4.GS- a. a Albert ^HJUfLO JCiH I U A, MAIllfl ggrf "1 STATE OF ILLINOIS HEXRY HORXER, Governor DEPARTMENT OF REGISTRATION AND EDUCATION JOHN J. HALLIHAN, Director DIVISION OF THE STATE GEOLOGICAL SURVEY M. M. LEIGHTON. Chief URBANA REPORT OF INVESTIGATIONS — NO. 58 A STUDY OF THE EQUILIBRATION METHOD OF DETERMINING MOISTURE IN COAL FOR CLASSIFICATION BY RANK O. W. Rees, F. H. Reed, and G. W. Land ^.^ ' r^ t •" PRINTED BY AUTHORITY OF THE STATE OF ILLINOIS URBANA, ILLINOIS 1939 STATE OF ILLINOIS n. Henry Horner, Governor OF REGISTRATION AND EDUCATION . John J. Hallihax, Director BOARD OF RESOURCES AND CONSERVATION Johx J. Hallihax, Chairman Henry C. Cowles, Ph.D. Forestry Arthur Cutts Willard, D.Engr., LL.D., President of the University oloxy of Illinois D.Sc, bLOGICAL SURVEY DIVISION Urban a M. Leightox, Ph.D., Chief .-xley, M.S., Assistant to the Chief Jane Titcomb, M.A., Geological Assistant GEOLOGICAL RESOURCES Coal G. H. Cady, Ph.D., Senior Geologist L. C. McCabe, Ph.D. James M. Schopf, Ph.D. Earle F. Taylor, M.S. Charles C. Bolf.y, M.S. Industrial Minerals I. E. Lamar, B.S. H. B. Willman, Ph.D. Douglas F. Stevens, M.E. Robert M. Grogan, M.S. J. S. Templeton, A.B. Oil and Gas A. H. Bell, Ph.D. G. V. Cohee, Ph.D. Frederick Squires, B S. Charles W. Carter, Ph.D. F. C. MacKnight, Ph.D. Frank E. Tippie, B.S. Roy B. Ralston, B.A. Areal and Engineering Geology George E. Ek.bi.aw, Ph. D. Harry McDermith, B.S. Richard F. Fisher, B.A. Subsurface Geology L. E. Workman, M.S. J. Norman Payne, Ph.D. Elwood Atherton, Ph.D. Merlyn B. Buhle, M.S. Gordon Prescott, B.S. Stratigraphy and Paleontology J. Marvin Weller, Ph.D. Chalmer L Cooper, M.S. Petrography Ralph E. Grim, Ph.D. Richards A. Rowland, Ph.D. Physics R. J. Piersol, Ph.D. Donald O. Holland, M.S. Paul F. Elarde, B.S. Jack Tuttle GEOCHEMISTRY Frank H. Reed, Ph.D., Chief Chemist W. F. Bradley, Ph.D. G. C. Finger, Ph.D. Helen F. Austin, B.S. Fuels G. R. Yohe, Ph.D. Carl Harm an, B.S. Non-Fuels J. S. Machin, Ph.D. James F. Yanecek, M.S. Analytical O. W. Rees, Ph.D. George \Y. Land, B.Ed. P. \V. Heni.ine, B.S. Mathew Kalinowski, B.S. Arnold J. Veraguth, M.S. MINERAL ECONOMICS \Y. H Yoskuil, Ph.D., Mineral Economist Grace N Oliver, A.B. EDUCATIONAL EXTENSION Don L. Carroll, B.S. PUBLICATIONS AND RECORDS George E. Ekblaw, Ph.D. Chalmer L. Cooper, M.S. Dorothy Rose, B.S. Kathryn K. Dedman, M.A. Alma R. Sweeney, A.B. Frances Harper Lehde, M.S. Meredith M. Calkins Consultants: Ceramics, Cullen Warner Parmelee, M.S., D.Sc, Universitv of Illinois; Pleistocene Invertebrate Paleontology, Frank Collins Baker, B.S., Universitv of Illinois. Topographic Mapping in Cooperation with the United States Geological Survey. This Report is a Contribution of the Section of Geochemistry, Analytical Division. September 1, 1939 ILLINOIS STATE GEOLOGICAL SURVEY (A7270— 1M— 10-39) 3 3051 00005 6881 CONTENTS ■ PAGE Introduction 5 Acknowledgments ' 6 F.xperi mental work t 7 Apparatus " Procedure . . ,. i 8 Results 9 Discussion of results 9 Conclusions 23 Bibliography 28 Appendix A — Determination of the vapor pressures of the saturated salt solutions used in equilibration tests 29 Apparatus 29 Procedure 31 Calculations ^^ Discussion of results 34 Bibliography 34 TABLES PAGE 1. Equilibration data tor first six samples studied 10 2. Analyses and moisture values obtained by different procedures for first six samples studied 12 3. Analyses of fifteen samples used for further equilibration studies 13 4. Data on samples equilibrated in evacuated desiccators in constant temperature 30 C. water bath 14 5. Data on samples equilibrated in atmosphere of nitrogen at atmospheric pressure in con- stant temperature (30 c C.) air bath 15 6. Moisture values obtained by different methods for fifteen samples 17 7. Residual moisture values after second equilibration ^determined* and per cent increase in weight of samples during second equilibration. Water bath — desiccators evacuated. 18 8. Residual moisture values after second equilibration 'determined) and per cent increase in weight ot samples during second equilibration. Air bath — nitrogen atmosphere — no evacuation 19 9. Moist mineral-matter-free calorific values as calculated using moisture values obtained by different methods 25 I. Appendix A — Experimental vapor pressure values 32 [3 ILLUSTRATIONS FIGURE PAGE 1. Air-hath thermostat 7 2. Water-bath thermostat 8 3. Moisture-humidity curves tor first six coals studied 11 4. Moisture-humidity curves — low-moisture samples, desiccators evacuated, water-hath thermostat 13 5. Moisture-humidity curves medium-moisture samples, desiccators evacuated, water- hath thermostat 16 6. Moisture-humidity curves high-moisture samples, desiccators evacuated, water-bath thermostat 20 7. Moisture-humidity curves — low-moisture samples, nitrogen atmosphere 21 8. Moisture-humidity curves — medium-moisture samples, nitrogen atmosphere 22 9. Moisture-humidity curves- high-moisture samples, nitrogen atmosphere 24 A. Diagrammatic sketch of manometer assembly, side view 30 H. Diagrammatic sketch of manometer assembly, front view 30 C. Diagrammatic sketch of manometer, scale, and alidade, side view 31 I). Manometer assembly 32 E. Diagrammatic sketch showing the geometrical relationship involved in the calculations. ^>^ [4] A STUDY OF THE EQUILIBRATION METHOD OF DETERMINING MOISTURE IN COAL FOR CLASSIFICATION BY RANK (). W. Rees 1 , F. II. Reed 2 , and G. \V. Land 3 INTRODUCTION The establishment, in 1936, of the ten- tative specifications for" Classification of Coal by Rank" (l) 4 and the advance of these specifications to standard in 1938 (2), with the provision that the lower rank bituminous coals be classified according to their moist mineral-matter-free B.t.u. values, made imperative a reliable method for determining true bed moisture. Most of the methods employed heretofore have dealt with the empirical determination of the amount of moisture in the coal sample as presented at the laboratory without consideration of possible excess surface moisture or of moisture lost during sampling. Thus values have resulted which may or may not represent the bed moisture, which we consider to be the moisture content of the coal bed as it occurs naturally under the conditions of temperature and humidity in the mine. The equilibration procedure attempts to determine the moisture in the coal at 100 per cent humidity, which is assumed to be the true bed moisture. Such a pro- cedure has been used in various connec- tions but of chief interest here was its use by Lavine and others (3, 4) in studies on peat and lignite and by Stansfield and Gilbart (5) in studies on coal. At the meeting of the Coal Classifica- tion Committee of the American Institute of Mining and Metallurgical Engineers in Chemist and Head, Analytical Division. 2 Chief Chemist, Geochemical Section. ■'Research Assistant, Analytical Division. 'Numbers in parentheses refer to bibliography at end of report. New York in 1936 it was proposed that the equilibration procedure, as outlined by Stansfield and his coworkers, be made the standard procedure for determining moisture in coals having visible surface moisture for the classification of North American coals by rank. It was the opinion, however, that the reliability and convenience of the method should be further verified before it be accepted as standard. Because adoption of such a pro- cedure as standard might invalidate mois- ture values obtained in the past by other methods if the values were greatly differ- ent, it therefore seemed desirable to secure information concerning the reliability of values obtained by the equilibration pro- cedure when applied to high-moisture coals, such as are common in Illinois, and the relationship of equilibration values to values obtained by the present stand- ard procedure. This work was undertaken to provide such information. In this paper are presented the results obtained by applying the equilibration procedure to 21 Illinois coals ranging in moisture content from 4 to 18 per cent, as well as values obtained by other pro- cedures, with which comparisons are made. A brief review of the development of the equilibration procedure first proposed by Stansfield and Gilbart in 1932 (5) follows. Dissatisfied with the usual A.S.T.M. method for air-drying coal they attempted to devise a test which would serve the dual purpose of preparing the sample for laboratory handling and for evaluating the moisture-holding capacity of the coal. [5] EQUILIBRATION METHOD OE Their first procedure, developed in 1907 (6), required the exposure of crushed coal in shallow trays in a box which also held trays containing a solution of calcium chloride of 1.30 sp. gr. At ordinary tem- peratures this solution has a vapor pres- sure about 60 per cent that of water at the same temperature. The coal was weighed from time to time until a mini- mum weight was recorded. As this was a prolonged procedure, requiring six weeks in extreme cases, the method was aban- doned in 1910. The second method, de- veloped in 1923 (7), used an air-drying apparatus which Stansfield and Gilbart constructed in their laboratory. Crushed coal was dried in a rapid stream of air of 60 per cent humidity. The humidity was controlled by causing the circulating air to pass up a tube in which calcium chloride solution ran down a number of lamp wicks. By this method 48 hours were sufficient for the attainment of practical equilibrium. This apparatus was improved from time to time until 1930, but meanwhile it was found that temperature control was essential if the accuracy required for classification pur- poses was to be attained. Therefore, early in 1931 a new apparatus was con- structed (8), in which a uniform tem- perature of 30° C. could be maintained and natural gas of 60 per cent humidity could be circulated. Equilibrium was reached in this new equipment within 48 hours. The moisture value obtained after air-drying at 60 per cent humidity by the method developed in 1923 was adopted for coal classification by the Canadian Department of Customs and Excise (9). In 1931 Stansfield and Gil- bart developed a vacuum-desiccator pro- cedure for drying at constant temper- ature at various definite humidities (5). Their procedure consisted in equilibrating different portions of coal at different re- lative humidities in desiccators at a tem- perature of 30° C. Various saturated salt solutions were used for the corresponding relative humidities as shown in appendix A, table 1. Residual moisture in the equilibrated samples was determined by heating at 105° C. for 3 hours in a vacuum oven in which an inert atmosphere of natural gas was maintained at an absolute pres- sure of about 3 inches of mercury. These moisture values so obtained were plotted against relative humidity values and the curves were extrapolated to cut the 100 per cent humidity axis. The moisture values corresponding to 100 per cent humidity were taken as the "true" or "capacity" moisture values of the coal. Some work was done in which these authors attempted to obtain a satisfac- tory moisture value by equilibrating at only one relative humidity. In the earlier work 60 per cent humidity was used but erratic results were obtained so that in later work a relative humidity of 97 per cent was used. Moisture values obtained at a single humidity level were used only after one or more entire moisture humid- ity curves for the various coal areas in the province had been obtained. A study of 54 entire humidity curves of coals with moisture content ranging from 1 to 32 per cent showed that the moisture re- tained at 97 per cent humidity averaged 98.6 per cent of the extrapolated "true" moisture value and seldom varied far from this value. They therefore calcu- lated "true" moisture by dividing the moisture retained at 97 per cent humidity by 0.986 in all cases where they did not prepare the entire curve. These values were termed moisture "by calculation." In some cases the moisture humidity curves could be extrapolated easily where- as in other cases irregular curves were obtained, the extrapolation of which was impossible. A distinct difference between dehydration and rehydration curves was found. Acknowledgments The writers wish to acknowledge the cooperation and assistance of the manage- ment of the mines in collecting samples. To Dr. G. H. Cady they express their appreciation for many helpful suggestions during the preparation of this report. The help of Dr. L. C. McCabe and C. C. Boley in securing samples, and of J. W. Robinson, P. E. Grotts and M. L. Kalinowski in securing analytical data is gratefully acknowledged. To Dr. M. M. Leighton, Chief of the Survey, is due the credit for making this investigation possible. DETERMINING MOISTURE IX COAL Fig. 1. — Air-bath th ermostat. EXPERIMENTAL WORK The experimental work of this report was an application of the equilibration method of Stansfield and Gilbart for the determination of bed moisture of Illinois coal. The procedure adopted and the equipment used for the first six coals studied were as nearly as possible those outlined by Stansfield and Gilbart. Mois- ture determinations for these six samples of coal were made, using a large air oven to provide constant temperature during equilibration. Desiccators were evacuated every twelve hours. The discrepancies in the results obtained indicated inadequate temperature control. Therefore, a con- stant-temperature water-bath was sub- stituted for this air oven and equilibra- tions on 15 other coal samples were made. Again, desiccators were evacuated every twelve hours. In addition, equilibrations were made at three or four humidities on these 15 samples using a nitrogen atmos- phere without evacuation. Below is a description of the apparatus and an out- line of the procedure followed in the work. Apparatus A double-walled box approximately 72 by 26 by 24 inches, outside dimen- sions, was built to serve as a constant temperature oven for the equilibrations. The walls of this box were insulated with about three inches of rock wool, heat was provided by light bulbs so placed that air trom a fan was heated at one end of the box, passed between the upper walls to the other end and was admitted to the chamber proper through a plate in which many holes were drilled to give good dis- tribution. A cooling coil, through which cool water was passed continuously, was provided so that the heating bulbs worked against this cooling arrangement. A mercury thermo-regulator was used to control the temperature of the box at 30° C. ± 0.5° C. The box (fig. 1) was equipped with a side door over its entire length for convenience in putting in and taking out desiccators. Later equilibra- tions were made in a large wafer-bath thermostat (fig. 2) at 30° C. + 0.1° C. Pyrex vacuum desiccators as shown in the oven (fig. 1) were used for the equili- EQUILIBRATION METHOD Of \ '4 « ^fe> V Ji K iJisapj!^! *^: I y^ / Fig. 2. — Water-bath thermostat. brations. For determining moisture after equilibration a Cenco vacuum oven was used. Small petri dishes were used to hold the samples during equilibration. Lids were provided for these dishes to protect the samples during weighing and transfer from one container to another. Bottles containing saturated solutions of the same salts as those used in the desiccators were used to provide properly humidified air to desiccators in bringing the pressure to atmospheric pressure at the end of the equilibration periods. Procedure In studying the application of the Stansfield and Gilbart method to the de- termination of moisture in Illinois coals, it was decided to use fresh face-samples of low, medium, and high moisture con- tent. Accordingly, six samples of coal were collected at intervals of about one week and each treated as follows: 1) A sample was cut down from the face, crushed and ground to J4 inch size, riffled to one quart in the mine and brought to the laboratory where it was further crushed to minus 14-mesh, and portions were weighed and equilibrated for 48 hours at each of the nine different humidities recommended by Stansfield and Gilbart. After equilibration, the samples were reweighed and then dried at 105° C. in the vacuum oven for three hours. The pressure in the oven was maintained at about three inches of mercury by ad- mitting a slow stream of nitrogen. At the end of the three-hour period the samples were removed from the oven, cooled, weighed, and the moisture values calcu- lated. The samples of coal were then re- equilibrated for 48 hours at the same humidities after which moisture values were determined again. Both first and second equilibration moisture-values were plotted against humidity, the curves were extrapolated to cut the 100 per cent humidity axis, and the moisture value obtained was taken as the "true" mois- ture value of the sample. DETERMINING MOISTURE IX CO A I. 2) Another sample was collected ex- actly the same as that for 1) for a proxi- mate analysis. On this sample air dry loss and the regular A.S.T.M. (10) mois- ture were determined to give total mois- ture. 3) A 50 to 75 lb. gross sample was taken from the freshly exposed face and brought to the laboratory in a sealed container, where it was crushed to }/i inch size and riffled down to a quart sample. Moisture was then determined by the usual A.S.T.M. procedure includ- ing air dry loss, etc. 4) Two 20-mesh samples (approxi- mately 5 grams each) were prepared in the mine and placed in weighing bottles which had been previously weighed in the laboratory. These samples were brought to the laboratory where they were weighed and total moisture values were determined by the A.S.T.M. procedure (10). 5) Two 20-mesh samples were pre- pared in the mine exactly as in 4) and these samples were used for total moisture determinations in the vacuum oven fol- lowing the same procedure used for de- termining moisture in the equilibrated samples. In all cases except 3) duplicate por- tions were taken as a precaution and duplicate determinations were made in all cases. Later, fifteen additional coal samples were obtained and treated by the same procedure as outlined above, with the exception that a water-bath thermo- stat was used for the equilibrations in place of the air-bath thermostat. In ad- dition, portions of these samples were equilibrated at three or four humidities in a nitrogen atmosphere without evacua- tion in the air-bath thermostat. Results Equilibration results for the first six coals studied are given in table 1 and shown graphically in fig. 3. Table 2 pre- sents proximate analyses of these coals together with a comparison of moisture results obtained by different procedures. Tables 3, 4, 6, and 7 and figures 4, 5, and 6 present data for the 15 samples equili- brated in the water thermostat. Tables 3, 5, 6, and 8 and figures 7, 8, and 9 pre- sent data for the 15 samples equilibrated in nitrogen atmosphere, with no evacu- ation, in the air thermostat. DISCUSSION OF RESULTS Six coals were first equilibrated at nine different humidities, and the moisture content of each sample was determined in the vacuum oven. These values are listed in table 1 as "Residual moisture in per cent after first equilibration." The dry samples were then re-equilibrated and moisture values were determined again. These are listed in table 1 as "Residual moisture in per cent after second equili- bration." A comparison of equilibration values shows that they are in close agree- ment in the low humidities, diverge con- siderably at the intermediate humidities, and converge again at the high humidi- ties. The values obtained for the second equilibration are, in general, lower than those of the first equilibration, the varia- tion being greatest for the intermediate humidities. Extrapolation of the moisture-humid- ity curves to give 100 per cent humidity values gave results of questionable valid- ity. In the case of sample C-1901A (fig. 3), the slope of the curve at the high humidities was so steep that extrapolation gave a result which appeared question- able. In the case of sample C-1864A (fig. 3), strict adherence to all values gave a curve proceeding downward at the point of crossing the 100 per cent humid- ity axis. While the moisture value for the sample equilibrated at 97.7 per cent humidity was lower than that for the 96.8 per cent humidity by an amount within experimental error, it leaves one in doubt as to which way to extend the curve to cut the 100 per cent humidity axis. Data obtained on the second equilibration series on samples C-1901A and C-1904A (fig. 3) gave such steep curves at the higher humidities that values obtained by their extrapolation seem questionable. Attempts to draw smooth curves through the points as plotted were un- successful, as wavy curves resulted which were not easily extrapolated. Therefore, all points were connected by straight lines. The slope of the curve as extra- polated to 100 per cent humidity was influenced little if any by the low or intermediate humidity values, but was determined by the last two high humidity moisture values in most cases. 10 EOUI LIB RATIOS METHOD Ol rvj o 3 NTtNH0MDHr-lr-lO 1/ 5O' u ^ioOh\OnnnoO»oo > C/} w ^ ^ oiNOOOHioa(NO«)ir,ao r— oaacocM^ooaooaaaa u. -1 o c < i^- o On 2 w <; £ ^ OOOCCONNH(NOX-tHOO 00 © io O O © f> *— i *— I'tinoONOi o oococaoMCM^^aax O o o s On rH ,— ' ,_H ^ H ,—l f^^TtONO'-'O^aotio o 't'tocot^ooao^oooc ,_,,_,_, ,-h rtrlHHHrl ^ fOOfOONM-HNOONM r— ^t^OONOOr^-OOOoot^ ON ^H ^H ,_, _ ,-H ^ _ _ ,_ _ 00 roMTtONNO'tMOOO no ■^rooOOOOiOONONOOt^ o> HHrtH *— i i—t ^h»-< H Z o nO On *— lOfO- ifONONfOO W ON rOHO^OOO^OOONi*) U 6, 00 *-H T— 1 *— 1 *— 1 ■<— I »— l i— 1 ID OOOOOrO'Hi^aO'+OO'tt^ CA 00 rONOM»\OiOiOOaOvOO w -J *H oOOO^OOOLOfO-HONOO^O to 1^ ^Tj.ooi^'tTtvOOO't'tO 5 t^- •t rn rt Qv O (N io aa N ro O0 5 ON Of^OOf^'OfOl^f^OfONfO 5 u-) PC oo TfvO'H'tOOOO^t^OOfn nO -f(NOfOrOM'OfC , tM-tr / ; -t o © © I fOCN-+' v l't , -'-tO MM(NMTHr-iM(N(NM(NfN 'tfNO' l OfO'NI^lON'tt^'f 1—1 '5 •Jl o u _3 Ph co 3C oo X <<;<<<< a. ro^O^^-t a/) ox. 5 o nOnOOOOO rtrt 3* 00 00 On On On QN i-, »-. »-H l— t »— 1 *— 1 >"H tH > L L u u U U •*" < c c o o oi a! U u J3.D 3 3 cr cr ■ C/3 W 0* l_ h * o iV *3 O Uh Q fe < a -+ o o o o. OC w ta ^onoi-hoo 2: Ol t-h -t ,-n pq Q MMfOMMfN ^00^010 j On ^-h ~h O O PO «s NCtO>00 O "-i C-l On *-i On t-H t— ( y—t i— 1 >. "Ttt^t^(Nrt a: Q -t-tM-t-t-rf 3 — ^ On On LO 00 t^- <"0 C/) < g In r*5 f> (N C5 ro f5 z c t^ 00 ro r^ sO On 03 rj< t^. Tf nO rt 1 no rf Tf lO ^ Tf rf< Q -a -hNNOOM X < u OOMOXON to J- ro -rfi i/~i r -+ -+ -+ c 3 CN O <"0 LO ^h t- i > < £ N\O'*'*00(N u f) f) f5 f5 fO f) ^_ 00 O O On -f NO a *-h <^1 On »H CN CN I < ^ % , H00-t00OO < £ ooooa^o V- _ _| ^H ^H O U CN o lc NO NO NO c 3 o U . C- (y ~p Tf O *-i .2^ S ^^^ SCO H^rr:- ^3 o b ^ > > b£ ^^ 03 iy • 3 - J h^ U5 M o O • . 2 w ' "" tJ T3 r/ -, fiofl Eg si " id + - > a) a) C3 .5 3T3t3 ^ ^ O ~Au,aj .b .b u - c aj .— . ri; A- ,c ^ 3 to 0.0,^^ a; O <^ DETERMINING MOISTURE IS COAL 13 20 30 40 50 60 70 RELATIVE HUMIDITY - PER CENT DC Fig. 4. — Moisture-humidity curves for low-moisture samples, desiccators evacuated, water-bath thermostat. Table 3. — Analyses of Fifteen- Samples Used for Further Equilibration- Studies T -ill Coal Ash Volatile Fixed Carbon Sulfur B.T.U. \'n County bed No. As rec'd Dry As rec'd Dry As rec'd Dry As rec'd Dry As rec'd Dry Unit coal C-2009 Vermilion .... 7 10.1 11.7 35.9 41.4 40.6 46.9 3.10 3.58 10.981 12.675 14.634 C-2010 La Salle 2 9.0 10.5 37.0 43.2 39.8 46.3 3.04 3.54 11.085 12.920 14.691 C-2022 Loaan 5 11.4 13.3 34.0 39.7 40.2 47.0 2.843.32 10.571 12.344 14.530 C-2023 Saline 5 9.3 10.1 32.8 35.5 50.2 54.4 2. 98|3.22 12. 106 13.113 14.833 C-2030 Gallatin 5 11.1 11.5 36.7 38.2 48.4 50.3 3.39 3.52 12, 640|13. 13645. 133 C-2031 St. Clair 6 11.6 13.0 37.0 41.5 40.5 45.5 3.69 4 14 11.005 12.355 14.517 C-2034 Franklin 6 8.4 9.3 33.1 36.6 49.0 54.1 1.16 1.28 11.804 13.038 14.535 C-2037 Christian 6 11.2 12.9 35.9 41.3 39.9 45.8 4.48 5.15 10.624 12.205 14.353 C-2039 Marion 6 9.8 11.3 34.9 39.9 42.8 48.8 3.31 3.79 11.000 12.572 14.444 C-2042 Randolph .... 6 10.7 12.0 35.1 39.4 43.3 48.6 2.80 3.14 11.101 12.462 14.424 C-^046 1 Knox 6 9.1 11 .1 32.2 39.2 40.7 49.7 2.953.60 10.248 12.492 14.311 C-2059 1 La Salle (*) 6.5 7.7 35.1 41.5 43.1 50.8 2.15 2.54 11.355 13.411 14.714 C-2064 Gallatin 5 9.0 9.4 36.7 38.4 49.8 52.2 3.193.34 12.736 13.330 14.956 C-2069 Saline 5 11.5 12.1 35.0 36.7 48.8 51.2 3.30 3.46 12.248 12.846 14.904 C-2080 Saline 5 12.3 12.9 36.2 38.0 46.7 49.1 3.49 3.67 12.101 12.706 14.899 'Strip mine. 2 Loeal below coal No. 6. 14 EQUILIBRATION METHOD OE Q -^roNOiOPOOoONOON ^H ,-H rf CN 'HiOfO'tsOfO'HOOOOtNOOlO'tiO'tM ^_ W w 3 fOfOi*i , iO^\OOMfnoOOOfOfOi , '*'H'-i«)oOioi"tT)'Tt"tT)"* H s* cs 3 » S O > CO Nir,MO00O(Nr0OOtOM^-tT|"rtio00>Off)fNl00i , T) n© icaO , ONOOO f 0^ f O^OOO>XN'HrtiooOfOO i OOOO r t , 0'- , ON p r- O^ J O OnOnOOnOnOnOOOnOnOnOnOOOnOnOOOnOnOnOnOnOnOnOnOOOOOOnOnOn r^ O o > H S3 On '" H r ~ ' »— 1 1— 1 ^H ^ _ _ ,_ ^ ^ oaNfoooooa'HfooON^oiNOooiNtN^ofoooooNON r- ON^ioOOOfONiooOOioOMONioOONfOOOfOOnioio o OOOOOOOOO^OOOOOOCOOOOOOOOOOOOOOO O v© o § ON *"■ ,— ' i— 1 *— 1 T— 1 1— 1 «— 1 *— 1 *-H *— 1 6 o ^rHO'- | tNa i OfO'HfO^OMNf^i"H(Na' H, Of^O>'*l^(NfONO'-l rOfO'* i'tfONiOO'- | !»a'*i , 't'*'*'*^'* ,— 1 ,— 1 ,-H ,— 1 ,— 1 ,— 1 ^H -^ ,-|*~Iv-Hi-I«-ItHi-H1— l*HvH r^ aNOO^N^lOfOOOONOXNONMOOOOOOOlN^fNfOtNO^ O (N(NTf f^'t'tOOtNfC ^00f0l^'*^ l ^(NfNifO^O00 vO rfrO'*rt\OOfOfOOOOO(NfNlrHfOOONNi' ^ rf< rf ^t< Tf t*i ro z ON _ ^H _ llH HH ^^___<^H,— |__,-H ^H ^rtrHCNNa^t^OOOO'HfOriOOOl^NfO't'tN'tNOMWOOOOOl'; u a! W O o- (NTHf^Mf^(NOlO(N(NaOvO^^'- |rt '- | '-'00\NNrOOt^fO'tfCfOfO <* iOi^'tOOHNOO!NiOir,ir / rHO't't^(N(NCfCOOONMi.|5,oO(N -V rHNrnoofno'o^tNfNoot^^oO'HO'-iOOoooofnNfOfrifnfo^fo < 00 ___,_, t _| t H tHt H.^H t H,_ I ,H -* 0'-i-tioOOOfOiON'0 0'1'NO"t | OOMOO f OO> ( *! H 00»NO"tt N • O O "O O iO ^H^ ir ; -t M M 00 "O 00 O O '^ O^N O^ O iO vO >i iO f^ N fO f*5 f3 M oo 00 r*~, • 0' / )OHMooo-to i ooa | OHOoNfO'- | -tO f O' H -t , HfNi ,"*■ On O ro • 00 -h -t ro r>i ,-H sO J h- c R •r- R 1 -5 | "S c &H U S Ph ^ _0J C 2 a JS (L c ; ; I '. '. '. '. '. « '• us ' « > • • i ONO<» r OO'- H '+ t ^ONfNNOON'*ONO OJ 1 K - 1 5 6 0»-ir^r^fO> rt DETERMINING MOISTURE IX COAL 15 Table 5. — Data on Samples Equilibrated in Atmosphere of Nitrogen* at Atmospheric Pressure in Constant Temperature (30° C.) Air Bath Lab. No. C 2009 C-2010 C-202 2 C-2023 C-2030 C-2031 O2034 C-2037 C-2039 C-2042 C-2046^ C-2059 4 C-2064 C-2069 C-2080 Average Average \'ermilion La Salle . Logan Saline. . . . Gallatin . St. Clair Franklin. Christian Marion Randolph. Knox. . . . La Salle. . Gallatin. . Saline. . . . Saline. . . . Coal bed No. Equili- bration C 1 ) Humidity values (per cent) 11.2 1.6 1.4 1.6 1.4 1.7 1.5 1.6 1.3 .9 .8 1.9 1.3 2.0 1.7 1.9 1.6 2.0 1.6 1.9 1.5 1.9 1.7 1.1 1.0 1.0 .9 59.8 7.1 3.4 2.2 1.7 7.9 4.7 8.8 4.9 9.4 5.2 8.4 4.1 9.2 4.2 7.8 3.4 3.2 2.4 3.2 2.3 3.0 2.1 84.4 11.2 7.2 13.2 7.8 13.4 9.6 6.0 5.1 2.7 2.4 9.4 6.8 9.0 8.4 11.6 9.6 11.4 10.1 10.0 8.3 16.4 9.5 13.3 7.1 3.8 3.3 3.S 3.3 3.5 3.0 97.7 13.0 12.5 14.2 13.4 14.4 14.1 6.5 6.2 3.0 3.0 10.7 9.9 10.1 9.9 13.2 12.7 13.7 12.9 10.8 10. 17. 16. 14. 13. 4. 4.1 4.2 4.1 3.9 3.7 100^ 13.3 13.3 14.4 14.3 14.6 14.8 6.6 6.4 3.1 3.1 10.9 10.4 10.3 10.1 13.5 13.2 14.1 Ratio of moisture values 97.7% 100% 13.4 10.9 10.9 17.8 17.6 14.7 14.3 4.3 4.2 4.3 4.2 4.0 3.8 .977 940 .986 937 986 953 985 969 968 968 982 952 981 980 978 962 972 963 991 963 989 938 986 930 977 976 977 976 .975 .974 .981 959 Calcu- lated moisture value 3 13.3 13.0 14.5 14.0 14.7 14.7 6.6 6.5 3.1 3.1 10.9 10.3 10.3 10.3 13.5 13.2 14.0 13.5 11.0 10.9 17.9 17.2 14.8 13.9 4.3 4.3 4.3 4.3 4.0 3.9 l A Residual moisture after 1st equilibration. B Residual moisture after 2nd equilibration. 2 By extrapolation. 3 Factors 0.981 and 0.959 used respectively for 1st and 2nd equilibration values. 4 Strip mine. s Local below coal Xo. 6. In table 2 are tabulated moisture values for these six samples obtained by six procedures including first and second equilibrations on each sample. These values, in general, check very well with the exception that the second equilibra- tion value for sample C-1903A is higher than the other values and both the first and second equilibration values for sample C-1904A are lower than the other values. The first and second equilibration values check reasonably well in four of the six samples but for samples C-1901A and C-1903A there is considerable variation. The difficulties which arose in applying the equilibration procedure and inter- preting the results on the first six coats led us to study the vapor pressures of the solutions in desiccators used lor equili- bration. Vapor pressures were determined lor each solution as used in the desiccator and then relative humidity values were calculated. A description of apparatus and procedure used in these determina- tions is presented in appendix A. Infor- mation was obtained in the course ol these determinations which has a distinct bearing on the equilibration procedure. EQUILIBRATION METHOD OF 30 40 50 60 70 RELATIVE HUMIDITY - PER CENT 100 Fig. 5.— Moisture-humidity curves for medium-moisture samples, desiccators evacuated, water-hath thermostat. DETERMINING MOISTURE IX COAL 17 Table 6.- -Moisture Values Obtained by Different Methods for Fifteen Samples Lab. No. County Coal bed No. Moisture values obtained by different methods 1 (per cent) Mine humidity (per cent) 2 A B C D E F G H C-2009 C-2010 C-2022 C-2023 C-2030 C-2031 C-2034 C-2037 C-2039 C-2042 C-2046 4 C-2059 4 C-2064 C-2069 C-2080 Vermilion. . . . La Salle Logan Saline Gallatin St. Clair Franklin Christian Marion Randolph. . . . Knox La Salle Gallatin Saline Saline 7 2 5 5 5 6 6 6 6 6 6 ( 6 ) 5 5 5 13.3 14.3 14.4 7.7 3.8 10.9 9.5 13.0 12.5 10.9 18.0 15.1 5.0 4.8 5.0 12.6 13.1 13.8 7.7 3.7 10.4 9.4 12.0 12.3 10.2 17.9 15.1 5.0 4.8 5.0 13.5 14.4 14.6 7.3 3.1 ' 9A 12.9 12.5 3 11.1 17.6 14.5 4.8 4.7 4.8 14.0 14.9 15.1 8.1 3.7 11.5 10.0 13.4 13. I 3 11.6 18.8 15.1 5.1 5.1 5.1 13.1 14.0 15.2 6.5 3.1 10.4 9.8 14.3 17.1 10.9 18.6 14.9 4.5 4.3 4.0 13.1 14.1 14.9 6.3 3.3 10.6 9.7 13.4 15.2 11.1 19.3 14.4 4.2 4.7 4.1 13.3 14.4 14.6 6.6 3.1 10.9 10.3 13.5 14.1 10.9 17.8 14.7 4.3 4.3 4.0 13.3 14.3 14:8 6.4 3.1 10.4 10.1 13.2 13.4 10.9 17.6 14.3 4.2 4.2 3.8 95 96 94 97 92 95 97 97 93 92 88 5 48 : - 97 96 96 J A Sample ground to J^-inch. in mine. Total moisture by Air Dry Loss + Regular Moisture. B 50-lb. mine sample crushed to 5^-inch in laboratory. Total moisture by Air Dry Loss + Regular Moisture. C 5-gram sample crushed to 20 mesh in mine. Moisture by A.S.T.M. oven, air atmosphere, 105° C. \ X A hours. D 5-gram sample crushed to 20 mesh in mine. Moisture by vacuum oven 3 inches Hg., X 2 atmosphere, 105° C., 3 hours. E 100 per cent humidity extrapolated moisture values — -1st equilibration, evacuated. F 100 per cent humidity extrapolated moisture values — -2nd equilibration, evacuated. G 100 per cent humidity extrapolated moisture values — 1st equilibration, nitrogen atmosphere. H 100 per cent humidity extrapolated moisture values — 2nd equilibration, nitrogen atmosphere, determined with sling psychrometer. ^Crushed to 20 mesh in mine; brought to laboratory in large sample bottle and transferred to weighing bottle. 4 Strip mine. Outdoor humidity. "Local below coal No. 6. An attempt to use the original air oven was not successful since it was impossible to hold the temperature of the desiccators sufficiently close to 30° C. by this means. It was therefore necessary to use a water- bath thermostatically controlled at 30° C. varying not more than + 0.1° C. Evacuation of the desiccators lowers the temperature 5° C. or more and in the air oven, several hours are necessary to regain the proper temperature. Such lowering of temperature with slow return to proper temperature, is important both in making vapor pressure determinations, and in the actual equilibration of coal samples. Lowering of the temperature by evacuation results in condensation of moisture on the coal sample and unless it is left until equilibrium is regained, the moisture of the sample will be too high. This is particularly important at the higher humidities. The use of a water- bath thermostat was found to effect re- turn to the desired temperature within a short time (about one-half hour) after evacuation and proved satisfactory in the vapor pressure determinations. The humiditv values calculated from the de- termined vapor pressure values checked closely with those used by Stansfield and Gilbart. The humidity values used by them and those determined in this labor- atory are compared in table I, appendix A. Having proved to our satisfaction that the relative humidities in the desiccators used were satisfactory, provided there was adequate temperature control, ad- ditional coal samples were studied. Fif- teen samples representing low, medium, and high moisture coals of the State were obtained and treated as outlined above. Analyses of these samples are shown in table 3. The samples were equilibrated for 48 hours in desiccators placed in a water-bath thermostat, with evacuation every 12 hours. Results of these determi- nations are tabulated in table 4 and shown graphically in figures 4, 5 and 6. As stated above, it was learned that evacuation lowered the temperature in the desiccators and that the return to the desired temperature was slow in an air-bath thermostat. It was thought that elimination of evacuation and the use of an inert atmosphere at normal pressure 18 EQUILIBRATION METHOD 01' V c w h H < A UJ u U <; hi W Ph 1/3 Bj Q O y. h < < u Q o uj I« X w Q <* 1 o w r -1 O c h UJ < « Q cc ft O o UJ C/3 w « UJ Q H /". Ph O < u UJ uj X 2 J ?; < > * uj Q <5 <; ■j) <• U. ;_; O Q h + N ir; o i^ 00 MMfOIN^rO^fTl OOONOOiOOMfT) M^^MrtfOrOfO ro<^ro r^^i -M— O ■ '-' O l^ On ^ -* On ~+ (N O 'i O On *t OO ■ h (N M) O r 1"0 i 1 't^-t'00'ta'OooNi'OOo t>>Mi/)i/)ir ( 000000^ iO"T^ JHU^ c ajrto"3 rt 4-;^- r: ^H rt 1 rro« ra K! 0\OM' y )0'- ,, tt^ONfNOON , t^O O'HMfNfOfOfOfO'^'+'+'OOOOO ooooooooooooooo (N(N(N(^l(N(N(N(N(N(NrN(NfN(NM I I I I I I I I I I I I I I I uuuuuuuuuuuuuuu u ft S 2 X ^«io DETERMINING MOISTURE IN COAL 19 Table 8. — Residual Moisture Values After Second Equilibration (determined) and Per Cent Increase in Weight of Samples During Second Equilibration. Air Rath — Nitrogen Atmosphere— No Evacuation Lab. No. Count 1 Coal bed No. Humidity values (per ceni I 11.2 Wt. 1 R.M. 2 59.8 Wt. 1 R.M 84.4 97.7 Wt. 1 R.M. 2 Wt R.M 2009 2010 C-2022 C-2023 C-2030 C-2031 C-2034 C-2037 C-2039 C-2042 C-2046 3 C-2059 3 C-2064 C-2069 C-2080 Vermilion La Salle . Logan Saline. . . . Gallatin . St. Clair Franklin . Christian Marion Randolph Knox. . . . La Salle. . Gallatin . Saline . . . Saline. . . . 7 2 5 5 5 6 6 6 6 6 6 (') 5 5 5 1 1 1 1 1 1 1.4 1.0 1.0 .91 1.4 1.4 1.5 1.3 .77 3 7 6 6 5 7 3.3 3.4 1.7 1.4 1.1 1.0 .90 2.4 2.1 4.7 4.9 5.2 4.1 4.3 3.4 2.3 2.3 2.1 7.2 7.7 9.4 5.3 2.4 6.8 8.5 9.6 9.8 8.2 9.3 7.0 3.3 3.3 3.0 7.2 7.8 9.6 5.1 2.4 6.8 8.4 9.6 10.1 8.3 9.5 7.1 3.3 3.3 3.0 12.3 13.1 13.9 6.0 2.6 9.9 9.8 11.8 12.8 10.6 16.5 13.4 4.1 3.9 3.8 12.5 13.4 14.1 6.2 3.0 9.9 9.9 12.7 12.9 10.5 16.5 13.3 4 1 4. 1 3.7 ^Veight increase, per cent. -Residual moisture, per cent. :i Strip mine. 4 Local below coal Xo. 6. within the desiccators might eliminate this difficulty. Accordingly, portions of the 15 samples were equilibrated at three or four humidities in a nitrogen atmos- phere, using the air-bath thermostat. The results obtained are tabulated in table 5 and shown graphically in figures 7, 8, and 9. Results obtained in this way are similar to those obtained in the former series of determinations. Table 6 presents a comparison of moisture values obtained by various procedures for these coals. As a means of checking whether the vacuum oven procedure for determining moisture accounted, within experimental error, for the increase in weight of samples during second equilibration, a comparison was made of vacuum oven moisture values with the per cent increase in weight of the dry samples during the second equili- bration. The comparison of these values for the 15 samples equilibrated in the water-bath thermostat is shown in table 7 and for those equilibrated in the nitro- gen atmosphere, air-bath thermostat, in table 8. Reference to these tables will show that the vacuum oven procedure accounts, within experimental error, for all moisture taken up by the samples dur- ing equilibration. Figures 4, 5, and 6, in which are plotted the moisture-humidity data for the 15 samples of coal equilibrated in the water- bath thermostat, show irregular curves similar to those obtained for the first six coals. These curves are difficult to extra- polate, as they are inclined very steeply upward toward the 100 per cent line in some instances and downward in others, leaving doubt as to their proper direction. The irregularities are greater than the allowable limits of deviation in the pro- cedures used, which makes it impossible to draw smooth curves through the points. The low and intermediate humid- ity values have little or no influence on the slope of the curve as extrapolated, for this slope is determined by the last two high humidity values. In figures 7, 8, and 9 are plotted the moisture-humidity data for the samples equilibrated in a nitrogen atmosphere. These curves appear smooth- er than those of figures 4, 5, and 6 be- cause fewer points were plotted. How- ever, here again the slope of the curve as extrapolated to cut the 100 per cent hu- midity axis is determined by the last two points. It would seem, therefore, that equilibrations at low and intermediate humidities are entirely useless for extra- polation. Curves whose extrapolation 20 EQUILIBRATION METHOD OF 30 4 50 RELATIVE HUMIDITY 60 70 PER CENT Fig. 6. — Moisture-humidity curves for high-moisture samples, desiccators evacuated, water-hath thermostat. DETERMINING MOISTURE IN COAL 21 -^ ***i - ****** ^*^ °^ a«**J>' ^o& '^\-""^ >" „ c -^. - - - r-^ -. ^__ o^ , <^ i^S 2? .• 5^^ _-•—--" -^ "^ o— — 10 20 30 40 50 60 70 RELATIVE HUMIDITY - PER CENT 80 90 100 Fig. 7. — Moisture-humidity curves tor low-moisture samples, nitrogen atmosphere. slope is determined by only two or possi- bly three points would seem to be un- reliable. Stansfield and Gilbart (11) suggested that after characteristic curves had been run on many representative coals it was possible to caluclate an average ratio between the 97 per cent humidity mois- ture value and the "true" moisture value obtained by extrapolation. This ratio could be used in calculating true mois- ture values from the 97 per cent humidity moisture values. By such a procedure equilibration at only one humidity was necessary and the value so obtained was termed moisture "by calculation." The factor as determined by Stansfield and Gilbart for this calculation was 0.986. Similar factors have been calculated from the data in this report. The factors together with calculated moisture values are shown in tables 1, 4, and 5. Reference to tables 1 and 4 will show factors and calculated moisture values obtained from the retained moisture values at both 96.8 and 97.7 per cent humidities. The aver- age factor of first equilibration values at 96.8 per cent humidity for the first six samples studied (table 1) is 0.971 and for second equilibration values it is 0.898. Factors for 97.7 per cent values are 0.979 and 0.930 respectively for first and second equilibration data. Considerable deviation of individual values from the averages are apparent. In calculating moisture values, average factors for first equilibration values were used for first equilibration data and average factors for second equilibration values were used for second equilibration data. The mois- ture values calculated from 96.8 and 97.7 per cent humidity values check reason- ably well but appreciable deviations of these values from the 100 per cent ex- trapolated values are noted. Factors for the 15 samples of table 4 are 0.950 and 0.952 respectively for first and second equilibration data at 96.8 per cent humidity and 0.960 and 0.957 respectively for first and second equili- data at 97.7 per cent humidity. Again, in- dividual factors deviate appreciably from the average factor. Calculated moisture values for the two series check verv well 22 EQUILIBRATION METHOD OE 20 30 40 50 60 70 80 90 RELATIVE HUMIDITY - PER CENT Fig. 8. — Moisture-humidity curves for medium-moisture samples, nitrogen atmosphere. 100 DETERMINING MOISTURE IN COAL 23 but vary considerably from the 100 per cent extrapolated values in some cases. Factors tor results on samples equilibrated in a nitrogen atmosphere are available only for the 97.7 per cent humidity data. These are shown in table 5. Values ot 0.981 and 0.959 respectively were ob- tained tor first and second equilibration data. Calculated moisture values seem to check with the 100 per cent extrap- olated values somewhat better than do those of tables 1 and 4. In table 6 are tabulated moisture values tor the 15 samples obtained by different procedures including first and second equilibration values obtained using evac- uation and nitrogen atmosphere. In general the results obtained by various procedures seem to check reasonably well. For the most part method B, or total moisture by Air Dry loss -f- Regular Moisture, gave lowest results. Equili- bration results for samples C-2023, C- 2030, C-2064, C-2069, and C-2080 are, in general, lower than results obtained by other procedures but the equilibration re- sults for sample C-2039 are distinctly higher than the other results for this sample. As a test of the suitability of the mois- ture values obtained in this study tor use in the rank classification ot Illinois coals, moist mineral-matter-free B.t.u. values for all 21 coals studied were calculated on the basis of each different moisture value. These values, together with ash, sul- fur, and B.t.u. values reported on the "as received" basis according to each in- dividual moisture value, are shown in table 9. Reterence to this table will show devia- tions in moist mineral-matter-free B.t.u. values in only four of the 21 coals studied, which are in the critical range for classifi- cation. Critical moist mineral-matter-free values are 11,000, 13,000 and 14,000 B.t.u. (1, 2). On sample C-1901A, methods A, B, C, and E produced moisture values leading to calculated heat values greater than 11,000 B.t.u. whereas methods D and F gave values leading to calculated heat values of less than 11,000 B.t.u. On sample C-1904A, methods A, C, and D produced moisture values leading to cal- culated heat values less than 11,000 B.t.u. while methods B, E, and F gave moisture values leading to calculated heat values greater than 11,000 B.t.u. For coal C- 2034, methods A, B, and C gave moisture values resulting in calculated heat values greater than 13,000 B.t.u. but methods I), E, F, G, and H gave moisture values lead- ing to calculated heat values less than 13,000 B.t.u. On sample C-2080, all methods but one gave moisture values re- sulting in calculated heat values greater than 14,000 B.t.u. Method D gave a moisture value which resulted in a calcu- lated heat value less than 14,000 B.t.u. Coals C-1901A and C 1904A are border- line coals, and agglomerating and weath- ering characteristics would have to be con- sidered for classification before such de- viations would be controlling factors. In three of the four cases where the heat values are around critical dividing points, the deviations ot moist mineral-matter- free B.t.u. values are within experimental error and are therefore not significant. CONCLUSIONS The study ot the equilibration pro- cedure has led to the following conclu- sions: 1) The equilibration procedure as applied to the Illinois coals studied pro- duces data which, when plotted, give irregular curves. Satisfactory extrapola- tion of these curves is impossible. 2) Low and medium humidity val- ues do not appear to affect the slope of extrapolated curves. 3) Other procedures which are much simpler, but may involve slight modifica- tion of the standard procedure for samp- ling, appear to produce results suitable for classification by rank. Although the general equilibration method does not appear suitable for de- termining 100 per cent humidity moisture values for rank classification, it may be useful for studying the nature of mois- ture in coal. The am hors plan to use this method in studying the nature of the moisture of banded ingredients of repre- sentative Illinois coals in an attempt to throw further light on the nature of moisture in coal. 24 EQUILIBRATION METHOD OP Fig. 9. 30 40 50 60 70 RELATIVE HUMIDITY - PER CENT -Moisture-humidity curves for high-moisture samples, nitrogen atmosphere. DETERMINING MOISTURE IS COAL 25 Table 9. — Moist Mineral-Matter-Free Calorific Values as Calculated Using Moisture Values Obtained by Different Methods. (All Coals contain less than 69 percent dry Mineral-Matter-FREE Fixed Carbon) Coal Moisture Ash Sultur B.t.u. B.t.u. Lab. No. Count) bed Method 1 as as as as moist No. Rec'd Rec'd Rec'd Rec'd M.M. Free C- 1863 A La Salle 2 (A) 14.6 10.1 3.88 10786 12181 (B) 14.2 10.1 3.90 10837 12240 (C) 14.4 10.1 3.89 10811 12210 (D) 14.7 10.1 3.87 10773 12165 (E) 14.3 10.1 3.89 10824 12225 (F) 14.4 10.1 3.89 10811 12210 C- 1864 A Perry 6 (A) 10.5 10.7 3.91 11012 12535 (B) 9.9 10.8 3.94 11086 12637 (C) 10.2 10.8 3.92 11049 12593 (D) 10.5 10.7 3.91 11012 12535 (E) 10.4 10.8 3.92 11024 12564 (F) 10.0 10.8 3.93 11074 12623 C- 1900 A Saline 5 (A) 7.2 8.4 2.51 12333 13632 (B) 6.6 8.4 2.52 12413 13722 (C) 7.1 8.4 2.51 12346 13647 (D) 7.2 8.4 2.51 12333 13632 (E) 6.7 8.4 2.52 12400 13708 (F) 7.0 8.4 2.51 12360 13663 C-1901A Henry 6 (A) 18.4 9.7 3.84 9884 11088 (B) 17.7 9.8 3.87 9969 11199 (C) 18.7 9.7 3.82 9848 11046 (D) 19.1 9.6 3.80 9779 10976 (E) 18.1 9.7 3.85 9921 11130 (F) 19.0 9.6 3.81 9812 10991 C- 1903 A St. Clair 6 (A) 10.5 11.1 3.76 11098 12694 (B) 10.6 11.1 3.75 11087 12681 (C) 10.4 11.1 3.76 11110 12708 (D) 10.8 11.1 3.75 11061 12651 (E) 10.4 11.1 3.76 11110 12708 (F) 11.9 10.9 3.70 10924 12459 C- 1904 A Henry 6 (A) 20.7 10.0 3.25 9636 10838 (B) 19.5 10.1 3.30 9782 11019 (C) 20.4 10.0 3.26 9672 10879 (D) 20.8 10.0 3.25 9624 10824 (E) 18.6 10.3 3.34 9891 11172 (F) 18.8 10.2 3.33 9867 11131 C-2009 Vermilion .... 7 (A) 13.3 10.1 3.10 10989 12398 (B) 12.6 10.2 3.13 11078 12516 (C) 13.5 10.1 3.10 10964 12369 (D) 14.0 10.1 3.08 10901 12297 (E) 13.1 10.2 3.11 11015 12443 (F) 13.1 10.2 3.11 11015 12443 (G) 13.3 10.1 3.10 10989 12398 (H) 13.3 10.1 3.10 10989 12398 C-2010 La Salle 2 (A) 14.3 9.0 3.03 11072 12324 (B) 13.1 9.1 3.08 11227 12515 (C) 14.4 9.0 3.03 11060 12310 (D) 14.9 8.9 3.01 10995 12222 (E) 14.0 9.0 3.04 11111 12368 (F) 14.1 9.0 3.04 11098 12353 (G) 14.4 9.0 3.03 11060 12310 (H) 14.3 9.0 3.03 11072 12324 26 EQUILIBRATION METHOD OF Table 9 — (Continued) Coal Moisture Ash Sulfur B.t.u. B.t.u. Lab. No. County bed Method 1 as as as as moist No. Rec'd Rec'd Rec'd Rec'd M.M. Free C-2022 Logan 5 (A) 14.4 11.4 2.84 10566 12103 (B) 13.8 11.5 2.86 10641 12206 (C) 14.6 11.4 2.84 10542 12075 (D) 15.1 11.3 2.82 10480 11988 (E) 15.2 11.3 2.82 10468 11974 (F) 14.9 11.3 2.83 10505 12017 (G) 14.6 11.4 2.84 10542 12075 (H) 14.8 11.3 2.83 10517 12031 C-2023 Saline 5 (A) 7.7 9.3 2.97 12103 13535 (B) 7.7 9.3 2.97 12103 13535 (C) 7.3 9.4 2.98 12156 13612 (D) 8.1 9.3 2.96 12051 13476 (E) 6.5 9.4 3.01 12261 13732 (F) 6.3 9.5 3.02 12287 13779 (G) 6.6 9.4 3.01 12248 13717 (H) 6.4 9.5 3.01 12274 13763 C-2030 Gallatin 5 (A) 3.8 11.1 3 . 39 12637 14472 (B) 3.7 111 3.39 12650 14487 (C) 3.1 111 3.41 12729 14580 (D) 3.7 111 3.39 12650 14487 (E) 3.1 11.1 3.41 12729 14580 (F) 3.3 11.1 3.40 12703 14549 (G) 3.1 11.1 3.41 12729 14580 (H) 3.1 111 3.41 12729 14580 C-2031 St. Clair 6 (A) 10.9 11.6 3 . 69 11008 1 2668 (H) 10.4 11.6 3.71 11070 12741 (C) (I)) 11.5 11.5 3 . 66 10934 1 2564 (E) 10.4 11.6 3.7,1 11070 12741 (F) 10.6 11.6 3.70 11045 12711 (G) 10.9 11.6 3 . 69 11008 12668 (H) 10.4 116 3.71 11070 12741 C-2034 Franklin. 6 (A) 9.5 8.4 1.16 1 1 799 13004 (B) 9.4 8.4 1.16 11812 13018 (C) 9.4 8.4 1 .16 11812 13018 (D) 10.0 8.4 1.15 11734 12931 (E) 9.8 8.4 1.15 11760 12960 (F) 9.7 8.4 1.16 11773 12975 (G) 10.3 8.3 1.15 11695 12873 (H.) 10.1 8.4 1.15 11721 12917 C-2037 Christian 6 (A) 13.0 11.2 4.48 10618 12165 (B) 12.0 11.4 4.53 10740 12340 (C) 12.9 11.2 4.49 10630 12180 (D) 13.4 11.2 4.46 10570 12109 (E) 14.3 11.1 4.41 10460 11964 (F) 13.4 11.2 4.46 10570 12109 (G) 13.5 11.2 4.45 10557 1 2093 (H) 13.2 11.2 4.47 10594 12137 C-2039 Marion 6 (A) 12.5 9.9 3.32 11001 12385 (B) 12.3 9.9 3.32 11026 12414 (C) 12.5 9.9 3.32 11001 12385 (D) 13.1 9.8 3.29 10925 12283 (E) 17.1 9.4 3.14 10422 11649 (F) 15.2 9.6 3.21 10661 11951 (G) 14.1 9.7 3.26 10799 12123 (H) 13.4 9.8 3.28 10887 12239 DETERMINING MOISTURE IN COAL 27 Table 9- (Concluded) Coal Moisture Ash Sulfur B.t.u. B.t.u. Lab. No. County bed Method 1 as as as as moist No. Rec'd Rec'd Rec'd Rec'd M.M. Free C-2042 Randolph .... 6 (A) 10.9 10.7 2.80 11104 12616 (B) 10.2 10.8 2.82 11191 12733 (C) 11.1 10.7 2.79 11078 12586 (D) 11.6 10.6 2.78 11016 12499 (E) 10.9 10.7 2.80 11104 12616 (F) 11.1 10.7 2.79 11078 12586 (G) 10.9 10.7 2.80 11104 12616 (H) 10.9 10.7 2.80 11104 12616 C-2046 Knox 6 (A) 18.0 9.1 2.95 10243 11401 (B) 17.9 9.1 2.96 10256 11416 (C) 17.6 9.1 2.97 10293 11458 (D) 18.8 9.0 2.92 10144 11275 (E) 18.6 9.0 2.93 10168 11302 (F) 19.3 9.0 2.91 10081 11204 (G) 17.8 9.1 2.96 10268 11429 (H) 17.6 9.1 2.97 10293 11458 C-2059 La Salic Local (A) 15.1 6.5 2.16 11386 12286 below (B) 15.1 6.5 2.16 11386 12286 6 (C) 14.5 6.6 2.17 11466 12388 (D) 15.1 6.5 2.16 11386 12286 (E) 14.9 6.6 2.16 11413 12330 (F) 14.4 6.6 2.17 11480 12404 (G) 14.7 6.6 2.17 11440 12360 (H) 14.3 6.6 2.18 11493 12418 C-2064 Gallatin 5 (A) 5.0 8.9 3.17 12664 14107 (B) 5.0 8.9 3.17 12664 14107 (C) 4.8 8.9 3.18 12690 14137 (D) 5.1 8.9 3.17 12650 14092 (E) 4.5 9.0 3.19 12730 14200 (F) 4.2 9.0 3.20 12770 14245 (G) 4.3 9.0 3.20 12757 14231 (H) 4.2 9.0 3.20 12770 14245 C-2069 Saline 5 (A) 4.8 11.5 3.29 12229 14066 (B) 4.8 11.5 3.29 12229 14066 (C) 4.7 11.5 3.30 12242 14082 (D) 5.1 11.5 3.28 12191 14021 (E) 4.3 11.6 3.31 12294 14160 (F) 4.7 11.5 3.30 12242 14082 (G) 4.3 11.6 3.31 12294 14160 (H) 4.2 11.6 3.31 12306 14174 O2080 Saline 5 (A) 5.0 12.3 3.49 12070 14028 (B) 5.0 12.3 3.49 12070 14028 (C) 4.8 12.3 3.49 12096 14059 (D) 5.1 12.2 3.48 12058 13996 (E) 4.0 12.4 3.52 12198 14198 (F) 4.1 12.4 3.52 12185 14183 (G) 4.0 12.4 3.52 12198 14198 (H) 3.8 12.4 3 53 12223 14228 'A Sample ground to M-inch in mine. Total moisture by Air Dry Loss +Regular Moisture. B 50-lb. mine sample crushed to 34-inch in laboratory. Total moisture by Air Dry Loss -(-Regular Moisture. C 5-gram sample crushed to 20 mesh in mine. Moisture by A.S.T.M. oven, air atmosphere, 105° C. l 1 ^ hours. D 5-gram sample crushed to 20 mesh at mine. Moisure by vacuum oven 3 inches Hg. ,X> atmosphere, 105° C, 3 hours. E 100 per cent humidity extrapolated moisture values — 1st equilibration, evacuated. F 100 per cent humidity extrapolated moisture values — 2nd equilibration, evacuated. G 100 per cent humidity extrapolated moisture values — 1st equilibration, nitrogen atmosphere. H 100 per cent humidity extrapolated moisture values — -2nd equilibration, nitrogen atmosphere. 28 EQUILIBRATION METHOD OF BIBLIOGRAPHY 1. Classification of Coals by Rank (Tentative) D 388-361", A.S.T.M. Standards on Coal and Coke p. 98 (1936). 2. Classification of Coals by Rank, I) 388-38, 1938 Supplement to Book of A.S.T.M. Standards pp. 157-162. 3. Lavine, Irvin, and Gauger, A. \Y., Ind. Fng. Chern. 22, 1226 (1930). 4. I.orian, M., Lavine, I., Mann, C. A., and Gauger, A. W., Ind. F.ng. Chem. 22, 1231 (1930). 5. Stansfield, Edgar, and Gilhart, K. C, Trans. Amer. Inst. Min. and Met. Kngr., Coal Div., 101, 125-143 (1932). 6. Investigations of the Coals of Canada, Dept. of Mines, Ottawa, Rept. 83, p. 130 (1912). 7. Stansfield, E., Trans. Can. Inst. Min. and Met. 26, 292 (1923). Also Research Council of Alberta, Fourth Annual Report 39 (1923). 8. Stansfield, F., and Gilhart, K. C, Trans. Amer. Inst. Min. and Met. Fngr. Coal Div., 101, p. 138, Appendix I (1932). 9. Appraisers Bull. 2814, Ottawa (Aug. 25, 1923). 10. Standard Methods of Laboratory Sampling and Analysis of Coal and Coke, A.S.T.M. Desig. I) 271-33. 11. Stansfield, F., and Gilbart, K. C, Private Communication. DETERMINING MOISTURE IN COIL 20 APPENDIX A DETERMINATION OF THE VAPOR PRESSURES OF THE SATURATED SALT SOLUTIONS USED IN EQUILIBRATION TESTS By O. W. Rees and G. W. Land The problem of making vapor pressure determinations on the saturated salt solu- tions used in the equilibration tests was complicated by the necessity of obtaining values for them as used in the moisture study rather than in cells especially con- structed for vapor pressure determina- tions. Following is a description of the apparatus and procedure used as well as the results obtained. Of the many procedures available for making vapor pressure determinations two were selected as promising. The first procedure tried was the static method used by Frowein (1) which used an oil- filled siphon manometer. This was found to be unsuitable so the manometer method applied by Rayleigh (2) to the measure- ment of low pressure was tried. This procedure, with certain modifications, proved to be satisfactory. A description of the apparatus used in this study follows. APPARATUS The double-arm manometer was con- structed of glass, and the two arms were connected at both the top and bottom with glass tubing. From the tube connecting the bottoms of the mano- meter arms, a long glass tube extended down to a mercury well which could be raised and lowered to control the mercury level in each arm. The upper connecting tube was provided with a vacuum stop- cock making it possible to close off one arm from the other. Each arm was pro- vided with a sealed-in-tungsten contact point. These points were connected through push buttons and dry cell battery to a lamp and scale galvanometer for use in levelling the manometer while en- closed in an air-bath thermostat. The manometer assembly was mounted on a framework pivoted at the center to pro- vide tilting in either direction. Tilting was controlled by a screw extending through the bottom of the air-bath thermostat. A mirror for projecting the angle of tilt was mounted at the hub of the tilting device half way between the two arms of the manometer. The mano- meter was connected to a McLeod gage by a long slightly flexible glass U-tube. This gage measured the absolute pressure in the system at the beginning of each determination. The flexible glass tube also connected to the system the desicca- toi containing the unknown solution. The temperature of this solution was controlled at 30°C. ± 0.1°C. by immersing the desiccator in a water-bath thermo- stat. A light was mounted on a milli- meter scale four meters from the mano- meter mirror in such a way that the angle of tilting was projected by the mirror through a glass window in the door of the air-bath thermostat on to the scale and read by means of a telescope. Fig- ure A gives a diagrammatic sketch of the manometer assembly, side view or view along the axis of rotation of the mano- meter. Figure B is a diagrammatic sketch of the manometer assembly, front view or view perpendicular to the axis of rotation of the monometer. Figure C is a diagrammatic sketch of the mano- meter, scale, and alidade, side view. Figure D is a photographic view of the manometer assembly. "This figure shows a Leeds & Northrup lamp and scale in position to read the projected tilt of the manometer. This was later replaced with a surveyor's alidade which was used for all measurements reported. A rubber tube connection to the levelling reservoir is also shown in this figure. This was later replaced with a glass tube extending straight down to the reservoir. Vacuum desiccators of the type shown in figure I) were used to contain the saturated solu- tions whose vapor pressures were meas- ured. 30 EQUILIBRATION METHOD OF GALVANOMETER GLASS U-TUBE TO MC LEOD GAUGE AIR BATH BOX MERGURY LEVELING BULB MERCURY SEALED RUBBER CONNECTION WATER THERMOSTAT DESICCATOR MERCURY RESERVOIR DRYING TUBE GALVANOMETER TO PUMP* MC LEOD GAUGE TO WATER PUMP -0*— MIRROR AIR BATH BOX i MERCURY RESERVOIR GLASS U-TUBE B Fig. A (above). — Diagrammatic sketch of manometer assembly, side view. Fig. B (below). — Diagrammatic sketch of manometer assembly, front view. DETERMINING MOISTURE IN COAL 31 TO MC LEOD GAUGE =!-•— ALIDADE HORIZONTAL SCALE Fig. C. — Diagrammatic sketch of m*an6'metir, scale, and alidade, side view. PROCEDURE Saturated solutions for the vapor pres- sure measurements were prepared by two procedures. In the first procedure a saturated solution was prepared and placed in the bottom of the desiccator. An excess of the salt was placed in direct contact with this solution to insure com- plete saturation at all times. In the second procedure the dry salt was placed in the desiccator together with a small beaker of distilled water. The desiccator was closed, evacuated, and allowed to stand until the water in the beaker had passed over to the salt. While the first procedure was more rapid, it produced solutions containing more entrapped air bubbles than the second. However, it was possible to remove these air bubbles by several alternate evacuations and re- turns to atmospheric pressure. Duplicate determinations were made on solutions of each salt prepared by each method. The procedure for the actual vapor pressure measurements was as follows: The air and water-bath thermostat were brought to the desired temperature (30°C). The desiccator containing the saturated solution was evacuated by using an aspirator pump, and placed in the water bath. When it had reached the 30°C. temperature (determined by a small Anschutz thermometer placed in- side the desiccator) it was connected to the manometer system through the mer- cury-sealed rubber connection (fig. A), stopcocks G and I (figs. A and B) were closed and the system was partially evacuated with a Hyvac pump. Stop- cock G was then opened and the system, including the desiccator, was further evacuated. Pumping was continued for about ten minutes after which stopcock G was again closed and with stopcock I opened, the system was evacuated until the McLeod gage reading was suitably low. This value was recorded and stop- cock J (fig. B) was closed. The mano- meter was then levelled by adjusting the tilt of the beam with the special screw (fig. A) until the contact points indi- cated that the mercury in both arms was equal in height in the arms. It was neces- sary to adjust the height of the mercury 32 EQUILIBRATION METHOD 01 Fig. I). Manometer assembly Table I. Experimental Vapor Pressure Values 1 2 3 4 5 6 7 Differ- 8 Average 9 Vapor 10 Saturated salt solution Solution prepared by direct com- Solution prepared by distillation ence between ot results pressures used by Differ- bination method Vapor pressure, mm. method Vapor pressure, mm. results from two methods from two methods Stansflcld and Gilbart between columns 8 and 9 1st 2nd avg. 1st 2nd avg. m m . mm. m m . KCIO, 31.116 31.066 31.091 31.091 31.070 .021 K 2 S0 4 30.801 30.751 30.776 30.751 30.785 30.768 .008 30.772 30 . 800 .028 BaCU.2H 2 28.687 28.637 28.662 28.677 28.639 28.658 .004 28.660 28.600 .060 KC1 26.796 26.890 26.843 26.879 26.874 26.877 .034 26.860 26.900 .040 NaCl 23.997 24.007 24.002 23.972 23.927 23.950 .052 23.976 23 . 900 .076 NH 4 NO, 19.090 19.045 19.068 18.978 19.032 19.005 .063 19.037 19.000 .037 Ca(N02)2.4FhO. . . 14.907 14.967 14.937 14.907 14.931 14.919 .018 14.928 14.900 .028 CH,COOK 6.529 6.482 6.506 6.535 6.606 6.571 .065 6.539 6.365 .174 UCI.H2O 3.575 3.523 3 . 549 3.573 3.532 3.552 .003 3.551 3 . 560 .009 DETERMINING MOISTl'Rl-: IN COAL 33 i~^ r ex I QC3 a. 2 Fig. E. — -Diagrammatic sketch showing the geometrical relationship involved in the calculations. reservoir (figs. A and B) during this levelling process. When proper levelling was secured the projection from mirror C (fig. C) was read on the scale by use of the alidade. Stopcock K (fig. A) was then closed, shutting off" the back arm of the monometer from the rest of the sys- tem, stopcock I was closed, and stopcock G was opened, admitting the unknown vapor pressure to the front arm of the manometer. The rocker arm was then tilted and the height of the mercury reservoir adjusted until the contact points in both arms just touched the mercury. The mirror projection was read and re- corded and the manometer was adjusted at five-minute intervals until a constant reading was obtained. From the data obtained as outlined above and from certain fixed dimensions of apparatus vapor pressure calculations were made. CALCULATION'S The dimensional data of the apparatus (figs. C, E) as used in this work, which were necessary for vapor pressure calcu- lations were: (1) Distance AB = 290.6 mm. (by cathetometer measurement) (2) (3) (4) (5) (6) Distance DC = 413.39 cm. Scale position (D) level with cen- ter of mirror = 70.10 cm. Scale reading (E) corresponding ing to level position of mano- meter. Scale reading (F) corresponding to tilted position of manometer (due to vapor pressure of solu- tion). Absolute pressure of system at beginning of measurement (Mc- Leod gage reading). Figure E is a diagrammatic sketch showing the geometrical relationships in- volved in the calculations. Distance AB is the distance between manometer arms and indicates the level position of the manometer. A'B' indicates the mano- meter position tilted to compensate for the vapor pressure (VP) of the unknown solution; CD is the horizontal distance from mirror C to the scale; DE and DF are distances on the scale determined respectively by the level and tilted positions of the manometer. Angle a 3 is the angle passed over by the reflected beam of light and it is equal to the sum of angles a y and a 2 . Values for these 34 EQUILIBRATION METHOD OE two angles may lationships tan «, be obtained by the re- DE and tan as = DC DF DC Numerical values may then be obtained by reference to standard trigonometric tables. The angle of triangle B'A'L is the actual angle of tilt and it determines the distance B'L which measures the change in height of the mercury level in the manometer due to the vapor pressure of the saturated salt solution. By the laws of reflection, the angle passed over by a reflected beam of light is twice that passed over by the reflecting surface when the latter is rotated about an axis parallel to the surface, therefore: angle = -~- + 2 Having obtained the value for angle 6 and knowing the distance B'A' of triangle B'A'L it is possible to calculate distance B'L sin