I LLINOI S UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN PRODUCTION NOTE University of Illinois at Urbana-Champaign Library Large-scale Digitization Project, 2007. A Vol. Ax '' "' ..OCTOBER 16, 19tl1. No;. '<2. ; S 4 . - [Entered Frb.1T*96 a'Yrbaa, i.,,as sec . matt under Acto Congrss o July 16, 1894] tTL ETIN NO. 50 - TESTS, OF A SUCTION GAS PRODUCER -BY 1C. M GARLAND AND A. P.- KRATZ '- A" ^ ^ S^^ l '5 °-ff" ^ ..^;.y .- ^ _ )' ..*- <.--: "f ^'A-:i i^ . '^^S^sy^ ^.^y^^y^,.I i--5^1 ^ 1 ,^iy'si^ -5, 5' A A'-' NIVERSITY OF ILLINOis- E BERING 'AE XPEMENTSTATION .-URBANA, ILLINOIS PUBLISHED BY -wa N VyrtSmP A' '^^ ^es&^tS- 5 ' ^''^ - *'* Ay ^^ " *'^1*-;"~^-' *."^ ^ '1'" 1^ -*,"^? -; ^1^ . ''"f"1^ ''1''--,'^ ^ ;1'::' ;A ."'< :''y:.5 * 5:';': I5-.~fSII^ 5(45. ":'^^^ % :'A *e -iir^ ^ it(!' '5 AN-fi ' -, 'S- -; S -< 'S A' ''A, -." - y-' **r. ^,tr :" -;" f -f *"*': '?:-L.t" 'V $1' cover arE "ca, ' + ,tions - umber c ,++ i, !1'++ 1' +: +++ +2+ +e 7 :+l0++ ^a ^q^XVsDI V +l + ,2/ ; 1t u E Engineering Expriment Station was established by the purpose of the Station to carry on investigations S *along v.arious lines of engineering and to study problems facturing, railway, mining,Bconstructional, and industrial interests of theState.1 The control of the Enguieering Experiment Station is vested' :,in ,the hieads of the several departments f the College of' Engineering. T'hese constitute the Station Staff, and with the D irector, determine the character ofthe investigations to be under taken. The work is carried on under the supervision of the Staff, sometimes by research fellows as graduate work, sometimes by members-of the instructional force of the College of Engineer- ing, but mor frequently by investigators belonging. to the Station corps. The results of these ivestigations are published in the Station's own staff of investigators. There will also be issued from time to tie in the form of ,tiulars, compilations giving the results of the efpe tme tsi of, engineers , industrib. works, technical.institutions, 1and governientaltestinga departments. The volume and"numberat the top of the title page of the tote n?:. 2tiae: :orni o einpaw1-i' lt8; Of '^i .. ... 4f + e er l teciitcAna utos  i  ...........~ g e~rtenta + ++,I' ' - ,-++++:+++ * m~lnst~i++1* V V AV' ~ ~---.~ -'I- '-A S ~ I" I V -S ,~-,- 4 V ',sV~ ~ $ 2~~j~ V ~~V'<~ 4L ~ ~' ' A'' ~ 'I- 'V4' V s^^y'-^ ^..^^fe'"^1 ^s *^^'^ ^*'^-^*'v^ , . '-',' '. '' **''w ' * * ' * .'>l-' >y$^:%'''^ j^-^'^^ ur1^^,-;^^ ^g^^.1^^.^- u-%^^' Igaj^ ^^&\ WES;'^^ ^S^ ^..H'^'^^ s,^y^ W^- a'-"^';^:^-''^ ^^'*i^ :*.".,,'?"" ''^'."^**. ii^'^'A Ifi.^'^1-?'^- ^*fe,;:'-'^'.;: aot1,,^?:-^'. ^^%^ *sv-y^^^. ";:'':':-v*i^"»' y-:.^-.;^^; ^-^^-n^; H..^.-^^1-^ ^*.l.v^"**-3^^> s»^-- **^^''. 1-llSI p?^"'^^?-^ lilllS ..:^*-";**^*'y^'i.*\?:* -l*\'-*:;A.--.t*,^..*l; '^??^:"11^ Nys?^ ^^^^ &|IT%1 ^*;;>.-"Sfa-»< UNIVERSITY OF ILLINOIS ENGINEERING EXPERIMENT STATION BULLETIN No. 50 OCTOBER, 1911 TESTS OF A SUCTION GAS PRODUCER By C. M. GARLAND, FORMERLY INSTRUCTOR IN MECHANICAL ENGINEERING AND A. P. KRATZ, INSTRUCTOR IN MECHANICAL ENGINEERING CONTENTS I. THEORETICAL DISCUSSION Page 1. Introduction...................................... 3 2. Simple Carbon Monoxide Producer ........ ........... 3 3. Producer Using Water Vapor.......................... 5 4. Conditions in the Actual Producer ..................... 8 5. The Carbon Monoxide Producer ....................... 10 II. DISCUSSION OF THE METHODS OF EXPERIMENTATION 6. Object of the Tests............ ........ .............. 11 7. The Producer .................... .......... .. ..... 12 8. The Plant........................... ........... . 15 9. Method of Conducting Tests.......................... 15 10. Duration of the Tests ....... . ..... .................. 35 11. Forms for Computation and Discussion of Items of Table 5 38 12. Unaccounted-for Loss............ ........ .......... 40 III. RESULTS OF THE TESTS 13. General Discussion of Tests.... ...................... 41 14. Percentage of CO.. ............................... 48 15. Effect of the Size of the Fuel......................... 51 16. Effect of Capacity ............................... ... 58 17. Coal Per Square Foot of Grate Area....... ........... 60 18. Clinker ................................ . . .. .... 61 19. Weight of Water Required for the Producer...... ..... 63 20. Effect of Fuel Bed Temperature...... ............... 66 IV. CONCLUSIONS V. APPENDIX TESTS OF A SUCTION GAS PRODUCER I. THEORETICAL DISCUSSION 1. Introduction.-The chemical reactions occurring within the fuel bed of the gas producer, while well understood by the metallurgist and the gas engineer, are probably not so clear to the mass of mechanical engineers. Hence, it may be pertinent in the presentation of the results of these tests, to review briefly the theory involved in the conversion of a solid fuel into a gaseous fuel through the agency of the gas producer. 2. Simple Carbon Monoxide Producer.-In its simplest form, the gas producer consists of a closed retort in which carbon is burned in a limited supply of oxygen. Fig. 1 illustrates such a producer. Air enters the retort at the base, and passing up through the grate into the bed of incandescent fuel, comes in contact with the carbon in the lower zone of the fuel bed, where the oxygen of the air unites with the carbon to form CO2. The formula expressing this reaction is: C+2O= CO2...................(1) Since the atomic weights of carbon and oxygen are, respectively, 12 and 16, the above formula indicates that 12 parts by weight of carbon require 32 parts by weight of oxygen and that 44 parts by weight of carbon dioxide are formed; i. e., C + O = CO, 12 + 32 =44 or 1 + 2* = 31 This reaction is exothermic, i. e., it gives out heat, the amount of heat given out being 14 540 B. t. u. per lb. of carbon. The CO formed, remaining in contact with the incandescent carbon, begins immediately to take up more carbon, since at tem- peratures above 1100°F., the CO becomes an oxidizing agent. This reaction is expressed by the formula: COs +C = 2 CO ...............(2) which is reversible, i. e., it may take place in either direction, de- pending upon the temperature. The above formula shows that 44 lb. of CO2 unite with 12 lb. 4 ILLINOIS ENGINEERING EXPERIMENT STATION of C and produce 56 lb. of CO. Thus: CO, + C = 2 CO 44 + 12 = 56 83 + 1=41 This reaction is endothermic, i. e., it absorbs heat from the fuel bed. The amount of heat absorbed per pound of carbon taking place in the reaction is 10 100 B. t. u. Outlet FIG. 1 As the 10 100 B. t. u. represents also the amount of heat that would be given out on combustion of the CO formed, the theo- retical thermal efficiency of the carbon monoxide producer will be, therefore, if all of the carbon is assumed to be converted into 10100 carbon monoxide, 14540-= 69.5 per cent. The remaining 80.5 per GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 5 cent of the heating value of the fuel has been lost in radiation, conduction, and in sensible heat in the gases leaving the producer. The gas will consist of CO and N2. 3. Producer Using Water Vapor.-In the actual commercial producer, the loss in sensible heat inherent in the simple carbon monoxide producer is reduced by utilizing this heat in the decom- position of some other agent, either the vapor of water or carbon dioxide, introduced into the fuel bed with the air. The water vapor is usually supplied by evaporating water either in an external boiler, which utilizes the sensible heat of the gases leaving the producer, or by some form of vaporizer in the form of a water jacket surrounding the fuel bed, and utilizing the heat from the fuel bed for the vaporization of the water. The steam generated in either case mixes with the air entering the producer, and passes with it into the fuel bed. The reaction between the carbon and oxygen of the air has been considered, and it has been pointed out that for each pound of carbon burning to carbon dioxide, 14 540 B. t. u. was given out; also, that if this carbon dioxide were decomposed into the monox- ide, its formation would take up 10 100 of the B. t. u. originally given out. Consequently, 4440 B. t. u. would be left in the form of sensible heat, part of which might be utilized in the decom- position of water into oxygen and hydrogen. The hydrogen lib- erated through such a reaction would pass through the fuel bed partially intact, thus adding a combustible constituent to the gas, while the oxygen would unite with the carbon to form either the dioxide or the monoxide, depending upon the temperature at which the reaction occurred. It is considered, ordinarily, that the following reactions occur within the producer when the vapor of water is introduced with the air, viz: 2 H2O+C=2H-2+CO2................ (3) and HIO+C=H2+CO..............(4) Reaction 3 predominates in the. region of low temperatures be- tween 1000°and 1600° F. Above 1600° F., reaction 4 predominates. At a temperature close to 10000 F., there will scarcely be a trace of reaction 4. If it is considered that the fuel bed is of a uniform temperature, in the neighborhood of 10000 F., so that reaction 3 holds, the following relation exists between the constituents en- 6 ILLINOIS ENGINEERING EXPERIMENT STATION tering into the reaction. 36 lb. of water+ 12 lb. of carbon produce 4 lb. of hydrogen and 44 lb. of carbon dioxide, i. e., 2:H20 + C = 2H, + CO, 36 + 12 = 4 4- 44 or 3 +1=*+31 and 1 lb. of carbon burning to CO2 liberates 14 540 B. t. u., 1 lb. of hydrogen burning to H20 liberates 62 000 B. t. u. 62000 Since * lb. of hydrogen is formed through this reaction, 62000 20 660 B. t. u. disappears or is absorbed. For each pound of car- bon entering into the reaction, 20 660 - 14 540 = 6120 B. t. u. dis- appears. Evidently, the above heat deficit must be supplied by additional carbon in the fuel bed burning to CO through the pres- ence of oxygen supplied in the air and according to reaction (1) C + 02=C02, since for the assumed temperature of the fuel bed, it will not be possible for more than a trace of the carbon monox- ide to form.* From equation (1), 14 540 B. t. u. is liberated per pound of carbon, therefore 6120= .42 lb. additional carbon, burned to CO,0 14540 is necessary to supply heat for the completion of reaction 3. Consequently, 1.42 lb. of carbon are necessary for the formation of k lb. of hydrogen, all of the heat of the carbon being utilized in the production of hydrogen, which is the only combustible con- stituent of the gas, the other constituents being CO and N2. In considering reaction 4, which predominates at the higher temperature, it will be well to refer again to the experiments of Harries. *Haber's "Thermodynamics of Technical Gas Reactions", page 138. This has been illustrated very clearly by the experiments of Harries. Harries passed water vapor over incandescent carbon in a tube at various temperatures and found the relation between the CO, CO02, H20 and H2 in the resulting gas. At a temperature of 12400 F., the follow ing results were obtained: H2 percent = 8.41; 002 per cent = 3.84; CO percent = 0.63; H20 per cent = 87.12 The extent of this reaction, doubtless, depends not only upon the temperature, but also upon the time of contact of H20 and CO02 with the carbon, and while the latter variable has evidently been neglected in the experiments, the results are sufficiently decisive to justify the above assumption that at the low temperature, the percentages of CO formed will be a negligible quantity. GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 7 At a temperature of 20500 F., the following relation was found between the constituents of the gas leaving the tube, H2 per cent = 50.730 H20 per cent = 0.303 CO2 per cent = 00.600 CO per cent = 48.340 99.973 This clearly illustrates the predominance of reaction 4. If it is assumed again that the temperature throughout the fuel bed is uniform and is such that only reaction 4 occurs, we have 18 lb. water + 12 lb. carbon producing 2 lb. hydrogen and 28 lb. CO or HO2 + C = Ha + CO 18 + 12 = 2 + 28 or l+ 1 = + 2 1 lb. of carbon burning to CO liberates 4440 B. t. u., and 1 lb. of hydrogen burning to HO20 liberates 62000 B. t. u. 62000 Consequently, the formation of J lb. of hydrogen absorbs 6-- = 10330 B. t. u. A deficit of 10330 - 4440 = 5890 B. t. u. re- sults. This may now be made up by carbon burning to CO through the oxygen supplied in the air according to reactions 1 and 2. For each pound of carbon burning to CO 4440 B. t. u. is liberated. Consequently, in order that reaction 4 be completed, 5890 -- =1.33 lb. of carbon must burn to CO. The total heat from 4440 the burning of 2.33 lb. of carbon to CO results in the production of I lb. of hydrogen. 11 lb. of water are theoretically neces- sary, or .64 lb. of water per lb. of carbon. The gas resulting from the above reactions consists of CO, H112, and N2. The percentage by volume may be obtained in the following manner: Total weight of H2 = lb. Total weight of CO = 2.33 x 21 = 5.44 8 ILLINOIS ENGINEERING EXPERIMENT STATION Total volume of H2 at 62° F. and 30 in. Hg. = - - 0.0053 - 31.5 cu. ft. Total volume of CO at 62° F. and 30 in. Hg. = 5.44 - 0.0736 = 74.00 cu. ft. Total volume of 02 from air 1I x 1.33 -- .0842 = 21.1 cu. ft. Total volume of air 21.1 - 0.21 = 100.3 cu. ft. Total volume of N =79.2 cu. ft. Total volume of gas leaving producer 31.5 + 74 + 79.2 = 184.7 cu. ft. Total volume of gas per lb. of carbon 184.7 -- 2.33 = 79.3 cu. ft. Theoretical Gas Analysis High Heating Value B. t. u. per cu. ft. H. per cent = 17.0...... ...... ................ .. . 55.8 CO per cent = 40.0................. ....... ...... . ...... 127.8 N2 per cent= 43.0......................... ...... . ......... 000.0 -- total 183.6 4. Conditions in the Actual Producer.-In the actual producer, reactions 3 and 4 are doubtless taking place continuously in dif- ferent parts of the fuel bed. It would hardly be possible to operate a producer at so low a temperature as to produce the results ob- tained according to reaction 3. It would be impossible to operate at such a temperature as to prevent reaction 3 occurring. The conditions that probably maintain are as follows. The moisture laden air, comparatively cool, passes into the fuel bed; on entering, it cools down the first layer of fuel, or the combus- tion zone, as it is usually called, to such an extent that reactions 1 and 3 probably result, part of the oxygen supplied by the air uniting with the carbon to form CO2, while part of the moisture decomposes and forms H2 and CO. The carbon dioxide, formed according to reactions 1 and 3, passes up into the hotter portion of the fuel bed and takes up other atoms of C to form CO, according to reaction 2. The moist- ure which is not decomposed and the 02 which is not combined in the combustion zone pass into the hotter portion of the fuel bed known as the decomposition zone, or dissociation zone, where re- actions 1 and 2 probably take place in immediate succession, in the case of the oxygen, while part of the moisture is combined according to 4, producing H112 and CO. The gases leaving this portion of the fuel bed are therefore composed of H112, CO, CO0 and small quantities of 02, and vapor of water, the last two constituents having either passed through the fuel bed intact or having resulted from dissociation. Theoretically, at the higher GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 9 temperatures there should be only a trace of CO2; actually, this quantity will vary from 2 to 17 per cent of the volume of the gas, depending upon a number of conditions. The gases passing from the dissociation zone or layer of high temperature enter into the distillation zone, which is at a lower temperature, and which is so named for the reason that in this zone the volatile matter is distilled from the fresh fuel by the hot fuel bed beneath. This volatile matter for anthracite coal con- sists of small quantities of H2, HO20, CH4, C0H4 and condensible hydrocarbons in the form of tar, etc. The CH4 and C0H4 in the producer gas made from anthracite are inconsiderable; the former will not represent more than 2 per cent of the volume of the gas, while the CzH4 will probably not exceed 0.1 per cent by volume. The condensible hydrocarbons are also inconsiderable. In producer gas made from bituminous coal, these distilla- tion products may represent 40 per cent of the heating value of the gas, and are not only desirable but necessary constituents, when the gases are to be used for reverberatory and other metal- lurgical furnaces where high temperatures are desirable. The following reaction, which may occur between the con- stituents of producer gas, is reversible and depends upon the temperature. CO + H2O=CO2 + H2.... ..............(5) From the results of Hahn's' investigations, it was shown that at a temperature of 15200 F., H,2 and CO became equally strong reducing agents. At lower temperatures, the carbon mon- oxide is the stronger. This means that if the gases on leaving the dissociation zone enter a zone at a temperature lower than 15200, there will be a tendency for the CO to react on the water vapor present and form CO and H,. At higher temperatures, there will be a tendency toward the formation of CO and water. The velocity of the reaction and the extent to which each takes place depend upon the temperature and the depth of the fuel bed. High temperatures and deep fuel beds tend to produce a gas low in CO2 and H2. Low temperatures or shallow fuel beds produce a gas low in CO, and high in CO and H2. It should be understood that the above reactions depend very largely upon the presence of the fuel which acts as a catalyst. Allner2 has shown that if the above reaction in the presence of a 1Haber: Thermodynamics of Technical Gas Reactions. page 145. 2 Haber: Thermodynamics of Technical Gas Reaction, page 309. 10 ILLINOIS ENGINEERING EXPERIMENT STATION catalyst is established, or if equilibrium maintains at a tempera- ture above 22000 F., and if the gases are cooled without the presence of the catalyst, the reactions are "frozen", that is, the removal of the catalyst does not permit the readjustment of equi- librium due to the lower temperature that would otherwise have occurred had it not been removed. If, on the other hand, the catalyst is not removed and the gases are cooled, the reaction will continue until a temperature of about 1400°F. is reached. This latter is the condition that maintains in the producer as the gases flow from the dissociation zone to the distillation zone. The extent of this reaction depends upon the velocity of the gases through the distillation zone, the depth of this zone, and upon the temperatures within the zone. After the gases have left the zone, i. e., are out of the presence of the catalyst, the reaction is "frozen", and no further change in the composition of the gas occurs. 5. The Carbon Monoxide Producer.-The simple carbon mo- noxide producer has been considered in the discussion of the theory of the producer. In contradistinction to this, there is a commercial producer known as the carbon monoxide producer, which is being used to some extent. The differentiation of this producer from other commercial producers lies in the substitu- tion of CO, taken from some outside source, for the vapor of water, for the purpose, of conserving part of the 30 per cent heat loss inherent in the simple carbon monoxide producer. The pro- ducer is used principally for the driving of gas engines. The supply of CO is obtained by piping the exhaust from the engine to the ashpit of the producer. From reactions 1 and 2, and when one pound of carbon burns to CO, it is known that 4440 B. t. u. is given out, which may be utilized in the dissociation of CO.2 14 540 B. t. u. is liberated per pound of carbon entering into the reaction, and since in the pro- duction of CO, 4440 B. t. u. is liberated, evidently 10 100 B. t. u. remains in the CO. Therefore, in the reduction of COz, 10 100 B. t, u. is absorbed per pound of carbon entering into the reaction, 10100 while 3 3966 B. t. u. is absorbed per pound of CO. The 3i GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 11 B. t. u. free for the carrying out of this reaction per pound of carbon is 4440. The number of pounds of CO necessary for the 4440 utilization of this heat is 4440 = 1.11 lb. Since each pound of 3966 carbon in the producer appears principally as CO in the exhaust gas of the engine, there will be 31 lb. of CO2 produced in the exhaust per lb. of carbon in the producer which is much more than is necessary for the above reaction. On investigation, this system would seem to offer a number of advantages over the system in which the producer uses steam for the production of H2. The gases leaving the producer con- sist principally of CO and N2, together with small quantities of EH2, CH4, and C2H4, distilled off from the green fuel. The gas, therefore, will be likely to be more uniform in quality than H2 enriched gas, and will, consequently, be less likely to cause pre- mature ignition in the engine cylinder. This will be due to the large quantity of nitrogen present in the gas and to the small quantity of hydrogen. It will also be possible to use a much higher compression in the engine cylinder, and this will tend to offset any loss in efficiency in the producer that may be caused by the sensible heat lost in the nitrogen. Since part of the engine exhaust is delivered to the producer, the sensible heat in the exhaust is utilized in addition to the heat contained in the unburned products resulting from incomplete combustion in the engine cylinder. These latter may represent anywhere from 5 to 20 per cent of the heating value of the original gas. The disad- vantage of the process doubtless lies in the regulation of the amount of CO2 delivered to the producer. II. PURPOSE OF THE TESTS AND DISCUSSION OF THE METHODS OF EXPERIMENTATION 6. Object of the Tests.-In the field of small isolated power plants, the suction gas producer using anthracite coal of the finer grades has for the past few years been fighting for place. As in the case of almost all new apparatus, its ultimate success or fail- ure has been retarded largely on account of the lack of impartial data on the efficiency, cost of operation, reliability, etc., together with the natural prejudices against change, .and the difficulty in 12 ILLINOIS ENGINEERING EXPERIMENT STATION securing capable operators. The tests herein described were made for the purpose of obtaining impartial data on the efficiency, reliability, and operation of suction gas producers of small size, using anthracite coal as fuel. They were conducted on the producer in the Mechanical Engineering Laboratory, of the Uni- versity of Illinois. Incident to the main object of the tests was the development of a method of studying and testing the producer, the standardization of forms for the presentation of the results of the tests, and probably the most important of all, the derivation of formulas for the necessary computations. Twenty-five tests were made and four grades of fuel used. The results of all tests are included in Table 5, and the forms and formulas are given in the Appendix. It is hoped that the results of the tests themselves will be of value to manufacturers and to users of small power; the method and forms, a partial discussion of which has appeared in a previous article1, may be of value to builders of gas producers and to engineers engaged in development and testing. 7. The Producer.-The producer under consideration was installed by the Otto Gas Engine Works of Philadelphia, and is Known as their No. 3 producer. In the specifications, it was stated that the producer had a maximum capacity of supplying gas for 60 horse-power, and that this was equivalent to a maximum pro- duction of 8100 cu. ft. of gas per hour. It was also stated that the producer was designed for intermittent operation and that the length of runs should not be greater than 12 hours. TABLE 1 DIMENSIONS AND PROPORTIONS 1. Dimensions of grate, ft ........ ................................................ 1.25 x 1.33 2. Grate area, sq. ft................................................................ 1.666 3. Mean diameter of fuel bed, ft........................................... ....... 1.545 4. Depth of fuel bed, ft........ ...... ............................................ 2.21 5. Area of fuel bed, sq. ft................................ ......................... 1.877 6. Height of discharge pipe above grate, ft ................................... ... 2.875 7. Approximate width of air spaces in grate, in ..................... ............. 0.5 8. Area of air space, sq. ft .......................................................... 0.722 9. Proportion of air space to whole grate area, per cent........................... 43.3 10. Area of discharge pipe, sq. ft................................................... .165 11. Outside diameter of shell, ft............... ....................... .............. 2.833 12. Length of shell from base to top of magazine, ft ..... .......................... 7.125 13. Ratio of minimum draft area to grate area, 1 to ............................... 48.8 1Journal A. S. M. E., Dec. 1909. GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 13 The producer, the principal dimensions of which are given in Table 1, was of the contained vaporizer type and was provided with a plain bar grate. The plant, as installed by the Otto Gas Engine Works, con- sisted of the producer, a wet scrubber, a gas receiver and a 23 horse- power producer gas engine known as their No. 7 engine. A section through the producer is shown in Fig. 2. A diagrammatic sketch of the plant as modified for testing is shown in Fig. 3. FIG. 2 14 ILLINOIS ENGINEERING EXPERIMENT STATION GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 15 In order to relieve the producer test of an engine running in conjunction with it and to make the former independent of the latter, it was decided to blank off the engine from the gas main and produce the necessary suction by means of a steam ejector. This reduced the labor of operating to minimum, made regulation positive, under immediate control, and produced conditions tending to insure greater accuracy and satisfaction. It has been urged that producer tests can not be run like boiler tests, independently of the engine, principally on account of the prejudice existing among prospective buyers of producer installations. This objection would doubtless hold in the case of acceptance tests, but even so, it would be desirable in all such cases to run the engine tests in conjunction with the producer tests, and under the normal conditions of everyday operation, However, for the purpose of studying the producer when operating under different conditions, or for studying the action of different fuels in the producer, or for obtaining data on the efficiency and composition of the producer gas generated, the present method has decided and obvious advantages. 8. The Plant.-Referring to Fig. 3, the producer A is shown, provided with a fan blower P for starting, a two-way cock 0, a waste pipe leading to the roof and a water seal in the connection between the producer and the first wet scrubber B. The wet scrubber B was filled with coke and provided with an overflow at Q. The scrubber water entered at R. The Schutte-Koerting steam ejector used for producing the draft, was located at F and provided with the steam connection as indicated. The ejector has a capacity of 12 000 cu. ft. of gas per hour. The steam used by the ejector is condensed in the second wet scrubber G and the suspended mois- ture is removed by the separator N and the dryer H. The latter is simply a gas bell filled with straw. Its use was made necessary by the Westinghouse meters located at I and J. These meters are of the "wet" type, so that the suspended moisture in the gases, before the use of the dryer, tended to collect in the meter and raise the level of the sealing fluid. The gauge box L was blanked off from the pipe line during the producer tests and was used only for the calibration of the meters. This calibration was effected by the use of air introduced at the compressed air connection indicated. 9. Method of Conducting the Tests.-In conducting the tests, 16 ILLINOIS ENGINEERING EXPERIMENT STATION the items to be considered first were the weight of coal, the heating value of the coal, the volume of gas, and the heating value of the gas. The measurement of these quantities will be considered in the order named. (1). Determining the Weight of Coal Fired.-The correct deter- mination of the weight of coal fired in testing producers of the intermittent type, where tests of short duration are necessary, is a problem of no small importance, owing to the difficulties in obtaining the weight of coal in the producer at the start and at the close of the test. The method of starting that suggests itself as being the most readily carried out, especially by the engineer, is to clean the ash and clinker out and work the fuel bed, by means of the poking bar, into a uniform condition; then, in closing, to dupli- cate these conditions as nearly as possible. In any actual case, it is always possible to compute from a number of actual trials the average weight of coal required to fill the producer to a given level. If this mean value is taken as the true value or true weight of coal required to fill the producer, then the maximum variation from the mean or true value in the case of any one trial will indicate the probable maximum error that will be made in filling the producer. Evidently this same maximum error may be made in bringing the fuel bed to the starting condition, irrespective of the weight of coal required to bring the fuel bed to this condition. An example will illustrate this better. Suppose that for a number of actual trials, the average weight of coal required to fill the producer to a given level is 600 lb.,- suppose the maximum variation from the mean is 15. The maxi- mum error is 21 per cent, based on 600 lb. Suppose, further, that in a given test, 200 lb. of coal were burned. Evidently in bringing the fuel bed to the starting condition and in filling, a maximum error of 15 lb. may be made; i. e., in filling, it is possible to make an error in this particular test of 71 per cent. In order to determine the error approximately in estimating the weight of coal during the present tests, the producer was filled four separate times, and the weight of coal required noted in each case. The average of the four weights was taken as the mean weight or true weight of coal required to fill the producer. The results are given in Table 2. GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 17 It will be seen from this table that the maximum variation from the mean is 8.75 lb. or 1.7 per cent. It will be seen from TABLE 2 WEIGHT OF GREEN COAL REQUIRED TO FILL THE PRODUCER Variation Trial Weight from Variation No. pounds Average per cent Weight 1 669.25 8.75 1.70 2 676.25 1.75 .26 3 683.25 . +5.25 .77 4 683.25 +5.25 .77 Total 2 712.00 Aver. 678.00 Table 5, giving the results of the tests, that the smallest weight of coal fired on any test was for test 26, and that this weight was 146 lb. In bringing the fuel bed to starting conditions and in filling the producer at the close of the test, it would, therefore, have. been possible to make an error of 8.75 lb., or, in this case, an error of 8.75 x 100 146 = 6 per cent. This is, of course, the purely mechanical error in filling, and does not take into consideration the difference that may maintain in the condition of the fuel bed in starting and stopping. There is always present in the fuel bed a larger per cent of ash in stopping than in starting, and if this is in the form of clinker, a large error may result. If it is in the form of a powder, the error will probably not be so great, as the ash will tend to pack into the interstices around the coal; and while the fuel bed may contain a larger per cent of ash in stopping than in starting, the volume of the fuel bed, and the weight of carbon present may re- main practically the same. The error, therefore, in the estima- tion of the weight of coal due to the presence of the ash will not be large. The error in determining the weight of ash is unimpor- tant, as this may be determined from the analysis of the coal and the total weight of coal fired, with a greater degree of accuracy than could possibly be determined from the weight of the ash and refuse taken from the producer. The above method of starting and stopping the test has been 18 ILLINOIS ENGINEERING EXPERIMENT STATION used throughout the present work. There is another and much more accurate method of procedure, which is, however, both tedi- ous and expensive. Fill the producer with a weighed amount of wood and coal and start the fire. During the period of starting, measure the volume of gas discharged, and at the same time take a continuous sample for analysis. Also weigh the ash and refuse accumulating during this period, and take a sample for analysis. Reduce the total volume of gas to standard gas, i. e., gas at 620 F. and 30 in. Hg., and calculate from the analysis the volume of CO2, CO, 1H, CH4 and C2H4. From the specific weights of the constituents of producer gas as given in the Appendix page 84, item 121, calculate the total weight of each of these constituents. Let a, b, c, d, e, be the weights of the respective constituents. Then the total weight of carbon that has been taken from the fuel bed during the starting period will be, W= A a + i6 + d + -e+ w, where w- = the total weight of carbon as determined from the analysis and the weight of the ash and refuse. The total weight of hydrogen that has been taken from the fuel in the producer and that appears in the gas may be deter- mined by the formula S= c+ id + -e This neglects the hydrogen that may be formed by the decompo- sition of water, and will, therefore, introduce an error, the amounr of which will depend upon the length o£ time between the lighting of the fire and the starting of the test. With small producers, this will be very small. In large producers, tests are usually of such length and the weight of coal fired of so large an amount, as to make the error in the estimation of the coal in starting and stop- ping a negligible amount, consequently, such a method as the one under discussion would not be necessary. In closing the test, the coal should be burned low in the pro- ducer, and immediately after closing, the fuel bed should be drawn and weighed, the incandescent coals quenched and then sampled, and an analysis made from the sample. The total weight of equivalent coal fired during the test may be obtained from the following formula, (W + Wi) 14560 + W, x 62000 We = Ws + Wt- H where GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 19 We = total weight of equivalent coal fired during the test. Ws = total weight of equivalent coal in producer at the start. Wt = total weight of coal fired during test. W = total weight of carbon appearing in the gas before start- ing. W2 = total weight of hydrogen appearing in the gas before starting. Wf = total weight of carbon within the fuel bed at the close of test. H = The heating value of the coal per pound. (2) Heating Value and Analysis of the Coal.-The heating value and the analysis of the coal were determined by the chemist, and a discussion of the chemical method is unnecessary here. The chemical work is probably accurate within one per cent. The greatest error in the analysis or in the heating value is likely to be made in the sampling of the coal. The weight of the coal sample should be not less than 10 per cent of the total weight of coal fired, and the samples taken in the present tests represented from 10 to 15 per cent of the weight of coal fired. These samples were mixed and quartered until a sample sufficient to fill about eight quart jars was obtained; this sample was then reduced by grinding until it would pass a i in. screen. This was again mixed and quartered, and a sample taken sufficient to fill a pint jar. This sample was sent to the chemist for analysis. The coal used in the tests was stored indoors and was practi- cally air-dry as fired. Proximate analyses and heating values were obtained from the coal used in every test. Ultimate analyses were obtained for at least three samples from each grade of coal used. From these ultimate analyses and the proximate analyses from each test, the ultimate analysis for each test was determined by a method of approximation. (3) Measuring the Volume of the Gas.-The volume of the gas generated was measured by Westinghouse meters. These were of the wet type, so that the gas should contain no suspended moisture. This moisture was removed by the dryer shown in Fig. 3. In or- der to insure accuracy in the measurement by the meters, a gauge box L, Fig. 3, was connected to the gas main so that the meter or meters could be calibrated from time to time. This calibration was effected by the use of compressed air. Blind flanges were 20 ILLINOIS ENGINEERING EXPERIMENT STATION placed at X and Y, and the air was admitted at the connection shown. This air, expanding, passed to the meter and then to the gauge box which contained a thin plate orifice. The pressure and temperature of the air were taken at the meter and also at the gauge box. From the pressure and temperature in the gauge box the volume of air passing the orifice was computed from the work of R. J. Durley.1 From this volume, and the pressure and tem- perature taken at the meter, the volume passing the meter was computed. The calibration was made at different capacities and a calibra- tion curve for the meter plotted. This latter was practically a straight line. In obtaining the volume of gas generated during a test, the volume was taken from the calibration curve of the meter and then reduced to standard gas by the formula given on page 86, of the Appendix, item 125. The volume of gas as obtained from the meter, as just explained, was checked by calculating the volume of gas generated from the weight and analysis of fuel and the analysis of the gas. The formula for this computation is given in the Appendix, page 86, item 126. The volumes as obtained by computation ordinarily checked within 5 per cent of the volumes determined by the meters. Where the sampling of coal and ash is carefully carried out, this method of computing the volumes from the analysis is reliable for pro- ducers using hard coal. It is based on the fact that the weight of carbon contained in the coal must equal the weight of carbon con- tained in the gas plus the weight contained in the ash and refuse, plus the weight lost in tar and the weight lost in the gas absorbed by the scrubber water. The carbon lost from the hard coal pro- ducers in the form of tar and condensible hydrocarbons is very small, doubtless less than one per cent. The weight lost by the absorption of CO, and CO in the scrubber water is also very small, so that the carbon in the coal should be accounted for within 5 per cent at the most. The method may even be used for the deter- mination of the volume of gas generated by bituminous coal pro- ducers. In the case of these producers, 10 per cent of the carbon may be lost in the tar and other heavy hydrocarbons which con- dense and are deposited in the scrubbers and gas mains. 1 Trans. American Society Mechanical Engineers, Dec. 1905. GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 21 The percentage of carbon so lost may be found by taking a continuous sample of the gas as it leaves the producer, and before it is cooled, and drawing it through some form of condenser in which the condensible matter will be thrown down while the gas passes to a small meter for measurement. From the weight and analysis of the condensed matter and the volume of the gas sam- ple, the per cent of carbon lost in this way may be obtained. The determination is tedious and requires considerable outlay, but it may often be used to advantage in the testing of large producers where mechanical means of measuring the volume of the gas gen- erated are impossible. The value of the volume of gas computed from analysis as a check on the accuracy of the test has been dis- cussed on page 39. (4) Sampling the Gas'.-The correct sampling of the gas for analysis and for the determination of the heating value is impera- tive. In the present tests, the form of sampling device is illustrated in Fig. 4. This device takes the sample from three separate points in the main and practically the same volume of gas from each point. In mains of larger diameter, more nipples should be used. A good rule'is to use one sampling nipple for each inch of diame- ter of main. FIG. 4 Credit is due Mr. 0. A. Carnahan for his painstaking and conscientious work in the analysis of the gases, for all the tests, and his ingenuity and suggestions along this particular line. Credit is also due Mr. J. P. Clayton for his assistance in computing, and in operating the calorimeter on a large number of tests. 22 ILLINOIS ENGINEERING EXPERIMENT STATION The above device was inserted beyond the second scrubber and at S, Fig. 3. In this position, the gas is under pressure so that it is necessary only to connect the aspirator bottle used for collecting the sample of gas to the nipple leaving the sampling device. Where the gas is under a pressure lower than atmospheric, it is necessary to connect an aspirator of the ordinary laboratory type to the sampling device and draw the gas from the main by the aspirator and force it into a vessel provided with a water seal. Samples of the gas may be taken from this vessel both for analy- sis and for the calorimeter. Fig. 5 illustrates the arrangement. The principal objection to its use is that small quantities of gas, principally CO, are absorbed by the water in the aspirator. r Gas m eleY 5amp- tt a FIG. 5 GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 23 The sample of gas taken from the sampling device for analysis was collected for the present tests in ordinary aspirator bottles and over water previously saturated with producer gas. These bottles held about 5 litres. 600 cc. of the sample was transferred to a special glass sampling tube containing mercury. This sample was then sent to the chemist for analysis. The sample of gas for the Junker calorimeter was also taken from the above sampling device by connecting the burner of the calorimeter through rubber tubing to the nipple. When operating the calorimeter, Gaz FIG. 6 24 ILLINOIS ENGINEERING EXPERIMENT STATION samples were taken continuously, one after the other. In order to obtain correct samples of gas, they should be drawn from the main continuously and at a uniform rate of flow. With the aspirator bottles above mentioned, the condition of uniform flow was practically impossible at the time of sampling. Later, the gases were collected in glass sampling tubes containing mercury. Each tube held about 600 cc. The tubes were connected directly to the nipple leading from the sampler, and the mercury was drawn off through a glass cock located at the bottom of the tube. The mercury passed through the cock and dropped from a glass tube, the opening in the end of which had been made of such size as to allow the mercury to run from the tube in about two hours' time. Fig. 6 illustrates this arrangement. The apparatus has the advantage of collecting the sample over mer- cury, which eliminates small errors that alwaysresult from the absorption or liberation of gases by the water, and in addition per- mits the observer to see at a glance whether or not stoppage of the flow of mercury has occurred. (5) Determining the Heat Value.-The heating value of the gas has been determined by the Junker calorimeter and by compu- tation from the analysis of the gas. The latter value is assumed to be more nearly the true value, and has been used in the computations. It is assumed to be more nearly the true value for the reason that the gas samples from which the analyses were made were continuous samples taken over the entire period of the test, while samples for the Junker calorimeter were more or less intermittent, due in the case of some of the earlier tests on the Philadelphia and Reading coal, to poor gas, which would not burn at times, and again, to the necessity of using the operator on other work. There are, also, certain errors such as radiation and conduction and possibly errors in the meter which enter into the calorimeter determinations and for which allowance cannot readily be made. Again, the thermometers used in measuring the tem- perature of the entering and leaving water read only to tenths of a degree so that an error of 0.5 of one per cent could easily be made in estimating to hundredths, as the rise in temperature of the water was frequently not greater than two or three degrees. It was found by actual trial that radiation and conduction could easily affect the readings to the extent of two per cent in GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 25 making determinations upon producer gas.. The temperature of the water used in the calorimeter in the case of all tests was below the room temperature, so that the heat was conducted into the calorimeter. This tended to make all determinations about two per cent high. Upon calibrating the gas meter used with the calorimeter, it was found to read a fraction over two per cent low. The error in the meter was, therefore, considered to offset the error due to conduction and radiation. On account of the above causes, it is believed that the heat- ing value of gases whose principal constituents are simple gases, such as CO and H2, may be determined with a greater degree of accuracy by computation from the analysis rather than by the calorimeter, especially where the latter is operated under con- ditions existing on a test and possibly by one unfamiliar with the errors likely to be made in using the instrument. In the case of all the present tests on which the calorimeter was run continuously, the heating value as determined by the calorimeter checked the heating value as determined from analy- sis within less than three per cent, and on a number of these within one per cent. This seems to indicate that either method of obtaining the heating value is accurate if proper care is used. (6) Error Due to Vapor Pressure of Water.-An error that caused considerable discrepancy between the heating value as given by the calorimeter and that as/given by the analysis, and which resulted in about ten days of investigation upon the accuracy of the Junker calorimeter, was the neglect of the effect of the vapor pressure of water contained in the gas passing through the meter. This source of error had been recognized in the deriving of the formula for the calorimeter that appears in the Appendix, page 83, item 120, and had been allowed for, but in temporarily working up the heating value on some of the tests, this formula was not used, and this error was lost sight of. The result was an error of from two to four per cent, which could not be accounted for after allowing for radiation and other errors in the use of the apparatus. As we have since found that this is an error that is not usually recognized or allowed for by engineers, it will not be out of place to discuss and indicate its magnitude. The gas entering the Junker meter is saturated with the vapor of water. Consequently this vapor is under the pressure due to 26 ILLINOIS ENGINEERING EXPERIMENT STATION the temperature of the gas. According to the laws of Dalton, if a mixture of gases is inclosed within a vessel, the volume of each gas occupies the entire volume of the vessel and each gas is under its own pressure. The sum of the pressures of the constituents of the mixture is equal to the total pressure within the vessel. The total absolute pressure within the meter is the barometer pressure plus the pressure in inches of water indicated by the U tube attached to the pressure regulating device that accompanies the meter. The temperature of the gas within the meter is obtained from the thermometer. Assume that the total pressure in the meter is 30 inches of mercury absolute, and that the tem- perature of the gas is 70° F. We find from the steam table that the vapor pressure of the water is .73 in., the total pressure of the mixture is 30 in., consequently, the pressure of the dry gas is 30.0-0.73=29.27 inches. If the temperature in the meter is 100° F., the pressure of the vapor is 1.94 in., and the pressure of the dry gas is 30.0-1.94=28.06 in. of mercury. From these examples, it will be seen that an error of about 73 - x 100 = 2.4 per cent will be made at 70°F. by neglecting the 30 effect of the vapor pressure in computing the standard volume of 1.94 gas used by the calorimeter and that an error of about 1.94 x 100 = 30 6.5 per cent at 100° F. will be made. The same error will be made in computing the total volume of gas discharged from the producer, as indicated by the gas meters. The effect of the above error is to give a low value for the heat of combustion of the gas, and to give a high value for the total volume of gas generated by the producer. If, therefore, no correction is made, the total heat contained in the gas gen- erated by the producer, since it is the product of the two quan- tities, will not be in error, and the efficiency of the producer will consequently not be in error. If, however, the total volume of gas generated has been corrected for vapor pressure, and the volume of gas used by the calorimeter has not been corrected, which is frequently the case, the efficiency of the producer will be low by the amount of this error. (7) Anatyzing the Gases.-The analysis of the gas was made in the Hempel gas analysis apparatus and over water. The burette used was water-jacketed to prevent changes in temperature GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 27 affecting the results, while the water used in the burette was saturated with gas. In some of the later tests, the gases were collected in gas sampling tubes holding about 600 cc. and were collected over mercury. The analysis was also made over mercury. It is believed that this was at least, a safe precaution, as the water in the aspirator bottles previously used for sampling, and in the burette, even if saturated with the gas beforehand, might either absorb or give off small quantities of the gas. There are also other advantages in the use of the "accurate" method which need not be discussed here. (8) Measuring the Weight of Air Used.-As the weight of air is not used in calculations of the first importance, it was not con- sidered necessary to measure this quantity, as it could be computed from the weight of nitrogen appearing in the gas with an accuracy well within 5 per cent. This has been done for all tests and the result appears under item 97 of Table 5. (9) Measuring the Temperatures.-All temperatures with the exception of the temperatures in the fuel bed and the temperature of the gases leaving the producer, were taken with mercury ther- mometers, all of which had previously been calibrated and calibra- tion curves plotted. All mercury thermometers were used with their bulbs in direct contact with the medium whose temperature was to be measured. In calibrating, the Reichsanstalt standards of the department of Physics were used. In making the com- putations, the corrections were taken from the calibration curves. (10) Temveratures of Gases Leaving Producer.-The measure- ment of the temperature of the gases leaving the producer was effected by the use of a platinum-rhodium thermocouple and a Siemens & Halske millivoltmeter calibrated to read direct in degrees Centigrade. This couple and voltmeter were compared with a Reichsanstalt standard, reading in degrees Centigrade and to 5000. After making the proper stem correction for the standard, the readings of the thermocouple practically agreed with the readings of the standard. For higher temperature, the thermo- couple readings were taken at the melting point of zinc, 4190 C., the melting point of silver, 9610 C., and the melting point of cop- per, 1084 ° C., and were found to be practically correct. In using the platinum-rhodium couples, considerable care had 28 ILLINOIS ENGINEERING EXPERIMENT STATION to be taken in order not to have the couple contaminated with gases, as such contamination would destroy the calibration and ultimately make the couple worthless. As a couple of this type five feet in length costs about $50.00, it is evidently necessary to use some care in this respect. In order to protect the present couple, the hot junction was placed in a quartz glass tube about i-in. inside and i-in. outside diameter. This tube was then in- serted in a brass thermometer cup which was screwed into the gas main leaving the producer at 0, Fig. 3. The cold junction of the thermocouple was placed in melting ice. After a number of tests had been run, it was considered pos- sible for the conduction of the brass thermometer cup to lower the temperature as given by the thermocouple. In order to test this, a mercury thermometer, and later a Hoskins thermocouple, the latter having been compared with the platinum-rhodium couple, were placed in direct contact with the gases. The producer was then run at different capacities, in order to produce different temperatures in the leaving gases, and simultaneous readings of the Hoskins couple exposed to the gases and of the platinum- rhodium couple protected by the thermometer cup and quartz glass tube, were taken. The curve of Fig. 7 was plotted from the data taken on this experiment. The results were somewhat startling. As the points fell on practically a straight line, there could be no doubt of the effect of the thermometer cup and the quartz glass tube upon the reading of the thermocouple. The tem- peratures used in the computations were taken in degrees F. from curve 1 of Fig. 7. The actual readings of the platinum-rhodium couple in degrees Centigrade were plotted as abscissas while the corrected readings in degrees F. were plotted as ordinates. Curve 2 of Fig. 7 is a curve showing the relation between degrees Centigrade and degrees Fahrenheit, and was used in transfer- ring temperature from one scale to the other. The corrected average temperature of the gases leaving the producer, in degrees F., is given in Table 5, item 39. (11) Temperatures in the Fuel Bed.-The temperatures in the fuel bed were obtained on several tests by the use of a platinum- rhodium thermocouple and the Siemens and Halske galvano- meter. The hot junction of this couple was placed in a i-in- GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 29 K i~i * '41' I~z K 30 ILLINOIS ENGINEERING EXPERIMENT STATION quartz glass tube about 30 in. long, the quartz tube in turn being placed in a i-in. iron pipe closed and pointed at one end. The pipe containing the hot junction of the thermocouple was then TABLE 3 TEMPERATURE THROUGH FUEL BED Temperatures °F . * Test Tim e 3 inches No. No. from near 6:40 6 to 6:50 1:52 8 to 2:20 11:30 18 to 12:40 11:00 19 to 11:40 11:45 21 to 12:30 10:05 25 to 10:55 10:30 27 to 11:40 8:43 23 to 23 9:35 3:50 Side of Lining 1 1652 2 2192 3 4 1832 1 2490 2 1832 3 4 . 1 2 2350 3 .... 1 2000 2 2300 3 2350 1 700 2 1560 3 2075 1 2100 2 2350 3 1 1075 2 1580 1 1237 2 1500 3 1600 1 1950 3 inches from far At Center Side of Lining 1535 1688 1742 1782 1922 2282 1048 2120 2060 2247 2327 2282 1975 1724 1922 2300+ 2163 2250 2380 2375 1970 1950 2160 2300+ 2137 2250 575 375 1550 1475 1875 1725 2037 2025 2225 2275 2200 2400 1075 1650 2000 2000 1610 1625 1812 1900 2037 2225 1675 1700 *Zone No. 1 = 24 in. above grate. Zone No. 2 = 18 in. above grate. Zone No. 3 = 10 in. above grate. Zone No. 4 = 3 in. above grate. t indicates that the temperature rose above the softening point of the thermo- couple and hence was not obtained. GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 31 inserted in holes drilled through the walls of the producer. These were drilled one above the other at 3 in., 10 in., 18 in. and 24 in., re- spectively, from the grate. Temperatures were taken first through the upper hole, at 3 in. from the near side, at the center and 3 in. from the far side, and in the same positions through each of the other holes. The platinum-rhodium couple did not prove entirely satis- factory for this work, on account of the difficulty in properly pro- tecting it. When the end of the iron pipe entered a zone where clinker was forming and where there were large holes in the fuel bed through which the air was passing, the temperature rose rapidly and either melted off the end of the pipe or bent it so that it was difficult to remove. After losing several inches from the end of the couple and at the same time contaminating it with the gases from the fuel bed, it was decided to try the Hoskins couples. These are a patented thermocouple of considerable mechanical strength, supposed to be unaffected by furnace gases. The composition of the elements forming the couple is not known. The temperature as indicated by the Hoskins couple is read in degrees Fahrenheit direct from the galvanometer provided. This latter reads to 2500° F. It was found, however, that the couples softened at 2250° F., so that it was not possible to go higher than this. These couples are much heavier than the platinum- rhodium couples, and, according to the manufacturers' statement, could be used in the fuel bed without protection. It was found better, however, to enclose them in iron pipes. The temperature n the fuel bed was taken by means of the Hoskins couple from test 18 to test 31. The temperature so observed will be found in Table 3. (12) Measuring the Water Used.-The water fed to the pro- ducer from the vaporizer in tests 2 to 8 inclusive and in tests 31, 32 and 38, was obtained by weighing a tank filled with water which rested on platform scales. This tank was placed on the charg- ing platform above the producer. The water was allowed to drip from the tank into the vaporizer. A constant level was main- tained on the vaporizer by allowing the water to overflow into a second tank resting also on platform scales. The weight of water fed to the producer was therefore the difference between the weight of the supply tank and the weight of the overflow tank. 32 ILLINOIS ENGINEERING EXPERIMENT STATION In all other tests the vaporizer was blanked off and a steam jet supplied the moisture necessary in the operation of the producer. The jet passed through a small thin plate orifice. The pressure on the orifice in pounds per square inch was taken, and after the test, the orifice was calibrated under the pressure observed on the test. The jet was located beneath the grate and in the center of the ashpit. Any steam condensing in the ashpit from the jet overflowed, and was collected and weighed at the end of the test. The weight of this overflow was small and in a number of tests there was no overflow. The weight of water supplied to the pro- ducer was obtained from the pressure on the orifice and the cali- bration at this pressure. The actual weight that entered the fuel from the steam jet was the difference between the above weight and the weight of the overflow. The former quantity is given under item 79A of Table 5. Besides the water from the vapori- zer, there is also the moisture carried in by air and by the coal. These weights are given under items 79B and 790 respectively. The computations for the weight of moisture carried in by the air are given in the Appendix, page 87, items 104 and 105. The water supplied to the first scrubber was measured by a water meter, which was calibrated before the test. The water used by the second scrubber was not considered, as this was only incidental to the method of testing. (13) Weight of Moisture in the Gas.-The weight of moisture in the gas leaving the producer was determined by two separate methods.. First, by drawing a sample of gas through a calcium chloride tube, by means of a water aspirator. A small wet gas meter was placed between the calcium chloride tube and the aspirator from which the volume of gas was determined. The moisture accumulating in the tube was weighed. From the data obtained from this apparatus, the per cent of moisture in the gas leaving the producer was determined. In the second method, the weight of water decomposed in the fuel bed was computed from the analysis of the gases, Appendix, page 87, item 86. This weight is given in Table 5, item 80. The total weight of moisture entering the producer is given under item 79; the difference between the two items gives the weight of water in the producer gas. It is believed that the percentage of moisture in the gas based on this determination is the more ac- curate. GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 33 (14) Ash and Refuse.-The ash and refuse taken from the grate during the test, together with that resulting from the clean- ing at the end of the test, were collected and weighed. The whole was then quartered and sampled, and the sample sent to the chemist for analysis. The weight of carbon lost through the grate was determined from the total weight of ash and refuse and the per cent of carbon as given in the chemist's report. It will be noted from item 49 of Table 5 that in several of the tests, the per- centage of ash and refuse obtained is lower than the percentage of ash in the coal as given by the chemist's report, item 53. This is due to the impossibility of removing all of the ash from the fuel bed at the close of the test. This ash may have been either in the form of clinker which accumulated on the lining of the pro- ducer, or in the form of a soft white powder which packed into the interstices around the coal. The failure to obtain the correct or true weight of ash formed during the test, as explained under the measurement of the weight of coal fired, is not of great importance, as the weight of ash and refuse is used only for the determination of the weight of carbon lost through the grate. In a number of tests, ultimate analyses were made on the ash and refuse for the purpose of investigating the formation of clinker. These analyses are given under item 63 of Table 5. The percentages are based on the weight of "earthy matter." (15) Soot and Tar.-The amount of soot and tar escaping the scrubber of the hard coal producer is, under normal conditions, very small, and as a rule very little trouble results from their presence. No attempt was made in the tests to determine their weight. After about 600 hours' operation and the generation of about 2 000 000 cu. ft. of gas, the horizontal pipe leading from the wet scrubber to the steam ejector was taken down and examined. It was found to contain a coating of tar and soot less than v in. thick on the inner wall. As this period of time would correspond to about two months, running at 12 hours per day, it may be as- sumed that the scrubber capacity is sufficient. (16) Stand-by Losses.-Small producers of the present type are usually designed for intermittent operation and are not well adapted for continuous runs of longer than twelve hours' duration. This is a disadvantage due to the size of the producer, and results from the inability to thoroughly clean the ash from the fuel bed 34 ILLINOIS ENGINEERING EXPERIMENT STATION without lowering the heating value of the gas to such an extent as to interfere with the operation of the engine. In larger producers the volume of gas generated per unit of time is large: the fuel bed is also large, so that the air admitted in poking and cleaning the fire represents a very small percentage of the gas volume gener- ated. Also, the proportion of the fuel bed disturbed at one time is relatively small compared with the total area of the fuel bed. On account of these conditions, the fuel bed of a large producer may be cleaned and poked without seriously affecting the quality of the gas, so that continuous operation is a comparatively easy matter, while with a smaller producer of the same type continuous operation would be impossible. The specifications accompanying the present producer state that it is to be used for runs of not greater than 12 hours. Conse- quently, there will be a stand-by of twelve hours' duration if the engine is being operated twelve hours per day. As this is the condition that maintains in commercial work, it is of importance that the losses due to this stand-by be known. In order to determine these, a stand-by test of 120 hours start- ing October 11, 1909 at 10:20 a. m. and running until October 16, at 10:20 a. m. was made. The test was started by running the pro- ducer until a normal fuel bed was obtained, and then cleaning the fires and filling the producer with coal. The producer was then closed down, and at the end of twelve hours, the fires were cleaned, and the producer operated until the gas would burn at the try cock. The producer was again closed down for twelve hours, and at the end of the period, the fires were again cleaned and the above cycle of operations repeated. During the 120-hr. run, the fires were cleaned and the gas producer operated until a working gas was obtained ten times, that is, once every twelve hours. At the close of the test, 280 lb. of coal were required to fill the producer. This represents a stand-by loss of 28 lb. of dry coal per twelve hours. If we consider that 500 lb. of coal per 12-hr. run is about the normal capacity of the producer, the stand-by loss represents about 5.5 per cent of the dry coal fired or about 6 per cent of the combustible. The results of the stand-by test are given in Table 4. The stand-by loss is divided into four parts; (1) that which is lost through the grate due to cleaning the fire; (2) that which is GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 35 lost through the poor gas generated from the time the blower is started until the gas is sufficiently rich for use in the engine; (3) that which is lost in the small quantity of gas generated during the actual stand-by period; and (4) that which is due to the con- duction and radiation of heat from the producer during this stand-by period. The loss through the grate may be obtained from the data of Table 4. It was not deemed sufficiently important to separate the second and third losses, and it was not possible to obtain the fourth other than approximately. TABLE 4 Results of Test for the Determination of Stand-by Losses. Time of start, 10:20 a. m., October 11, 1909. Time of stop, 10:20 a. m., October 16, 1909, Duration of the trial, 120 hours, Producer started 10 times. Fuel used, anthracite. Commercial name, Scranton pea. FUEL 1 Size and condition.............................. ..... ..... ..................... Pea clean 2 Weight of coal as fired, lb .................... .. ............ ....... 288.0 3 Percentage of moisture in coal.................... . ................... ..... 2.92 4 Total weight of dry coal fired, lb....... ........................... ........... 280.0 5 Total ash and refuse, lb..................................................... .... 51.0 6 Quality of ash and refuse ......... ....... ...... ............................ 7 Total weight of combustible, lb ....................... . .............. 212.0 8 Percentage of ash and refuse in dry coal, per cent............................. 18.2 ULTIMATE ANALYSIS OF DRY COAL 9 Carbon (0 ) ................. .... ..... ........................ per cent .......... 79.83 10 Hydrogen (H2)... .......................................... " .......... 2.59 11 Oxygen (02) ............. .............................. " . ......... 2.20 12 N itrogen (N ) ............................. .......... " .... " .......... .82 13 Sulphur (S) ........................................ .. ...... " . ......... 1,39 14 A sh .. .................. ............... ............... . .. .. . " . ......... 13.17 15 Moisture in sample of coal as received ...................... ' .......... 2.92 ANALYSIS OF DRY ASH AND REFUSE 16 Carbon, per cent. ........................ .. ............................ ......... 43.00 17 Earthy m atter, per cent............. .. .................... .................... 57.00 FUEL PER HOUR 18 Dry coal fired per hour, lb .......... ......... ........................ .......... 2.34 19 Combustible consumed per hour, lb ....... .................................... 1.77 20 Dry coal per sq. ft. of grate area per hour, lb................................. 1.40 21 Combustible per sq ft. of grate area per hour, lb.... ...... ................... 1.06 22 Dry coal per sq. ft. of fuel bed per hour, lb..................................... 1.25 23 Combustible per sq. ft. of fuel bed per hour, lb ........... .................. .94 10. Duration of the Tests. -Owing to the conditions under which the tests were run, it was not possible to make runs of longer than 36 ILLINOIS ENGINEERING EXPERIMENT STATION twelve hours' duration. Fortunately, however, it will be found, from the discussion under the measurement of the weight of coal fired, that with the exception of the very low capacity tests, runs of this length were sufficient to reduce the probable error in de- termining the weight of coal fired, to about two or three per cent. It has also been noted that owing to the small size of the produc- er, tests of longer duration are hardly practicable. This is due to the fact that toward the end of all 12-hr. tests, the accumula- tion of ash in the fuel bed necessitated such thorough cleaning as to seriously interfere with the uniformity of the heating value of the gas. As the producer is provided with a plain bar grate and all the cleaning must be accomplished by opening the ash doors, this particular type of producer, while operating with a fair degree of satisfaction on runs of not greater than 12 or 15 hours' duration, would not be satisfactory for continuous runs of longer duration. This has been recognized by the builders and was so stated in their specifications. As the object of these tests was to show the actual operating efficiency of the producer, and as under these conditions the pro- ducer would not be run for greater than 12 continuous hours, the present tests have been made to conform to those conditions as nearly as possible. The producer necessarily shows a lower efficiency under these conditions than it would under continuous operations. The reason for this lower efficiency is the' stand-by losses. These are composed of four separate losses, as has been pointed out. One of these affects directly the results of the tests. This is the loss due to the radiation and conduction of heat from the fuel bed and producer during the stand-by period. It results from the fact that at the close of a run or test of 12 hours' duration, the fires are cleaned, and in cleaning, a large quantity of incandescent ash and carbon is removed from the fuel bed; this mass is replaced by the green fuel from above so that after cleaning, the temperature of the fuel bed is low and remains so until the starting of the new run or test. The pro- ducer lining, shell water jacket and water also lose heat during this stand-by period. On again starting the producer at the be- ginning of a test, the fire is blown until a gas sufficiently rich for operation is produced, and the test is then started. The average temperature of the fuel bed and -producer has not reached the GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 37 temperature of normal operation. At the close of the test, this temperature has been reached, consequently, a large quantity of heat must be used in raising the average temperature of the producer at the begining of the test to the average temperature at the close of the test. In the case of the present test, the pro- ducer was not in operation continuously and it was not always possible to bank the fires from one test to another, for this reason. The fire in the producer was usually started on an average of about two hours before the test began. Gas sufficiently rich for operating could be obtained in about thirty minutes from the time of starting the fires, the additional time permitting the producer to warm up. Notwithstanding this, however, the difference be- tween the average temperature of the producer at the close and the average temperature on the starting of the test may have been anywhere from 200 to 7000F. The magnitude of this loss and the extent to which it probably affected the heat balance of tests may be best illustrated by an example. Assume that the average temperature of the producer at the starting of the test is 1000° F., assume that the average temperature at the close of the test is 1600°F. The difference in temperature is about 600° F. The 0 / Sa K LA(W W"2LUY ow>D WNrTN0ow Caw z 0 P&/L af^o /f /R~ f?£1Vf COAL x 4 S C 7 d 9 o // /2 /. /f /N /Ac FIG. 8 total mass in the present case represents about 1200 lb. If we assume the specific heat to be .25, the total heat that has been lost during a run in bringing the average temperature of the producer at the start of the test to the average temperature at the close of the test is 1200 X 600 X I = 180 000 B. t. u. On the high capacity tests this loss is a very small percentage of t o x 0 -3-- 7 Al -1 10_ -n 38 ILLINOIS ENGINEERING EXPERIMENT STATION total heating value of the fuel, on the lower capacity tests it is a very large percentage. Suppose 150 lb. of coal are fired on one of the light load tests. The heating value of this coal is 12 500 B. t. u. per lb. The total heat supplied to the producer is 12 500 X 150 = 1 870 000. The per cent loss due to the above is, there- fore, rso% 0 = 9.6 per cent. As this loss is largely inherent in the operation of the producer, it was deemed advisable to indicate its magnitude and not eliminate it from the results of the tests. The curves of Fig. 8 have been plotted from the results of all tests and show approximately how this loss varies with the weight of combustible gasified. These curves would also seem to indicate that the radiation and conduction losses from the pro- ducer are between 0.7 and 1 per cent, since the curves become parallel to the Y axis in this neighborhood. 11. Forms for Computation and Discussion of Items of Table 5.-Three forms have been drawn up for use in making the computations for the tests. Table 5 is made according to form 1, which is used for the presentation of the results of the test; a blank of this form is shown in the Appendix. Form 2 was made to include all items involved in computation of the results and form 3 contains the derivation and discussion of the formulas used. In working up the tests, the average total quantities are taken from the original log sheets, corrected and placed on form 2. The item number of form 3 refers to the item number of this form. The results are computed in the order of form 3 and placed on form 2, from which the results compiled are placed on form 1. In order to find the formulas used in computing any result given in Table 5, this result must be found on form 2, and bhe item number corresponding to this result on form 2 cor- responds to the item number of form 3 under which the form- ulas used will be found. For example, suppose that it is de- sired to find the formulas used in computing the cold gas efficiency as given in Table 5. Referring to form 2, Appendix, it will be found that the cold gas efficiency on this form appears under item 155. By referring now to item 155 of form 3, the formulas required will be found. One of the important considerations in developing the forms has been to include such items as will give proof of the accuracy GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 39 of the work, and that will point out the nature of any serious errors that are always likely to creep into experimental work. The gas volume has been determined from a previously cali- brated meter. This volume has been checked in most tests by computing the volume of gas from the weight of coal consumed, the analysis of the coal and the analysis of the gases. If the vol- ume of gas as given by the meter is not more than 5 per cent lower than the computed volume, it is assumed that the gas vol- ume as determined by the meter is correct. If there is a greater variation than this, the error may be either in the weighing or estimation of the coal, in the meter, or in the analysis and samp- ling of the gas. If the heating value of the gas as computed from the analysis checks the heating value as given by the Junker calorimeter, the conclusion is that the error is on the side of the coal. If the volumes check, and also the heating value of the gas as determined from the analysis and from the calorime- ter, the conclusions are that the only serious error that can exist must be in the measurement of the temperature of the gases leaving the producer. Even this temperature may be checked by computing from the weight of scrubber water, item 90, and from the rise in temperature of the scrubber water, the heat lost to this by the gases leaving the producer. The word combustible has been used in these tests according to the definition appearing in the Appendix, forms 2 and 3, item 54. As this word has been used in an arbitrary sense by a number of writers and also by the engineering societies, its definition is of considerable importance. The calorific value of the gas is given under items 104 to 105 Table 5 and form 1. The high value is defined as the total heat given out by a cubic foot of gas at 620 and 30 in. of mercury when it is burned in oxygen and the products resulting from com- bustion are brought back to 620 F. The net or effective value is equal to the "high" value minus the latent heat contained in the water that is formed by the combustion of the hydrogen. It will be seen that if the gas is used under such conditions that it is exhausted at a temperature greater than 2120 F., this latent heat is of no use in the cycle of operation. This is the case where the gases are used either in the gas engine or for firing boilers. The low or effective value is given under 40 ILLINOIS ENGINEERING EXPERIMENT STATION item 104a which is computed from the analyses of the gases. Under item 126 is given the grate efficiency. This is a very variable quantity and depends upon the fireman, the size of coal fired, the amount of ash contained in the coal and upon the num- ber of times the fires require poking and cleaning. With large producers and careful firing, this item should not be less than 95 per cent. With small producers of the type used in this work 95 per cent is an excellent result. With the coals used on the tests, from 80 to 90 per cent is probably a fair value for this item. The hot gas efficiency based on the high heating value of the gas is given under item 127. This value varies from 100 per cent principally by the percentage of the radiation, conduction and unaccounted-for loss and by the percentage of the heating value of the fuel lost through the grate. The hot gas efficiency is of no special importance where the gases are used for power pur- poses. The cold gas efficiency based on the high heating value of the gas appears under item 128, again under item 128a, based on the low or effective heating value of the gas. This efficiency for both high and effective heating values has also been based on 100 per cent grate efficiency or upon combustible and is given in items 128c and 128d. This latter value has been computed in order to show certain relations that are independent of the grate efficiency. 12. Under unaccounted-for loss may be grouped the follow- ing losses. a. Radiation b. Conduction c. Loss due to tar and soot and to the absorption of the gases by the scrubber water. d. Loss due to difference in temperature between the initial and final conditions of the fuel bed. e. Minor losses which may be either positive or nega- tive such as the loss due to the sensible heat in the fuel and in the ash. f. Experimental errors made on the tests. I i -* AC -2e- 233212 5222sag8 44 «x a - a.. » - - . . ;a , ass28 s s ass2s 2n n j '-" a 9 333' S p a 3 =- 3-aA-a . gaa a - pta S SagFR s£ gs e- - --- - a;i- * . » S - a s's - ,. S4 saa s S . .I . ; .- . . 4R a i2 2 |t - a3-a -. . 22222 SBE 88 a-S2 2 S2 2 I 2 .2 ",s 2 2 asawas s2 2 as2s a3s S- '| 2 2a8 2-2S 2 S--- a-2-222 22 2223 2 * 2 S aa 2 2 2 283s s2a Rse as- 2 2 53 2g22 12""""2 3 5222 2 2 g '- 'i. a.«»s 3 2 . ... 2sa 2 2 2a2a"2 2. 2i2a32 2 s 2 2 S I,,, "a, ES£BS!a S"85 c s - r - Y- 4S 2.axia d „ g Ia- m ; ,s i ^""i " «- s » s "*. s sa1 s ss as8ses s a E i | i3 ' £ " g a s 23 '-- a-" "s- 23 s2ia3 ! a. a .a a3 2 2 22a 21as2sa ss s2a2 as 2a I2 2 * £ 5 sa 22 2' a -s s."- 2- '2 ssa -; a-2n 2s : is ja 2j3 2 n anI 2a -;ar E a22ese -' s as ss-"sa so assss 8 I__ | In_ a N_ " "___i i-jI______ 1 2 2 2 22 2"";" "-""2- sa S2-1,2 s "-a 22 2 l2";' s B"^' s-"2 as sa,8,g 2 - 5 8"2 4-, ---- 2 - -- 222 23.22 2- z 3; 52'- 23 B;2SS SS3SS2 2 2 h3 2 S2"5»,o o ; ss22 2"2"- 8-»-a22 2 2O *-M I - --2 _______ ;""2 ;2;2;222 5 I" 33 52 2'2 i "" " 22 222222 12 2 I22 3"38 22 B2SK S ft;«s'-2 22 22322 |s smsiB 2 3"" 233 - '" ' 3 "2 222 2 --1. i i ii I i.-a iii i232 32;: 322 "'at! k22a:a" i:8~ :22 ¶ ;-' i % a 3 2~i ii ti" 2312 '-ai Waia s' " '""'" 2 a2 Bai arS ~2:g tl ;a'·2 '$ 2- a¾12' Iti ,2r1O& Z 222. [a0 < a.8 a .Q .482 b a~a. ..l28 " i 2.a22 :a --;282882 '..a~aa822! I $~ B | Sa "nn--oca 2 22 I I || I 2"sss^ " - - 2 22 2 3"53~~ . '- " I ,. a ..afxi S2 i3'; " S S "a---- 2 29 2 3"jt3':2 2 2a^»«' is 23 S | i |~ r Iz~adlsa '' '"g "'aaaaaaaSl 2 sb32 s I «»*!B^ 2 S"! -S as 22 2 2 2 - 3I 2 U¾; "% j i IP I 8 ' "- 3" S t3312a33'^^ - -,-,.,,„ a9 2 - S , a- . ,,.8 ,s Sl SS ~ B 82a s ~ E3^*a i '- | as s - In·a **S f 2-eS 22llS S § E SSsN'SS' '' - g ^«n«eg SS S S - -P H 3 ^. S~ 35 :- _E f ^..^.rs'si 2 ;i2 2, 2"if'1222~ £8 22 32 a. 12 33. 2 2 A3 3 3 3~ 2 ~a. a - 2:% .2"2~ ~2 2;23 "2" 2"' 2"" 12 a22A C '2 S 22 22 2 1122 N: a w l8 SK Sp .S * ma« a 5aa XSi « . s - 22 aa a a " - 2 2S aS3 2a "'2 "'22'a 2;23 '33ss'a -a8 8 ,221 223 "--3" -a - "'23r " S'52 at 32 ¥s 2 33 - '- 3 2" ' 22 12 -2321 S "23 "32"''0 a . 2a2 E a 22 i .,,1 a.a22s - - .^2 -a a 2 A 2 ' "22a 223'-' a - - - g; 2g 5. aa aa 22'l sS ' 2 ^=sS "2 S s 3 2 2 a» %2a a |- 2'- < 3 g " ' a 'ET " 'ai- - - - 2; -; - .- 22a 2" s aa A -l?'-! - ,- -~ 22 s aa- a a "22 sl" -- 2 2 2" 12"' -S ;:- a a a S 3^ 1 "'at A .all a^a aaaa 33 "a;2-; si1 'a sa assss ~ ,. rEY a2 S" - 3 2 2~~~4 a22!aa-s - $ 28~22fIj'; 22222 asaa "a.22 sa2e2 2 .., 2 *sawi 2..-"2 2r-s a 2"2'-S c I22o 221212 :222 222a 22"3 222a 2 1221 2 32' 12 32. 2 32 2 2' C" 4siUrIl 4rnl 1 3 ::: ,a . 2 ," ' A' 3 a'tssil ta.rg 2a~ l sj ° "S2s~az Sa S O"S53 5 S2 2 ap sj 5·. J: 8 22 ; 212 12-.- 2 22 2 122 2 212 12 22 12 212 12 122 2 2; 2 Aa 2 22 12 22 2 22 12 23 2 22 2 22 .a2 2~a 2"", a a 2i'2 2 a "''2 32 - 2" a a'-2 aa 12 a a at """at.. 2 '3 2 a a 122223 2 - l or -aa22 "' -' -a 8 ~d ·- 2 22 222 s 2 2 :: 122; y 3c 2 ?3 sS 132" .212 g'a"" 1 "2 I3E I" a2 2 12 2 2 12 S 2 S* IS 2 at" i Is2 aNs 32 A93very low temperatures, and by the time the temperature of the fuel bed rose to an efficient working temperature, the formation of clinker on the lining of the producer was so rapid as to inter- fere with the quality of the gas. By the use of larger quantities of steam, or at lower capacities, it would doubtless be possible to use this coal in a producer of larger size with satisfactory results. With a producer of the size and type used on the tests, satisfactory operation is practically impossible. This series of tests is of little value except to illustrate the effect of a fuel containing a fusible ash upon the operation of a small producer. Owing to the formation of clinker on the sides of the producer, the fuel bed is greatly reduced in area, and there is the tendency for the formation of holes which permit the air to rush through, resulting in high temperature and poor gas. Owing to the frequency of poking, large quantities of incandescent coal drop into the ashpit, thus reducing the grate efficiency. Fig. 9 is a graphical log of test 6. (2) Tests with Lehigh Valley Anthracite.-Tests 15 to 21 inclu- sive were run on Lehigh Valley anthracite, chestnut size. This was extra large chestnut and clean in appearance. Ninety per cent of this coal would pass through a 2-in. mesh, while all of it would practically pass over a 11-in. mesh. The average percent- 42 ILLINOIS ENGINEERING EXPERIMENT STATION FIG. 9 GRAPHICAL LOG TEST No. 6' 56IF118,6E;T W,974ýF -ZO6- 4ýý96--alx-r 6 711 IAI"6- bye-lVes WI?72F-/P 7L'W--fý,9 74,M-6- 1 -0 -w-X N I J,/ -,,-Z It GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 43 age of ash as taken from the analysis was approximately 15 per cent. The ash taken from the producer at the close of the tests was soft and white and contained small pieces of clinker. There was practically no tendency for the formation of large clinker arch- ing across the fuel bed at the higher temperature. The conditions that maintained in this series were fairly uni- form, considering the size of the producer. Fig. 10 and 11 show the graphical logs for tests 17 and 20, respectively. FIG. 10 GRAPHICAL LOG TEST NO. 17 44 ILLINOIS ENGINEERING EXPERIMENT STATION FIG. 11 GRAPHICAL LOG TEST No. 20 The average heating value of the gas for all of the tests was 110. The variation of the heating value on all of the tests, while probably sufficient to cause considerable trouble in the operation of an engine driving electric lighting machinery, would not inter- fere seriously with the operation of an engine where close regu- lation was not a prime consideration. (3) Tests with Scranton Pea Coal.-Tests 23 to 27, including tests 31, 32 and 38 were run on Scranton pea anthracite. This coal was clean in appearance and like the other two coals, con- tained on an average about 15 per cent ash, while the calorific value corresponded to the calorific value of the other coals and GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 45 was about 12 800 B. t. u. per lb. of dry coal. It would all pass over a 1-in. mesh and through a 1-in. mesh. Analysis of the ash is given under test 23. This shows it to be an ash consisting of about 96 per cent of the oxides of silicon, iron and aluminum, while the remainder is made up of the oxides of calcium and mag- nesium, which are present in insufficient quantities to make the ash readily fusible at the temperature maintaining in the fuel bed. The ash as taken from the fuel bed was soft and fine, mixed with small pieces of clinker. There was very little tendency to- ward arching and the tests were the most satisfactory of the series. The conditions were as uniform as could be expected, FiG. 12 GRAPHICAL LOG TEST NO. 25 46 ILLINOIS ENGINEERING EXPERIMENT STATION and the efficiency reasonably high. Fig. 12, 13, and 14 are graph- ical logs from the tests of this coal. FIG. 13 GRAPHICAL LOG TEST NO. 31 Fig. 13 shows the results of test 31, which was a variable load test run for the purpose of showing the effect of suddenly varying the rate of gasification. For the first three hours of the test, the producer was operated at the rate of about 3300 cu. ft. of gas per hour; for the next three hours at the rate of 1155 cu. ft.; for the third 3-hr. period at 4500 cu. ft.; and for the last 3-hr. period at a rate about 2200 cu. ft. (4) Tests with Gas House Coke. -Tests 28, 29, and 30 were run on GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 47 FIG. 14 GRAPHICAL LOG TEST No. 32 gas house coke containing about 14 per cent ash. The appearance was dirty and the size very irregular. Thirty per cent of this would doubtless pass a h-in. mesh, 50 per cent a 2-in. mesh, while the remainder would pass a 4-in. mesh. There was some trouble due to the formation of clinker, and considerable trouble caused by a tendency of the coke to pack in the hopper, thus preventing its descent into the fuel bed. In spite of this, the conditions that maintained on the tests were fairly uniform, and it is believed that very little difficulty would be experienced in using a coke of uniform size and quality in this type of producer. Fig. 15 is a graphical log plotted from the re- sults of test 80. 48 ILLINOIS ENGINEERING EXPERIMENT STATION FIG. 15 GRAPHICAL LOG TEST NO. 30 (5.) Test No. 14.-Test 14 was a special test run on the Phil- adelphia and Reading coal to determine if the ash in this coal could not be made sufficiently fusible to run from the fuel bed. This test will be discussed under the formation of clinker. 14. Percentage of CO2.-The percentage of CO, in the gas is ordinarily considered to be a measure of the efficiency of a producer, low percentages indicating high efficiency while high percentages indicate the reverse in efficiency. This is to a certain extent true in most producer practice, but it is not necessarily so, as the fol- lowing will illustrate. If the heat resulting from the formation of CO. is utilized in the decomposition of the vapor of water for GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 49 the production of H,, which enriches the gas, then no appreciable loss results. If, however, the CO, results through poor operat- ing methods, such as insufficient or improper poking, resulting in holes in the fuel bed and thus permitting the air to rush through, then the presence of a high percentage of CO2 is indicative of a low efficiency, and this condition will manifest itself in excessive- ly high temperatures of the leaving gases. This latter condition is illustrated by the results of all the tests on the Philadelphia and Reading coal (see Table 5). The temperatures of the gases leaving the producer, item 39, are ex- cessive, the CO2 item 117, is high, while the efficiencies, items 126- 128d, are low. The former condition, in which both the CO2 and the effi ciency are high, is illustrated by the tests on the Lehigh and Scranton coals. The principal results of test 23, which have been placed in Table 6 for convenience in discussing, illustrate this condition. The per cent of CO in the gas in this test averaged 10.15. The temperature was, however, low, 800'F., and the efficiency high, thus showing that the fuel bed was in good con- dition but at so low a temperature, due to the use of a large amount of steam in the air fed to the producer, as to produce a predominance of the reaction 3 which has been discussed in page 6 of the theory. Test 24, also placed in Table 6, shows practically the same efficiency but with 4.04 % C02 in the gas. This is due to a high temperature in the fuel bed which tends to cause reaction 4 to predominate, resulting in a high percentage of CO, practically the same temperature of the leaving gases as was obtained in test 23, and practically the same efficiency. I. S. *0 5- EFFIC/ENC/ES BASED ON H/6H HEAT VA4UC rI SCRANTON PF4 COA4L f 0 ,.fEHI6H V4LLEY eOAtL 19 sa PHIL AVO EffFPING CoAL I i 70 ' - -- - --O-- - -- - -- CARSON RATIO If iv I ' FiG. 16 50 ILLINOIS ENGINEERING EXPERIMENT STATION The curves of Fig. 16, 17, 18, and 19 also show that the per cent of CO2 has very little effect upon the efficiency of the pro- ducer. The theoretical efficiency curve showing the efficiency for different percentages of CO on the assumption that the heat k k coIN 6Ws vRcAvr ~Y vo/ vA- FIG. 17 C02-EFFICIENCY CURVES FOR PHILA. AND READING COAL <3 <3 '4 (4 -4 <3 CO /IN GAS PERCENT BY Yr/LUMe FIG. 18 CO--EFFICIENCY CURVES FOR LEHIGH VALLEY COAL liberated through the formation of the C02 is lost in sensible heat, has been drawn and is shown in Fig. 19. The difference between the actual curves and the theoretical curve is a measure of the amount of heat conserved through the decomposition of water vapor. The efficiency curves on this figure have been affected by the unaccounted-for loss which has been discussed on page 37, and has resulted in lowering the points in the low capacity tests to /00 EFF/CIE£HCIE BASED ON H/6H HEAT VrL. I4 8ASE ON /0 X G/.ATe £FF 90 - - fO ACTUAL. EFFICIEmCY 80 Boo 7o - 7 - -- --a _ 60 - 4oo S' 6 7 8 9 /0 GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 51 such an extent as to make these curves concave to the axis. The curves of Fig. 18 are also affected by the unaccounted for loss, but in this case the gas from the low capacity tests ran high in CO, owing to the use of a large quantity of steam, which resulted in causing the curves to drop off at the high percentage of CO2 faster than they otherwise would. The curves of Fig. 16, showing the relation between the effi- ciency and the carbon ratio, or the ratio of the weight of carbon to the weight of hydrogen in the producer gas, illustrate the point under discussion, showing that the efficiency of the producer is practically independent of the amount of hydrogen in the gas, consequently, of the per cent of CO in the gas. Fig. 20 shows how the per cent of CO2 in the gas depends upon the weight of water decomposed per pound of combustible. I a BASED ON loo/%, TEEFt. H H/OH HEAT VALiU H0 ACTUAL EFFICIENCY -- THEORETICAL- EFFICIENCY -Tr-_r-rrrrr CO, IN &AS- PerCEAFNT Y VOLL/AfE FIG. 19 CO2-EFFICIENCY CURVES FOR SCRANTON PEA COAL 15. The Effect oj the Size of the Fuel. -The determination of the effect of the size of the fuel upon the composition of producer gas and also upon the efficiency of the product is complicated by the difference in operating conditions that may not be recognized and upon the difference in the surface effect of the incandescent fuel upon, first, the reactions of oxygen and hydrogen and later on the reaction of CO on this fuel. The exact nature of the surface 00 0 0 4 S- e 7 6 9 /0 8 11 10 52 ILLINOIS ENGINEERING EXPERIMENT STATION effect, which is called "catalytic action", and under which a number of different effects are grouped, is very little understood at the present time. It probably results from chemical reaction on the surface of solids in contact with gases, which depend upon the state or form of the solid, such as charcoal, lampblack, coke or FIG. 20 anthracite coal, and in the case of these fuels upon any impurities contained in them. The result of catalysis is to hasten any reac- tion that may be taking place between the solids and the gases in contact with them, or between the gases themselves. It is be- lieved, however, that the ultimate extent or that the equilibrium of the reaction is not affected. It is probable that the difference in the catalytic action of different fuels, such as were enumerated above, may be very great. It is also probable that this effect is not so great between the same types of fuel, as, for example, between two grades of anthracite, or between two grades of coke. The ash in the fuel, as above noted, may also have a decided effect upon this action. . The difference in the catalytic action of charcoal, coke and anthracite is illustrated by Fig. 21, 22, and 23, which are taken from Dr. Clement's work, "On the Rate of Formation of CO in Gas Producers".' The results were obtained by passing CO2 over charcoal, coke, and anthracite coal, respectively. In each case the material was ground to a certain size, about 0.2 in. on the side, and placed in a porcelain tube that could be maintained at different temperatures by means of an electric current, traversing a wind- ing of nickel wire placed around the tube. The velocity of the gas, the per cent of CO formed, and the temperature of the tube 11Bull. No. 30, Eng. Exp. Sta., J. K. Clement, L. H. Adams and C. N. Haskins. hi I i> t ^ s GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 53 were the variables. Fig. 21 shows the relation for charcoal be- tween the per cent of CO formed and the velocity of the gas through the tube. Fig. 22 and 23 show the results for coke and anthracite coal, respectively. The velocity, which is the recipro- cal of the time of contact, is expressed in feet per second. /1f'C/iPfocRL OF T7*AI or cow/TWcr FIG. 21 FIG. 22 It will be seen from these figures that the percentage of CO formed by the reaction C02+ 0 =- 2 CO, depends upon the tem- perature, upon the velocity or the reciprocal of the time of contact 54 ILLINOIS ENGINEERING EXPERIMENT STATION of gases with fuel, as it has been expressed on the figures, and upon the catalytic action of the fuel. This latter effect is shown by the difference in the per cent of CO formed at the same tem- perature and under the same velocity for the different fuels used .o .-o .30 .4W .50 .60 .70 .So .90 V .oo o-o e 7t r, rF o'-- c-awrcr - FIG. 23 in the experiments. For charcoal, Fig. 21, the per cent CO for a velocity of 1 ft. per second and a temperature of 11000 C., is 85 per cent, for coke and anthracite the percentages under the same condition, (Fig. 22 and 23) are 9 and 11 respectively. For a tem- perature of 13000 C. and a velocity of 1 ft. per second the percent- ages of CO formed are 100, 74, and 45 for the above fuels. Tests 18, 19, 24 and 25 have been selected as representing most decidedly the effect of the size of the fuel upon the results of the performance of the producer. Tests 19 and 24 were run under practically the same operating conditions, the only known variable being the size of the fuel, catalytic action excepted. Tests 18 and 25 were also run under essentially the same conditions of operation, with the size of the fuel constituting the only known variable. The principal results of the tests have for convenience been transferred from Table 5 to Table 6. GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 55 TABLE 6 1 Test N o ............. ..................... 19 24 18 25 23 2 Fuel. ...................................... Lehigh Scranton Lehigh Scranton Scranton Size of fuel, mean diameter in in.... ... 1Y% V 1% I % V 3 Capacity in cu. ft. per hr ................. 2730 2941 4545 4795 2980 4 Water decomposed per lb. of combust- .429 .358 .446 .355 .689 ible .... ..................... 5 Ratio of water decomposed to water supply .323 .429 .408 .639 .402 6 Percent C02 in the gas...... .............. 9.4 4.04 9.44 4.2 10.15 7 Per cent H2in the gas .................... 9.3 10.80 8.95 10.40 16.96 8 PercentCOinthe gas................. .... 17.43 27.3 17.13 27.01 17.39 9 B.t.u. per cu. ft. (High) ................. 106.9 138.2 101.5 138.1 128.8 (Low or effective) 101 131.6 96,0 131.7 119.4 10 Cold gas effi,.iency based on 100 per cent grate efficiency (High)................ 66.9 77.2 72.7 79 9 75.4 (Effective).................. ......... 63.3 73.5 68.6 70.3 69.s 11 Per cent loss in sensible heat and in mois- ture ................................... 19.3 12.7 22.7 14.8 15.3 12 Unaccounted- for loss ..................... 11.6 9.4 2.1 4.4 6.9 13 Temperature of gases leaving the producer °F ........................... ....... . . 970 809 1201 1108 800 14 Temperature in the fuel bed °F, average.. 2200 2300 2300 1740 15 Equivalent velocity of gas, feet per second 8.7 3.14 14.46 5.12 2 It will be seen from this table that the per cent of CO., for tests on the Lehigh chestnut is about 9.4 per cent, while the per cent C02 for the tests on the Scranton pea is about 4 per cent. The heating value of the gas is low in the former tests, as is also the efficiency, while the temperature is a little high. The relative per- centage of C02 in the gas in the case of these tests is therefore to a certain extent a measure of efficiency of the producer. The difference in the percentage of C02 appearing in the gas from the two coals in the tests under consideration may be due to two causes, viz., (a) the difference in the catalytic action of the two fuels, as above discussed, and (b) the difference in the size of the fuel. In the case of the Scranton pea which, as has been noted, would pass a i-in. mesh, the amount of catalytic surface exposed to the action of the gases is much greater than the sur- face exposed in the case of the Lehigh coal, which will pass over a 11-in. mesh and through a 2-in. mesh. The following will illus- trate this. The mean diameter of the Scranton coal may be taken 56 ILLINOIS ENGINEERING EXPERIMENT STATION as I in., the mean diameter of the Lehigh coal may be taken as 1j in.; since the number of pieces contained in a given volume is inversely as the cube of the mean diameter, and since the surface of each piece is directly as the square of the mean diameter, the ratio of the total surfaces exposed in the fuel bed of the producer will be as 3 and 1 for the Scranton and Lehigh coals, re- spectively. Theoretically, the size of the fuel will have no effect upon the velocity of the gas through the fuel bed for the same rate of gas- ification. This is due to the fact that the number of pieces of coal required to fill the producer varies inversely as the cube of the mean diameter of the pieces, while the volume of each separate piece varies directly as the cube of the mean diameter, the volume of the coal present, and the volume of voids; consequently, the velocity of the gas through the fuel bed is independent of the size of the fuel. That the voids are, in the case of the two coals, practically of the same volume, has been determined by water displacement. The volume of voids, in the case of the Scranton coal, was found to be 46 per cent of the volume occupied by the coal and 43 per cent for the Lehigh. The velocity of the gases through the fuel bed is therefore practically the same for the two coals for the same tem- perature and rate of gasification. The amount of draft required to drive the gases through the bed will, however, be different, ow- ing to the increased frictional resistance due to the smaller voids in the Scranton coal. The purely mechanical effect of the increase in the size of the fuel will be, from the above, if the volume of the fuel bed remain constant, to decrease the area of surface exposed to the gases, to decrease the number of pieces of coal in the path of the gases, and to decrease the drop in the pressure of the gases through the fuel bed. The decrease in the drop in pressure may be assumed to have no effect upon the composition of the escaping gases within wide limits. The extent of the effect of decreasing the area of the catalytic surface exposed to the gases and of decreasing the num- ber of pieces of coal in the path of the gases upon the composition of the producer gas generated and upon the reaction within the fuel bed is unknown. Dr. Clement's work shows the effectof the velocity of the gases upon the reaction of CO with incandescent GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 57 carbon. Our tests No. 24 and 25, Table 6, on the Scranton coal, seem to show the effect of the velocity of the gases on the perfor- mance of the producer, tests 18 and 19 also show the same. It will be noted from the above that the increase in the rate of gas- ification of about 100 per cent has not resulted in any marked change either in the composition, or the heating value of the gases in the case of tests 24 and 25. Tests 18 and 19 on the Lehigh coal show a slightly greater difference, and it is believed this is due to the lesser catalytic surface which results in a shorter time of contact of gases with fuel. From these tests and the curves of Fig. 21, 22 and 23, and beyond the velocity of 1 ft. per second (the curves are drawn only to this velocity) which corresponds to the velocity of the gases obtaining in the fuel bed of the producer, it will be seen that the effect of the velocity on the reaction of C02 on carbon is small. As before pointed out, the difference in the velocity of gases through the fuel bed for the same rates of gasification for the two fuels, if any, must be small. The great difference in the CO2 or in the composition of the gases from the two fuels for the same rates of gasification as above noted, must therefore be due to one or more of the following causes: (1) to the greater catalytic surface exposed to the gases as in the case of the Scranton coal; (2) to the greater number of CO molecules coming in contact with the coal,due to the greater number of pieces; (3) to the smaller size of the voids in this coal which would tend to cause a more intimate mixture of gases with the coal; or (4) to a greater catalytic effect. It has been remarked that it is probable that the difference in catalytic action between two grades of anthracite coal is small. We will therefore assume that the difference in composition of the gases from the two fuels under discussion, when the rates of gasifi- cation and operating conditions are the same, is due to the differ- ence in the amount of surface exposed to the action of the gases or to the difference in the number of pieces of coal in the path of the gases or to a combination of the two, or, in other words, to the effect of the size of the fuel. The above conclusion is also supported by operating experi- ence. Shallow fuel beds, which are equivalent to the use of large fuel, or small catalytic surface, produce high CO, high tempera- ture of gases and low CO. The tests 18 and 19 show these condi- tions. The efficiency does not drop off, consequently the tempera- 58 ILLINOIS ENGINEERING EXPERIMENT STATION ture of the gas is not high; for the excessive heat liberated, as evinced by the high CO, is utilized in the production of hydrogen. 16. The Effect of Capacity.-The effect of capacity, or of the rate of gasification upon the efficiency of the producer, when op- erating with the different fuels, is illustrated by Fig. 24, 25 and 26. FIG. 24 CAPACITY-EFFICIENCY CURVES FOR PHILADELPHIA AND READING COAL CAP,4CITY- CUF' STD. GAS PER HOUlr. FIG. 25 CAPACITY-EFFICIENCY CURVES FOR LEHIGH VALLEY COAL As there were only three tests for the gas house coke, no curves have been plotted from these tests. For each of the other fuels, three curves have been plotted, one for the actual efficiency GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 59 based on high heating value of the gas, one for the efficiency based on 100 per cent grate efficiency, or, which is the same, based on combustible and upon the high heating value of the gas, and a third based on combustible and the low heating value of the gas. This latter curve is shown dotted on the figures. Fig. 24 shows the curves for the Philadelphia and Reading coal. The curve for the actual efficiency drops off at the higher capacities. This is principally due to the more rapid formation of clinker, as the capacity and the temperature increase, which re- sults in the necessity for more frequent poking, consequently, a much higher grate loss. The curve for efficiency based on 100 per cent grate efficiency and the high heating value of the gas drops off slightly. The curve based on 100 per cent grate efficiency and upon low heating value is practically a straight line. The curves of Fig. 25 show the results for the Lehigh Valley coal. The drop in the curve for the lower capacities is partially due to the sensible heat lost to the producer walls and fuel bed, on ac- count of the lower temperature of the fuel bed at the start of the test. This has been pointed out in the discussion of the unaccounted- for losses. If a correction were made for this unaccounted for loss, the curves would approach a straight line. The drop at the other end of the curve is largely due to the effect of the velocity of the gases through the fuel bed, or to the shorter time of contact of gases with fuel owing to the larger size of this fuel. CAPACITY- CU FT STD. 6AS PER HOUR FIG. 26 CAPACITY-EFFICIENCY CURVES FOR SCRANTON PEA COAL ILLINOIS ENGINEERING EXPERIMENT STATION The curves for the Scranton pea coal, Fig. 26, show very lit- tle tendency to drop off at the highest capacity. This, it is be- lieved, is explained by the small size of the fuel, consequently, by the larger area of fuel surface exposed to the gases. The rapid falling off of the actual efficiency curve on these tests is due to the sensible heat lost to the producer walls and fuel bed as previously noted, and also to the fact that in the case of the Lehigh and Scranton coals, which had no tendency to clinker, that the weight of ash and refuse tended to remain con- stant and was independent of the capacity. As the composition of the ash and refuse remained constant, the actual grate loss tended to remain constant, which would, therefore, cause the actual efficiency at light loads to drop off. The curve based on 100 per cent grate efficiency is practically a straight line slanting toward the low capacities; this slant is also due to the sensible heat lost to the walls and the fuel bed of the producer. The results of Table 6 bear out the conclusion drawn from the curve as to the effect of capacity. These results together show that capacity, or the rate of gasification, within very wide lim- its has little effect upon either the efficiency, the composition of the gas, or upon the heating value, so long as the steam supplied to the producer is so regulated as to maintain the same temperature within the fuel bed. 17. Goal Per Square Foot of Grate Area-The weight of coal gasified per square foot of grate area is related directly to the rate of gasification and depends upon this. Owing to the general use of the term and to the interest attached thereto by engineers, it probably merits a separate discussion. It will be seen from Table 5, item 66, that this quantity varies in the tests on the Lehigh and Scranton coals from 7.67 lb. per hour, in the case of test 21, to 49.8 lb. in the case of test 17. Within this range, the producer has been operated with satis- factory results. Anthracite producers in this country are rated on a basis of from 10 to 15 lb. of coal per sq. ft. of grate area per hour by the manufacturers. In European practice, the rating is from 20to 30 lb. of coal per square foot of grate area. This difference in rat- ing is doubtless due to the difference in the fuels used in the two countries. From the results of the present tests, it must be concluded GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 61 that the coal gasified per square foot of grate area depends almost entirely upon the nature of the coal. Were it not for practical con- siderations there seems to be no reason why the coal gasified should depend upon any condition other than the depth of the fuel bed. From the works of Dr. Clement previously referred to, and of Boudouard and others, it will be seen that a ceitain length of time is required for the completion of the reaction within the fuel bed of the producer, also that the time required for their comple- tion depends upon the temperature and upon the catalytic action of the fuel. Consequently, in order to attain high rates of gasifica- tion, it is necessary that the fuel bed be increased in depth in or- der that the proper time of contact of gases with fuel obtains. In the actual operation of the producer with a given depth of fuel bed, the coal gasified per square foot of grate area depends upon the amount of ash, the nature of the coal, and upon the cat- alytic action of the fuel. The larger the percentage of ash in the fuel, the more fre- quently the producer requires cleaning, so that with high rates of gasification the removal of the ash alone would place a limit. The nature of the ash, that is, the temperature of fusing, also places a limit for a given depth of fuel bed. An ash fusing at a low temperature requires a low temperature within the fuel bed of the producer. This results in requiring a longer time of contact of gases with fuel in order that the reaction within the fuel bed may be completed. A fuel with slow catalytic action pro- duces the same result. Since, from the above, the coal gasified per square foot of grate area depends upon the depth of the fuel bed, we have used the expression "rate of descent of dry coal through the fuel bed" or " coal burned per cubic foot of fuel bed", to express the per- formance of a producer in addition to the expression under dis- cussion. 18. Clinker.-The formation of clinker is due to the presence of incombustible matter in the ash which fuses at the temperature maintaining in the fuel bed of the producer. There are two ways of dealing with this, the more practical one being the operation at such capacities and with such an amount of water, or CO2, in the case of the carbon monoxide producer, as to keep the tempera- ture below that of the fusing point of the earthy matter contained 62 ILLINOIS ENGINEERING EXPERIMENT STATION in the fuel. In doing this, the efficiency of the producer will be somewhat lowered, as the reaction of CO on carbon will be greatly lessened, and large quantities of steam must be used to keep the temperature down, through the formation of hydrogen and the heating of the moisture that escapes decomposition. The heating of this moisture results in the lowering of the efficiency of the producer. The curves of Fig. 27, 28 and 29 indicate that the ^ ~_ =__ -- -- - -t_ Xa as9so 0 &o 100 ^ "a O Wrr" £46/1 iv'My Zr _ 0 M, Mcr. fI-ev - ror/h p,-v' '^0 II 700 900 //00 /300 /150 1700 7rAf,?W'7 -, A- 97ff L Z.'V/A9' RPuVC6A0 - D5., FIG. 27 TEMPERATURE-EFFICIENCY CURVES FOR PHILADELPHIA AND READING COAL 0 0 k I$ if EFF/clENC/ES SASED ## &/5// HEAT VA HO ACTUAL EFF/c/ENCY #4o i o0 7oo 900 1100 1500 rEMPEfATULRE OF GAS LEAVINM/ PO/PUcEff - DEG F FIG. 28 TEMPERATURE-EFFICIENCY CURVES FOR LEHIGH VALLEY COAL L^oM --.---- ------- SEFFCINCES BASED ON/ H/IH HEAT VALUE 0 I A BASED 0N 1F00'/, rRATE EFFIC/ENCY f0O ACTrUAL EFFHlIE/VC ' °4 .H 3* 500o 700 soo loo 30o0 TEMPERATULRE OF 6A5 LEA VING PRODVCER -DEC F FIG. 29 TEMPERATURE-EFFICIENCY CURVES FOR SCRANTONPEA COAL -- --'-- -ST 0 n 0 F' LC .1 1 GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 63 efficiency of the producer tends to increase with the higher tem- peratures of operation. This increase, however, is comparatively small within certain limits and since reliability of operation is the desideratum, and not necessarily efficiency, the obvious means of decreasing the irregularities due to the formation of clinker is by operating at lower temperatures. It is possible that fuels may be found in which the formation of clinker begins at a temperature so low as to prevent the reactions within the producer taking place. This is, however, not probable. The water vapor in practically every case tends to disintegrate both fuel and clinker. If an excess is used, it will tend to cause the fuel bed to become mushy and will result in poor gas, and a shut down if engines are operating on the gas. A second method of dealing with a coal containing a fusible ash is to add a flux in the form of a limestone or other cheap ma- terial which will render the clinker so fusible as to cause it to run from the fuel bed of the producer. Test 14 on the Philadelphia and Reading coal was run in this manner. It was determined from an analysis of the ash that about 60 lb. of limestone (CaCO8) would be required per 100 lb. of coal in order to form a highly fluid clinker. The fuel charged on this test was, therefore, mixed with limestone in the ratio of 100 lb. of coal to 60 lb. of limestone. As the limestone when heated gives out 44 per cent of its weight in CO, this CO0 takes the place of the water vapor and keeps the temperature down through the CO, reacting on the C to form CO. The efficiency of the test is low, due partly to the absorption of heat in driving off the CO from the limestone, but largely to the loss through the difference in temperature of the fuel bed at the start and close of the test. The graphical log sheet, Fig. 30, shows the uniformity of conditions that maintained. The fuel bed required practically no poking during the test, as the clinker fused with the limestone and trickled into the ash pit. At the close of the test, however, when the fire doors were opened for cleaning, the entire fuel bed "froze", and had to be broken up at the expense of considerable time and labor. The above method is mentioned simply as a matter of interest, and indicates a possible though hardly practical means of dealing with clinker. 19. Weight of Water Required for the Producer.-The weight of 64 ILLINOIS ENGINEERING EXPERIMENT STATION water required by the producer depends very largely on the fuel that is being used, and to a certain extent upon the proportions of the producer. With a fuel that has no tendency to clinker, the highest efficiency will be obtained by using such an amount of water as sCWSEr wA -9LBS. Jrdg/dWo - CV-fT AT. 6s /,W£-SI-M- ,cWS gwr*v rrM-W"r ff rEPrv4'wE of rws , , I I ', s FIG. 30 4-A ----^--y-----------------^------- __ -- %% H "N XIIi A- ---  -- .-- A -- 1- -" -A.^^ .--'-§ -- _ >i -_-_A\ _ ^ .4->- ^^ - - -V- -- ' _ _ < L .._ _.- -^ - j . ___ fc===dJ~fc^±-4;-===N GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 65 to result in between 2 and 4 per cent CO2 in the gas leaving the producer. This of course, requires a deep fuel bed and a high temperature within the decomposition zone to insure the comple- tion of the CO reaction. It is a condition of operation that can scarcely be obtained in a producer as small as the one that has been used on the present tests, for the reason that while there may be no formation of clinker, there will be, owing to the small diameter of the fuel bed, a tendency for the fuel to pack, which will necessitate poking from time to time in order to insure the proper descent of the fuel. This packing or arching of the fuel, as it decreases the time of contact of gases with fuel, results in increasing the percentage of C02, and raising the temperature of the gases, which tends to lower the efficiency. If the producer had been provided with some form of shaking grate, which is quite possible in pro- ducers of this size, the efficiency would have been increased several per cent and the uniformity of operation increased much more. In the case of fuels containing an ash fusing at temperatures close to 2200° F., it will be necessary to use sufficient water to lower the temperature of the fuel bed below that temperature at which the formation of clinker begins. This may result in a slightly lower efficiency of operation, but as it is usual that such operation is necessitated through the use of a poor, consequently a cheap fuel, the efficiency is of less importance than the relia- bility of operation. In the case of such operation, the hydrogen in the gas will be high, the COM high and the CO low. Test 23 of Table 6 as compared with Test No. 24 shows the effect of operating with a large quantity of moisture, on the CO,, heating value and efficiency. As has been indicated a number of times, the reaction of water on incandescent carbon to form hydrogen takes place much faster at the lower temperature than does the CO reaction, conse- quently, when a producer is operating with a shallow fuel bed, or with a larger fuel, as in the case of the Lehigh coal, or with a fuel of slow catalytic action, in order to keep down the temperature of the gases and to keep up the efficiency, it will be necessary to use larger quantities of water than would be the case when operating with a deeper fuel bed, as there will be insufficient time for the completion of the CO reaction unless the producer is operated at 66 ILLINOIS ENGINEERING EXPERIMENT STATION a greatly lowered'capacity. '.4 '41 '4- '4- ~&1 S WATER DECOMPOSEP PER LB OF COM5BJTrBLE -LBS. FIG. 31 The curves of Fig. 31 show that within the accuracy and range of these tests, there is practically no effect upon the efficiency of the producer due to the increase of the water decomposed. If the range were increased in either direction, there would be a great falling off in efficiency. W*t1? O~CO4'PO$4D ~ £5 OP C usr/&f - ZBS FIG. 32 Fig. 32 shows the effect of the water decomposed per pound of combustible upon the volume of gas generated per pound of combustible, and Fig. 33 shows the effect of the gas generated per pound of combustible upon the heating value of the gas per cu. ft. 20. The Effect of Fuel Bed Temperature. -The effect of the fuel bed temperature upon the composition of producer gas and upon the efficiency of the producer from the previous discussion has been pointed out. From the results of Harries' experiments, (see page 6). the experiments of Dr. Clement, illustrated GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 67 by Fig. 21, 22, and 23, and from the results of Boudouard's experi- ments, the general effect of the fuel bed temperature can be pre- dicted with considerable certainty. The exact effect, as it de- pends upon the catalytic action of the fuel, the size of the fuel, and other variables, can not, however, be determined for the different fuels without direct experiment. Within comparatively GRS &-ey/e.4PR7-o 40. V. o Z coZ8vsr/a e - CV A-' FIG. 33 wide limits, and, speaking in general, the efficiency of the pro- ducer, as has been pointed out, is practically independent of the amount of steam fed to the producer or the amount of H2 in the gas. With low fuel bed temperatures, it will be seen from Dr. Clement's curves that the CO will be low and the CO2 high. From the experiments of Harries, it will be seen that the reaction of HO20 on carbon has reached equilibrium at a temperature as low as 12400 F. with 8.41 per cent of H2 formed. So therefore in the formation of H2 at low temperatures and the formation of CO at the higher temperatures, the principal effect of changing the temperature of the fuel bed is to cause a shifting in the relative percentage of CO, C02, and H2 present in the gas generated. If the change is so made that there is no great difference in the tem- perature of the leaving gases, the efficiency of the producer is very slightly affected. Tests 23 and 24 of Table 6 illustrate this. The fuel bed temperature in the case of test 23 is 17400F., the tem- 68 ILLINOIS ENGINEERING EXPERIMENT STATION perature of the fuel bed for test 24 is approximately 22000F. The temperature of the gases leaving the producer is practically the same, as is also the efficiency. The percentage of CO has shifted from 27.3 to 17.39, while the H2 has shifted from 10.8 to 16.96 per cent. The heating value of the gas for the lower tem- perature has fallen off, while the number of cubic feet of gas per pound of combustible has increased from 92.6 for test 24 to 101.41. This latter is due to the lesser density of the hydrogen. The fuel bed temperatures, as obtained for the different tests, are given in Table 3. The radiation and conduction losses will probably not exceed 1 per cent. The loss due to tar and soot and to the absorption of the gases by the scrubber water will not exceed 1 per cent. The losses due to the difference between the initial and final tempera- tures of the fuel bed are variable and depend upon the capacity and to a certain extent upon the length of time required to blow the fires and produce a gas sufficiently rich for operating con- ditions. These losses have been discussed on page 37, and it was shown here that they would doubtless vary from .5 to about 9 per cent, depending upon the above conditions. The experimental errors made in the measurement of different quantities may be either positive or negative and the magnitude of the probable error varies from about 2 or 3 per cent in tests at capacities between 4000 and 2800 cubic feet of gas per hour to be- tween 4 and 7 per cent for tests between 2000 and 750 cubic feet per hour. GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 69 IV. CONCLUSIONS 1. A producer of the size and type tested is a practical piece of apparatus. With proper care in operating, and in the selection of the fuel, it is also a reliable piece of apparatus possessed of a fair degree of efficiency. 2. The efficiency of the producer is, within wide limits, prac- tically independent of the rate of gasification. 3. The efficiency of the producer, within comparatively wide limits, is only slightly affected by the relative percentages of H2 and CO appearing in the gas. With low temperature in the fuel bed, due to the use of a large quantity of steam, there is a high percentage of H2 and CO.. With high temperature in the fuel bed, the H2 and CO drop off while the CO increases. 4. From the above, it follows that the amount of steam to be used under any given conditions depends entirely upon the nature of the fuel. Clinkering coals require large quantities of steam in order to keep the temperature below that of the formation o01 clinker, while non-clinkering coals require much less steam. 5. The size of the fuel for a given temperature of fuel bed and a given depth of fuel bed has a marked effect upon the heating value and composition of the gas, and upon the efficiency of the producer. An increase in the size of fuel tends to cause the efficiency, heating value of the gas, and the percentage of CO to drop off. 6. The graphical logs show that the conditions (with the exception of the tests on Philadelphia and Reading coal) were sufficiently uniform to prevent trouble in the operation of engines driving machinery that requires no great degree of sen- sitiveness in regulation, such as the driving of pumping machin- ery, or shafting for machine shops, provided these engines have overload capacities of from 15 to 20 per cent. The conditions were not sufficiently uniform to permit the use of the gas in en- gines designed to drive electric lighting and other machinery requiring sensitive regulation. This is due, however, to the very small size of the unit. With larger units, the uniformity has proved entirely sufficient for the operation of engines driving electric generators. 70 ILLINOIS ENGINEERING EXPERIMENT STATION 7. From item 111 of Table 5, it will be found that the volume of gas generated per lb. of coal averages about 76 cu. ft. The average low heating value of the gas, if the tests on the Phila- delphia and Reading coal are not used, is approximately 110 B. t. u. per cu. ft. Since the average engine requires about 10,500 effective B. t. u. per b. h. p. hr., 95 cu. ft. of this gas will be required per b. h. p. hr. or #-|- = 1.25 + lb. coal. Since the coal contains about 12,800 B. t. u. per lb., the efficiency of 2545 x100 the plant will be 254 x 1.0 = 16 per cent. approximately. The 12800 x 1.25 thermal efficiency of a steam plant of this same size will be about 3 per cent, assuming a boiler efficiency of 60 per cent and an engine efficiency of 5 per cent. If the anthracite fuel costs $5.00 per ton, soft coal for the steam plant must cost $0.94 per ton in order that the fuel cost remain the same per h. p. hr. The cost of attendance, mainten- ance and repair will be practically the same for each plant. 8. Under the above conditions, with anthracite at $5.00 per ton, one b. h. p. could be produced at a fuel cost of .31 cent per hr., or 3.7 cents per 12 hrs. 9. The effect of the capacity, the amount of steam used, the effect of the size of the fuel, the amount of CO in the gas, and other items of a general nature, and the conclusions drawn from such items,are entirely applicable to producers of all sizes. It is believed that these items, with the forms and formulas, consti- tute thelmost valuable portions of this paper. APPENDIX GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 73 APPENDIX * FORM 1 RESULTS OF GAS PRODUCER TRIALS 1 Test number ... ............. .............................. 2 M ade by ........... ..... ... . .. .. ........................................ 3 At ....... ...... ... .. ............ ...... The University of Illinois...... . ... 4 K ind of producer ........... ...... ..... Otto ................................. 5 To determine ........................... Efficiency ... .. ............. 6 Principal conditions governing trial.... Uniform load..... ............. 7 Kind of fuel............................ Scranton-Anthracite, ............. 8 K ind of grate........... .... .. .... .. .. Plain ....... ...................... 9 Method of starting and stopping test.... Alternate..... ................. 10 Type of producer..... .... .. ... ... .. .. Suction ............................... 11 Form of blower-ejector................. Schutte & Koerting.................. 12 Date of trial ................... .................................. 13 D uration of trial ..................... ..... ...... .............. ................ DIMENSIONS AND PROPORTIONS 14 D im ensions of grate, ft .. . . ...... ........... . ............. . . ........ . 15 Grate area, sq. ft ............... ................................. 16 M ean diam eter of fuel bed, ft........................................ 17 Depth of fuel bed, ft ... ........... ......... .. ....... . ......... 18 Area of fuel bed, sq. ft................... . .............................. 19 Height of discharge pipe above grate. ft............... .................... 20 Approximate width of air spaces in grate, inches... ..................... 21 Area of air space, sq. ft..................... ........................... ... 22 Proportion of air space to whole grate area, per cent.................... 23 Area of discharge pipe, sq. ft . . ..... ...... .... ...................... 24 Water heating surface in vaporizer, sq. ft....... ..................... 25 Outside diameter of shell, ft.......... ....... ............................. 26 Length of shell from base to top of magazine, ft....... ................ 27 Ratio of water heating surface to grate area,-to 1. ............... ...... 28 Ratio of minimum draft area to grate area, 1 to .......... ................. AVERAGE PRESSURES 29 Draft in ashpit, inches, water... .... ................... ......... 30 Suction at producer outlet, inches, water... .. ..... .............. 31 Pressure at meters, inches, water... .............................. 32 Corrected barometer reading..... . ..... ............................. 32.1 Steam pressure. lb. per sq., in. gage .. .................................... AVERAGE TEMPERATURES 33 Of fire room , deg. Fahr................. ...................................... 34 Of steam leaving vaporizer, deg. Fahr.............. ...................... 35 Of feed water entering vaporizer. deg. Fahr. ..... ................... 36 Of overflow from vaporizer, deg. Fahr. ........................................ 37 Of water entering scrubber, deg. Fahr. ..................................... 38 Of water leaving scrubber, deg. Fahr ....... ........................ 39 Of gases leaving producer, deg. Fahr........ ........................ 40 Of gases leaving scrubber, deg. Fahr........ ....................... 41 Of gases entering meter, deg. Fahr.... . ......................... FUEL 42 Size and condition ...... .. .................. ........ .. ............ 43 W eight of coal as fired. lb ...... ....................... ....................... 44 Percentage of moisture in coal..... ......................... ...... 45 Total weight of dry coal fired, lb ................ ....... ...................... *In the following Appendix, the forms used in computing the results are given for the benefit of those who may have occasion to use them in the working of results of producer tests. The use of the forms has been explained on page 38. 74 ILLINOIS ENGINEERING EXPERIMENT STATION 46 Total ash and refuse, lb. .................. ..................................... 47 Quality of ash and refuse..... .. .... ......... ........................ 48 Total combustible consumed, lb.......... ............................ 49 Percentage of ash and refuse in dry coal.. .. ..................... PROXIMATE ANALYSIS OF COAL 50 F ixed carbon .... .. . ................................................ 51 V olatile m atter....................... .......................................... 52 M oisture ......................................................................... 53 A sh ..... ............ ............... ................... ............... ...... 54 Sulphur, separately determined.. ................................. ULTIMATE ANALYSIS OF DRY COAL 55 C arbon , C .................................................. .... 56 H ydrogen, H 2............ ............... ...... .. . ............................... 57 Oxygen, 0 2..................................................................... 58 Nitrogen. N2 ..................................................... 9 Sulphur, S .................................................... . ................ 60 A sh ......................................................... .................... 61 Moisture in sample of coal as received........ ....................... ANALYSIS OF DRY ASH AND REFUSE 62 Carbon, per cent ....... .... ....... ........................... 63 Earthy matter, per cent........ .......... ....................... ...... a SiO2 ..... ......................... .................. Al20... .............................................. Q Fe203........ . . .. .......................... ............... c M g O ............. .............. ...................... ............... d C aO .................................................. FUEL PER HOUR 64 Dry coal fired per hr. lb ......... .......... ........ ... ...... ............... 65 Combustible consumed per hr. lb .............. ................... 66 Dry coal per sq. ft. of grate area per hr. lb......... ..................... 67 Combustible per sq. ft. of grate area per hr. lb... .. ............... 68 Dry coal per sq. ft. of fuel bed per hr. lb....................................... 69 Combustible per sq. ft. of fuel bed per hr. lb................................. 70 Rate of descent of dry coal through fuel bed, lb per ft. per sq. ft. per hr ..... 71 Rate of descent of combustible through fuel bed. lb. per ft. per sq. ft. per hr. CALORIFIC VALUE OF FUEL 72 Calorific value by oxygen calorimeter per lb. dry coal, B. t. u................ 73 Calorific value by oxygen calorimeter per lb. of combustible B. t. u........... 74 Calorific value by analysis per lb. dry coal. B. t. u...................... 75 Calorific value by analysis per lb. of combustible, B. t. u ...... .............. WATER 76 Totall weight of water fed to vaporizer, lb... ....... ............... 77 Total weight of overflow from vaporizer, lb.................................. 78 Water1 actually evaporated in vaporizer, lb...... ..................... 79 Total weight of water fed to producer. lb............ .................... a From vaporizerl ........................ ................ ....... ..... b In air ....... ......... . . . .................................. c In coal..................................................... 80 Total weight of water decomposed. .......................................... 81 Total weight of water in gas leaving producer, lb..... ................. 82 Ratio of water decomposed to water supplied ...... ................... 83 Weight of water decomposed per lb. gas generated, b ..... ............. ISteam fed to producer, where vaporizer is not used. GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 75 84 Weight of water decomposed per lb. of dry coal fired. lb ................... 85 Weight of water decomposed per lb. of combustible consumed, lb............ 86 Weight of water decomposed per lb. of air supplied, lb....... ............ 87 Weight of water supplied per lb. of dry coal fired. lb.................. 88 Weight of water per lb. of combustible consumed, lb...... ............ 89 Weight of water supplied per lb. of dry air used, lb....................... 90 Total weight of scrubber water, lb ... ... ........................ WATER PER HOUR 91 Water evaporated per hr. in vaporizer, lb.... ................. ..... 92 Water evaporated per hr. per sq ft. of water heating surface in vaporizer, lb. 93 Weight of water decomposed per hr., lb.............................. 94 Total weight of water fed to producer per hr., lb...... ................. 95 Weight of scrubber water used per hr., lb .................................. QUANTITY OF AIR 96 Per cent of moisture in air, per cent of dry air...... ................. 97 Total weight of dry air, lb......... ....... ................................... 98 Total weight of dry air per hr. lb. ........ .................................. 99 Weight of dry air used per lb. of dry coal fired, lb.......................... 100 Weight of dry air used per lb. of combustible consumed, lb................... 101 Weight of dry air used per lb. of dry gas generated, lb........ .......... GAS 102 Per cent moisture in gas leaving producer, per cent of dry gas................ 103 Per cent of soot and tar in gas leaving producer....... ................ 104 Calorific value of standard gas from analysis (high value) B. t. u. per cu. ft.. 104a Calorific value of standard gas from analysis (low value) B. t. u......... 105 Calorific value of standard gas from calorimeter, (high value) B.t,u. per cu. ft. 106 Specific weight of standard gas, lb. per cu. ft...... ................... 107 Specific heat of dry gas leaving producer ................................... 108 Carbon ratio C/H ................................. .... ..... .. .... ........ 109 Total volume standard gas, cu. ft.. ................................ 110 Volume of standard gas per hr. cu. ft... ............................. 111 Volume of standard gas per lb. of dry coal....... .................... 112 Volume of standard gas per lb. of combustible....... ................. 113 Total weight of standard gas, lb....... ...... ............................. 114 Weight of standard gas per hr.. lb ..... ............................... 115 Weight of standard gas per lb. of dry coal fired, lb......................... 116 Weight of standard gas per lb. of combustible consumed, lb.................. GAS ANALYSIS BY VOLUME 117 Carbon dioxide, C02............................. ....... ......................... 118 Carbon monoxide, CO.... .. ...................................... .. 119 Oxygen, 02 ...................................................... 120 Hydrogen, H2.......... ........ ....... ......... .... .. ......... 121 M arsh gas, CH4 . .. . . ...... ............... ............ ................... 122 Oleflant gas, C2H4 .............................................................. 123 Sulphur dioxide, S02...... ......................................... 124 H ydrogen sulphide, H2S .............................. ........... ............ 125 Nitrogen, N2, by difference........ ..................... .... ..... EFFICIENCY 126 Grate efficiency, per cent..... ...................................... 127 Hot gas efficiency, based on high beating value, per cent.... ........... 128 Cold gas efficiency, based on high heating value, per cent.... ........... 128a Cold gas efficiency, based on low heating value, per cent..... ........... 76 ILLINOIS ENGINEERING EXPERIMENT STATION EFFICIENCY BASED ON 100 PER CENT GRATE EFFICIENCY 1286 Hot gas efficiency, based on high heating value, per cent.... .......... 128c Cold gas efficiency, based onhigh heating value, per cent.................... 128d Cold gas efficiency, based on low heating value, per cent................. COST OF GASIFICATION 129 Cost of fuel per ton delivered in producer room............................... 130 Cost per 1000 cu. ft. of standard gas, cents.. ... ........ ......... 131 Cu. ft. scrubber water per 1000 cu. ft. gas............... ............. POKING 132 M ethod of poking ...................... .......... ........... .................... 133 F requency of poking ............. ................................ ............ FIRING 134 M ethod of firing ............................................................. 135 Average intervals between firing .............. ................................ 136 Average amount of fuel charged each time. lb...... ..................... HEAT BALANCE DEBIT B. T. U. a Total heat supplied per lb. dry coal............. b Total heat supplied by air per lb. dry coal...... c Total heat supplied by moisture in air per lb. dry coal .......... . ......................... d Total heat supplied by moisture in coal per lb. dry coal....... ...................... e Total heat supplied as sensible heat in coal per lb. dry coal.......................... f Totall heat supplied by water in vaporizer per lb. dry coal........ ................. T otal .. ............................. CREDIT B. T. U. Per CONT a Heat contained as sensible heat in dry gas..... 6 Heat contained in moisture..................... c Heat contained in dry gas (heat of combustion) d Heat in unburned carbon...................... e Heat contained in ash and refuse as sensible heat f Heat lost in overflow from vaporizer........... g Heat lost in radiation and conduction.......... Total ... ................................ FORM 2 RESULTS OF' GAS PRODUCER TRIALS NO. OF TEST DATE, TIME OF START, TIME OF STOP DURATION OF TRIAL, HRS. KIND OF FUEL. DIMENSIONS AND PROPORTIONS 1 D im ensions of grate, ft ... .... ...................... ... ....................... 2 G rate area sq. ft........ .. .. . .. . ........ .......................... 1Supplied in steam, where vaporizer is not used. GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 77 3 M ean diam eter of fuel bed, ft ................. ................................... 4 Depth of fuel bed, ft........................ ..................................... 5 Area of fuel bed, sq. ft ........ ..... .. .......... ........................... 6 Height of discharge pipe above grate, ft...................................... 7 Approximate width of air spaces in grate, inches...... ................. 8 A rea of air spac . sq. ft ........................................ . ............... 9 Ratio of air space to whole grate area, % ........................... 10 Area of discharge pipe, sq. ft........ .... . ................................. 11 W ater heating surface in vaporizer. sq. ft .................................... 12 Outsidediameter of shell, ft....... . ............ ..................... 13 Length of shell from base to top of magazine, ft...... ................. 14 Ratio of water heating surface to grate area -to 1...... ................ 15 Ratio of minimum draft area to grate area to................................ AVERAGE PRESSURES 16 Average barometer reading, inches Hg .... ........................... 17 Average corrected barometer reading, inches Hg ..... ................... 18 Draft in ash pit, inches water.................................................. 19 Suction at producer outlet, inches water ..................................... 20 Absolute pressure at producer outlet, inches Hg.......................... 21 Suction1 at orifice, inches water........... ................................. 22 Absolute pressure1 at orifice, inches Hg....... ............... ........ 23 Pressure at meters, inches water... .... . .. .............................. 24 Absolute pressure at meters, inches Hg....... ............ ...... ...... 25 Vaporpressure at meters, inches Hg... ...... ........................ 26 Dry gas pressure at meters, inches Hg.... ..... ........................ 27 Suction at meter for dryer, inches water... ...... ...................... 28 Absolute pressure at meter for dryer, inches Hg .............................. AVERAGE TEMPERATURES 29 At barom eter, deg. Fahr ..... .................................................. 30 Of fire room , deg. Fahr .............................................. : ....... .. 31 Of fire room , deg. absolute Fahr ............ .................................... 32 Of steam, deg. Fahr.... ....... ............. . ............................... 33 Of feed water entering vaporizer, deg. Fahr .................................. 34 Overflow from vaporizer, deg. Fahr ............................................. 35 Rise in vaporizer, deg. Fahr ................................................... 36 Of water entering scrubber, deg. Fahr........ ........................ 37 Of water leaving scrubber, deg. Fahr ............................................ 38 Rise in scrubber, deg. Fahr ............ .......... ........................... 39 Of gases leaving producer, deg. Fahr.. .. ......................... 40 Of gases leaving producer, deg. abs. Fahr .................................. 41 Of gases leaving first scrubber, deg. Fahr...................................... 42 Of gases leaving first scrubber, deg. abs. Fahr...... ................... 43 Drop in temperature of gases in scrubber, deg. Fahr ..... ............. 44 Of gases entering meters, deg, Fahr ................................. .......... 45 Of gases entering meters. deg. abs. Fahr................................. 46 Of gas at meter at dryer, deg. Fahr ................................. ........... 47 Of gas at meter at dryer, deg. abs. Fahr.................. ................ FUEL 48 Size and condition... . ... . ....................................... 49 W eight of coal as fired, lb ......................... .............................. 50 Percentage of moisture in coal...... .... .......... .................... 51 Total weight of dry coal fired. lb ............................................... Steam pressure may be substituted here in case the water is not supplied from the vaporizer 78 ILLINOIS ENGINEERING EXPERIMENT STATION 52 Total ash and refuse, lb ......... .................................. 53 Quality of ash and refuse .......................................... 54 Total weight of com bustible, lb....................... ......................... 55 Percentage of ash and refuse in dry coal, per cent... ................... PROXIMATE ANALYSIS OF COAL 56 Fixed carbon, per cent........................................................ 57 Volatile m atter, per cent....................... .. ............................. 58 M oisture, per cent............ ........ .................................... 59 Ash. per cent . ......................... . .... ..... ........................ 60 Sulphur, separately determined, per cent.................................... ULTIMATE ANALYSIS OF DRY COAL 61 Carbon, C, per cent ............................... ...... .... ..................... 62 Hydrogen, H2, per cent ............................................. 63 Oxygen, 02. per cent .......... ....................... ....................... 64 Nitrogen, N2, per cent .................................................... 65 Sulphur, S. per cent ................ ............................................ 66 Ash, per cent................................ .................... 67 Moisture in sample coal as received, per cent............................ ANALYSIS OF DRY ASH AND REFUSE 68 Carbon, per cent ... .. .... .. ..................... ........................ 69 Earthy matter, per cent... . .... ........ . ........................... a Si02 b A20lO Fe203 c MgO d CaO FUEL PER HOUR 70 Dry coal fired per hr.. lb........ .. .............................................. 71 Combustible consumed per hr., lb... ................................ 72 Dry coal per sq. ft. of grate area per hr., lb....... ..................... 73 Combustible per sq. ft. of grate area per hr., lb ....... ................ 74 Dry coal per sq. ft. of fuel bed per hr., lb..... .............................. 75 Combustible per sq. ft. of fuel bed hr.. lb....................................... 76 Rate of descent of dry coal through fuel bed. lb. per ft, per sq., ft. per hr....... 77 Rate of descent of combustible through fuel bed. lb. per ft. per sq. ft. per hr.. CALORIFIC VALUE OF FUEL 78 Calorific value by oxygen calorimeter per lb. dry coal. B. t. u............. 79 Calorific value by oxygen calorimeter per lb. combustible B. t. u.............. 80 Calorific value by analysis, per lb. dry coal, B. t. u...... .............. 81 Calorific value by analysis, per lb. combustible, B. t. u................... ... WATER 82 Total1 weight fed to vaporizer, lb.......................... ............. ...... 83 Total weight of overflow, lb.......................................... ......... 84 Water' actually evaporated in vaporizer, lb.................................. 85 Weight of water fed to producer, a From vaporizer1...... ..... ..................... b In air............................................ c In coal .......... .............. ............................... Total ................................... .. .. ...................... 86 Total weight of water decomposed from analysis, lb...... ............... 87 Total weight of water decomposed as used in calculations, lb.................. 1 Steam fed to producer where vaporizer is not used. GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 79 88 Total weight of moisture in gas leaving producer. lb......... ................... 89 Ratio of water decomposed to water supplied... ........................... 90 Weight of water decomposed per lb. of gas generated, Ib....... ............... 91 Weight of water decomposed per lb. of dry coal fired, lb....................... 92 Weight of water decomposed per lb, of combustible consumed, lb............. 93 Weight of water decomposed per lb. of air supplied, lb...................... 94 Weight of water supplied per lb. of dry coal fired.lb ........ .................. 95 Weight of water supplied per lb. of combustible consumed, lb.................. 96 Weight of water supplied per lb. of air used, lb............................. 97 Total weight of scrubber water, lb ................................. 98 Total weight of water absorbed from sample by dryer, grams.... ......... WATER PER HOUR 99 Water evaporated per hr. in vaporizer, lb...... ............................. 100 Water evaporated per hr. per sq. ft. of water heating surface in vaporizer, lb.. 101 W eight of water decomposed per hr., lb.......................................... 102 Total weight of water fed to producer per hr.. lb.................. ............... 103 Weight of scrubber water used per hr., lb......... ...................... QUANTITY OF AIR 104 Relative humidity of air, per cent... ................................... 105 Per cent of moisture contained in air, per cent by weight of dry air................ 106 Total weight of dry air by analysis, lb....... ................. ............. 107 Total weight of dry air by orifice,lb......... ......................... 108 Total weight of dry air as used in calculations, lb........ ................ 109 Weight of dry air per hr. from total used in calculations ..... ............. 110 Weight of dry air used per lb. of dry coal fired, lb.............................. 1ll Weight of dry air used per lb. of combustible consumed, lb ...................... 112 Weight of dry air used per lb. of dry gas generated, lb....................... GAS 113 Volume of gas sample passing through meter at dryer, cu. ft.... ........... 114 Volume of standard gas passing through meter at dryer, cu. ft.... .......... 115 Total weight of gas passing through dryer meter, lb....... ............... 116 Percentage of moisture in gas leaving producer, from dryer, per cent dry gas... 117 Percentage of moisture in gas leaving producer, from water fed to producer, per cent dry gas ..... .............................................................. 118 Percentage soot and tar in gas leaving producer, per cent... .............. 119 Calorific value per cu. ft. standard gas from analysis B.t.u. (high value)......... 119a Calorific value per cu. ft. of standard gas. by analysis (low value)............. 120 Calorific value per cu. ft. of standard gas from calorimeter, B.t.u. (high value) 121 Specific weight of standard gas, lb. per cu. ft........... ...... .......... 122 Specific heat of dry gas leaving producer ....................................... 123 Carbon ratio C/H . . . ............. ....... . ....................................... 124 Total volume of gas from meters, cu. ft ..................................... 125 Total volume of standard gas, from meters, cu. ft...... .................. 126 Total volume of standard gas, from analysis, cu. ft.............................. 127 Total volume as used in calculations, cu. ft... ........................... 128 Volume of standard gas per hr. from total used in calculations.... ......... 129 Volume of standard gas per lb. of dry coal from total used in calculations, cu. ft. 130 Volume of standard gas per lb. of combustible from total used in calculations, cu . ft ................................................. ...... . ................... 131 Total weight of standard gas from total used in calculations, lb................ 132 Weight of standard gas per hr.. lb................ ......................... 133 Weight of standard gas per lb. of dry coal, lb ....... ....... ........... 134 Weight of standard gas per lb. of combustible, lb.............................. 80 ILLINOIS ENGINEERING EXPERIMENT STATION GAS ANALYSIS BY VOLUME 135 Carbon dioxide, CO 2 ................. .................................. .......... 136 Carbon m onoxide, CO ....................... .. ............... .............. .... . 137 O xygen, 0 2......................... ............................................. 138 H ydrogen, H 2 ............................. .................................. ....... 139 Marsh gas, CH4... .. .. .... ............ ............... ............. 140 Oleflant gas, C2H 4 .......................... ........ .......................... . ..... 141 Sulphur dioxide, S02.......................... .... ........ ........................ 142 H ydrogen sulphide,H2S................................ ............ ...... ....... 143 Nitrogen, N2 by difference...... .. ............................. ........ GAS ANALYSIS BY WEIGHT 144 Carbon dioxide, C 2. ........................................ .................... .. 145 Carbon m onoxide, CO....... ......................................... 146 Oxygen, 02 ............ ................... ......................... 147 H ydrogen, H 2 .................... ................. . .................... ....... 148 Marsh gas, CH4....... ..... ........ ... .. .................... ... 149 Oleflant gas, C211H4 ...... .. ... .................................................. 150 Sulphur dioxide, S02 .................................................... . ...... 151 Hydrogen sulphide, H2S.... . ......................................... 152 Nitrogen, N2 by difference.......... ........ ... ............................ . EFFICIENCY 153 Grate efficiency, per cent .... . .... ..................................... 154 Hot gas efficiency, based on high heating value, per cent..... ............. 155 Cold gas efficiency, based on high heating value, per cent...................... 155.1 Cold gas efficiency, based on low heating value, per cent................... EFFICIENCY BASED ON 100 PER CENT GRATE EFFICIENCY 155a Hot gas efficiency, based on high heating value, per cent.. ............... 155b Cold gas efficiency, based on high heating value,per cent.... .............. 155c Cold gas efficiency, based on low heating value, per cent.... ............. COST OF GASIFICATION 156 Cost of fuel per ton delivered in producer room........ .................. 157 Cost per 1000 cu. ft. of standard gas, cents.......... ................... 158 Cu. ft. scrubber water per 1000 cu. ft. standard gas....... ............... POKING 159 M ethod of poking.... ........... .......................... .. .................. 160 Frequency of poking.................. ........... . .. ....................... FIRING 161 M ethod of firing ........... .............. ......... .............. . ................ 162 Average intervals between firings........ .............................. 163 Average amount of fuel charged each time... .. ...................... HEAT BALANCE DEBIT 1 T.U. a Total heat supplied per lb. dry coal............................. b Total heat supplied by air per lb. dry coal............. ..... c Total heat supplied by moisture in air per lb. dry coal........... d Total heat supplied by moisture in coal ..................... e Total heat supplied as sensible heat in coal..... ............ f Total1 heat supplied in vaporizer water........................... Total...... . ..... ............................. I Supplied in steam, where vaporizer is not used. GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 81 PER CREDIT B.T.U. CENT at Heat contained as sensible heat in dry gas.................. > Heat contained in moisture... ............................. o Heat contained in dry gas (heat of combustion).................. d Heat in unburned carbon......................................... e Heat contained as sensible heat in ash and refuse................ f Heat lost in overflow from vaporizer...... ................ g Radiation and conduction, by difference..... ............... Total ..................... ............. ............. FORM 3 GUIDE SHEET CONTAINING ALL FORMULAS AND THEIR DERIVATION The item numbers refer to the items of Form 2, and are arranged in the order of computa- tion. Item 4. "Depth of fuel bed" is to a certain extent arbitrary. In order that the term may have a fixed and definite meaning we will define it as the distance between the upper edge of the ash zone and that section of the fuel b( d from which the gases separate and leave the fuel. The upper edge of the ash zone can ordinarily be readily determined by inspection. Item 16. This reading is the average of the barometer readings for the test and is not corrected. Item 17. Item 16corrected. The follo wing formula may be used: Let H = corrected barometer reading. t = temperature,deg. fahr. It = barometer reading corresponding to temperature t. Then 11= (1.00254 - 0.000079t) Item 17. = Item 16 (1.00254 - 0.000079 X Item 29) Item 18. = Observed. Item 19. = Observed. Item 20. = Item 17 - Item 19 X 0.0735 Item 21. = Observed. Item22. = Item 17 - ltem 21 X 0.0735 Item 23. = Observed. Item 24. = Item 17 + Item 23 X 0.0735 Item 25. = Taken from steam tables using temperature in Item 14, 1 lb. per so. in. = 2.04 in. Hg. Item 26. = Item 21 - Item 25 Item 27. = Observed. Item 28. = Item 17 - Item 27 X 0.0735 Items 29 to 47 inc. The observed temperatures should be corrected from the calibration curves before being placed in Form 2. The absolute temperature = the observed temperature + 460 deg. Item 39. This item is observed in deg. Cent. and should be transferred into deg. Fahr. Deg. Fahr. = deg. Cent. + 32 Each observation must be transferred. Item 50. Taken from Item 67. Item 5l. Item 49 ( - Item50 ) Itent 5". Taken from ash sheet, correction being made for any moisture taken up in the ash- pit. 82 ILLINOIS ENGINEERING EXPERIMENT STATION Item 51. In these tests the total weight of combustible consumed will be taken as the total weight of dry coal fired, minus the weight of ash computed from the analysis, minus the weight of nitrogen- 8 X the weight of oxygen, minus the weight of carbon contained in the ash and refuse and equals Item 51 Item 51 X Item 66 Item 51 X Item 64 S Item 51 X Item 63 Item 51 -- -- 10 --- -- --- 0 --- - - --- i -- 100 100 100 Item 52 X Item 68 100 Therefore, Item 54 = Item 51 1 - Item 66 + Item 64 + S Item 63- Item 52 X Item 68 L 100 I 100 Item 52 X 100 Item 55. Item 51 Items 56 to 69. From chemist. Items 69, a, b, c, d. The ultimate analysis of the ash will be made only in special cases to obtain data on the formation of clinker. Item 51 Item 70. -hours hours Item 54 Item 71. = hours--- Item 70 Item 72. --Item Item 2 Item 71 Item 73. = Item Item 2 Item 70 Item 74. =Item 5 Item 5 [ten Item 71 Item 5 Item 76. "The rate of descent of dry coal through the fuel bed," or "The dry coal per cu. ft. of fuel bed per hour," which is the same, offers a means of comparing the rate of gasification in different producers that seems to be better adapted for the purpose than the expressions taken from boiler practice, viz: ''coal per sq. ft. of grate area." or ''coal per sq. ft. of fuel bed," the latter having been used in producer practice. Item 74 Item 76. = Item 4 Item 7 Item 77. = Item 4 Item 78. Taken from chemist's report. Item 78 X Item 51 - Item 52 X Item 68 X 145.40 Itm 79. = Item 54 Item 80. = Item 61 X 145.40 + Item 65 X 40.00 + [Item 62 - 1 of Item 63] X 620.00 Item 80 X Item 51 - Item 52 X Item 88 X 145.40 Item 81. Ie 54 Item 54 Item 113. Total volume of gas passing through meter at dryer. Observed. Item 114. Total volume of standard gas passing through meter at dryer. Neglecting the effect of moisture, GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 83 Let pi = absolute pressure in inches Hg. at dryer meter. ti = absolute temperature, deg. fahr. at dryer meter, vi = total volume of gas passing through meter. P. V, and T, be the condition of standard gas. P = 30 in. Hg. T = 460 + 62 = 522 Then Jol)WP V ti T or = T plvil X 522 17.4 plvo or Pti 301t - t from which the value of Item 114 follows. S114. = 17.4 Item 28 X Item 113 Item 114. = 17.4 - tem 47 Item 47 Item 118. Not considered in these tests. Item 119 and 119a. One cubic foot of standard gas, i. e. gas at a temperature of 620 Fahr. or 522' absolute and a pressure of 30 in. Hg. gives up the high value in the following table when the products of combustion are brought back to this temperature, and the moisture condensed. If the moisture is not condensed it gives the low value. H2 = 328 (high) or 276 (low) B. t. u per cu. ft. of standard gas. C2H4 = 1610 (high) or 1510 (low) B.t.u. per cu. ft. of standard gas. CO - 319 (both high and low) B. t. u. per cu. ft. of standard gas. CH4 = 1010 (high) or 910 (low) B.t.u. per cu. ft of standard gas. Item 120. This quantity is the average of all the calorimeter determinations. Each sep- arate determination by the calorimeter must be computed and the heating value obtained. The following formula may be used. The calorimeter readings are taken in centigrade units with the exception of the meter reading and pressure. Let t2 = temperature of entering water, deg. cent. tl = temperature leaving water, deg, cent. r - rise in temperature of water, deg, cent. W = weight of water used during the intervals = 8 litres for all tests. Gi = cu. ft. of gas used from meter tg = temperature of entering gas, deg. cent. pg = pressure entering gas inches Hg. absolute, corrected for vapor pressure of water (see Item 25). H = heating value per cu. ft. of standard gas (62 deg. Fahr. or 16.7 deg. Cent. and 30 in. Hg.) ts = temperature of standard gas = 62 deg. Fahr, or 16.7 deg, Cent. ps = pressure of standard gas = 30. in Hg. Gs = cu. ft. of standard gas. Glpgg Gst)s TgXcs Where Tg and Ts are in absolute deg. cent., (1 - t2 = r Total heat per cu. ft. standard gas in B.t.u. = H Total heat absorbed by water = W X r 84 ILLINOIS ENGINEERING EXPERIMENT STATION WTV X rX 3.968 WTV X rX 3.968 SGs O- X pg X Ts 7g X ps 8 X r X 3.968 X Tg X 30 Gi X pv X (16.7 + 273) Tg X r X3.29 G1 X Pg where 3.968 is the conversion factor. In this formula it is assumed that the exhaust products are brought back to 620 F. This is not strictly true but the error introduced is negligible, when the error in the use of the apparatus is considered. There is another error due to the exhaust products carrying out more or less vapor of water than was brought in by the entering gas and air. This error will also be small and may either be positive or negative depending on con- ditions. The entering gas will in most cases come from direct contact with water and will therefore be saturated. The air ordinarily will not be saturated. On combustion, moisture will be formed by the union of the oxygen and hydrogen, there will be a con- traction in volume of the gases due to the combustion, and also a contraction or expan- sion due to a change in temperature after combustion. In whichever direction the change in the weight of moisture in the out-going gas from that brought in by the en- tering gas may occur, this. change may be considered very small; for the contraction on combustion will be comparatively small, and this contraction will partly offset the un- saturated condition of the air used for combustion. Also the change in temperature of the out-going gas from that of the entering gas will be small. The heating values as given in Items 119 and 120 are the hilh values. The values obtained from the analysis will be more accurate and will be used in all computations. Item 121. The specific weights of the following gases at 62 deg. and 30 in. Hg. are CO02 = 0.11610 CH4 = 0.04278 CO = 0 07362 C2H4 = 0.07370 02 = 0.08418 SO2 = 0.16380 H2 = 0.00530 H2S = 0.08682 N72 0.07400 Item 121. = [Item 135 X 0.1161 + Item 136 X 0.07362 + Item 137 X 0.08418 + Item 138 X 0.00530 + Item 139 X 0.04278 + Item 140 X 0.0737 + Item 141 X 0.1688 + Item 142 X 0.08682 + Ite n 143 X 0.0740] io Item 144 to 152. Calculation of the gas analysis by weight from the analysis by volume. As- sume that we have one cubic foot of gas at 62 deg.'Fahr. and 20 in. Hg. of the following composition: VOLUMETRIC ANALYSIS SPECIFIC WEIGHT ANALYStS BY WEIGHT PER CENT CO2 = a percent 0.1161 = Ha A =- aXa CO = b 0.07362 = T 3 = WXb 02 = c 0.08418 = Wc C = H2 =d 0.00530 = Wd D = -- GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 85 CH4 = e 0.04278 = We E =-- W C2H4 = f 0.07370 = Wf F = - SO2 = a 0.16380 = Tg G =WgX W H2S = h 0.08682 = lh H = -Wh N2 = i 0.07400 = Wi [ = - W Where W [aX Wa+b X Wb+ c X W+eX We...... +iX Wi], = Item 121. Item 122. The specific heats of the gases vary according to the pressure and temperature As the pressure used throughout the experiments is atmospheric we have only to consider the variation with the temperature. The following formulae taken from Zeuner, vol. I. page 147, give the specific heat for constant volume Cv. ICO2, ma y 6.50+ 0.00774t....... ....... ...... . ............... (1) H20, mac = 5.78+ 0.00572t... .................. ....... ........... (2) 02 H2N2, CO,m Cv 4.76 + 0.00214t ........... ............ . ............ .....(3) mCp - mUv = 1.9934.... ........ ........... .... . .......... ......(4) For the specific heat of marsb gas CH4, our other constituent, we will use the value Cp = 0.6. This is approximate, but as the quantity of CH4 is small the resultant error is consequently small. In the above formula, m is the molecular weight of the gas, t the temperature in deg. cent. and Cv the mean specific heat between zero and t deg. cent. Cp is determined from formula (4). From the above formulas, the analysis by weight as determined be- low and the temperature of the gases leaving the producer, the specific heat of each constituent in a unit weight of the gas may be determined. The specific heat of the gas will be the sum of the specific heats of the constituents. Substituting the value of mCv from formula (4). and the value of m, and changing to deg. fahr. we have from the above formulas: For CO02, Cp = 0.19 + .0000977t................. ........ . ............ ...a H 20. Cp = 0.426 + .000176t ............... ............. . ....................... ....... b H 2, Cp = 3.355 + .000678t .... ................................................. ...c CO , Cp = 0.24 + .000048 t ................ ...................... ................ d N2, Cp = 0.21 + .0000484t ................................................... e CH4. Op - 0.6 ..................... .... ... ...... ........ ............. ... ....f 02, Cp = 0.21 + .0000424t......................................................... a Let a. b. c. d, e and f, represent the mean Cp for the above gases between 32 deg. and t deg. Fahr. Then the Cp of the producer gas = the sum of the products of the constituents of the gas by weight X the specific heat of the constituent. That is, Item 122= [a X Item 144 +c X Item 147 + d X Item 145 + e X Item 152 + fX Item 148 + X Item 146] 100 Item 123 1 2 3 4 CO2 = 02 + C CO = C + O CH4 = C + 2H2 C2H4 = 2C + 2H2 44 = 32- +12 28 - 12 + 16 16 = 12 + 4 28 = 24 + 4 Total weight of carbon appearing in a unit weight of gas from the above = per cent by weight CO2 X - + per cent by weight CO X -- + per cent by weight CH4 X + per cent by 1100 700and Le Chateliers Formulas. 'Mallard and Le Chatelier's Formulas. 86 ILLINOIS ENGINEERING EXPERIMENT STATION weight C2H4 X- - 700 The total weight of H2 appearing in a unit weight of gas = per cent by weight +per cent by weight CH4 X -- + per cent by weight C2H4 X 40 700 or Item 123 = [Item 144 X 0.273 + Item 145 X 0.429 + Item 148 X 0.75 +Item 149 X 0.858] - [Item 147 + Item 148 X 0.25 + Item 149 X 0.143] Item 124. Observed. Item 125. Let G = total volume of gas as measured by the meters. p = absolute pressure of this gas in inches Hg. as observed. T = absolute temperature in deg. fahr. t = observed temperature. The volume of gas G as measured by the meter is saturated with water vapor at the temperature t. Let pi = pressure of this vapor in inches as obtained from the steam table. Then as the pressure p is the total pressure of the mixture, the actual or partial pres- ure of the dry gas is P-- p =pl. Let ps, Gs. and Ts, be the condition of standard gas. Then Gs X ps- G X or _ G X p X Ts _ GX X 522 _ X p2 X17.4 T. T T X rps T X 30 T Therefore Item 125 equals Item 124 X Item 26 X 17.4 Item 45 Item 126. Calculation of the volume of the gas from the analysis of the gas and the analysis of the coal. Evidently the total weight of the carbon appearing in the gas should be equal to the total weight of carbon in the coal minus the weight that is lost through the grate and the weight lost in soot and tar. This latter is small for the hard-coal producer and will be neglected. Let P = Per cent carbon by weight in dry coal. W= total weight of dry coal. WI = total weight of ash and refuse. Pi =Per cent by weight of carbon in the ash and refuse. W2 = total weight of carbon that should appear in the gas, or the weight-of carbon utilized in the producer. P2 = PF-Pi Wi 100 This carbon is contained in the CO2, CO, CH4. C2H4. The proportion by weight of C in CO2 is 3/11, of C in CO is 3/7, of C in CH4 is 3/4 and of C in C2H4 is 6/7. Therefore the total weight of C contained in a unit weight of gas will be 3/11 A + 3/4 E + 3/7 F + 6/7 G 100 Where A, E, F, and G are the per cent by weight of CO02, CH4, CO, and C2H4 from the gas analysis. The per cent of this carbon contained in the gas as CO02 is -- 3/11 A The actual weight of this carbon will be W2. Since W2 is the total weight of car- WI X 100 bon utilized, from the fuel. One pound of carbon on burning produces 3% lb. of CO2. 3/ll A WI2 X W3 / 0 X 3. = total weight of CO2 in the gas. Let Ws = the specific weight of CO02 at 62 deg. and 30 in. Hg. See Item 121. The standard volume Vs of CO02 will therefore be, A W2 100 X Ws X Wf Let this volume equal a per cent (from the volumetric gas analysis) of the total volume of gas delivered by the producer. The total volume of standard gas from the gas analysis is therefore GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER . 87 100 Vs a Vis A x 12 a X W3 X Ws Item 126 therefore equals Item 144 X (Item 51 X Item 61 - Item 52 X Item 68) 0.116 X Item 135 X (0.273 Item 144+0.75 Item 148 +0.429 Item 145+0.858 Item 149) Item 127. Item 126 should be used as a check on Item 125. The difference between the two values should not exceed 5 per cent. Item 125 should be used in all computations, Item 127 Item 128 = hou hours Item 127 Item 129 = Item 127 Item 51 Item 127 Item 130 Item 54 Item 54 Item 131 = Item 127 X Item 121 Item 131 Item 132 hours hours Item 133 Item 131 Item 51 Item 134 = Item Item 54 Item 135 to 143 From chemist. Items 104. The relative humidity, or per cent saturation is observed by means of a hair h y- grometer. This may also be obtained from a wet and dry bulb thermometer, and a set of psy- chrometric tables. Item 105. See Kent. page 484 for weights of air and moisture. Letp = per cent saturation, or relative humidity, Item 104. n = weight of moisture contained in one cu. ft. of saturated air at the observed tem- perature, Item 29. - = weight of moisture in 1 cu. ft. of air as used. 100 If m = weight of 1 cu. ft. dry air at the observed temperature, then Item 105 -= - -X 100 = 0 n = Item 104 X n loom m m This formula is in error due to neglecting the vapor pressure of water; this is, however, neg- ligible in the present case. Item 82. Observed. Item 83. Observed. Item 84. = Item 82 - Item 83. Item 86. The weight of water decomposed in the producer is evidently 9 times the weight of hydrogen formed, since 1 lb. of water on decomposition yields 1 lb. of hydrogen and 8 lb. oxygen. The total weight of hydrogen formed is equal to the total weight of free hydrogen appearing in the gas, plus the total weight of hydrogen appearing in the CH4 in the gas. minus the total weight of hydrogen that is not in combination with oxygen in the coal. Item 86, therefore, equals 9 [Item 131 (Item 147 +0.25 Item 148) - Item 51 (Item 62 - % Item 63)1 100 88 ILLINOIS ENGINEERING EXPERIMENT STATION Item 87. Owing to the difficulty in obtaining the weight of moisture in the gases leaving the producer with aproper degree of accuracy by the use of a dryer, it will ordinarily he better to use Item 86 for this item. ftem 106. Obtained from gas analysis by weight, Items 144 to 152 inclusive. Let A = per cent C02 Let D = per cent H2 B = per cent 02 E = per cent CH4 C = N2 per cent F = per cent CO 1 2 3 C +02= C2 C+= CO= CO H2+O= H20 12+32= 44 12+16= 28 2 +16= 18 + =i + =i -1-+-.8_, 11 +11 7 + 7 9 9 From equation (1), one lb. of 002 requires iA lb. of Ofor its for.nation From (2) one lb. CO requires 4 lb, of 0 for its formation. The total amount of 0 appearing in 1 lb. of the gas is therefore 8 1A + F+B IO0 This 0 comes from that contained in the air, that contained in the coal, and from the water decomposed. The oxygen contained in the coal, however, is supposed to be united with hydrogen, and is therefore contained in moisture, which has been allowed for in the water de- composed. Let TV = total weight of gas. Then the total weight of 0 used is 'V0(UA+ 4 F+B) 100 , 11 7 / Let W2 = weight of wa er decomposed. From (3). 1 lb. of water decomposed lib- erates -- lb. of 0. 9 8 Weight of 0 supplied by decomposition of water = W'2 Let W3 = total weight of 0 supplied by air. From the above equation we have, (-- A + F + B) - 2 + W3. (11 7 100 9 or Ws = A+ F+ ) -8 WI ....... (4) W113 The weight of air used is therefore )3, since the proportion by weight of 0 in air is 23. or 03 W4= 3 ( - A+ F +B F+B - W2...........(5) Therefore Item 106 = Item 131 8 Item 144 + Item 145 + Item 146 100 11 14+ - X Item 87 023 ......................... (6) The above computation may be made from the weight of nitrogen appearing in the gas. The nitrogen comes from the air used and from the nitrogen introduced with the fuel. Let Cper cent = weight of N2 from analysis. Let W as before = total weight of gas, GARLAND-KRATZ--TESTS OF A SUCTION GAS PRODUCER 89 CW Then - 10 = total weight of N2 in the gas. W1 HI The weight of N2 supplied by fuel will be too . where W1 equals the total weight of dry coal and Hi is the per cent by weight of N2 contained in the coal. We have therefore. CW WiHI 100 100 where M'W4 = total weight of N2 in the air. The weight of air supplied is therefore W, c "W W \Hl 1 1 /1; CW - WIHl 0.77 100 100 0.77 77 or Item 106 = (Item 131 X Item 152 - Item 51 X Item 64)--............. .....7) The weight of air derived by formula (6) will be liable to error, due principally to the error in the determination of the total quantity of water decomposed, which may be large, and also to the neglecting of the S02 formed. The weight determined by formula (7) will be in error due principally to the taking of the weight of N2 from the analysis by difference. The results obtained from formulae (6) and (7) should check within 5 per cent. The results obtained by (7) are believed to be more accurate and will be used in all com- putations. Item 107. This may be obtained direct from the calibration curve of the orifice. It should be compared with the two values obtained above. Item 108. This will ordinarily be taken from Item 106. Item 108 Item 109 Item 108 Hours Item, 110 Item 108 Item 51 Item 111 Item 108 Item 54 Item 112 Item 108 Item 110 --Itm- Item 131 Item 85 = Item 84 + Item 85b + Item 85c. Item 85b = Item 108 Item 105 Item 85c = Item 49 X Item 50 100 Item 88 = Item 85 - Item 87 Item 89 = Item 87 Item 85 Item 90 = Item 87 Item 87 Item 91 = Item 8 Item 51 Item 87 Item 92 em 54 Item 93 = Item 87 Item 108 Item 85 Item 95 = Item85 Item 54 Item 85 Itm 96 = Item 108 Item 97 = Observed Item 98 = Observed 90 ILLINOIS ENGINEERING EXPERIMENT STATION Item 99 =Item 84 Hours Item 100 = Item 99 Item 11 Item 101 = Item 87 Hours Item 102 = Item 85 Hours Item 103 = Item97 Hours Item 115 = Item 114 X Item 121 Item 116 = Item 98 X 0.2205 Item 115 Item 117 = 100 Item 88 Item 131 Item 153 The grate efficiency is 100 times the ratio of the total B.t.u. in the fuel minus the B.t.u. in the fuel lost through the grate; to the total B.t.u. contained in the fuel. Therefore Item 153 = Item 51 X Item 78 X 100 - Item 52 X Item 68 X 14540 Item 51 X Item 78 Item 154. The hot gas efficiency is 100 times the ratio of the total heat of combustion of the gas, plus the sensible heat of the dry gas, plus the total heat contained in moisture in the gas to the heat of combustion of the dry coal, plus the beat given by the entering air. by the coal as sensible heat, and by the moisture or steam supplied in air. Item 154 = 100 Item 119 X Item 127 + Item 122 X Item 131 (Item 39 - 62°.) + Item 88 [1116 + 0.6 (Item 39 - 212)] Item 51 X Item78 + Heat supplied in steam. The heat given the producer by air, and sensible heat in coal may be neglected if the room temperature is within 20° of the standard temperature 620. With a producer of the contained vaporizer type or one which utilizes the sensible heat of the gas to make the steam the term "Heat supplied in steam" drops out of the equation. Item 155. The cold gas efficiency is 100 times the ratio between the total heat of combustion of the gas, to the total heat of combustion of the coal plus heat supplied from out- side sources. That is, Item 155 = 100 Item 119 X Item 127 Item 51 X Item 78 The efficiency based on 100 4 grate efficiency = Item 155 Item 158 Iem 157 Item 156 X Ttem 49 0.02 X Item 127 Item 158 Item 97 X 1000 Item 97 62.5 X Item 127 0.0625 X Item 197 HEAT BALANCE DEBIT Item a. Obtained from Item 78. Items b, c, d, e, f. Using as a standard the temperature of 62* F., the heat given to the producer by the items b to f inclusive is in most cases negligible. The error at a tem- perature of 100* F. is less than 1 per cent for a producer of the contained vaporizer type. However, the formulas will be given for computation of these items. Item b = Item 110 X 0.24 ( Item 30- 62° F.) Item c =tem 85bItem 5(H-1070where H = the total heat in 1 lb. saturated Item 51 GARLAND-KRATZ-TESTS OF A SUCTION GAS PRODUCER 91 steam at the temperature of the fire room. Item d = Item 49 X Item 50 (Item 30 - 620 F.) 100 X Item 51 Item e = 0.24 X (Item 30 - 62* F.) Item 82 ( Item 33 - 620 F.) Item Item 51 CREDIT. Item a = Item 122 X Item 133 X (Item 39 - 62° F.) Item b = te 117 X Item 133 [(Item39 - 2120 F.) X 0.6 + 1116] 100 Itemc = Item 119 X Item 129 Item d = Item 52 X Item 68 X 145.40 Item 51 Item e This is very small and may be neglected. Item f Item 83 (Item 34 - 682 F.) Item a = Sum of Items on debit side - (Item a + Item"b>+ Item c + Item d + Items e and!.) PUBLICATIONS OF THE ENGINEERING EXPERIMENT STATION *Bulletin ýo. 1. Tests of Reinforced Concrete Beams, by Arthur N. Talbot. 1904. *Circular No. 1. High-Speed Tool Steels, by L. P. Breckenridge. 1905. * Bulletin No. 2. Tests of High-Speed Tool Steels on Cast Iron, by L. P. Brecken idge and Henry B. Dirks. 1905. *Oircular No. 2. Drainage of Earth Roads, by Ira 0. Baker. 1906. *Circular No. 8. Fuel Tests with Illinois Coal. (Compiled from tests made by the Tech aologic Branch of the U. S. G. S., at the St. Louis, Mo . Fuel Testing Plant, 1904-1907,) by 1,. P. Breckenridge and Paul Diserens. 1909. * Bulletin No. 3. The Engineering Experiment Station of the University of Illinois, by L P. Breckenridge. 1906. *Bulletin No. 4. Tests of Reinforced Concrete Beams. Series of 1905, by Arthur N. Talbot. 1906. *Bulletin No. 5. Resistance of Tubes to Collapse, by Albert P. Carman. 1906. *Bulletin No. 6. Holding Power of Railroad Spikes, by Roy I. Webber. 1906. *Bulletin No. 7. Fuel Tests with Illinois Coals, by L. ,.P. Breckenridge, S. W. Parr and Hanry B. Dirks. 1906. *Bulletin No. 8. Tests of Concrete: I Shear: 11. Bond, by Arthur N. Talbot. 1906. *Bulletin No. 9. An Extension of the Dewey Decimal System of Classification Ap- plied to the Engineering Industries, by L. P. Breckenridge and G. A. Goodenough. 1906. *Bulletin No. 10. Tests of Concrete and Reinforced Concrete Columns, Series of 1906, by Arthur N. Talbot. 1907. *Bulletin No. 11. The Effect of Scale on the Transmission of Heat through Locomotive Boiler Tubes, by Edward C. Schmidt and John M. Snodgrass. 1907. *Bulletin No. 12. Tests of Reinforced Concrete T-beams, Series of 1906, by Arthur N. Talbot. 1907. *Bulletin No. 13. An Extension of the Dewey Decimal System of Classification Applied to Architecture and Building, by N. Clifford Ricker. 1907. *Bulletin No. 14. Tests of Reinforced Concrete Beams, Series of 1906, by Arthur N. Talbot. 1907. * Bulletin No. 15. How to Burn Illinois Coal without Smoke, by L. P. Breckenridge. 1908. "Bulletin No. 16. A Study of Roof Trusses, by N. Clifford Ricker. 1908. *Bulletin No. 17. The Weathering of Coal, by S. W. Parr, N. D. Hamilton, and W. F. Wheeler. 1908. *Bulletin No. 18. The Strength of Chain Links, by G. A. Goodenough and L. E. Moore. 1908. *Bulletin No. 19. Comparative Tests of Carbon, Metallized Carbon and Tantalum Fila- ment Lamps, by T. H. Amrine. 1908. *Bulletin No. 20. Tests of Concrete and Reinforced Concrete Columns, Series of 1907, by Arthur N. Talbot. 1908. * Bulletin No. 21. Tests of a Liquid Air Plant, by C. S. Hudson and 0. M. Garland. 1908, *Bulletin No. 22. Testsof Cast-Iron and Reinforced Concrete Culvert Pipe, by Arthur N. Talbot. 1908. * Bulletin No. 28. Voids, Settlement and Weight of Crushed Stone, by Ira O. Baker. 1908. Bulletin No. 24. The Modification of Illinois Coal by Low Temperature Distillation, by S. W. Parr and C. K. Francis. 1908. Bulletin No. 25. Lighting Country Homes by Private Electric Plants, by T. H. Amrine. 1908. Bulletin No. 26. High Steam-Pressures in Locomotive Service. A Review of a Report to the Carnegie Institution of Washington, by W. F. M. Goss. 1908. Bulletin No. 27. Tests of Brick Columns and Terra Cotta Block Columns, by Arthur N. Talbot and Duff A. Abrams. 1909. Bulletin No. 28. A Test of Three Large Reinforced Concrete Beams, by Arthur N. Talbot. 1909. Bulletin No. 29. Tests of Reinforced Concrete Beams: Resistance to Web Stresses, Series of 1907 and 1908, by Arthur N. Talbot. 1909. Bulletin No. 30. On the Rate of Formation of Carbon Monoxide in Gas Producers, by J. K. Clement. L. H. Adams, and C. N. Haskins. 1909. * Out of print. PUBLICATIONS OF THEB ENGINEERING EXPERIMENT STATION- (Continued) Bulletin No. 31. Fuel Tests with House-heating Boilers, by J. M. Snodgrass. 1909. Bulletin No. 82. The Occluded Gases in Coal. by S. W. Parr and Perry Barker. 1909. Bulletin No. 33. Tests of Tungsten Lamps, by T. H. Amrine and A. Guell. 1909. Bulletin No. 34. Tests of Two Types of Tile Roof Furnaces under a Water-tube Boiler by J. M. Snodgrass. 1909. *Bulletin No. 35. A Study of Base and Bearing Plates for Columns and Beams, by N. Clifford Ricker. 1909. Bulletin No. 36. The Thermal Conductivity of Fire-Clay at High Temperatures, by J . K. Clement and W. L. Egy. 1909. *Bulletin No. 37. Unit Coal and the Composition of Coal Ash. by S. W. Parr and W F. Wheeler. 1909. Bulletin No. 38. The Weathering of Coal, by S. W. Parr and W. F. Wheeler. 1H09 Bulletin No. 39. Tests of Washed Grades of Illinois Coal, by C. S. McGovney. 1109. Bulletin No. 40. A Study in Heat Transmission, by J K. Clement and C. M. Garlar d. 1910. Bulletin No. 41. Tests of Timber Beams, by Arthur N. Talbot. 1910. Bulletin No. 42. The Effect of Keyways on the Strength of Shafts, by Herbert F. Moore. 1910. Bulletin No. 43. Freight Train Resistance, by Edward C. Schmidt. 1910. Bulletin No. 44. An Investigation of Built-up Columns under Load, by Arthur 1'. Talbot and Herbert F. Moore. 1911. Bulletin No. 45. The Strength of Oxyacetylene Welds in Steel, by Herbert I. Whittemore. 1911. Bulletin No. 46. The Spontaneous Combustion of Coal, by S. W. Parr and F. W. Kres-- mann. 1911. Bulletin No. 47. Magnetic Properties of Heusler Alloys, by Edward B. Stephenson. 1911. Bulletin No. 48. Resistance to Flow through Locomotive Water Columns, by Arthur N. Talbot and Melvin L. Enger. 1911. Bulletin No. 49. Tests of Nickel-Steel Riveted Joints, by Arthur N. Talbot and Herb ert F. Moore. 1911. Bulletin No. 50. Tests of a Suction Gas Producer, by C. M. Garland and A. P. Kratz. 1911. UNIVERSIET OF SL TT TNQIS T!E STATE UNIVEPSITY THE UNMERSITY uLU1PES THE COLLEGE OF LITERATURE AND ARTS (Ancienit anld Modern SLanguagesan Literatures, Philosophical and Political -Sei ence Groups of Stuadi~es,gEconomics, Commle~rce and Industry.).. COLLEOG OF ENGINEBRINQ Graduate and undergraduate courses in Arhitecbure Architectural Engineeeringr Civil 1Engineering; Electrical Engineering- Mechlnieal Enginteer- ing; Mining Engineering; Municipal and Sanftary Eng- ineering; Railway Engineering. COLLEGE OF SCIENCE (Astronomy, Botany Chemistry, Ge-. ology,, Mathematics, Physte ePhysiology, Zoology). COLLEOG OF AGRICULTURE (Animal Husbandry, Agronomy, Dairy Hflusbandry, Horticulture, Veterinary Sciencej House- hold Science. COLLEGl OF LAW (Three years' course).- COLLEGE OF MEDICINE (College of Physicians and Surgeons, Chicago). (Four years' course). COLLEGE OF DENTISTRY (Chicago), (Three years' tourse). SCHOOLS-T-GRADUATE SCHOOL, MUSIC (Voice Pilano, Vio- lin), LIBRARY SCIENCE_ PHARMACY (Chicago), EDU- CATION, RAILWAY ENGINE(EBRING AND ADMINISTRA TION. A Summer School with a session of eight weeks Is open durig the summer. A M*iliary Reghent Is organized.at the Univer.ity for Ifistruco Mion MIlitary Science. Closely connected with the work . of UnIverlty re students' orantizations fodducational and social purposes. (Gles and Mandolin Clubs, Literary, SoeItC, an Techeal Societies and Clubs, Young Men's and Young Womee's Christian A Iotions)., United Ste E xpeinet Staion at0 otory of Iu'ra I~tsory Bidlogicl Exeriment Statn. on Ilita Ridvr, I inetigate prdblem of impo ce tothe engineerin n sitnulfctg rnitgtere*ts Mf the otte. at e L b a r y c o I v en 8 5 , 0 0 G s o l u m e u r e y For 13g 'a Infor t tiop onK ~t CI M Us'faan Ilea Frc4taloa4b noriito ddea >7'77 A -"~ 77~7>~ 747777 7 47 7-777 7777777 77 7747. 777 7 7- 7 7 >7 '.7>7 77 777 £ ~74'' '777777 4 777; 7 '7 477 <' 77.j-> 7777 77 '77 .7 7 7777 47 77 7477 7>7 777 7 47 7 97 4777 '77 77 777 74 7 777 77 77 7 77 7. 7) 747 7> 7 77 7 7777 7777 7797 7 7>7 7 747 37 ~> 77;~ (7 (~777' 777 777. ' 77 77 7- 7777 7.7 1 77~777777 7~77 97 7< 7 77' 7 77 2' 777 >7 7 7 7 7 7 7. 7777 77777 7'. 7 ->7 <~~7 77> 77 977 7777 7 7 77777 7- 7. 7 77 777 4 7 7 7' 7 77 ~'7'.7. >7 777 .7 97 ' 7 77' 7'> 777777 ,777- >7>7 7774777 7 777777 >7 7-77 7. -7 7 7 47 7 77'. 7 7 7 7777 77 7 7 7 7 777 777777 777 77 47 7 777 4 7.7.7 * ':^'*- ii! '* :.' ,i :  < ' :;' i-i¸ - -' "  : I,  -,jK: -i ! 77 7> 7 773 77.7 77 477 7777 -7 >' 77 7 7 7> 7 74 777 7 7 77 7 77777 >7 77 7 777 77 77>7 7 7 -77 777 It 777 7-7 N 7>77 77777 - 77 77~~774 - ,, IF 77 7'77 7> 77~777 717 -y Ž-'. '.I 71 7>7>7~ 7 77 7777 -~ ''>77'. 7 77 77>77 4~ .47 '.7 7~7 77 '77 797 ~>7 77 74 >77 ~7 / ,,~7-77 7 77.7 ~ 7? >~'~ 'A r-~ 477 777 '7 4797 ~N >7774777 ~~>' 777777 >7/7 47 7.7 >77>7<7.777 '774" 4 7. 777 '7 N,-7< 4>7 >77777 W'> 7$ 77777 '7-7r >7<» 77 77 7'. 7~77-S7 '74' ~>-7 ¾. >777> 4 ~ 7 7777 * 7777777 ~ '~ ~ 7)7 77737. 47 '>~t" ~7-'.> 7> >7 ~>)77 77 .~ >1 >~>777 777 '4777'. >7 77717 77 ~ 7,7 7~~)77 7 77 77 7 77 7 >7 4977 >7 .7.77. 77797 7 7 7 777 7 77777 ~77> 7.477" '.->7 'ti> ~v~>> 7777> 47. >77t t '.7>7477 ' ** '*' I'- '^''Sv i^,.- '"..'-'".'y? ; ^-^ 'T VN ^; ^;.-^--