke is the residual product obtained from heating coal to a high temperature, and consists chiefly of carbon and inorganic ash-forming consti- tuents. The hydrogen and oxygen content is less than 1 pel’ cent., and the carbon is fused, so that the great number of very finely divided particles of the original coal substance, which afforded a vast surface system for the condensation of moisture films, are united. It therefore seems probable that the circulation of dried air is unnecessary; that larger temperature limits than 104 degs. to 110 degs. Cent, are permis- sible ; that no appreciable gain in weight, due to absorption of oxygen, takes place on heating longer than one hour ; and, finally, that practically all the moisture should be removed on heating the unpul- verised coke to constant weight. If the latter procedure is permissible, the ever-present danger of loss of moisture during pulverisation will be eliminated, and it will be possible to determine the moisture content of a shipment of coke with the same ease and simplicity as for iron ores. Comparatively few reports of experiments on the determination of moisture in coke have been published. Most textbooks recommend the same pro- cedure as for coal. A few investigators have recog- nised that coke may he dried to constant weight in much the same manner as ore is dried. Thorner recommends that the gross sample of coke be pulverised and quartered to hazelnut size imme- diately after collection. A representative portion of 1,200 to 1,500 grms. is placed in a galvanised-iron can, sealed moisture-tight, and shipped to the laboratory. At the laboratory a glass tube is inserted in a suitable opening in the can and pushed down until the end is near the bottom. The can is thou weighed, placed in an air oven at 105 to 110 degs. Cent., and heated to constant weight while a current of air is drawn through the tube. Usually 8 to 10 hrs. are required to remove all the moisture. The 1912 report of the committee on standard methods for determining the constituents of foundry coke specifies that the gross sample of coke shall be crushed to % 111 • size, mixed and quartered. One quarter shall be reserved for the determination of total moisture. This determination consists in dry- ing to constant weight at 104 to 107 degs. Cent. Wagener found that saturated lump coke contain- ing 16 to 20 per cent, water, lost 4 per cent, water while being pulverised and put through a sieve having 4 mm. meshes. Unsaturated coke containing 9 per cent, water lost from 1 to 2 per cent, water. To avoid this loss of moisture, Wagener recommended that the gross sample of coke be quickly broken down to fist size with a hammer, then mixed and quartered. Two opposite quarters weighing about 60 kilogs. are weighed on a tared galvanised-iron pan, placed in a large oven, and dried for 24 hrs. at a temperature of 150 degs. Cent. The loss in weight is the total moisture. This dried sample is then pulverised and analysed, it being assumed that moisture is reabsorbed from the atmosphere during the pulverising process. Effect of Dry-air Circulation and Duration off Heating. The first series of tests by the authors of this paper were made to ascertain whether the circulation of dry air through the oven is necessary to remove the maximum amount of moisture. Twelve samples of different metallurgical cokes from various sources were prepared by crushing the air-dry coke to 60 mesh. Moisture determinations were made in duplicate by the standard method used by the Bureau of Mines for determining moisture in coal.f Samples weighing * U.S. Bureau of Mines, Technical Paper 148. t F. M. Stanton and A. C. Fieldner, “ Methods of Analysing Coal and Coke ” (Technical Paper, 8, Bureau of Mines, 1913, pp. 12-13). 1 grin, were placed in J in. by 1J in. porcelain capsules and in sets of 12 heated for one hour at 105 degs. Cent, in a constant-temperature oven. A current of air, dried by concentrated sulphuric acid, was passed through the oven and over the samples rapidly enough to change the total volume of air in the oven two to four times in one minute. At the end of the heat- ing period the capsules were covered with closely fitting, flat aluminium covers, and cooled in separate desiccators over concentrated sulphuric acid. Each capsule was rapidly weighed to the nearest 0-5 mg. immediately upon removal from the desiccator. Duplicate determinations were also made in sets of 12 in the same oven without using dry-air circulation, and a fifth set of determinations was made in which the period of heating was extended to two hours, in order to ascertain whether any gain in weight throngii absorption of oxygen took place. Coal samples heated for two hours usually show less loss in weight than samples heated for one hour. As shown in Table 1, no appreciable difference in moisture was obtained by circulating dry air through the oven. The difference for the 12 samples was only 0*005 per cent. Neither was there any appreciable change in weight on heating the samples for two hours. Table 1.—Results of Determining Moisture at 105 degs. Cent, in 60-mesh Coke, with and without Dry Air Circulation. Percentage of moisture. Difference. Heated 1 hour in current Heated 1 hour in Heated Coke No. of dry air. stagnant air. 2 hours Columns Columns - . — in stagnant 4 and 7. 7 and 8. Set 1. Set 2. Average. Set 3. Set 4. Average. air. 1 2 3 4 5 6 7 8 9 10 13651 0*15 0*16 0*16 0*15 0*16 0*16 0*19 0*00 + 0*03 13672 0*26 0*27 0*27 0*26 0*30 0*28 0*33 4-0*01 + 0*05 13707 0*20 0*17 0*19 0*16 0*21 0*19 0*21 0*00 + 0*02 13743 0*63 0*68 0*66 o*6o ; 0*67 0*64 0*71 -0*02 +0*07 13784 0*07 OTO 0*09 0*07 on 0*09 0*06 0*00 -0*03 13829 0*12 0*13 0*13 0*12 0*10 0*11 009 —0*02 - 0*02 13388 0*11 0 10 0*11 0*10 0*10 0*10 0*10 -0*01 0*00 14172 0*09 0*13 0*12 0*12 0*12 0’12 0*08 0*00 — 0*04 14183 0*33 0*33 0*33 0*32 0*33 0*33 0*30 0*00 -0’03 14730 0*44 0*48 0*46 0*47 i 0*53 0*50 0*47 +0’04 -0*03 15276 0*21 0*22 0*22 0*21 0*21 0*22 0*22 0*00 0*00 16933 0*37 0*36 0*37 0*30 0*31 0*31 0*37 -0*06 + 0*06 Average -0*005 + 0*007 Effect off Temperature. The second series of tests were made to ascertain the effect of heating the 60-mesh coke at temperatures higher than 105 degs. Cent. Three sets of determina- tions were made, namely, set No. 1 by the standard method at 105 degs. Cent., as in first series; Set No. 2 by heating in stagnant air at 150 degs. Cent.; and set No. 3 by heating in stagnant air at 200 degs. Cent. Each set was heated for a period of 1 hr., then cooled over concentrated sulphuric acid, and weighed; then it was returned to the oven, heated for an additional 30 mins., and again weighed. The results are given in Table 2. Table 2.- Results of Determination of Moisture in 60-mesh Coke at Various Temperatures. Percentage of moisture. Difference- ( * Heated in current of Heated in stagnant Heated in stagnant Coke No. dry air at 105 degs. air at 150 degs. air at 200 degs. Col- Col- Cent. Cent. Cent. A 3 and 3 and r 7 5. 7. 1 11 1 u 1 U hour. hours, hour. hours, hour. hours. 1 2 3 4 5 6 7 8 9 13651... ... 0*15.. . 0*15... . 0*30.. . 0*31. .. 0’36. .. 0*37... . 0*16.. . 0*22 13672 .. ... 021.. . 0*22... 0*42.. . 0*42. .. 0’51.. . 0*53... 0*20.. . 0*31 13707... ... 0*20.. . 0*22... 0*28.. . 0*29. .. 0’32. .. 0*34.. . 0’07.. . 0*12 14172... ... 0’12., . 0*13... 0*23.. . 0*23. .. 0’30 .. 0*30.. . 0*10.. . 0’17 13743... ... 0’66.. . 0*70... 0*84.. . 0*89 0*95.. . 0’96... 0*19.. . 0*26 13784... ... 0*08.. . 0*09... 0*11.. . 0 11. .. 0*16.. . 0’18... 0*02.. . 0*09 14183... ... 0*31 . . 0*31... 0*49.. . 0*47. .. 0*62.. . 0’62... 0 16.. . 0’31 16933... ... 0*36.. . 0*39... 0*51.. . 0*51. .. 0*59.. . 0*59... . 0*12.. . 0*20 14730... ... 0 54.. . 0*55... 0*65.. . 0*65. .. 0*79.. 0*80... . 0*10.. . 0*25 15276... ... 0*26.. . 0*26... 0*35.. . 0*36. .. 0*45.. .. 0*47... . OTO.. . 0*21 13888... ... 0*11.. . 0*11... 0*15.. . 0*15. .. 0’23. .. 0’25.. . 0*04.. . 0’14 13829... ... 0*13. .. 0*13... . 0*22.. . 0*22. .. 0*28. .. 0*28.. . 0’09.. . 0*15 Average ____________............................... 0’11... 0*20 For each of the three temperatures, 105 degs., 150 degs., and 200 degs. Cent., the loss in weight for one hour was practically the same as that attained through additional heating for 30 mins. There was, however, a slightly greater loss at the higher temperatures. The greatest difference between 105 ’degs. and 150 degs. Cent, was 0-20 per cent., and between 105 degs. and 200 degs. was 0*31 per cent. The average dif- ferences were 0J11 and 0-20 per cent, respectively. It is probable that the greater loss at the higher temperatures is due to the removal of residual traces of water and gases condensed in extremely thin films on the surfaces of the coke particles. Mack and Hulett (Am. Jour. Sci., vol. 43, 1917, pp. 89-110) have called attention to these films in connection with the water in coal, and have pointed out that such water would probably have no measurable vapour pressure at 100 degs. Cent., but would gradually, as the tem- perature increased, develop sufficient vapour pressure to permit removal of additional moisture at higher temperature. However, the differences that were obtained on heating the coke powder between limits of 105 degs. and 150 degs. Cent, may be regarded as negligible; even the maximum difference of 0’31 per cent, at 200 degs. Cent, is within the unavoidable moisture variations likely to be incurred in preparing the gross sample for analysis. Direct Determination of Moisture in Lump Coke. Experiments on the direct determination of moisture in 1 in. lumps and also in 4-mesh coke at various tem- peratures were made in the following manner: — March 22, 1918. ______________________________________________________ A large sample of coke was taken from a pile that was wet from a recent rain and was crushed to 4-mesh size. This sample was then thoroughly mixed and divided into three representative portions, Al, A2, and A3, which were treated as described below. Sample Al was air-dried to constant weight at 30 to 35 degs. Cent., then rapidly crushed to 10 mesh by passing it through crushing rolls. The 10-mesh pro- duct was quickly reduced on a riffle sampler to 400 grms., which was pulverised to 60-mesh in an air-tight porcelain Abbe ball mill containing flint pebbles. The 60-mesh sample was then reduced to 60 grms. on a small riffle sampler and placed in a rubber-stoppered bottle. The residual moisture in this air-dried sample was determined at 105 degs. Cent., according to the standard method previously described in this paper. The total moisture “as received ” in sample Al was computed as follows: — Moisture “ as received ” = Moisture in 60-mesh coke _ (100 — air-dry ingloss) . . . X q-7 ■ ■ 4- air-drying loss. Ln order to check any possible change of moisture content in the air-dry sample, while it was being pul- verised and reduced from 4-mesh to 60 mesh, samples for moisture determination were taken at various stages of the sampling procedure as follows: — Table 3.—Percentage of Moisture in Air-dry Coke at Various Stages in the Preparation of the 60-mesh Laboratory Sample. Kind of sample. Moisture. Per cent. 4-mesh sample taken immediately after air-drying ; 341*5 grammes dried to constant weight at 150 degs. Cent______________________________ 0*32 10-mesh sample taken immediately after passing . through rolls; 433*6 grammes dried to constant weight at 150 degs. Cent. ................... 0*23 60-mesh sample taken immediately after opening ball mill; 1 gramme dried one hour at 105 degs. Cent, in a current of dry air......... ........... 0'31 Final 60-gramme laboratory sample; determina- tion as above ............................. 0*35 The maximum variation in the different samples was 0-12 per cent, which is of the same magnitude as the error in the moisture determination. Hence it may be assumed that the “as received” moisture as deter- mined in sample Al is the true moisture content of the coke, and may be used as a standard of reference for judging the accuracy of the results that were obtained on samples B and C by shorter methods. Sample A2 (at 4-mesh size) was dried to constant weight at a temperature of 105 degs. Cent, in a Freas electrically heated oven without dry-air circulation ; and sample A3 (at 4-mesh size) was similarly dried to constant weight in an ordinary gas-heated oven at a temperature of 150 degs. Cent. The results obtained on the three samples are given in 'Fable 4. together with the results on a second set of three samples, Bl, B2, and B3, which were similarly pre- pared from another lot of coke. Table 4.—Results obtained by Heating 4-mesh Coke to Various Temperatures. a' § §2 *^s Sample. Fineness. S1®*^ ” O O O 1-2 fe C-I o Degs.C. Hours. Perc. Sample Al, 2,056 grms. 4-mesh ... 30-35 ... 120 ... 10*1 Sample Al, 1 grm.... 60-mesh ,. 105 ... 1 ... 0*35 Total moisture “ as ---- received” ................................. 10*4 Sample A2, 1,089 grms. 4-mesh ... 95-105 ... 2*5 ... 10*5 Sample A3, 514 grms.... 4-mesh ... 150... 2’0 ... 10*5 Sample Bl, 1,630 grms. 4-mesh ... 30-35 ... 72 ... 9*7 Sample Bl, 1 grm..... 60-mesh ... 105 ... 1 ... 0’38 Total moisture